234 76 25MB
English Pages 464 Year 2008
Theorizing Narrativity
Narratologia Contributions to Narrative Theory/ Beiträge zur Erzähltheorie
Edited by/Herausgegeben von Fotis Jannidis, John Pier, Wolf Schmid Editorial Board/Wissenschaftlicher Beirat Catherine Emmott, Monika Fludernik Jose Angel Garcia Landa, Peter Hühn, Manfred Jahn Andreas Kablitz, Uri Margolin, Matias Martinez Jan Christoph Meister, Ansgar Nünning Marie-Laure Ryan, Jean-Marie Schaeffer Michael Scheffel, Sabine Schlickers, Jörg Schönert r
12
w DE
G Walter de Gruyter · Berlin · New York
Theorizing Narrativity Edited by John Pier and Jose Angel Garcia Landa
w G_ DE
Walter de Gruyter · Berlin · New York
© Printed on acid-free paper which falls within the guidelines of the ANSI to ensure permanence and durability.
Library of Congress Cataloging-in-Publication
Data
Theorizing narrativity / edited by John Pier, Jose Angel Garcia Landa. p. cm. - (Narratologia ; 12) Includes index. ISBN 978-3-11-020244-1 (alk. paper) 1. Discourse analysis, Narrative. 2. Narration (Rhetoric) 3. Communication. I. Pier, John. II. Garcia Landa, Jose Angel. P302.7.T49 2008 40Γ.41—dc22 2007049146
ISBN 978-3-11-020244-1 ISSN 1612-8427 Bibliographic information published by the Deutsche
Nationalbibliothek
The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de.
© Copyright 2008 by Walter de Gruyter G m b H & Co. KG, D-10785 Berlin All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Printed in Germany Cover design: Christopher Schneider, Berlin
Contents JOHN PIER AND JOSE ANGEL GARCIA LANDA
Introduction
7
GERALD PRINCE
Narrrativehood, Narrativeness, Narrativity, Narratibility
19
MEIR STERNBERG
If-Plots: Narrativity and the Law-Code
29
JOHN PIER
After this, therefore because of this
109
PETER HÜHN
Functions and Forms of Eventfulness in Narrative Fiction
141
WERNER WOLF
Chance in Fiction as a Privileged Index of Implied Worldviews: A Contribution to the World-Modelling Functions of Narrative Fiction
165
BEATRIZ PENAS IBANEZ
A Pragma-stylistic Contribution to the Study of Narrativity: Standard versus Non-standard Narrativities
211
DAVID RUDRUM
Narrativity and Performativity: From Cervantes to Star Trek
253
JUKKA TYRKKÖ
'Kaleidoscope' Narratives and the Act of Reading
277
MICHAEL TOOLAN
The Language of Guidance
307
6 ANSGAR NÜNNING AND ROY SOMMER
Diegetic and Mimetic Narrativity: Some further Steps towards a Transgeneric Narratology of Drama
331
MONIKA FLUDERNIK
Narrative and Drama
355
MARIE-LAURE RYAN
Transfictionality across Media
385
JOSE ANGEL GARCIA LANDA
Narrating Narrating: Twisting the Twice-Told Tale
419
Author/Name Index
453
JOHN PIER AND JOSE ÄNGEL GARCIA LANDA
Introduction Why is a narrative a narrative? What makes a narrative more or less narrative? What elements in a semiotic representation can be qualified as properly narrative? Which formal and/or communicative resources can be exploited or elaborated in specifically narrative ways? What features of a discourse otherwise considered a description, an argument or a dialogue can legitimately be regarded as narrative? How is narrative affected by its realization in the different media? These questions stake out, at least in part, the problem of narrative specificity, or narrativity. The contributions to this volume address these and related issues from various angles, seeking to shed light on what can be regarded as an increasingly crucial topic in narratology.1 Indeed, the multitude of paradigms and approaches spawned by "postclassical" narratological theories together with the broadening scope of cultural phenomena seen as coming within the purview of "narrative" is not only a fact of the history of the discipline, but it also underscores the necessity of reflecting anew on criteria that are appropriate and necessary for theorizing narrative in its manifold realizations. In its early stages, narratology devoted considerable attention to identifying and formalizing the underlying structures shared by all narratives, and it produced a number of influential narrative grammars and other such models that sought to capture the minimal and "immanent" features of narrative. Since then, and more perhaps than any other factor, the various attempts to account for the contexts of narrative, from the sides of both production and reception, have transformed the horizons of narratological research and indeed the concept of narrative itself, bringing about a number of reorientations in perspective and closer attention to criteria previously not taken into consideration. Given these 1
The present publication was initiated at the time of the Narratology Seminar convened by us at the 7th Congress of the European Society for the Study of English held at the University of Zaragoza in September 2004. Papers were read by Jos£ Angel Garcia Landa, Peter Hiihn, Ansgar Nünning, Beatriz Penas Ibäflez, John Pier, Michael Toolan and Jukka Tyrkkö.
8
John Pier and Jose Angel Garcia Landa
evolutions, questions such as those asked above have become more prominent, and it is around narrativity, not unlike the "literariness" of literary theory and poetics in earlier times, that some of the most consequent issues in contemporary narratological research are concentrated. The attempt to define narrative as such, independent of the specific genres or of literary fiction that has dominated the study of narrative throughout its long history, was inaugurated by structuralist narratology and continues to be a subject of debate today. 2 One of the significant consequences of this development, reflected in much discussion about narrativity, is that stories have come to be seen in a double perspective: "being narrative" and "possessing narrativity." 3 In the present collection, this question is taken up most explicitly by Gerald Prince, who speaks of narrativity in terms of narrativehood, an extensional category which designates a class of entities meeting certain conditions (e.g. the logically consistent representation of two synchronous events that presuppose or imply one another), and of narrativeness, an intensional category designating one of the qualities of an entity. Characterized both by kinds and by degrees, narrativity is also influenced by contextual factors and, to account for this, Prince adopts the notion of narratibility (tellability in the terminology of other narratologists): What is the "point" of this story? It would be an exaggeration to state that Prince's delineation of the components of narrativity provides a blueprint for the other essays, for each of the contributions develops a cogent series of arguments that stands on its own, articulated in function of a particular problematics. Yet readers may at times find this system a useful point of reference, for it sets out in a clear and well-balanced fashion the dimensions of narrative generally acknowledged by present-day research—its formal requirements, its modalizations, its communicational potential—through the perspective of narrativity. And indeed, this characterization largely corroborates the definition of narrativity provided by A Companion to Narrative Theory·, "the formal and contextual qualities distinguishing narrative from non-narrative, or marking the degree of 'narrativeness' in a discourse; the rhetorical principles underpinning the production or interpreta-
2
3
Cf. Marie-Laure Ryan, "Narrative," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan (London and N e w York: Routledge, 2005), p. 344. Ibid.: 347.
Introduction
9
tion of narrative; the specific kinds of artifice inherent in the process of narrative representation."4 With these general criteria in mind, we now wish to briefly review the essays of this volume, with no pretension, however, to an adequate account of the tightly-wrought arguments and no strictures as to alternative readings each contribution is sure to accommodate. The ordering of the contributions we have adopted reflects what appears to us to characterize the principal areas of narrative theory that have felt the effects of the rise of interest in narrativity and, no doubt, bear witness to the far-ranging consequences of the notion for narratological research. Broadly speaking, some contributions focus on or take into consideration narrativity and its neighboring concepts as a means to single out what differentiates narrative from other forms of cultural representation while others, in contrast with but not necessarily in opposition to or in neglect of this "centralizing" perspective, investigate issues that seemingly lie on the "periphery" of these considerations but which in fact come under a new light when viewed from the angle of narrativity, revealing a true potential for transforming the organization and aims of narrative theory. Following Prince's opening text, the articles are loosely divided into five groups in such a way as to bring out similar and/or complementary lines of investigation, even though some of the issues debated within each group are relevant in various ways to concerns addressed in other groups. Narrativity in a dynamic, anti-immanentist frame of reference Meir Sternberg inquires into the pertinence of the law-code to narrative. The corpus juris, he notes, is frequently perceived as timeless and immobile, bearing on the general as opposed to the particularities of the cases for which the law is used to adjudicate, so that its codified and static nature is often seen as being at odds with the dynamic and dialogical character of narration. Sternberg maintains, however, that the law is factcontingent and thus necessarily rooted in narrative. Among the examples discussed (all taken from the Old Testament) is the following: "If thou buyest a Hebrew slave, six years he shall serve." The " I f . . . then ..." pattern of this mini "law-tale" is both contingent and future-oriented: if suchand-such an action is accomplished, then a certain course of action must 4
Edited by James Phelan and Peter Rabinowitz (Maldon, MA, etc.: Blackwell Publishing, 2005), p. 548.
10
John Pier and Jos6 Angel Garcia Landa
follow. Such "antecedent futurity" (of which there are three basic forms) exemplifies perfectly well the trial-and-error dynamics of narrativity described by the author in his other publications: suspenseful prospection, curiosity-driven retrospection, surprise-generated recognition—all of which, as illustrated by the if-plot, serve to invest narrative with fluctuating degrees of uncertainty. One interest of the if-plot, built into the lawcode, is that it highlights a feature of legal storytelling which is present in narrative as a whole, and thus forms an integral part of a comprehensive theory of narrative. Taking exception to the "immanentist" theories of narrativity, John Pier adopts the logical fallacy of the post hoc, ergo propter hoc as a key to the concept of narrativity. An examination of Roland Barthes' suggestive proposal to treat this fallacy as "the mainspring of narrative activity" reveals that, for Barthes, the principle must ultimately be abandoned in favor of an "atemporal matrix-like structure" (Levi-Strauss), since it rests on a "logical error"—that of a "confusion between consecutiveness and consequence." However, on consulting the relevant writings by Aristotle (on müthos in the Poetics and on "non-cause as cause" in the Rhetoric) and in consideration of the fact that from the perspective of informal logic, fallacies may not be errors in all cases, it becomes apparent that Barthes (who is not alone in this regard) mistakenly superimposed errors in syllogistic reasoning on the problem of causality in the natural sciences (it is assumed that if Β follows A then Β is caused by A). A more fruitful way of looking at the relevance for narrative theory of the fallacy in question is to place it within a context of inferential reasoning in which, during the reading process, inferences of varying degrees of probability are drawn, subsequently to be confirmed, invalidated or revised, possibly in the light of unforeseen factors. To the extent that reading "linearly," from beginning to end, generates hypotheses, conjectures, etc. (a "naive" reading) that are then "tested" retrospectively (a "critical" reading), the theory of abductions (Peirce, Eco) provides a text semiotic basis for accommodating the formula "after this, therefore because of this" as a key to the processual nature of narrative. At the same time, the process triggered by the fallacy is also relevant to what Meir Sternberg has described as the "dynamics" of prospection, retrospection and recognition, this dynamics thus serving as a bridge between the heuristics of a "naive" reading and the semiotics of a "critical" reading of narrative.
Introduction
11
Narrativity and the perception of eventfulness and chance in narrative Working with the concept of "eventfulness," a key notion among the Hamburg narratologists, Peter Hühn begins with the observation that structuralist models of narrativity tend to be based on "mediacy" (or "perspectivity") and "sequentiality" (or "temporality").5 Some models, he notes, consider that change in narrative involves a violation of expectations, and he associates this feature with "point," insisting on the connection between narrativity and tellability: narrativity combines gradational qualities with binary narrativehood, while eventfulness is contextsensitive, an event being qualified as such according to its social, cultural and historical setting. Further specifications of eventfulness are provided with the incorporation of Lotman's concept of sujet (texts departing from a norm due to violation of an established boundary within a "semantic field") and of cognitivist schema theory (frames and scripts inferred from the text by the reader, deviation from which underscores the degree of noteworthiness of an event). For these reasons, "tellability (and ultimately narrativity) can be said to depend on eventfulness." Based on these criteria, it is possible to assess how events in narrative are perceived according to their cultural and historical contexts. The article concludes with an analysis of the manifest and revolutionary eventfulness of the story of social ascension in Richardson's Pamela as compared with the staged and illusory eventfulness in Joyce's "Grace," the story of the non-occurrence of the protagonist's spiritual rehabilitation. Werner Wolf takes up the problem of chance in narrative fiction. Noting that one of the functions of narrative is precisely to come to terms with chance in a world governed by change, he argues that the ways in which chance is positioned in relation to a given narrative, be it ruled out, bracketed or highlighted, can provide valuable insight into the worldview implicit in that narrative, worldview thus forming an overarching complement to the text's system of causation, its implied author and its perspectival structure. In narrative, chance occurrences, whether in the form of contingency, accident or coincidence, can materialize either at story level or at discourse level, but they are also apprehended in relation to a number of "filter factors" such as the individual reader, cultural and historical context and narrativity. Narrativity is of particular relevance to 5
Hühn's article also provides a lucid overview of narrativity-related research.
12
John Pier and Josi Angel Garcia Landa
chance in fiction to the extent that beginnings are inherently contingent, for instance, or that, given the constraints of narrative teleology, chance occurrences at the end of a story have a significance different from those found in the middle; moreover, the status of chance occurrences must be qualified in yet other ways if they result in a coincidental recognition in comedy or if they are overdetermined by the deus ex machina of tragedy. Wolf completes his essay with a highly pertinent analysis of chance and implied worldview in Pandosto, a sixteenth-century romance, and Tess of the d'Urbervilles. The former illustrates the inconstancy of fortune and the mutability of worldly affairs in interplay with providential justice. Hardy's novel, written in the age of realism, pits numerous natural incidents of chance against the secular, even Darwinian, urge for explanation and overdetermined plot-based causality with somewhat ambiguous overtones of tragic fate. Narrativity from pragmatic and performative perspectives Beatriz Penas Ibafiez integrates a "natural language text "-based concept of narrativity and a "natural language use(r) "-driven concept of narrativity. She underlines the consequences of assuming a perspective on narrativity either as a product's property or as a process evolving out of use and proposes an integration of both. Her balanced perspective on narrativity results in a backgrounding of issues characteristically dealt with in classical narratology and a foregrounding of other features derived from an understanding of narrativity as semiosis. Specifically, Penas Ibäfiez understands narrativity as a process of narrative meaning-making and interpreting "irreducible to form and grammar," reliant on a variety of factors which condition the configuration of a narrative. Place, time, society, a particular context of situation and culture, a tradition and an individually stylized performance are factors of diversity through which narrativizing takes place for specific linguistic-literary communities of readers and writers. Diverse ways of narrativizing lead to diverse narrativities, some standard and some non-standard. She finds prestige attached to the non-standard narrativities of authors like Vladimir Nabokov and Ernest Hemingway. While the proportionality between surface-to-bottom narrativity is deviational in Hemingway's "iceberg" texts, in the Nabokov text it is the proportionality between center-to-margins narrativity that is nonstandard. Their measure of deviation is read as a trope in accordance with the literary register that their narratives adhere to, while outside the pres-
Introduction
13
tigious literary register, the same or a lesser degree of deviation can easily lead from non-standard to plainly sub-standard narrativities. David Rudrum proposes a critique of the traditional theories of narrativity, theories which, he contends, overemphasize what narrative is—the "intrinsic" properties of a text that provide an "essentialist" view of narrative—at the expense of what narrative does—the performativity of narrating and its perlocutionary effects. Attempts to define narrative in terms that distinguish it from non-narratives, seeking to specify the degree of narrativity a narrative might possess, or theories that grant centrality to causation in narrative overlook the importance for storytelling of conventions and expectations found in linguistic communities. Although more recent reflection on narrativity has addressed the question of how a text is processed by readers and has introduced such notions as narratibility, tellability and point, these innovations, Rudrum argues, are irreconcilable with attempts to define narrative on the basis of formal or structural features, independently of context. In place of theories that embrace types, degrees and modes of narrativity, and in accordance with the Wittgensteinian notion of "family resemblances," he thus favors a pluralistic conception of narrativity in place of a general unifying conception, "identifying narrativity with Austin's 'perlocutionary effects' and thus deliberately going to the opposite extreme of 'essentialist' theories of narrativity." Narrativity and the demands placed on readers through given and/or multiple ordering of texts Jukka Tyrkkö discusses "kaleidoscope narratives," that is, fragmentary texts, printed or electronic, that exploit the potential for multilinearity and thus call for high levels of participatory involvement by the reader. More than the standard unilinear text, these narratives are highly reliant on the "links" ("decision-points"), or absence thereof, between textual segments or fragments, confronting the reader with alternative orderings of the text (as in Marc Saporta's Composition No. 1, a set of unbound and unnumbered pages) and thus displaying a variety of potential and possibly concurrent or even alternative plots. This situation poses a serious challenge for causal connectivity between narrative events, making the narrativity of kaleidoscopic narratives dependent on the perception of the reader. "Hypertextual links," writes Tyrkkö, "function not only as verbal connectors and as functional sites of interaction, but also as de facto narrative connections between fragments of the text." As a result of textual
14
John Pier and Jose Angel Garcia Landa
fragmentation and of the fragility or absence of links, readerly performance is intensified as the reader is left to explore potential narratives and untold stories. Among the works discussed are Richard Horn's Encyclopedia, in which fragments are arranged alphabetically rather than chronologically, and Geoff Ryman's electronic hypertext, 253, consisting of 253-word descriptions of 253 passengers on a London underground line, allowing for an untold number of potential links between the fragments and of the possible courses of action open to the characters. Michael Toolan focuses on the experience of narrativity, which he equates with the process of reading rather than with the conclusion or resolution arrived at by the reader. Conceding that one naturally reads for the plot, he points out that this takes place, in part, "against the grain"— that of the narrative texture—since uncertain predictions and tentative speculations are adopted while the effects of prolonging and postponement can be captured only at subsequent points of the text. To determine how the textual features of a text "guide" the reader to the thematic strands of that text, Toolan, working within a text linguistic framework, adopts techniques from corpus linguistics in an analysis of the opening paragraphs of a story by Alice Munro. The study shows that the link between the compound keywords of this story and its themes, structures and point are indirect but that, projected against the backdrop of the collocations established by a comparator corpus, the choice of certain words proves nevertheless to be crucial in the selection of interpretative possibilities. One of the qualities of this discussion is that the author is careful to take account not only of what is said (i.e. the key words that appear in the text), but also of what remains implicit within the context of utterance and of factors contributing to the "narrative interest" of Munro's story. The corpus analytical approach is admittedly restricted to the textual surface, ignoring intratextual links, for instance, and Toolan is careful to point out that the methodology employed is fairly rudimentary and in need of greater elaboration. Even so, his demonstration convincingly suggests that "[n]arrativity reflects our sense, as we read a narrative, not that there will certainly be an outcome, but only that matters as presented may 'come out', that change may occur."
Introduction
15
Narrativity and the transmedial expansion of narrative theory: transgeneric and transfictional dimensions In a reflection on transmedial narratology and in response to the need for a greater understanding of the kinds and degrees of narrativity, Ansgar Nünning and Roy Sommer take up the neglected question of narrativity in drama. A brief survey reveals that theories of narrativity tend to focus on narrative types (e.g. narrative as opposed to description or argument), but it is also noted that some theories are text-oriented (narratives possess certain properties that distinguish them from non-narratives) while others are reception-oriented (emphasizing cognitive experience o f an anthropomorphic nature). Adopting the view, essential to transmedial narratology, that narrative is not indissociable from a verbal act produced by a narrator, the authors call into question the traditional view that drama is distinguished from fiction by the lack of narratorial mediation. They also question the extent to which plot-oriented narratologies apply to drama, claiming that "plays do not just represent narratives (i.e. a series of events), they also stage narratives in that, more often than not, they make storytelling, i.e. the act o f telling narratives, theatrical." These realignments, which lay the way to a narratology of drama, also result in a distinction of considerable interest for the analysis o f theatrical representation: "mimetic narrativity" pertains to the causal and/or temporal sequence of events independent o f a given medium or communicative situation as opposed to "diegetic narrativity," bearing on verbal rather than visual or performative communication and thus "dramatizing" the act of narration in ways that are unavailable to written narratives, for example. The distinction also alerts us to the fact that narrativity functions differently from one medium to another and from one genre to another. Monika Fludernik, for whom drama is a narrative genre whose narrativity remains largely unexplored, draws attention to a paradox of traditional approaches that include drama among the narrative genres on the basis of plot while at the same time excluding it from narrative due to the absence of a narrator or of a narrative function. More satisfactory than a theory based on plot, she maintains, is a theory focusing on fictional worlds and/or experientiality. From the perspective of her "natural" narratology, she thus postulates as a minimum requirement for narrativity in drama the presence of a character on stage, the effect o f which is to place a consciousness within a space-time frame and to evoke a fictional world. Where narrative fiction and drama (together with cinema) differ is not so
16
John Pier and Josi Angel Garcia Landa
much in the represented world as at the "discourse" level, i.e. in the mediation of the represented world: in verbal narratives, setting is necessarily described but is more readily presented visually on the stage; drama opts for ellipsis where verbal narrative can resort to summary; etc. Also discussed is the alternative between dramatic discourse as "performance text" as opposed to "dramatic text" and dramatic discourse as synonymous with playscript, with the result that discourse is located between plot and staging. The problem of performance, rarely taken into account by narratologists, reveals a blind spot in communication models of narrative since, for example, the medieval bard is neither an author nor a narrator, but a narrative performer. Fludernik concludes with an examination of the potential for interchangeability between narration in drama (e.g. narrativization of stage directions) and performance in narrative (e.g. the use of stage directions in the "Circe" episode of Ulysses). Marie-Laure Ryan investigates the features of fictionality in narrative as they appear in one medium or another an also explores the consequences of these various manifestations from the perspective of narrativity. She first outlines four basic conditions of transfictionality: transfictionality involves a relation between two distinct texts by different authors; the texts are distinct but related to each other; it is assumed that the reader is familiar with the transfictionalized text; transfictionality seeks to preserve the immersive power of the narrative text. She then reviews the characteristics of these conditions with regard to the major technological periods in the history of communication: the oral age; the age of manuscript writing; the age of print; the electronic age; the digital age. It is noted, for example, that transfictional practices intensified during the age of print in reaction to the closely knit relations between author, text and fictional world. The digital age has important consequences for transfictionality to the extent that digital texts acquire something of the volatility of oral texts, the computer serves as a medium of both expression and transmission and the relations between text and fictional world are rendered problematical. The author studies several computer games that push the resources of transfictionality to the limit, observing that "[t]he Protean texts of digital culture may thus comply with the spirit of transfictionality, without respecting its letter." The article concludes with a discussion of the controversy surrounding the narrative status of computer games. Noting that texts, unlike life, can both be a narrative and have narrativity, Ryan approaches the issue by adopting a "fuzzy" conception of narrativity, locating computer games between life
Introduction
17
and the mimetic mode. On the one hand is "semantic narrativity" with certain core conditions (a story "conjure[s] the mental image of a world populated by individuated entities [that] undergo changes of states caused by events"), but also marked by a number of more or more or less constraining features (e.g. a story must have a beginning, a middle and an end). On the other hand, "pragmatic narrativity" concerns the medium or situation in which a story takes form, linguistic media being the best adapted, followed by film and drama and then by digital media where computer games, for instance, allow for greater inferential leeway. Narrativity and retelling Jose Angel Garcia Landa approaches narrativity from the vantage point of narrativization and narrative doubling, understood as interactional communicative phenomena. Narrative, as a dialogic phenomenon, is a rearrangement of previous narratives in order to articulate a new one, more complex or more to the point in a given interactional exchange. Effects of doubling ("narrated narratings," stories within the story, etc.) add semiotic intensity and suggest that repetition and retelling are basic to narrativity. Narrativization is therefore a remaking of previously narrativized events. Notions like "tellability," "point of view" and "event" need to be redefined in view of this interactional situatedness of narratives. Discursive phenomena involving the response to narratives, their use in conversation or criticism or their theoretical analysis also partake of this interactional situatedness. The connectedness between events characteristic of narratives (and which is subject to reinterpretation and retelling) is shown to be relative to the communicative dynamics of discourse interaction. Several definitions of narrativity are examined and criticized in order to emphasize the configurational dynamics of narrativity—a dynamics of constant remaking through communicative interaction. In this light, Garcia Landa addresses the retrospective dimension of narrative, in particular the "narrative fallacy" and its diverse aspects, such as the post hoc/propter hoc confusion, hindsight bias, foreshadowing, sideshadowing, the double logic of narrative, simulated contingency, etc. His narratological analysis extends into intertextuality, cognitive theory and hermeneutics and ends by taking up the question of narratives which retell or represent narrative acts. Literary works which narrate acts of narrating keep us aware of the continuity between everyday conversation and elaborate literary genres and build bridges between them, re-appropriating orality for literature and
18
John Pier and Josi Angel Garcia Landa
constructing complex interactional forms precisely through a return, with a difference, to the source of narrative interaction. The barebones summaries of the articles offered above are but a glimpse at the complete arguments developed by each of the authors. Of the many features of this collection sure to strike the reader is the fact that without, in most cases, adopting the same precepts and criteria and without arriving at the conclusions or thematic unity one might expect of a school of thought or of a monograph, the contributions provide insight into the phenomenon of narrativity from several different angles in ways that are sometimes complementary and sometimes divergent, but at all times illuminating. The points of comparison to be made are too numerous and open-ended and the implications and cross references to be investigated too subtle and far-reaching for these questions to be taken up in any meaningful way in these introductory comments. We leave it now to the reader to explore the contributions to this volume and to theorize narrativity. Tours/Paris and Zaragoza, July 2007
GERALD PRINCE (Philadelphia)
Narrativehood, Narrativeness, Narrativity, Narratability Narratology studies certain objects called narratives, what they have in common as well as how they may differ from one another, and there have been various (more or less restrictive) definitions of these objects, various answers to the question "What is a narrative?" Some theorists and researchers define narratives as verbal productions recounting one or more events while others define them as any kind of representations of events (including non-verbal ones relying on still or moving pictures, for example, gestures, or a combination thereof). Some maintain that they must involve causality, that they must be populated with individual beings and things, that they must be anchored in human experience, that they must constitute a whole. Others do not agree with most, some, or any of these prescriptions.1 As for myself, I will adopt the following definition in my discussion: an object is a narrative if it is taken to be the logically consistent representation of at least two asynchronous events that do not presuppose or imply each other. This definition, which is both flexible and limiting, has many virtues (apart from not conflicting with widely held views about the nature of narratives). For instance, the definition distinguishes between narratives and non-narratives (and, more specifically, between narratives and the mere representation of an event or activity, the mere description of a process or state of affairs). It also distinguishes between narratives and so-called antinarratives (e.g. Robbe-Grillet's Jealousy) by assigning consistency to the former. Moreover, however constraining it may be, it leaves numerous aspects of the objects it isolates indeterminate and makes room for a good deal of diversity. It does not, for example, broach the truth or falsehood of narratives, their factuality or fictionality, their ordinariness or artistry. Nor does it specify the kind of topics they address and the kind of themes they develop. Nor does it limit their poten1
See, for example, Bal (1997 [1985]); Barthes (1975 [1966]); Fludernik (1996); Genette (1980 [1972]); Herman (2002); Jannidis (2003); Revaz (1997); Ricreur (1984 [1983]); Ryan (1991); Schmid (2003).
20
Gerald Prince
tial magnitude or restrict their modes and media of representation. In addition, it will serve my demonstrative purposes. 2 While this and other definitions answer the question "What is a narrative?", indicate which entities constitute narratives (specify the class of objects designated by the noun "narrative"), and thus characterize "narrative" extensionally (an aspect of narrativity that I will call narrativehood), they also answer a slightly different question—"What is narrative?"— pertaining to a quality rather than an entity, focusing on an adjective rather than a noun, designating a set of traits rather than objects, and thus characterizing narrative intensionally (an aspect of narrativity that I will call narrativeness): any object is qualifiable in indefinitely many ways, as indefinitely many things, and an object is (qualifiable as) narrative if it exhibits (some of) the traits associated with narratives and, more specifically, if it is the logically consistent representation of at least two asynchronous events, etc. Thus, narratologists have been significantly (though not exclusively) interested in what makes an object—and, in particular, a narrative—narrative. As early as 1964, in "Le message narratif," Claude Bremond remarked that "what Propp studies in the Russian tale [...] is an autonomous layer of signification, endowed with a structure which can be isolated from the whole of the message: narrative."3 In Le recit, JeanMichel Adam underlines and develops this point. After bracketing the question of representational media, he identifies the narrative layer of signification as one of several textual types (the descriptive, the explicative, the conversational, etc.) found in texts. He also points out that texts are characterized by the combination of several types and the dominance of one of them. 4 Similarly, in Coming to Terms, Seymour Chatman argues for thinking of texts in terms of text-types, which he defines as "underlying (or overriding) structures that can be actualized by different surface forms." 5 For the purposes of his argument, Chatman focuses on three such structures: Narrative (which he capitalizes and explicitly distinguishes from "a narrative"), Description, Argument; he notes that they can coexist in the same text; and he shows how each can operate at the service of the others. For instance, Narrative operates at the service of Argument in La 2
3 4 5
For my definitions and distinctions (though not for my terminology), I have particularly profited from Fludernik (1996); Herman (2002); Ryan (1991), (1992), (2005). I have also relied on my own work (1982), (1999), (2005). Bremond (1964: 4). Adam (1999 [1984]: 9-10). Chatman (1990: 10-11).
Narrativehood, Narrativeness, Narrativity, Narratability
21
Fontaine's "Phoebus and Boreas" whereas the reverse is true in the "Ithaca" episode of Ulysses.6 There is, indeed, much more than narrative in narratives (comic effects, touching images, psychological insights, philosophical developments), and narratology aspires to account for Narrative (among other things)—that is, for entities insofar as they are narrative, for narratives in their narrativeness. Given my definition of a narrative, for example, the presence (or absence) of events, their number, and their temporal, causal, or implicational links would be narratologically pertinent, but comic effects or psychological insights would probably not. Several points can be made on the basis of these various takes and remarks about narrativehood and narrativeness, about narratives, narrative, and narrativity. On the one hand, consider an object A satisfying five of the six conditions for narrativehood singled out in my definition: it is the (1) logically consistent (2) representation of (3) two (4) asynchronous (5) events that presuppose or imply each other (given that last specification, it does not satisfy condition 6). A is not a narrative, and the same goes for an object Β satisfying only four of the six conditions or for objects C, D, E, and F satisfying none of them. However, it would not be entirely unreasonable to maintain that A is probably closer to constituting a narrative than Β and certainly closer to that state than C, D, E, and F. But things are perhaps more complicated. It is possible that the six specified conditions do not each have the same weight or importance: if event representation is central, asynchronism may weigh more than consistency, for instance, and the two-event requirement may weigh more than either. It is then possible for an object satisfying three conditions (beyond the central ones) to be closer to constituting a narrative than another object satisfying a different set of three conditions; and it is even possible for an object satisfying two conditions to be closer to constituting a narrative than an object satisfying three. On the other hand, if text-types are more or less dominant in different texts, it is not only possible for the Narrative type to be dominant in one text and ancillary in another; but it is also possible for the Narrative type to be dominant (or ancillary) in two different texts, albeit more so in one than in the other. Furthermore, it may occur that, because of its strong dominance in a given text, a quasi-Narrative type (say, a structure realizable by texts satisfying only five of my six definitional conditions) would give more narrative character or weight to that text (thus making it more
6
Ibid.: 6-21.
22
Gerald Prince
narrative, as it were) than the Narrative type would bestow on other texts because of its marked ancillariness in them. The various possibilities I have mentioned point to the existence of degrees of narrativity, i.e. degrees of both narrativehood and narrativeness. Some objects are narratives; some are quasi-narratives; and some are not narratives. Some narratives are more narrative than others; some nonnarratives are more narrative than others; and some are even more narrative than narratives. Now, though these various possibilities are a function of quantitative factors related to necessary and explicit definitional conditions (number of conditions satisfied, total weight of conditions satisfied, relative weight of Narrative type in a textual economy, etc.), some possibilities can (also) be a function of other, qualitative rather than quantitative, factors related to optional or definitionally implicit elements as well as necessary and definitionally explicit ones. These qualitative factors more or less significantly foreground the narrative identity of a given text, the nature or the import of a Narrative structure underlying it, that of some structural constituents, or that of some properties of the latter, and they thus affect the (degree of) narrativity of that text beyond any strict definitional considerations. Or else these qualitative factors derive from the fact that certain textual components without satisfying this or that definitional condition required for any narrative, tend more or less clearly to do so. In other words, the narrativity of a text is not merely equivalent to its narrativehood and narrativeness: it depends more generally on whatever makes a text more (or less) immediately identifiable as (a) narrative. One textual qualitative factor would be the positiveness of the events depicted, since narratives are representations of events and not of their mere possibility or of their negation. Narratives live in certainty: this happened and then that; this happened because of that; this happened and it was linked to that. Though they need not preclude hesitations or speculations or negations—in fact, these factors can generate suspense or function as signs of objectivity or emphasize the qualities of what actually happened—and though, at least in their verbal manifestations, they can be hospitable to interrogative, conjectural, and negative sentences, narratives can perish under the effect of sustained indecision and ignorance. Another factor has to do with the discreteness of asynchronous events, since the more such discreteness obtains, the more asynchronism is emphasized. Consider, for example, "She ate and then she went to bed" as opposed to "She went to bed after eating," or "She had a million dollars and then she lost it" as opposed to "She lost the million dollars she had." Furthermore,
Narrativehood, Narrativeness, Narrativity, Narratability
23
as indicated above, a definitional element may not be present in a given text but something more or less similar to it may be. Thus, even if a textual segment like "it rains and hails and snows" or like "it rains and rains and rains" does not quite function as a representation of asynchronous events, it is perhaps not that far from doing so. Of course, these qualitative factors—like the quantitative ones—may not be equally important. Of course too, their effect may vary from one textual economy to another. Just as various textual factors (quantitative or qualitative, necessary or optional, operating at the level of the narrated or that of the narrating) make some objects closer to narratives than others or more prototypically narrative or more narrative-like than others, various contextual factors make some objects more readily viewable or processable as narrative(s). Consider, for instance, a text like "She was very rich and then she developed a passion for poker and then she lost all her money" as a response to "Tell me a story about her" or as a response to "Give an example of an English sentence made up of three conjoined sentences."7 Context (along with text, no doubt) also influences the effectiveness qua narratives of various objects, their narrative value (as distinct from their narrative identity and character or their narrativity), what is characterized as their reportability or tellability (or narratability, as I will call it), and context—along with text—helps to answer questions such as "What makes a narrative worth producing?" or "What makes an object a successful, appealing, valuable narrative?" In his classic discussion of the "point" of a given narrative—the reason for which that narrative is produced and the essential matter it is getting at—William Labov emphasizes that "[p]ointless stories are met (in English) with the withering rejoinder 'So what?' Every good narrator is continually warding off this question: when his narrative is over, it should be unthinkable for a bystander to say 'So what?' Instead, the appropriate remark would be 'He did?' or similar means of registering the reportable character of the events of the narrative."8 Now, it is often the case that, to some extent at least, the raison d'etre of a narrative and its import, its significance, its worth, are indicated or suggested by so-called evaluative devices. Thus, the reflections of one character or another may underscore the unusual quality of the situations represented, the repetition of certain events may foreground their 7
8
There have been several interesting discussions of various factors affecting (categories akin to) narrativity. See, for instance, Coste (1989: 52-72); Herman (2002: 85-113); Ryan (1991: 48-174); Schmid (2003). Labov (1972: 366).
24
Gerald Prince
value, the use of disnarrated elements—explicitly referring to what did not take place but could have—may emphasize the quality of what did take place, and commentary by a narrator may directly state the point of the narrative. 9 Furthermore, it may be that certain topics and themes are inherently interesting and that their mere narrative deployment makes a narrative worthwhile. A sexist French recipe for successful narratives valorizes ingredients like mystery, religion, sex, and aristocracy: "My God, said the Duchess, I am pregnant. Who done it?"; 10 and an old Readers ' Digest formula is supposed to stress—along with sex and religion— personal experience, travel, money, and the animal kingdom: "How I Made Love to a Rich Bear in the Alps and Found God." Moreover, perhaps the very fact that any narrative brings together distinct events and links them temporally in a non-contradictory manner always sheds some light on the configurations or operations of time, thereby making a point. Yet, as Labov's remarks (and my own experience) more than suggest, there are perfectly well-formed narratives (satisfying every condition for narrativehood) which depict and account for changes in situations and totalize them into sequences of interrelated parts but which are greeted nonetheless with "So what?" rather than "He did?" Moreover, important topics and themes like sex and mystery are neither specific to nor characteristic of narratives as opposed to non-narratives, and they do not always play equally well in Paris and Philadelphia. Conversely, perhaps any topic or theme—however local or insignificant—not only constitutes potential material for a narrative (is narrativizable) but also can be made narratively appealing (is narratable). Perhaps any set of events—however unimportant or trivial—not only is susceptible of narrativization but also can be endowed with narratability. In any case, that is what much fiction endeavors to show; and I, for one, find interesting any narrative that pertains to me. As for evaluative devices, they may not always prove quite appropriate or persuasive. After all, claiming that (sequences of) events are unusual, extraordinary, bizarre, unfortunately does not suffice to make them so. In short, even if narratability often depends on strictly textual factors—e.g. the nature of the events narrated and of their configurations (or, at the level of the narrating, the nature of the focalizations selected, of the speeds exploited, of the orders of presentation adopted)—it also always 9
On the notion of the disnarrated, see Prince (1988). This is Margaret Boden's English rendition of "Mon Dieu, dit la Marquise, je suis enceinte et ne sais pas de qui" (Boden 1977: 299), quoted in Ryan (1991: 154).
Narrativehood, Narrativeness, Narrativity, Narratability
25
depends on contextual ones. Sometimes, a (spatio)temporal feature proves particularly significant: "Napoleon died" functions differently in 1821 France or England and in 2005. Sometimes, alethic status plays a crucial role, as Marie-Laure Ryan explains: "extraordinary events work better in factual than in fictional narrative, because they are too easy to make up." 11 Sometimes too, the medium and circumstances of manifestation are important. Thus, two different manuscripts of the same narrative may be at different stages of disintegration; or the radio transmission of a story may be more powerful in the attic than in the cellar; or a narrative painting may be more visible in the morning light than in the evening. Frequently (not too say always), narratability is a function of the receiver. As Mary Louise Pratt points out, something like "Bill went to the bank today" is narratable if Bill is known to be a miser who has always mistrusted banks, but much less newsworthy if he is not.12 More generally, different persons (or the same person on different occasions) can greet the same story with "So what?" or with "He did?" I may have heard this or that tale a hundred times or not even once; I may long to hear it anew, but then again I may not; I may want to learn what happened and why, but you may not; I may find an account of what took place fascinating, and you may disagree. Note that there are, of course, many other contextual factors of narratability to take into account (context is inexhaustible and, however much a text contextualizes what it represents, it never does so completely). For example, there is the actual sender of a given narrative, as opposed to its author, its narrator, or this or that performer of the narrative. Suppose that someone mails me a narrative written by X featuring a narrator Y and which is to be read aloud to me by Ζ. The identity of the mailer may well affect my response to the narrative. Or, to focus on the author instead, suppose that X's fame is great. This, too, may influence my evaluation. Note also that the narrative value of an object is not equivalent to its overall value. Whatever appeal Sartre's Nausea may have, for instance, at least some of it is probably not narrative, but philosophical. Further narratologically pertinent discriminations can no doubt be effected, and further narratologically pertinent domains can no doubt be discovered or explored. Still, even this brief discussion not only addresses basic questions regarding the character, form, and functioning of narratives, but it also generates many other (narratological) questions worth 11 12
Ryan (2005: 590). Pratt (1977: 135).
26
Gerald Prince
exploring: on the nature, weight, and interaction of the factors affecting narrativity, for example; on the relation between particular representational media and narrativity (or narratability); on the acquisition of the capacity to manipulate narrativity; on the different kinds of factors relevant to narratability that different groups (or different genres) exploit; and so on and so forth. Indeed, the systematic study of what in an object and its setting (what in a text and its context) encourages its identification as (a) narrative, triggers the calling up of certain grids for processing it as such, and affects the value of such an invocation can clarify the character and specificity of narrative semiosis, the meaning and meaningfulness of narratives and narrative moments. References Adam, Jean-Michel 1999 Le Recti. 2 nd ed. (1 st ed. 1984). Que sais-je? (Paris: Presses Universitaires de France). Bal, Mieke 1997 Narratology: Introduction to the Theory of Narrative. 2 nd ed. (1 st ed. 1985) (Toronto: University of Toronto Press). Barthes, Roland 1975 [1966] "An Introduction to the Structural Analysis of Narrative," in New Literary History 6: 237-62. Boden, Margaret 1977 Artificial Intelligence and Natural Man (New York: Basic Books). Bremond, Claude 1964 "Le message narratif," in Communications 4: 4-32. Chatman, Seymour 1990 Coming to Terms: The Rhetoric of Narrative in Fiction and Film (Ithaca and London: Cornell UP). Coste, Didier 1989 Narrative as Communication. Foreword by Wlad Godzich. Theory and History of Literature 64 (Minneapolis: University of Minnesota Press). Fludernik, Monika 1996 Towards a 'Natural' Narratology (London and New York: Routledge). Genette, Girard 1980 [1972] Narrative Discourse: An Essay in Method, translated by Jane E. Lewin. Foreword by Jonathan Culler (Ithaca and London: Cornell UP). Herman, David 2002 Story Logic: Problems and Possibilities of Narrative. Frontiers of Narrative (Lincoln and London: University of Nebraska Press).
Narrativehood, Narrativeness, Narrativity, Narratability
27
Jannidis, Fotis 2003 "Narratology and the Narrative," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 35-54 (Berlin and New York: Walter de Gruyter). Labov, William 1972 Language in the Inner City (Philadelphia: University of Pennsylvania Press). Pratt, Mary Louise 1977 Toward a Speech Act Theory of Literary Discourse (Bloomington: Indiana UP). Prince, Gerald 1982 Narratology: The Form and Functioning of Meaning. Janua linguarum, Series Maior 108 (Berlin, etc.: Mouton). 1988 "The Disnarrated," in Style 22: 1-8. 1999 "Revisiting Narrativity," in Grenzüberschreitungen: Narratologie im Kontext / Transcending Boundaries: Narratology in Context, edited by Walter Grilnzweig and Andreas Solbach, 43-51 (Tübingen: Gunter Narr). 2005 "Narrativity," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 387-88 (London and New York: Routledge). Revaz, Franfoise 1997 Les textes d'action. Recherches textuelles 1 (Metz: Universiti de Metz). Ricceur, Paul 1984 [1983] Time and Narrative, vol. 1. Translated by Kathleen Blarney and David Pellauer (Chicago and London: The University of Chicago Press). Ryan, Marie-Laure 1991 Possible Worlds, Artificial Intelligence, and Narrative Theory (Bloomington and Indianapolis: Indiana UP). 1992 'The Modes of Narrativity and Their Visual Metaphors," in Style 26: 368-87. 2005 "Tellability," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 589-91 (London and New York: Routledge). Schmid, Wolf 2003 "Narrativity and Eventfulness," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1:17-33 (Berlin and New York: Walter de Gruyter).
MEIR STERNBERG (Tel Aviv)
If-Plots: Narrativity and the Law-Code 1. Introduction 1.1. Statute or Story: Why Exclude the Law-Code from Law-telling? In Aristotle's Rhetoric, and the discipline associated with it, narration makes an integral and conflictual part of forensic oratory. To argue or prove one's case, one must represent the events at issue, and better than the opposite side. Argument entails telling, or retelling, what happened, to the desired effect. Subsequently, however, this idea of legal discourse and the discipline itself changed, once the lawyers took over: the court drama got reconceptualized into the esoteric, hence also more prestigious, terms of jurisprudence. With advocacy no longer a persuasive art open to all aspirants (the manual's implied readership) but a specialized profession, the interlinkage between its two elements went underground. Reasoning came to overshadow telling in the guild's eyes, theory, education— regardless of practice. Not least a matter of image and self-image, this, because professionals always seek to distance themselves from what comes naturally. Over the last decades, however, the approach to the legal process as narrative-driven has been regaining currency in various quarters. Its exponents range from jurists to linguists to semioticians to cognitive psychologists to literary-minded scholars and narratologists. Like the respective disciplinary viewpoints, the participants in the legal process that have been treated as storytellers vary, or multiply, among the analysts. Taken together, they extend from the lawyers to the laity involved: from counsel and client in consultation, through the adversary attorneys pressing their cases at the trial and/via the witnesses on either side, to the jurors and the
30
Meir Sternberg
judge or, with the appellate courts, judges. 1 "This endless telling and retelling [...] is essential to the conduct of the law. It is how law's actors comprehend whatever series of events they make the subject of their legal actions. It is how they try to make their actions comprehensible again within some larger series of events they take to constitute the legal system and the culture that sustains it."2 Further, among "the many layers of storytelling involved in any adjudication before the law," the higher the court, the higher the level of narrative engagement. The top judiciary "must pay particular attention to the rules of storytelling, the conformity of narratives to norms of telling and listening," or "must braid together the story of the particular case at hand and the history of constitutional interpretation." 3 This narrativist line of inquiry is still young and uncoordinated: a mirror image, for better or worse, of its unitary rhetorical predecessor. It actually branches out into assorted sublines—divided by training or range or subject matter or culture or geography—nor do those more akin often meet. The usual atomism of the humanities. Yet enough has been achieved by the law-as-narrative program for it to start (re)establishing itself alongside existing traditions in the field, whether the natively jurisprudential or the imported law-as-language approach. 4 The more so because these traditions themselves contain practitioners who have taken up the narrativist idea or independently rediscovered and applied it. However, across the various methods and instruments and emphases of the most recent trend, the legislator is conspicuously absent among the I wish to thank John Pier and Jose Angel Garcia Landa for their careful reading and editing, as well as for their patience. Thanks also to Otto Pfersmann for sending me a number of his papers on the law, especially the one cited in this final version. 1
For a recent overview, see Brooks (2005); for a list of references, see Amsterdam and Bruner (2000: 3 5 5 - 5 6 , η. 2). Though useful, both focus on the American scene and on directions related to their own; typically so. Compare the last paragraphs o f 1.2 below.
2
Amsterdam and Bruner (2000: 110). Brooks (2005: 416). On the latter (and of course much later) enterprise, see useful overviews and bibliographies in Danet (1980), (1985), (1990); Maley (1987), (1994); Gibbons (1999), (2003), or the wide-ranging collection in Gibbons (1994). Note that some of these appear in volumes or journals devoted to discourse at large: a measure of parallel establishment in the linguistic field. Similarly, Brooks (2005) has made its way into A Companion to Narrative Theory and Porter Abbott's lively chapter on "Narrative Contestation" (2002: 1 3 8 - 5 5 ) into a textbook.
3
If-Plots: Narrativity and the Law-Code
31
group of legal narrators. Or perhaps inconspicuously absent, because orthodoxly so—the common legacy of all earlier projects, rhetorical, juristic, or language-oriented. You might think that storytelling begins only after the law-code, with the passage from letter to life: from the statute book to the troubled arena of law enforcement and administration, all the way to the final sentence. Indeed, that marked absence, or dividing line, often surfaces in the form of pointed exclusionary opposition: It is in the nature of the corpus juris to prescribe general rules about what is permissible and impermissible in delineated spheres of human activity and to establish institutions and procedures for constantly instantiating those rules with reference to specific cases. [...] Narrative in its very nature makes it humanly possible to relate the Grand and Timeless principles of a corpus juris to the current particularities of the cases we adjudicate or arbitrate, or those in which we help our clients comprehend the circumstances of their lives within the framework of the law or vice versa. [...] Narrative's inherent structure fits it for this task.5
Observe how the opposition grows here qualitative, ungradable, and the mutual exclusion absolute, because cast in essentialist terms: as though "inherent" in the law-code's "general rules" vis-a-vis the lawyerly "narrative" that alone associates them with living "cases ... circumstances," told between humans about humans. The very "natures" of the opposites are allegedly poles asunder, with (non)humanity thrown in for good measure. To sharpen the polarity further yet, the corpus juris appears in the aggregate—as if its "rules" were indivisible—and "narrative" in the singular, implying its linkage to this or that case at issue. The rhetoric of binary essentialism apart, however, the authors in effect suggest, or echo, more arguable and oft-invoked points of contrast. One lies in the frequently alleged "Grand [i.e. 'general'] and Timeless principles of a corpus juris" as against the "particularities of the cases we adjudicate," time-bound by implication. To spell out the latent argument: not only is each of these cases (e.g. of libel or homicide) "current," newly arisen, while the rules stand, ever-applicable, governing them all and their yet unborn equivalents. Such cases also necessarily relate to events and circumstances located in definite past spacetimes: an act of killing, say, perpetrated on a certain individual, there and then, by another. But contrast the statutes in examples (l)-(3) in section 3 below. They govern everyone "who kills a beast" or "has newly taken a wife," every place5
Amsterdam and Bruner (2000: 140^11).
32
Meir Sternberg
ment "on trial" or "cowardice" in battle, anytime, anywhere. The lawcode must generalize the events, agents, coordinates into "timelessness," as well as otherwise departicularize them, in order to regulate every possible act of the kind. But narrative—so the argument would conclude— inherently, hence exclusively, concerns temporalized "human activity." Statute is one thing, as it were, story another. This conception of the law-rule is in fact highly traditional, and oddly so, because it recirculates the old legistic wisdom that you would expect a narrativist inquiry to rethink, if not to reverse. (Concerning the key term itself, What is a statutory "rule" a rule o/?, is the first question to ask.) Even beyond jurisprudence, this tradition boasts, inter alia, Vaihinger's famous Philosophy of 'As i f . To him, the law is as general as mathematics, as remarkably suitable "for the deduction of logical laws and their illustration, or for the discovery of logical methods." Hence the need for "juristic fictions" to subsume the unlaw-like variant under the rule by analogy. "Since laws," juristic or mathematical, "cannot include within their formulae all particular instances, certain examples of an unusual nature are treated as if they belonged to them": as if a grandfather were a parent, say, or an umbrella a weapon. 6 Logicized to this extent, with misfits analogized and assimilated for extra coverage, the statute is in effect again denarrativized, indeed dechrono-logized. The one-sidedness accords with the widespread tendency to throw the two Aristotelian components out of balance: law-telling neglected in favor of reasoning. Apropos the statute, however, the imbalance persists, even when it has otherwise been newly redressed. After all, Aristotle himself would find it in order here. Among the newer approaches, these lines of argument are equally typical, down to the ground for the exclusionary polarity. The above quote about the corpus juris, from a jurist working together with a known cognitive psychologist, thus matches the following statement from a notable practitioner and promoter of law-as-language. "Law is language," which manifests both "a code of laws and processes for applying them and disputing their application. This distinction, between the static or codified, and the dynamic and dialogic aspects of the law is useful": of the two, the second has become established "as narrative," so that "prosecution and defense cases [are viewed] as containing competing narrative representa6
Vaihinger (1924: 3 3 - 3 5 ) . One could multiply parallels from other sublines within the narrativist circle itself. For example, Martha Minow opposes the law's "commitments" to those of "narrative," which "revel in particularity, difference, and resistance to generalization" (1996: 35).
If-Plots: Narrativity and the Law-Code
33
g
tions of the same reality." "Static or codified" evidently corresponds to "Timeless principles"; "dynamic and dialogic," to "the current particularities of the cases we adjudicate" through narration. By the same criteria, one might, or should, also exclude from the genre such discourse kinds as parable and fable and joke. All would appear likewise departicularized and virtually "timeless" in scope; the first two are, moreover, likewise normative and intended to regulate "human activity" in general. Yet they all count as narrative subgenres, and no approach to the law would disqualify them along with the statute. On the contrary, Amsterdam and Bruner themselves illustrate from the fable the terminal element in their Austere Definition of narrative: ρ"the coda—say, for example, Aesop's characteristic moral of the story." Same feature, opposed groupings, and to the limit: what rules out the one discourse kind makes the other exemplary. Why the unequal treatment? It may be that the family likeness escapes notice, or that surface or quantitative variations get elevated to differences in kind. But perhaps still another feature lurks behind the respective groupings, to polarizing effect, right or wrong. For example, consider what the following statement assumes and implies: "To the extent that law is fact-contingent, it is inescapably rooted in narrative." 10 This means that lawyers construct stories with a claim to factuality, historical narratives about some past affair. By implication, this factuality may elsewhere give place to fictionality—without detriment to narrativity—as long as the fictional substitute remains fact-like, history-like, in being past-directed (e.g. events in an Aesopian fable and in Remembrance of Things Past alike). Inversely, the law-code deals not with any accomplished happenings, factual or fictional, but with possible futures ("what is permissible and impermissible in delineated spheres of human activity"). These possible futures are by definition not yet actualized—with the "impermissible," hopefully never—and conditional upon another, antecedent futurity at that, often expressly so. "If thou buyest a Hebrew slave, six years he shall serve" (Exodus 21: 2): first prospective buying, then service. Hence, to the extent that law-as-narrative is "fact-contingent," the statute's multiply future " I f . . . then" contingency wouldn't qualify for storyhood. 8
Gibbons (1999: 1 , 3 ) .
9
Amsterdam and Bruner (2000: 114), after Labov; Bruner (2002: 20, 117, η. 2). Amsterdam and Bruner (ibid.: 111). Or more explicitly, "Law looks to the actual, the literal, the record o f the past," so that "legal pleading" is "historically established" (Bruner 2002: 6 1 - 6 2 ) .
34
Meir Sternberg
All these arguments pro and con that run across approaches to the law—including the implication just spelled out—find overt large-scale parallels in narrative theory itself. Thus, Narrative shies away from abstraction and thrives on concreteness, it concentrates on the particular and not the general. [...] Or, to put it differently, narrative prefers tensed statements (or their equivalent) to untensed ones. [...] Narrativity [...] is also a function of the extent to which their [the represented events'] occurrence is given as a fact (in a certain world) rather than a possibility or probability. The hallmark of narrative is assurance. It lives in certainty: this happened then that; this happened because of that."
The correspondences leap to the eye: particular not general, tensed not untensed, fact not contingency, only that all become genre-wide here. And the n a r c o l o g i s t ' s relative hedging goes with a stronger emphasis on the criterial value that the genre allegedly ascribes to factuality, assurance, certainty "rather than possibility or probability." In another, more specialized shorthand, narrative thus gets firmly anchored (on top of everything else) in categorical as against modal discourse.
1.2. Statute and Story: Aligning Narrative with Modalized Discourse I have elsewhere challenged this anchorage on several fronts, beginning with my very redefinition of narrative since the 1970s. Against the Aristotle-old consensus, narrativity accordingly shifts its locus and life from the represented process to the interplay between the represented and the discoursive process. It lives not in the mimesis, far less categorical mimesis, of events, happening, action, development etc. per se, but in the (re)construction of those events along the gapful sequence as told and read, hence in the threefold dynamics of closure on the move by trial and error: suspenseful prospection, curiosity-driven retrospection, and surprise-generated recognition. To these universale of narrative we'll come back. For now, let me just point out their inherent modality. Arising from a gapped future or past, all three dynamics entail multiple (ambiguous, uncertain, hypothetical, reversible) gap-filling, necessarily a matter of "possibility or probability" rather than "fact." And with permanent gaps— like an unresolved conflict or enigma—so our attempts at settling them must remain to the end. 11
Prince (1982: 149). In a note ad loc. (ibid.: 172, η. 4), "historical narrative," recounting "a series of facts from the past," again counts as paradigmatic. See also note 14 below.
If-Plots: Narrativity and the Law-Code
35
Given such orientation to the dynamic interplay between times—our eventful tentative processing in sequence of an ambiguous event sequence—not only narrativity and narrative are reconceived but modality itself as well. In or beyond this genre, it no longer depends on the markers of necessity and possibility (e.g. must, may, will, probably, perhaps, unlikely, Suppose ..., If ..., then) overfocused by logicians: their occupational bias toward the manifest (grammaticalized, lexicalized) and analytic formalism has unhappily restricted and confused the tracing of modality in, on, below, along living discourse, literature and law included. Rather, even when no modal operators surface in the language, or allow tidy notation, our mind always can, and often cannot but, supply them in response to the uncertainties entailed by world-making as a process of sensemaking underpressure: from human epistemic limits, (con)textual exigencies, artful licenses. The simplest, most categorical-looking representation ("A did Β," "X happened"), therefore, leaves why-when-where gaps, and unspoken forked futurities that project ahead, for us to settle as best we can. Once you observe and generalize this universal process of tentative worldmaking in sequence, there emerges a network of further basic distinctions, hence options, alien to any logic-based idea and repertoire of modality. Qualified discourse thus varies with the gaps responsible for it, as does, inversely, its categorical opposite, or in other words, the problem of deciding between them. Either subdivides according to whether it is a feature (1) apparent and temporary, in effect deceptive (arising only to complicate, even reverse itself) or genuine and enduring (to the end); (2) perceptibly ambiguous (i.e. gapped, whether into suspenseful prospection or curiosity-driven retrospection) or imperceptibly so (with a view to belated surprising recognition); (3) disambiguated in the sequel or left equivocal; as well as marked in the text or inferrable in context. All our discoursing/reading about the world entails movement, or hovering, among these interlinked polarities of modality and/vs. categoricalness—now with quick, even automatic resolution, now with doubt and effort, even in(de)terminably. All, moreover, in keeping with my Proteus Principle, whereby the same function (e.g. modalizing) lends itself to different forms (e.g. verbal or mental, explicit or implicit, univocal or equivocal), as vice versa. And all along the line, narrative is the exemplary case, because generically richest in the ends and means and interplays of troubled processuality.
36
Meir Sternberg
This definitional modalizing of the genre applies, inter alia, to the subgenre of legal (hi)storytelling: the narratives recounted at the trial by counsel, witness, judge, for example, all generating in the addressees much the usual dynamics of uncertainty. That they stand or fall by "reasonable doubt" is enough to assimilate them to the genre's rule, more institutionally than usual, if anything. So their ostensible polarity with the legislator's statute already begins to reverse into deep, operational analogy. Further, and again including the entire legal branch, the generic modality also extends at will to various particular forms manifest on the surface of narrative. Nor are these limited to occasional markers that qualify a sentence here or there. In the common sequence of telling and retelling that I call repetition structure, the "member of forecast" (wish, prophecy, order, or indeed law-directive) freely interacts with that of "fulfillment" (narrated after the event) to build up a plot. Man proposes, for example, God disposes, or vice versa. But even where unfulfilled in reality—left modalized—the forecast still operates as a discourse event: no less than does a categorical retrospection on history past and with equal consequentiality as such. A wish thwarted, a proposal declined, a prophecy unrealized, a command(ment) spurned, has been voiced nevertheless and may therefore always trigger some response, too. An act, verbal as otherwise, and of whatever modal status, is an act, down to its latent causal force, 12 impact, repercussions in the world. Again, reported discourse cuts across the categorical/modalized polarity, sometimes opting for one pole, sometimes for the other, sometimes for their tense, interactive juxtaposition. This licensed variability shows even in direct report, allegedly a verbatim copy of the original and so most factbound. Instead of reproducing a past utterance or thought—as the "direct speech fallacy" would have it—such quotation freely turns ^reproductive ("He will/shall/may say: ' . . . " ' ) or otherwise qualified, not just away from factuality but even into counterfactuality: "He would/should/might have said: Along a different line of qualification, there is literal nonfactuality—quoting under a negative sign. Thus the nonevent that I call "the minus-insideview," such as "she did not think: ' . . . " ' or "He did not know that ..." Both of these nonevents are remarkable. In the former, her nonthought assumes an immediate, monologue-like directness; in the latter (free) indirectness, his epistemic minus presupposes an ontic plus. There, 12 13
Sternberg (1978: 6 7 - 8 6 ) , (1985: 3 6 5 ^ 4 0 ) , (1998: Index under Repetition). E.g. Sternberg (1982: 137ff.); Fludemik (1993: 4 0 9 - 1 4 ) .
If-Plots: Narrativity and the Law-Code
37
the nonfactuality of his knowing goes with the factivity of "know" (as against "think") and so with the objective factuality of the knowledge quoted: the truth given in the that-clause is denied to the subject by the main clause. 14 Again, the usual nonfactivity of a quoting verb like "hear" may reverse into encoded factivity, or vice versa. 15 And to compound the distance from the polar categorical report, the assorted modalizing lines freely cross: e.g. "She couldn't have said/thought/known/heard ..." The consequences reach much further than appears. Quotation makes, of course, a central narrative act, operating in the fullest sense of the word and over a broad range. It includes all the discourse-events (vocal, mental, inscriptional, speech, talk, writing, thought, hearing) re-presented anywhere in any of the various forms and, again, modalities available to discoursing at second-hand. It also can (and in fiction, must) run to an entire narrative, as when Sterne directly quotes Tristram; or better, if we consider the modalizing fictional remove, quotes what Tristram is supposed to have narrated in writing. Yet compare an editor reproducing a genuine testimony, or yourself citing a day dream, in life itself. Further, quotation intersects with repetition structure, whose members of "forecast" and "report" originate in characters speaking/thinking before and after the event, respectively, with the former qualified by nature. If anything, the qualifying modes open to quoting apply a fortiori to the telling about nondiscoursive objects (occurrences, entities, states). For, as such, their representation can never emulate the semblance of "verbatim" exactness and certitude given to the direct re-presentation of discourse by discourse. That the semblance often proves formal and misleading there, or indeed grows qualified, only strengthens the point of contact between quoting and narrating. But their strongest, unbreakable contact here is generic. For quotation also entails ambiguity, as a montage of discourse-events (the quoted and the quotational, with their respective voices and viewpoints); hence entails uncertainty in the unpacking. Which of the given elements belong to the quotee, which to the quoter? The reader can only hypothesize, the way we necessarily do vis-ä-vis the first-order world of action unfolding in the rest of the text. So, even what isn't formally marked as modal in either domain (and wouldn't count as such among logicians or their liter14
15
Sternberg (1979: 116-17), and (2001a: 2 2 6 - 3 9 ) , within a larger theory of negation. The minus-quote relates to what Prince (1988) later called "the disnarrated"; but whether and how this accords with his tying of narrativity to "fact ... certainty" (as seen above) is another question. Sternberg (1979: esp. 115ff.), (1986), (1998: esp. 3 1 7 - 2 4 , 4 0 5 - 1 1 ) .
38
Meir Sternberg
ary and other followers) assumes operational modality within the genre's dynamics of gapping and gap-filling. In all these regards, the directspeech fallacy typifies the larger narrativist, or narratological, bias against mere "possibility or probability." More recently, I have extended the principle to legislation itself as— next to the genre-constituting universale—both the most decisive and the widest front. 16 The most decisive, because a law is always, necessarily, modalized out of factuality in the ways outlined; and also because it invariably compounds this feature with what its disqualifies label as generality, timelessness, staticness, or the like. In brief, the variables open to repeating or quoting grow jointly constant here, from beginning to end. And the widest front, because a law co-occurs, often "interplots," with others to form a law-code; and also because it incorporates the narrativities of repetition and quotation among its makings, workings, aids to processing and understanding. For now, just consider how the Bible's Moses quotes to the people the laws spoken to him by God, or repeats, indeed re-quotes, in Deuteronomy statutes already given in Exodus, with tell-tale variants. Among later parallels, King John's Magna Carta quotes to the nation the liberties granted by him at Runnymede; the so-called enacting formula that legislatures use (e.g. "Be it enacted by the Queen's most Excellent Majesty ... as follows") in effect places the new law enacted thereby in a deontic quotational frame; 17 and the subsequent promulgation frames that frame, along with the inset, in reporting the new enactment verbatim on public record. Further, the inset statutes may in turn quote the gist of the agent's prospective thought (e.g. malice) or, with language crimes, speech (e.g. slander) or writing (e.g. libel), always to yet another modalized, if now telescoped, effect. In such light, finally, the stories told along the legal process itself stand revealed, once again, as far from sheer factuality: they not only quote the law but also model themselves on it in developing their versions of the actual, historical past— repeat it in another key—and can even turn hypothetical themselves for the sake of argument or contingent judgment. The age-old binarism there-
16
Sternberg (1998: 4 2 6 - 6 3 8 ) , esp. chapter 9: "Law, Narrative, and the Poetics of Genesis."
17
More on enacting formulas, within a speech-act approach to the law, novel at the time but otherwise traditionally Austinian, see Kurzon (1987: 9 - 1 2 ) . Austin (1973 [1962]: 22, 7 0 - 7 1 , 104, 121) notoriously discounts quotation as "parasitic" on ordinary language, together with joking or fictionalizing.
If-Plots: Narrativity and the Law-Code
39
fore needs to be rethought at the supposed pole of the case, or the court trial, as well as that of the code's narrativity. This last extension (Sternberg [1998]) focuses on the Hebrew slave law, in all its variants, and interested readers may want to consult the detailed analysis of its manifold narrative art. Here, with the shift to a theoretical focus, I will occasionally draw on this exemplary case in generalizing the principle to encompass the codified law-tale as such, across its numberless contingent manifestations from antiquity onward. Also, without detriment to generality, I will often cite other Biblical instances, for much the same reasons: from the presentational to the substantive to the comparative. These reasons include the sheer economy and lucidity of illustrating from relatively brief statutes; their fame and accessibility and historic importance, as with the Ten Commandments; and their linguistic multiformity even amid legislative (directive, communicative, interpretive) unity. Further, the very differences, genuine or apparent, of the Bible's law-coding from the modern rule (nowadays taken for granted or canonical or advanced) prove it in both senses. They test the latter-day norms, or forms, and help to establish a higher common ground: to draw a line between fundamentals and historical or ideo-cultural variables, especially in narrative terms. This line is all the more important to draw as an antidote to its frequent blurring, if not erasure, in various quarters for various reasons. Note 1 above has already indicated one such typical ground, and, it so happens, apropos narrativist theorizing: the reference to some particular contemporary legal system (e.g. the American one). Amsterdam and Bruner (2000), or Bruner (2002), also typify the appeal to a desired value-scheme. Others carry it to the point of zeal for or against, critical, revisionist, justificatory, but always with a culture-specific target. Legal storytelling itself has often been advocated (or attacked) in America as a weapon for social reform. In mainstream jurisprudence, a major writer like Ronald Dworkin regards a "useful" theory of law as one geared to "a particular stage of a historically developing practice" and "addressed to a particular legal culture"—the theorist's own—with a view to providing "the best moral justification" for it. 18 No wonder particularists tend to dismiss, or just ignore, premodern ("primitive") legal systems, on top of everything else outside their coordi-
18
Dworkin (1986: 90, 102).
40
Meir Sternberg 19
nates. Below we will therefore see prescriptivism at work, restricting what counts as law (or due process) not only by focal delimitation but even by fiat—linguistic, anachronistic, naturalistic, moral, penal, ideological, psychological, and so forth. Together or apart, these biases contuse surface forms that come and go with the deep forces that underlie their assorted manifestations. 20 Hence the need for a strong antidote. Among other thought-provoking divergences, Biblical law is part of a larger canon, which frames it within an ongoing story of a people's history. This challenges our current ideas of legal context, intertext, precedent, provenance; it also brings out the multiform generic continuity of the two (law- and history-)tellings amid their different subgenres. 21 Again, unlike the usual impersonality of modern law, the Bible alternates "thou/you" with "he" references; and the second-person referents alternate or hover in turn between Everyman and 22
judiciary. This uncommon choice highlights the problematics of addresseeship and communication at the heart of the law. At the same time, more generally, the deictic features alleged to except the statute from the legal circle of telling thereby compound, via person, with another longignored narrative form, the you-tale, which has only recently come into 19
22
Inversely with H. L. A. Hart, who calls his approach, as I would mine, "general and descriptive" (1994 [1961]: 239). Within narrative study itself, an equivalent would be novel-centrism, at times further reduced to novelistic (postmodernism. Such privileging, or obliviousness, likewise excludes millennia of world-wide generic practice that make a difference to the idea of narrativity and narrative scope. At the opposite, nonliterary extreme, as demonstrated in Sternberg (2003a) and (2003b), cognitivist story analysis in psychology and AI favors instead textoids and otherwise simplified tales, with equally predictable results. Since antiquity, the Bible's law has been divorced from the narrative on various grounds; and even modern attempts at interrelating them still tend to assume their generic difference. For some discussion and references see Sternberg (1998: esp. 522-25 and the notes on pp. 664-67). Inter alia, the well-known Cover (1983) exemplifies there the problematic ideas of narrative opposed or related to the law. But this entire canonspecific tradition invites comparison to the generalized story/statute dualism with which I started. The practice belies, several times over, latter-day sweeping generalizations to the contrary. E.g. "the sovereign power addresses its audience, its subjects" in the third person: "legislative language uses the category of person in a characteristic way, completely eschewing first and second person forms," with a view to impersonality plus distance (Maley [1987: 33, 40]). Yet the Commandments begin with "I am the Lord thy God, who brought thee out of the land of Egypt" (Exodus 20: 2), followed by a string of direct imperatives.
If-Plots: Narrativity and the Law-Code
41
23
theoretical notice. Similarly with attributes like God-givenness, ostensible accessibility, coexistence of judicial with sociomoral or otherworldly retribution, and so forth. Throughout, what is exceptional by modern criteria defamiliarizes automatisms, or provincialisms, or agendas, brought to the law and/as narrative. 2. Statute as Story I
2.1. The Law's Represented Action: Master tale, Built-in Gaps, Lines of Closure What, then, is a statutory rule a rule of? It governs action, as against immobile existence or a quasi-entity like numbers, and so must represent what it governs. The lawyer's argument entails narrating what happened; the legislator's rule, what shall or shall not happen. The one is familiarly retrospective, the other markedly, yet as actionally, prospective. A statute accordingly tells a story in the legal key. For short, the Rabbis even named it by its action, as a Do [aseh\ or Don't [al ta'aseh]. No less suggestive is the current term for law-making, "enactment," with the Latin for "thing done" behind it. The thing done there is actually twofold. It stretches from the legislative act to the legislated activity, from regulating to rendering a world—in brief, from the discourse to the object made by the legislator. The statute is indeed "enacted," even on the objective level—our initial concern—because habitually storied: as a representation of an event, an agency at work, a patient who suffers, a change of fortune, against or under the law, all variously combinable. All are also open to the world's numberless representables. The statutory story may narrate any kind of doing, from treason or homicide to parking, from criminal to legislative utterance, and any kind of counterdoing, from execution downward. But its focus generally remains well-defined; its course (e.g. "If'-to"then") determinate, or determinable, unraveling included; its referents (happenings, actors, arenas) typological; and its given extent as short as its prospective extension in time is long and its jurisdiction wide. Far from homogeneous, this complex of features already spells inner discord, but hardly to the detriment of the law-tale's narrativity. Take the inverse ratios between temporal and textual length, binding force and 23
See the collection of articles in Fludemik (1994).
42
Meir Sternberg
open-ended scope. A statute is not "timeless"—as is a law of mathematics, its alleged parallel in Vaihinger—but rather omnitemporal: it represents a development that may arise an unlimited number of times in the future, binding upon all these instances from the moment the legislation takes effect onward. "A perpetual statute throughout your generations," Numbers 18: 23 calls it; and this is also what a jurist like Hart means b^ "standing orders to be followed time after time by classes of persons." Whenever a wife is newly taken, a beast killed, a Hebrew enslaved, a person to be indicted, such and such shall happen, according to our examples below. The difference from "timelessness" (or, for that matter, the "untensed") is crucial. For the timeless exists beyond time altogether, hence outside all agency and counteragency, contingency and command. On the other hand, the omnitemporal is, in a nutshell, "time-bound, but temporally unre25 stricted." The law is time-bound, since it represents some event(s) happening in the world, and therefore subject to contingency, mobilizing agents and counteragents, and exhibiting the peculiarly temporal features of sequentiality, directionality, and duration. But the law is also timeunrestricted, in that (unless stipulated otherwise) its life-span extends indefinitely. It represents, not an occurrence located at a particular, unique spacetime juncture—as does standard narrative—but an occurrence-type given to numberless instantiations in time to come, and so representable in little: typified, the many activities projecting ahead shrink into minimum discourse. An action remains an action, though, however multiplied in the 27
world told about and miniatured in the tale's wording. Were the statute other than time-bound amid unrestrictedness, it could not bind on the fruits of time. In short, omnitemporality is here a condition for generality and normativity, rather than a bar to narrativity. Quite the contrary, if anything, where the generalized bound and binding form exercises boundless story-generating power. Such a mini-tale 24
Hart ( 1 9 9 4 [1961]: 23).
25
Lyons (1977: 680). More on this span, in correlation with other temporal domains, see Pfersmann (1994). Elsewhere, happenings so departicularized out o f "specificity" and "concreteness" would indeed count as summary, and if antedating the rest o f the event sequence, as exposition rather than action proper (Sternberg [1978: 14ff.]). But then, they count as such there only by way of opposition to their particularized neighbors or, if expositional, sequents. In the absence of such differential value, as here, the departicularized events assume the role of action themselves—discourse, like nature, abhors a vacuum—aided and abetted by special narrativizing forces.
If-Plots: Narrativity and the Law-Code
43
counts as a master tale in every sense, at once requiring and problematizing correlation with the host of its offspring: whether the more particular scenarios imagined by the legal theorist or the unique actualities refracted and recounted by the rival storytellers in law enforcement. 28 All of them variants of the master tale, indeed born of it, they yet remain variant from it, let alone from one another, and some must count as bastards. The problematics mainly concerns the given discourse's limits of representation or, in interpretive parlance, of specification and extendibility. Who's who in the statutory discourse, what's what, when and where and why? Among the candidates for the role, possible or actual, which fit the slot of traitor, perjury, breach of contract, malice aforethought? If one should love one's "neighbor" as oneself (according to Leviticus 19: 18), who is one's neighbor, who isn't? In "If thou buyest a Hebrew slave, six years he shall serve ...," do all modes of acquisition obtain and is the term for the acquired object genderic or generic, exclusive or inclusive of women? Generally, how to map such a bare-bones yet binding reality-image as the lawscenario onto the fullness of reality, even in principle, or onto the hard yet debatable singularity of any real case? How to match the type of (wrong)doing at issue with its genuine tokens, which variously flesh it out to the same operative effect, and what becomes of those deemed untypical? (Compare Jesus's estoteric parables.) Inference alone can determine the reference of the umbrella words to the world, the master tale's applicability to its imaginable modifications in law and contested manifestations in life. From the narrative standpoint, however, the minus of uncertainty—like all gaps—reverses into a plus. Among the examples above, let me illustrate, very briefly, from one that I 29
have elsewhere detailed as a case study. "If thou buyest a Hebrew slave ..." leaves the mode of "buying" (acquiring) unspecified. Does this suggest an open-ended act of enslaving a fellow citizen—any which way—a gap amenable to all possible closures in theory and reality? Or, quite the opposite, a limitation to some particular mode assumed by the legislator (e.g. debt-slavery or self-sale), as has long been thought, and if so, which? The very doubt, embodied in the ongoing scholarly dispute, renders the elusive wording perceptible—a kind of Jakobsonian poetic function—and the elliptic telling multigenerative. Yet any closure by an interpreter en28
Unless otherwise noted, this phrase includes courtroom application by the different parties, from attorneys to judge, as well as the work of the executive branch. Sternberg (1998), especially the last chapter.
44
Meir Sternberg
riches the given bare-bones tale as well. Authoritatively so (as nowhere else in text land) if it comes from a judge; and doubly so, if it appeals to earlier or other narratives for intertextual reference, filling, determinacy, support. On such an interreading, the slave law thus rules out acquisition by force: all imaginable pre-scenarios of the co-national ("Hebrew") violently pressed into servitude, let alone perpetual servitude, the way Egypt treated the entire Israelite nation before the Exodus. Even a negative answer like this specifies the "buying" act—along with the parties involved, the background to that transaction, the eventual seventh-year release, the underlying ideology—while weaving together past, present, and future into a larger, echoing narrative continuity. Again, does "slave" co-refer to the woman? Is its masculine gender all-embracing or exclusive? Uncertainty, and controversy, recur to much the same effect, as does their resolution, especially by appeal to extralegal narrative grounds. My argument 30
for a cross-gender reading shows it encoded, in turn, in retrospects on actual variants of the law-scenario. All of them aids to decoding, these analogous life stories are common knowledge to the original parties to the law's enunciation and, for us, recorded in canonical telling, especially the Genesis/Exodus affair with the Egyptians. The given statutory discourse implies historic precedents of bondage and bondspeople from the audience's experience or annals—within living or collective memory—where women prominently figure as more than equal to their menfolk under servitude. Thus, decades before, with the people enslaved in Egypt, its newborn males killed, and its spirit broken, Moses (then a baby, now the law-speaker) survived thanks to his mother's and sister's outwitting the oppressor (Exodus 2). The law is, as it were, foretold, and to inclusive, even woman-highlighting effect, in a different, fuller, real-life narrative key. Another allusive interplot develops across the subgenres, for elucidation with remembrance and continuity. As a canon, a poetic and multigeneric one at that, the Bible affords the law enunciated, or its understanders, special but not exceptional resources. Compare the modern appeal to legal precedent, or the resolution of one statute via another: both are equally forms of what I call interplotting, only kept within the law-circle itself. More universal yet is particularizing the reference at issue without cross reference to any other text: by inference from style, usage, context, frequency, axiology, practical logic, the code's spirit, the sense of justice. 30
Ibid.: last chapter.
If-Plots: Narrativity and the Law-Code
45
Look at three disambiguations of a proverbial referring term. Who's one's neighbor is a question much debated since the Leviticus 19: 18 command to love him as oneself. The Rabbis identify him as a fellow Israelite, one of "the sons of thy people" mentioned earlier in the verse. But Jesus, responding to a lawyer's query on the point, defiantly opts instead for any Good Samaritan, no matter what his nationality (Luke 10: 25-37): a characterization ethical and freestanding, rather than ethnic and contextualized. Again, in a renowned judgment by Lord Atkins, the gapfilling principle changes from morality to legal pragmatism: "The rule that you are to love your neighbour becomes, in law, you must not injure your neighbour; and the lawyer's question, who is my neighbour? receives a restricted reply," namely, "persons who are so closely and directly affected by my act that I ought necessarily to have them in contemplation as being so affected." 31 Three readings, three role-castings, in polemical sequence. Each later closure, though allusive, would appear to quarrel with its precursor's ground and viewpoint as well, metatextually, rather than to adduce a precedent or parallel by way of supportive intertext. Yet "neighbor," hence the conduct toward him, does become particularized on each who's-who, and their quarrel gives the text density and salience, just as with the cross references above. Further, these two cardinal aids to elucidation work together, in harmony or otherwise. One may also characterize the "Hebrew" within the immediate (con)text—by, say, the "wife" with whom he arrives or cohabits (Exodus 21: 3-5)—and one may also identify the "neighbor" from his precedent or parallel occurrences exterior to the Leviticus verse. So the rule holds throughout. The built-in trouble for the legal interpreter and practitioner (on which more below) is the discourse's gain in thickness and interest: such mapping/matching entails further plot-making, beyond the givens, between the lines, the respective event-lines notably included.
2.2. Generic and Legislative Narrativity: Universals Specialized On the other hand, all these features, capacities, and resources are variously special or exceptional in attaching to legislative narrativity, to the statute as a story in language. Here precisely discoursive meets disciplinary specialty to argue for an interdisciplinary theory, which alone would cover what the above multivocal phrases encapsulate. 31
Quoted in Maley (1994: 4 5 ^ 6 ) ; also Hart (1994 [1961]: 264).
46
Meir Sternberg
In a way, the hermeneutic issues outlined thus far are hardly unfamiliar to the law student and practitioner—another measure of their centrality. They have occasioned a great deal of analysis and controversy, which the literary scholar may visit with profit. Across variations in approach or emphasis, though, the basic viewpoint there remains inadequate: too narrow and too wide at once, because focused on a single, albeit important, component—the verbal one—that statutory discourse shares with others. The medium then overshadows the matter, a fortiori the compounded peculiarity of their union. And, in the light of the key issues just broached, we're now in a better position to appreciate the unhappy difference made by such a hierarchy. To take up my introductory remarks, it all begins with the point of departure, namely: the statute's misgrouping, hence underspecification, as a piece of language, only encoded (or subcoded) in an uncommon idiom. Further, that idiom's uncommonness is taken to involve unstory-likeness. Recall how legal telling gets polarized with the law's "timeless" ruling; or how Gibbons (1999), not content to generalize "law is language," in effect proceeds to denarrativize it by equating the "codified" with the "static." (The same divide recurs, most often, by way of silent omission, as if law and law-telling were obviously poles apart.) Far from self-contained, or an end in itself, however, the codified language with all its distinctive attributes is geared to a narrativity that articulates, multiplies, and organizes those attributes into the wanted statutory operation. Words, world, and working compose functionally, their respective peculiarities yoked together in the service of a unique goal. This already marks off the legislative composite from all extralegislative varieties, namely: from the whole of nonnarrative discourse, even when likewise world-imaging, and, to a lesser extent, from the rest of narrative, as well as from nonlinguistic communication. Storied, the statute generically polarizes with nonnarrative (e.g. descriptive writing) in that it represents a mobile object. It tells in time about the dynamics of a world in time: a crime-to-punishment event-sequence, say, with agents who break the status quo and next turn patients at the hands of counteragents, along specified lines. As dynamic representation necessarily includes its antipole, the states of affairs changed by it—but not vice versa—the elements, spacetimes, patterns, opacities, inferences, experiences, complications all mount accordingly. The distance only widens visä-vis weakly representational writing, such as philosophical argument, never mind the nonrepresentational extreme. Even amid linguistic com-
If-Plots: Narrativity and the Law-Code
47
monality, narrative thus remains a law to itself, and the codified law-tale a subgenre within it, whose own markedness has already begun to emerge: the bare yet binding and thus productive yet limitable type-action, above all. Inversely, on the reductive "law-is-language" premise, all these generic hallmarks, oppositions, and subgroupings vanish from sight, together with the differences they make to everything. So they do, by another route, when the law's words are flattened and frozen into "timelessness," hence out of dynamism. The increasingly discriminate implications that narrativity, verbal narrativity, statutory narrativity carry for theorizing and reading—if they are at all drawn—will then at most enter here and there by the backstairs: unawares and unannounced as a rule, let alone integrated with the official conception. A major case in point is the problematics of reference just discussed. In a highly influential statement, among jurists, H. L. A. Hart assimilates the issue to the nature of language, particularly when used for rule-making. How to bring specific instances under the general terms (statutory or otherwise) before us? In every rule, we distinguish "clear central cases, where it certainly applies and others where there are reasons for both asserting and denying that it applies. Nothing can eliminate this duality of a core of certainty and a penumbra of doubt. [...] This imparts to all rules a fringe of vagueness or 'open texture'." What marks "a general feature of human language" (inevitably rich in class terms, such as "neighbor" or "slave" above) is exploited by legislation to regulate beforehand an unknowable spectrum of concrete possibilities that might arise. 32 Hart's thesis has since been debated in certain re33
spects; yet concepts like generality, vagueness, openness—with their opposites—widely recur in jurisprudence and have predictably gained still firmer verbal anchorage from linguistic approaches to the law. 3 Small wonder that resolving the "vague" term is then said to involve corresponding verbal factors: "the limits of synonymy, the proper scope of hyponymy or inclusiveness of meaning; of change of meaning; of contextuality; of intertextuality." 35 True, no doubt, yet sadly underspecific, be32
34
35
Hart (1994 [1961]: 119ff.). For some objections and responses, see the Postscript to Hart (1994), which has itself kindled further dispute. E.g. Endicott (2000); Maley (1987: 38-46), (1994: 2 7 - 3 1 ) ; Bhatia (1994: 137fF.) on syntax as well as lexis, all with earlier references. In volume and diversity, though, the work bearing on the topic is second only to that on literary reading. Maley (1994: 31).
48
Meir Sternberg
cause these factors enter into all linguistic communication. Instead of mere vaguenesses, I would claim, what we encounter in the statute are gaps in a represented world's time. They are left yawning, due not just to the medium's "open texture," but to the discontinuities between the telling and the told sequence: between the narrative we receive and the narratives we may in theory reconstruct from it about an open future (e.g. "If thou buyest a Hebrewess ...") or those we, as trial lawyers, would reconstruct on that narrative's pattern after the event that it debatably envisages. If anything, it is the exigency of keeping the law's master tale open to new instantiations in real life that determines, or explains, the open texture of its telling; just as the need to argue or judge in court its applicability to some newly told instantiations—to distinguish the genuine token from the misfit, the operative from the unintended statutory variant—presses for closing tale and texture together, regarding the gap at issue at least. This woman among others, this child as such, this man "bought" this way, qualifies (or not) for "slave": so will the lawyerly role-casting then go. A matter of picking and placing characters in action, role-casting already shows that the gaps involved—found, debated, closed—are mapped onto the discourse but still irreducible to its open verbal texture. Certainly not if openness equates with "vague" synonymy, hyponymy, change of meaning, and the like. The irreducibility becomes more evident when we pass from object/state-designating nouns to verbs with their built-in dynamism: from single referents to world-changing, interactional, concatenated (e.g. "If ... then") reference. Or when we turn from pieces of world manifest in the statute's words, however ambiguously, to those altogether missing—like why the Hebrew gets enslaved, or the unspoken antecedents and consequents to be discussed next. Or from gaps in the text's enacted world to gap-filling via cross reference. Intertextuality is one thing, available to all language use and reducible at will to verbal play, stylistic incongruity, and allusive snob appeal. Interplotting is quite another thing, narrative-specific in dovetailing the intertext's event-lines: the action of the given precept with that of the gap-resolving precedent, for example. Throughout, storied indeterminacy entails a different, richer, because multilevel, problematics and productiveness. But nor do the words themselves lose their differential force in the thickened world-making. Oddly enough, with this productiveness there goes the statute's extraordinary discourse-sensitivity, as unusual in and beyond story land, poetry-like. Legislative codifying is inseparable from its linguistic encoding. It resists translation not only into another medium,
If-Plots: Narrativity and the Law-Code
49
intersemiotically, but also into a synonymous-looking message, intralinguistically, by way of paraphrase, rewording, etc. (Except in legislative amendment, of course, wherefrom emerges some related untranslatable text.) A far cry from the ordinary narrative kernel (e.g. the Oedipus tale) or plot-type (e.g. the detective story's), which is itself medium-free and accordingly manifestable at will in artistic development. Again, a finished novel or play, though surface-bound, may yet transform into a recognizable cinematic version, equivalent in overall narrative dynamics and experience. The law contrasts even with its own tokens within the legal field proper. The storytelling done in rivalry at the trial both freely incorporates or alternates nonverbal semiotic codes (gesture, intonation, audiovisual aids) and enjoys verbal license. On the other hand, the verdict pushes verbal codification to the limit of formulaicness, Guilty/Not Guilty, thereby excluding all generativity. The comparison with these nearest relatives brings out how every difference in the given statutory language essentially makes a difference to the life projected, or projectible, which any divergence from the text would mar. Productivity and sensitivity of this order are each rare by themselves, rarer even as a discourse kind's criterial features, and rarest of all outside verbal art or its lyrical branch. How much more so their constant twinning and interplay, for theirs is a meeting of extremes. A narrative bound in word yet without sharp boundaries in the world? Codified, in effect closed, because fixed, yet all too open to inference? Master and sovereign yet dependent on how underlings take it? This twinship otherwise sounds like an oxymoron. The statute book is accordingly the culture's largest anthology of mini-narratives, whose smallness and fixity is in inverse ratio to their thematic diversity, their enduring impact on the world, and their power to generate further narrative by way of readerly gap-filling or lawyerly application. 2.3. F r o m Narrativity to Narrativity Plus The story told may rise yet farther above the threshold of narrativity, higher than guaranteed by the distinctive built-in pressure for inferring the limits of its reference. We will find it, for example, emplotted into a strong (e.g. crime-and-punishment, tit-for-tat) chain; bearing a conflictual relation to other law-scenarios; given a subjective dimension (e.g. "malice aforethought") via insideviews of the agents; invested with high tellability, in and out of legaliterary context; retrojectible into past history (e.g.
50
Meir Sternberg
the event that inspired and now interprets its enactment, the way the people's bondage in Egypt did and does the Hebrew slave law) as well as projecting forward; or, via space/time coordination, branching into multiple paths cum parallels, reminiscent of modernism's "spatial form" and Borges's plural reality. (Not to mention yet the kind's second constant, whereby law-speaking within the world joins forces with law-telling about the world.) All these assorted energies of "narrativity plus," however, converge on, or radiate from, the dynamic kernel of eventhood. As I put it elsewhere, "the law enunciated projects a course of action to be followed, in the narrative along with the normative sense. It cannot divorce itself without surrendering its raison d'etre In turn, a dynamics of the told entails a dynamics of telling/reading. Indeed, as I've always defined it, narrativity lives between these twin temporalities, whose interplay launches and maintains the processes that are as universal to the genre as they are unmatched elsewhere. They consist in suspense, curiosity, and surprise: the mental (inter)dynamics of prospection, retrospection, and recognition. These processes not only attend but experientially attest and distinguish a represented action—the genre's hallmark—which has eluded the objectivist approaches to narrative current since Aristotle. And elusive it must remain there, because "objective" defining terms—geared to the narrated world or sequence alone—are not only endlessly and fruitlessly disputable in theory, but also unworkable in practice. Thus, what they would isolate on some formal basis as an action, event, or change of state, hence as a marker of narrative, often makes a description in context, and vice versa. A hopeless business, this quest for objectively (and, in Structuralism or recent cognitivism, scientistically) formalized generic differentials. Instead, the key lies in whether the reading subject processes the given object into actionality, rather than into any alternative, static design: as things in general are known by their fruit, so is narrative known by its effects on our mind. A represented action is what it peculiarly does to us, under any formal guise, in the reading; and what it does must involve the three generic universale, whose peculiar narrative energy assimilates and mobilizes any component (e.g. words, themes, spatiality, viewpoint)
36
Sternberg (1998: 530-31). Given this kernel, the analysis there proceeds to trace the lines of narrativity-plus developed by the exemplary law, on a scale impossible to replicate here.
If-Plots: Narrativity and the Law-Code
51
shared with the rest of discourse. They all follow from our progressive encounter with a time sequence that narrates another time sequence, in or out of event order. Either way, gaps en route are inevitable. Caught on the move between past and future, between ignorance and knowledge of the events told, how can the reader help wondering about the opaque developments ahead, or wanting to settle mysteries left behind, or bumping against unexpected disclosures? 37 So, if the ultimate proof of narrativity lies in these effects and operations of reading between the times, the law enactment certainly qualifies. Just observe how, throughout the penal code, suspense (i.e. forked expectation about the outcome) necessarily attaches to the movement from crime to punishment. By the same rationale, this expectancy governs all analogous early-to-late developments: " I f ' to "then," premise to proposition, buying a slave to letting him/her go free, though tempted to act otherwise. The mind, tensed up by the "If ...," gains release from the "then ..." Normally ordered by chrono-logic, the law privileges a futuredirected, hence suspenseful, teleo-logic of communication. But this does not exclude the opposite, "curiosity"-governed dynamics, born of an opaque past. We have seen how the omnipresent uncertainties of reference leave gaps about the crime (or the enslavement) itself and its antecedents, pulling our mind backward—in quest of closure—even as we go forward. In turn, busy retrospection may adjoin or change to the surprise of belated recognition: an unknown precedent comes to light, an unexpected yet a posteriori defensible referent springs to the law-interpreter's mind. Across the three dynamics, further, the gaps encountered and the inferences made ("modalized") to close them are not simply referential but also temporal—the elusive objects forming part of a mobile world in mobile disclosure—nor are they simply encoded in the language but also in a plot logic typical of the codified tale. Across the three dynamics, again, the distinctiveness of this chrono-logic extends from the plot's built-in gapping throughout and multiple hypothetical closures to its (and their) amenability to definitive resolution by constituted authority, legislative or judicial. No gaps are doomed to permanence in the statute, alone of all stories and representations. Finally, those specialized universal dyna37
For detailed arguments and case studies, see especially my (1978), (1985), (1990a), (1990b), (1990c), (1992a), (1998), (2001b), (2003a), (2003b), with earlier references there. For extensions to other arts and to interart traffic, see for example, Bordwell (1985), (2002), (2004) on film and Yacobi (1995), (2000) and Kafalenos (2005) on ekphrasis.
52
Meir Sternberg
misms interact with the ample energizing resources of narrativity plus— whether constant or optional, whether having analogues elsewhere in the genre, like some devices already outlined, or peculiar indeed and yet to emerge. Among the latter would be the anomaly of retroactive legislation, as outrageous in temporal as in social terms, and the incomparably absolute world-making force of enactment. Nothing about the law's discourse, in brief, remains either static or undifferentiated from other verbal (a fortiori extraverbal) messages. 3. F o r m s of the Law-tale and Narrative If-Plotting A principled continuity thus reveals itself here with the defining feature of the genre at large: with narrative in its action-based narrativity, complete with the effects generated by its emplotment as such along the sequence we meet and process and experience. This narrativity persists even in the tiniest of the statute book's mini-stories, or the mini-storied statutes, and regardless of the linguistic form in which the drafter casts them. On scrutiny, the least promising instances by these criteria not only pass generic muster to reinforce as well as illustrate the argument above. They also luminously encapsulate a double unity in variety: the storied genre's unity across the statute variation and the statute kind's unity across its own assortments as a subgenre. Here are the law-tale's three commonest forms in a nutshell: ( l a ) Remember the day o f the Sabbath to keep it holy. (Exodus 20: 8) ( l b ) In future no official shall place a man on trial upon his own unsupported statement, without producing credible witnesses to the truth o f it. (Magna Carta [1215]) (2a) If a man has newly taken a wife, he shall not go out with the army and shall not be charged with any business; he shall be free at home one year and gladden his wife whom he has taken. (Deuteronomy 24: 5 - 6 ) (2b) If a seignior wishes to divorce his wife who did not bear him children, he shall give her money to the full amount of her marriage-price and he shall also make good to her the dowry [ . . . ] (The Code of Hammurabi, 138) (3a) He w h o kills a beast shall pay for it, a life for a life. (Leviticus 24: 18) (3b) Every person w h o through cowardice shall in time of action withdraw or keep back [ . . . ] shall suffer death. (Articles of War [1749])
These three linguistic schemata and their relations will be further explored as we proceed. For now, let us first briefly compare (1) with (2) under the names (though not the analysis) they've received in Biblical
If-Plots: Narrativity and the Law-Code
53
studies: apodictic vs. casuistic law. 38 The former issues an absolute, or absolute-looking, directive about a certain behavior pattern, like the observance of the Sabbath, which implies in turn a host of subdirectives, of course. Apart from the implicit subdoings, the observance projected by this Commandment (la) entails a larger double change in the world: vis-avis both the pre-law state of free nonobservance and the pre-Sabbatical course of the workaday week, equally free of this mandatory "holiness" even for the observant. From what it tells to the telling/reading, the apodictic schema is always packed with more dynamism than appears, and further reinforcements will emerge before long. Charged with similar energy, it may force the proceedings of the law enforcer himself, as does the meta-law in (lb). More generally, it prescribes ("Remember ... to keep holy") or prohibits (e.g. "No official shall place ...") or permits some future line of conduct, without any visible qualifications, such as If/When/Where ... Of all the forms, therefore, the apodictic scenario is not only the most unqualified but also the most reducible—down to a single-clause terse imperative, as in the (originally two-word) Commandment "Thou shalt not kill." Even so, the minutest of the law's mini-tales must represent a course of action (and in [la], directly to the envisaged agent, Everyman, within Israel at least) in regulating it. The name of Do/Don't still, or perhaps best, fits the instances that approach its laconic imperative. In the "casuistic" style of the next example, both of these apodictic features vanish together. The directive turns openly conditional and its syntax complex to match. It runs from subordinate ("If a man has newly taken a wife ...") to main clause ("he shall not go out ..."): specifically, from protasis to apodosis; or, logically, from antecedent to consequent, with the chain mapped seriatim onto narrative time, dynamism, reasoning, chronologic, in short. Indeed, this joint difference highlights the logic of action that the apodictic style buries underground. The "if ... then" advance both draws together two action-units (here, wife-taking and home-making, exempt from public duty) and formalizes in their consequentiality the Suppose Principle that underlies and modalizes all narrative, most evidently the fictional kind. For comparable highlightings, think of Balzac's reference to his dense openings as "premises to a proposition," or of Alice's favorite games of make-believe ("Nurse! Do let's pretend that I'm a hungry •JO
Alt (1989 [1934]) and Paul (1970) are classic references.
54
Meir Sternberg
hyaena, and you're a bone"), or of Scholem Aleichem's titular "If I were Rothschild" and Italo Calvino's If on a winter's night a traveler..., or of "The Three Laws of Robotics" in Isaac Asimov's science-fiction. Each lays bare the genre's omnipresent, endlessly manipulable chrono-logical movement from exposition as antecedent pro tem—entertained regardless of belief—to the particular dynamics narrated as a chain of consequents that follows therefrom. Where the chronology gets deformed in the telling (exposition elided, delayed, distributed), the effect on the reader is the more disorienting to the extent that it overtakes the action logic: with the premises absent, or temporarily gapped, we can't make reasonable sense of what happens, far less an unambiguous plot sequence. This is why expositional deployment is the paradigm case of the three processual universale—prospection, retrospection, recognition—all triggered by such ambiguities. 39 Nor does the Suppose Principle except historical emplotment—and its mates in everyday storytelling—despite its claim to truth as well as its adherence to the chronology of events. By the same token, this fact-bound emplotment must involve narrative's universale of processing amid uncertainty: the hypotheses about the real narrated action that we form, adjust, shift on the move throughout inevitably appeal and in turn themselves amount to Suppose's. So do, besides, the modern historian's own empirical, probabilistic reconstructions of what happened—and, especially, why—under the constraints of human knowledge. In the stricter parlance I already used and will take up later, there's no escape from the modality built into the genre, nor any limit to the variable or extra shapes it can assume. The exercises in historical imagination, which start with a counterfactual postulate ("If William had lost the battle of Hastings ..."), read like the make-believe game that Scholem Aleichem founds on his imaginary premise above: they only estrange the normal practice of history telling to reveal the principle at work throughout factography as well as fantasy. The ontic opposition between the two kinds narrows down to a difference in whether the antecedents, however told, invite acceptance as true per se—and the whole chain with them—or just entertainment pro tem. (The difference also correlates with the respective overall acts of telling, because fiction's telling is itself fictional and accordingly has antecedents of its own for us to entertain: "Imagine such and such a narrator ...") So, all narrative involves if-plotting of some kind, with the condition39
Sternberg (1978) and, on film, Bordwell (1985).
If-Plots: Narrativity and the Law-Code
55
ality generally buried out of sight, though never out of the processor's mind: we always proceed, reflectively or unthinkingly, upon some ontic basis, right or wrong or equivocal. But what surfaces here and there all over the genre is visibly the rule of the legislative subgenre's casuistic tale. The family likeness even strengthens, among other ways, 40 because the "if ... then" action logic proves to underlie not only the rest of the narrative genre but of this very subgenre, too. In apodictic discourse, clean against formal appearances and scholarly opinion, it lurks below either end of the given statute, as a kind of implicit hypothetical-consequential frame. The joint told/telling dynamics, in which narrativity consists, here outreaches the generic minimum on yet another front toward narrativity plus, and along (or between) lines integral to the subgenre. For one thing, no matter how absolute-looking on the surface, the apodictic directive generally carries with it unspoken presumptions and provisos, stipulations and exceptions, latitudes and loopholes, or else the law would become too general for pragmatic discriminateness. It would lose touch with life's numberless realities, exigencies, contingencies. Of all its potential tokens, the type-story must accommodate some while excluding others as nonmatching with its assumed, differential groundwork. A fit involves accord with the assumed typal context, no less than with the given words that it silently encloses and determines. The question is only which of the attachable strings attach to the Do/Don't wording, how a particular event-sequence fares on contextual scrutiny; and the answer may of course grow disputable. It never remains so in principle, because always decidable within the system (by legislative amendment or the judge at the trial). The universal gapping falls here under a unique metarule of closure. But the closure would only settle which of the implicit expositional gap-fillings applies. So, either way, these tacit understandings operate as a web of antecedents that ground and limit the action overtly imposed, in effect turning it into a consequent. "[If/Unless such and such is the case, then] thou shalt..." 40
E.g. the special case o f so-called "legal fictions," whereby an agent or entity o f one type counts for some purpose as another: a woman, say, as a man (Fuller [1967]; cf. Vaihinger [1924: 3 3 - 3 5 ] , already discussed above). Only, those fictions (unlike art's) target and change the real world, even against nature, by their very legal enactment, a fortiori their enforcement. On which more below, when w e come to the power that lawdiscoursing as a verbal act exerts upon the reality discoursed about and the action within it.
56
Meir Sternberg
For comparison, recall how the casuistic "If thou buyest a Hebrew slave," though formally leaving the mode of buying wide open—and in Scripture's lexicon, "buy" ranges as widely as "get"—still rules out acquisition by force. Or how this silent "exclusive of ..." finds an inclusive counterpart in the woman's subsumption under "slave" (or an exact one in the interpretive attempts made to keep her out by taking the noun's gender literally). In either instance, on any reading, the unspoken antecedent will thus have been determined—as it must somehow be—to establish the law's narrative horizons, productivity, and of course applicability: its socalled intention. Far from a "verbal" issue, this, given the bearings of either word on the world: on the narrated frame of reference, on what can or cannot happen in it, hence on what shall happen to actual people in actual life. Resolution means enabling or disabling all the law's possible storied tokens that (e.g.) begin with a woman cast in the role of object ("slave"), actional as well as grammatical, to "thou buyest." Similarly with apodictic law, except that what requires closure here is not some indeterminacy within the given " i f ' -premise but the entire " i f ' premised in silence. And again, this missing groundwork is not a question of elliptic wording or juristic reasoning alone, because antecedence necessarily assumes here its fullest sense, temporality included. As this hole in the action logic unsettles the chrono-logic, it launches a retrospective dynamics of inference for wholesale repair. From such a narrative viewpoint, the bare apodictic directive then jumps in medias res\ nor does it provide the gapped exposition thereafter at that, behind time, leaving closure to the reader instead. We must provide the antecedents ourselves, start afresh, as it were. Various contextual premises (e.g. "in normal circumstances," "other things being equal," "subject to higher priorities") are so rudimentary and extensive as to go without saying, even to escape cognition. Out of sight, out of reflective mind, along with the text's operative worldview—except to the outsider who, uninformed, finds the enactment strange, unmotivated, oversweeping, and the like, or except to the insider himself when under pressure for close reading in trouble. The former, where he does not just misread or underread, will then see groundlessness and indiscriminateness in his ignorance of the ground rules; the latter, confronted with a real specific difficulty, will instead try to figure out the missing groundwork by or within those rules. What is the juristic doxa presumed, applicable, extendible here? Such figuring out of the antecedent in retrospect, like all gap-filling,
If-Plots: Narrativity and the Law-Code
57
means enriching the narrative, often with striking or unexpected plot entanglements by way of closure. Or, significantly and distinctively, interplot entanglements. For, whatever your approach to the rest of the genre or to discourse in general— regarding the "autonomy" of the text—no law is an island cut off from the system. In the Bible's canon as a whole, we have already found examples of intertextual, interactional working between problematic scenarios and precedent (hi)story, with a strong analogy to the currently better-known inference or argument from legal precedent proper. But much the same interpretive drive brings various scenarios into contact for resolution within the law-code itself. Think of the need, or will, to harmonize discordance between "absolute" statutes. When a matter of will, their discord is incidental, superficial, or even contrived by the twinning of unrelated absolute plots; but the harmonizing method remains the same. To keep them both operative, the apodictic D o ' s at odds will be so twinned that their contradiction will read instead as casuistic subordination: as though one of them repaired the other's jump in medias res. For the one taken as overriding then necessarily conditions the other, antecedent-like, to generate a two-link interplot. On the Rabbis' interplotting between Commandments, for example, you must honor your parents (Exodus 20: 12), but not if they would have you desecrate the Sabbath (ibid.: 20: 8). Why the linkage, a fortiori why such linkage, between autonomous Do's, not even adjacent?, you may wonder. The trigger for this interplot is the later re-juxtaposition of the two fiats, now in immediate continuity: "Every one of you shall fear his father and his mother, and you shall keep my Sabbaths" (Leviticus 19: 3). The juxtaposition reads as a latent clash, the "and" as "provided that," hence as both an ideo-logical priority—Sabbath first—and a resolution in the case that the dilemma of which Do to observe materializes. The inferred agonistic linkage and priority between the Do's then becomes retrojectible into their original, discontinuous co-occurrence among the Exodus Commandments. There, too, short of this implicit negative ("not if ...," "unless ..."), conflict-solving antecedent—or the positive way round, given a law-abiding home of Sabbath keepers—the filial behavior commanded allegedly ensues. 41 41
In principle, the two may also interplot otherwise, via a different trigger and rationale but to the same effect, within the framework of the Commandments themselves. By this alternative inference, the semantic link between "honor" and "keep holy," across the intervening distance, evokes the possibility of opposed allegiances: in which case, or in-
58
Meir Sternberg
Likewise, the next Commandment's forbidden "killing" gets delimited and the term itself defined in silent retrospect vis-a-vis all capital offenses that follow, so as to except the death penalty: no life-taking action (the reader infers for harmony) outside judicial counteraction. Elsewhere, more straightforwardly, clash between laws reverses into the coexistence of an apodictic with a casuistic form of the same law. The underlying premise's source of legibility thus ranges from umbrella-context to interplotting to multiform plotting, and its targets from word to world to the workings of the legal system. Moreover, the naked absolute directive that we encounter finishes as it began, in medias legis. Its appearance of self-contained integrity again deceives the eye, perhaps, but never the integrating mind faced with what amounts to a narrative torso. Indeed, the torso-likeness would count against it a priori on certain approaches. From an opposed viewpoint and discipline, unmindful of narrative dynamics, some legal theorists and lawas-language analysts in effect rule out the apodictic style altogether: not just from storyhood (the usual common fate of all the law-maker's styles) but from the statutory circle itself. Where and why would it be found deficient? First, vis-ä-vis an essentialist conception of the "legislative sentence" prevalent since the midnineteenth century. 43 "Legislative statements typically begin with fairly terscenario, the resolution will go in favor o f the Do with the more heavily loaded term. The Sabbath wins out afresh. In larger principle yet, this rationale suggestively compares with how narrative implies an analogy between characters or plot lines in order both to generate and to dispel suspense about which o f the forks ahead (e.g. the outcomes o f a showdown) will materialize. For cases in point, see Sternberg (1978: chap. 3, esp. 68ff.) and (2001b). In turn, all this ranges beyond the theoretical concerns o f jurisprudence. Such discordant law-scenarios often also figure in the notoriously discordant law-stories told at the trial, where the very bids for harmonizing the code redouble the discord in pulling different ways. Either side will then manipulate the interlaw gaps as best suits his overall retrospect on what happened, vis-ä-vis the adversary's. Further, as if to belie once again the accepted opposition of the "static" code to the "dynamic" case, the one clash parallels the other. The analogy tightens when the former's voices get embodied, too. (Think o f how, as in the Talmud or our Parliaments, the law itself is debated among would-be legislators, face-to-face and to much the same agonistic effect associated with lawyerly conflict.) To crown it all, both analogues fall under the law-specific procedure o f adjudicating between rival (inter)tales via an authoritative (internale that possibly accords with neither: a decision mechanism exercised by the law-maker on one clash, by the jury on the other, and by the judiciary on both. 43
The term dates back to Coode (1973 [1843]).
If-Plots: Narrativity and the Law-Code
59
long initial case descriptions," in "the form of 'where,' ' i f or sometimes 'when'" clauses. 44 The alleged motivation behind such an opening ranges from the directive's applicability to its penal logic. As "very few legislative statements are of universal application," Bhatia goes on to argue, "it is absolutely crucial for the writer to specify the kind of case description(s) to which the rule applies." So much so that, unless qualified, "the provision will lose its essential nature." 45 The apodictic style, therefore, does not even get mentioned. Unintroduced the casuistic ("casedescriptive") way, it would remain too "universal" for enforcement. But other disqualifiers, especially legal philosophers, may go still further and deeper than the linguist in appealing to the "essential" nature of the thing. Thus Hans Kelsen explicitly contrasts the two forms and excludes the if-less one, not on the grounds of overextensive reference but on that of incompatibility with the retributive spirit of the law. Unconditional and sanctionless, the plain Do/Don't would go against the law's normative (i.e. casuistic) format. According to Kelsen's influential binarism: "One shall not steal; if somebody steals, he shall be punished [...]. If at all existent, the first norm is contained in the second norm, which is the only genuine norm [...]. Law is the primary norm which stipulates the sanction." 46 Inter alia, the Ten Commandments—definitely "existent," and one of them actually echoed by Kelsen—would not qualify for statutes, as they might elsewhere not qualify for stories on related formalistic/essentialist grounds, like unreality or actionlessness or deficient emplotment. Nor would a cardinal meta-rule like the Magna Carta's (lb). 4 7 Still less would their rhetoric of mental enforcement be taken into account. While the lawyer's persuasive arts have been foregrounded from the Rhetoric to the narrativist return to it, those of the law escape notice, as though the soverign's m i | h t not simply outranks and outpowers but discounts the subject's mind. 8 Here my narrative functionalism is therefore poles asunder from both jurisprudential and narratological formalism. Pursuing this test case will tell against either, with general implications for the principle at issue, across disciplinary lines.
44 45 46
48
Bhatia (1994: 1 4 4 ^ 5 ) . Ibid.: 146. Kelsen (1949: 61); cf. Hart (1994 [1961]: 39ff.). Actually, even modern law-codes, which otherwise privilege the casuistic norm— whence the very name penal code—often draft meta-rules in apodictic form. Sternberg (1998: 471-638).
60
Meir Sternberg
Not only are such conceptions of law, narrative, and law as narrative circular, hence arbitrarily exclusive (with the essentialist "nature" or "genuine" thrown in). Applied to our case, I'll argue, they also break down on their own grounds, since what they would exclude turns out to meet the formalist's own requirements another way: via the discourse as encountered and understood in the communicative process. Operational likeness persists amid ostensible variance, all within the family, so that the products come out much the same: if-plots on or below the verbal surface. That the examples in (1), or "Thou shalt not steal," end without a visible consequent (to form a so-called legislative sentence, or Kelsen's "If ..., then sanction," or an Aristotelian chain) is beyond dispute. Whether the absence is superficial or not, purposive or not, fillable or not, analogous to the casuistic sequence or not: here lie the questions. According to the general rule of narrativity formulated above—and just specifically applied to the missing if-exposition—the ultimate proof of this terminal gap is in the reading process invited. The matter hinges not on the show or deficiency of formal completeness, nor on some preconceived idea of "action," but on the narrative dynamics latent in the interplay of the developments told (regardless of how formed, enacted, pieced out) and their given deployment along the tale. As the beginning with the apodictic scenario generates curiosity about the antecedents—launches a process of retrospection in an attempt to deduce them for intelligibility, stability, and here also applicability—so the finish with it galvanizes suspense about the invisible but inevitable consequences, left to our prospective operations under the same exigencies. Both the initial and the terminal quest for a missing plot-link may also arouse surprise (the dynamics of belated recognition, with a new patterning to match) when the closure that emerges is unexpected: as with the Rabbinic interplotting of the Sabbath/filiality observances above, and often elsewhere. A nice operational symmetry thus reveals itself between the two in medias cutoffs, all along the line. Once the Do/Don't ends without any counterdoing for better or worse, toothless, as it were, the suspenseful "and then?" questions left open clamor for answers. What will follow from compliance, we wonder, and, above all, what from noncompliance? Nor is this suspense idle, whether pragmatically or processually, any more than was the curiosity (and/or surprise) before. Processually, it even grows stronger—more perceptible, more kinetic, more demanding, more transformative with regard to the givens themselves—than in a straightforward casuistic equivalent. There, as in (2), the "If ..." of course arouses
If-Plots: Narrativity and the Law-Code
61
expectancy about a "then," but the "then ..." duly follows to meet and stabilize it: the newlywed husband kept at home, for instance, and the seignior's divorcee given her rights. End of story, or so it appears from the overt, orderly emplotment. But it isn't the end in law, really, any more than in life, considering the discourse's failure to provide a sequel of and to, especially, law-breaking: noncompliant agency, whether the husband's (say) or the authorities'. What if either disobeys? Will the one wife then be left solitary, and the other get thrown out empty-handed, with impunity? Or will the Hebrew suffer enslavement for life without redress or reprisal? The more distant future thus yawns open, the suspenseful gaps about it no longer temporary, much less short-lived, but permanent—unless resolved by our own prospection under the law's guidelines (including meta-rule) of closure, with a view to some retributive finish. Thus, as the Hebrew's perpetual enslavement requires an official ceremony to ensure its voluntariness (Exodus 21: 6), the officials must presumably intervene where he's kept on against his will; and should they fail him in turn, God will pay back all the law-breakers involved, and does so with a vengeance (Jeremiah 34). Two contingency provisions for disobedience, one earthly, one heavenly, lurk between the lines here; but either of them shows why and how we move from the statute's immediate and given future (Do observed) to the remoter one (Do broken, then counterdone) that it has elided. Interplotting of some kind evidently comes into demand again to supply a further, now wordless, conditional in response to the breach: an understood " I f . . . then," "or else ...," "on pain of ...," and the like. The apodictic form, on the other hand, already opens with a link missing, well before its terminal double, so that the quest for closure escalates by every standard. The lesser the givens, the earlier and the greater the demands on us, against formalist preconception. What greets us here, as in (1), is a bare Do/Don't, altogether if-less and then-less as well as toothless in a yet longer view. It accordingly undercuts beforehand any analogous impression of completeness, so as to highlight and radicalize the law's Suppose game: now, the suspenseful prospective dynamics on the move toward the futures at issue. In the process, the very directive (e.g. to keep the Sabbath holy, not to put a man on trial in the absence of credible witnesses) changes form and action-logical status. Having wordlessly premised some antecedent(s), this directive itself now reads as an antecedent ("If ...") to some tacit consequent ("then
62
Meir Sternberg
..."): t o a sanction, whether positive or punitive, whether legal or moral or social or, in the Biblical and other two-world codes, divine. This mixed lot of inferables belong together here, since in them the bipartite logic of retribution finds its otherwise missing second term. They all operate to, literally, the same end: to supply closure and pressure at once, terminal enchainment with mental enforcement. Not that their retributive unity in variety would be readily admitted, either. Against it, the "if ... then only" formalism ä la Kelsen compounds with a yet wider restrictive tendency among modern jurists (especially so-called positivists like him) apropos the "then" itself: an exclusionary attitude toward anything other than the strictly and empirically legal sanction—a fortiori if the alternative reaction is implicit or gracious, never mind both. Low legalism, one might call it, by analogy to low realism. One also detects there yet another essentialism, reifying the law instead of observing how lawdiscourse produces a comparable effect on plot and people by different means. This fails to accord either with the normative or with the narrative power of the "extralegal" sanction: its hoped-for impact on the lawaudience and (indeed, due to) its finalizing emplotment of the law-tale operate regardless. A threat of odium or even, in the appropriate context, of heavenly wrath, is a threat; a promise of approval, or reward, a promise; and both, however encoded and inferred, are accordingly as consequential as their low-legal equivalents. 49 The difference between the two relates more to ideo-culture (or to the theorist's disciplinary horizons) than to action logic. 50 Again, across this ostensible divide, the ever-latent consequent sanction may follow even within the legal, including low-legal system, and in either court or code. Thus a judge deciding, after the event, how to punish the violation of a bare Do/Don't in the spirit of the law. (An act, this, not essentially different from the judicial habit of law-making in the shape of law-interpreting, and akin to the more ordinary exigency of fitting the 49 Similarly, and even more widely so, with other directives issued in life: they carry as a rule implicit tail-ends of various kinds to back them up, yet all understood and pressuring. Indeed, for us addressees to hear there the unspoken "If ... then" and "... or else," the D o itself needn't so much as appear in imperative, law-like form: w e then ourselves supply the directive along with the consequential force behind, say, the polite question. Imagine your friendly taxman asking, "Will you pay up by the end o f the week?" In Tom Jones (1749) (XII: 12), Fielding, a judge by profession, thus favorably contrasts the gypsies to "all other people" in that "they look on Shame as the most grievous Punishment in the world": to this difference—just shown there in judicial action—"their Happiness is entirely owing."
If-Plots: Narrativity and the Law-Code
63
stipulated punishment to the crime.) The same, down to the newly emergent consequent's equivocal status between retroactive and retrospective (merely explicative, as it were) gap-filling, will happen within the lawcode itself. The sanction may in turn break surface there, not least regarding the paradigm of the apodictic, the Ten Commandments. Thus, following the absolute directive to honor one's parents, the lawspeaker overtly and variously resumes it with a consequential vengeance: to threaten its breach with the death penalty (Exodus 21: 17, Leviticus 18: 8 - 9 , 20: 9, 11 Deuteronomy 21: 18-21; cf. Proverbs 20: 2). Likewise even with the immediate follow-up along the Commandment itself, only here in terms of retribution both positive and divine, for a change: "Honor thy father and thy mother, so that thy days may be long" (Exodus 20: 12). The resumptive scenarios that ensue in the Pentateuch therefore look like the original's mirror image, converting the filial into an unfilial agent and, with it, the (high-legal) promise of longevity into the (low-legal) threat of summary death. However, the original itself already suggests a doubleedged conditional, via the binary rationale of the opening (either you honor or you don't) followed by a purpose clause ("... so that ..."). Its gracious language actually reads: "If thou honorest ... thy days ..." or, by a reversal of ground and outcome, "If thou wishest that thy days ... then honor ..." Either way, a deducible cause-and-effect retributive linkage fortifies suasion along with narrativity, and this besides the initial tacit antecedent (e.g. "given a law-abiding, Sabbath-keeping home"). So it legibly does in turn when the grounding, hence the enchainment, goes underground to leave the surface apparently absolute, or conditionless, on both flanks. No premise before the apodictic directive, no promised or threatened response after. Still, the torso is part of a narrative and normative body, otherwise hidden from view yet asking for mental articulation as such. The naked command, exemplified by (1), then implies the whole threefold chain. Faced with the imperative of remembering to keep the Sabbath holy, we thus figure out both what grounds and what compels it, via appropriate intertexts. "Remember" harks back to the Creator's rest on the seventh day, for example, and inversely, the noncompliant addressee's oblivion looks ahead to the death penalties that the Code specifies and the history stages in resuming the matter (e.g. Numbers 15: 3 2 36). The former provides a motivation of God-likeness, the latter an earthly sanction. One can also invoke an "unless ..." (e.g. some overriding emergency) to qualify the memorial "because . . . " implicit at the start, and one can likewise figure out a reward (e.g. divine blessing, group approval)
64
Meir Sternberg
to balance the latent threat at the end. A persuasive mini-plot accordingly projects itself, as usual, onto the future, now below the precept altogether. Also as usual, that future has its own earlier and later and latest, symmetrical to those of a past narrative action. The given Do/Don't, taken as intermediate link, reaches behind one in medias cutoff for exposition and beyond the other for stick and/or carrot. More generally, the apodictic style at its barest least formalizes and forecloses (hence most challenges) Aristotle's overexplicit "beginning —» middle —• end" whole, the paradigm of gapless enchainment to this day. In the two-clause examples (3), even the formal likeness to casuistic ordering and syntax and enchainment grow more visible, with all the relevant narrative implications. At second glance, at least. On the face of it, the grammatical subordination transforms from the conditional's protasis/apodosis to subject-clause/main-clause, where the former identifies the agent (or patient) whom the latter concerns. Despite the shared hypotaxis, one preposing the subordinate term at that—it may therefore appear—the subject just runs to an entire clause: "He who kills a beast ...," "Every person who through cowardice shall in time of action withdraw or keep back ...," or, elsewhere, "Whoever does X ...," "All Y's that ...," and so forth. Below this surface, however, the commonality extends to the deeper meaning and chrono-logic. Look again at our examples (3). The subject clause there reads like an antecedent whose subject defines the agent by the act (beast-killing, withdrawal through cowardice) that makes him subject to the counteraction (payment, death) imposed by the main clause in response, to the same conditional effect as (2). Given the premise, the consequent "shall" follows. The overt conditional clause in (2) itself operates as a restriction of the ensuing directive to the antecedent circumstances—or, recalling the etymology of "casuistic," to the case premised. It might as well open with "He who has newly taken a wife ...," "Every husband newly married ...," "All seigniors wishing to divorce ..." Predictably, the two forms interchange on record: Leviticus 24: 17-18, 21, for example, actually alternates them in consecutive laws, then apropos the very same laws, including that of (3a). 51 Either arrangement may also predominate in the different law-codes on world-wide record. All a matter
51
"If a man kills any human being, he shall be put to death. He who kills a beast shall pay for it, a life for a life [...]. He who kills a beast shall pay for it, and he who kills a human shall be put to death."
If-Plots: Narrativity and the Law-Code
65
of secondary choice, or stylistic variation, overlying and expressing the firm hypotactic cum action-logical unity. Under their divergent, protean guises, therefore, all three forms exemplified in (l)-(3) encode and communicate the if-plot that one of them flaunts. 4. Genarrativity As such, in directing (prescribing, prohibiting, permitting) activity, what the three forms all represent is some occurrence to come (to arise, not to arise, to arise at will) under the relevant (given, assumed) conditions, as opposed to one already arisen and now told after the event. As such, in other words, where they all appear to vary from standard narrative, the historical or history-like "Once upon a time ...," is in being modally qualified rather than categorical. Even so, the difference would then lie not in the action represented—the nuclear "propositional content," if you will—let alone in the extra action-logical interlinkage distinctive of it by the statute's rationale; nor in the peculiar generic effects triggered by our own eventful movement through the discourse sequence that emplots it; but in one of the assorted variable features that, however significant, do not make or break narrative. Instead, they reinforce or subordinate or just specialize it; and the law-scenario gains on balance multiple reinforcement amid variation, due to the very feature at issue and its corollaries or concomitants. So, even were we to grant the modalized statute's variance from the (hi)storied norm, it would not take away the commonalities reinforced. The disregard that this subgenre has suffered on account of its ifplot, inter alia, therefore betrays another prejudice hindering the wanted advance from fundamentals toward a comprehensive theory of legal storytelling and of the narrative genre as a whole. An odder prejudice than most, at that, even going only by what has emerged thus far. Ironically, the very appearances to which the dogmatist may look for cons turn out to be pros; and, on closer analysis, the reversal happens because, not in spite, of this antipolarity to our favorite, unmodalized stories of fact and fiction—or their appearances as such. If anything, the law-tale's modest scale, typal scope, and conditional yet binding status give it a rare narrative-generating capacity, with interest to suit. The gaps it leaves both require filling in terms of its own modal, statutory discourse, or interdiscourse, and also receive it via massive application cum expansion in practice: in all the apparently categorical stories (always about an actual happening once upon a given time) developed, invoked, examined,
66
Meir Sternberg
challenged, adjudicated by the lawyers who refer to the statute, possibly for centuries. 5 From police detection through courtroom drama to the ultimate judicial decision, the way of law enforcement is paved with variant retellings that would map the if-plot onto history. Always subject, of course, to the narrative rules, or meta-rules, in force—of evidence, alternation, interpretation, gap-filling, quoting, (cross)examining, argument, discourse, verisimilitude—but these guidelines only contextualize the fundamental generative must. One side needs to construct, the other to deconstruct and if possible to re-construct, and the adjudicator to sort out, a tale of the past that runs true to its enacted type, or rings like it. As history tellers, that is, the prosecutor works for identity to the legislative scenario, whereby the accused in the dock embodies the if-agent (or the ifagent typifies the accused) and the envisaged consequence should duly follow in the life-plot; the defense (re)tells for misidentification, hence inconsequence and overall inapplicability, and/or for the application of another law-scenario, or of this scenario to another person than the accused, all to Not Guilty effect; and the judge/jury decides, via replay and possibly independent typology, whether the balance measures up to legal truth, satisfies the onus of proof. At the very least, this fourfold repetition structure—with its distinctive members, narrators, tensions, ordering, turn-taking, interlinkage—is built into the legal process: a thrice-told history emerges from the scenario on the way to the sentence. Nor is the scenario's modal peculiarity—a conditional enchained with a command—least responsible for this emergence. The trial of narrative strength between the law-retellers after the event centers in whether the conditional did, and the command must, realize itself: in whether their emplotment has adumbrated the event at issue from (say, criminal) beginning to (punitive) finish. And so in every new case where the law's iffy potentials breed versions of life past, all made in its image, or mirror image, and vying for the strongest likeness. History52
They refer the case to the code, that is, not and certainly not only to sheer "common law" precedent, which narrativist approaches (inter alia) tend to overemphasize instead. With the statute excluded wholesale from narrative, one might predict the exponent's focus on case-to-case reference, where both analogues are obviously storied. Given an inclusive theory, however, the cross references aren't mutually exclusive, either, but can jointly thicken and strengthen the lawyer's narrative: two relevant action models of different kinds serve better than one. (Whether invoking a precedent does not itself amount to extracting a general law-like rule, or common denominator, is a question we needn't go into here.)
If-Plots: Narrativity and the Law-Code
67
likeness itself—and with it the ruling at stake, about what did and will happen—turns on the law-likeness: the relative fit (match, mismatch, sufficient match) between the given modalized type and its alleged categorical tokens. 53
This generative, or genarrative, power stretches yet further. Witness the bare-bones law-tale's interplotting with its neighbors (already exemplified apropos Commandments) in sense-making, harmonistic theory and in antagonistic lawyerly dispute, or the interplotting with its earlier application to some actual case(s) as precedent, or with its foregoing historical parallel in, say, "Hebrew" bondage. Or consider the stories of how the enacted law-story arose, as meta-told by legal historians and, in the Bible, tellable from antecedents within the canon. Or consider its amendment, restrictive or inclusive but always obliquely hermeneutic, by the legislator into a variant master tale in force henceforth. Compared with this enacted center and generator of an ever-widening, multiform (including crossmodal) narrative circle, the rest of the genre offers plenty of genarrative equivalents but few rivals and no superior. The analogy I suggested with the parable's invitation to decoding, gapfilling, translating into actuality, realizing in the world—the more so on Jesus's authority—is therefore the rule of generic genarration, not a special counterpart to a special case. Indeed, a composite text of immense manifest and ongoing genarrativity regarding every component, the Bible as a whole literalizes the rule behind this analogy. Among the discourse kinds that juxtapose, crosslink and reproduce there, take only our narrative pair: the codified scenario and the historical account. Both subgenres, apart or together, have over the millennia inspired countless follow-ups along all the lines of textual response, development, afterlife known to humanity. These stretch from Rabbinic retelling and New Testamental resumption (themselves notoriously proliferative in turn ever since) to the 53
A portmanteau term first introduced in Sternberg (1998: 525ff.), with extensive application, to designate the telling about the (law-)tale's historical or history-like process of genesis: how a narrative, statutory or otherwise, generates the meta-narrative of its own generation. Thus the Hebrew slave law refers back to the enslavement in Egypt, Homer invokes the Muse, a historian or a fictive (auto)biographer names the sources drawn upon: to the limit of foregrounding the quest for information, as in Nabokov's The Real Life of Sebastian Knight (1941), or the writing process, as in Tristram Shandy 1759— 67), where the reading features too. (Compare also the story of genesis that Sonnet [1997] traces along Deuteronomy.) Here, the term appears in a wider sense to include every enlargement that a narrative generates, complete with all the responses it implies or provokes.
68
Meir Sternberg
most recent fleshing out of the ancient discourse in today's assorted styles: interpretive, text-critical, homiletic, archaeological, legislative, sociopolitical, novelistic, cinematic, pictorial, for example. If a statute book is the largest anthology of mini-tales within the culture, then those anthologized in Scripture have elicited whole libraries from posterity. But so have their historied fellow miniatures, owing to the same drive toward elaborating and otherwise building upon what counts as a master tale. (Always with one key difference: no law-like mechanism of authorized closure, which would limit future genarration. Another productive minus.) In this light, it would also pay to think of a myth or a novel or a film as the center of its own varied products. Readings, intertexts, pre-texts, histories of genesis, revisions, sequels, adaptations, referrals to life, wholesale alternative mappings included—all form variants, or tokens, of an openended story cum type. Likewise worth comparing with the statute is the story reduced to genarrative type. I mean action schemas and formulas of all kinds, especially those given prescriptive, law-like control over discourse (here, narrative) performance, whether by the literary establishment or lowbrow taste or show business. Thus the relationship between tragedy, usually defined by what happens in it, or should happen, and conceivable or particular tragedies born of this nuclear event-sequence. When originating from a legislator deemed authoritative, as Aristotle was in the early modern era, the schema even hardens into a text: the Poetics on tragic action, complete with its inheritors' quarrels over the sense, reference, and updating or extendibility of the operative terms—as bitter as any waged in legal theory and practice. Consider the fortunes of "mimesis," or the arguments for and against "wholeness" defined by "beginning-middle-end" linkage, or the Three Unities read into the master text, or the ongoing wonder at the scaling of the happy above the unhappy closure—with their notable effects on the production and reception of tragic plots. All this doubtless shows a striking likeness to the other genarrative centers-and-circles instanced. Even so, the basic narrative schema in question (a character injuring, or about to injure, someone closely related) leads a different existence from its two analogues. It constitutes a story-generating mechanism, for tragic pity and fear, without counting as a story in its own right—the way both the law-tale and the novel do, amid polar modality. Unlike these, the very readers of the Poetics as a master text do not read it for the tale. With the representational form of narrated agency and much besides common to
If-Plots: Narrativity and the Law-Code
69
all three genarrators, the communicative function yet draws a line among them. 54
5. The Rule of Uncertainty: If-Plot Constants, Variables, Effects 5.1. Modality across (Legal) Narrative But then, narrative theory cannot escape the generic challenge perceptibly embodied in the law's modal features, either, by domesticating them into a special case or branch, off-center. They match, even highlight and radicalize, a set of omnipresent forces, workings, processes, effects—notably including those long established as definitional in my reconception of the genre, tested against an assortment of corpora, or counterviews, and now further extendible and refinable in turn by yet another subgenre, more peculiar than most. This two-way traffic is exactly the point of the argument. As a conditional directive, the law's if-plot foregrounds the Suppose Principle underlying all action logic (not excluding the historian's movement, often tacit, from whatever antecedents supposedly obtain to inferred consequents). Fiction, we recall, even carries this principle from the narrated activity to the supposed overall narrative act (e.g. Tristram's) and every discourse-event framed by it. In a yet wider overview, the multiform iffiness is integral to narrative communication, always gradually, ambiguously, hence modally proceeding between represented and reconstructed time, between ignorance and knowledge of the narrated world. Throughout the telling/reading process, the genre's universal dynamics of suspense, curiosity, and surprise all thrive on uncertainty. What leaps to the eye in the statute, then, energizes and dominates the mind grappling with other stories. And on top of it, literally, because visible on the surface, the discourse of modality proper abounds in the rest of the genre. Two widespread forms have already been exemplified: quoting speech or thought that the quotee (real or fictive) has not (yet) actually produced, and forecasting even nondiscoursive events along a repetition structure. So much for the untypical, if at all storied, appearance of the law; or, the other way round, for the preconception that narrative events, even if fictional, must be "given as a fact (in a certain world) rather than as a possibility or probability." But nor is it true—to revert to a more specific 54
More on the schema, including definitions of narrativity itself, see Sternberg (2003a), (2003b) and references there.
70
Meir Sternberg
denarrativizing claim—that the corpus juris alone breaks this alleged must, which the legal process observes: "fact-contingent, it is inescapably rooted in narrative." In the rest of the law field itself, rather, practitioners modalize their own stories or the other's all over, especially in epistemic terms. Some do it at will, some by fiat, all according to codified roleassignment. Lawyerly argument, or judicial analysis, involve of course logical reasoning toward anything between possible and necessary consequence. But reasoning goes here with telling, itself variously qualified or qualifiable, as well as downright factual on the surface. Indeed, the very reasoning modulates into the omnipresent narrative inference that we call gapfilling, all the way to the purely iffy. For gap-filling, always inferential by nature and so at best probable (in the histories the rival tellers fill out as elsewhere) is here even launched, sustained, assessed, and countered on sheer hypothetical premises. The range of modalities at work diversifies accordingly. Aiming for the benefit of doubt vis-ä-vis the prosecution's categorical tale of guilt, the defense will thus put forward a suppositional alternative, in the hope of lowering the probabilities below the high threshold required for the judge or jury to convict. Inversely, en route to a decision, the latter qualify (i.e. quantify by degrees of certainty) everything they have heard from the two parties—the "categorical" histories along with the openly modalized Suppose's of guilt vs. nonguilt. They approach and weigh the entire lot (with whatever care, skill, impartiality) as possible versions of the truth, and may opt for yet another storied variant of their own devising as the best-fitting, likeliest, most law-like according to their lights. In turn, their own exercise in comparative narratology leads, after all, to further modal hedging, here amid authorized decision. The verdict, though univocal between Guilty and Not Guilty, expresses not certitude but the next best thing: the absence or presence of reasonable doubt. The sentence may even be qualified outright, now deontically at that, because "suspended": an if-penalty hanging over actual life to come. And both are in their turn vulnerable to appeal, hence also inherently contingent, a priori, on higherlevel judicial comparativism, etc. Not that the extraordinary decisionmechanism built into the legal system—the dream of some literary interpreters, the nightmare of others—thereby compromises its authority or uniqueness. Rather, that very unique power of adjudicating (among tales, via an authorized tale, by appeal to an enacted tale) highlights the room systemically left for modality throughout the process.
If-Plots: Narrativity and the Law-Code
71
Analogous instances, then, proliferate far and near. Only, what is integral and/or always open to narrative representation—in every form, order of magnitude, aspect of possibility and necessity—recomposes afresh into the law-tale's differentia specifica. 5.2. How the Statute Radically Qualifies Its Story However, this specific modality—its factors, coherencies, workings, permutations—is not easy to describe. Take the prevalent threefold classification just illustrated, more or less explicitly, from judicial proceedings. According to it, modality subdivides into the alethic, the epistemic, and the deontic logics: concerning the necessity or possibility of propositions, of claims to knowledge, and of acts to be carried out, respectively. Now, identifying the statute as "deontic" 55 would be true, easy enough, and, for the narratologist, even promising—actions, definitionally at work here, are his business—yet insufficient. Not only does this subdivision encompass and merge numerous varieties (e.g. from order to advice to request to plea, from imperative to declarative to question form), which themselves ramify and crisscross, sometimes in ways yet to be understood. The "enacted" kind at issue also belongs to the least visited corner, where the directive mixes with the conditional, prescribed with permitted and always provisional agency in the world. This tricky deontic juncture awaits careful study, all the way to such finer relations as those between the Supposes of (in, behind) a military and a divine command, or between the legislator's if-plot and the judge's if-penalty. One may wonder whether the usual formalistic tools brought to modality could do the job anyway—and I for one doubt their very applicability to living discourse 56 —but they haven't really been put to the test here.
55
E.g. Maley (1987: 30). For reasons I have often detailed elsewhere: as in (1982), apropos quotation, or in (2001a), apropos the workings o f inference, especially (factive) presupposition. But those arguments, for and against, would readily stretch to our crux, inter alia, and will below. For now, just observe that the underground modalizing detected thus far— whether in the genre-wide processing or in lawyerly strife, in the apodictic style or in the trial's verdict—turns on a functional, holistic, which here also means narrative, approach to the discourse. Such teleo-logic is poles asunder from the questionable assimilation of modal to formal logic common among linguists, pragmaticists, and philosophers of language. (See also note 58.)
72
Meir Sternberg
Experts in modal logic and related fields have tended to evade the conditional imperative (among other nondeclarative kinds) even when uncomplicated by the extra specifics of the law. David Sanford disarmingly admits as much in his well-known book, I f P , then Q. Having cited examples like "If you see her, say hello" or "If music be the food of love, play on," he observes: "These are commonly neglected by theoretical treatments of conditionals, and I shall continue this tradition of negligence." The reason offered in excuse, though, looks somewhat odd. "A conditional imperative, after all, seems to pose no vexing problem," being "hypothetical rather than categorical." 5 The ground for avoidance, one suspects^is the very opposite of this hypothetical^ alleged nonproblematicness. A fortiori within the legal if-plot. Its specific modality involves a complex of factors that need sorting out to be appreciated as an integral whole vis-ä-vis the rest of discourse and particularly of narrative discourse. Factors, I mean, beyond the all-encompassing invariant: the "qualified" representation that definitionally opposes all modal (alethic and epistemic as well as deontic) utterances to "categorical" statements—these associated, inter alia, with the paradigm of history telling. Nearer yet to home, moreover, the factors involved in the Do/Don't if-plot need specifying relative to the commonalities within the deontic family proper: beyond what they share with the nonlegal varieties of directive and even with the subvariety of conditional imperative exemplified by Sanford. The question is how precisely the law-tale in statute form gets qualified. Along what axes and lines? Via what cross of doing and Do/Don't, actional chrono-logic with modal or activity-regulating logic? To what effect on the world's movement and our own in the role of subject, target, audience, possible actor, all in one, not to mention law-practitioner? It's best answered in an ascending order of differentiality and, for brevity, with technicalities kept to a minimum. First, a new time-feature joins those already found to be characteristic of the statute's narrativity, each with its peculiar grounds, effects, and implications for the whole. As regards sequence, we encountered an unusual, tense duality: the chrono-logic attached to the "If ... then" ordering is yet riddled with gaps, owing to the typal, hence skeletal plot, the in 57
Sanford (1992: 5). More realistic as well as more concerned, Pfersmann (1994) starts, instead, by emphasizing the difficulties that attend its unpersuasive and contested "formal" treatment in the references he cites.
If-Plots: Narrativity and the Law-Code
73
medias cutoffs, and the ambiguities of language or just its unequality to the plenitude of life. While reading in orderly sequence, we must look ahead and backward for closure. Next, as regards duration, there is the omnitemporality associated with the law as a standing Do/Don't apropos a scenario liable to recur indefinitely, from its enactment by the sovereign power onward. To these hallmarks, the modality involved adds a marked temporal reference. Like all types and subtypes of deontic utterance, and in polar contrast to standard historical or history-imitating narrative, the law-tale is on the whole future-oriented, representing a prospective action. The "In future" that opens example (lb), from the Magna Carta, is obviously dispensable, being (re)encoded in the sequel's "no official shall place on trial ..." and even in less overt forms of command. Whether expressed by an appropriate phrase or verb, or twice signaled, as here, or left unmanifested at all, the "in future" bearing goes as a rule with the deontic modality. Its redundance here presumably means to evoke and negate by implication past injustices, rather than to express the self-evident futurity of the new limit set to a citizen's indictment henceforth, and with it to officialdom's arbitrary power. But self-evident on what ground, under what rule? In the current study of modality, the restrictive linkage of all flats to futurity appears forced by sheer logical necessity (hence across all possible worlds) on pain of nonsense. Needless to say, as it were, "we cannot rationally command some59
one to do something in the past." More exactly, however, this cannot be done in the so-called realistic human condition, with its irreversible arrow of time, and not even in "supernatural" worlds and law-worlds such as the Bible's, where an otherwise omnipotent God stops short of reversing the arrow himself, as the Almighty always might. 60 But then, he always might in reason, and his might is even actualized (usurped, after a fashion) by the earthly retroactive legislator—an outrageous measure, but not impossible nor, alas, unattested.61 In principle, therefore, the "connection be59
61
Lyons (1977: 751). And exceptionally does, as though to prove for once both the capacity and the usual restraint chosen. In the sign given to the doubting King Hezekiah, the shadow cast by the sun on the dial turns back ten steps, just as it might advance ten steps if the king preferred a jumping-ahead miracle (II Kings 20: 8-11). This time-reversing show o f omnipotence has a durational equivalent, likewise unique: the sun, and with it time's flow, gets arrested in mid-sky to enable Joshua to complete his victory (Joshua 10: 1 2 14). Still, jurists would domesticate its abnormal temporality into chrono-logical (if not moral) normalcy: formalism dies hard, especially when appealing to common sense as
74
Meir Sternberg
tween deontic modality and futurity" is not "intrinsic" or "rational (as misgeneralized by Lyons on the accepted view 6 2 ) but ontology-bound or empirical. Not at all a matter of formal logic, in short, but of reality-frame. A significant difference, this, whether for retro-legislation or for our wishful thinking with hindsight or for the poetic license of fictionality: the past does essentially lend itself to deontic (as well as to otherwise modalized and to ostensible categorical) backreference, anticlockwise as it may be. Like all time travel to things foregone, an order (in both senses, the directive plus the directional) involving regress is not a contradiction in terms. 63 Here, actually, the deontic time-feature aligns in principle with its ordinal and durational mates. Laws always can, and some do, invert the "If ... then" sequence and/or restrict its effective span, even prescribe a oneoff action, unique in spacetime (e.g. dissolving Parliament). Both hallmarks, though well-motivated, are characteristic, not constant, much less inherent to the kind's discourse or world or their relations; and so is the reference forward. What makes narrativity, legislative as otherwise, is the change of state (fore)told and experienced, rather than its direction. In common law-making practice, at any rate, a legal directive about the world, however revolutionary, is indeed subject to the unidirectionality of the natural world-line, and accordingly future-oriented. By the same
well. (See Pfersmann [1994: section II] for some arguments and references.) But nor does the abnormality end here. A variant of retroactivity, in inverse form, occurs with legislation governing a state or development (e.g. technology) that doesn't yet exist in reality, and may never come into existence, possibly even representing it as if it does and did. In short, the enactment anticipates an event unrealizable at the time, so that, if ever realized, it will find itself already subject to regulation, conditioned at birth. For an example, with science-fictional twists, see David Nimmer (2001): the jurist applies there my narrative guideline "retrospective incoherence signals (guarantees, invites) prospective coherence" (1985: 280) to make sense of the Digital Millennium Copyright Act passed by Congress in 1998. 62
Cf. Lyons (1977: 746, 751, 824). Outside the law, recall the histories starting, or restarting, with a counterfactual "If ..." Fictionists imagining possibilities (or re-imagining actualities) may also flaunt such an antecedent. Thus, John Wyndham's story "Chronoclasm," whose heroine travels back from the twenty-second to the twentieth century—she studied it in her history lessons— rests on an explicit conditional Suppose. It belongs to a collection of science fictions described by the author as "experiments on the theme: Ί wonder what might happen if . . . ? " ' (1959: 8). A novel like Kingsley Amis's The Alteration (1976), premised on an English pope, as if Henry VIII never broke with Rome, is a cross between the two— only minus the opening retroactive formula that would instruct us to start anew.
If-Plots: Narrativity and the Law-Code
75
(onto)logic, futurity goes with the law's assorted deontic fellows: commands at large, demands, requests, recipes, pleas, warnings, exhortations, invitations, dissuasions, prohibitions, etc. It also runs through the still larger category of what I would call prospectives: hopes, fears, wishes, plans, prophecy, day dreams, fortune-telling, or the scenarios evoked in the receiver's mind by narrative suspense. This is where, and why, the envisaged law-action stands to its categorical tokens (e.g. the prosecutor's backward-looking narrative of guilt) as the member of forecast (even if just wishful or predictive) stands to its fulfillment (accurate or divergent) within any extralegal repetition structure. Another counterpart is the relation between quotings that anticipate and that follow the speech/thought quoted, either again performable outside the law's domain altogether. Within that domain, inversely, practitioners by nature tell and countertell and retell, quotes included, what already happened. They may hypothesize stories (the detective's tentative reconstruction of the crime or the defense's indication of an alternative criminal) but must yet orient them to the past. Being hypothetical, these stories obviously never claim the hardness of fact; yet, being reconstructive, history-like, nor are they images, much less directives, of futurity. The statute therefore finds no wholesale equivalent even in the Suppose's constructed during the process of law enforcement—all both retrospective and epistemically rather than deontically modalized—except for the judge's suspended sentence. None but the tales projecting ahead at the two ends of the legal system and sequence meet here, as though to close the circle of the open future. 64 In turn, the future being open and opaque vis-ä-vis the normally closed and at least theoretically accessible past, futurity involves doubtful representation: contingent, hypothetical, provisional, ever vulnerable to chance and challenge. Whether what is projected to happen will in fact happen remains uncertain. Twice uncertain, let me emphasize. Objectively so, given the world's numberless potential turns (always under the common ontology, excluding the representer's omnipotence, hence any failsafe projection); as well as subjectively uncertain, given human ignorance (with the corresponding epistemic proviso, exclusive of omniscience, hence of guaranteed clairvoyance). God alone can tell for sure whether the action ordered, requested, encouraged, threatened, planned, wished-for, dreaded will ever (or that prohibited, never) transpire in life: seeing to it takes Almightiness, foreseeing it entails the All-Knower's viewpoint. 64
Both the negative and the positive analogies will soon turn out more complex.
76
Meir Sternberg
Once again, the Bible serves as a defamiliarizer of what appears selfevident, a corrective to hasty generalizing from sublunary empirical limits in the name of reason. Nothing is unreasonable about "Let there be light, and there was light" in Genesis, or about the preannouncement of the plagues inflicted on Egypt: both perfectly accord with the operative reality-model, which just happens to vary from that unthinkingly taken for granted in discourse (e.g. deontic) theory. That God can foretell with absolute certainty makes another principled exception, which further contextualizes the rule itself to undercut any "logical" package-dealing of the factors involved: here, of prospective and provisional world-imaging. In the human state, though, the axes indeed coincide. For us, deontic goes together with epistemic modality, as well as with an irreversible and uncontrollable ontology. How much more so in the law's directives, across formal schemata. Besides the conditionality associated with all prospectives as such, deontic or otherwise, their if-plots entail two further sets or series of conditions— one general and tacit and unique to the statutory law-tale as a whole, one specific, whether expressly or obliquely, to each instance. The former consists in the silent if-midplot, the three unspoken cumulative premises on which all penal consequents (with various analogues elsewhere in the law) hinge: "if detected and if tried and if convicted." In fact, this threelink middle subdivides at every juncture to extend the chain of hypotheticals and radicalize the action's ifflness, incomparably so. 65 Besides, statutory if-plots grammaticalize or imply conditions specific to each, like those analyzed in our examples. Wife-gladdening, free of public service, depends on the time of marriage; honoring one's parents, on Sabbathkeeping; the death penalty, on cowardice in battle; and so forth. In a sense, as also demonstrated already, analogous particular conditions govern all narrative, including the (hi)storytelling done at every trial. No narrative without exposition (overt or implicit, preliminary or deferred, block-like or distributed) in which to ground the movement of its world: to establish the reality key, the norms and web of society, the agents' characters and biographies thus far, the initial state of affairs at large. Only, a significant variance now emerges between these omnipresent Suppose's, all anterior to the narrated movement in time past, and their 65
The normative majesty of the law, embodied in the apodictic Commandments, would rather minimize the contingency that its enforcement involves as an operational process. Burying the extra antecedents underground thus reconciles the conflicting demands.
If-Plots: Narrativity and the Law-Code
77
counterparts in the future-oriented statute. The respective expositions (the past relative to the past events told vs. the premise to the eventuality foretold) markedly differ, and the two chrono-logics as a whole with them. Given an accomplished, already-performed action, historical or fictional, the antecedent Suppose's operate to make it intelligible, entertainable, followable, plausible, acceptable—or, when gapped or otherwise wrenched, the contrary—rather than performable in the first instance, the way the law's iffy antecedents make their consequents. The former premises condition the telling/reading, while the latter also condition the told/read happening itself: the discoursive as against the very actional dynamics, respectively. Here lies the point of the following exchange between two detectives confronted with an unexpected murder: "But no!" said Dr. Bottwink wearily. "It is impossible! By all the rules of logic and reason, it is impossible!" "But it has happened, Dr. Bottwink," said Rogers. 66
The "impossible!" is expressive (of incredulity) rather than logical. What happened in reality happened, never mind whether and how we learn, understand, evaluate the why's behind it all. But for the statute's "then" to happen, the " I f ' must realize itself first—wife taken, cowardice shown—or else the entire two-phase action will remain purely hypothetical, literally impossible, no matter how well understood. A fortiori with the unspoken general if-midplot taken into account: every single one of its required links must duly ensue for the counterdoing to become possible within the scenario, never mind enforcement in reality. The "If s" conditional force thus hardens in the law to maximize the "then's" uncertainty—beyond the common suppositional deontic and narrative grounding, even when rolled together. Nor is God's law-making in the Bible an exception. He needn't figure in his laws at all, and never forces their conditional on the subjects, willy-nilly, as he always might, with a view to the consequent. Nothing like the ultimate agent provocateur, so to speak. His supernatural powers, unprecedented and unmatched, only ensure the workings of retributive justice on earth, by a divine superintendence. God will counteract in person where the need for 66
Hare (1957: 147). Outside the "If ... then," to be sure, the consequent may realize itself on its own (e.g. wife-gladdening beyond the first year, the death penalty for offenses other than cowardice). But the law makes it, and with it the whole event-chain, contingent on the premise.
78
Meir Sternberg
balance arises in response to a prior human act—to an antecedent materializing for better or worse—rather than interfere to get that act performed and so initiate the chain. Thus the longevity guaranteed to the filial child, or the death to the oppressor of the weak. As part of the same normative revolution whereby he gained omnipotence along with omniscience, the Bible's God leaves the choice and the initiating move to humanity: addressed to agents created free, in the divine image, his "If ... then" assumes special ideological, on top of action-logical, point. Among all arenas, here even God's extraordinary control and knowledge of the world is realized only in enforcement after the fact, and so validates the rule of the scenario's uncertain occurrence in the law-world. The nearest equivalent to the statute here is the judge's suspended sentence, its actualization rigorously contingent on the guilty party's behavior during the future stipulated. But even relative to the judge in turn, a fortiori to various other predictive hypothesizers, often ill-wishing, the law would generally much rather that the unfavorable contingency should never arise. The entire penal code, for example, aims to deter wrongdoing and save everyone the counterdoing in the process of law enforcement, so as to make the whole action —> reaction sequence still less likely to occur. But even when pressing for compliance with the Do/Don't—via the envisaged tale's rhetoric or, much less often, a beneficent retribution foretold—the statute expresses a marked preference. Dissuasive or persuasive, the code is always normative, more thoroughly so than the (un)happy ending of a past affair. In deontic terms, it brackets the directive with the desiderative submodalities, conditional agency with wishful thinking. Nor, again, does this rule except God's law. Designed to form a chosen people who will live up to its high standards, it manifests the normative and educational code par excellence, with a virtuoso art of suasion to match. 68 The preferences, suasion included, newly divide with the sides in a trial. Compared to other storytelling, legal as extralegal, theirs is remarkable for its multiple intermediacy. What these adversaries tell stands between the law invoked and the law application imminent, between fleshing out the " I f ' and anticipating the "then" in real-life terms to a real-life audience. Hence, more generally, they also stand, once again, between the categorical and the variously modalized, as we have already observed them doing elsewhere: apropos the Supposes and gap-filling integral to
68
Sternberg (1998: 471ff.); cf. (1985: 4 4 1 - 5 1 5 ) , (1992b) on the corresponding narrative rhetoric.
If-Plots: Narrativity and the Law-Code
79
narrative, the mixture of truth-claiming with sheer hypothetical reconstruction, the appeal to special rules of evidence, to the legal measure of proof, to precedent and precept alike. Now, either of the lawyers stands in addition between the past events told and the sequel foretold by them, between the history-like and the hortative (persuasive or dissuasive) combined with the hopeful. Only, their hopes pull in opposite directions, governing the respective tales to stage a tug-of-war, a trial of strength, literally. The prosecutor's narrative of guilt wants to establish that the antecedent envisaged by the law has arisen—the case meets the condition, the token fits the I f s stipulated type—and so the consequent should follow. The defense counsel would have it believed, instead, that this " I f ' doesn't apply, or if it does, not to his client or not beyond reasonable doubt—a mismatch or insufficient match—and therefore the "then" shouldn't apply, either. They would each evoke a past history tailored to a desired future; yet between them that ending is left open, uncertain, polarized, and suspenseful accordingly. While foreclosure is their embattled business, the closure of the whole enchainment—of the future as well as, and in line with, the past—falls to the adjudicator. And the verdict encapsulates not only the retrospectioncum-prospection authorized by and (in theory) accordant with the law but also the normative response it makes, for better or worse, to what officially happened. Guilty vs. Not Guilty, too, express opposed value judgments—not least hopes, fears, wishes, solidarities, antagonisms—in and through their legal enforcement: at times, the normative judgment even comes ahead of this enforcement, conditioning it against the standards of law-likeness. (The O. J. Simpson trial is a notorious case in point.) Either way, the desiderative modality persists to the finish, and there recombines with the deontic in the Do/Don't ("The accused shall/shan't ...") of the judicial sentence. Or, better, in the sentence's Do/Don't vs. Doable, polarized between another Must and May, necessity and contingency, imposed and suspended future. Here, the two ends of the law-system's grand chain meet once again, possibly to the extent of sharing the modal qualification by incertitude as well. For the judiciary has its own conditional directive: the if-penalty that applies the if-plot to a specific case in history with a loophole attached, as though hope of better things never dies. The if-plot's uncertainty, however, is not only definitional but also given to further radicalization, along three main lines, all current and combinable in legislative practice. First, there are the qualifying clauses addable to the initial premise (and to its tacit umbrella follow-ups, our "if-
80
Meir Sternberg
midplot") in the sequel: 69 each addendum narrows yet further the scenario's range of applicability, hence lowers its chances of actualizing itself. The more specific the type, the fewer its tokens; the more conditions piled on the antecedent If, the more iffy the consequent as well. Second, the qualifications often systematically diverge to subgroup the possibilities, rather than accumulating for sheer restrictiveness. In narrative terms, there arise branching, alternative scenarios ("If ... then ..., but/or if ... then ...") or, as in the Hebrew bondage law, subscenarios ("Ifi ... and if2 70
... then ..., but/or Ifi and if3 ... then ..."). These forks all run parallel within the law world, in a way comparable to a multilinear or multiworld novel. But the sharper the forkedness, the lower the prospective frequency of (co-)occurrence. Third, and most fundamental, there is the statutory equivalent, or precursor, of the judge's suspended sentence. Both foreground the unknown, except that the codifier not only retains but also wants even less control over the outcome. The code generally includes permissions alongside obligations and prohibitions. "When thou goest into thy neighbor's vineyard, thou mayest eat thy fill of grapes ... but thou shalt not put any in thy vessel" (Deuteronomy 23: 24): subject to the veto on carrying anything away, "you" hereby gain the right to eat there, even gorge yourself, without counting as a thief. While equally regulating behavior, such licenses entail of course a change in the deontic modal operator, from "shall (not)" to "may." This means a reversal in the action's boundness—all the way from necessity to possibility, from the law's control to the agent's empowerment. Inversely, the event's likelihood weakens afresh even 71 within the scenario. Given the antecedent, a command will leave the subject no option in law—the act must (or, if negated, mustn't) ensue— whereas a permission builds subjective optionality, and with it objective open-endedness, into the consequent itself. For permission is by nature double-edged. Granted the right to something, you may also choose not to 69
See examples in Bhatia (1994: 144^9). "7/i thou buyest a Hebrew slave, six years he shall serve and in the seventh he shall go out free for nothing. I f i he comes in single, he shall go out single; if he has a wife, his wife shall go out with him. If his master gives him a wife ... then ... And i/4a the slave plainly says ... then ..." (Exodus 21: 2-6; my subscripts). In overall balance, though, the silent if-midplot, attached to all Do/Don't's, needn't apply here, for obvious reasons. It reapplies, in modified form, only when the empowered subject is charged with exceeding the limits of permission: a comparable if-series then follows, but ends in an if-acquittal. A conviction would of course reverse the whole chain into the appropriate Don't scenario.
If-Plots: Narrativity and the Law-Code
81
exercise it, doing as you please in the matter, possibly to the limit of total inactivity: no self-indulgence, ever, at a neighbor's expense. The licensed eating scene, a fortiori the qualifying veto at its end, will then remain theoretical in your case and every abstainer's. That is why jurists call the permissive rule "discretionary," as opposed to mandatory or directory.72 The freer the agent, in short, the more doubtful the action; the less pressure, the greater the plurality. Inherent and variously given to extra weakening, then, the uncertainty of the statute's if-plot culminates in the passage from the Do/Don't to the Doable. But, as with the gapping of particular reference, only on a larger, scenario-wide scale, (de)ontic and epistemic uncertainty makes for heightened narrativity. The projected world-image's loss in surety vis-a-vis "categorical" representation is the projecting discourse's gain in interest and thickness and unrest, owing to the multiple, often rival futures left suspended in our mind. Not that categorical narrative lacks forked prospection (among other common, genre-wide dynamics), given the universality of suspense, with its multiple Suppose's about what lies ahead. But this constant grows intricate and radical here, along various distinctive lines. Paradoxically, at the same time that the law's suspenseful "If' ensures closure by a "then"—a temporary gap, resolved next—it also entails deeper, larger, permanent open-endedness by its own duality: it initiates a scenario all poised between occurrence and nonoccurrence, with their polarized normative (desiderative) values to boot. Recall my earlier analysis of this if-plot's suspension between the limits of deontic necessity and impossibility, hinging on whether the antecedent will obtain. "Duality," or "limits," I say, because the universe of codified must's rules out the inbetween spectrum of probabilities: either the whole shall happen or nothing. According to (2a), it is either a citizen newly wed and therefore exempted from service or an old-timer doing his civil duty—rather than any cross of terms and any calculus of maybe's. (That we readers can quantify such likelihoods by experience—from the chances that the antecedent will occur to those that the consequent will duly follow—only entangles and heightens the prospection, now on epistemic grounds: the way of the world as known to us. 73 ) In turn, the suspense guaranteed by the " I f ' will 72
Maley ( 1 9 9 4 : 2 0 - 2 1 ) . "Murder will out" and "Crime pays," for example, generalize opposed empirical views of what is likely, even bound, to happen: offenses will incur and elude the stipulated punitive counteraction, respectively. Other estimates relativize this opposition into
82
Meir Sternberg
rise and modulate according to the variables: the further qualifications, the branchings and parallel alternatives, the discretionary may's, which translate into actional "may or may not happen's." All these, finally, gain another lease on suspensefiil life in the courtroom narratives that dispute and decide their applicability to a singular real happening. Throughout, you will recall, the dynamics of suspense in the law is anything but a matter of idle narrative thrills, processually or pragmatically, even less so than in our actual and artistic experience of forked futurity. Life and death may hang in the balance of how the given enacted prospective gets interpreted, complicated, applied, adjudicated.
6. Uncertain Law Action, Certain Law Enactment 6.1. An Exemplary Told/Telling Bipolarity Moreover, the overall narrative effect gets complemented and complicated from the opposite direction, too, thus having the best of both limitresources. The uncertainty of the statute's represented action in the world finds a counterpart as extreme in the absolute certainty of its discoursive action upon the world. Here the dynamism of the enacted if-plot reveals a new two-facedness, as elusive as it is crucial and generally instructive. It may escape notice altogether, because of an unfortunate ambiguity in "conditional directive (or imperative)." The phrase is liable to suggest an iffy working across the board: as though everything about such a deontic utterance were qualified out of necessity into mere possibility. This would conflate the two levels of action that I've just contrasted, the extralinguistic happening modalized (and variously so at that) with the failsafe effect of the deontic modality that binds it in the telling. 74 Instead, given a conditional directive, the conditionality applies to what is directed—ifX ob"perhaps," "provided that ..." or the statistics o f law enforcement. But all such forecasts, never mind their hardness or ground or rifeness or objectivity, generate yet another play between event-lines: an interplay of the empirical with the legal scenario, always complicating the reader's suspense. A similar conflation may befall narrative at large, as though the told and the telling events couldn't modally diverge, far less polarize. But my opening glances at repetition and quotation already indicate that they can, and variously do. Thus, I can foretell a narrative performance (my own or another's) that will look back to some historical development, or report one that looked forward to some hypothetical eventuality. The law, then, is also a paradigm case o f this noteworthy told/telling interplay.
If-Plots: Narrativity and the Law-Code
83
tains, then Y ensues—not at all to the force of the directing language, which regulates the whole agency involved in the scenario, no matter what. The directive's "shall (not)" imposes on us (citizens, law enforcers, arbiters, depending on the immediate addressee) a new categorical duty to act as specified: the obligation to gladden the bride, for example, or the veto on killing, both modifying life under the law in their respective areas. Even the permissive "may" grants an inalienable right to choose between lines of action, with the same novelty effect on existence, only in reverse: it makes us freer agents. Therefore, whether the scenario foretold by the legislator will materialize in any world remains uncertain several times over; yet the provisos and privileges he legislates in foretelling it stand for all time in our world. This bipolarity is encapsulated in the Do/Don't shorthands, which twin the unqualified command (or, if a Doable, license) with the qualified status of any future human doing, commanded as otherwise; likewise, we find the bipolarity miniatured in "enactment," readable as both the "thing done" by the law-maker vis-ä-vis the addressee, instantly, and the thing to be done by the addressee eventually, if at all. I suggested as much at the beginning, and the argument since has comparatively defined and explored the narrativity peculiar to one of these twinned extremes: the statute as foretold story, chrono-logic, master tale. We now turn to the statute's other face, the deontic power unleashed on reality in the foretelling. It is no easier to theorize and correlate, perhaps even to get acknowledged in the present state of narratology and related disciplines (especially legal theory, philosophy of language, pragmatics, with their interfaces since the 1960s). By the same token, though, a careful analysis of this twinned discourse pole again promises to yield more general, or generic, insights as well. 6.2. Speech A c t Theory Tested against Law-Speaking The suggestions of law-discourse activity packed into the ancients' Do/Don't or today's "enactment" cross with miscellaneous related hints in folk wisdom, artistic practice, literary study. Thus the countless variants of the saying that life and death hang on the tongue, or the faith and fear attached to certain expressions, or the universe-creating utterances in Genesis, or the pregnant silences in dialogue whose minus yet charges with tense energy the interchange they suspend. Literary critical parallels surface here and there, reaching as far back as Aristotle's ideal of poetic speech. Ideally, the words spoken will all get assimilated to the mimesis
84
Meir Sternberg
of action: the way drama emplots verbal interchange—later, even soliloquy—while the epic keeps the teller's voice disembodied and unincorporated, apart from the dramatis personae's. (Similarly with the emphasis on the impact of language use, notably including forensic discourse, in his Rhetoric and throughout this persuasion-oriented line.) Modern examples range from the neo-Aristotelian Elder Olson on poetic speech as a mimetic act of saying to the Russian Formalist Victor Shklovsky on the speaker's eventful address, "with complex peripeties," in a love poem by Pushkin. 75 Inversely, both attack theorists—William Empson and Roman Jakobson, respectively—who concentrate on the meaning or make-up of the poetic word as distinct from, or opposed and superior to, the mere action unrolling in the world. So insights classical and latter-day, targeted on everyday life and verbal art and moving rhetoric, point the same way. That they all bear on the power of nondirective language (formally outside the deontic realm) in effect helps to generalize the point. The implications for directive, command, statute as "enactment," would only follow a fortiori. In short, this reveals the law as the quintessential theater of discoursive agency, while underlining afresh our need to pinpoint its speciality. But the idea has now famously gained voice in Austin's speech-act theory, clean against the positivism of the philosophical tradition. We always "do things with words" and the things done always outreach sheer mimesis, description, representation, truth-bearing, propositionaling, veriflability, informational value, saying something—depending on one's favorite jargon. Even the most categorical (Austin's "constative" or "descriptive") statement, geared to some particular object in reality under the hardest truth-conditions, actively does as well as merely say things, or rather does them in and through the saying. It will thus engage in assertion, or perform an assertive act, with the attendant commitments. The more evidently so when the utterance shifts the say/do balance toward performativity. For instance, whatever becomes of the neglected "conditional imperative," you might expect the performance of the imperative as a deontic utterance (the law's included) to come into its own, alongside other forms and forces adaptable to narratology. The expectation redoubles owing to the early contact between the fields. Austin himself already glances in passing at the legal system and
75
Olson (1952); Shklovsky (1991 [1970]).
If-Plots: Narrativity and the Law-Code 76
85
jurisprudence, doubtless as a result of his association with the distinguished philosopher of law H. L. A. Hart at Oxford. Hart returns the compliment: for instance, his best-known work, The Concept of Law (1994 [1961]), draws a running comparison between the obligations imposed by the law and in the exemplary speech act of promising. These early beginnings have been widely followed up since, over the last two decades in particular. (Kurzon's "It Is Hereby Performed... ": Legal Speech Acts [1986] offers the first monograph-length treatment.) Even with narrativity and narrative study left out of the picture, as usual, a fruitful traffic between the other domains—legal and analytic linguistic philosophy— would therefore make it easier to repair the omission: to advance from twofold toward threefold interdisciplinarity on the basis of a shared concern with (discourse) agency. Unfortunately, even the limited premise, or promise of fruitfulness, has yet to materialize. The hope of interdisciplinary exchange lays bare holes instead—conceptual and empirical—not just a gulf. Speech-act theory has indeed got into notorious trouble on its home ground as well, in pragmatics and the philosophy of language. From a broader and more holistic vantage point—to cut a long analysis short—it remains immensely suggestive, yet too weak to capture much of the power of discourse as a whole: let alone that of narrative discourse, the strongest and richest of all in "performances" on its multiple levels. Austin, with his systematizers and appliers to the law, even fail to appreciate this generic difference— though important and, you would think, especially inviting, to speech act theory itself as such. For example, their use of "description" vis-ä-vis linguistic action, but regardless of whether the language represents an action, is symptomatic. As usual in philosophy, the term indiscriminately subsumes all "propositional contents"—the narrative alongside the nonnarrative, including the "descriptive" in the narrow generic, antipolar sense that literary study has encoded and explored. This fuzziness is not so bad as the outright denial of energy to the law-code that we have already encountered, but the two give away much the same unawareness of narrativity. Regarding deontic modality proper, we hear that an order to carry out some act does not express "a proposition which describes the act itself. What it describes is the state-of-affairs that will obtain if the act in question is performed." 77 True,
76 77
E.g. Austin (1962: 4, 7, 22, 31, 33, 150ff.), (1970: 235-36). Lyons (1977: 823).
86
Meir Sternberg
except for the recurrent yoking together of "act" with "state-of-affairs" as targets of "describe." Conflated thereby are representations of the world in movement and at rest, of agentive dynamics and existential statics, of doing or becoming and being as objects. Thus, when I say "Eat!," I describe nothing, but I perform an act of ordering, and so foretelling, an act or activity of eating; while in the utterance of "He is well," my statement describes a state. (Similarly with less tell-tale overinclusive synonyms, Austin's favorite "constatation" among them.) A pity, this missed opportunity to correlate, rather than bracket, the speech doing with the doing spoken about or its absence. The lacunae extend, often in chain reaction, from the represented to the communicative arena, a fortiori process. A host of functions and responses and movements—the three universal narrative effects among them— altogether elude the saying/doing, constative/performative binarism or axis. Nor is this very axis, our main focus now, adequately theorized by and after Austin. The "illocutionary" doings it covers generally prove at best low-level, underdeveloped, isolated, and anemic: as far from what the narratologist or the common writer/reader means by "action" (and we would now extend to discoursive interaction) as from what the legislative system performs by "enactment." Apropos verbal (inter)action itself, their weakness is a far cry from, say, the "plot, with complicated peripeties" that Shklovsky traces in the lover's address to his one-time beloved. Found wanting under the test, these speech "doings" only throw our concerns and desiderata into sharper relief. Thus, to claim that God, in saying "Honor thy father and thy mother" or "Thou shalt not kill," performs the illocutionary act of law-making, or 78
issuing an absolute directive, or codifying an obligation, and so forth, would not much advance our inquiry. Such pigeon-holing almost sounds like a tautology. It amounts to alleging that God has shown an intent to communicate a law: if not quite a misnomer, this hardly bears per se the weight and justifies the name of "(en)act," let alone "force," with their dynamic implications. Nor, to judge from the analytic record, would we find the key in the performance's breakdown along the lines attempted by Austin et al.: into subacts (e.g. locutionary, illocutionary, referential, predicative, etc.) or felicity conditions (preparatory, sincerity-demanding, essential). Apart from the doubts about which requirements bind which act, which in real78
As already in Austin (1973 [1962]: 4), with numberless echoes since.
If-Plots: Narrativity and the Law-Code
87
ity's living practice and which in a philosopher's Utopia—still unresolved—need they apply at all, across the board? If they did, then God's law, for example, would be null and void, since its making neither follows nor even assumes (or knows) any "conventional procedure" that should allegedly, "felicitously," govern the words, moves, parties, thoughts, and circumstances involved. (Nor, for that matter, does the Magna Carta, or the American Constitution, or any enactment foundational or revolutionary, again in direct ratio to their impact: exactly because they, too, mark some unprecedented turn, or leap or breach, in the plot of history. And even established legislatures may choose to abridge, or suspend, their own formal procedure. ) By the same token, one wonders how the Bible's world could possibly emerge out of the void, failing any world-creating decorums for God to observe: the revolutionary creation through the divine string of imperatives (Genesis 1-2) would become inconceivable to faith and fantasy alike. 80 Either reduction to absurdity also warns against overgeneralizing the human theater, all the more complex and versatile for the nonomnipotence of its agents, at speech as otherwise. For example, what is the common denominator of statutes made by a tyrant, an absolute king, a minority in power, a popular junta, and a democratic legislature, all enforced with equal vigor? Certainly not any (genuine) dependence on Austinian prerequisites for "felicity," except within an ideologist's or anachronist's vicious circle. There, the democratic kind would alone qualify as statutory, because alone attached to a procedure, and only when that attachment has materialized in full—clean against the teleology of speech as activity, judged by its doing things, empirically. The overconditioning of our human arena by the theory grows doubly perceptible vis-ä-vis divine omniscience. While God uniquely sees into his characters, our own minds—like theirs and their novelistic fellows'—remain opaque to conversational partners or observers in general. How, then, can social doing and interdoing (e.g. law-making) hinge on secret thinking (e.g. "sincerity")? A closer look, taken below, will only further expose the basic unreason and suggest a viable alternative. 79
80
In World War I, the Speaker o f the House of Commons ruled that "Parliament had a right to enact a statute in any w a y it thinks fit," and "the court declined to enter into the way Parliament regulates its own affairs" (Kurzon [1986: 15]). Equally significant, in contradiction to this fact, Kurzon himself would later (ibid.: 44) impose "a prescribed manner" on such enactment, in the name of Austin's felicity conditions. Sternberg (1985: 106-8).
88
Meir Sternberg
Inversely, even where more tenable or undisputed, the Austinian components of illocution are still too atomized and rarefied to capture (inter alia) the power exerted by the law on and in this world, as well as on God's. A typical example would be the following key generalization, applied to one of Austin's paradigmatic acts: The illocutionary act 'takes effect' in certain ways, as distinguished from producing consequences in the sense o f bringing about states of affairs in the 'normal' way, i.e. changes in the normal course of events. Thus Ί name this ship the Queen Elizabeth' has the effect o f naming or christening the ship; then certain subsequent acts such as referring to it as the Generalissimo Stalin will be out of order. 81
Students of narrative action-logic may well have spotted the hole in reasoning, which betrays a distinction overdrawn by wholesale compartmentalizing. True, an act does not subsume all the consequences it produces, but nor does it remain apart from all of them, far less to an equal extent. How else would (discourse) action sequences arise, let alone "well-made" plots, or for that matter, "logical" chains of statement and "closely reasoned" arguments? How else would we tell what follows strictly—along this or that sequence, in this or that context—what plausibly, what loosely or accidentally? Where does the Aristotelian "necessary" vary from the "probable" linkage and both from the "improbable" and "impossible," or propter hoc from post hoc? The strength of consequentiality makes the difference. Hence its gradability, or relativity, speaks against the absolute, atomizing disjunction of the illocutionary act from whatever it "produces." Between all and nothing, there is some, and "producing" covers a multitude of discriminate linkages. Thus, a statute's enactment not only envisages and intends but (teleo)logically entails what Austin would keep out, i.e. particular "consequences in the sense of bringing about states of affairs" and "changes in the normal course of events." This bringing about is exactly its raison d'etre and the inevitable result (or corollary or concomitant, we'll not split hairs) of its "taking effect": there arises a brave new world of social action, interaction, counteraction, along 82
the enacted lines, from the immediate discoursive Now onward. To put it in Austin's own metalanguage: a legislature can't possibly enact a stat-
81
Austin (1973 [1962]: 116). The terminus a quo can only range between the time of enactment, publication, or circulation among enforcers and a stipulated target date, later or, in retroactivity, earlier, all with stipulable termini ad quern as well.
If-Plots: Narrativity and the Law-Code
89
ute yet disclaim the "consequences" without inconsistency, any more than you can assert a proposition and deny its necessary implications. Details will soon follow, but observe that much the same applies to Austin's own favorite instance, as elsewhere. How to divorce "effect" from "consequence" at large, when naming the ship "the Queen Elizabeth" entails that referring to it by another name (e.g. "Generalissimo Stalin") will henceforth be a misnomer? And doesn't the entailment operate owing to the state of affairs generated or altered by the naming? In this case, as it happens, the attempted divorce would also leave this act remarkably thin, more so than usual on the speech-act approach. For one thing, because of Austin's argument in the foregoing paragraph. If the illocutionary act is to come off, "an effect must be achieved on the audience," namely, "the securing of uptake,"83 Discourse action hinges on its (re)cognition at the other, receiving end. But as the ship makes no audience—unlike the receiver of a statute or promise or warning—the "effect" of naming it would at best split between inanimate target and human crowd. Formally, the utterance casts neither of them as addressee, hence as cognizer to-be, far less responder: even the humans present merely attend rather than interact. Though deemed integral, uptake is yet left in the air. For another thing, the very favoring of this instance suggests rarefication: a performative act of naming that would at best rule a variant nomination out of order. It is all within and about linguistic (mis)behavior as such, on the narrow front of reference to a single thing, and actually purified of further referential implications at that. The narrowness compares with that of an act whereby a statute labels an object, or stipulates a term's meaning, or equates referent Ζ with Y by a legal fiction; except that such metarules are here, in the law-code, parts or preliminaries to wider and more direct statutory action upon the extralinguistic world outside the system. Those metalinguistic acts serve to determine (fix, newly orient) the bearing of the terms in question throughout the code, not just against the "vagueness" endemic to language and general terminology especially, but at times even against standard usage. Their effect doesn't end with the lexicon and its semantics, however. They legislate the mapping of the words concerned on the field of reference, hence on the substantive changes, enacted by first-order laws: what shall become henceforth of the 83
Ibid.: 115-16. And there's no genuine security without an ensuing acknowledgement ("I see," "Fine," "Done!" ...).
90
Meir Sternberg
object-type so (re)labeled or of the Υ + Ζ (re)grouping. Other comparable metalinguistic enactments are more directly and forcefully operative as verbal acts governing verbal action on pain of reaction. Thus laws made about language crimes: breach of promise, threat, libel, offensive speech, name-calling. Within ordinary legal rules, the antecedent operates in a similar way, on both the code, nominally, and the case, narratively. Defining (or, inversely, reading) "murder" as type-agency X will not stop there, either, now by force of the very "if murder, then ..." penal chrono-logic. Further, the term's (re)definition may not only contextualize but also itself imply or constitute or adjust the law at issue, teeth and all.84 That is why, even beside the dynamism of (re)naming as a prelude to transforming the name-bearer itself, the supposedly exemplary performative act of naming a ship per se appears so rarefied. As concerns the life of language, put your trust in logicians or philosophers and you'll lose touch with it. Taken together, the Austinian claim and exemplar bespeak an Ivory Tower. If the choice were between extremes, the antipole would be Aristotle's over-holistic yet under-communicative doctrine: action as a continuous enchainment of necessary or likely propter hoes, all anterior to verbal discourse. Finally, the consequences kept out by Austin supposedly transfer to the "perlocutionary act," performed through the utterance. But that act is marginal within the theory and, as handled there, even less relevant to our concerns, because of its definitional hopeless variability on every front. The perlocutionary act involves what happens to happen (or would desirably happen) when someone wants to get someone else to do, feel, believe, realize something. In the absence of any built-in perlocutionary force— means associated with the effect—all remains contingent, nothing rulegoverned. The "act" (i.e. impact) in view may transpire or not, success may come for reasons other than expected, or encoded, failure may range between nonresponse and counter-productiveness: a well-meant promise engenders alarm, say, instead of reassurance. Unlike the illocutionary aspect, which is relatively determinate (as a promise, advice, request, etc.) and, if anything, overdetermined by the theory, here lurk together pellmell "all those effects, intended or unintended, often indeterminate, that 85
some particular utterance in some particular situation may cause." That we, humanly ignorant, often cannot even tell whether any effects have 84 85
E.g. Bhatia (1994: 1 3 9 ^ 0 ) . Levinson (1983: 237).
If-Plots: Narrativity and the Law-Code
91
been caused, and which, radicalizes the polarity to the attested and traceable impact of the statute as legislative speech.
6.3. Constitutive Dynamism: Word, World, and Mind under the Legislator's Act This impact consists primarily in objective change, the remaking of the world once the law takes effect. Other commands than the law's, including its nearest deontic relatives, merely aim for such a change. The directive words enacted by the legislator, however, do bring about a shift in normative ontology, willy-nilly, at times a reversal. To illustrate from the effects of a historic statutory discourse-event, the world governed by the Ten Commandments—or life in it under them—is nothing like the same as earlier. It has grown different, all at once, in existential and psychological and sociomoral coordinates, in the retributions met and the futures imagined (or chosen) and the sense of the present alike, in the self s public image and solidarity as in its private identity, in the shapes forced and lines drawn upon brute reality, with their interior representations. Within or without the mind, these shapes and lines determine what agency now counts as a Do, what as a Don't, what as a Doable, and of what exact type of doing, complete with the specific if s. Further, they not merely classify and regulate what we do or avoid or plan anyway but often also generate types of activity that would otherwise remain unperformable, because unborn as well as unbound, simply nonexistent (e.g. paying or evading taxes). The law then constitutes, in the enactment, the world that it represents in action within the limits imposed. Its generative force accordingly extends from the discourse to the very object discoursed about: from the telling/reading to the happening told, from thick interpretability to God-like creativity. Indeed, if law-making extends to worldmaking in this fundamental sense, then so does the analogy to the Creator ex nihilo by the power of the word and the rules (deepest-seated, inviolable "laws") whereby he binds the created: only with a target-switch from nature to culture. God himself switches arenas thereafter, when speaking as cultural legislator, first humanity's, then the Chosen People's. Heavenly or earthly, the legislative typology generates even basic concepts and entities, doings and dealings, often taken for natural, self-evident within the culture that has absorbed them, unless defamiliarized by an alternative segmentation. (Take an act so well-defined in our eyes, apparently inde-
92
Meir Sternberg
pendent of enactment, as murder. On the Icelandic Saga's code, for example, murder is undeclared killing, regardless of intentionality. By legal fictions, nature itself may reverse: a woman will count as a man, say, or a stranger displace a parent. 86 ) But then, every legislator would have the code naturalized among the people, rather than defamiliarized into the artificial system it constitutes. Therefore, while other prospectives and deontics (e.g. wish, fear, scenario, military command) represent actions not yet realized in the world, if ever, the act of law-enactment gives such actions a unique objective reality, often subjective and immediate, too—even before their literal realization by anyone. It maps them on the world as the things done or not done or eligible according to the code under which everyone lives, as binding agentive forces and rules and models in an all-embracing public sphere constituted by their joint directives. Uncannily, the action is there (at work on and within and around us) even when it isn't (in evidence, on history's record or the police's or the court's or our own). It joins the parameters that make up "Such is existence" and include natural "If ... then's" of corresponding generality. Imagining a literally lawless culture, such as Homer's Cyclops land, will help to estrange the difference that lawgovernedness makes, and its evolution modifies, even to the habitual lawbreaker. Moreover, the dynamic force of the statute's enactment (a fortiori, of the whole code) outreaches this constitutive and regulative effect on existence, as well as the I f s consequent within the enacted law-world. Having gone through the virtual and the existential domains, that force turns operational. It also bears consequences for the real world, where the words get applied by legal practitioners: interpreted with a view to the matching of the " i f ' and the actualizing of the "then" in due case-specific response. Whether the application is right or wrong doesn't affect the if-plot's further consequentiality beyond the hypothetical chain: its extension to another life by virtue of the discourse that codifies it into a web of narrative
86
The ado made by Tristram Shandy, in one of his best comic scenes, about whether the law regards the father and/or mother as being "of kin" to their child is not just a laughing matter, nor a quaint anachronism. In America of the 1980s, an appellant who had fathered a daughter on a married woman demanded visitation rights after she returned to her husband, and the Supreme Court denied him not merely any such rights but the very claim to paternity. "[T]he law, like nature," Justice Scalia wrote, "recognizes only one father"—here, to the a priori exclusion of the natural in favor of the legal-cultural one (cited in Amsterdam and Bruner [2000: 77ff.]; Bruner [2002:40]).
If-Plots: Narrativity and the Law-Code
93
difference. Either way, the taxonomy proceeds to govern the counterdoings of law enforcement. Who will be indicted and who spared a trial, who will live and who will die, who will go to prison and who slip through the net and who get a medal—or, historically, who in fact encountered the respective treatments—for the commission or omission of what would look to a Martian the same act: indeed, of what might yesterday count as the same, or still does elsewhere. And so forth, with evergrowing refinement and always in relation to the culture's, or to each member's, extralegal ontologies, not least the sense of justice. To the extent that our language system conditions our reality picture, hence our experience and behavior, this is due partly to the value-and-action-laden conceptual networks that it overlies, assimilates, inculcates, mirrors, triggers in verbal form. As world-makers, the linguistic and the legal codes thus interdepend, down to their evolutionary dynamics. In evolution, how public stands to private remodeling under the new statute also grows clearer by analogy to social vs. interior language. Ideally, the law having come into effect—a tell-tale phrase—state of being and state of mind undergo change together throughout its jurisdiction. The ideal of twinned, outer/inner development via law-speaking, and its normal achievement, more or less, should be obvious enough, once you think of it by reference to the twofold linguistic contract. Here, the Ten Commandments (Exodus 20: 1-14) again offer a paradigmatic example, if only because willingly and vocally accepted by the whole Israelite audience at Sinai as the basis of its covenant with God. (Also because they concern matters of revolutionary weight, of life and death, of ideonational—thereafter, still wider—importance, and of relevance to the entire law-code.) "All that the Lord has spoken we will do": thus the unanimous Israelite self-commitment before and after the public law-speaking (ibid.: 19: 8, 24: 7). In opting for the covenant set before them, the people undertake to adjust lifestyle and mindscape alike to the Do's and Don't's codified there: to incorporate into both realms a web of normative terms (e.g. monotheism, Sabbath-keeping) unfamiliar to them as to the rest of humanity or forgotten during the centuries of Egyptian servitude. Then, in the law-speaking, the rhetoric (e.g. the I-to-thou address, the plain language, the backreference to the miraculous Exodus) would clinch this acceptance, channel the mental self-enforcement. The last Commandment even directly regulates internals. "Thou shalt not covet ..." foregrounds the emotion as a negative on its own, apart from its liability to cause social misbehavior. A telling measure of how, for the
94
Meir Sternberg
voluntary "thou"-party, observance begins in the will to observe, though it is the actual doing that ultimately weighs most. (Nothing like the Pauline inverse balance of "faith" vs. "works.") Compare the saga's Icelanders having their foundational code recited to them at the Althing, and thus implicitly reaffirmed, in a three-year cycle. As the command of a language means internalizing its commands to yourself, so does the command of the law. Yet the principle of the law's world-making, a fortiori the law-code's, evidently holds in less harmonious cases (and outer/inner disharmonies overtake the Sinai covenant itself before long). Even if unwelcome or unjust, imposed in the worst sense, the change of state will transpire notwithstanding, given that the enactment affects its subjects' being and living regardless of their subjective feeling. Also regardless of subjective awareness. Ignorance of the law will not excuse its breach, any more than a plea of one's linguistic incompetence cancels, let alone justifies, a solecism. Nor will a plea of misunderstanding help those familiar with a statute to avert its application and consequences. If anything, the law presumes all subjects/addressees to be competent (as well as innocent) and its modern rhetoric will accordingly lurk deeper and aim higher than the Bible's. When we turn from the ideal, presumed to the actual state and from contingent to essential workings, then, the "receiving" end of the lawdiscourse receives not necessarily the words but their objective effect on its world and itself: whether or (as usual) not a genuine partner to the act, the public is always a patient of the enacted action. The subject's mind as good as remains beside the legislative point at any juncture, its variables immaterial to the tale of mutability between junctures. This is worth emphasizing as another principled disproof and corrective to the Austinian building of interiority into all performance, with extra and elaborated mental strings (emotive, cognitive, ethical) attached to 87
specific performance types in the name of "felicity." 0 ' Thus the allinclusive demand for "illocutionary uptake": the successful performance of a language act would thereby hinge on the addressee's identifying and generally understanding it. No communicative doing performed in trans87
This recalls the inward turn of literary modernism, whereby "felicitous" (artistic, tellable) narrative got conditioned, or even centered, on the characters' subjective life: in Jamesian poetics, most articulately. Within recent narratology, sequels include definitions of narrativity as "experientiality" (Fludernik [1996]) or goal-pursuit (Ryan [1991]). These interior strings are further discussed in Sternberg (1978: 276ff.) and (2003b: 58Iff.), (2005), respectively.
If-Plots: Narrativity and the Law-Code
95
mission, as it were, without the appropriate cognitive doing in response. But nor is the transmitter himself spared mental baggage; rather, it piles up on his side. For felicity, both the "sincerity" and the "essential" conditions load each would-be speech performer with some assortment of feelings, beliefs, designs, undertakings judged integral to his act: he must think his statement valid, his promise reliable, his advice good, and so forth. Actually, all discoursive doers travel lighter, free of such private must's, though bound to what they have manifestly done in and through the (assertive, promissory, advisory) speaking. The internal extras piled on them are unrealistic in the human condition, hence even less reasonable as must's than the attendant institutional, procedural felicities have already shown themselves to be vis-a-vis a superhuman law-and-universe-maker, or his equivalent on earth. Funnily, Austin et al. in effect treat human (interaction as if it were a novelistic affair conducted between two God-like 88
mind-readers and witnessed as well as analyzed by fellow omniscients. An ironic nemesis, this blindness, considering that Austin deems literary fiction "parasitic" upon ordinary language. The whole legal system—the law's force and coming into force along with all the stages and roles of law enforcement—provides the ultimate counterexample to these fanciful must's. The very examples drawn from it by speech-act analysts, or their following among law students, boomerang. "If X makes a statement which he knows or believes to be untrue, he thereby perpetrates the abuse that we refer to as lying or prevarication; 90 and, if he does so on oath in a court if law, he commits perjury." Now, X himself may "know" it to be "untrue" from within, but how do "we" outsiders and the court know? Perjury unknown, or even suspected yet undemonstrated, let alone undemonstrable, doesn't count as perjury but as an
88
Just as funnily, elsewhere in pragmatics and related fields, the discourse subject's mind is all too often conspicuous for its absence, as if the choice were between extremes: secret vs. social life. For details, within a theory that aims to integrate perspective with plot, see Sternberg (2005) and references there. Below we'll see also h o w the telling in the legal system offers another interesting example of these flexible relations, with law89 enactment as limit case. I illustrate here only from experts in the theory, rather than from its misapplication to law by amateurs, as in Bruner (2002: 2 4 - 2 5 ) . Compare Gorman ( 1 9 9 9 ) on its abuses in literary study. 90 Lyons (1977: 734).
96
Meir Sternberg 91
act of genuine testimony. Getting the speech act wrong, moreover, involves a threefold perspectival confusion: (1) between the moralistic (if you will, deontic) and the epistemic viewpoint; (2) between omniscience and earthly restrictedness; (3) between knowledgeable self-view and the fallible insideview of another's self. Inversely, if a testimony counts in law as perjury, it is only after the witnessing event and from the outside even so: the act hinges not on what the convicted perjurer knew or believed to be true, but on what the jury/judge declares him to have known in the light of the evidence, itself always possibly false or misunderstood. So the declared perjurer might in fact have told the truth, after all, to the letter or to the best of his knowledge, just as the credible, altogether unsuspected witness might have lied in his teeth: too bad, if nevertheless decided otherwise within the rules of the game. This act turns, in short, on an outsider's rule-governed, public, binding epistemic judgment of a private epistemic state—and all by a decision procedure between tellings, or readings, unique to the legal system. The more unique, in that (unlike fiction or inspired history) the adjudicator enjoys and exercises the authority of mind-reading by force of enactment, without pretending to any supernatural omniscience and yet with instant necessary consequences for real life. Likewise, only worse, is illustrating the breach of "sincerity," with the stigma of "abuse" attached, from a juror who "finds the defendant guilty 92
when he knows him to be innocent." Here the juror himself cannot possibly "know," from without, the other's innocence, any more than we observers (e.g. law-officers) can gain 93 access to his own alleged inside knowledge and insincere performance. Similarly with the parties to a 94
trial in the Icelandic saga's world, all indifferent to homicidal intent; or with the latter-day prosecutor, who needn't always establish a motive for 91
92
A s one cynic generalizes: "It isn't a crime if nobody knows about it. It never becomes a crime until someone finds out" (Higgins [2000: 354])—"and proves it," he might add. Here, the silent if-midplot again enters and conditions the process. Levinson (1983: 230). Happily, therefore, the question of what might be done if w e could fails to arise. On pain o f offending against language itself, no earthly observer involved possibly "knows," given that this verb (studied in Sternberg [2001a]) presupposes, and would grant the subject, the very truth that is at issue during the trial. Even the adjudicator never knows in this absolute sense, being humanly limited, fallible, nonomniscient like the rest. By the system's dispensation, as already indicated, he only pronounces the legal truth for now, reversible in the appellate court or in the light of fresh evidence. On the typology, or "semantic field," o f homicide, see Maley (1985).
If-Plots: Narrativity and the Law-Code
97
the alleged offense, though supplying such a cause would help to fill out his story. And even where the defendant's inner state makes a vital difference, it remains subject to the law-officers' natural epistemic limitations along with the law's artificial and culture-specific operations: the case of the trial for perjury extends to, say, malice aforethought, deliberateness {mens rea) at large, criminal knowledge, or the Bible's "coveting," unless judged by God himself. (Recall also the foregoing comment on the epistemic modality that variously qualifies the guilt/innocence tales narrated and opposed and assessed during the judicial process.) No wonder, then, that this rationale is also, indeed most, in evidence where it all begins. The law-code fits and exceeds, perhaps inspires, the rest of the system's objective modus operandi: each statute does its world(re)modeling, complete with the various implications for law and life thereafter, independently of the mental state of either the transmitting or the receiving party. This privileging of sheer discourse agency at the expense of the discourser's interiority is likely to appear outrageous to many, far beyond the law and/or language experts who have endeavored to redress the balance, ideally into a meeting of minds. It goes against literary critical articles of faith (whether communication or interpretation, the primacy of one's own reading or the character's subjective experience), just as it runs against common ideas of fairness, justice, intentionality, authenticity, free agency or, once more, speech ethics and decorums. Thus Kurzon, following Searle, attaches to the legislative command the sincerity condition "S[speaker] wants H[hearer] to do A." 9 5 On top of the unearthly, mind-reading epistemology assumed, as before, the demand hardly makes sense in its own institutional terms. For all we know, the legislature enacted A not because they wanted it done but under democratic pressure: from the media, the government, the judiciary, public opinion, an interested lobby, their constituents or spouses—all within the rules of the game and without detriment to the effective outcome. (Not to mention unholy, and still less knowable, motives and forms of duress.) We simply have to accept another rationale of discourse interaction, here visible at its purest, and adjust our theories accordingly. As with the transmitter, so with the receiver: the law enactment imposes itself on the world in disregard for the subjects' knowledge, understanding or opinion of it. The legislative register has always been notorious for its compounding of esoteric opacities: archaic, specialized, heter95
Kurzon (1986: 8); Searle (1969: 6 4 - 6 6 ) , also (1979: 14).
98
Meir Sternberg
olingual, circumlocutory terms, intricate syntax, lust for overwriting, tacit or explicit cross-reference. As a Report on the Preparation of Legislation in Britain describes it: "To the ordinary citizen the provisions in the statute book might sometimes be as well written in a foreign language for all the help he may expect to obtain there as to his rights and duties under the law. And this in an age [...] when the statute law has a growing effect on practically every sphere of daily life." 96 This, more remarkably yet, when the statute addresses as well as affects the ordinary citizen who finds it unintelligible, or would if alive to it. He fares much like the named ship, only worse, given his humanity and speech role vis-ä-vis the enacting addressor. He finds another parallel, older and deeper, in nature being created (with eternal law-like regularities imposed on each sphere) by word after divine word. This includes the human creature himself, though singled out as the addressee of God's fiats, where all nonhumanity had been reduced to third-person targets. However favored with verbal contact, Adam had no say in the matter, nor any genuine understanding of the Do's and Don't spoken to him. What could the poor innocent make of the order forbidding him to eat of the Tree of "Knowledge," on pain of "death," when the very words outreach his mental range? For his nonspecialist descendants, the common must of publishing the law, if not a hollow gesture, is similarly charged with irony. Still nearer to home is the nightmare that haunts the modern imagination. The paradox of the unknowing subject in life deemed knowledgeable in law may well appear Kafkaesque, only that Kafka carries our sociolegal reality to its logical extreme. In The Trial (1925), K. is literally denied access to law books and kept ignorant of everything, from the charge against him to the proceedings to the verdict to the death sentence. To crown it all, the action-first logic that holds for and between the two discourse participants stretches to the one acting and acted upon at once. Unlike the typical ("you"-pressing) deontic speaker, every legislator generally doubles as self-addressor, subject to what his vote has helped to enact. But, though doubly involved and presumed competent, he needn't (sometimes couldn't or just doesn't) master the enactment in either role. Understanding is dispensable again. The vote itself counts as uptake cum assent; the address to the public, with the self-address, as binding per se, 96
Quoted in Maley (1987: 25). Here lies the apparent paradox of Stendhal's disclosure that, while writing La Chartreuse de Parme, he took his cue from the Civil Code, of all styles, and with a view to directness, of all qualities. He read two or three pages of it every morning, "afin d'etre toujours naturel."
If-Plots: Narrativity and the Law-Code
99
without so much as a comparable behavioristic response by a show of hands. 97 The ostensible inverse, predictably rare, comes to the same thing. Far from esoteric, the Bible strives for nation-wide "foolproof communication" in historical and law-making narrative alike. 98 Hence the code's simple language, often deceptively simple but always accessible to all comers, relative to the "foreign" parlance of the modern guild. Of the two subgenres, the legal also flaunts its audience-mindedness. The God-given code is relayed by Moses directly to the assembled people. It is often explicitly phrased to address them ("thou" or "you" substitutes for the usual third-person agency, whence the Rabbinic term Do/Don't). No less uncommon, it is vocally accepted by them, as a solemn bond with the lawgiver, and emphatically presented as an object of reception, study, memory throughout the culture. "This book of the law shall not depart out of thy mouth, but you shall meditate on it day and night" (Joshua 1: 8) is itself a recurrent command issued to leader and laity alike, nation-wide. Amid all this exceptionality, which befits a covenant, the law yet applies without regard to its legibility or to the actual sense made of it by an individual. An Israelite thus suffers death for violating the Commandment of Sabbath-keeping in a manner that was never specifically forbidden before and that Moses, the intermediate law-speaker, must himself refer to God after the event, namely, wood gathering (Numbers 15: 32-36). The enactment from heaven that ensues, along with its fatal enforcement, hov99 ers therefore between specific elucidation and retrolegislation. Even so, 97
98
99
Hart (1994 [1961]) would in effect except the legal (especially judicial) system itself from the laity's predicament: he assigns to its members, controversially, an "internal" viewpoint and "rules of recognition." Even so, what insider knows and understands, never mind approves, the whole law-code? Only God the Law-giver cum supreme judge and the Law-Speaker in the saga world, founded on a rudimentary code. (The question indeed gains dramatic point in both narratives.) The law's world-making, then, doesn't quite depend on the jurists, either, except perhaps as a collective of minds and masteries that among them work out how it applies. Sternberg (1985: 48ff. and Index), (1991), (1998: 509-629 passim). Cf. note 96 on Stendhal above. Such hovering abounds in earthly equivalents, given the fuzzy line that divides the judiciary's reading from its making (in effect, antedating, with retroactive application) of the law. In the sequel to the Bible's law itself, examples on record proliferate as early as ancient Rabbinic legaliterature. There, we often encounter a specific case followed by a general or easily generalized ruling that appeals to divine authority—much like the history-to-statute pattern, with the latter's elusive status, in the Wood Gatherer episode.
100
Meir Sternberg
the Wood Gatherer should have known better, as it were, and pays with his life for his poor or permissive reading. The very exception, then, ultimately proves the law-code's outer/inner balance of power. When it comes to essentials, or to a choice, the operation of the enacted on the exterior world outweighs its reception within the secret life of the addressee cum prospective agent. Though attachable to statutory discourse, the mental strings for which theorists press in the name of consensus, morality, democracy, or felicity needn't attach to it, nor indeed do they in standard practice over the ages. Approval, accessibility, sheer awareness: each and all together remain dispensable for once. Not even the barest form of response associated with the audience's illocutionary "uptake" is obligatory in this discourse framework, which carries to the limit the rule that a social act goes by its public effect, or what I called the presumption of competence. Law enforcement itself stops short of this extreme, prior to the end at least. The sentence and the verdict do remain as indifferent to the defendant's comprehension as the scenario, though fortunately easy to comprehend. In Oliver Twist (1837-39), Fagin is too dazed to make out the judgment pronounced at his trial, but he has been sentenced and will hang, regardless. The foregoing stages of the legal process, however, variously provide for mutual intelligibility and balanced or reciprocal speech activity. At these phases, the insiders are supposed not only to know the law and its proceedings but also to make them known to outsiders presumed ignorant. More so, in fact, than to one another.100 And the turn toward sharing the esoteric knowledge means that such action requires cognition in verbal performance, not least an informed viewpoint on the otherwise autonomous enactment: the network, rules, and choices peculiar to the law-world. For example, just as a bet needs to be ratified through "Done!" from the taker—another thematic shorthand—so with the arresting and interrogating police officer's need for a vocal "Yes": under the Miranda regulations, they must ensure the suspect's (including a fellow professional's) express awareness of his rights to silence and counsel. As a matter of duty, in turn, the defendant's counsel will spell out the options before the
100
Yet none the less so when the insider counts as outsider. Hence the saying that the lawyer w h o represents himself has a fool for a client: the presumption of competence varies with the legal role.
If-Plots: Narrativity and the Law-Code
101
trial and provide a running commentary on it. Other participants also initiate, receive, exchange guidance. Thus the judge making sure that the witness understands the cross-examiner's question (as well as vice versa with the answer) or explaining to the jury the relevant points of law, the issues in dispute, the onus of proof, the benefit of doubt, and the weight of the evidence. Not that everything in the process becomes transparent to everyone (as ongoing attacks and bids for reform indicate), but that communication grows far less one-sided, storytelling more interactive, relative to the discourse of sovereignty unlimited. The relative distance widens in literary practice, especially in fiction. Here, the statute occupies the polar extreme to narratives, legal as otherwise, that value subjectivity and can exploit the license of omniscience for mind-reading at will. Applied to court scenes, this license authorizes inside views of everyone concerned: of the defendant's motives then and tension now, the witness's hidden bias, the prosecutor's will to win, the interlocutors at cross purposes, for example, all opposable to the fafade of words spoken and related behavior on view. Legaliterature, from trial dramas to crime stories to detective tales, also richly plays off the law's private against its public bearings, or the sense of justice against the codified letter, and times of change in either assume a special interest there. Across all such interplays within arenas that disclose, prize, and empower the secret life, however, the newly enacted sociolegal ontology will continue in force until replaced by a newer one to establish (or extend) a multiphase shift on the ground: plotted on, or through, existence itself, traceable along history. 7. Conclusion Accordingly, whatever the correspondence between lawscape and mindscape, that ontic plot exhibits a definite likeness (on top of a generative linkage) to the if-plot. Just as the action represented by the law has its before and after, earlier and later —offense, then response—so has the act of law-discoursing about that action: it kinetically, chrono-logically, intervenes between the pre-law and post-law states, only without if s and but's. Recall how even a command so undetailed and uneventful-looking as that to keep the Sabbath holy reaches back to two anterior states at once, in forcing their inversion from nonobservance to observance. The law-telling address evokes by its novelty the pre-Commandment, Sabbath-less existence of the addressed "thou," when he was free to do as he pleased
102
Meir Sternberg
throughout the week; the told scenario evokes the free bustle of the preSabbatical week by way of another opposition-and-change, now to the "thou"-agent's God-like rest on the seventh day. Likewise, only more visibly, with the Magna Carta's "In future, ..." opening in (lb). It looks back to pre-1215 official tyrannies, henceforth outlawed, as well as to preindictment suspicious doings that are met by the law-enforcer's counterdoing in the form of inquiry. So the enunciated if-plot coextends with the enunciatory plot that constitutes it. The two inherent movements reinforce each other, thickening the overall narrative dynamics from different and to some extent opposite quarters. Not only does the action told on the surface mix with the teller's ipso facto enactment of a new order (pun intended), but the former's hypothetical, recurrent, and generally extensive time-sequence also merges with the latter's actual, one-off, and instant consequentiality. Unlike the if-plot, under whatever guise, the act of law-plotting has no Suppose's about it, for a change, no strings inherently attached to its effect of change. 101 Crime X shall incur punishment Y (and/or Z) within a world already remade by fiat into the appropriate penal ontology. On this elusive disparity amid twinship, our paradigm has further light to offer. Note how the Commandments addressed by God the legislator in Exodus recall and supplement the commands ("Let there be light" and so forth) whereby the Creator made the universe in Genesis. By the power of the directive Word, two worlds spring into existence ex nihilo·. first, that of "nature" (so-called, because itself created), then that of "culture," expected to become second nature for those who live under the covenant. The terms of their being and doing have got revolutionized by mutual agreement, newly created in effect; and whether they will henceforth live up to the covenant determines their fortunes, to the bitter end. There, Jerusalem's culminating offense against a single ideonational law, that of the Hebrew slave, exposes the people "to the sword, to pestilence, and to fam101
Not even the principled liability to legislative amendment or repeal and judicial review after the act. Unless and until that happens, the statute keeps its force. Likewise with procedural conditions, o f the kind Austin et al. so multiply. In fact, these strings are only attachable to law-making by a special meta-law, and are always detachable by humans as well as by a God. Likewise, again, with the in-between option: postdating a statutory act both needs to be explicitly stipulated and finds its converse in retroactivity. Nor is the latter so uncommon as may appear or as its perpetrators would have it appear: the judiciary when daily making new law in the guise of interpreting the old with authority, or, less often, the legislature when anticipating an event currently unrealizable. (See notes 61 and 99 above.)
If-Plots: Narrativity and the Law-Code
103
ine," to loss of sovereignty and mass deportation: enslavement for enslavement. The Sinai covenant itself is declared broken, and with it the 102 framework of existence, pending a new dispensation (Jeremiah 34). So the code's logic of retribution is no less metamorphic, and its workings no less determinative, than the physical laws of nature. (If anything, the physics of the created world can be visibly suspended in the interest of its fellow code's making or administration.) And what happens under an omnipotent God, for better or worse, serves as usual to defamiliarize our man-made equivalents. Among the equivalents themselves, landmarks such as the Magna Carta and the American Constitution similarly help to foreground the undramatic yet definite workings of more ordinary legislative acts and the cumulative difference they make. Finally, this objective change, regardless of the subjects concerned, implies a further analogy between the law-code and the entire narrative genre (or for that matter, discourse, literature, or art) as a system in time. Not that the analogy would be welcome to all, for reasons already mentioned. Generalizing the presumption of systemic competence to embrace narrative at large, its diachronic movement included, would go against various literary critical values and vogues. Nowadays, it would, above all, risk the charge of elitism or downright tyranny: as though self-constituted authorities presumed to meddle with the individual reader's liberty. A strong counterargument is the family likeness to the indisputably normative systems above, as in how a new arrival (whether a statute or otherwise storied) modifies the entire picture, from the code itself to the world around it. For short, just recall T. S. Eliot in "Tradition and the Individual Talent" (1919): What happens when a new work of art is created is something that happens simultaneously to all the works of art which preceded it. The existing monuments form an ideal order among themselves, which is modified by the introduction of the new (the really new) work of art among them. The existing order is complete before the new work arrives; for order to persist after the supervention of novelty, the whole existing order must be, if ever so slightly, altered; and so the relations, proportions, values of each work of art toward the whole are adjusted. 103
The reference to "what happens when ..." applies with particular aptness to the genre defined by happening and prefigures Austin's idiom of doing. At any rate, the extralegal order (and suborders) will again change
102 103
Sternberg (1998: 629-38). Eliot (1961: 15).
104
Meir Sternberg
irrespective of your knowledge; but you'd better know, if you aspire to competence or, negatively, to self-defense against the penalties of ignorance. In highlighting a narrative's effects on the narrative system and universe, as an additional dynamic force, the law points to yet another moral for narratology.
References Abbott, H. Porter 2002 The Cambridge Introduction to Narrative (Cambridge: Cambridge UP). Alt, Albrecht 1989 [1934] "The Origins of Israelite Law," in Essays on Old Testament History and Religion, translated by R. A. Wilson, 81-132 (Sheffield: JSOT Press). Amsterdam, Anthony G. / Bruner, Jerome 2000 Minding the Law (Cambridge, Mass: Harvard UP). Austin, John 1973 [1962] How To Do Things With Words, edited by J. O. Urmson (New York: Oxford UP). 1970 Philosophical Papers (Oxford: Oxford UP). Bhatia, Vijay 1994 "Cognitive Structuring in Legislative Provisions," in Language and the Law, edited by John Gibbons, 136-55 (London: Longman). Bordwell, David 1985 Narration in the Fiction Film (Madison: University of Wisconsin Press). 2002 "Film Futures," in SubStance 31: 88-104. 2004 "Neo-Structuralist Narratology and the Functions of Filmic Storytelling," in Narrative across Media: The Languages of Storytelling, edited by Marie-Laure Ryan. Frontiers of Narrative, 203-19 (Lincoln and London: University of Nebraska Press). Brooks, Peter 2005 "Narrative in and of the Law," in A Companion to Narrative Theory, edited by James Phelan and Peter Rabinowitz, 415-26 (Oxford: Blackwell). Brooks, Peter/Gewirtz, Paul (eds.) 1996 Law's Stories: Narrative and Rhetoric in the Law (New Haven: Yale UP). Bruner, Jerome 2002 Making Stories: Law, Literature, Life (Cambridge, Mass: Harvard UP). Coode, George 1973 [1843] "On Legislative Expression," in Drafting, edited by S. Robinson, 335-98 (Sydney: Butterworth). Cover, R. M. 1983 "Nomos and Narrative," in Harvard Law Review 97: 4-68. Danet, Brenda 1980 "Language in the Legal Process," in Law and Society Review 14: 445-564.
If-Plots: Narrativity and the Law-Code 1985
105
"Legal Discourse," in A Handbook of Discourse Analysis, edited by Teun van Dijk, vol. 1: 273-91 (London, etc.: Academic Press). 1990 "Language and Law: An Overview of Fifteen Years of Research," in Handbook of Language and Social Psychology, edited by Howard Giles and W. Peter Robinson, 537-59 (Chichester: Wiley). Dworkin, Ronald M. 1986 The Law's Empire (Cambridge, Mass: Harvard UP). Eliot, T. S. 1961 Selected Essays (London: Faber and Faber). Endicott, Timothy Andrew Orvville 2000 Vagueness in Law (Oxford: Oxford UP). Fludemik, Monika 1993 The Fictions ofLanguage and the Languages of Fiction (London: Routledge). 1996 Towards a 'Natural' Narratology (London and New York: Routledge). Fludernik, Monika (ed.) 1994 Second-Person Narratives. Special Issue of Style 28.3. Fuller, Lon L. 1967 Legal Fictions (Stanford: California UP). Gibbons, John 1999 "Language and the Law," in Annual Review of Applied Linguistics 1 9 : 1 5 6 - 7 2 . 2003 Forensic Linguistics: An Introduction to Language in the Justice System (Oxford: Blackwell). Gibbons, John (ed.) 1994 Language and the Law (London: Longman). Gorman, David 1999 "The Use and Abuse of Speech-Act Theory in Criticism," in Poetics Today 20: 93-119. Hare, Cyril 1957 An English Murder (Harmondsworth: Penguin). Hart, H. L. A. 1994 [1961] The Concept of Law (Oxford: Clarendon Press). Higgins, George V. 2000 The End of Day (New York: Harcourt). Kafalenos, Emma 2005 "Effects of Sequence, Embedding, and Ekphrasis in Poe's 'The Oval Portrait'," in A Companion to Narrative Theory, edited by James Phelan and Peter Rabinowitz, 253-68 (Oxford: Blackwell). Kelsen, Hans 1949 General Theory of Law and State, translated by Anders Wedberg (Cambridge, Mass: Harvard UP). Kurzon, Dennis 1986 "It Is Hereby Performed . . . " in D. K.: Explorations in Legal Speech Acts (Amsterdam: John Benjamins). Levinson, Stephen C. 1983 Pragmatics (Cambridge: Cambridge UP).
106
Meir Sternberg
Lyons, John 1977 Semantics (Cambridge: Cambridge UP). Maley, Yon 1985 "The Semantic Field of Homicide," in Advances in Discourse Processes, edited by James D. Benson and William S. Greaves, XVI: 152-68 (Norwood, NJ: Ablex). 1987 "The Language of Legislation," in Language and Society 16: 25-48. 1994 "The Language of the Law," in Language and the Law, edited by John Gibbons, 11-50 (London: Longman). Minow, Martha 1996 "Stories in Law," in Law's Stories: Narrative and Rhetoric in the Law, edited by Peter Brooks and Paul Gewirtz (New Haven: Yale UP). Nimmer, David 2001 "Back from the Future: A Proleptic Review of the Digital Millennium Copyright Act," in Berkeley Technological Law Journal 16: Supplement. Olson, Elder 1952 "William Empson, Contemporary Criticism, and Poetic Diction," in Critics and Criticism, edited by Ronald S. Crane, 45-82 (Chicago: University of Chicago Press). Paul, Shalom M. 1970 Studies in the Book of the Covenant (Leiden: Brill). Pfersmann, Otto 1994 "Temporalite et conditionnalW des systemes juridiques," in Revue de la Recherche Juridique 19: 221-43. Prince, Gerald 1982 Narratology: The Form and Function of Narrative. Janua Linguarum, Series Maior 108 (Berlin, etc.: Mouton). 1988 "The Disnarrated," in Style 22: 1-8. Ryan, Marie-Laure 1991 Possible Worlds, Artificial Intelligence, and Narrative Theory (Bloomington: Indiana UP). Sanford, David H. 1992 If P, then Q: Conditionals and the Foundations of Reasoning (London: Routledge). Searle, John 1969 Speech Acts: An Essay in the Philosophy of Language (Cambridge: Cambridge UP). 1979 Expression and Meaning (Cambridge: Cambridge UP). Shklovsky, Victor 1991 [1970] The Bow-String: Three Excerpts, translated by Benjamin Sher (http : / / w w w .websher.net/srl/bow.html). Sonnet, Jean-Pierre 1997 The Book within the Book: Writing in Deuteronomy (Leiden: Brill). Sternberg, Meir 1978 Expositional Modes and Temporal Ordering in Fiction (Baltimore and London: Johns Hopkins UP).
If-Plots: Narrativity and the Law-Code 1979
107
"Between the Truth and the Whole Truth in Biblical Narrative: The Reordering of Inner Life by Telescoped Inside View and Interior Monologue," in Hasifrut 29: 110-46. 1982 "Proteus in Quotation-Land: Mimesis and the Forms of Reported Discourse," in Poetics Today 3: 107-56. 1985 The Poetics of Biblical Narrative: Ideological Literature and the Drama of Reading (Bloomington: Indiana UP). 1986 "The World from the Addressee's Viewpoint: Reception as Representation, Dialogue as Monologue," in Style 20: 295-318. 1990a "Telling in Time (I): Chronology and Narrative Theory," in Poetics Today 11: 901-48. 1990b "Time and Space in Biblical (Hi)storytelling: The Grand Chronology," in The Book and the Text: The Bible and Literary Theory, edited by Regina Schwartz, 81-145 (Oxford: Blackwell). 1990c "Time and Reader," in The Uses of Adversity: Failure and Accommodation in Reader Response, edited by Ellen Spolsky, 49-89 (Lewisburg, PA: Bucknell UP). 1992a "Telling in Time (II): Chronology, Teleology, Narrativity," in Poetics Today 13: 463-541. 1992b "Biblical Poetics and Sexual Politics: From Reading to Counterreading," in Journal of Biblical Literature, 111: 463-88. 1998 Hebrews between Cultures: Group Portraits and National Literature (Bloomington: Indiana UP). 2001a "Factives and Perspectives: Making Sense of Presupposition as Exemplary Inference," in Poetics Today 22: 129-244. 2001b "Why Narrativity Makes a Difference," in Narrative 9: 115-22. 2003a "Universale of Narrative and Their Cognitivist Fortunes (I)," in Poetics Today 24: 297-395. 2003b "Universale of Narrative and Their Cognitivist Fortunes (II)," in Poetics Today 24: 517-638. 2005 "Self-Consciousness as a Narrative Feature and Force: Tellers vs. Informants in Generic Design," in A Companion to Narrative Theory, edited by James Phelan and Peter Rabinowitz, 232-52 (Oxford: Blackwell). Vaihinger, H. 1924 The Philosophy of 'As i f , translated by C. K. Ogden (New York: Harcourt). van Dijk, Teun (ed.) 1985 Handbook of Discourse Analysis, vol. 1 (London: Academic Press). Wyndham, John 1959 The Seeds of Time (Harmondsworth: Penguin). Yacobi, Tamar 1995 "Pictorial Models and Narrative Ekphrasis," in Poetics Today 16: 599-649. 2000 "Interart Narrative: (Un)Reliability and Ekphrasis," in Poetics Today 21: 4 0 8 47.
JOHN PIER (Tours, Paris)
After this, therefore because of this Perhaps no concept in the study of narrative casts its web so wide as narrativity, and perhaps no other defies definition more stubbornly. A narrative of any kind—verbal, non-verbal, fictional, non-fictional, etc.—as well as other cultural artifacts containing elements of narrative—lyric poetry, painting and sculpture, but also conversation, computer games, etc.— would hardly be thinkable in the absence of narrativity or, given the fairly recent coinage of the term, some principle or function fulfilling this role. But what constitutes narrativity? Can narrativity be defined by its formal features? Is it one narrative category among others? Are there types of narrativity?, degrees of narrativity? Do narratives possess narrativity or do they exhibit narrativity? Do they produce narrativity or are they produced by narrativity? Does narrativity in, say, a novel differ from narrativity in a short story or a film? Can narrativity be perceived in different ways? While consensus on such questions is unlikely ever to be achieved, the fact that narrativity and its impact on narrative theory have grown since the inception of the notion in the 1960s and 1970s, understood at the time as an "immanent" property of narrative by some and as a marginal consideration by others, into something compellingly more multifaceted reflects the historical mutations of narratology as researchers have expanded the object of their analyses from written narratives to include a broader spectrum of cultural artifacts and as narratology itself has developed into a bewildering array of theories and methodologies.1 Further evidence of the evolution of narratological paradigms can be found in the rise of "transgeneric" narratology for the analysis of poetry2 and drama.3 Moreover, the relations between narrative, music and the visual arts have given rise to an "intermedial" narratology,4 while "transmedial" narratology focuses on 1 2 3 4
For a helpful inventory of present-day "narratologies," see Nünning (2003). E.g. Hühn/Kiefer (2005); Schönert/Hühn/Stein (2007). E.g. the contributions of Nünning/Sommer and Fludemik to the present volume. E.g. Wolf (1999), (2003), (2004); Nünning/Nünning (2002).
110
John Pier
how the properties of the various semiotic media (writing, image, sound, electronic media) influence the form and affect the experience of narrative.5 Indeed, the expansion of narratology to include non-epic forms within its purview would never have come about without the corresponding emergence of reflection on the nature of narrativity. As Marie-Laure Ryan has explained: [...] if we accept the possibility of narrativity in drama, movies, and painting, the presence of a narrator is no longer a necessary condition, and if lyric poetry, philosophy, legal cases, and simply life are able to display the property of narrativity, we may have to reconsider the dependency of this property on the explicit presence of a plot. From a property w e could take for granted, the narrativity of the text becomes a problematic issue. 6
In response to this situation, Ryan outlines a theory of twelve "modes of narrativity" which regards story not as a coherent and knowable point of departure to be described at the discourse level but proceeds, rather, "from the text to the story, taking the story as a problematic meaning to be recovered from the text."7 Three principles form the "basic conditions of narrativity": 1. A narrative text must create a world and populate it with characters and objects. Logically speaking, this condition means that the narrative text is based on statements asserting the existence of individuals and on statements ascribing properties to these individuals. 2. The narrative world must undergo changes of state that are caused by physical events: either accidents or deliberate human action. These changes create a temporal dimension and place the narrative world in the flux of history. 3. The text must allow the reconstruction of an interpretive network of goals, plans, causal relations, and psychological motivations around the narrated events. The implicit network gives coherence and intelligibility to the physical events and turns them into a plot.
As principles underlying the conditions of narrativity, these criteria are incontrovertible, for it could hardly be the case that a story is told which does not populate a world with characters and objects, in which no change takes place and which is expressed in a medium that does not enable read-
5 6 7 8
E.g. Ryan (2004), (2005a). Ryan (1992: 368-69). Ibid.: 369. Ibid.: 371.
After this, therefore because of this
111
ers9 to determine the coherence and intelligibility of events formative of plots. Some of the modes are plot-typological while others involve "the mental operations necessary to retrieve and/or properly evaluate the narrative structure,"10 and among the modes some are binary, others relative. Following a somewhat different set of criteria and emphasizing degrees rather than modes of narrativity, Gerald Prince distinguishes narratives from non-narratives as an aspect of narrativity that he calls "narrativehood" (what, extensionally speaking, defines narratives as entities) as opposed to "narrativeness" (the ways in which the traits of a narrative qualify it intensionally), and he further differentiates these textual features from contextual factors contributing to the "narratibility" of a story." Another notion, closely related to narrativity, is that of "eventfulness" according to which, within a narrative world, events are both "real" and "resultative" (necessary conditions), but with respect to context-sensitive factors are also more or less "eventful" to the extent that changes of state are relevant, unpredictable, persistent, irreversible and non-iterative.12 The present paper seeks to identify the ways in which narrativity is both born out of and a condition of stories. In order to achieve this aim, it grants particular interest to the post hoc ergo propter hoc fallacy—"the mainspring of narrative activity," as Roland Barthes put it in a suggestive formula.13 The attractiveness of Barthes' metaphor lies in the idea that narrativity is not so much a category to be included among the properties of narrative as it is a source of the dynamic action of narrative, the emphasis being on the "how" of narrative rather than on the "what." However, Barthes fails to exploit the full potential of his insight, and for this reason the present contribution will seek to provide a more satisfactory theoretical framework for the fallacy as a key to the principle of narrativity.
9
10
13
Or listeners, spectators and viewers. Except where "reader(s)" is used specifically with regard to written texts, this term will be understood, for the sake of economy, as the receiver of narratives in any semiotic medium. Ibid.: 382. See Prince's article in this volume and Prince (2005). Cf. Schmid (2003) as well as Peter Hühn's contribution and Hiihn (forthcoming). Barthes (1966: 10). The English translation of Barthes' article gives "the mainspring of narrative" for "le ressort de l'activiti narrative" (1977: 95)—revelatory of a confusion in reflection about narrativity from the outset between "categorial" and "functional" approaches. For Prince (2003 [1987]: 78), Barthes identifies the fallacy in question as "the mainspring of narrativity."
112
John Pier
1. From an Atemporal Toward a post hoc/propter hoc Narrative Logic Approaches to narrativity such as those referred to above have by no means been adopted by everyone who has studied the question. In one of the earliest explicit formulations of the concept, dating from 1969, A. J. Greimas spoke of "an immanent level, constituting a sort of common structural core where narrativity is located and organized prior to its manifestation," i.e. prior to (and distinct from] the "apparent level of narration, [...] whatever the language chosen for manifestation." 14 An even broader claim is made when it is stated that narrativity is "the very principle of all narrative [...] and non-narrative discourse" and that "generalized narrativity [is] the organizing principle of all discourse." 15 The dubious claim of the inherent narrativity of all discourse (both narrative and non-narrative) aside, it has been suggested, somewhat erroneously, that Greimas's identification of immanent story structure with narrativity follows in the footsteps of Propp. 16 It must be pointed out, however, that in his morphological study of the Russian folktale, Propp sought to classify the thirty-one "functions" of the dramatis personae forming "the basic components of the tale": "Function is understood as an act of a character, defined from the point of view of its significance for the course of the action." 17 At the same time, he deliberately set aside the "style of the tale" and laid no claim as to the status of these functions with regard to something comparable to Greimas's "apparent level" or to the "signifying structures" focused on by various linguistic and semiotic theories. A milestone in the modern study of narrative and one of enduring influence, Propp's contribution is nonetheless marred by reductionism insofar as, unlike anthropologists and socio-linguists, he worked not in the field where these oral tales were actually told and listened to, but with written transcriptions. By bracketing these factors out, he necessarily failed to take account of the "performance" of oral storytelling and of the possible impact on the un14
15 16 17
Greimas (1970 [1969]: 158). Cf. Henault (1993 [1979]: 144): "It can be said there is narrativity when a text describes, on the one hand, a state of departure in the form of a relation of possession or dispossession with a valorized object and, on the other hand, an act or a series of acts productive of a new state, exactly the opposite of the state of departure." (All translations from the French mine) Greimas/Courtis (1979: 248-49). Rimmon-Kenan (2002 [1983]: 7). Propp (1968 [1928]: 21) (italicized in the original).
After this, therefore because of this
113
folding of the story of the interaction between the teller and his audience through gesture and other proxemic features, phatic and deictic expressions, the social relations among participants, etc. The alternative to identifying narrativity with immanent story structure18 dates from Labov and Waletzky's (1967) important article, subtitled "Oral Versions of Personal Experience," introducing such notions into the pragmatics of narrative as "point" and "tellability" that have become key terms in the analysis of conversational storytelling.19 As a result of adapting methodologies to account for written as opposed to conversational narratives, for example, rather than attempting to identify a formal deep structure that embraces all manifestations of narrative, narrativity can no longer be considered a property of or synonymous with the "immanent structure" of narrative.20 Theories that ground narrativity in deductive and axiomatic models inspired by what Paul Ricceur characterized as "semiotic rationality" seek to provide an achronic simulacrum of narrative. In doing so, they allow for little if any distinction in principle between fictional and historiographic narratives or, we might add, narratives as they appear in legal discourse, psychoanalytic discourse, conversational discourse, etc.21 Indeed, the "elementary structure of signification" that enters into Greimas's "constitutional model," based on the logical categories of contrariety, contradiction and presupposition, postulates a set of formal requirements that cannot be met by narratives: were these conditions to be fulfilled, Ricceur notes, "nothing would happen. There could be no event, no surprise. There would be nothing to tell."22 In fact, readers and listeners of narra18
20
21
22
The critiques of the over-extension of Propp's method by structuralist narratology to include complex literary works, for example, are too numerous to cite here. For the necessity of taking account of the interactional dimension of narrative in the elaboration of a "transmedial" narratology, see Herman (2004). For a brief survey of the notion of "tellability," see Ryan (2005b). On the "open-endedness" of conversational storytelling, weak in the narrative levels and teleologically-ordered temporal sequences exploited by written narratives, for instance, see Ochs/Capps (2001). Fludernik's (1996) "natural" narratology, in which "experientiality" rather than "mere sequentiality and logical connectedness" (ibid.: 26) lies at the root of narrativity, derives in large part from the ramifications of this form of narration for narrative theory, including its narratological varieties, traditionally bound to the features and parameters of written narratives. Ricceur (1985 [1984]), especially chapter 2: "The Semiotic Constraints of Narrative." The criticism bears on the Ecole de Paris, but also pertains to other formal models. Ibid.: 57. For Claude Bremond (1973 [1972]: 99), the major defect of Greimas's method is that "it ignores a law of narrativity: the option, or rather the obligation, to de-
114
John Pier
tives are constantly confronted at the surface level of texts with quasicontradictions, quasi-contrarieties and quasi-presuppositions that serve to trigger the operations of "narrative intelligence." Now, for me, it is precisely out of such quasi-logical or even logicallydefective operations that narrativity is born—a dynamic process engaged during the unfolding of a narrative rather than a built-in textual property. Narrativity operates through trial-and-error inferential reasoning in which suppositions and conjectures are entertained, consciously or not, subsequently to be confirmed, invalidated or revised, or perhaps left in suspense, displaced by unforeseen or improbable incidents, contradicted by incompatible or inconclusive developments, etc. Involved is a process of "heuristic" reading completed by "semiotic" reading which, we will see, has important implications for narrativity whose driving force, as already suggested, is the fallacy "after this, therefore because of this." It is enlightening to note, however, that while this fallacy, or something resembling it, has long been associated with reflection on narrative, its emergence as a principle of narrativity can in no way be characterized as resolute. Closer examination of Barthes' article reveals that the word "narrativity" occurs three times, but in conjunction with "signs" and "signifiers" rather than with the Scholastic formula. 23 In fact, the author is careful to take his distance from "the very confusion between consecution and consequence, what comes after being read in narrative as caused by," so that "narrative activity" is ultimately founded on a "logical error." 24 In line with Levi-Strauss's dictum that "[t]he order of chronological succession is absorbed into an atemporal matrix-like structure," 25 he goes on to state that the tendency of narrative analysis at the time (the 1960s) was to "dechronologize" and then "relogicize" narrative content: from the perspective of "narrative logic," time is a "chronological illusion" included
24
25
velop as a series of choices made by the narrator at each moment of the narrative between several ways of continuing his story." I owe this observation to Sturgess (1992: 156). Barthes (1966: 10). It is paradoxical that the "functionality" of "catalyses" (chronological) and the "double functionality" of "cardinal functions" (chronological and logical)—the "smallest narrative units"—are characterized as a "logical error." It should be noted, however, that in his S/Z (1974 [1970]), Barthes sought to redress such difficulties through the introduction of the "proairetic" code (sequence of actions) and the "hermeneutic" code (an enigma that arouses a sense of curiosity and suspense in the reader. For a commentary, see Baroni (2007: 70-74). Quoted in Barthes (1966: 12).
After this, therefore because of this
115
within a more general "referential illusion."26 Barthes' argument thus simultaneously points to the potential of the formula "after this, therefore because of this" for a theory of narrativity and, in the name of the logical form of narrative content, built up from the "smallest narrative units" deployed along the paradigmatic and syntagmatic axes postulated by structuralist linguistics, relegates it to a parasitic status in the name of "an atemporal logic behind the temporality of narrative,"27 thereby underrating the powerful synthesizing nature of narrativity. The search for such an atemporal narrative logic, in line we might note with Greimas's notion of "temporalization" (which "consists, as its name indicates, in producing the meaning effect [effet de sens] 'temporality', and thus of transforming a narrative organization into a 'story'" 28 ), predictably encounters a serious objection: "Surely, the logic of causality does not oppose (break, override, replace) but presupposes and tightens the line of chronology."29 Closer to the perceptions of readers, and without appeal to an underlying "atemporal logic" of narrative, is Gerard Genette's discussion of the "arbitrariness of narrative," a feature of storytelling found by Valery to be both irritating and fascinating.30 Focusing on the functional nature of narrative units, he adopts Saussurean terminology in conjunction with Russian Formalist concepts, opposing an "arbitrariness" of direction—the possible-at-each-instant, or contingency of linear succession—to retrograde determinations, or the determination of causes by effects, the end commanding the beginning teleologically in such a way that an a posteriori justification or "causal alibi" constitutes the "motivation" of narrative. In this way, fiction is governed by a "finalist determination": "the because assigned to making one forget the for what?—and thus to naturalizing fiction or realizing it (in the sense of causing to pass for real) by concealing how it is artificial." According to the degree of motivation, he differentiates two types of narrative, akin respectively to recit (or histoire) and discourse "non-motivated" narrative, subdivided into "arbitrary" ("The
"ml 27 28
30
Ibid. Greimas/Courtes (1979: 388). Sternberg (1990: 916). And of course Ricceur's Temps et recit, from another perspective, is predicated precisely on refuting the evacuation of time from the narrative art. Genette (1968: esp. 17-21). Terms taken from Emile Benveniste for the two modes of linguistic enunciation and adapted by Genette (1966: 159ff.) to designate, respectively, "pure" narrative and language in its "natural" form, showing traces of subjectivity.
116
John Pier
marquise asked for her coach and went to bed") and "verisimilar" ("The marquise asked for her coach and went for a ride"), as opposed to "motivated" narrative ("The marquise asked for her coach and went to bed, for she was quite capricious"). Although expressed in a language characteristic of the structuralist "moratorium on representational topics" 32 and despite the objection that the linkages "depend [...] only on the institutionalized beliefs of the reading public," 33 Genette's examples do serve nonetheless to bring out a process of reasoning which is unmistakably narrative: having read that the marquise has ordered her coach, one is more likely to assume, prospectively, that she will take a ride than go to bed, take a bath or dress for tea; should the marquise do the unexpected, however, an explanation is in order so as to justify, a posteriori, the unforeseen course of action by stating that the marquise is capricious, for instance, or by inferring, retroactively and through circumstantial evidence, that she is capricious or possibly that she has suddenly been taken ill or that the coachman is nowhere to be found. Now, it is both noteworthy and instructive that Genette, in keeping with structuralist narratology generally, does not incorporate these insights into his Narrative Discourse. Defining narrative as "a linguistic production undertaking to relate one or several events," he proposes (as is well known, but bears repeating for present purposes) 1) to treat narrative as "the expansion of a verb. I walk, Pierre has come are [...] minimal forms of narrative"; 2) to analyze narrative discourse according to the categories of the verb: tense (temps); mood {mode)·, voice (yoix)\ 3) to examine the relations between story (histoire: "the signified or narrative content"), narrative (recit: "the signifier, statement [enonce], discourse or narrative text itself') and narrating (narration: "the producing narrative act and, by extension, the whole of the real or fictional situation in which that action takes place"). 34 In this narratology geared to "discourse" rather than to "story" (cf. Bremond's recit racontant as opposed to recit raconte) or, in Genette's terms, a "modal" narratology ("analysis of narrative as a mode of 'repre32 33 34
Pavel (1986: 6). Sternberg (1983: 161). Cf. Genette (1980 [1972]: 30-1, 27). I have adopted "to relate one or more events" for "la relation d'un ou plusieurs 6venement(s)" in place of the published translation's "to tell o f ' and "narrative act" for "acte narratif' in place of "narrative action."
After this, therefore because of this
117
sentation' of stories," ruling out drama, for example),35 narrative is considered the "expansion of a verb."36 Thus, "Pierre has come," relating a single event, forms a "minimal narrative," further reduction of which would result in nothing to narrate at all and which, from the utterer's standpoint, is susceptible to limitless if not aimless expansion. However, unlike "The marquise called for her carriage and went to bed," it allows little if any place to narrativity in the makeup of narratives, leaving the question of linkages in suspense with meager hope, for the reader, of resolution: Why has Pierre come? Having come, what will he do next?... Gerald Prince goes some way toward settling this issue with the notion of "minimal story," which consists of "three conjoined events," as in: "He was unhappy, then he met a woman, then as a result, he was happy." The first and third events are Stative, the second is active. Furthermore, the third event is the inverse of the first. Finally, the three events are conjoined by the three conjunctive features in such a way that (a) the first event precedes the second in time and the 37 second precedes the third, and (b) the second event causes the third.
As the term "minimal story" indicates, the focus here is on an idealized narrative content, purged of the parasitical discursive variables that intervene to a greater or lesser extent in actual narrative communication and in stories that are not reduced to their bare minimum. One might also wonder about the meaning of "stative event" and whether stories truly portray such symmetrical inversions without further ado, either at the beginning or at the end.38 Be that as it may, Prince has identified what, in principle, would occur in narrative in its "purest" form, namely, the unqualified coincidence of causality and chronology (no fallacy here), meeting the standards we are entitled to expect of any reliable narrator worthy of his name: "The hallmark of narrative is assurance. It lives in certainty: this happened then that; this happened because of that; this happened and it was related to that."39 Of uncontested conceptual clarity, the criteria for minimal story seem overly stringent and restrictive vis-ä-vis actual narratives, and it is perhaps for this reason that Prince later proposed to define narrative (sans "minimal") as "the representation of at least two real or fictive events or situa-
35 36 37 38 39
Genette (1988 [1983]: 12). Genette (1980 [1972]: 30). Prince (1973: 31). Cf. Sternberg (1992: 465-66). Prince (1982: 149).
118
John Pier
tions in a time sequence, neither of which presupposes or entails the other." 40 This definition is more accommodating to Genette's marquise narratives, but at the same time it may underestimate the role of presupposition and entailment, opening the way to a potentially unrelated sequence of events devoid of any narrative cohesiveness. To the extent that Marcel becomes a writer can be expanded to encompass Proust's entire Recherche, it is presupposed, firstly, that Marcel exists within a certain universe of discourse and secondly, that the statements contained within that novel pertain predominantly to Marcel, notwithstanding the impact of stories about secondary characters, potentially disruptive violations of the basic presupposition, etc. Similarly, Marcel becomes a writer is entailed by "Marcel learns to tell stories, writes a novel and publishes it," but also by "Marcel enrolls in journalism school, earns a diploma and joins a newspaper staff' as well as by a number of other possible scenarios that might be encountered in stories. Proceeding in the opposite direction, moreover, engages one in an inductive process whereby on reading Marcel becomes a writer it is possible to infer, but only with a certain degree of probability, to be verified subsequently, that he becomes a novelist, a journalist or a screenwriter. 41 This being the case, it may be desirable to amend Prince's definition of narrative to read: "the representation of at least two real or fictive events or situations in a time sequence, either of which might presuppose or entail the other." All in all, then, it seems that these "minimal" forms—narrative or story—serve to broach questions that can be adequately explored only in larger stretches of discourse. Their very artificiality (unless one wishes to construe them as "mini-stories") limits the extent of their applicability to texts as such, if for no other reason than that they virtually squeeze out the possibility of narrativity as an active or dynamic process engaged when reading, viewing or listening to stories. Perhaps the main interest of these minimal forms is that they underscore the necessity of reflection on narrative linkages. A second trait of Genette's narrative model, also characteristic of structuralist narratology in this regard, is the absence of the principle of 40
41
Ibid.: 4 (emphasis in the original). Cf. Prince (2003: 58) which defines narrative as "[t]he representation (as product and process, object and act, structure and structuration) of one or more real or fictive EVENTS communicated by one, two, or several [...] NARRATORS to one, two, or several NARRATEES," thus downplaying the earlier formal definitions of event. Cf. Pier (1980: 332-33).
After this, therefore because of this
119
plot. As his definitions of "story" and "narrative" show, the assimilation of narrative content and narrative text into concepts of structural linguistics shifts the focus away from the notion of plot. This is confirmed in the chapter titled "Mood," devoted to the degrees of affirmation and their modal variations in narrative ("regulation of narrative information"), where Genette takes up Plato's discussion of the poet speaking "in his own name" ("pure narrative," or haple diegesis) and speaking "as if he were someone else" ("imitation," or mimesis), and he accordingly adopts the terms "narrative of events" ("transcription of the (supposed) non-verbal into the verbal") and "narrative of words" ("imitation" of character discourse).42 Nowhere in Narrative Discourse, however, is there a mention of "plot," or muthos"—"the arrangement of the incidents" (sunthesis tön pragmätöri)—in other words, of the fact that, for Aristotle, tragedy (and stories generally, as tradition would have it) is the "imitation or representation of the action [praxis]" of people.43 The absence of plot in modal narratology has a number of consequences. Since it includes no differentiation comparable to that between "simple" plots ("an action which is one and continuous" and in which "the change of fortune takes place without Reversal of the Situation [peripeteia] and without Recognition [anagnorisis]") and "complex" plots (an action "in which the change is accompanied by such Reversal or Recognition or by both"),44 the problems of causality and chronology do not arise. For Aristotle, it is precisely because of the distinction between simple and complex plots that propter hoc can be opposed to post hoc, for reversal and recognition, it is pointed out, should arise from the internal structure of the plot [i.e. the arrangement of the incidents], so that what follows should be the necessary [anagke] or probable [eikos] result of the preceding action. It makes all the difference whether any given event is a case of propter hoc or post hoc.*5
It stands out here that causality is not to be confused with chronology, although this is not to say (as Barthes would have it) that narrative is absorbed into an atemporal logic or that causality and chronology cannot cohabit. A second point is that the above passage must be read against the backdrop of Aristotle's holistic conception of tragedy, based on the rela42 43 44
Genette (1980 [1972]: 162ff.). Poetics 1450a. Ibid.: 1452a.
120
John Pier
tions of necessity and probability between the parts of the action within the context of the whole, the resultant functional tension between holos and müthos being an important but little-explored theme of the Poetics,46 "A whole," says Aristotle, is that which has a beginning, a middle, and an end. A beginning is that which does not itself follow anything by causal necessity, but after which something naturally is or comes to be. An end, on the contrary, is that which itself naturally follows some other thing, either by necessity or as a rule, but has nothing follow it. A middle is that which follows something as some other thing follows it. A well constructed plot, therefore, must neither begin nor end at haphazard, but conform to these principles. 47
Thus, when holos and muthos coincide (as in "He was unhappy, then he met a woman, then as a result, he was happy"), that is, when causality and chronology are coterminous with beginning, middle and end, no particular emotion is produced in the reader (such as pity or fear in the case of tragedy). However, an unexpected chain of events will come as a surprise ("He was unhappy, then he met a woman, then as a result, he was bankrupt"), touching off a series of questions, conjectures, suppositions, etc. in an attempt to get a better grasp of what has occurred. Surprise is heightened, moreover, when such events "follow as cause and effect" rather than "of themselves or by accident; for even coincidences are most striking when they have an air of design." 48 Put another way, the relations of cause and effect in plots can be determined only in function of the whole, and not within their local context. This being the case, it is notable that Aristotle's commentary on simple and complex plots (quoted above) does not employ the word ergo, the emphasis being on the difference between what happens next and what happens as a result. In other words, Aristotle does not introduce here the logical mechanism that would allow us to infer, fallaciously or not, that because the marquise went to bed after ordering her coach, she is capricious. Or that, on reading "The king died and then the queen died," the queen died of grief. I will not dwell on Ε. M. Forster's illustrations of story ("a narrative of events arranged in their time-sequence") and plot ("also a narrative of events, the emphasis falling on causality"), 49 as familiar to students of narrative as "The cat is on the mat" is to several generations of logicians. 46 47 48 49
Cf. Sternberg (1992: 474ff.). Poetics 1450b. Ibid.: 1452a. Forster (1962 [1927]: 27).
After this, therefore because of this
121
However, I would like to draw attention a curious and insurmountable ambiguity of the two sentences that has perhaps not been sufficiently taken into account. Given that both story and plot are abstractions that must be reconstituted and that plot can be dispensed with, but not story, if a text is to be considered a narrative, what is the textual status of Forster's sentences? Are they intended to summarize a larger stretch of discourse? Are they verbalizations of underlying textual structures? Or are we to suppose that they actually occur in the text? Whatever the answer, it is clear that the two sentences contain a number of unspoken assumptions: the king is the legal heir to the throne and both the king and queen in question are monarchs of the same realm; they are lawfully married to each other and are of the same or approximately the same generation. But now let us imagine a story that begins as follows: "The king died and then the queen died." Toward the end, we learn the identity of the two individuals: Charles I of Spain (died in 1558) and Elizabeth I of England (died in 1603). The historical facts are correct, but can we conclude in this case that what comes after is caused by what comes before? And can we entertain the idea that Elizabeth I died of grief as a result of the death of Charles I? A cleverly crafted text may well lead us down this path for many pages until the facts are revealed, at that point casting the intervening pages in a very different light and possibly prompting us to reread. And what if, after this opening sentence and several pages of narration, it were to turn out that the "king" and the "queen" are the nicknames of two exotic characters living in the Haight-Ashbury district of San Francisco? In another imaginary story, the situation would be different yet again if, somewhere in the middle, it became evident that "The king died and then the queen died" were a set of instructions given by a kindergarten teacher to her pupils for acting out a fairy tale. In all of the above cases, the formula "after this, therefore because of this" applies to the sentence in question, but with effects distinctly at variance with those of the sentence in its isolated form. More precisely, the formula applies provided some causal relation between the death of the king and the death of the queen is inferred, which, however, would tend to blur the boundary between story and plot as intended by Forster. Applied over the longer stretch of discourse in the narratives imagined above, the situation is altered markedly: the causal relation assumed within the initial sentence is projected onto the text to come, conditioning not only how the local sequences will be apprehended, but also how the relations between the parts and the whole will be understood, only to be reevaluated in retro-
122
John Pier
spect. Clearly, then, something more is at issue here than what can be gleaned from Forster's story and plot: involved is a process that can be related to what Genette, borrowing from structuralist terminology, describes as "arbitrariness" vs. "motivation" of narrative, but which will be examined here in terms of what Louis O. Mink calls "the contingencies of forward references" as they are "canceled out" by "the necessity of backward references," culminating in "configuration," or an act of "seeingthings-together" which is peculiar to the understanding of stories. 50 In an earlier study, I analyzed configuration from an intertextual perspective, showing how narrative (using Nabokov's Lolita [1955] as a test case) assimilates intertextual frames to a greater or lesser degree through operations of inferential reasoning based on Umberto Eco's text-semiotic theory of abductions. Lower-order abductions, it was argued, are triggered in the process of linear or "heuristic" reading, while retroactive or "semiotic" reading engages higher-order abductions. 51 Working largely within the same parameters, the present study shifts the emphasis to narrativity. With these considerations in mind, let us now return to the fact that in his account of complex plots Aristotle is careful to separate causality and chronology conceptually, but provides no link between them, no "therefore" from which to derive a causal chronology. Even so, the possibility of relating the two is not ruled out—nor, interestingly, is there any mention of the fallacy in the Poetics. For a discussion of this point, it is necessary to turn to two other texts, firstly, the Sophistical Refutations, a treatise on sophisms ("contentious arguments") that has recently come to be seen as a key to understanding the long-neglected dialogical aspect of Aristotle's logic. 52 A refutation is defined as "a proof of the contradiction" and sophistical refutations as "what appear to be refutations, but are really fallacies instead," 53 in other words, arguments that appear to be valid but that are not. Among the thirteen fallacies employed in sophistical refutations is that of "treating as a cause what is not a cause," 54 a form of "deduction ad
50 51
54
Mink (1970: 55Iff.). Pier (2004); cf. Pier (2003: 85-93). C.L. Hamblin's Fallacies (2004 [1970]) is regarded as a watershed in this revaluation and comes within the context of the development of informal logic, the theory of argumentation and pragmatics. The following discussion draws on ibid.: 37-8, 78-80 passim; Allwood/Andersson/Dahl (1977); Lukasiewicz (1957 [1951]); Mackie (1967); Walton (1995); Woods/Walton (1989 [1978]). Sophistical Refutations 168a, 164a. Ibid.: 167b.
After this, therefore because of this
123
impossibile" that consists of inserting "irrelevant matter" into the argument—a move that may threaten the validity of that argument, somewhat like a violation of the Gricean Cooperative Principle. "Non-cause as cause" is taken up again in the Rhetoric, this time not as an error in the form or the premises of syllogisms, but in a way that bears more directly on the problems of narrative: Another line [i.e. topic] consists in representing as causes things which are not causes, on the ground that they happened along with or before the event in question. They assume that, because Β happens after A, it happens because of A.55
In a way, it is easy to see why this passage was not included in the Poetics, for it says that a causal relation inferred from a temporal sequence is unwarranted—hardly a basis on which to get a good story going! This form of "non-cause as cause," as the Rhetoric calls it—later to be known as the post hoc, ergo propter hoc fallacy—corresponds to "false cause" in the modern sense of the natural sciences, not the logical sense. It has generally been acknowledged since Hume that the logical necessity of a causal connection can never be demonstrated: the fact that Β follows A in time does not mean that A entails Β or that Β can be derived deductively from A. To compensate for this situation, the sciences have developed methods of inductive testing which include checking up with adequate control cases so that it may be possible to establish a high degree of probability values or of correlation coefficients between A and B; in this way, it may be found that causality and temporal succession coincide in a certain percentage of cases. With insufficient statistical data, however, there is a risk of making unwarranted inductive generalizations—"the fallacy of jumping to a conclusion." 56 On the other hand, even a generic statement with a high level of correlation such as "Eating fast food causes cardiovascular disease" does not provide an adequate basis for establishing a causal relation in any particular instance of individuals who both eat fast food and suffer from cardiovascular disease. Now, the differences between the two forms of our fallacy—the logical, which appears in syllogistic form, and the causal, pertaining to the natural sciences—has to my knowledge been largely if not entirely overlooked by theoreticians seeking to employ it in the analysis of narrative. 57 55 56
Rhetoric 1401b. Salmon (1963: 56); quoted in Hamblin (2004 [1970]: 46). The two forms are fully conflated in Barthes' "Introduction," and since this model rests on the deductive procedures of structuralist linguistics, the fallacy must ultimately give
124
John Pier
Syllogistic fallacies, whether deductive, inductive or retroductive (hypothetical), serve to structure arguments, not narratives, and strictly speaking the causal form is not a fallacy but an insufficiently tested natural law, the characteristics of which are unlike those of narratives, whatever their semiotic media. In any case, narrative, as any other cultural object, is constituted only through and in the various semiotic media, so that what is at issue in narrative cannot be the nature of causality per se but rather, "that of the function and representation of what is interpreted as 'causality' in narrative (and in common speech)"; for this reason, causality is inevitably "relativized" in narrative—a function of the inferences that can be drawn from the formula "after this, therefore because of this" within a narrative context.58 2. Heuristic Reading, Semiotic Reading Given the considerations debated so far, I would like now to suggest that there are considerable advantages to analyzing narrativity as a negotiation between the fallacy in question (as presented in the Rhetoric) and the theory of abductions. While the former is an error of induction, the latter is an adaptation of the syllogism in Peircean semiotics to the procedure of scientific inquiry,59 but it is also highly relevant to the theory of textual way to an "atemporal logic." Robert Scholes proceeds somewhat differently. Defining narrativity as "the process by which a perceiver actively constructs a story from the fictional data provided by any narrative medium," he then argues that narrativity is based on "a mental operation similar to a logical fallacy: post hoc ergo propter hoc. What is a fallacy in logic is a principle in fiction: that a cause-and-effect relationship links the temporal elements in any sequence. I am not suggesting that fiction itself is fallacious in some way, but rather that it is constructed so as to make this fallacy a feature of the fictional world" (Scholes [1982: 60, 62]). The definition seems closer to that of "narrativization" (cf. Alber [2005]) than it does to one of narrativity, and the mental operation involved posits the fallacious nature of causal and temporal relations in narrative sequence as a characteristic of the ontology of fiction, the implicit assumption being, among other things, that fiction and narrative are synonymous; in light of more recent research on fictionality, this assumption is no longer tenable (cf. Schaeffer [1999]). Genot (1980: 19). The principles focused on in the present comments seek to clarify the conditions according to which causality is constituted in narrative. Conceptually, they lie upstream of what Brian Richardson (2005) in his excellent overview of causality in narrative describes as the connections among narrative units, events and character action in contrast to the system of causation, along with space and time, that govern the ontology of a storyworld. See, for example, Eco (1983).
After this, therefore because of this
125
communication. The interest of the theory of abductions for textual matters, and more particularly for the problems of narrative, lies in the fact that it leaves considerable room for non-deductive forms of reasoning, in other words for inferences that may be valid with a greater or lesser degree of probability. Such reasoning is better adapted to the analysis of narratives than deductive inferences, which are valid only when their conclusions are logically entailed by their premises, for stories, contrary to the claims of advocates of narrative grammars, are not the emanations of axiomatic systems. The various forms of abduction offer a promising framework for demonstrating how specification and semiotic mediation of the fallacy in narrative contexts proceed. (For want of space, the following discussion, which lies against the backdrop of my two articles already mentioned, will remain brief.) In an "overcoded" abduction, an inference occurs in a somewhat mechanical fashion when, in a given context of utterance and co-text, one proceeds unquestioningly from a general rule to a specific case. Referring back to Aristotle's non-cause as cause, the tendency to assume that if Β happens after A, then Β happens because of A (an assumption that may not be fallacious in all cases) "overcodes" the segment in question; as a matter of hypothesis, this pattern is then projected forward, possibly covering the entirety of the narrative. If the overcoded abduction is not subsequently confirmed (B is not caused by A, or its cause is unspecified), a plausible alternative will be sought out.60 This is an "undercoded" abduction, an inference to a probable rule possibly involving an "inferential walk" outside the text in search of an appropriate solution. Reaching backward, undercoded abductions seek the causes of effects (marquises who go to bed after ordering their coach are capricious), but at the same time project forward, seeking confirmation that may never come. These operations take place as the reader progresses, engaging a process of "heuristic" reading. But as this process advances, a retroactive or "semiotic" reading takes form, of greater or lesser intensity according to the complexity of the narrative in question. On encountering either indeterminacies ("gaps" to be filled in in the form of a "fair guess" as to the 60
For Schmid (2003: 19-20), the minimal definition of narrativity does not require causal connections on the grounds that literary texts rarely specify an explicit cause for a change of state and that the possible causes of a change of state lie open to interpretation. Empirically speaking, this cannot be disputed, but it does not explain away the trial-and-error attempts of readers of stories to formulate valid hypotheses in quest of narrative coherence, however tentative, incomplete or inconclusive.
126
John Pier
future course of events, subsequently to be confirmed, or perhaps a missing element, perceived in hindsight) or heterogeneous elements that cannot be accounted for by overcoded or undercoded abductions, "creative" abductions come into play, that is, conjectures which hypothesize facts and/or a general rule that must then be submitted to inductive testing ("meta-abduction"). The latter case can be illustrated with the cinematic version of Cyrano de Bergerac, when Cyrano says: "This is the nose that launched a thousand battles." 61 Although it is known from the story itself that Cyrano's long nose is at the origin of many troubles, some viewers may feel perplexed or somehow left out of the picture by this statement; but those who are familiar with Marlowe's Dr. Faustus (1592-93) are likely to recall the sentence "This is the face that launched a thousand ships," in reference to the beauty of Helen—the cause of the Trojan War. Here, an inferential walk outside the film in search of a work already known (or in any case presupposed) initiates an inductive inference regarding physical traits and their fateful consequences, but it also triggers undercoded abductions bearing on the relevance of the parallel between Helen's love for Paris in Homer and in Marlowe and Roxane's love for Cyrano in the film as well as on Cyrano's "Faustian" character, etc. The two readings—heuristic and semiotic—-include series of projections forward and reaches backward, with a dynamics of prospection, retrospection and recognition being set up as from the time of heuristic reading. But the process does not stop here. As shown by the example above, an intertextual allusion or some other heterogeneous element, even when perceived in the course of a heuristic reading, can ultimately be accommodated only as the result of a semiotic reading. 62 It is here that a configuration is achieved, a "seeing-things-together" that includes, but extends beyond, the purview of heuristic reading—a total that adds up to more than the sum of its parts, so to speak. The potential for such configurations is evidently greater in narratives with a powerful intertextual dimension than it is in those of a more "straightforward" nature, although even here there might be more than meets the eye at a leisurely reading. With configuration, the "after this, therefore because of this" encountered in heuristic reading—itself more all-embracing than a Forster-type
61
The example is taken from Ben-Porat (1976:112-15). Cf. Eco (1979: 205) which speaks of "a naive and a critical reading, the latter being the interpretation of the former."
After this, therefore because of this
127
"plot"—reaches backward and projects forward beyond the limits of the narrative at hand, situating it within a web of intertextual relations on which, in one way or another, all narratives rely. The system outlined above provides a possible theoretical framework aimed at explaining how the fallacy "after this, therefore because of this," the "mainspring" of narrativity, is deployed in and plays a vital role in the constitution of narratives. In this process, abductive reasoning is engaged, consciously or not, but in either case and whatever their degree of sophistication and cultivation, readers are invariably drawn to wondering "What will happen next?", "Why did such and such happen?", etc., possibly being prompted to revision or to rereading on the emergence of unforeseen factors. In this regard, the perceiver's experience of a narrative proves to play a crucial role in that it excites the affective forces of suspense, curiosity and surprise that set into motion a dynamics of prospection, retrospection and recognition. It is in fact in these latter terms that Meir Sternberg defines narrativity: "the play of suspense/curiosity/surprise between represented and communicative time (in whatever combination, whatever medium, whatever manifest or latent form)."63 Or again: "narrativity lives in the dynamic interplay between the told and the telling, the represented and the communicative time-sequence."64 And it is striking that, unlike many theoreticians, he does not begin with a definition of narrative to which he then appends the notion of narrativity as something of an afterthought, but that it is out of the dynamic forces of narrativity that narratives are generated: narrative is "a discourse where such play dominates," and consequently narrativity is promoted to "the status of regulating principle, first among the priorities of telling/reading."65 To clarify these points, the following summary of Sternberg's theory is worth quoting in full: This interplay between temporalities generates the three universal effects/interests/dynamics of prospection, retrospection, and recognition—suspense, curiosity, and surprise, for short. Suspense arises from rival scenarios about the future: from the discrepancy between what the telling lets us readers know about the happening (e.g. a conflict) at any moment and what still lies ahead, ambiguous because yet unresolved in the world. Its fellow universals rather involve manipulations of the past, which the tale communicates in a sequence discontinuous with the happening. Perceptibly so, for curiosity: knowing that we do not know, we go forward with our mind on
63 64 65
Sternberg (1992: 529). Sternberg (2006: 129). Sternberg (1992: 529).
128
John Pier
the gapped antecedents, trying to infer (bridge, compose) them in retrospect. For surprise, however, the narrative first unobtrusively gaps or twists its chronology, then unexpectedly discloses to us our misreading and enforces a corrective rereading in late re-cognition. The three accordingly cover among them the workings that distinguish narrative from everything else, because they exhaust the possibilities of communicating action: of aligning its natural early-to-late development with its openness to untimely, crooked disclosure. 66
We see here that the various "effects/interests/dynamics" that characterize narrative are born out of the disjunctive relations that occur between "actional and communicative, told and telling/reading sequence";67 such is the case, for instance, of stories beginning in medias res. As a result of these distortions and twisting of chronology, "[t]he literary text can be conceived of as a dynamic system of gaps,"68 prospection, retrospection and recognition thus becoming so many operations for "filling in" these gaps, to the extent this is relevant and possible. In effect, these three "strategies" function as "basic sense-making operations [for] the construction of rival hypotheses with which to fill in the gaps opened up by the sequence about the world's affairs and whatever attaches to them by nature or art, which in narrative means everything."69 Returning now to the question of heuristic reading and semiotic reading, based on the theory of abductions through which, among other things, the fallacy "after this, therefore because of this" as it appears in narratives can be examined anew, two general observations are in order. First, it is in the sequentiality of heuristic reading that the ordering, disordering and reordering of the told in relation to telling/reading generates the narrative interests of suspense, curiosity and surprise.70 Suspense, turning on hope or fear, prompts a dynamics of prospection, a strategy which, in inferential terms, employs undercoded and/or creative abductions in order, for example, to formulate expectations about the protagonist's future course of action in light of what is already known. In re66 67 68 69 70
Sternberg (2001: 117). Ibid. Sternberg (1978: 50). Sternberg (1992: 531-32). These comments draw partly on Sternberg (1978: 236-^6) and (1985: 264-320). For another reading of these narrative interests, see Baroni (2007: 121 ff. passim), which integrates them into three phases of "narrative tension": 1) the "knot" of an intrigue that elicits a questioning; 2) a "delay" that generates expectation; 3) the "denouement" resolving the tension. This excellent work appeared too late to be given its due in the present paper.
After this, therefore because of this
129
sponse to curiosity, retrospection is engaged in an effort to find the missing link, and it is perhaps here that the assumption that if Β happens after A, then Β happens because of A stands out most prominently. In large part, this is no doubt for the simple reason that stories, in their vast majority, purport to relate a series of events or happenings that have occurred at some time prior to their telling. 71 In any case, given the arguments put forth in this paper, causes can be inferred with varying degrees of validity and are not to be confused with the causal fallacies that the scientific method strives to eliminate with empirical testing. Nor can causes be explained away with an appeal to (post)structuralist notions of "referential illusion": had Oedipus not killed a man in the road, would there be a story of how he came to commit incest with his mother? 72 In heuristic reading, surprise occurs with the belated revelation of a gap, setting up a dynamics of recognition when, in the course of a series of overcoded abductions (inferences drawn from general rules to particular cases), possibly adjusted in the wake of plausible alternatives arising from undercoded abductions, an apparently settled course of events is upset by the irruption of an unforeseen element. Up to this point, the element in question may have remained ingeniously camouflaged, imperceptible even to the most astute of observers; but it may also pop up out of nowhere in the form of an intertextual trace—a trace, moreover, whose ramifications might not be evident at first sight. It is at this point that undercoded abductions and inferential walks give way to the conjectures of creative abductions in an attempt to comprehend a state of affairs that may initially appear to be, or even remain, inscrutable. Our second observation, then, is that prospection, retrospection and recognition, which are initiated in the sequential perception of narratives, form a bridge between heuristic reading and semiotic reading. They represent strategies for binding together a "naive" reading and a "critical" reading. But it must be stressed that these two readings are not necessarily separated from one another in time as two distinct "acts" of reading, for to one degree or another the former is presupposed by, included within, the latter. Thus in practice, heuristic reading and semiotic reading may take place more or less concurrently, although in the case of more complex and "gappier" narratives, particularly those involving an inferential walk of an 71
For a discussion of the "hindsight bias" in narrative, see Josi Angel Garcia Landa's contribution to this volume and Garcia Landa (2002). In narrative, even chance, contingency and coincidence may not be what they seem; on this point, see Werner W o l f s article. See also Richardson (1997).
130
John Pier
intertextual nature, a semiotic reading may emerge only with multiple readings. 3. (Re)ordering Cause and Effect The possibility of illustrating these principles with particular texts is clearly limited within the scope of this paper. Even so, I wish to look briefly at a work that seems of emblematic interest for our purposes, suggestive of further avenues for analysis: James Joyce's A Portrait of the Artist as a Young Man (1916). At one level, this multifaceted novel lends itself to the fairly unobstructed reading of a story in the tradition of the Künstlerroman: Stephen Dedalus's schoolboy adventures at a Jesuit school; his first sexual encounters followed by repentance and his last-minute decision not to become a priest; rupture with his family and finally the blossoming of his artistic vocation at university. However, with closer attention to textual details and knowing that, initially at least, Joyce's intention was to structure his previously written "epiphanies" into an analogical ordering of five chapters, each subdivided into sections and with smaller portions of text being intricately interrelated among themselves, then the apparent flow of the narrative proves to be somewhat illusory. The text is in fact highly selective with regard to the incidents and chronological ordering of Stephen's story which, to a large degree, is structured as a succession of present moments. The result is that the cause-and-effect relations habitually associated with plot structure are quietly played down in the overall organization of Joyce's novel. A case in point is the first section of chapter I, slightly more than a page in length but acting as a "microcosm" that reverberates through much of the novel. Scenes spread out over a number of years during Stephen's infancy evoking the awakening of his five senses together with his growing sensitivity to his surroundings are compacted to such a degree that, for lack of temporal or any other form of deictic demarcation, with the boundaries further blurred by the use of free indirect discourse, these scenes could easily be taken for a single scene; but in fact, they are a series of distinct scenes, each narrated iteratively (i.e. narrated as though it occurred once, although it actually occurs several times), punctuated with "hypothetical" ellipses of indeterminate duration. 73
73
Cf. Genette (1980 [1972]: 109, 116-17).
After this, therefore because of this
131
In what ways are the features of this introductory section manifested in other parts of the novel? How does this page engage the processes of prospection and retrospection? Before looking at a few interconnected extracts, it should be pointed out that within the overall composition of the novel, this section represents not so much a "beginning" in the chronological sense as an ad hoc attempt from the vantage point of a narrating instance situated later in time to determine the causes from the effects. What appears on the opening page is thus constituted partly in function of what is portrayed later, and it is only in confronting that page with specific points in the subsequent text that narrative significance emerges. The first passage is from a song, apparently sung in the Dedalus household on several occasions: O, the wild rose blossoms On the little green place. He sang that song. That was his song. O, the green wothe botheth,74
Three words—rose, green, place—will take on an importance in subsequent parts of the novel that it is all but impossible to detect at this point. A few years later, at the Jesuit school, the teacher (a priest) says to Stephen and another student during an academic competition: —Now then, who will win? Go ahead, York! Go ahead, Lancaster! (12)
The reference is to the War of the Roses (1445-85) during which the Yorkists (supported by Ireland), whose emblem was the white rose (Stephen, leader of one team, wears a white rose), were defeated by the Lancastrians, whose emblem was the red rose (the leader of the other team, Jack Lawton [an English-sounding name], wears a red rose). The reverberations of this easily overlooked connection with historical fact throughout the novel are too numerous to be commented on here. In any case, they seem to be lost on the young Stephen, who instead muses about the "beautiful colours" of roses: Perhaps a wild rose might be like those colours and he remembered the song about the wild rose blossoms on the little green place. But you could not have a green rose. But perhaps somewhere in the world you could. (12)
74
Joyce (1968 [1916]: 7); page references cited hereafter in the text. Indispensable information will be found in the "Explanatory Notes" (ibid.: 484-550) and in Gifford (1982: 127-287).
132
John Pier
The "little green place" reappears, but also the improbable "green rose," while the conjecture about the possible existence of such a rose seems only to confirm a pattern set up earlier: "the green wothe botheth" of "his [Stephen's] song" superimposes "green place" on "wild rose" in such a way as to produce "green rose." At the same time, the grammatical ambiguity of "blossoms"/"botheth" (noun or intransitive verb?) is echoed on a larger scale: is "the green rose blossoms" to be taken as a noun phrase or as a sentence? But now one other detail must be accounted for, namely, that the song, popular in nineteenth-century Ireland, is misquoted. The original reads as follows: [ . . . ] N o w the wild rose blossoms O'er her little green grave [...]. 7 5
Clearly, this is something quite different from the song as we have it in the Portrait, which says: "the wild rose blossoms on the little green place. So what does all of this mean within the context of the novel? To answer this question (in part), it is necessary to go to the final section of the book (leaving out relevant intermediate steps, for economy's sake), where Stephen writes in his notebook: Crossing Stephen's, that is, my green [...]. (249)
Referred to here is St. Stephen's Green, located near University College, Dublin, St. Stephen (Stephen's namesake 76 ) being the first Christian martyr, stoned to death for blasphemy. With these various details in mind, it is now possible to draw a number of conclusions with regard to the convergence of the passages in question. First of all, it is clear that the grave in question is not the "green grave" of the popular song (i.e. a grave which is green), but rather a "grave on the green," i.e. a grave located on or at the green. Similarly, "green place" does not correspond to "green grave" but designates "St. Stephen's Green," the word "Green" being used in the sense of "place" or city square. On the other hand, St. Stephen's Green is not to be understood as the physical location of the saint's grave, but as the place where his martyrdom is commemorated. The accretion of meanings we see here together with their subsequent specifications—mere potentialities at the 75
Quoted in Gifford (1982: 133). For brevity o f the argument, I bracket out the surname Dedalus, charged with equally resonant overtones.
After this, therefore because of this
133
beginning of the novel—take form with what we have been calling a "semiotic" reading and are thus open to analysis only by taking account of the complementary perspectives arising out of earlier and later portions of the text. The second point is that, as is often the case in Joyce, polyvalent lexico-semantic and syntactic patterns of the type just examined are inextricably interwoven with the actions and happenings at the story level. Thus, the itinerary followed by Stephen, culminating in the affirmation of his vocation as an artist, is, metaphorically, a "blossoming" of "the green rose": in hindsight, "the green wothe botheth" of Stephen's first song proves to augur his future, the "cause" of that song being materialized in its "effects." The two sequences (lexico-semantic/syntactic and actional) converge in "Crossing Stephen's, that is, my green [...]," in other words, when Stephen lays claim to his name and thus to his identity as an artist. In the end, the blossoming of "the green rose" and the act of crossing St. Stephen's culminate in the dissociation of "green" from "grave," producing a configuration of meanings which coincides with the story of how Stephen progressively discovers the meaning of his name. In view of the criteria adopted for analysis of these extracts from the Portrait, what is perhaps most striking is that despite the number of pages between them (from five to well over two hundred), the linkages are demonstrable. This is so thanks to the strategies of prospection, retrospection and recognition mobilized in the course of heuristic reading and semiotic reading discussed in the previous section. One aspect of this schema is that it lays the ground for tracing the processes of abductive reasoning which, within narrative contexts, put the fallacy "after this, therefore because of this" into a particular light. Indeed, it would hardly be illuminating or even relevant to claim that the textual succession from "green grave" to "green place" to "St. Stephen's Green" or that of "the green wothe botheth" to "my green" mark a causal and chronological link within the narrative world of the novel. As we have seen, the filling of gaps takes place here nearly the other way around. 4. Conclusion Among the themes we have dealt with is causality in narrative, a subject that raises many questions not addressed in this paper. A fundamental question, causality is also, notes Brian Richardson, "one of the most ne-
134
John Pier
glected and under-theorized topics of narrative theory and criticism." 77 Richardson's monograph, itself a significant step toward overcoming this neglect and the unexamined ad hoc associations that causality has so often inspired among students of narrative, systematically explores various dimensions of the question, three of which are relevant to a more extensive treatment than it has been possible to undertake here. The first is that cause, together with time and space, forms an integral part of narrative setting. Second is the fact that cause is open to thematic treatment in narrative and can even appear as an ideological construct. Thirdly (and most important for Richardson's study) is that the causal settings of narrative, whether dominated by fate, providence, determinism, fortune, chance or necessity, break down into "four basic types of probability that govern fictional worlds: supernatural, naturalistic, chance, and metafictional systems of causation." 78 The emphasis, then, lies on causation as a constitutive element of narrative worlds and worldviews. In this paper, by contrast, the focus has been on the connection between causality and narrativity, that is, on how causes peculiar to stories are inferred in the processes set in motion by heuristic reading and semiotic reading. There is clearly a need to investigate the links between these complementary points of emphasis more closely. This being the case, however, the processes involved in inferring "after this, therefore because of this" are, in my view, more suitable to the theory of narrativity than Richardson's working definition of cause: "a condition that occasions a change in events." 79 This definition can be likened to Lakoff and Johnson's definition of "skeletal literal causation," 80 although it does not incorporate any further refinements or analysis of the various kinds of causation such as those outlined by these authors—which, in any case, are of a different order from the "relativized" and media-bound causes characteristic of narrative.
77 78 79
80
Richardson (1997: 14). Ibid.: 15. Ibid.: 36. "[A] cause is a determining factor for a situation, where by a 'situation' we mean a state, change, process, or action. Inferentially, this is extremely weak. All it implies is that if the cause were absent and we knew nothing more, we could not conclude that the situation existed. This doesn't mean that it didn't; another cause might have done the job. The only implication is entirely negative: Given a lack of such a cause and a lack of any other knowledge, we lack a justification for concluding anything" (Lakoff/Johnson [1999: 177]; emphasis in the original). It must be noted that this book appeared two years after Richardson's.
After this, therefore because of this
135
In her recently published Narrative Causalities, Emma Kafalenos adopts a more interpretive approach to causes, organized in large part around a methodology of "function analysis" derived from Propp. Here is no place to go into the complete argument of this rigorous work, but it can be noted that a number of points it develops bear on questions examined in the present contribution. Defining narrative as "a sequential representation of a sequence of events [in any medium]," 81 Kafalenos assigns primary importance to chronology: "This chronologically ordered, finite set provides the context in relation to which we interpret the causes and consequences of individual reported events. [...] all narratives unavoidably shape readers' (listeners', viewers') interpretations of the causes and effects of those events." 82 Chronology also plays an essential role in our own considerations. However, it is our contention that interpretations of causes and effects in narrative are determined chronologically only in part, and that to some extent they are governed by the fact that, as Aristotle puts it, we tend to "assume" that if Β happens after A, then Β happens because of A—an assumption subsequently to be confirmed, overthrown, in some way revised or even determined retroactively. Another point of interest lies in the interpretation of functions. According to Kafalenos, the function of an event is interpreted in relation to a configuration (in Mink's sense) at a given point in a narrative and is possibly reinterpreted progressively as the reading of the narrative progresses, culminating in a "complete configuration" or "final fabula" (although conflicting interpretations may give rise to "incomplete configurations"). 83 From the perspective of narrativity as outlined here, however, the question is not so much one of interpreting the causes and effects of events as a story progresses as it is that of the array of inferences—only some of which are causal—that can be drawn from the interlocking forms of abductive reasoning and the strategies of prospection and retrospection mobilized by heuristic reading and semiotic reading.
81 Kafalenos (2006: 2). An event is defined as "[s]omething someone does and something that happens [...]" (ibid.: 1). See the definition of event by Prince (2003 [1987]: 28): " A c h a n g e o f STATE m a n i f e s t e d in DISCOURSE b y a PROCESS STATEMENT in t h e m o d e o f
Do or Happen." Kafalenos (2006: viii) (emphasis in the original). Ibid.: 151-52. Interpretation is defined as "the process of analyzing the causal relations between an action or happening and other actions, happenings, and situations one thinks of as related" (ibid.: 1).
136
John Pier
These brief observations on research into the problem of causality in narrative, itself a crucial subject calling for further examination by theorists of narrative, also suggest that causality per se does not provide a sufficient basis on which to account for the dynamic functioning of narrative. It has been the aim of this paper to show that a more all-embracing perspective is provided by narrativity.
References Alber, Jan 2005 "Narrativization," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 386-87 (London and New York: Routledge). Allwood, Jens / Andersson, Lars-Gunnar / Dahl, Östen 1977 Logic in Linguistics. Cambridge Textbooks in Linguistics (Cambridge, etc.: Cambridge UP). Aristotle 1932 Peri Poetikes /Poetics. Bilingual edition, in S.H. Butcher, Aristotle's Theory of Poetry and Fine Art. 4 th ed. (1st ed. 1895), 1-111 (London: Macmillan). 1987 [1866] Aristotle on Fallacies; or The Sophistici Elenchi, translated with notes by Edward Poste (New York: Garland). 2004 [1908] Rhetoric, translated by W. Rhys Roberts (Mineola, NY: Dover Books). Baroni, Raphael 2007 La tension narrative. Suspense, curiositi et surprise. Coll. Poetique (Paris: Editions du Seuil). Barthes, Roland 1966 "Introduction ä Γ analyse structural des recits," in Communications 8: 1-27. 1974 [1970] S/Z: An Essay, translated by Richard Miller (New York: Hill and Wang). 1977 "Introduction to the Structural Analysis of Narrative," in R.B.: Image Music Text, translated by Stephen Heath, 79-124 (New York: Hill and Wang). Ben-Porat, Ziva 1976 "The Poetics of Literary Allusion," in PTL 1.1:129-51. Bremond, Claude 1973 [1972] "Le « modele constitutionnel» de A.-J. Greimas," in C.B.: Logique du recit. Coll. Poitique, 81-102 (Paris: Editions du Seuil). Eco, Umberto 1979 The Role of the Reader. Explorations in the Semiotics of Texts. Advances in Semiotics (Bloomington: Indiana UP). 1983 "Horns, Hooves, Insteps: Some Hypotheses on Three Types of Abduction," in The Sign of Three. Dupin, Holmes, Peirce, edited by Umberto Eco and Thomas A. Sebeok. Advances in Semiotics, 198-220 (Bloomington: Indiana UP). Fludernik, Monika 1996 Towards a 'Natural' Narratology (London and New York: Routledge).
After this, therefore because of this
137
Forster, Ε. Μ. 1965 [1927] Aspects of the Novel (Harmondsworth: Penguin Books). Garcia Landa, Jose Angel 2002 "Catastrophism and Hindsight: Narrative Hermeneutics in Biology and Historiography," in Beyond Borders: Re-defming Generic and Ontological Boundaries, edited by Ramon Plo-Alastrue and Maria Jesus Martinez-Alfaro, 105-19 (Heidelberg: C. Winter). Genette, Gerard 1966 "Frontieres du recit," in Communications 8: 152-63. 1968 "Vraisemblable et motivation," in Communications 11: 5-21. 1980 [1972] Narrative Discourse: An Essay in Method, translated by Jane E. Lewin. Foreword by Jonathan Culler (Ithaca, NY: Cornell UP). 1988 [1983] Narrative Discourse Revisited, translated by Jane E. Lewin (Ithaca, NY: Cornell UP). Genot, Gerard 1980 "Spatio-Temporal Aspects of 'Narrative Causality'," in G.G.: Problemes du calcul du recit IV: Quatre essais de Narrativique, 13-27 (Universite de Paris XNanterre: Centre de Recherche de Langue et de Littirature Italiennes). Gifford, Don 1982 [1916] Joyce Annotated: Notes for Dubliners and A Portrait of the Artist as a Young Man. 2 nd ed., revised and enlarged (1 st ed. 1967) (Berkeley, etc.: University of California Press). Greimas, Algirdas Julien 1970 [1969] "Eliments d'une grammaire narrative," in A.J.G.: Du sens. Essais semiotiques, 157-83 (Paris: Editions du Seuil). Greimas, Algirdas Julien / Courtis, Joseph 1979 Semiotique: Dictionnaire raisonne de la thiorie du langage. Langue - Linguistique - Communication (Paris: Hachette). Hamblin, C. L. 2004 [1970] Fallacies, Preface by John Plecnik and John Hoaglund, current bibliography by Michael F. Schmidt and Hans V. Hansen (Newport News, VA: Vale Press). Henault, Anne 1993 [1979] Les enjeux de la semiotique: introduction ά la simiotique. Formes semiotiques (Paris: Presses universitaires de France). Herman, David 2004 "Toward a Transmedial Narratology," in Narrative across Media: The Languages of Storytelling, edited by Marie-Laure Ryan. Frontiers of Narrative, 4 7 75 (Lincoln and London: University of Nebraska Press). Hiihn, Peter Forthcoming. "Events and Eventfulness," in Living Handbook of Narratology, edited by Peter Hühn et al. (Berlin and New York: Walter de Gruyter; Hamburg: Hamburg UP). Hühn, Peter / Kiefer, Jens 2005 The Narratological Analysis of Poetry: Studies in English Poetry from the 16'h to the 20fh Century. Narratologia 7 (Berlin and New York: Walter de Gruyter).
138
John Pier
Joyce, James 1968 [1916] A Portrait of the Artist as a Young Man. Text, Criticism, and Notes, edited by Chester G. Anderson. The Viking Critical Library (Harmondsworth: Penguin Books Ltd.). Kafalenos, Emma 2006 Narrative Causalities. Theory and Interpretation of Narrative (Columbus: The Ohio State UP). Labov, William / Waletzky, Joshua 1967 "Narrative Analysis: Oral Versions of Personal Experience," in Essays on Verbal and Visual Arts, edited by June Helm, 12-44 (Seattle: University of Washington Press). Lakoff, George / Johnson, Mark 1999 Philosophy in the Flesh: The Embodied Mind and its Challenge to Western Thought (New York: Basic Books). Lukasiewicz, Jan 1957 Aristotle's Syllogistics from the Standpoint of Modern Formal Logic. 2 nd enlarged ed. (1 st ed. 1951) (Oxford: Clarendon Press). Mackie, J. L. 1967 "Fallacies," in The Encyclopedia of Philosophy, edited by Paul Edwards, vol. 13: 169-79 (New York: The Macmillan Press and The Free Press; London: Colliers-Macmillan). Mink, Louis O. 1970 "History and Fiction as Modes of Comprehension," in New Literary History 1.3: 541-58. Nünning, Ansgar 2003 "Narratology or Narratologies? Taking Stock of Recent Developments, Critique and Modest Proposals for Future Uses of the Term," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratoiogia 1: 239-75 (Berlin and New York: Walter de Gruyter). Nünning, Vera / Nünning, Ansgar (eds.) 2002 Erzähltheorie transgenerisch, intermedial, interdisciplinary. WVT-Handbücher zum Literaturwissenschaftlichen Studium 5 (Trier: WVT). Ochs, Elinor / Capps, Lisa 2001 Living Narrative: Creating Lives in Everyday Storytelling (Cambridge, Mass. and London: Harvard UP). Pavel, Thomas 1986 Fictional Worlds (Cambridge, Mass. and London: Harvard UP). Pier, John 1980 "Some Implicit Principles of 'Minimal Narrative'," in Kodikas / Code. An International Journal of Semiotics 3: 321-35. 2003 "On the Semiotic Parameters of Narrative: A Critique of Story and Discourse," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratoiogia 1: 73-97 (Berlin and New York: Walter de Gruyter).
After this, therefore because of this 2004
139
"Narrative Configurations," in The Dynamics of Narrative Form. Studies in Anglo-American Narratology, edited by John Pier. Narratologia 4: 239-68 (Berlin and New York: Walter de Gruyter). Prince, Gerald 1973 A Grammar of Stories: An Introduction. De Proprietatibus, Series minor 13 (The Hague and Paris: Mouton). 1982 Narratology: The Form and Functioning of Narrative. Janua Linguarum, Series maior 108 (Berlin, etc.: Mouton). 2003 A Dictionary of Narratology. Revised ed. (1 st ed. 1987) (Lincoln and London: University of Nebraska Press). 2005 "Narrativity," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 387-88 (London and New York: Routledge). Propp, Vladimir 1968 [1928] Morphology of the Folktale, translated by Svatava Pirkova-Jakobson, Preface by Louis A. Wagner, Introduction by Alan Dundes. 2 nd revised ed. (1 st ed. 1958) (Austin and London: University of Texas Press). Richardson, Brian 1997 Unlikely Stories: Causality and the Nature of Modern Narrative (Newark, NJ: University of Delaware Press; London: Associated University Presses). 2005 "Causality," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 48-52 (London and New York Routledge). Ricceur, Paul 1985 [1984] Time and Narrative, vol. 2. Translated by Kathleen McLaughlin and David Pellauer (Chicago: The University of Chicago Press). Rimmon-Kenan, Shlomith 2001 Narrative Fiction: Contemporary Poetics. 2 nd ed. (1 st ed. 1983) New Accents (London and New York: Methuen). Ryan, Marie-Laure 1992 "The Modes of Narrative and Their Visual Metaphors," in Style 26.3: 368-87. 2005a "On the Theoretical Foundations of Transmedial Narratology," in Narratology beyond Literary Criticism: Mediality, Disciplinarity, edited by Jan Christoph Meister. Narratologia 6: 1-23 (Berlin and New York: Walter de Gruyter). 2005b "Tellability," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 589-91 (London and New York: Routledge). Ryan, Marie-Laure (ed.) 2004 Narrative across Media: The Languages of Storytelling. Frontiers of Narrative (Lincoln and London: University of Nebraska Press). Salmon, Wesley Charles 1963 Logic (Englewood Cliffs, NJ: Prentice-Hall). Schaeffer, Jean-Marie 1999 Pourquoi la fiction? Coll. Poetique (Paris: Editions du Seuil).
140
John Pier
Schmid, Wolf 2003 "Narrativity and Eventfiilness," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 17-33 (Berlin and New York: Walter de Gruyter). Schönert, Jörg / Hühn, Peter / Stein, Malte 2007 Lyrik und Narratologie. Text-Analysen zu deutschsprachigen Gedichten vom 16. bis zum 20. Jahrhundert. Narratologia 11 (Berlin and New York: Walter de Gruyter). Scholes, Robert 1982 Semiotics and Interpretation (New Haven and London: Yale UP). Sternberg, Meir 1978 Expositional Modes and Temporal Ordering in Fiction (Bloomington and Indianapolis: Indiana UP). 1983 "Mimesis and Motivation: Two Faces of Fictional Coherence," in Literary Criticism and Philosophy, edited by Joseph P. Strelka. Yearbook of Comparative Criticism, Vol. X: 145-88 (University Park, PA and London: The Pennsylvania State UP). 1985 The Poetics of Biblical Narrative: Ideological Literature and the Drama of Reading. Indiana Studies in Biblical Literature (Bloomington: Indiana UP). 1990 "Telling in Time (I): Chronology and Narrative Theory," in Poetics Today 11.4: 901-48. 1992 "Telling in Time (II): Chronology, Teleology, Narrativity," in Poetics Today 13.3:463-541. 2001 "How Narrativity Makes in Difference," in Narrative 9.2: 116-22. 2006 "Telling in Time (III): Chronology, Estrangement, and Stories of Literary History," in Poetics Today 27.1: 125-235. Sturgess, Philip J. M. 1992 Narrativity: Theory and Practice (Oxford: Clarendon Press). Walton, Douglas 1995 A Pragmatic Theory of Fallacy. Studies in Rhetoric and Communication (Tuscaloosa and London: The University of Alabama Press). Wolf, Werner 1999 The Musicalization of Fiction: A Study in the Theory and History of Intermediality. Internationale Forschungen zur Allgemeinen und Vergleichenden Literaturwissenschaft 35 (Amsterdam and Atlanta, GA: Rodopi). 2003 "Narrative and narrativity: a narratological reconceptualization and its applicability to the visual arts," in Word & Image 19.3: 180-97. 2004 "'Cross the Border — Close that Gap': Towards an Intermedial Narratology," in EJES — European Journal for English Studies 8.1: 81-103. Woods, John / Walton, Douglas 1989 [1978] "Post Hoc, Ergo Propter Hoc," in J.W. and D.W.: Fallacies: Selected Papers 1972-1982. Studies of Argumentation in Pragmatics and Discourse Analysis, 121-41 (Dordrecht and Providence, RI: Foris Publications).
PETER HÜHN (Hamburg)
Functions and Forms of Eventfulness in Narrative Fiction Most structuralist narratological models stipulate mediacy and sequentially or, in other words, perspectivity and temporality as the basic constituents of narrativity. These two dimensions seem to be inherent—in various ways—in the binary opposition (and their three-level or four-level extensions) underlying the majority of narratological models such as histoire/ricit, story/discourse, fabulalsjuzet, story/text. 1 Strictly speaking, this two-dimensional definition refers to what some authors term "narrativeness" or "narrativehood," a dichotomous category concerning the question of whether a text is or is not a narrative, as distinct from narrativity in a narrower sense, which they define as a question of degree: texts may be more or less narrative. 2 Of these two parameters, however, sequentiality alone possesses a differential quality, since all discourse (description, argumentation or explanation as much as narration) necessarily presupposes some form of mediation and perspectivising, narratives being distinguished from other text types by the temporal organisation of what is mediated. 3 As the organising principle(s) for the sequential dimension, narratologists usually name chronology, coherence and/or causality. 4 Thus, Genette (1980 [1972]) introduces mainly formal terms for sequentiality on the level of discourse: order, duration and frequency. Bal describes fabula as the logical and chronological organisation of events, the progression from possibility via realisation or event to result. 5 For Rimmon-Kenan, chro1
2
For a critical overview o f these parameters of narrative and a useful discussion of their theoretical difficulties and problems, cf. Pier (2003). E.g. Prince (1999: 43-44); Herman (2001: lOOff.). See also the distinction made by Prince in his contribution to the present volume.
3
See e.g. Sternberg (2001: 115-16). For a general overview, cf. Schmid (2003), who contrasts classical and structural narrative theories as concentrating either on mediacy or on sequentiality (temporality) on the story-level with respect to a definition of narrativity.
5
Bal (1997 [1985]: 5).
142
Peter Hühn
nology is sufficient as an ordering principle for a minimal story: causality, she argues, can be reformulated in chronological terms, while for a description of sequentiality on the text-level, she refers merely to Genette's formal categories. 6 Chatman defines narrativity by a doubly temporal logic ("story'V'discourse") along with the additional principle of causality, typically concentrating on mediacy rather than sequentiality, which goes for a great number of other narratological approaches as well. 7 Martinez and Scheffel emphasise coherence and connectedness as a formal feature. 8 Several theorists, however, posit change as an additional requirement for narratives, mostly in general formalist terms, such as the progression from conjunction via disjunction to a new conjunction, 9 the succession of equilibrium, disequilibrium and equilibrium, starting with equilibrium or with disequilibrium and thereby constituting either a story of punishment or conversion, respectively, 10 or the logical sequence of possibility, process and outcome, with potential bifurcations at each stage and two different directions: improvement or deterioration." To name some further models, partly indicative of differences in approach: Labov distinguishes five stages (on the discourse level): orientation, complication, evaluation, conclusion, coda; 12 Abbott differentiates between event and series of events and between constituent and supplementary events without further specification; 13 Sternberg's description of the sequential dimension—the interplay of the three universal narrative effects of prospection, retrospection and recognition (or suspense, curiosity, surprise)—is primarily a definition of change in terms of the temporality of reader perception. 14 Toolan refers to the movement of the trajectory of incidents towards closure. 15 Some models specify this feature of change as a violation of expec6 7 8
10
12 13 14 15
Rimmon-Kenan (2001 [1983]: 17-19). Chatman (1990: 9). Martinez/Scheffel (1999: 25). Cf. Greimas for whom "narrativity [is] the irruption of the discontinuous into the discoursive [sic] permanence of a life, a story, an individual, a culture, disarticulates that discoursive permanence into discrete states between which it sets transformations" (1987 [1983]: 104). Todorov (1977 [1968]: 119). Bremond (1973: 134 passim). Labov (1972: 369). Abbott (2002: 20-22). Sternberg (2001); cf. also Sternberg (1990), (1992). Toolan (2001: 4-8).
Functions and Forms o f Eventfulness in Narrative Fiction
143
tations, connecting it with the notion of tellability. 16 Thus Pratt, drawing on Labov's (1972) concept of "natural narrative" and Grice's "Cooperative Principle," concentrates on the component "evaluation" as the central dimension for defining relevance and the "point" of what is being told, 17 stressing that the point typically consists in the unusual, the problematic, the wonderful, the amusing, the terrifying, etc., 18 in close connection with the quality of "tellability." Herman argues that narrativity is gradable, its degree being dependent on the violation of expectations, while violation is closely connected with the concepts of tellability and "point." 19 I would like to pursue these suggestions further towards a more precise specification of the criteria of narrativity, arguing firstly that readers and listeners typically demand a particular "point," namely some "event" or decisive turn in the sequential organisation of a text in order to consider it as properly narrative, and secondly that this additional requirement of eventfulness is context-sensitive and consequently culturally as well as generically specific and historically variable. 20 1 will then turn to an analysis of eventfulness in two literary examples taken from different periods, revealing different forms and degrees of eventfulness. First, however, two of the terms mentioned here require further clarification. 1) The notion of narrativity as used in this paper combines the two senses differentiated by Sternberg (2001) and Prince (1999) in the opposition between binary narrativehood and gradational narrativity: eventfulness as the differential criterion of narrativity is both a binary category (a text is either narrative or not narrative, i.e. either features or does not feature events) and a scalar category (texts can be more or less narrative and thus rank higher or lower on the scale of eventfulness). 2) The context-sensitivity of eventfulness is a complex phenomenon which comprises the following different aspects and needs to be specified accordingly when analysed with respect to concrete narratives: the relevance—to the event—of the social and cultural setting depicted in the 16 17 18 19
See Bruner (1991: 11-13). Pratt (1997: 4 6 - 4 7 , 63ff.). Ibid.: 137, 1 4 0 ^ 1 . Herman (2002: 9 0 - 9 2 , 100-03). The present essay is closely linked to Schmid's (2003) article on "Narrativity and Eventfulness," originating in the work being carried on in project PI I: "Ereignis und Ereignishafligkeit in der englischen und russischen Literatur aus kulturhistorischer Perspektive" ["Event and Eventfulness in English and Russian Literature from the perspective of cultural history"] within the Hamburg Research Group "Narratology" funded by DFG. Project Ρ11 is conducted by Wolf Schmid and myself.
144
Peter Hühn
text; the relation of the event to social and cultural (or literary) phenomena outside the text; and the status of such an event within the contemporary world (whether common, rare or new). One particularly important type of context consists of other literary texts which may serve as a frame of reference for the constitution of eventfulness in a narrative. 1. Modelling Eventfulness: Schema Theory and Lotman's Concept of Sujet Intuition and empirical evidence suggest that with most, if not all narrative text-types and genres, recipients expect some decisive change or turn and that a narrative lacking this crucial "point" will elicit the bored question "So what?" dreaded by every storyteller. Such a "point" generally functions as the raison d'etre of a narrative, constituting its "tellability" or "noteworthiness." This obviously goes for jokes, anecdotes, gossip and everyday conversational story-telling about personal experiences, but also, and more emphatically, for the great variety of literary narratives extending from highly schematised popular genres (including detective, crime and spy fiction as well as love and adventure stories) to canonical novels, short stories and ballads and even for drama and poetry.21 Critics and theorists have addressed the feature of tellability in narratives from a variety of perspectives. Pratt22 and Prince23 apply this notion to literary texts. Prince discusses message and point as part of narrative pragmatics, linking the term "point" (pointless vs. pointed) to the phenomenon of relevance and stressing its context-dependence under two aspects: for the text and for the receiver. The need for a reconstruction of the context for the receiver is due to the fact that literary texts are not closely tied to the context of their production.24 Fludernik refers to the dynamics between narrativity and tellability or point, drawing on Labov and other discourse analysts.25 Though based on her specific concept of natural narratology and experientiality, her thesis that events ("point") do not constitute narrativity in themselves but through their emotional and
21 22
24 25
Cf. Hühn/Kiefer (2005); Schönert/Hühn/Stein (2007). Pratt (1977: 136, 144ff.). Prince (2003 [1987]: 97, 83). See also Prince's contribution to the present volume. Prince (1983: 534). Fludernik (2003b: 245^t6).
Functions and Forms o f Eventfulness in Narrative Fiction
145
evaluative overload can be linked to the approach adopted further on in the present paper for a description of eventfulness. The requirement of a "point" for narratives has been most extensively discussed by scholars analysing everyday or "natural" storytelling. 26 Polanyi seeks to answer the question "What do Americans tell stories about?", using point (the noteworthy, the narratable, the interesting) as the main focus. 27 The point of narrative, she argues, emerges from context reference, which is dependent on narrative or event structure (time), descriptive structure (material) and, most importantly, evaluative structure. She further distinguishes between the culturally interesting (with the broadest appeal to the entire culture), the socially interesting (appeal to a group) and the personally interesting, noting that the interesting (in America) can be equated with the odd or the unexpected, the violation of a norm, 28 although from a cultural perspective, it is highly variable. Pratt discusses this point by applying Grice's maxims to literature, especially to novels, and she analyses the "display text" (i.e. the thematisation of the unusual), tracing the similarity between novels and natural narratives in this respect to the notions of the detachability of the display text and its susceptibility to elaboration. 29 The common premise underlying these various approaches consists in the assumption that something more crucial than mere succession and change is demanded for proper narratives: an unexpected, exceptional or new turn in the sequential dimension, some crucial departure from the established course of incidents, 30 what Bruner succinctly refers to in his formula: "canonicity and breach." 31 Such decisive or crucial turns I will call "events." Thus, tellability (and ultimately narrativity) can be said to depend on eventfulness. To be sure, other features of the narrative text 26 27 28 29
31
E.g. Labov (1967), (1972); Ochs/Capps (2001); Prince (1983). Polanyi (1979: 209). Ibid.: 2 1 1 - 1 2 . Pratt (1977: 51ff., 136, 144-48). There appears to exist another broad type of narrative, what I will tentatively call "process narration," which is employed as a more descriptive and neutrally informative way of tracing and communicating developments, processes or changes without necessarily raising expectations of surprising or unpredictable turns or deviations. Process narration seems to be used, for example, in the natural sciences, in historiography, in lawsuits, but also in such informative genres as recipes and instruction manuals (to be sure, historical developments and court cases may take dramatic turns, but need not do so, and are thus not a necessary prerequisite). Bruner (1991: 11-13).
146
Peter Hühn
(including the particular choice and rendering of characters, incidents and setting, the specific style, use of wit and irony, etc.) also contribute to its tellability, with the occurrence of at least one event, however, remaining indispensable in all cases. 32 As for the definition of eventfulness, those narratological models that posit change as an additional constituent of narrativity, e.g. the models of Greimas (1987 [1983]), Bremond (1973), Todorov (1977 [1968]) or Sternberg (2001), tend to be too general or lacking in specificity to account for the structure and status of what constitutes a change. 33 Further specifications are required with respect in particular to what counts as a decisive change and whether the measure of decisiveness can be determined on a purely textual basis. The most elaborate proposal so far for a definition of eventfulness is Wolf Schmid's (2003) list of seven criteria. Schmid names two absolute preconditions: reality and resultativity, i.e. changes must actually occur and they must be complete in order to qualify as events. He adds five relative or gradational criteria: relevance, unpredictability, persistence, irreversibility and non-iterativity, i.e. changes can be more or less relevant, unpredictable, persistent, irreversible and singular, and their degree of eventfulness will vary accordingly. 34 I would like to supplement and contextualise such a definition of eventfulness by elaborating an analytical approach which combines two concepts of modelling sequentiality in narrative texts: cognitivist schema theory and Jurij Lotman's semiotic model of the artistic text. Both concepts stress the semantic dimension of eventfulness as well as its correlation to textual and extra-textual order and value systems, thereby allowing for a precise explication of the relation of event to context. Also relevant in this regard are Pratt's emphasis on the context-dependence of relevance, pointedness and thus, by implication, eventfulness, although she does not use the term, referring instead to Labov's notion of the "unspoken permanent
32
Cf. the general overview of dimensions and aspects of tellability provided by MarieLaure Ryan in her entry in Routledge Encyclopedia (2005): 589-94. For an overview of different types of event in the sense of change, see David Herman's entry "Event and Event-Types" in Routledge Encyclopedia (2005): 151-52. Such a notion of event as (mere) change has to be clearly distinguished from the more specific concept of event proposed in the present article: event as a decisive and unexpected change. Schmid develops these criteria mainly on the basis of Chekhov's poetics.
Functions and Forms of Eventfulness in Narrative Fiction
147
agenda." 35 Similarly, Culler and Bruner extensively discuss the context sensitivity of narrative meaning and that of genre. 36 Schema theory 37 enables us to show how recipients make sense of narratives, as of any other kind of text, 38 by drawing on their world knowledge as determined by cognitive structures and semantic patterns existing in their minds and hence already meaningful. Schank/Abelson, for instance, introduce a distinction among scripts: stereotypic sequences of actions or incidents of a situational, personal or instrumental kind; plans'. procedure of achieving goals, conducted by actors; and themes: similar to what is normally called a [thematic] frame—the situational context of a person's behaviour—provides valuable insight into the relations between these categories when he states: "I assume that definitions of a situation are built up in accordance with principles of organization which govern events—at least social ones—and our subjective involvement in them: frame is the word I use to refer to such of these basic elements as I am able to identify." 39 Among these schematic patterns, two broad types are particularly useful for the analysis of narratives: firstly, static schemata or frames, i.e. situational or thematic contexts into which a text is placed so as to be understood; and secondly, with particular importance for narratives and their plots, dynamic schemata or scripts, i.e. sequential or procedural patterns which underlie or organise a course of action or a succession of incidents. Scripts describe the abstract structures of stereotypic sequences of actions (as, for example, the prototypical steps of a conventional courtship procedure) to which plots (for instance, the love story in a novel) may refer, but from which they will deviate to a greater or lesser degree on account of the concrete circumstances of setting and characters. A narrative text that conforms closely to a schema is not noteworthy and thus not eventful, since such a text only reproduces what is already known and expected. Eventfulness thus involves departure from a schematic pattern or script activated in the text.40 Moreover, it is clear that departure from an established pattern may vary in degree by dint of being more or less unexpected or exceptional, so that the concept of eventfulness is nec-
35 36 37
39 40
Pratt (1977: 72-73). Culler (1975: 113ff„ 140ff.); Bruner (1991: 14-17). See also Bakhtin (1981: 259-422). Cf. e.g. Schank/Abelson (1977); Schank (1990); Herman (2002); Cook (1994). The same goes for authors, of course, in the act of producing the text. Goffman (1986 [1974]: 10-11). See Herman (2002: 85-86).
148
Peter Hühn
essarily gradational. 41 Also to be taken into account are two important conceptual consequences entailed by cognitivist theory. Schemata (frames and scripts) are not inherent in textual structures in any ontological sense, but must be inferred by the reader from specific cues or signals in the text and activated in his mind on the basis of his world knowledge and correlated to the narrative. 42 In addition, relevant schemata will vary both diachronically and synchronically, the analysis of eventfulness thus involving identification and specification of those schemata that can be shown to be culturally, historically and generically germane: what counts as an event will thus have to be assessed with reference to the context in terms of genre, 43 culture, social group, historical period and author. Furthermore, since the actual relevance of schemata and the status of events depend on the consciousness of a perceiving subject for which a schema is relevant and, consequently, an incident appears to be eventful, eventfiilness will vary in relation to the entity or level in the textual set-up (character, narrator, author or reader). Thus, a development may be eventful for a character but not for the narrator and the reader or vice versa. The cognitivist description of eventfulness and tellability can be supplemented by Lotman's concept of sujet, which grants central importance to the notion of event. 44 Lotman, too, construes eventfulness as a departure or deviation from a norm, the violation of an established order. He describes the basic order typically underlying a narrative text by using such metaphorically spatial terms as "semantic field" with a boundary separating the scope of a particular set of norms and values from the domain of a different order with opposite norms. Crossing the boundary is normally impossible for, or prohibited to, the figures within this sub-field, and if it occurs, brought about by the protagonist (who thereby becomes a mobile figure), this fact is something noteworthy or significant in itself and thus constitutes an "event." The elaboration of this semantic field and its binary subdivision is always culturally and historically specific. Like 41
44
Cf. Herman's (2002: 86, 91, lOOff.) scalar definition of "narrativity" in contrast to the binary category of "narrativehood." Schmid (2003) also conceives of eventfulness as a gradational category. See Herman (2002: 91, 95-96): the text cues recipients to activate certain kinds of world knowledge. See also Todorov (1977 [1968]) and Barthes (1977 [1966]). Herman (2002: 105) describes genres as script-based macrodesigns; Pratt (1977: 86) speaks of genres in terms of communicative conventions and felicity conditions See also Schaeffer (1989). Lotman (1977 [1970]); cf. Shukman (1977).
Functions and Forms of Eventfulness in Narrative Fiction
149
Schmid, Lotman also defines events in gradational terms, for the extent of "resistance" a boundary puts up to being crossed (i.e. its degree of inviolability) conditions the degree of eventfulness. These categories enable the analyst to explicate the relation between the narrative text and its social or cultural context(s), thus determining the relevance and semantic significance of the narrated change. Lotman's model allows for the conceptualisation of successive and progressive events, but also of regressive events. The protagonist may become immobile in the new subfield or progress further, in which case the semantic field is re-defined, the previous second sub-field changing into a new first sub-field, which in turn is delimited by another boundary that the protagonist, if he/she continues to be mobile, may cross and so forth and so on. But it is also possible for the protagonist, having crossed the boundary, to retrace his/her steps, as it were, re-entering the initial field and revoking or cancelling the event. The two approaches—the cognitivist and the semantic—complement each other in that schema theory offers a rigorous conceptualisation of the sequential coherence of actions, happenings and incidents, while Lotman provides a clear description both of the semantic framework and of the decisive turn within the sequential dimension as well as a more explicit definition of the status of the event in cultural and historical terms. Systematic research into the cultural and historical development of what counts as an event has been lacking so far. 45 As a modest contribution to such a research program, but also to illustrate the approach outlined above, I will now discuss two examples representing different forms and degrees of eventfulness conditioned by widely differing contexts: Samuel Richardson's Pamela; Or, Virtue Rewarded (1740), a prototypically eventful novel which initiated a novelistic convention centring on a particular type of decisive change or "event," and James Joyce's short story "Grace" (1914), set in quite another socio-cultural context and incorporating conventions that surreptitiously problematise notions of eventfulness. 46
45
The need for such a research programme into diachronic and synchronic variability is mentioned by Herman (2002: 107-13; see also 86) and Fludemik (2003a). My analyses o f these texts will focus exclusively on the structure o f the plot, the question of its eventfulness and its dependence on the cultural and historical context. They are not meant to be comprehensive interpretations.
150
Peter Hühn
2. Completed Eventfulness: Richardson's Pamela In Richardson's Pamela, the "semantic field" or overall context is defined primarily in social terms and is subdivided hierarchically by a strict class boundary into the aristocracy, represented by Mr. B. and his family (his mother Lady B., recently deceased, and his sister, Lady Davers), on the one hand, and on the other, the lower classes, having to work for their living, represented by the servants—among them, fifteen-year-old Pamela Andrews and her hard-working poor parents, but also the housekeeper, Mrs. Jervis, and the clergyman, Mr. Williams. All characters (with one exception) are immobile in that their positions are firmly and unquestionably fixed in their respective fields. Only Pamela's place in the lowerclass sub-field turns out to be unstable and insecure from the start. This is due to two aspects of her condition, both connected with her sex and age: her role as an unprotected servant girl in an aristocratic family after her benevolent mistress's (Lady B.'s) death and her quality as a sexually attractive nubile young woman. As far as her relation to the other sex is concerned, two scripts are available to Pamela (according both to social norms and literary conventions): marrying someone from her own class or being seduced by a member of the aristocratic class.47 Whereas the first alternative is apt to complete and consolidate her social integration in the first field, the second option, while seeming to permit the crossing of the boundary into the higher aristocratic sub-field, would eventually result in her moral and social degradation and cause her to be expelled from her original social position into a morally inferior second field (she would be "ruined," as the usual phrase puts it). In the course of the development of the novel, these two alternative scripts are in fact presented to Pamela as practical possibilities in the form of two different plotlines. The main plotline is constituted by Mr. B., the libertine son of her former mistress, persistently and perfidiously trying to seduce her, against her strong virtuous resistance, through flattery, the use or threat of force and, ultimately and most elaborately, by tempting her with the promise of an established position in aristocratic circles as his kept mistress with the prospect of marriage sometime in the future. At one point in this plot progression, she is offered marriage by a member of her 47
These two scripts actually underlie the plot patterns o f two types of the female novel of love prevalent at the time: the seduction/rape tale and the courtship novel. See Doody (1974: 18ff.).
Functions and Forms of Eventfulness in Narrative Fiction
151
own class, Mr. Williams, 48 as a means of warding off the threat to her moral integrity and securing a safe and firm position for her in the first field. With respect to the second alternative, the seduction plotline and the concomitant prospect or illusion of a boundary crossing, the novel provides explicit definitions of the boundary from either side, as it were, corroborating the rigorous separation between the two sub-fields. At the very beginning of the novel, Pamela's parents, warning her against Mr. B.'s designs, specify the constitutive norms of the lower-class sub-field in opposition to the corrupt standards of the aristocracy as moral integrity and religious strictness: honesty, goodness, hard work, virtue, piety and trust in God and his Providence. Virtue is considered superior to wealth and even to life: "we had rather see you all covered with rags, and even follow you to the church-yard than have it said, a child of our's preferred any worldly conveniences to her virtue." 49 While the parents thus define the boundary from below, Lady Davers, later in the novel, demarcates it from above, again with direct reference to Pamela's person and position, when her brother contemplates taking her as either a mistress or a wife: "ruining" her would be wicked, she writes, but marrying her would be "utterly inexcusable," a gross violation of the superior state of the noble and old family tradition: "[our family] is as ancient as the best of the kingdom; and for several hundred years, it has never been known, that the heirs of it have disgraced themselves by unequal matches" (293); "[a] handsome man, as you are in your person, so happy in the gifts of your mind; and possessed of such a noble and clear estate; and very rich in money besides, left you by the behest of fathers and mothers, with such ancient blood in your veins, untainted. I cannot bear to think of your thus debasing yourself." (293-94) Because of their literary and social conventionality, the practical realisation of either of these two scripts would not rank high on the scale of eventfulness. Marrying Mr. Williams, on the one hand, would merely confirm and perpetuate Pamela's original membership by birth of the first sub-field, corroborate the existing order and thus prove not eventful at all, which is highlighted with the explicit advocacy of such a marriage by the two staunch defenders of the inviolable boundary between the fields mentioned above: Pamela's father (198) and Lady Davers (294). If Pamela 48 Though Mr. Williams's position as a clergyman is somewhat higher up the social scale than that o f the servant girl Pamela, the society of the novel clearly places them in the same class, far beneath the aristocracy. Richardson (1980 [1740]: 46); page references cited hereafter in the text.
152
Peter Hühn
were to succumb to seduction, on the other hand, this would ultimately lead to her moral and social debasement, thereby undermining her place in the first field and relegating her to an altogether inferior position. This course of action would constitute a negative event, but one of a relatively low degree on account of its common occurrence, both in literature and in life. For a variety of reasons, however, Pamela refuses to follow either of the two conventional scripts. The seduction script is resolutely and persistently rejected by her on purely moral grounds in spite of her unconscious but growing love for Mr. B., and even after she gradually becomes aware of it: " Ό sir, [...] my heart will burst; but on my bended knees I beg you to let me go tomorrow [...]: and don't offer to tempt a poor creature, whose whole will would do yours, if innocence would permit'" (116). 50 Mr. Williams's marriage proposal she declines because, as the reader may gather from her behaviour, of her latent strong attachment to her "master," even though she explicitly states as her reason that she has no plans for marrying yet: "There is not a man living that I desire to marry. To keep myself honest, and to be a comfort and assistance to my poor parents, is the very top of my ambition" (181). Pamela's rejection of the two conventional scripts and concomitant roles eventually enables her to cross the normally impassable rigid class boundary, i.e. marry Mr. B. and thereby move into the superior field of the aristocracy—an event prepared for over a long period of time and finally brought about as a result largely of decisive changes in Mr. B.'s attitude: conception of love for Pamela (115); offer of cohabitation with a vague prospect of marriage after a test year (230-31); recognition from her journal and letters of her unselfish loving attachment to him in spite of his brutal and deceitful behaviour (285-86); and finally, unreserved declaration of love and proposal of marriage (295ff, esp. 307) as well as acknowledgement of her moral superiority as a model for his own life (305ff): "let me tell my Pamela, that, after having long been tossed about by the boisterous winds of culpable passion, I am not now so much the admirer of your beauty, all charming as you are, as of your virtue. My love therefore must increase, even should this perishable beauty fail, as the station of life you are now entering upon, will afford you augmented opportunities to display your virtue" (372). In the last analysis, the crossing of the boundary is effected essentially through overriding social with moral norms. 50
See also Pamela's vigorously outspoken rejection of Mr. B.'s "proposals" (227ff.).
Functions and Forms o f Eventfulness in Narrative Fiction
153
Although Pamela's transition from the first to the second semantic field, i.e. from the lower to the higher social status, is ultimately achieved with the brief but legally and religiously binding ceremony of the wedding in the chapel (374-75), the boundary is crossed not as a sudden or random occurrence, but as an integral part of a protracted process which, in addition to the preliminary psychological and moral changes, includes the customary wedding preparations and especially the long drawn-out efforts to overcome various serious obstacles resulting from family and social circumstances, such as Lady Davers's hostile opposition and the neighbouring gentry's reservations or outright antagonism. Nor is the achievement definitive and final: the two-volume sequel of the novel published a year later (1741) reveals that Pamela's position in the second sub-field is not secure yet, but has to be earned and further confirmed through her exemplary behaviour. All these difficulties testify to the inherent strength and stability of the subdivision of the semantic field and the firm resistance of the boundary to transgression. Therefore, the fundamental transformation of Pamela's social status represents a very high degree of eventfulness, since it constitutes a grave violation of the hierarchical social norms in favour of emotional and moral values as well as a radical deviation both from the seduction script and the conventional marriage script.51 What is socially and culturally significant about Pamela's eventful transgression of the boundary is her seeming passivity. She does not actively aspire to such a movement, but is granted its achievement as a reward for her active efforts in another respect, notably her unwavering rejection of Mr. B.'s advances and her moral exemplariness in connection with an underlying intuitive confidence in Mr. B.'s fundamental goodness. And even after his proposal of marriage and the subsequent preparations for firmly establishing her as his wife in the second field, Pamela continues to stress her humbleness, unworthiness and inferiority, invariably referring her social elevation to God's Providence (cf. e.g. 309ff., 364ff., 375-76). In keeping with her persistent reactivity, she discovers and acknowledges her love for Mr. B. only after he has first declared his love to her (307).
51
Deviation from the conventional marriage script is particularly striking when Mr. B. chooses Mr. Williams, whose marriage proposal to Pamela could have resulted in Pamela's integration into the first sub-field, to officiate at the wedding ceremony ( 3 4 2 - 4 3 , 374ff.).
154
Peter Hühn
The high degree of eventfulness represented by this marriage in violation of the rules is evidenced by contemporary reactions to Pamela. Satirical and polemical attacks on the novel, such as Henry Fielding's Apology for the Life of Mrs. Shamela Andrews (1741) and Eliza Haywood's AntiPamela: or Feign'd Innocence Detected (1742), question the protagonist's artlessness, naivety and passivity, accuse her of hypocrisy, scheming and manipulation and, thereby radically discrediting her morally, 52 seek to invalidate the eventfulness of Pamela's marriage and social rise by ascribing its achievement to dissimulation and deceit. These hostile reactions to Pamela are clearly formulated from an upper-class position (Fielding was an aristocrat) and thus indirectly testify to the context-dependence of the event in the novel. 53 Something similar goes for positive reactions to Richardson's novel, such as its many imitations and continuations (e.g. John Kelly's Pamela's Conduct in High Life [1741]),54 which conspicuously often give the heroine an aristocratic birth (initially unknown, but revealed later on). As a result, the high degree of eventfulness of Pamela's marriage is considerably reduced, apparently because the radical deviation from the established social order was felt to be unacceptable—another indirect indication of the dependence of the event on the class context. That such a marriage across the class-divide is both highly exceptional 55 and of profound relevance becomes apparent only through reference to the historical context, i.e. the social, cultural and religious conditions and practices which were current at the time of the novel's writing. Central contextual features are, firstly, the hierarchical class system in Britain and its practically impermeable boundary between the middle class and the aristocracy; secondly, the high level of ethos on which the middle class base their dignity and self-respect, the norms of honesty, working for one's living and, especially with regard to women, sexual "virtue," in contradistinction to the aristocratic values of rank, wealth, leisure, pleasure and, with particular respect to men, loose sexual morals; and thirdly, the ban on female initiative in pursuing love and marriage and 52
Cf. B o w e n (1999: 258ff.) and Doody (1974: 71ff.). "[Pamela] overthrew classical literary decorum in making a low, ungrammatical female its heroine; it overthrew social barriers in presenting a misalliance as not only possible but in given circumstances desirable. [ . . . ] Shamela shows what a revolutionary book Pamela could seem. Richardson's novel affronted the old Etonian in Fielding, and he registered the reaction of the Establishment" (Doody 1974: 74).
54
Cf. ibid.: 74 ff. Cf. Bowen (1999: 261).
55
Functions and Forms of Eventfulness in Narrative Fiction
155
the repression of female self-awareness with regard to feelings towards men, which delimits the agency of women as protagonists, confining their scope of action to reactivity and submissiveness. In addition, the notion of divine Providence guiding human destiny functions as an important overarching schema, especially for Pamela, since it shifts (or projects) the agency denied to women to a superior benevolent power and thereby rewards her violation of social boundaries with divine sanctions: in the end, she can "acknowledge, with thankful humility, the blessed Providence, which has so visibly conducted me through the dangerous paths I have trodden to this happy moment" (311). Only in retrospect is Pamela able to recognise the religious consistency of her eventful social transformation, which in the last analysis may be identified as a partly secularised version of the puritanical conversion-and-redemption script and is thus specifically dependent on the cultural and historical context of Richardson's novel. To be sure, the novel contains numerous indicators of these contextual norms and values. For instance, Lady Davers's letter to her brother forbids him to marry Pamela on account of the unbridgeable hierarchical distance between the two families (293-94), and Mr. Andrews states that he would not "own" his daughter if she were "dishonest," i.e. had allowed herself to be seduced (327). In fact, these two characters generally act as a mouthpiece for endorsing the values and principles separating the two social fields, with Mr. B.'s sister trumpeting the social superiority and exclusivity of the aristocracy and Pamela's father representing middle-class moral integrity and ideological self-assuredness (cf. also 45-46, 51-52, 85, 6 9 70). In spite of such indications within the text, however, eventfulness is ultimately not a textual property, but has to be inferred and constituted by readers through relating the textual cues to their knowledge of the period. Although the text can thus be seen to provide appropriate cues, readers may miss the point of the event, and in particular they may fail to appreciate the relevance and difficulty of the transition if they are ignorant of the contemporary social and ideological context of the novel or if they apply an inappropriate frame of reference. Richardson's type of eventful plot structure—the social rise of the heroine to the upper classes through marriage thanks to her moral exemplariness—subsequently contributed to the formation of a new literary convention based on a social and psychological script for later novels. Examples ranging from Jane Austen's Pride and Prejudice (1813) and Charlotte Bronte's Jane Eyre (1847) to William Thackeray's Vanity Fair
156
Peter Hühn
(1847-48), Charles Dickens's Great Expectations (1860-61) and George Eliot's Middlemarch (1871-72) reveal growing difficulties, however, due to the proliferation of obstacles on the part of the protagonists and/or of society and to the social circumstances that affect the actual possibility of eventful transformation, coming to prevent transformation altogether. 3. Staged Eventfulness: Joyce's "Grace" As for the second illustrative example, the plot of Joyce's "Grace" from Dubliners concerns the spiritual, moral and social rehabilitation (through benevolent intervention by his friends) of Tom Keraan, a middle-class businessman fallen into disrepute on account of bad company and undisciplined drinking. Contrary to Pamela, the semantic field in "Grace" and its division into sub-fields are not defined as a hierarchy of two social classes (the middle vs. the upper class), but with respect to the middle class alone as the opposition of exclusion vs. inclusion as determined according to social, moral and religious categories in the form of immoral behaviour vs. moral integrity, dissoluteness vs. righteousness, disreputableness vs. respectability, being isolated and forsaken vs. being integrated and supported by friends, and illustrated in spatial terms as the contrast between lavatory and pub 56 vs. bedroom in a bourgeois home (144ff.) and church (157ff). In all, Kernan performs a double trajectory. The text opens with the completion of his prior downward movement from the higher to the lower sub-field, his downfall from the status of middle-class respectability, literally a drunken fall down the stairs of a pub into the filthy lavatory in a state of unconsciousness, which even arouses the suspicion of the police (138-39). What the short story then goes on to narrate in detail is the process of Kernan's return to a state of "grace," his new rise and re-entry into respectable society—an upward movement deliberately engineered by his friends as Kernan's conspicuous crossing back over a boundary and his public re-admittance into the community of respectable businessmen and devout Catholics in the form of a religious retreat in the Jesuit Church. This movement conforms to the script of penitence and redemption, "making him turn over a new leaf," "making a new man of him" (143), and generally giving him a fresh start after his drunken fall (cf. also 149-50). Though Kernan's transformation does not involve drifting away from this positive script, it does qualify as 56
Joyce (1977: 138ff.); page references cited hereafter in the text.
Functions and Forms of Eventfulness in Narrative Fiction
157
eventful within the implied social and religious context (Catholic Ireland) because of its adherence to central (Christian) schemata of improvement and renewal against the powerful threat of the negative script of sinful decline and degeneration. One can ascribe a relatively high degree of eventfulness to this change because of the spiritual significance attached to such a return of a repenting sinner into the fold. 57 The specific context within which this transformation through repentance figures as eventful is implied in the verbal utterances of the characters by Kernan's friends when they talk about "owning up" (149-50), making Kernan "a good holy pious and God-fearing Roman Catholic," and renouncing "the devil [...] not forgetting his works and pomps" (156). The purported transformation is underscored in particular by the priest at the conclusion of a church "retreat" in his sermon on the parable of the unjust steward, 58 which he ends by voicing the repentant sinner's trust in God's forgiveness: "... I have looked into my accounts. I find this wrong and this wrong. But, with God's grace, I will rectify this and this. I will set right my accounts" (159). At the same time, however, a number of cues to the reader undermine the eventfulness—behind the characters' backs, as it were. Firstly, the moral superiority of the second sub-field is invalidated by the fact that all the friends are revealed to be troubled by disreputable circumstances which they are unwilling or unable to alter and which they try to conceal: Mr. Power's "inexplicable debts" (142), Mr. Cunningham's "unpresentable" wife who is an "incurable drunkard" (145) and Mr. M'Coy's clumsy habit of trying to remedy his precarious financial situation by borrowing and pawning (147); furthermore, the congregation during the retreat is shown to include Kernan's disreputable drinking companion, the moneylender Harford (158). Secondly, the solemn claim of spiritual and moral transformation is subverted by the pervasive evidence of the friends' and Kernan's and also the priest's fundamental worldliness and lack of genuine religious attitudes: for Mrs. Kernan, a belief in the banshee and in the Holy Ghost are equivalent (145); the friends' discussion about the history of the church, the popes and papal infallibility is shallow, boastful and full of ridiculous distortions (150ff.); Kernan's facetious wording of his rejection of a religious ritual betrays the superficiality of 57
58
Cf. the parable of the prodigal son in Luke 15: 11-32. This story of the son who, after squandering his inheritance abroad, returns home repentant and is joyfully welcomed by his father, is meant to illustrate God's grace and loving forgiveness. Luke 16: 1-12.
158
Peter Hühn
his participation in the retreat ("No, damn it all, [...] I'll do the retreat business and confession, and [...] all that business. But [...] no candles! No, damn it all, I bar the candles!" [157]); the priest is praised for his worldly attitude ("He's a man of the world like ourselves" [151]), and his sermon does, in fact, endorse eminently secular values, which is revealed by the choice of the parable of the unjust steward as his reference text with its employment of money, commercial success and even fraud as metaphors of redemption (158-59). Moreover, this association or even equation of spiritual and commercial values, both in the priest's sermon and in the businessmen's attitude, implies Calvinistic rather than Catholic tendencies and thus ironically further undermines the validity of the alleged religious transformation. Thirdly, the entirely superficial quality of Kernan's change is underlined by the pointedly non-religious, colloquial use of the word "grace." Instead of expressing his desire for God's loving forgiveness and his spiritual salvation, this term primarily refers to his newly cleaned hat. In accordance with Kernan's maxim during his former state of respectability before his fall ("By the grace of these two articles of clothing [i.e. a silk hat and a pair of gaiters], he said, a man could always pass muster" [142]), his recovery is reductively reflected in the material transformation of his silk hat from its dirty and battered condition on the lavatory floor in the beginning (138-39) to its "rehabilitated" appearance in the end (158). When the priest finally mentions the term "grace" in his sermon ("with God's grace, I will rectify this and this. I will set right my accounts" [159]), the phrase has been contaminated by its pervasive association with mere elegant outward appearance and sounds definitely hollow. The sarcastic implication of reductiveness also goes for the title of the short story and thus questions the plot development as a whole. And fourthly, the succession of the three spatial stations in Kernan's reformation—from the downfall in the lavatory of the pub via the recovery in his bedroom to the reception into the community of the retreat—can be read as an intertextual allusion to the script of Dante's Divina Commedia: the ascent from inferno via purgatorio to paradiso—a further ironical subversion of the seriousness of Kernan's reform. When combined, these various signals suggest to the reader that the superior values of the second sub-field (in opposition to the first) on which Kernan's plotted rehabilitation is meant to rely as eventful—moral integrity, religious seriousness, social respectability—are no longer intact and do not really exist anymore among the community which he is about to re-enter. The friends' dialogues in the course of their plot and the in-
Functions and Forms of Eventfulness in Narrative Fiction
159
formation about the characters provided by the narrator reveal these values as generally eroded by commercialism, hypocrisy, corruption and superficiality. In reality, Kernan does not cross a boundary at all, but basically stays within the same sub-field in which he first appears as fallen, merely undergoing a change in degree. The text of the short story invalidates the connection of the projected change to its enabling context of a religiously serious and morally upright middle-class society, activating instead a different kind of context, the petty-bourgeois, bigoted, selfconceited urban society of Dublin at the turn of the century as Joyce sees it. The short story builds up the expectation and notion of a decisive event only to unmask it as superficial and hollow: in this kind of society, no eventful change is possible. 59 "Grace" thus features eventfulness at two levels: the diegetic level of the characters and their consciousness, on the one hand, and on the other, the extradiegetic level of the narrator together with, by implication, the reader. That Kernan's eventful change is "plotted" by the friends within the story ("[Kernan] was quite unconscious that he was the victim of a plot" [144; cf. 145]) essentially undermines its relevance. Kernan remains completely passive ("a victim"), even facetiously rejecting part of the role prepared for him (barring the candles), in clear contrast to Pamela who, though forbidden to aspire directly to marriage as an aim, effectively acts nevertheless through her strong and persistent resistance to Mr. B.'s advances and through her conscious struggle for moral self-clarification and self-definition by constantly articulating herself in writing. The detailed narration of the friends' plotting design allows the reader to see through Kernan's intended change as manipulation, which qualifies it as superficial and relative. To the characters—both Kernan and his friends—the event seems valid, even though for the narrator and the reader nothing really take place. In Richardson's novel, by contrast, the marriage constitutes an event at both levels: for the protagonist as well as for the reader and, of course, for the author. 4. Conclusion Eventfulness must be conceived not as an absolute, but as a relative quality—relative not only to the context(s), but also to the point of reference 59
This kind of stagnation is generally characteristic of Dublin and Ireland in the entire collection of Dubliners, symbolically foreshadowed by the word "paralysis" on the opening page of the first short story, "The Sisters" (1977: 7).
160
Peter Hühn
within the narrative set-up, whether this point of reference is ascribed to character, narrator, author or reader. In this respect, Pamela possesses a homogeneous structure, since its eventfulness is endorsed by the novel as a whole and on all of its various narrative levels, whereas "Grace" turns out to be heterogeneous: the status of the event established on one level is surreptitiously subverted on others, which may be taken as a sign of the short story's modernity and its place in modernist literature. Just as the eventfulness in Pamela is not a textual property but depends on the appropriate contextual correlation, so the non-occurrence of an event in "Grace" has to be construed by referring to the relevant context (the stagnant state of middle-class society and the role of Catholicism in contemporary Ireland, in Joyce's critical assessment). Clues in the text, as pointed out above, help to direct the reader's attention. Ultimately, the two texts can be said to constitute two different types of eventfulness. In Richardson's novel, the event is located primarily on the histoire level, in Joyce's short story, on the level of the implied authorial meaning and, consequently, on the level of reception. The reader is meant to recognise the failure of the event to actually occur with respect to the protagonist of the story and to perform it vicariously in his own mind, as it were, i.e. to mentally overcome the paralysing stagnation in the social situation depicted and acknowledge the need for a fundamental change, for an eventful transformation. 60 Thus, "Grace" ultimately does, in fact, refer to an event, although the eventful transformation remains a virtual one as far as the narrative world of the text is concerned. 61 References Abbott, H. Porter 2002 The Cambridge
60
Introduction
to Narrative
(Cambridge: Cambridge UP).
For a detailed systematic differentiation of the various types o f events, cf. the discussion of eventfulness in poetry in Hühn, "Conclusion," in Hühn/Kiefer (2005: esp. 246ff.). Here, four types are differentiated in accordance with the four levels of protagonist, narrator, text and reader in the narrative set-up: events in happenings (located on the histoire level); presentation events (referring to the narrator); mediation events (referring to the textual medium); and reception events (referring to the intended reader activity). The other possible events in this kind of stagnant society seem to be total collapse (as in the short story "Eveline") or total rejection through emigration and exile (as envisaged in J o y c e ' s ^ Portrait of the Artist of a Young Man [1916]).
Functions and Forms of Eventfulness in Narrative Fiction
161
Bakhtin, Μ. M. 1981 The Dialogic Imagination: Four Essays, edited by Michael Holquist, translated by Caryl Emerson (Austin: University of Texas Press). Bal, Mieke 1997 Introduction to the Theory of Narrative. 2 nd ed. (I s 1 ed. 1985) (Toronto: University of Toronto Press). Barthes, Roland 1977 [1966] "Introduction to the Structural Analysis of Narratives," in R. B.: Image Music Text, translated by Stephen Heath, 79-124 (New York: Hill and Wang). Bowen, Scarlett 1999 '"A Sawce-box and Boldface Indeed': Refiguring the Female Servant in the Pamela-Antipamela Debate," in Studies in Eighteenth-century Culture 28: 257-86. Bremond, Claude 1973 Logique du recit. Coll. Poetique (Paris: Editions du Seuil). Bruner, Jerome 1991 "The Narrative Construction of Reality," in Critical Inquiry 18(1): 1-21. Chatman, Seymour 1990 Coming to Terms: The Rhetoric of Narrative in Fiction and Film (Ithaca, NY and London: Cornell UP). Cook, Guy 1994 Discourse and Literature: The Interplay of Form and Mind (Oxford: Oxford UP). Culler, Jonathan 1975 Structuralist Poetics: Structuralism, Linguistics, and the Study of Literature (Ithaca, NY: Cornell UP). Doody, Margaret Anne 1974 A Natural Passion: A Study of the Novels of Samuel Richardson (Oxford: Clarendon). Fludernik, Monika 2003a "The Diachronization of Narratology," in Narrative 11: 331-48. 2003b "Natural Narratology and Cognitive Parameters," in Narrative Theory and the Cognitive Sciences, edited by David Herman. CSLI Lecture Notes Nr. 158: 243-67 (Stanford: CSLI Publications). Genette, Gerard 1980 [1972] Narrative Discourse: An Essay in Method, translated by Jane E. Lewin. Foreword by Jonathan Culler (Ithaca, NY: Cornell UP). Goffman, Erving 1986 Frame Analysis: An Essay on the Organization of Experience. 2 nd ed. (1 st ed. 1974) (Boston: Northeastern UP) Greimas, Algirdas 1987 "A Problem of Narrative Semiotics: Objects of Value," in On Meaning: Selected Writings in Semiotic Theory, translated by P. J. Perron and F. H. Collins, 84-105 (Minneapolis: University of Minnesota Press).
162
Peter Hühn
Herman, David 2002 Story Logic: Problems and Possibilities of Narrative, esp. chap. 3: "Scripts, Sequences, and Stories." Frontiers of Narrative (Lincoln: University of Nebraska Press). 2005 "Events and Event-types," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 151-52 (London and New York: Routledge). Hühn, Peter / Kiefer, Jens 2005 The Narratological Analysis of Lyric Poetry: Studies in English Poetry from the Sixteenth to the Twentieth Century, translated by A. Matthews. Narratologia 7 (Berlin and New York: Walter de Gruyter). Joyce, James [1914] "Grace," in Dubliners, edited with notes by Robert Scholes, 138-59 (Frogmore and St. Albans: Triad/Panther). Labov, William 1972 Language in the Inner City (Philadelphia: University of Pennsylvania Press). Labov, William / Waletzky, Joshua 1967 "Narrative Analysis: Oral Versions of Personal Experience," in Essays on the Verbal and Visual Arts, edited by June Helm, 12-44 (Seattle: University of Washington Press). Lotman, Jurij 1977 [1970] The Structure of the Artistic Text, translated by G. Lenhoff and R. Vroon (Ann Arbor: University of Michigan Press). Martinez, Matias / Scheffel, Michael 1999 Einfuhrung in die Erzähltheorie (Munich: Beck). Ochs, Elinor / Capps, Lisa 2001 Living Narratives: Creating Lives in Everyday Storytelling (Cambridge, Mass: Harvard UP). Pier, John 2003 "On the Semiotic Parameters of Narrative: A Critique of Story and Discourse," in What Is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 73-97 (Berlin and New York: Walter de Gruyter). Polanyi, Livia 1979 "So What's The Point?," in Semiotica 25: 2 0 7 ^ 1 . Pratt, Mary Louise 1977 Toward a Speech Act Theory of Literary Discourse (Bloomington: Indiana UP). Prince, Gerald 1983 "Narrative Pragmatics, Message and Point," in Poetics 12: 527-36. 2003 A Dictionary of Narratology. Revised ed. (1 st ed. 1987) (Lincoln and London: University of Nebraska Press). 1999 "Revisiting Narrativity," in Grenzüberschreitungen: Narratologie im Kontext / Transcending Boundaries: Narratology in Context, edited by Walter Grünzweig and Andreas Solbach, 43-51 (Tübingen: Narr).
Functions and Forms of Eventfulness in Narrative Fiction
163
Richardson, Samuel 1980 [1740] Pamela; Or Virtue Rewarded, edited by Peter Sabor. Introduction by Margaret Moody (London: Penguin). Rimmon-Kenan, Shlomith 2002 Narrative Fiction: Contemporary Poetics. 2 nd ed. (I s 1 ed. 1983) (London: Routledge). Ryan, Marie-Laure 2005 "Tellability," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 589-94 (London and New York: Routledge). Schaeffer, Jean-Marie 1989 "Literary Genres and Textual Genericity," in The Future of Literary Theory, edited by Ralph Cohen, 167-87 (New York: Routledge). Schank, Roger C. 1990 Tell Me a Story: A New Look at Real and Artificial Memory (New York: Scribner). Schank, Roger C. / Abelson, Robert P. 1977 Scripts, Plans, Goals and Understanding: An Inquiry into Human Knowledge Structures (Hillsdale, NJ: Lawrence Erlbaum). Schmid, Wolf 2003 "Narrativity and Eventfulness," in What Is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 17-33 (Berlin and New York: Walter de Gruyter). Schönert, Jörg / Hühn, Peter / Stein, Malte 2007 Lyrik undNarratologie. Text-Analysen zu deutschsprachigen Gedichten vom 16. bis zum 20. Jahrhundert. Narratologia 11 (Berlin and New York: Walter de Gruyter). Shukman, Ann 1977 Literature and Semiotics: Α Study of the Writings of Yu. M. Lotman (Amsterdam: North-Holland). Sternberg, Meir 1990 "Telling in Time (I): Chronology and Narrative Theory," in Poetics Today 11: 901-48. 1992 "Telling in Time (II): Chronology, Teleology, Narrativity," in Poetics Today 13: 463-541. 2001 "How Narrativity Makes a Difference," in Narrative 9: 115-22. Todorov, Tzvetan 1977 [1968] "The Grammar of Narrative," in Τ. T.: The Poetics of Prose, translated by Richard Howard, 108-19 (Ithaca, NY: Cornell UP). Toolan, Michael 2001 Narrative: A Critical Linguistic Introduction. 2 nd ed. (1 st ed. 1988). The Interface Series (London and New York: Routledge).
WERNER WOLF (Graz)
Chance in Fiction as a Privileged Index of Implied Worldviews: A Contribution to the Study of the World-Modelling Functions of Narrative Fiction 1. Introduction: the analysis of implied worldviews as a neglected area of narratological enquiries into the world-modelling functions of fiction One of the most fascinating and important branches of literary studies, including narratology, is the analysis of the functions that texts can serve for individual readers, specific cultures and historical situations, but also for mankind in general. For it is through their functions that literary texts and the possible worlds created by them are anchored in life. 'Functions of literature' can refer to specific textual effects, aesthetic and otherwise, such as suspense, the immersion of readers into textual possible worlds (the creation or undermining of aesthetic illusion), the thematization of social injustices or the plea for their redress and so forth 1 , but they also include broader issues. In narrative literature, perhaps the most important among these broader issues is the contribution to the human endeavour to make sense of the experience of time in a world governed by change, and with this, of existence in general. It is with these general issues of sensemaking that the following contribution is concerned. All works of literature produce miniature models of the world. These models are inevitably partial and reductive, but their very partiality and reduction of the complexities of reality is one of the principal preconditions of the sense-making quality of literature, for it is only through finite models that infinite reality can be grasped. Owing to Jurij Lotman's discussion of the secondary world-modelling function of literary texts, 2 this 1 2
See Gymnich/Nünning (2005). Lotman (1977 [1970]: 21), (1972: 34-44).
166
Werner Wolf
is not a new issue in literary studies. The more recent narcological 'possible worlds theory' confirms this world-modelling quality, 3 for ultimately literary possible worlds are not mere flights of fantasy but—at least from a functional perspective—can be said to remain always rooted in the actual world. 4 One of the principal functions of literary texts as models of the actual world is their ability to illustrate or concretise worldviews. Narratives are privileged in this, as by definition they deal with what happens in or through time and thus always address fundamental challenges to human sense-making, notably birth, death and the experiences of transitoriness, but also accidents and other chance happenings. 5 The pre-eminence of narratives as an essential means of making sense of time and existence has repeatedly been stressed by narratologists, 6 and the fact that narratives imply, are based on, or at least relate to, specific worldviews is also well known. 7 Yet, in spite of the marked trend in narratology towards systematic theorising, the question of how worldviews are constructed and communicated in narratives has to my knowledge never been addressed in a systematic manner. An essential aspect of the worldmodelling function of narratives has thus escaped attention. Nor has narratology ever sought to provide a well-equipped tool-bag for the analysis of implied worldviews. What has been provided so far are isolated concepts, in particular the system of causation underlying a text, the controversial concept of the "implied author"9 (or—this is Ansgar Nünning's alternative—"die Ebene der Gesamtstruktur [...] eines Textes" 10 [the level of the total structure of a text]) as the origin of the implied norms emerging from
3
4 5
?
See, for example, Ryan (1991) and Doleäel (1998), but also Goodman's earlier discussion o f literary Ways of Worldmaking (1978). Cf. also Bruner (2002: 25). Cf. Paulson (1994). Besides almost ubiquitous fleeting remarks on the function o f narratives and narrativity for human sense-making, many studies have, from various angles, addressed this dimension in more depth, such as Adams (1996); M. Bell (1990); Bruner (1990), (2002); Herman (2003); Miller (1990); Mink (1978); Ricoeur ( 1 9 8 3 - 8 5 ) ; Stierle (1979); White (1987). The relationship between the world-modelling function o f literature and the fact that the models created inevitably betray a Weltanschauung is. for instance, formulated in Lotman's reflections (1977: 38), although he focuses on worldviews as a function of individual authors, and not of texts.
8 9 10
See Richardson (1997). Booth (1983 [1961]: 7 1 - 7 7 ) . Nünning (1993: 19).
Chance in Fiction as a Privileged Index o f Implied Worldviews
167
the text as a whole, and "perspectival structure" as one of the means of communicating these norms. 11 Yet, the questions of how the stances of 'implied authors' and of how implied norms and "perspectival structures" coalesce into worldviews have not found much attention. The reluctance to deal with this problem in more detail is, however, easy to understand, for virtually all elements of a literary text can contribute to the worldview implied in it. Consequently, a narratological theory of implied worldview analysis would have to address all aspects of stories (including their structure and content, explicit statements by characters, themes and their development, etc.) as well as all possible facets of the discourse level (including narrative situations, narratorial comments, imagery, the guidance of reader sympathies in favour of, or against, certain characters and so on). What is more, such a theory would not only have to account for a plethora of possible aspects, but also for the fact that implied worldviews, as the term indicates, tend to be tacit and implicit, rather than explicit elements of literary texts: as a consequence, the object of scrutiny comprises not only the entirety of what is present in a text, but in such an analysis it is frequently as important to take into account what is not mentioned, what has been omitted or suppressed. 12 In view of such difficulties, one option—and one that has been chosen with deplorable frequency—would be the capitulation of narratology in this area, the consequence being that implied worldviews would continue to be left to ad hoc analyses of individual texts carried out on the basis of objects and criteria chosen in a haphazard way, if not by means of a merely intuitive bricolage. Another option would be to take up the challenge and contribute to a future, albeit presumably always fragmentary, theory by tackling at least some facets of literary texts that are particularly apt for the analysis of implied worldviews. Such is the option chosen for the purposes of this essay. In particular, I propose to examine how narratives address the problem of chance as one of the privileged 'doors' that permit access to the hidden depths of implied worldviews. For the experience of chance—happenings which, at first sight at least, appear to be inexplicable—constitutes one of the central challenges to, and arguably triggers of, human sense-making and thus of worldviews. Therefore, 11
Pfister (1977: 9 0 - 1 0 3 ) ; Nünning (1989: ch. 3). This refers not only to the general fact that absences can always reveal something about the choices and preferences underlying a text, but in particular to tacit references to contemporary worldviews whose knowledge is taken for granted in a given text and therefore 'goes without saying'.
168
Werner Wolf
chance in this sense is a feature of most literary narratives, and the way it is dealt with may thus generally provide a good index of implied worldviews. In short, chance may not only "be used as a probe into the metaphysics of the novelist," as David Goldknopf once suggested with refer13 ence to the Victorian novel, but also, and perhaps more importantly, as a probe into the worldviews implied in the texts produced by novelists (and not solely those of the Victorian age). Chance, as an element of narratives, is inextricably linked with "causal connection" which Brian Richardson, in his seminal investigation of narrative systems of causation, has rightly called "a necessary condition of narrativity." 14 However, while Richardson concentrates on causation as part of a "poetics of narrative sequencing" 15 and reserves 'chance' to cases where, ultimately, no other form of sequencing applies, the following contribution will concentrate on chance appearances at first reading (which ultimately may even include a 'deleting' of chance) as revealing for the worldview implied in a narrative. As implied worldviews form a crucial part of the world-modelling function of narratives, the following discussion of possible ways of analysing textual worldviews is thus not only relevant for the interpretation of individual texts, but is also a contribution to elucidating an important facet of the world-modelling function of narratives in general. My focus will be on fiction, although some of the findings may also be relevant for a future, more comprehensive and perhaps even transgeneric and transmedial theory, at least as far as the representation of chance in (temporal) artefacts is concerned.
2. Chance and (implied) Worldview(s) as General and Narratological Concepts 'Chance' and 'worldview' are common concepts which are by no means restricted to fiction or even to narrative. Unfortunately, both are not only relatively unclear in common usage, but in addition tend to have slightly specialised connotations in literature. Therefore, it may be helpful to begin with some reflections first on their extra-literary and then on their intraliterary meanings. 13 14 15
Goldknopf (1969: 41). Richardson (1997: 37). Ibid.: 14.
Chance in Fiction as a Privileged Index of Implied Worldviews
169
The notion of chance in an extra-literary, everyday and also philosophical and scientific sense is of particular imprecision. 16 The common denominator of its various denotations is the idea that chance, or what results from it, defies explanation and is therefore unpredictable. The resistance to explanation is commonly equated with the impossibility of linking chance phenomena either to intentions and designs, that is, to teleology, or to foreseeable effects of known causes.' 7 From a logical perspective, causes and effects can be linked in a necessary or in a probable way. Chance is therefore also related ex negative to necessity and probability, which adds another dimension to its inherently inexplicable nature, rendering it applicable to phenomena that are not necessary and/or are subject to a high degree of (statistical) improbability. The imprecision of the concept 'chance' is revealed, among other things, in the impossibility of determining the exact degree of improbability starting from which it would be justified to apply this term. This imprecision also shows in the fact that chance can, in extreme cases, coincide with randomness, thus producing the impression of meaninglessness; then again, chance can also apply to coincidences which, on the contrary, may result in an overdetermination of meaning. In addition, the resistance to explanation which characterises chance can be either categorical or relative due to a lack of knowledge that could in principle be overcome. 18 To add yet another facet of meaning, chance applies not only to phenomena (happenings, occurrences), but may also be conceived of as a quasi agency ('as chance would have it') or even as a quality or attribute. This is true in particular where 'chance', as often happens, 19 is used as a synonym of 'contingency', although, as we will see with reference to narrative fiction, it is advisable to differentiate on this point. All in all, then, the notion of chance seems to be applicable wherever something defies exact explanation on logical grounds due to lack of relevant knowledge or to states of high improbability. 16 17
18 19
Cf. Wetz (1998: 27, 29-30); Kranz et al. (2004: 1408). Cf. J. Hardy (1996: 590). See also Kranz et al. (2004: 1408), and Pearsall (ed.) (1998: 303) s.v. 'chance': "the occurrence and development of events in the absence of any obvious design or cause." Of course, causes may be attributed to some chance occurrences retrospectively (this is when we say 'it was not a coincidence that...'), but at this moment the quality of the occurrence changes and is no longer chance. Generally, chance has an at least provisional experiential quality that is linked to moments when causal (or other) explanations are (not yet) available. In some cases such resistance to explanation persists even from a retrospective perspective. Cf. Kranz et al. (2004:1408). Ibid., 1408-09.
170
Werner Wolf
The resistance which 'chance' in this general sense presents to explanation is a challenge to what appears to be an essential human tendency, namely to make sense of everything we experience. This challenge is particularly discernible where the term 'chance' is not combined with the positive connotation it has in English in a second sense ("an opportunity to do or achieve something" 20 ), but with the negative connotation of 'accident'. As such, chance and the urge to explain it away—together, perhaps, with our sense of mortality—is one of the main triggers of worldviews, one of whose principal and general functions in turn is precisely to address the problem of chance in some way or other. As for the second term to be clarified in its everyday sense, namely 'worldview', it can be defined as any system of meaning, be it of a commonsensical, religious, philosophical, political, ideological or scientific nature, that seeks to provide answers to basic and general questions bearing on human existence in addition to addressing further, more particular issues (concerning, for instance, details of the conception of society, ethics, aesthetics, etc.). The domains into which worldviews extend—let alone their ramifications—are endless, so that in a more detailed description of the notion of 'worldview', some concentration on essentials is advisable. Obviously, among these essentials, 21 those which are most germane with regard to the present context involve the problem of chance and its possible containment. They, in particular, address the following questions: whether the world is based on some order or not; what role chance plays in the world; how an order, if it exists, can be evaluated (whether it is hostile, favourable or indifferent to man); whether, how and to what extent we can become aware of this order (this is also where 'worldview' is related to what Foucault called episteme22); whether a potential order can be attributed to some metaphysical sphere or being as a first cause or 'designer' or, alternatively, to certain scientific data which might also reveal randomness as an ultimate reality, and so on. Worldviews and chance are generally related to each other in three ways: 20
Pearsall (ed.) (1998: 303), s.v. 'chance'. These essentials include the problem of human identity, the possibility of an after-life, a potential meaning or tendency of history and, above all, a specific view of man: whether good, bad, morally mixed, perfectible, rational or not and so forth. Foucault (1966: 13 passim); in fact, Foucault's episteme, i.e. the basic configuration of knowledge and the beliefs regulating its acquisition, functions, etc.. essentially belongs to a period's or a culture's worldview.
Chance in Fiction as a Privileged Index o f Implied Worldviews
171
The first possibility is that worldviews seek to rule out or neutralise chance by insisting on some underlying principle that permits at least a basic explanation for all chance happenings (this is true both of deterministic religious beliefs, as represented by Calvinism, and of secular beliefs in the ultimate explicability of everything, as epitomised by Laplace's demon). The second possibility is a 'soft version' of this attempt to bring chance under control. It consists in bracketing or containing chance. Worldviews that follow this option admit the existence of chance, at least within a restricted perspective, but remain open to the possibility of integrating it into a meaningful or at least an intelligible whole. This form of reasoning about chance can, for instance, be observed in the belief in Fate as part of a metaphysical worldview in which individual occurrences elude human explanation but are thought to be ultimately traceable to some divine source, thus becoming meaningful within a larger whole. Another, secular version of the combination of admitting chance, and at the same time taming it by positioning it within a meaningful or structured whole, is modern chaos theory, quantum physics or the notion of chance mutation as a part of the theory of evolution. The third, albeit relatively rare, relation between worldviews and chance consists in highlighting chance as the ultimate and ineluctable reality. Such foregrounding can be seen particularly in absurdist worldviews of the twentieth and the early twenty-first centuries, but also in any kind of radical agnosticism whose followers acknowledge the existence of chance but are unable or unwilling to account for it with reference to any larger system of meaning. Worldviews are constructed and communicated through the use of several semiotic macro-modes, discourse types or 'frames' such as description, (religious) instruction and (philosophical or scientific) argumentation. However, the discourse type that is most pervasive and powerful for creating worldviews is certainly narrative, one of whose main aims, as has been pointed out time and again, is to come to terms with the problem of 23
chance and to contribute to what in German terminology is called Kontingenzbewältigung (mastering of contingency). 24 Narrative is indeed privileged in dealing with chance, since most 'narratemes' (that is, the
23 24
Paulson (1994). Warning (1980), (2001).
172
Werner Wolf 25
constituent elements of narrativity, as detailed by Gerald Prince and myself 26 ) can be seen to possess a close relationship to this phenomenon. This, for a start, is true of what I have termed "content narratemes" (or 27
"narrative building blocks" ), i.e., for example, a) the setting in which b) anthropomorphic characters are shown to be engaged in c) some multiphase action or events. 28 For chance is most interesting to us when it affects humans ('anthropomorphic characters'), and the occurrence of a chance event always involves a temporal dimension (and hence a temporal setting), namely a state before and after, and can thus contribute to narrative action. In addition, it is frequently due to chance that possible alternatives of a course of action are not realised and become, what according to Prince is also typical of narratives: "disnarrated elements." 29 Chance is moreover intimately related to "syntactic narratemes" 30 (the elements that link the 'building blocks' to each other and owing to which a meaningful story emerges as a discernible unity). This immediately becomes clear when we remember that besides the two basic syntactic narratemes, chronology and causality, 31 teleology figures among the most important "syntactic narratemes": for, although it has been said that, in general, not only causality but also teleology has a negative relationship to chance, chance in narratives (for instance, in accidental re-encounters of long-lost family members) often centrally contributes to the teleological goal of a story being reached as well as the story's unity. Finally, "general narratemes," which form the basis of all stories (namely the32qualities "representationality," "experientiality" and "meaningfulness" ), can also be seen to be connected with chance: the link between chance occurrences and experientiality is perhaps most obvious. For encountering the unexpected, particularly if it resists, or appears to resist, explanation, belongs to the most outstanding and memorable moments of human experience and triggers 25
26 27
29 30
32
Prince (1999: 46); for Prince's thorough description of narrativity, see also Prince (1982) as well as his contribution to the present volume. Wolf (2002b), (2004). Wolf (2004: 88). See ibid. Prince (1996: 98). Wolf (2004: 88). A s stated above, Richardson (1997: 96; cf. also 37) considers "causality is [...] a necessary condition of narrativity"; however, he at the same time and rather unconvincingly argues that it is "not clear" whether chronology or "temporal succession" is likewise a "necessary condition." Wolf (2004: 88).
Chance in Fiction as a Privileged Index o f Implied Worldviews
173
what has repeatedly been attributed to narratives in general, namely the urge to make sense of such experience (chance is thus related to the narrateme 'meaningfulness'). The relation between chance and both the experiential and representational qualities of narratives is manifest in the fact that stories typically involve not only a discursive representational act (which frequently implies a privileged retrospective position from which seeming chance can sometimes be dissolved into causally and teleologically explicable effects), but also a story level, a level of representation. Where this representational level is constructed so as to permit the recipients a (quasi-) experience of the story events themselves, chance, which in 33
itself has an experiential dimension, can become especially important and contribute centrally to the "tellability" 34 of the story as a whole. All in all, one can say that narratives are representations that focus on possibilities of making sense of series of temporal, eventful experience in which chance tends to play a more or less central role. In opposition to this, the descriptive, for instance (which, like narrative, is an important discourse type), while also involving representation, has fewer affinities with chance and its 'mastering', since it typically centres on paradigmatic "existents" rather than on syntagmatic "events." Owing to the nature of the discourse type 'narrative', the construction and transmission of worldviews in narratives proceeds in a particular way, bringing into play several textual levels, and indeed more levels than scientific or philosophical discourse typically rely on. 36 In texts classified as narratives due to the dominant discourse type employed, worldviews can thus be transmitted in different ways: in narrative fiction, to start with, they can appear explicitly on the story level, notably in arguments, norms or instructions attributed to individual characters, their utterances and thoughts. Such explicit worldview-relevant elements can also occur on the discourse level and are then mostly attributable to narrators and, more rarely, to fictitious readers. In all of these cases, we can speak of 'apparent worldviews', and there is frequently a plurality of them in any given fic-
33 34
See above, note 17. Pratt (1977: 132ff.); cf. also Ryan (2005). For this differentiation, see Chatman (1978: 19 passim). A s Virtanen (1992) has so clearly shown, "discourse types" (the dominant semiotic frames characterizing individual texts) can be realized by various "text types" (that is, sub-dominant frames which can be identical to, but can also differ from, the dominant "discourse type").
174
Werner Wolf
tional text, depending on individual perspectives. This multiperspectivity underlies what Bakhtin termed "dialogism" in the novel. 37 Apparent worldviews that are formulated explicitly by intradiegetic characters or extradiegetic agencies are optional realizations of the worldview-transmitting function of narratives. Alternatively, but also in combination with explicit forms, worldviews can take on an implicit form, and this is in fact what matters most in our context, since the overall worldview which forms the basis and one of the principal communicative effects of a textual world model is usually implicit. Implicit worldviews emerge from texts in their entireties, in particular from the positive norms and attitudes (whether implied or explicit) informing them, and are attributable to 'implied authors', if one wants to maintain this concept. A number of elements contribute to the emergence of implied worldviews. There is, first, the story level (e.g. implicit perspectives, attitudes and norms underlying characters' actions, utterances and thoughts as well as their consequences, the configuration of characters and what happens to them, but also, if applicable, the above-mentioned explicit formulations of apparent norms and worldviews). Next to the story level, the discourse level also plays a role for implied worldviews (this includes implicit or, again, explicit narratorial evaluations, the direction of sympathies for or against certain characters, etc.). Besides the totality of what is present on both the story and the discourse levels, implied worldviews are also shaped by what is omitted (what is conspicuously absent) as well as by contextual factors: the frames of reference and the world knowledge presupposed for the understanding of individual texts. All of these elements, of which apparent worldviews are only an optional part, form a broad basis for analysing the worldview that is expressed in, or through, a text, although it may not explicitly be formulated by any textual agency and thus may not appear on the text surface as a quotable statement or belief. Analysing an implied worldview thus resembles the description of the basic pattern of a carpet which emerges not only from individual elements, but also from their configuration, including the empty spaces between these elements, as well as from the frame surrounding the carpet. 'Worldview' in this sense only occurs in the singular in each text and is not subject to the perspectival distortion of individual discursive agencies below the level of the (implied) author, so that one should speak not of a plurality of apparent (and potentially untrustworthy) worldviews, but of the one implied worldview 37
Bakhtin (1981: 259ff.).
Chance in Fiction as a Privileged Index of Implied Worldviews
175
of a text. This singular does, of course, not rule out that the worldview implied in a given narrative is potentially ambiguous, open or inconclusive in certain details, nor that it may trigger—as in fact happens so frequently—markedly different interpretations. Yet, each interpretation will try to describe only one (version of an) implied worldview, or, alternatively, acknowledge openness or ambiguity as a feature of the worldview in question. The singular, which indeed imposes itself in the description of the overall worldview implied in a given text, is explicable by the fact that this worldview is the philosophical facet of a narrative as a miniature model of the world, for each narrative can form only one model. It is due not least to this philosophical world-modelling, which by definition addresses major questions of existence, that narratives (as well as other literary texts and works of art) are anchored in reality. Similar to 'worldview', 'chance' (as has already been said) is also manifested in fiction, but in ways that must still be specified in more detail, since not all facts of the concept as relevant to narratology can immediately be deduced from the meaning of chance' in everyday experience or discourse. To begin with, chance can occur in all the positions involved in narrative communication, most notably in the production and/or reception process and as part of the fictional text (see Figure 1 below). However, for obvious reasons, those occurrences of chance are most relevant as indicators of implied worldviews that have left traces in a text or work as results of intentional authorial activities, including the introduction of chance into the gestation of the work/text itself (which permits us to disregard involuntary accidents in production and reception processes38). These traces concern both the thematization and the illustration or representation ('telling' and 'showing') of chance, and can be broken down as follows:
38
One could argue, in particular from a psychoanalytical point of view, that the entire production (as well as reception) of artefacts also contains unconscious factors that may influence the occurrence of chance in them. While the importance of an author's unconscious cannot be denied, it is methodologically very difficult to grasp in literary texts (at any rate, more difficult than the published texts themselves) and would merit an investigation of its own. Therefore, the relationship between an authorial unconscious, its effects on the use of chance and implied worldviews will remain beyond the scope of this narratological essay; instead, what has been published is here assumed to be the effect of an authorial consciousness.
176
Werner Wolf
They can appear on the extraflctional level (discourse level I): chance may be either thematised or alluded to in paratexts 39 (such as the title of one of Joseph Conrad's novels, Chance [1912]), or it may leave traces in the process of composition and thus illustrate its workings. The intentional variant of the latter case occurs only rarely, as in experimental novels such as B.S. Johnson's The Unfortunates (1969) or, more recently, Robert Coover's short story "Heart Suit" (2005), where the individua^arts of the text can be physically reassembled ad libitum by the reader. Additionally, the illustration of chance, particularly in a storyworld (see below, no. 3), can also be promoted by the extra-fictional authorial choice of a specific literary genre. 41 Chance can occur, moreover, on the extradiegetic level (discourse level II): either in comments (thematizations) by a fictitious narrator (or reader) on chance in general, or on chance occurrences of the story; alternatively, chance can also be illustrated by the activities of an extradiegetic agency, particularly with regard to his or her role as a storyteller. Such illustrations can, for instance, be observed in the apparently haphazard progress (or retrogressive movement) of the discourse in Laurence Sterne's Tristram Shandy (1759-67) as well as in a narrator's metaleptic intrusions into the storyworld in order to avoid or promote chance developments (as in John Fowles's The French Lieutenant's Woman [1969], where the narrator tries to determine the progress of narration by the tossing of a coin in order to decide which of the two alternative endings of the novel should come c first). 42 Chance can also appear on the diegetic level (story level I): either in the form of a thematization by characters (including characters who are also authors or narrators of embedded stories), or where chance is represented as a shaping force of the storyworld, particularly when characters become objects of chance happenings or are themselves subjects of chance activities by throwing dice, for example, or by gambling (as the main characters in Paul Auster's novel Music of Chance [1990]) or by taking part in other random activities (including the haphazard production 39 Paratexts can be located on the extraflctional level owing to the prevailing expectation that they are non-fictional 'framings', unless otherwise indicated. In poetry, an analogy would be the apparently random text elements in a Dadaist poem, produced according to the principle of cutting up an existing text and reassembling the fragments in a haphazard way. See below the discussion o f Kontingenz-Gattungen by Preisendanz (1998a). For further examples, see Morson (1998: 302ff.).
Chance in Fiction as a Privileged Index of Implied Worldviews
177
of texts). A variant of the diegetic occurrence of chance (or the impression of its occurrence) can be linked to a specific discursive device, namely where, particularly in modernist texts, a stream-of-consciousness technique is employed in which random associations of ideas and the flux of time seem to dictate narrative sequences. Lastly, chance can, of course, also be manifest on (a) hypodiegetic level(s) (story level II)—again, either in the form o f thematization or in that of illustration or representation (as in the case of mises en abyme of thematizations and illustrations on the diegetic level). TEXTUAL/
THEMATIZATION
ILLUSTRATIONS and
COMPOSITIONAL
OF CHANCE ('telling')
REPRESENTATIONS OF CHANCE ('showing')
LEVELS 1. extrafictional level
A) paratexts
(discourse I)
B) traces of chance in the process of composition involuntary voluntary
2. extradiegetic level
A) e.g. narratorial discus-
B) narrator's random or other-
(discourse II)
sions of chance in general
wise chance-related activities, in
or comments on chance
particular as a storyteller
occurrences in the story 3. diegetic level
A) thematization of chance
B) chance as a shaping element
(story I)
by characters, including
of the storyworld characters as
author-narrator-characters
objects of chance occurrences
creating mises en abyme of
characters as subjects of random
positions 1A and 2A
activities including random text production creating mises en abyme of positions IB and 2B
4. hypodiegetic
A) mise(s) en abyme of
B) mise(s) en abyme of position
level(s) (story II)
position 3A
3B
Figure 1: Chance in fiction: levels and forms of occurrence
While levels 1, 2 and 4 certainly cannot be neglected, it is with level 3 — and here in particular with the occurrence of chance as a shaping factor of the story level—that we will be principally concerned in the following pages, for it is the influence of chance on the storyworld that best harmonises with the idea of narratives as miniature models of the world. Actu-
178
Werner W o l f
ally, the use of the term 'chance' in this context could be criticised, for all occurrences of chance in story worlds are only simulations. Stories are intentional constructs and as such by their very nature opposed to chance. This creates the paradox that chance happenings in storyworlds (but also chance phenomena linked to fictitious narrators) are wilfully determined by authors, and one therefore may indeed say with Brian Richardson that "chance in literature is never a chance occurrence." 43 Awareness of the intentionality of aesthetic artefacts and the 'faked' role of chance within their framework shapes the expectations of recipients, and in particular the expectation that here even chance is somehow (at least aesthetically) meaningful, for artistic intentionality and meaningfulness seem to go hand in hand. According to Lotman, "the tendency to interpret everything in an artistic text as meaningful is so great that we rightfully consider nothing accidental in a work of art." 44 This also constitutes one of the principal reasons why the way in which chance is dealt with in a text can indeed be regarded as a privileged access to the implied worldview. Yet, when exactly should one speak of a chance phenomenon in a storyworld? There is a widespread tendency among critics such as Morson to concentrate on the "role of chance [...] for the characters."*5 This reduction of chance to a mere facet of the story level is, however, only a part of what matters. For if implied worldviews are to be construed by readers, the dimension of reception, and in particular the simulated experience of chance which narrative texts can provide, must also be taken into consideration. As a consequence, chance in storyworlds (diegetic or, as the case may be, hypodiegetic chance) always involves the story level, but the discourse level can also play a critical role: the two levels can interact, providing particularly vivid experiences of chance for the reader, especially when chance occurrences in the storyworld are neither discursively prepared nor retrospectively explained. Even so, these levels can also enter into conflict, for what may appear as chance from a mere storyperspective can be explicable as a result of information provided later on by a narrator. Thus, chance-related phenomena on the (hypo-)diegetic 43
44 45
Richardson (1997: 18). This fact, together with the constraints of narrativity (see below), also leads Monk to formulate the exaggerated and one-sided thesis that "chance is that which cannot be represented in narrative" (1993: 9), meaning that chance cannot be "re-presented" or made present in an artefact, which, of course, does not imply that it cannot be represented as a part of a storyworld. Lotman (1977 [1970]: 17). Morson ( 1 9 9 8 : 2 8 9 ) .
Chance in Fiction as a Privileged Index of Implied Worldviews
179
level that may be relevant as indices of an implied worldview are all storyworld phenomena that appear as chance to characters and possibly to readers on a first reading; that is, they seem to fulfil the conditions of chance that readers would also apply to first experiences in real life, irrespective of later explanations and re-experiences (e.g. through filmic records). For even if the discourse later explains a chance occurrence away and thus actually deletes its 'objective' quality as chance, this very fact— the explicability of seeming chance—may be highly significant for the implied worldview. The first reading of a narrative is also important in another respect, for it is in this activity that the experience of uncanniness and surprise which often accompanies chance happenings is most vivid and marks such occurrences as salient and in all probability worldviewrelevant features in a story. 46 Only as far as the overall assessment of the role of chance in a given text is concerned does the perspective after a first complete reading have priority. From a narratological perspective, chance-related elements that are potentially relevant to the implied worldview thus comprise a wide range of textual phenomena: thematizations and (intentional) illustrations or representations of chance on the extrafictional, the extra- as well as the intraand hypodiegetic levels, and as far as representations are concerned, 'genuine' chance (for characters and readers) must be included as well as what at first, or on a first reading, may provisionally and subjectively only appear as chance to characters or readers. Among other options, those story elements are of particular relevance for the readerly experience of chance which on a first reading are not, or do not seem to be, causally (or ideologically) motivated by story elements or agents, in other words, phenomena that are not plausibly explicable as the obvious, necessary or probable result of previously introduced intentions nor of previously mentioned causes. Drawing on Richardson, one can say that these causes are usually taken from two different fields (which also show a worldview relevance in themselves), namely from the supernatural and the natural. 47 46
Morson (ibid.: 292) here pleads for a curious compromise and claims that "the position we adopt ought to depend on the sort of work we have before us," without specifying which kind of work would call for a preference for a first reading and which not. Cf. Richardson (1997: 62). A third, "metafictional," field (ibid.) is of comparatively lesser importance: it refers to usually metaleptic fulfilments of an extradiegetic (authorial or narratorial) will on the diegetic level. Richardson's fourth field of causation, "chance," stands in opposition to the other fields and applies when no reasons are given within these. Richardso's category "chance" thus seems immediately convertible into a
180
Werner Wolf
Thus, chance in fiction, as investigated below, applies to occurrences within diegetic (or hypodiegetic) worlds that on a first reading do not appear to be plausibly caused or motivated by anything within these worlds, neither by a supernatural agency nor by natural, secular factors—although it may turn out later on that factors from these fields are in fact responsible for the occurrences in question. 3. Textual Manifestations of Chance in Fiction as Indices of Implied Worldviews: Criteria of Analysis How, to what extent, and under what conditions can the occurrence of chance in storyworlds be indicative of (implied) worldviews, and of what kinds of worldview in particular? As chance breaks down into a number of variants that, as we will see, display different indexical values in this context, it is helpful to distinguish between these variants to begin with. The following three tend to occur in fictional worlds: 'Contingency' in the old Aristotelian sense of 'what may be or not be' (tyche):48 this attribute of phenomena is common and more or less inevitable at the beginning of stories, where the 'ground situation' (including setting, principal characters and basic facts) is laid out or possibly presupposed and where, in principle, no previous reasons or intentions are adumbrated (barring any intertextual allusions or possible extra-literary references, e.g. to history). The same kind of contingency may reoccur in the successive episodes of a story, although usually to a lesser extent. Contingency in this sense also informs the consequences of what Seymour Chatman (following Roland Barthes) calls "kernels" in plot develop49
ment, namely situations that allow for (at least) two alternatives, and thus imply that a narrated development may have occurred otherwise.50 'Accidents' in the sense of a sudden, unforeseeable and (seemingly) inexplicable occurrence, usually one that has important consequences for one or several characters. Accidents, either lucky or unlucky, can be reparticular worldview (and is not merely a provisional 'door' opening onto an implied worldview that may even prove to delete chance) and is thus narrower than the readercentred sense in which 'chance' is used in this essay. 48
Cf. Wetz (1998: 27).
49
Chatman (1978: 3 2 - 3 3 , 53ff.). The non-actualized potentials of such kernel-situations belong to what Prince has termed "disnarrated elements" (1996: 98).
Chance in Fiction as a Privileged Index of Implied Worldviews
181
garded as actualizations of contingency in the afore-mentioned sense, for in principle they could also not have happened. However, what differentiates them from general contingency is their nature not as an attribute of 'existents', but as happenings that are often characterised by surprising suddenness. 'Coincidence' is equally related to happenings, namely as the unforeseeable and (seemingly) inexplicable yet apparently meaningful intersection of two occurrences, sometimes even of two causal chains or sequences of happenings and events previously introduced into the storyworld, but without causal connection to one other. An example can be taken from the typical romance plot in which two lovers have been separated by accident, have been looking for each other for years and suddenly meet again, although this coincidental reunion was neither planned nor foreseen by either of the parties. A coincidence, which, again, may have positive or negative effects, is thus a special kind of accident (and therefore also a special actualization of contingency) characterised by the fact that it has to some extent been structurally prepared in a narrative and regularly produces the impression of an overdetermination of meaning, possibly with overtones of the uncanny. 51 Before outlining parameters for the analysis of chance in narrative fiction and its role as an index of implied worldviews, it is necessary to take account of certain 'filter factors', that is to say, elements that may alter, reduce or even delete the worldview-centred relevance of individual chance occurrences: A first filter factor is, of course, the individual reader and his or her awareness of (narrated) chance. Obviously, this sensitiveness can vary from one individual to the next according to numerous parameters (including the episteme and other facets of a reader's worldview) that are relevant for the reader, shaping his or her attitude towards events and their possible explanation. As in all reception-oriented theoretical and non-empirical discussions, the elusiveness of the actual reader is a problem that can be resolved only in part by appealing to an 'average informed reader' and textual givens. In the case of chance as a storyworld element, however, its significance may be highlighted (and the reader be alerted) by thematiza-
51
For various similar although partial definitions of this kind of coincidence, cf. Dannenberg (2004: 405), who emphasizes the fact that "an uncanny or striking connection" is created, or Hannay (1988: 89), who speaks of "a convergence through contiguity or similarity of causally unconnected events," and Wetz (1998: 28).
182
Werner Wolf
tion or discussion by intradiegetic characters or extradiegetic (e.g. paratextual) agencies. A second factor to be taken into account is the cultural and historical context in which a text is embedded, and in particular the extent to which this context contains chance-relevant aesthetics or provides of itself chance-sensitive worldviews that interact with a given text. Functionally, this interaction can mean accepting or even stabilising dominant extraliterary worldviews: in this case, texts follow what Lotman has termed an "aesthetics of identity"; 53 but it can also mean highlighting the 'blind spots' of extra-literary worldviews, that is, voicing a criticism or providing alternatives, 54 in which case texts develop an "aesthetics of opposition." 55 In all cases, the text-context interaction can be explicit or implicit. Contextual worldviews that are explicitly referred to in a text are comparatively unproblematic. It is when they are only implicitly alluded to or merely provide backgrounds that they can become problematic filter factors and point to the limits of all text-centred analyses of worldviews, for a lack of explicit reference can mean anything from irrelevance to the acceptance of conventional and therefore tacit, but nonetheless relevant presuppositions. It is here that contextual evidence and an awareness of historical dominants become crucial. For instance, chance occurrences that to twenty-first-century readers may appear improbable, over-contrived or otherwise salient might, if seen from a historical perspective, be perceived as remaining well within the order of things. Chance and its textual saliency are thus cultural and historical variables predicated on ruling systems of causation and belief as well as on epistemes. Another filter factor is the literary genre within whose frame a storyworld unfolds. There are in fact genres in which chance is more likely to occur and in which there is at the same time less emphasis on probability and motivation (always according to the ruling beliefs and epistemes mentioned above) than in others. In the 1970s, Erich Köhler claimed, somewhat inconclusively, that the novella is a genre that has a particular affin52
55
Cf. G o l d k n o p f s pertinent remark that the meaning of a chance incident in a story is largely dependent on "its narrative context" (1969: 44). Cf. Lotman (1972: 188). It is in particular where literary implied worldviews (or parts of them) are 'inscribed in' dominant extra-literary views that their implicit nature is easiest to understand, since what is taken for granted can be passed over in silence. Cf. Iser (1975: 304). In Iser's early reflections on the functions o f fictional literature, however, this potentially active role of literary texts was not sufficiently stressed. Cf. Lotman (1972: 191).
Chance in Fiction as a Privileged Index o f Implied Worldviews
183
ity with chance, since by its very economy it centres on an extraordinary event that remains only partially explained. 56 More convincingly, Wolfgang Preisendanz has since argued that there is a close relationship between comedy and chance. 57 He even dubs an extreme form of comedy, farce, a Kontingenz-Gattung (contingency-bound genre), 58 since it is destined to provide carnivalesque entertainment irrespective of the implausibility of coincidences or accidental occurrences. He also mentions other genres such as the travelogue, romance and the picaresque novel, without, however, granting them the lull status of Kontingenz-Gattung. It is indeed plausible that some genre-typical functions such as providing occasions for laughter, as in farce, but also a parodic intention, are to a large extent conducive to a proliferation of chance happenings in certain types of storyworlds, thereby weakening their indicative value for implied world views. However, in the case of comedy (as also in the epistolary novel, 59
another genre that is particularly open to chance ), this does not entirely rule out the possibility that the occurrence of chance events is related to a worldview-relevant tendency. As Ritter, 60 Warning 61 and others have argued, laughter in comedy frequently serves either to acclaim or to neutralise nonsense and tends to "positivise negativity," as Warning puts it. 62 In fact, in comedy, the inexplicability of chance may well be regarded as a kind of'nonsense' or 'negativity' which is neutralised. 63 Be that as it may, an awareness of generic conventions is always necessary for evaluating the relevance of chance for the implied worldview, even though in par-
56
Köhler (1993 [1973]: 126-27). See Preisendanz (1998b). One need only think o f the occurrence o f a deus ex machina (a positive form o f 'accident') as well as of recognition scenes (a form of coincidence) as conventional ways of winding up a comic plot to see that Preisendanz's thesis is in fact plausible.
58
Preisendanz (1998a: 453). This applies above all to epistolary novels in which the act of writing closely follows the narrated experience, for in this case the discourse appears to follow the contingency o f life in an especially truthful way that seems to suppress narrative constructedness to a maximum and to emphasize subjective reactions to the vicissitudes of life.
60
Ritter (1974: 80). Warning (1976: 3 2 5 - 2 9 ) . Ibid.: 325. It is noteworthy that Preisendanz himself hesitates between a merely 'technical' view of chance in farce as a mere enhancement of entertaining laughter and a perspective which links the proliferation of chance in this genre to a worldview that establishes "contingency as the absolute sign of reality" (see ibid.: 1998a: 451).
61 62
184
Werner Wolf
ticular cases it may be difficult to assess the relative proportions between genre-related and worldview-related chance. A final filter factor is provided by the general constraints of narrativity (some of which have already been touched upon in the discussion of the special affinity between narrativity and chance). This factor is particularly important because these constraints produce in both authors and recipients certain expectations (and therefore a certain tolerance) with regard to chance and at the same time may reduce the significance of individual chance occurrences for the implied worldview. As already indicated, these considerations concern the beginnings of all narratives, where an initial situation is posited that is to a large extent contingent. With a view to this, one may say that there is no narrativity without chance. Indeed, in most cases, the physiognomy of narrative beginnings is inherently contingent, since beginnings are arbitrary and tenuous constructs that could be different or altogether non-existent. In practice, however, their contingency will hardly be considered relevant for the implied worldview, since it is the result of structural necessity. 64 Chance occurrences towards the end of narratives are more difficult to assess: on one hand, they can be indicative of implied worldviews (e.g. when conducive to poetic justice); on the other hand, they must also be regarded as especially dependent on narrative constraints and therefore less so on worldview options, for it is here, at narrative conclusions, that plots have to be wound up, frequently prompting authors to resort to chance (e.^ coincidental recognitions or accidental deus-ex-machina interventions). Generally, the unity of narratives tends to bring about a higher degree of chance occurrence, particularly of coincidences, than would be likely in reality; since this higher frequency is technically motivated, it should often not be considered a significant symptom of a particular worldview. Another factor that also favours the intrusion of chance into narratives independently of worldview considerations has already been mentioned, namely the requirement of "tellability." 66 Tellability, together with suspense (which may enhance
64
It should be noted that the contingency of narrative beginnings affects the level of story and its experience, and not so much the level of discourse, for it is clear that in most stories' beginnings are 'written backwards' and—for the need of narrative coherence— start to prepare the outcome. This is what Goldknopf calls "a 'quick-and-dirty way' o f getting on with the narrative"
66 ( , 9 6 9 : 43 )·
See above, note 34.
Chance in Fiction as a Privileged Index of Implied Worldviews
185
tellability) 67 and concomitant considerations such as the sellability of works of fiction, creating attention for an author in the literary market, etc.—all this may account for the inclusion of spectacular chance events and 'adventures' in narratives. 68 As in the case of generic conventions, an awareness of the structural constraints of narrativity is indispensable for grasping the significance of literary chance phenomena, even though such an awareness can provide no more than a limited guidance for their relevance as worldview indicators. Although these various filter factors can influence the significance of chance occurrences in the analysis of implied worldviews, it is essential to identify the textual features that mark such occurrences. To this end, the following parameters for the analysis of chance in individual narratives and the steps in such an analysis may be helpful: I. Parameters for the analysis of individual chance occurrences in a first reading (where applicable, the more important actualizations of parameters are underlined'): 1. the nature of the chance phenomenon in the storyworld (contingent, accidental, coincidental); 2. the position of the chance occurrence (initial, internal, terminal): 3. the saliency of the chance phenomenon (high, low), depending on: a) the degree of its unexpectedness, (im)plausibility and/or uncanniness as a story element and the concomitant surprise effect; b) the degree of importance of the consequences of the chance occurrence for the plot and for the (principal or secondary) character(s); c) the degree of foregrounding attributed to the chance phenomenon by the discourse; 4. the evaluative quality of the chance element and its consequences with reference to particular characters and intentions (positive, negative; in accordance with the authorial or narratorial guidance of reader sympathies and poetic justice or in discordance with these fac67
Suspense is in fact also related to chance, and in particular to contingency, since it is based on the possibility that something can have an outcome that is different from the one expected. Interestingly, the etymology of 'adventure' seems to include the possibility o f a faterelated meaning, as the term is derived from Latin *adventura, future participle of advenire·. 'that which will arrive or happen'. However, the worldview implication o f a fate-directed cosmos has long been lost, and after the 17th century in most cases 'adventure' has simply been synonymous with "an unusual and exciting, often hazardous, experience or activity" (Pearsall [ed.] [1998: 25], s.v. "adventure").
186
Werner Wolf
tors; tending towards randomness and underdetermination of order and meaning or towards overdetermination); 5. attempted naturalization/explanation of the chance occurrence (nonexistent, implicit, explicit"). II. Steps towards an overall assessment of the role of chance for the description of the implied worldview after the first reading: A. assessment of the importance of chance in the given text (isolated occurrences or foregrounding of chance as a major textual element): a) frequency and extension of representations of chance (see Figure 1, positions IBb, 2B, 3Ba and b, 4B); b) frequency and extension of thematizations of chance (see Figure 1, positions ΙΑ, 2A, 3A and 4A); B. assessment of discernible tendencies in the individual chance occurrences as analysed according to parameters in I); C. taking into account of the intra-compositional/intratextual context relevant to the worldview (e.g. thematizations and/or illustrations of religion, destiny, certain philosophies, non-metaphysical attempts at explaining being(s), happenings and events, etc.); D. considering contextual aesthetics and worldviews or worldview elements that may be relevant to the (presentation of the) worldview implied in the text under scrutiny; E. taking into account the 'filter factors' mentioned above, if they have not as yet been considered (in particular, chance-related generic and narrative constraints or propensities); F. attempting to describe the implied worldview, taking into account parameters 1 to 5 and steps A to E. In most cases, no one-to-one relationship between individual parameters and certain worldviews will be possible, not least because of the multiplicity of elements, parameters and filter factors that must be considered. However, the parameters may indicate at least tendencies that provide some orientation in the assessment of implied worldviews. As already pointed out, for instance, contingency in initial position will not be very relevant in terms of the implied worldview (it may be revealing, however, if even this kind of conventional chance is 'naturalised' by the discourse through references to historical characters and situations, for instance, or to well-known settings). In contrast to initial contingencies, terminal and, above all, internal accidents or coincidences are usually more relevant.
Chance in Fiction as a Privileged Index of Implied Worldviews
187
Thus, chance occurrences that are consonant with the authorial or narratorial guidance of reader sympathies (particularly when these occurrences are in terminal position where, traditionally, poetic justice is achieved) as a rule indicate an optimistic view of some caring Providence, while negative consequences of accidents or coincidences, particularly if inflicted on positive characters, can be related to a wider range of possibilities (e.g. betraying a pessimistic belief in fate, agnosticism or nihilism). Moreover, a high frequency of thematizations and/or illustrations of chance as well as the saliency of individual accidents or coincidences renders it probable that chance is not only a plot feature, but also a thematic element and therefore especially relevant to the implied worldview. As far as the frequency of chance occurrences is concerned, it must be remembered, however, that narratives tend to accustom readers to them and that therefore only a relatively high number of occurrences will give this factor a relevance for the implied worldview. As for the saliency of individual chance occurrences in the reading process, which may also indicate a thematic relevance, this depends on the discernibility of one of two opposing factors: either on a high degree of distancing improbability, or on a high intensity of aesthetic illusion, enabling readers to experience the workings of chance vividly (perhaps in combination with a surprise factor). In the first case, this could, for instance, point to a will behind the text that may be worldview-relevant as laying bare the constructedness and artificiality of the plot, while in the latter case, the calculated immersion of readers into the storyworld could point, on the contrary, to a view of chance as a given, not a construct, in reality—with various further implications relevant for the implied worldview. Of special importance, of course, is the shaping of chance with reference to the question of its 'naturalization' or explanation. Where none or only a cursory explanation is attempted, this could point to an implied worldview in which chance is either no problem at all or an ultimately unaccountable phenomenon, whereas the attempt to naturalise chance and thus reduce it to a mere illusion may indicate either a secular or a metaphysical worldview in which ultimate meaninglessness either does or should not exist.
188
Werner Wolf
4. Historical Case Studies (I): Chance and Implied Worldview in Robert Greene, Pandosto: The Triumph of Time (1588) The guidelines for the analysis of implied worldviews presented above may well seem abstract. Therefore, the following two case studies will be devoted to illustrating their applicability in the practice of interpretation. In these case studies, no exhaustive readings of texts, which have already been copiously studied and have even been the objects of investigations centred on chance, 69 are intended. My purpose is merely to illustrate the parameters and steps of analysis developed in this paper. The first case study is an Elizabethan narrative, Robert Greene's Pandosto: The Triumph of Time. It is commonly known as the source of Shakespeare's A Winter's Tale, but is also interesting in our context, for in this text, which is half a tragedy of jealousy and half a romance, chance looms large and constitutes a privileged access to the implied worldview. The importance of chance in this work is particularly noticeable in recurrent accidents and coincidences that serve as shaping elements of the diegetic world. Thus, at the very moment that the hero, King Pandosto of Bohemia, having received a message from an oracle, becomes ashamed of his unjust jealousy of his wife Bellaria, which had driven him to abandon her allegedly illegitimate daughter Fawnia to the chances of the sea, "word [is] brought him that his young son Garinter was suddenly dead" and that hereupon his wife, too, "fell down presently dead." 70 Fawnia, however, does not drown but arrives, by chance, in "Sicilia" (173). Here, again by chance, a couple of shepherds find her and accept her as their foster child. Years later, Prince Dorastus, after being told that he is to marry some other woman, happens to meet Fawnia and falls in love with her at first sight. They get acquainted and intend to flee from "Sicilia" to Italy. However, this plan almost fails, for Fawnia's foster father, in order to prevent any complications because of the apparent class difference between his family and Dorastus, is about to go to the King and reveal the fact that Fawnia is not his daughter; by coincidence, he is intercepted by one of Dorastus's men and forced on board the ship on which Fawnia and Dorastus manage to flee. The ship, however, does not land in Italy, as
69
For Greene's Pandosto, see Dannenberg (2004: 427-28); for Hardy's Tess, e.g. Dessner (1992) and Monk (1993: 159-66). Greene (1987 [1588]: 171). Henceforth, page references to Pandosto will be indicated in the body of the text.
Chance in Fiction as a Privileged Index of Implied Worldviews
189
intended, but-—again by coincidence—in the country of Fawnia's birth, Bohemia. Here, after sundry complications (including an incestuous interest which Pandosto takes in Fawnia), another coincidence discloses Fawnia's identity to her father in a recognition scene, enabling her to finally be united with Dorastus. Yet, the truth, which time has disclosed to Pandosto, is apparently too much for him, and he now successfully commits an act which he had attempted once before: suicide. This rudimentary summary of a text that is only some fifty pages long already testifies to an unusual frequency of chance elements. Equally frequent in this text are the thematizations of chance, beginning with the mention of "fortune" in the full title of the work 71 and appearing mostly in the narratorial discourse as well as in the characters' speeches (including an extended reflection of Bellaria on her vulnerability to "fortune" because of her high, "fortunate" [164] estate). Thus, in Pandosto, chance is clearly foregrounded as a major thematic element and is therefore of particular relevance for the analysis of the work's implied worldview. In the manifold chance occurrences in Pandosto, which take place irrespective of their position in the text, the following tendencies are discernible: most occurrences in the storyworld are salient owing to their unexpectedness, improbability (perhaps not only for a modern mind) and relevance to the fate of the principal characters, but also due to the concomitant discursive thematizations of "fortune." It is an especially noteworthy tendency in Pandosto that most of the chance elements tend towards an overdetermination of meaning and are thus not mere random occurrences (with one notable exception, to which we will return presently). Whether positive or negative, they are remarkable in that they almost regularly alternate between having lucky and unfortunate effects, regardless of the guidance of reader sympathies for or against characters. This structuring principle is thematised on the discourse level, as for instance: "Fortune, who all this while had showed a friendly face, began now to turn her back and to show a louring countenance [...]" (176). Another tendency, also at discourse level, is equally striking (at least for the modern reader), namely the tendency not to 'naturalise' or explain accidents or coincidences otherwise than by thematising the intervention of fortune, usually before the 71 "Pandosto. The Triumph of Time. Wherein is discovered by a pleasant history that although by the means of sinister fortune truth may be concealed, yet by time in spite of fortune it is most manifestly revealed." (153) This by far exceeds mere "alternate phases of good and bad weather," as Dannenberg (2004: 427) claims.
190
Werner Wolf
happening in question is narrated. As "fortune" is to a large extent an allegory of 'chance', this does not explain chance through reference to a supernatural system of causation, but only highlights it, so that the reader is not given a more enlightened position than the characters and is left to experience chance in the same way. In some cases, notably in the curious double deaths of Garinter and Bellaria, both the story and the discourse levels cooperate in exposing chance to such an extent that "fortune" is not even mentioned. This 'going without saying' is also noteworthy with reference to contingent elements (elements that could also be different), such as Pandosto's jealousy or the perplexing location of Bohemia by the sea. As a rule, narrative 'facts' such as these are merely stated as givens, without any attempt at justification, as revealed in recurrent phrases such as "It happened [...]" or "it fortuned" (179). As for worldview-relevant intracompositional elements that appear to play a role in the motivation for narrative progress besides chance, the following items merit attention: the story is set in vaguely classical times, which justifies the repeated mention of gods, including "Apollo" (172), as well as the consultation of the "Oracle of Apollo" at "Delphos" (168) (once, however, there is also an anachronistic reference to "God" [177]). In addition, "destinies" are repeatedly referred to (166, 188, 199, 202), but only rarely are psychological and other kinds of secular causality mentioned. Interestingly, Pandosto's jealousy is at the very outset thematised in religious terms as a "hellish passion" (155), and Pandosto himself seems to be aware that its workings are implicated in a metaphysical context in which "revenge" (by the gods) plays an important role (172). As the anachronistic mentioning of "God," but even more so the recurrent references to "fortune" betray, the worldview-relevant frame of reference is not a belief in classical gods, but a much later system of meaning. In particular, it is the belief in Fortuna and her "wheel" (156), a "blind" (161) agency full of "inconstancy" (154), which appears so often in medieval and Renaissance thought, texts and iconography. In order not to overestimate the significance of chance in Pandosto, however, the relevant filter factors should be taken into account. This is advisable especially with regard to generic conventions: Pandosto is, after all, a romance in which chance and coincidence that lead to overdetermination of meaning have been traditional ingredients of the genre since its beginnings in classical antiquity. The same attention should be paid to cultural contexts, and in particular to the state of development of prose narrative in the sixteenth century, at which time secular causal explana-
Chance in Fiction as a Privileged Index of Implied Worldviews
191
tions had not yet reached the degree of sophistication displayed in the nineteenth-century realist novel. In addition, one should take into account the narrative constraints that tend towards closure and that may explain in part the coincidence of Fawnia's arrival in Bohemia rather than in Italy, a country which, in contrast to Bohemia, plays no role in the story up to this point. Even so, the treatment of chance in Pandosto is such that it does point to a specific worldview. Clearly, Pandosto constructs a model of reality in which chance or "fortune" is not glossed over, but rather foregrounded as a highly influential shaping factor of reality. The world portrayed by this romance in particular illustrates the "inconstancy" (156) of fortune in accordance with a belief for which Boethius is frequently invoked and which remained dominant throughout the Middle Ages as well as in Renaissance culture. 73 Significantly, this inconstancy and the resulting 'mutability' of earthly matters, which challenges the idea of justice and a meaningful order of the cosmos, is not left unquestioned, but is explicitly thematised by Pandosto when he self-accusingly berates his cruel jealousy and its consequences: Are the gods just? Then let them revenge such brutish cruelty. My innocent babe I have drowned in the seas; my loving wife I have slain with slanderous suspicion; my trusty friend I have sought to betray—and yet the gods are slack to plague such offences. Ah, unjust Apollo, Pandosto is the man that hath committed the fault! Why should Garinter, seely child, abide the pain? (172)
In the end, however, chance appears to be in the service of a just order of the world, since Pandosto's 'prayer' is heard: when he "call[s] to mind" all he has done and commits suicide "in a melancholy fit" (204), he is belatedly "plagued," as he had wished to be. While he thus punishes himself (or seems to submit to a slow 'revenge' of the 'gods'), his innocent victim Fawnia is rewarded with her love, Dorastus. So, at last, "fortune" appears to conform to her afore-mentioned tendency towards overdetermination of meaning and to have worked towards (poetic and providential) justice, thus substantiating a narratorial comment that appears earlier in the story: "But see how fortune is plumed with time's feathers, and how she can minister strange causes to breed strange effects." (178). Yet, this "tam[ing]" of chance 74 is not the entire story, for Pandosto as a whole challenges the claim, put forth by Leland Monk, that "the efface73 74
Cf.Assmann (1998); Haug (1998). Morson (1998: 289).
192
Werner Wolf
IS ment of chance is a structural feature of all narratives." As Pandosto shows, structure and the constraints of narrativity do not necessarily prevent the foregrounding of chance, nor is chance always 'tamed'. Indeed, the world model constructed in Pandosto contains chance events that cannot be "effaced" or explained away by being integrated into a simple teleology where fortune ultimately brings about poetic justice. Were this the case, how could the unjust deaths of two innocents, Garinter and Bellaria, be explained? It is revealing, in this regard, that Garinter's death is a random happening that leads to a marked underdetermination of meaning. It would seem, then, that in Pandosto the workings of chance and time appear to be containable within a meaningful frame only in part, for there remains a blind spot: chance happenings that run counter to an idea of 'just gods'. According to medieval Christian doctrine, the 'mutability' of fortune should be a mere means to teach man a lesson about the fallen world76and direct his gaze towards eternity, where the reign of fortune will end. In Pandosto, however, very little indicates the possibility of such a bracketing or 'taming' of fortune, and nothing points to its ultimate suspension in a metaphysical sphere. Where the gods as representatives of such a metaphysical sphere are mentioned, it is mostly by the characters, 77 while the narrator tends to talk about "fortune," a term which oscillates between denoting a goddess and the personification of an ultimately abstract idea. The existence and nature of a metaphysical sphere would thus remain curiously doubtful—were it not for the Oracle of Apollo. This oracle clearly tells the truth and thus represents an intrusion of the metaphysical sphere into the fictional world. However, it is remarkable that this is not an agency characteristic of the Christian faith, and that it is powerless to prevent the deaths of Garinter and Bellaria. In this respect, the choice of a non-Christian, classical setting is perhaps no coincidence. For the only attitude offered within the storyworld to such incomprehensible workings 75 76
Monk (1998: 4). Cf. Assmann (1998). Interestingly, this perspectival relativization also applies to the 'containment' o f fortune as an agency o f moral retribution, an idea given to Dorastus when he reproaches himself in prison for having disregarded his princely honour and disobeyed his father, thereby incurring the wrath of "the destinies": "Art thou not worthy for thy base mind to have bad fortune? Could the destinies favour thee, which hast forgot thine honour and dignities [in eloping with Fawnia, a seeming shepherdess]? Will not the gods plague him with despite that paineth his father with disobedience?" (199).
Chance in Fiction as a Privileged Index of Implied Worldviews
193
of chance is the Stoic reference to "patience." Interestingly, it is the innocent Bellaria herself, a positive character, who sums up reflections on chance with the idea that "patience is a shield against fortune" (164). Such patience no longer attempts to integrate fortune into a meaningful whole, but simply endures it. Thus, the treatment of chance in Pandosto points to a worldview which refers to the Christian doctrine of the wheel of fortune, but at the same time goes beyond this doctrine. It does so by offering the possibility of a perspective on chance as working primarily in favour of a meaningful whole, but also of counteracting this containment of 'fortune' by an unaccountable residue in which the goddess Fortune herself is subject to chance and only "by chance" serves "a just cause," as one of the narratorial formulations indicates: "Fortune, although blind, yet by chance [!] favouring this just cause [Fawnia's escape from Sicilia], sent them within 78
six days a good gale of wind [...]" (161). The worldview we have sought to identify is not explicitly formulated in this ambivalent way anywhere in the text. It is indeed 'implied', and its relative complexity, even in such an apparently unsophisticated narrative as Pandosto, shows that the analysis of implied worldviews is by no means a simple undertaking. Yet, as has hopefully become clear, investigating the role of chance opens a privileged access to this hidden, yet fundamental layer of literary narratives.
5. Historical Case Studies (II): Chance and Implied Worldview in Thomas Hardy, Tess of the d'Urbervilles
(1891)
Three hundred years after Pandosto, 'Fortuna' was no longer a current personification of chance. The Christian worldview in which Fortune could be contained had eroded for many intellectuals and novelists (not only in the well-known case of George Eliot), and the realist conventions of storytelling that had become dominant in the nineteenth century were largely antagonistic to pure chance as a propelling factor of plots, empha78
As Salzman (1987: xviii-xiv) justly remarks, the probably unintentional "ambivalent sentence" concluding the romance ("Dorastus [...] went with his wife and the dead corpse [of Pandosto] into Bohemia, where, after they were sumptuously entombed, Dorastus ended his days in contented quiet" [204; my emphasis]) "is not unfitting in view of the general tenor" of the text, for yet another unaccountable and unjust chance death would have emphasized the problem of the containment of'fortune' even more.
194
Werner Wolf
sising all kinds of alternative reasons for the existence and progress of the story world, thus producing what could be termed a veritable 'rage for 79
explanation'. Owing to the decline of supernatural systems of causation that formerly could be appealed to, most important in the explanations provided in realist novels are now non-metaphysical, secular factors whose workings often create the impression of determinism with reference to the characters. 80 Hippolyte Taine's formula of "race, milieu et moment [historique]" 81 is a famous (although incomplete) summary of the many factors now appealed to in order to explain developments and occurrences taking place in stories. However, this is not to say that realism as an important context of nineteenth-century fiction, both in terms of aesthetics and worldview, leads to the abolishment of chance (in the form of accident or coincidence) in contemporary storytelling altogether. Rather, the aesthetic commitment to a truthful mimesis of reality that underlies realism includes the admission of chance, even if 82 this creates tensions with the prevailing tendency towards explanation. Thomas Hardy's Tess of the d'Urbervilles is a case in point. It is a novel that is virtually permeated by chance. This not only applies to numerous minor elements, but also to no fewer than five major occurrences that all have decisive consequences for the plot. In addition, all of them, at least at first reading, appear as chance happenings, none of which belongs to the 'harmless' class of initial contingency (i.e. as part of the construction of ground situations): 1) The death of Prince, the horse of the Durbeyfields (1/4): this accident drives Tess's poor family further into poverty and motivates her to reconsider her aversion to visiting her alleged relatives, the d'Urbervilles, at Trantridge in order to look for work. She is indeed offered a job there and becomes the focus of Alec d'Urbervilles' attentions. 2) A quarrel that breaks out among the servant girls going home at night after a market dance near Trantridge (1/10): Tess, who is threatened by a fight, is led by this "juncture" 83 to accept Alec's 79 80 81 82
Cf. D.F. Bell (1993: 1-4). Cf. Wolf (1998). Taine (1893-95 [1863]: xxii). Cf. D.F. Bell (1993: 4-5). Hardy (1984 [1891]: 64). Henceforth, page references to Tess (including the notes) will be indicated in the body of the text.
Chance in Fiction as a Privileged Index of Implied Worldviews
195
offer and rides home with him on horseback. Alec, however, loses his way and leaves Tess alone while trying to find his whereabouts. Meanwhile, Tess has fallen asleep and cannot resist Alec, when, on his return, he deflowers her. 3) Tess's letter of confession, which does not reach its addressee (IV/33): in an attempt to flee her past (which includes the death of the baby Alec had fathered), Tess becomes a dairy maid at Talbothays, falls in love with Angel Clare and intends to make a clear breast of her history in a letter to him before their wedding. By chance, however, Angel never reads the letter, and Tess destroys it after having found out about the failure of communication. When Angel finally learns about Tess's 'fault' during their wedding night, this unexpected information makes him think that he married Tess on a false belief in her purity. Deeply frustrated, he rejects her and leaves for South America. 4) The coincidental meeting with Alec: one year later, Tess happens to come across Alec again, who has become an itinerant preacher but is still interested in her (V/44). 5) Yet further letters that do not reach their addressees as intended: Tess's letter to Angel, imploring him to come back to her in order to relieve her from her precarious position, reaches him too late, and a second letter arrives only when he has already returned to England (VI/48, VII/53). As a consequence, Tess is led to believe by Alec that Angel will never come back, and she becomes Alec's mistress. When Angel finally finds Tess again, Alec's alleged untruthfulness makes her hate him to the point of assassinating him. This leads to her arrest after a short, belated honeymoon with Angel and ultimately to her execution as a murderess. Among the discernible tendencies in these chance occurrences, one stands out in particular, namely that—in contrast to the situation in Pandosto—all are negative in their effects. In appearance, many belong to the class of 'underdetermined', random occurrences that simply happen (e.g. the accident of the horse or Tess's misdirected confessional letter), although when looked at more closely, they appear to convey some kind of negative overdetermination. However it may be, it seems as though the discourse, following the illusionist aesthetic of celare artem, is gauged to reducing this impression to the extent this is possible—although it can hardly be said to be completely successful in doing so.
196
Werner Wolf
This tendency towards a playing down of chance can already be seen in the area of thematization. Here, too, Tess deviates from Greene's romance in that chance is less frequently mentioned as such and never discussed in any depth. Interestingly, some thematizations even run counter to the impression of chance as a negative force which some incidents, but also some thematizations such as "misfortune" (e.g. 62) and "doom [...]" (e.g. 91), may create. This is the case when Angel says that Tess's letter reached him "by merest chance" (358), for the letter did in fact reach him as intended (albeit too late), or when, on an earlier occasion, Angel learns the pre-history of this Tess and thinks that she whom he believed to be the very embodiment of 'purity' has turned into something quite different, owing to "the contingencies of human experience" (227). Here, the narrator explicitly contradicts Angel ("He argued erroneously" [227]), implying that Tess—as the subtitle of the novel also indicates—is still "A pure woman" regardless of her sexual history. A further remarkable tendency concerning the use of chance in Tess (in yet another departure from Pandosto) can be seen in the many attempts at giving plausible explanations or at 'naturalising' chance. This is sometimes the consequence of previously narrated story elements, but is more frequently an effect produced at discourse level, where, before or after the incident in question, reasons are mentioned that reduce the impression of implausible chance. As has already been remarked with reference to Balzac's novels, discourse can thus serve to counteract the more or less manifest 'contingency' of the story. 84 Thus, a traditionally contingent plot element, such as the heroine falling in love with the hero, is in Tess no conventional 'love at first sight' (as in Pandosto), but is subtly rendered plausible on psychological grounds: Angel meets Tess a first time at a country dance, but without dancing with her (1/2). On their second meeting, he dimly remembers this incident, which makes him pay more attention to her. This psychological motivation is explicitly adduced on the discourse level as an explanation: "A casual encounter [...] the circumstance was sufficient to lead him to select Tess in preference to the other pretty milkmaids [...]" (116). As for the death of Prince, caused in part by Tess's falling asleep, this accident is rendered plausible by the fact that Tess had to stand in for her drunkard of a father after sleeping no more than two and a half hours the previous night. The fatal outcome of the accident is further accounted for afterwards by the description of the "morning-mail 84
Cf. Warning (1980: 4 2 ^ 3 ) .
Chance in Fiction as a Privileged Index of Implied Worldviews
197
cart": "[...] with its two noiseless wheels, speeding along these lanes like an arrow, as it always did, [it] had driven into her slow and unlighted equipage. The pointed shaft of the cart had entered the breast of the unhappy Prince like a sword [...]" (26). In a similar way, we get ample and verisimilar information about the incident which triggers the quarrel that brings Alec and Tess together on the fateful night (cf. 62); the same is true of Tess's confessional letter which she had thrust under Angel's door and about which we learn afterwards that he "obviously had never seen [it], owing to her having in her haste thrust it beneath the carpet as well as beneath the door" (204). Finally, Tess's fateful re-encounter with Alec, which stands out particularly, as for once it is an incident that is in no way prepared by the discourse or by previous story happenings, also receives a retrospective explanatory comment by the narrator when he points to the frequency of "transfiguration[s]" (295) from "sinner" to "saint" in "Christian history" (296). Remarkable, too, is the fact that from that point onwards, no major chance elements interfere with the plot. Unlike its distribution in Pandosto, chance thus occurs predominantly in an internal position (if we discount the usual initial contingent constructions of the fictional world), while the last 'act', as in many tragedies, appears to be the almost necessary result of previous developments and manages to unfold without the aid of (overdetermining) chance. The attempts at naturalising chance in Tess accord with other worldview-relevant tendencies that are equally in harmony with the context of realism. Among them are suggestions that a plethora of factors besides chance are at work in shaping Tess's life. Taine's triad of 'race', 'social milieu' and 'history' looms large in this context: as far as 'race' is concerned (which in Hardy's novel means a Darwinian influence of hereditary factors), the narrator says about Tess, a descendant of the d'Urbervilles, that she has a "slight incautiousness of character inherited from her race" (87), an inclination which is also shown in the murderous violence occurring in the "legend" of the "d'Urberville Coach" (344) and which Angel explicitly invokes when learning about Tess's having killed Alec ("he [...] wondered what obscure strain in the d'Urberville blood had led to this aberration" [375]). 'Race' also influences Tess in the form of general human nature, which means not only an "invincible instinct towards self-delight" (96), but also, as in Alec's dealings with Tess, "animalism" (296). As for 'milieu', it is repeatedly made clear in the novel that Tess's proletarian origin together with its staple realist attributions of poverty, alcoholism and ignorance is partly responsible for her tragedy.
198
Werner Wolf
Something analogous is true of Angel and his middle-class upbringing, which has deeply affected him with Wordsworthian optimism, bourgeois Victorianism and the Christian belief in purity and the ultimate goodness of creation. As for the role of history, it is perpetually present in the influence of Tess's past on her life, but also in the conventions of the 'historical moment' in which the story is set, particularly with regard to the (criticised) Victorian double standard in which premarital sex is much more reprehensible in women than in men. Additional shaping forces that are also relevant to the implied worldview—notably as factors that relativise the impact of chance—are contingent although not particularly improbable idiosyncrasies of the characters: most notable are Angel's false Romantic beliefs and Tess's interiorisation of Victorian norms resulting in her feelings of guilt—"mistaken creations of Tess' fancy" (82), as the omniscient authorial narrator critically qualifies them. Quite in the same line of secular and materialistic explanations is the fact that the idea of a benevolent agency influencing life, in particular the existence of the Christian loving God and a Wordsworthian belief in benevolent Nature and her "holy plan" (18), is repudiated over and again and is referred to by the narrator at best in an ironic way. 85 Yet, even when taking these realist filter factors into account, we get the impression that in Hardy's novel chance is more important than merely having a life-like share in shaping Tess's biography. This is true even if we consider further filter factors, in particular the constraints of narrativity (such constraints may account for Tess's meeting Alec a second time, for example, since the unity of the story, one of the basic narratemes, is thereby enhanced). In this context, one is also tempted to mention the generic conventions of tragedy, for Tess is clearly affiliated with this genre. However, this generic reference points to the deep tension that informs the worldview implied in this novel, for the secular realism which seems to dominate here is clearly antagonistic to tragedy, as this genre typically presupposes the belief in, if not necessarily an interference of, some transcendental agency. Yet, such an agency is conspicuously and disturbingly invoked by the narrator in his famous concluding comment 85 Ironically, too, Tess has the impression of a "Providential interposition" (217) when Angel precedes her intended confession by a confession of his own; however, the disastrous outcome of the scene belies the idea of Providence. In a similar way, Alec invokes "Heaven" as the ultimate cause of what happens to him (298); yet, here again, the consequences—namely his first luring Tess into an adulterous relationship and then his being killed by her—contradict this idea.
Chance in Fiction as a Privileged Index of Implied Worldviews
199
on Tess's execution: "'Justice' was done, and the President of the Immortals, in Aeschylean phrase, had ended his sport with Tess" (388). What is also disturbing is the fact that on closer inspection, the 'explanations' offered for the manifold chance occurrences sometimes account only for parts of these occurrences, while major elements are left in the dark. Thus, the re-encounter between Tess and Alec that initiates the final sequence of events is well explained as far as the question of how and why Alec was converted to Christianity is concerned, but it is unclear why Tess happens to come to exactly the spot where Alec is preaching. This remains a pure coincidence—with fateful consequences. What undermines the gesture towards secular and materialistic explanations in the novel yet further is the sheer number of minor 86 and major chance elements. Taken together, they produce a degree of improbability that seems barely compatible with what are traditionally taken as realist conventions 87 and rather foregrounds chance. Even so, chance does not seem to be a merely blind agency, either; this, at least, is what the repeated invocations of "Fate" (e.g. 216) as well as the mention of negative symbolism (concerning the colour red, for example 88 ) and "ill-omen[s]" (344) suggest. Admittedly, some of these mentions are linked to the unreliable perspectives of individual characters, but others are foregrounded to such an extent that one can hardly avoid the impression that there is after all some sinister agency which influences the fictional world. Particularly startling examples are the three afternoon crowings of a cock before Tess's wedding (cf. 209), which the dairyman interprets as an ill omen and which are mentioned twice in the following chapter (211, 215). The allusion is, of course, to Saint Peter's betraying Christ, the innocent lamb of God, and thus an ominous parallel, tinged with metaphysical overtones, is created to Angel's betrayal of the innocent Tess. The crowing is also connected with an attempted suicide of a secondary character, ironically triggering Tess's "final determination" to tell Angel about her "fault" (216). 86
88
Among the minor chance occurrences are the repeated coincidental meetings between Tess and farmer Groby, with whom Angel had a fight because of the farmer's insinuations about her, and the "odd coincidence" that Alec had a "rumpus with [Angel's] father" (183) and as a consequence of this became a Christian. Numerous parallels and mises en abyme (e.g. dreams) that seemingly occur 'by chance' further contribute to the novel's dense web of meaning. The fact that chance continues to play a role even in realist fiction has, however, gained some currency in research (cf. Goldknopf [1969]; Forsyth [1985]; Dessner [1992]; Monk [1993]). Cf. Tanner (1971 [1968]).
200
Werner Wolf
Thus, it appears that the most disturbing aspect of the employment of chance in Tess is not so much its uneasy combination with an overdetermination of possible causes for the plot development (to which repeated suggestions of a cyclical, non-teleological view of history in opposition to static nature must be added), but the overwhelmingly negative tendency of both immanent determinism and chance occurrences. In contrast to this, the positive potential of 'chance' is also mentioned, but characteristically not actualised (as when Tess wants to contact Angel's family whom she knows to be hostile towards her and "hoped for some accident that might favour her, but nothing favoured her" [288]).89 In fact, all her plans, all her attempts at "escaping] [a] past" (95), which proves to be "implacable" in "engirdl[ing] her" (297), are inevitably frustrated—or, as one is tempted to say, are 'doomed' to fail. In spite of powerful contrary tendencies, chance, as in Pandosto, thus remains a conspicuous factor in the fictional world, although it is transmitted differently and in its negative overdetermination betrays a different drift. Evaluating it, however, is particularly difficult, and even more so than in Pandosto, as it is implicated in a tension that defies resolution. On one hand, there is a clear tendency, in harmony with what one might expect of realism, to naturalise and contain chance as much as possible within a net of natural causes, so that it loses the overbearing quality it still had in Pandosto and so that truly unaccountable accidents and coincidences appear to have no larger a share than in real life. On the other hand, the high frequency of chance, its saliency as a shaping plot factor, the repeated mentions of popular 'superstitions' and omens, but also the symbolism and the fateful, tragedy-like vocabulary which the discourse never tires of using, seem to point towards something that goes beyond both immanent causes and the mimesis of a probable share of chance in everyday life: some negative transcendental agency, "something outside" the characters (274) and their world. It is well known that Hardy "denied [...] that he literally believed in an 'all-powerful being endowed with the basic human passions, who turns everything to evil and rejoices in the mischief he has wrought'" (414), and, remarkably, the narrator's terminal reference to Greek tragedy is not devoid of irony with reference to the premise of Greek tragedy itself, namely the existence of the gods and in 89
Monk's remark that "[h]ap is [...] always mwhap in Hardy's novels" and creates the impression of "an inverted Providence, which is indifferent or malevolent" (1993: 165, 158) is thus certainly true of Tess.
Chance in Fiction as a Privileged Index of Implied Worldviews
201
particular their king (not their "President"!), Zeus. Yet, all of this does not suffice to dispel the impression that chance, in Hardy's fictional world, is distinctly ambiguous, to say the least.90 It may to a certain extent be 'mere chance', but it may also be the effect of some Schopenhauerian 'Immanent Will', which Hardy later thematised in his poetry.91 The investigation into the role of chance in Tess thus reveals an implied worldview that is curiously ambivalent, wavering between immanent factors (be they materialistic, psychological or social) and the (perhaps merely poetic) suggestion of some kind of negativity or 'Fate'. At any rate, the negative determinism that reigns in Hardy's later novels conveys an aura of 'pessimism' which has dominated critics' responses ever since, and which Hardy himself had already commented on in his "General Preface to the Wessex Edition of 1912." In reaction against the reproach that his "philosophy of life" was "pessimistic," he interestingly denied that his novels were based on "a coherent scientific theory of the universe," about which he was rather sceptical anyway ("possibly there exists no comprehension of it anywhere"), and appealed to an experiential kind of "truth" ("impressions of the moment") which allegedly informs his writings (418-19). Revealingly, he sets this 'truth' in a dichotomy, the second term of which clearly has his preference: "Existence is either ordered in a certain way, or it is not so ordered" (419). In fact, the worldview implied in Tess does not show any 'certain order' of existence, but rather a web of factors that crush a more or less innocent victim whose last reaction, with her fatalistic acceptance of the state of affairs without trying to account for them, embodies a remarkable stance in view of an unknowable order of the world: "It is as it should be [...] I am ready [...]" (386).92 Among the factors that influence this world, one is conspicuously absent—a benevolent God (or "Tess's guardian angel" [69])93—while others (immanent rational causes as well as chance) certainly belong to the web of existence. Whether chance is 'pure chance' or is possibly 90
91
It must also be stressed that contextual evidence (in this case, Hardy's selfinterpretations and later views) is less important to the implied worldview in Tess than what is suggested by the text itself. Cf. Miller (1975); Monk (1993: 161-66). This "fatalistic" stance is also attributed to "Tess's own people down in those retreats" (70); on this acceptance of reality as such, see Tanner (1971 [1968]: 429). Cf. Angel's momentary reformulations of Browning's lines, namely that "God's not in heaven: all's wrong with the world" (246), which is all the more significant in that he is a believer and the son o f a clergyman.
202
Werner Wolf
linked to some immanent or transcendental negativity and to the relative importance of immanent determining factors (let alone the role left to something like free will in all of this) is not certain at all, and this very uncertainty points to a profound agnosticism. In this respect, another narratorial comment is revealing: "many thousand years of analytical philosophy have failed to explain [this problem of chance] to our sense of order" (69-70). 9 4 In spite of these uncertainties, one thing is certain: namely the fact that an analysis of the role of chance allows for a crucial insight into to the worldview implied in the work as well as into its ambiguities and ultimate "indeterminacy."95
6. Conclusion: basic functions of storytelling in dealing with chance, and the problematics of a textual theory of indicators of implied worldviews as a challenge to narratology The foregoing case studies could, of course, not cover the development of the treatment of chance in the history of English fiction in any detailed way. A more extended study of chance and narrativity therefore remains a veritable desideratum. At a minimum, such a study should also include sections on chance in eighteenth-century as well as in modernist and postmodernist fiction.96 As for postmodernism, an obvious example would be Paul Auster's fiction, e.g. City of Glass (1985) or The Music of Chance (1990), in which chance in the sense of a randomness with predominantly 94
This narratorial comment is triggered by the previous recounting of Tess's defloration and the question of "why so often the coarse appropriates the finer thus, the wrong man the wrong woman, the wrong woman the wrong man" (69) and revealingly leads to an ambivalence: "the possibility of a retribution lurking in the present catastrophe. Doubtless some of Tess d'Urberville's mailed ancestors rollicking home from a fray had dealt the same measure even more ruthlessly towards peasant girls of their time"; this metaphysical explanation is, however, immediately undermined: "But though to visit the sins of the fathers upon the children may be a morality good enough for divinities, it is scorned by average human nature; and it therefore does not mend the matter." (70). "[Uncertainty or indeterminacy" are the terms used by Waldorff (1979: 136) to describe the various suggestions offered in the novel to account for Tess's 'fate'. Richardson's (1997) ground-breaking inquiry into narrative causation contains a revealing chapter titled "The Fortunes of Chance" (1997: 20-31) as well as a historical overview (39—43) of "forms of causal setting" (39) which, in spite of his narrow conception of 'chance' (see above, note 47), form excellent starting points for a more detailed investigation beyond the confines of his centre of interest, namely modernist and postmodernist fiction.
Chance in Fiction as a Privileged Index of Implied Worldviews
203
negative effects is foregrounded as the ultimate reality and where former realist attempts at taming chance by setting it into a context of causal explanations are exposed as mere arbitrary fictions. For want of space, our case studies on Pandosto and Tess must, however, suffice here. 97 They are at any rate sufficient to illustrate some basic relationships between narrative and chance, and above all the overall role of storytelling in dealing with chance. In this respect, narratives can fulfil the following general functions: 1) Offering a way of coming to terms with chance by inserting it into a frame that provides it with meaning (this is the case in Pandosto); 2) Inducing the recipient to reflect on the role of chance as well as on other factors that may influence existence: fate, materialistic, psychological, historical and other causes (this aspect is prominent in Tess, and even more so in Auster's fiction, although with different and even more sceptical results); 3) Encouraging recipients to playfully exercise their tolerance of chance in real life 98 ; in this exercise, the reading of fiction is analogous to playing games involving chance (this relationship may also be seen in Pandosto and perhaps in Auster's fiction, too, while Hardy's seems too gloomy for such a playful aspect). Besides these three functions, one should also mention a fourth, which is present only in some texts, albeit in famous ones such as Boccaccio's II Decamerone (1349-51) and in Sterne's Tristram Shandy (1759-67): 4) Conveying the impression that reading or storytelling for that matter can temporarily suspend the effects of chance (in particular, the 97
For more details on Auster, see Woods (1995) and Einzinger (2005), to whose M.A. thesis written under my direction I am particularly indebted. Excellent further texts from the field of postmodernist fiction for testing the relevance of chance for the analysis of impled worldviews would be David Lodge's metafictional academic romance Small World (1984) as well as McEwan's most recent novel, Saturday (2005), in which chance not only plays a prominent role on the plot level, but is also repeatedly thematised in connection with evolution, the formation of individuals and other philosophical issues. Cf. also, in the field of film, Woody Allen's recent Match Point (2005), which may be said to be a tragic-comic meditation on the irreducible contingency of life. Cf. Haug (1998: 167), who also emphasizes the playful exercise of enduring contingency (Einübung in Kontingenz). In some texts, such as Marivaux's Le Jeu de l'amour et du hasard (1730), this playful aspect is already apparent in the title, which refers not only to the 'whims' of love and chance as subjects of the comedy, but is also a metafictional indication of the playful nature of their comic representation.
204
Werner Wolf
negative effects, including the most threatening aspect of chance, namely death). In this context, it is revealing that Boccaccio's genteel society of refugees from plague-ridden Florence and Tristram Shandy all flee from the contingency of death into storytelling. This is in fact more than a mere passtime, for it seems that while they, like Sheherazade, are occupied with their narratives and engage in the more or less inevitable construction of meaning which most if not all storytelling entails, death, as the ultimate challenge to the meaningfulness of our existence, has no power over them. These general functional relationships between chance, individual narratives and the worldviews implied in them, together with the possible findings of a yet-to-be-written history of chance in fiction, may appear as further complications of an already complicated matter, namely the interpretation of textual elements in narratives as indicators of implied worldviews. In fact, owing to the plurality and complexity of elements and factors that, as the foregoing theoretical remarks have shown, ought to be taken into account, there can be no clear one-to-one relationship between textual phenomena and certain worldviews. In particular, one must not forget that even among the more privileged textual factors, the use of chance is only one element. Others include the endings of narratives (in particular with a view to poetic justice), the occurrence and shaping of descriptions (notably the descriptions of characters' physiognomies"), the symbolism and imagery used in a text, the guidance of reader sympathies for or against characters who are given worldview-relevant opinions, the reliability and plurality of cognitive and evaluative perspectives in a narrative and—as far as implied aesthetics (an aesthetic counterpart to 'implied worldview') is concerned—metatextual (metafictional) elements. Especially interesting among these alternatives for our purpose are the ways in which the existence, nature and development of story elements (particularly concerning characters, their actions and what happens to them) are explained by, and thus point to, certain "systems of causation," 100 since chance, as stated above with reference to Brian Richardson (1997), is a prominent, mostly (but not exclusively) negative or disturbing phenomenon in the context of these explanatory systems.
99
For the relationship between descriptions of physiognomies and implied worldviews (in particular epistemes), see Wolf (2002a), (2002c). 100 Richard son (1997: 62).
Chance in Fiction as a Privileged Index o f Implied Worldviews
205
In sum, one must therefore admit that in the present state of narratology, and perhaps for reasons of principle, no formalised method of analysing implied worldviews is possible. Even so, this does not mean that contributing a foundation and some 'building blocks' to a theory of implied worldviews in narratives, even though it may ultimately remain fragmentary, is altogether useless. As modes of dealing with chance belong to the central issues of all worldviews, and as at the same time chance (including its containment or neutralisation) is such a fundamental element in storytelling, its analysis is certainly a particularly important building block, if not a cornerstone in such a theory. At any rate, chance in the various forms in which it occurs in narratives and with its manifold functions is a challenge to narratology, particularly in its present reorientation towards functional and historical issues. If, in addition to highlighting the textual manifestations of chance as one possible way of answering this challenge, the preceding remarks also help to raise an awareness that the functions of narrative world-modelling, both from a historical and a theoretical perspective, merit more critical attention than has been devoted to this issue so far, they have fulfilled their purpose.
References Adams, Jon-K. 1996 Narrative Explanation: A Pragmatic Theory of Discourse. American Studies 7 (Frankfurt am Main, etc.: Lang). Assmann, Aleida
Aachen British and
1998
"Let it be: Kontingenz und Ordnung in Schicksalsvorstellungen bei Chaucer, Boethius und Shakespeare," in Kontingenz, edited by Gerhart v. Graevenitz and Odo Marquard. Poetik und Hermeneutik 17: 2 2 5 - 4 4 (Munich: Fink). Bakhtin, Mikhail Mikhailovich 1981
The Dialogic Imagination, edited by Michael Holquist, translated by Caryl Emerson and Michael Holquist. University of Texas Press Slavic Series 1 (Austin: University o f Texas Press). Bell, David F. 1993
Circumstances: Chance in the Literary Text (Lincoln: University of Nebraska Press). Bell, Michael 1990 "How Primordial is Narrative?," in Narrative in Culture: The Uses of Storytelling in the Sciences, Philosophy, and Literature, edited by Cristopher Nash. Warwick Studies in Philosophy and Literature, 1 7 2 - 9 8 (London and N e w York: Routledge).
206
Werner Wolf
Booth, Wayne C. 1983 The Rhetoric of Fiction. 2 nd ed. (1st ed. 1961) (Chicago: University of Chicago Press). Bruner, Jerome 1990 Acts of Meaning, The Jerusalem-Harvard Lectures (Cambridge, Mass: Harvard UP). 2002 Making Stories: Law, Literature, Life (Cambridge, Mass: Harvard UP). Chatman, Seymour 1978 Story and Discourse: Narrative Structure in Fiction and Film (Ithaca and London: Cornell UP). Dannenberg, Hilary P. 2004 "A Poetics of Coincidence in Narrative Fiction," in Poetics Today 25.3: 3 9 9 436. Dessner, Lawrence Jay 1992 "Space, Time, and Coincidence in Hardy," in Studies in the Novel 24: 154-72. Doleiel, Lubomir 1998 Heterocosmica. Fiction and Possible Worlds. Parallax (Baltimore and London: The Johns Hopkins UP). Einzinger, Eva 2005 "Chance in Selected Novels by Paul Auster: Case Studies of Contingency in Postmodernist Fiction" (M.A. Thesis, University of Graz, Austria). Forsyth, Neil 1985 "Wonderful Chains: Dickens and Coincidence," in Modern Philology 83.2: 15165. Foucault, Michel 1966 Les Mots et les choses: Une archeologie des sciences humaines. Bibliotheque des sciences humaines (Paris: Gallimard). Goldknopf, David 1969 "What Coincidence Means in the Victorian Novel," in College English 31: 4 1 50. Goodman, Nelson 1978 WaysofWorldmaking (Indianapolis: Hackett). Greene, Robert 1987 [1588] Pandosto: The Triumph of Time, in An Anthology of Elizabethan Prose Fiction, edited by Paul Salzman. The World's Classics, 151-204 (Oxford: Oxford UP). Gymnich, Marion / Nünning, Ansgar (eds.) 2005 Funktionen von Literatur: Theoretische Grundlagen und Modellinterpretationen. Studies in English Literary and Cultural History 16 (Trier: WVT). Hannay, John 1988 "What Joyce Can Teach Us About Coincidence," in University of Dayton Review 19: 89-97. Hardy, Jörg 1996 "Zufall," in Metzler Philosophie Lexikon, edited by Peter Prechtl and FranzPeter Burkard, 590-91 (Stuttgart: Metzler).
Chance in Fiction as a Privileged Index of Implied Worldviews
207
Hardy, Thomas 1984 [1891] Tess of the d'Urbervilles, edited by James Gibson (London: Dent). Haug, Walter 1998 "Kontingenz als Spiel und das Spiel mit der Kontingenz: Zufall, literarisch, im Mittelalter und in der frühen Neuzeit," in Kontingenz, edited by Gerhart ν. Graevenitz and Odo Marquard. Poetik und Hermeneutik 17: 151-72 (Munich: Fink). Herman, David 2003 "Stories as a Tool for Thinking," in Narrative Theory and the Cognitive Sciences, edited by David Herman. CSLI Lecture Notes 158: 163-92 (Stanford: Center for the Study of Language and Information). Iser, Wolfgang 1975 "Die Wirklichkeit der Fiktion: Elemente eines funktionsgeschichtlichen Textmodells der Literatur," in Rezeptionsästhetik: Theorie und Praxis, edited by Rainer Warning. UTB 303: 277-324 (Munich: Fink). Köhler, Erich 1993 [1973] Der literarische Zufall, das Mögliche und die Notwendigkeit (Frankfurt am Main: Fischer). Kranz, Margarita / Knebel, Sven K. / Niessen, Alexandra / Hoffmann, Arnd / Schmidt, Jan 2004 "Zufall," in Historisches Wörterbuch der Philosophie, edited by Joachim Ritter and Karlfried Gründer, vol. 12: 1408-24 (Basel and Stuttgart: Schwabe). Lotman, Jurij M. 1972 Vorlesungen zu einer strukturalen Poetik, edited by Κ. Eimermacher, translated by Waltraud Jachnow. Theorie und Geschichte der Literatur und der schönen Künste (Munich: Fink). Lotman, Jurij 1977 [1970] The Structure of the Artistic Text, translated by Gail Lenhoff and Ronald Vroon. Michigan Slavic Contributions (Ann Arbor: University of Michigan Press). Miller, J. Hillis 1975 "Fiction and Repetition: Tess of the d'Urbervilles," in Forms of Modern British Fiction, edited by Alan Warren Friedman. Symposia in the arts and the humanities 2 : 4 3 - 7 1 (Austin and London: University o f T e x a s Press). 1990 "Narrative," in Critical Terms for Literary Study, edited by Frank Lentricchia and Thomas McLaughlin, 66-79 (London and Chicago: University of Chicago Press). Mink, Louis O. 1978 "Narrative Form as a Cognitive Instrument," in The Writing of History: Literary Form and Historical Understanding, edited by Robert H. Canary and Henry Kozicki, 129-49 (Madison: University of Wisconsin Press). Monk, Leland 1993 Standard Deviations: Chance and the Modern British Novel (Stanford, CA: Stanford UP). Morson, Gary Saul 1998 "Contingency and Poetics," in Philosophy and Literature 22: 286-308.
208
Werner Wolf
Nünning, Ansgar 1989 Grundzüge eines kommunikationstheoretischen Modells der erzählerischen Vermittlung: Die Funktionen der Erzählinstanz in den Romanen George Eliots. Horizonte 2 (Trier: WVT). 1993 "Renaissance eines anthropomorphisierten Passepartouts oder Nachruf auf ein literaturkritisches Phantom? Überlegungen und Alternativen zum Konzept des 'implied author'," in Deutsche Vierteljahrsschrift für Literaturwissenschaft und Geistesgeschichte 67: 1-25. Paulson, William 1994 "Chance, Complexity, and Narrative Explanation," in SubStance: Α Review of Theory and Literary Criticism [74] 23.2: 5-21. Pearsall, Judy (ed.) 1998 The New Oxford Dictionary of English (Oxford: Clarendon). Pfister, Manfred 1977 Das Drama. UTB 580 (Munich: Fink). Pratt, Mary Louise 1977 Toward a Speech Act Theory of Literary Discourse (Bloomington and London: Indiana UP). Preisendanz, Wolfgang 1998a "Gibt es Kontingenzgattungen?," in Kontingenz, edited by Gerhart v. Graevenitz and Odo Marquard. Poetik und Hermeneutik 17: 451-53 (Munich: Fink). 1998b "Komik als Komplement der Erfassung von Kontingenzen," in Kontingenz, edited by Gerhart ν. Graevenitz and Odo Marquard. Poetik und Hermeneutik 17: 383-401 (Munich: Fink). Prince, Gerald 1982 Narratology: The Form and Functioning of Meaning. Janua Linguarum: Series Maior 108 (Berlin, etc.: Mouton). 1996 "Remarks on Narrativity," in Perspectives on Narratology: Papers from the Stockholm Symposium on Narratology, edited by Claes Wahlin, 95-106 (Frankfiirt am Main: Peter Lang). Richardson, Brian 1997 Unlikely Stories. Causality and the Nature of Modern Narrative (Newark: University of Delaware Press; London: Associated University Presses). Ricceur, Paul 1983-85 Temps et Ricit, 3 vols. L'ordre philosophique (Paris: Editions du Seuil). Ritter, Joachim 1974 [1940] "Über das Lachen," in Joachim Ritter, Subjektivität. Bibliothek Suhrkamp 379: 62-92 (Frankfurt/M.: Suhrkamp). Ryan, Marie-Laure 1991 Possible Worlds, Artificial Intelligence and Narrative Theory (Bloomington and Indianapolis: Indiana UP). 2005 "Tellability," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn, and Marie-Laure Ryan, 589-91 (London and New York: Routledge).
Chance in Fiction as a Privileged Index of Implied Worldviews
209
Salzman, Paul 1987 "Introduction," in An Anthology of Elizabethan Prose Fiction, edited by Paul Salzman. The World's Classics, vii-xxxiv (Oxford: Oxford UP). Stierle, Karlheinz 1979 "Erfahrung und narrative Form: Bemerkungen zu ihrem Zusammenhang in Fiktion und Historiographie," in Theorie und Erzählung in der Geschichte, edited by Jürgen Kocka and Thomas Nipperdey. Theorie der Geschichte: Beiträge zur Historik 3: 85-118 (Munich: dtv). Taine, Hippolyte 1893-95 "Introduction," in Histoire de la littirature anglaise. 9 ,h ed. (I s t ed. 1863), vol. 1, iii-xlix (Paris: Hachette). Tanner, Tony 1971 [1968] "Colour and Movement in Hardy's Tess of the d'Urbervilles," in The Victorian Novel: Modern Essays in Criticism, edited by Ian Watt, 407-31 (London: Oxford UP). Virtanen, Tuija 1992 "Issues of Text Typology: Narrative—A 'Basic' Type of Text?," in Text 12.2: 293-310. Waldorff, Leon 1979 "Psychological Determinism in Tess of the d'Urbervilles," in Critical Approaches to the Fiction of Thomas Hardy, edited by Dale Kramer, 135-54 (London and Basingstoke: Macmillan). Warning, Rainer 1976 "Elemente einer Pragmasemiotik der Komödie," in Das Komische, edited by Wolfgang Preisendanz and Rainer Warning. Poetik und Hermeneutik 17: 2 7 9 333 (Munich: Fink). 1980 "Chaos und Kosmos: Kontingenzbewältigung in der Comedie humaine," in Honore de Balzac, edited by Hans-Ulrich Gumbrecht, Karlheinz Stierle and Rainer Warning. UTB 977: 9-55 (Munich: Fink). 2001 "Erzählen im Paradigma: Kontingenzbewältigung und Kontingenzexposition," in Romanistisches Jahrbuch 52: 176-209. Wetz, Franz Josef 1998 "Die Begriffe 'Zufall' und 'Kontingenz'," in Kontingenz, edited by Gerhart ν. Graevenitz and Odo Marquard. Poetik und Hermeneutik 17: 2 7 - 3 4 (Munich: Fink). White, Hayden 1987 The Content of the Form: Narrative Discourse and Historical Representation (Baltimore and London: The Johns Hopkins UP). Wolf, Werner 1998 " Ί must go back a little to explain [her] motives [...]'—Erklärung und Erklärbarkeit menschlichen Verhaltens, Handelns und Wesens in englischen Romanen des Realismus: Hard Times und Adam Bede," in Germanisch-Romanische Monatsschrift n.s. 48: 435-79. 2002a "Gesichter in der Erzählkunst: Zur Wahrnehmung von Physiognomien und Metawahrnehmung von Physiognomiebeschreibungen aus theoretischer und his-
210
Werner Wolf
torischer Sicht am Beispiel englischsprachiger Texte des 19. und 20. Jahrhunderts," in Sprachkunst 33: 301-25. 2002b "Das Problem der Narrativität in Literatur, bildender Kunst und Musik: ein Beitrag zu einer intermedialen Erzähltheorie," in Erzähltheorie transgenerisch, intermedial, interdisziplinär, edited by Ansgar and Vera Nünning. WVTHandbiicher zum literaturwissenschaftlichen Studium 5: 23-104 (Trier: WVT). 2002c '"Speaking faces'?—Zur epistemologischen Lesbarkeit von PhysiognomieBeschreibungen im englischen Erzählen des Modemismus," in Poetica 34: 389— 426. 2004 '"Cross the Border—Close that Gap': Towards an Intermedial Narratology," in European Journal of English Studies (EJES) 8.1: 81 - 1 0 3 . Woods, Tim 1995 " T h e Music of Chance: Aleatorical DisHarmonies Within the 'City of the World'," in Beyond the Red Notebook: Essays on Paul Auster, edited by Dennis Barone. Penn Studies in Contemporary American Fiction, 143-61 (Philadelphia: University of Pennsylvania Press).
BEATRIZ PENAS IBÄNEZ (Zaragoza)
A Pragma-stylistic Contribution to the Study of Narrativity: Standard and Non-standard Narrativities 1. D i s c u s s i n g Narrativity: Introductory Considerations on its F o r m a l and Intertextual D i m e n s i o n s The present chapter integrates a "natural language text"-based concept of narrativity and a "natural language use(r)"-driven concept of narrativity. It is aware of the consequences of assuming a perspective on narrativity either as a product-property or as a process evolving out of use. 1 It is also aware that an integrated perspective results in a certain displacement of the traditional European semiology of narrative (narratology) to make room for a more basically sociopragmatic approach to narrative semiosis. 2 Traditionally, the consideration of narrativity has resulted in the inseparability of two aspects—narrativity and a certain standard of textuality—synthesized within "the" narrative phenomenon and actualized in "the appropriately" narrative text. 3 The present discussion grows out of an awareness of the historical and theoretical antecedence of the standard structural text-based approaches to narrativity over the sociopragmatic-
2
This is not at all the same as Prince's distinction between "narrative as entity and narrative as quality" (Prince 2003: 1) following his discussion of Genette (1976 [1966]). Prince's (or Genette's) favoring of the philosophical terminology 'entity' over 'object or product' and 'quality' over 'effect' puts a clear focus on the ontological rather than the epistemic status of the narrative text, and on narrativity as an intrinsically immanent differential quality. Roger Fowler (1981) set the agenda for a new sociolinguistic theory of literary discourse that encourages more descriptive, empirical research in poetics while it exposes the poetic language fallacy and the objective fallacy predominant in structural poetics. Fowler can do so on the basis of his substituting a communication model of discourse for the formalist-structuralist one. His agenda for revising the theoretical tenets of structuralist poetics in general can help classical narratology to redesign its own agenda for a more comprehensive study of narrative discourse.
3
Cf. Sturgess (1992).
1
212
Beatriz Penas Ibänez
semiotic ones. The former approaches have been more reliant on bipolar accounts of the standard narrative text and its grammatically and semantically structured form-meaning complexities, while the latter have come to rely more and more on conceptualizing narrativity as semiosis, in other words, as a process of narrative meaning-making and interpreting, "irreducible to grammar," 4 and reliant on various discursive and stylistic possibilities through which narrativizing takes place for specific linguisticliterary communities. 5 Seeing natural language narrativity as an ongoing user-interconnecting communicative process allows us to disengage our perspective from the well-formedness of a recorded/reported narrative and to focus on the signspecific (visual, verbal or other) constraints that necessarily affect the narrativity of the semiotic artifact beyond the genre-specific constraints traditionally perceived by narratology as connatural to all narrativity, rather than as dependent on its production within a specific signic medium. It also opens the way to re-evaluating the relative (in)dependence (arbitrariness) of interpretation—the meaning-product derived from reading a text as narrative—with respect to form (the unappropriated reproducible text) and to the context of narrative in use. The unrecorded/unreported ephemeral narratives of memory and dream seem to be apt objects of a sociopragmatic study on narrativity. The fact that their elusive narrativities can be and have been transferred to other substantial media—think for instance of literary (or non-literary) written and spoken narrative texts, of filmic and musical narratives, or of the plastic narratives we find in sculpted friezes or in paintings—must perforce mean something, especially to cognitivists. As I see it, the issue of minimal narrativity and its translatability into other media (transmediality, intermediality) should be connected to perspective, point of view 6 and 4
Cf. Penas Ibäfiez (1998: 8). Sell's (2004) critique of structuralist poetics runs on a par with Fowler (1981). However, I take exception to Sell's mistaking of semiotics for semiology. American semiotics and European semiology do not share the same methods and goals, and the modeling role o f language on other signifying systems characteristic of Saussurean semiology is not part of Peircean semiotics. But I thoroughly agree with Sell's underlining of the sociopolitical and identitarian issues connected to literary and non-literary uses of language, which structuralist poetics was blind to. This is a controversial issue. Wolf Schmid, for instance, believes that "the category of point of view, or perspective, should [not] be included in the definition of narrativity" on the grounds that "implicit perspective is not unique to narration but is really a property of all modes o f representation" (2003: 20). It is unclear, however, why Schmid ex-
A Pragma-stylistic Contribution to the Study o f Narrativity
213
focalization. In sociopragmatics, the centrality of the notion of narrative (counter)voice / (counter)gaze / counterpoint cannot be overestimated. The interpretational ground for "counter" perspectives is social-discursive, ideological and cultural, not essentially formal. In the interpretation of narratives, formal syntax and semantics constitute simply one descriptive level. In addition to the told, the untold (but meant) must also be accounted for. This can be done in terms of more socio-pragmatic accounts of interpreted meaning that incorporate an understanding of narrative implicatures and explicatures 7 as well as of the sociosymbolic value 8 of linguistic-stylistic variation. The "counter" understanding is so essential for characterization-identity, narratorial stance and the reader's understanding of the narrated world that it often escapes attention and adequate explanation in narratological studies. Positionality is the precondition for the expressing and understanding of implicatures, explicatures and social value. It is in this sense that perspective and point of view are "natural" to narrativity. Because of its multidimensionality of meaning, narrative semiosis calls for a narratological explanation on the basis not only of natural language semantics, syntax and grammar (formal textuality), but also on that of pragmatics, sociolinguistics, stylistics, intertextuality and intermediality. As texts whose interpretation relies on the orchestration of situational constraints (register) and individual style as well as on conformity to sociocultural patterns of meaning-understanding within the scope of a natural language, literary narratives actualize their diverse narrativities along intertextual/intermedial paths that can be more or less standard or deviational, and their departure from the standard norm might show up at any point in their multidimensionality. In the following sections of this chapter, I shall be looking into the connection between my theoretical position and the narrative practice of two twentieth-century masters, Ernest Hemingway and Vladimir Nabokov. Both Hemingway's "iceberg" texts and Nabokov's narratives of displacement are non-standard types of narrative. They explore different eludes both explicit and implicit perspective from the definition o f narrativity. He offers an explanation for this exclusion which is partial and self-contradictory, considering only implicit perspective to be universal, and stating three pages later that narratorial perspective is the defining feature of narrativity in the strict sense of the term (2003: 7 8
23). Cf. Grice (1957), (1975), (1981). Cf. Fishman(1970).
214
Beatriz Penas Ibäflez
ways in which narrativity and intertextuality interact in order to create a branching text whose virtual space is a multidimensional network of possible routes. This is the virtual space that their readers must navigate as they move through the text. But Hemingway's and Nabokov's (post)modernist agendas differ (1) with respect to the kind of narrativity each of the writers exploits and to the degree of complexity of their textual and fictional constructions, and (2) in the demands they make on their readers—the kind and degree of inferential work their readings must assume in order to become suitably complex interpretations. Both authors start from the premise that they need not and would rather not tell their readers explicitly all that they are to understand when reading. Readers are invited to participate actively and to make the necessary interpretative effort. They are led into a process of narrative co-production whose complexity matches the composition process. Hemingway opts for an "iceberg" kind of text whose tip—the told narrative—signals the existence of a hidden narrative, one waiting to be retrieved by readers whose experience will allow them to read between the lines or, more specifically, to read intertextually. Because the proportionality between surface-tobottom narrativity is deviational in Hemingway's "iceberg" narrative text, reading must balance its textual and intertextual weight. The intertexts often fall within Hemingway's own ceuvre, since his other narratives give plenty of clues as to what is to be understood in the "pregnant" silences lurking under the surface of the text. These intertextual loops do not merely connect (segments of) textual forms; they also associate different moments of time: the present reading of the text and remembered readings of earlier source texts. As a consequence, narrativity ceases to be intrinsically contained in and associated with one particular structural form and reading time. Rather, it bears witness to the continuity of particular human experience, associating serial experiences of reading a narrative here and now with another reading there and then. Intertextuality expands the notion of narrativity beyond intrinsically formal considerations. Since intertextuality is a prominent feature of narrative writing that involves form, participants' effort, time and space, the definition of narrative needs to be expanded into a type of meaning that involves all these factors. And all these factors can provide a measure of narrativity. Regarding textual space, Hemingway's non-standard way of writing narrative plays vertically, relying basically on disproportion between surface-to-bottom narrativity. The surface narrative hides a more significant one that has not been narrated, but that can be read between the lines.
A Pragma-stylistic Contribution to the Study of Narrativity
215
As to Vladimir Nabokov's preferences in writing narrative, his experience as a creative writer of novels, short stories and autobiography in two languages, Russian and English, and as his own translator, gave him a privileged access to and a deep understanding of the issue of narrativity, both standard and non-standard. His deviational narrativity stems from different sources than Hemingway's. Nabokov's non-standard way of writing narrative plays horizontally, relying basically on disproportion between center-to-margins narrativity. In other words, Nabokov's narrativity accentuates not omission, but displacement. It does not silence significant narrative matter but rather masks it by displacing the most significant tale from the central textual space to the margins. Retrieval of the real matter requires inferential intertextual work on the reader's part, while in Nabokov's case, inference occurs through intermediality as much as it does through intertextuality, as the verbal and the visual narratives conjured in the reading experience are to be synesthetically associated in a complex "branching text," a multidimensional network of possible reading routes. This is a non-standard kind of narrativity as compared to that of traditional narratives and novels: "in non-branching texts, such as traditional novels, [the virtual space navigated by the readers, as they read the text,] collapses into a line". 9 My study of the various ways in which two major exponents of twentieth-century literary narrative cultivate non-standard forms does not contradict the fact that these non-standard narrativities have succeeded in becoming canonical and therefore standard within the literary realm. The following step would be to test how and when this change in the understanding and production of narrativity affects non-literary realms and critical narratological discourse and practice—another point of interest for the present study that will be taken up in the conclusions. In a way, the movement of the present discussion as it applies to natural language (literary) narrativity duplicates the movement of the twentieth-century history of the discipline from formalist communication models based on the conduit metaphor 10 to a sociopragmatic or Bakhtinian 9
Ryan (2003: 336). Communication goes beyond the mere passing of a message between an active sender and a passive receiver. Reddy (1979) has discussed this pernicious assumption under the rubric "the conduit metaphor" and offered an alternative metaphor: the "toolmaker's paradigm." The "conduit metaphor" communicative model is based on a static conception of what goes on between communicators and may have been considered useful at a time, but it is oversimplifying and, even more, blind if not blinding to the fact that
216
Beatriz Penas Ibdfiez
dialogical model of communication based on "translinguistic" sociopragmatic categories such as narrative voice shifting, dialogism, polyphony, and intertextuality." It is also a movement away from Saussurean semiology to post-Peircean semiotics, that is, from the dichotomous treatment of the "what" and the "how" of (narrative) meaning, traceable in extrapolations such as signifiant(-sjuzhef) versus signifii(-fabula) or in langue(motifeme) versus parole(-motif),n to the triadic modeling of interpreted meaning and the sign in which signifier and signified are inseparable from a culturally contextualized interpretant, and a "who" that is declined in a series of questions—questions about identity such as "Whose voice(s) do we hear in the narrative?"; interactional questions such as "By whom and to whom are those voices being addressed?"; environmental questions such as "Where and when is the voice heard?"; intentional questions such as "Why, what for, for whom is a voice speaking?" And so on and so forth. Shunning the excessive emphasis on the textual immanence of narrativity dominant in the theories that sustain traditional narratological approaches need not be seen as a rejection or dismissal of the concepts and tools we have inherited from formalism.13 Rather, it is a strategic move
11
speech is dialogic in Bakhtin's broad sense of the term. A meaningful stretch of speech is not the product of a single consciousness, but the interaction of many voices brought into focus by an individual's performance. Speech is a "blueprint" (Reddy [1979]) from which much may be inferred "about what the utterer believes about what the addressee believes, and about what effect the utterer intends the utterance to have" (Green [1989: 11]). By denouncing the so-called "objective fallacy," Fowler (1981) backs Reddy's position against the conduit metaphor view of language. The expression "objective fallacy" is not Fowler's own coinage, but a borrowing from Miner (1976). Cf. Penas IMflez (1996). I find John Pier's (2003) discussion of the two major dimensions of the structure of narrative, fabula/sjuzhet, and their narratological and semiotic reformulations illuminating. From the formalist structural model, Pier proceeds towards Eco's Peircean semiotic model in which the role of the reader cannot coincide with a static model of communication. Kindt and Müller (2003: 214-15) call for future developments of the theory of narrative and the ensuing concepts of narrativity, which should meet two criteria for adequacy: the criterion of continuity with classical formal narrative categories and the criterion of neutrality as regards the theory of meaning entertained by the critic. As regards the present discussion, we could say that there is no neutral theory of meaning, but that the more inclusive pragmalinguistic theories of meaning (Grice, Gumperz, Levinson) act as if neutral in relation to the earlier structural models, since pragmalinguistics integrates structural semantics into an extended sociopragmatic theory of meaning and interpretation.
A Pragma-stylistic Contribution to the Study o f Narrativity
217
forward that allows the critic of narrative discourse to ground both the interpretation of narrative meaning and the interpretation of a given text as being a narrative (its narrativity) on textual (formalizable) 14 as well as contextual (cultural and situational, therefore inferential) factors—factors that reach beyond the literalness of the text, though they may be more or less implicitly inscribed in it, and thus more or less accessible to a competent reading. In pragmatic accounts of the understanding of utterances and texts, there is no contradiction in saying that a minimal text like "Can you pass me the salt?" is a question from the standpoint of sentence meaning, but that from the standpoint of utterance, it is not a question at all, but a request. In a way, this is the same movement from linguistic meaning to intended meaning that occurs in tropes. 15 Moving away from an exclusively formal reading of narrativity to a sociopragmatic one will enable us to separate the sentence-meaning and grammar of the narrative text and consider its actual pragmatic meaning as intended by the language user, e.g. a satire, a parable or a parody. By "language user" is understood not only historical language users, i.e. authors and readers/audiences, but also their discursive replicas: implied author and reader as well as their diegetic and fictional persons, both narratorial and characteriological. On this basis, it is easier to see that what is meant (or understood to be meant) by a character need not coincide with what is meant or understood by the narrator; in the same way, narrative meaning need not coincide with what is either the author's or the reader's interpretation. A whole system of ironic unreliabilities can spring from this polyphony of language users and their capacity for narrative interplay. 14
15
"Narrative discourse" is an ambiguous expression. It might mean a form o f language, that is to say a variety or register with specifiable formal characteristics. Or it might mean a purpose of communication, a set of diverse conventions agreed on among authors and readers within diverse (more or less different) cultural or literary systems. The latter sense is less ethnocentric, while the former sense is also valid when the critic is thinking in terms only of one subset of a traditional set o f conventions—"the Western cultivated narrative canon," let's say. The meaning of tropes and user's meanings/intentions are two very different things in rhetoric, semantics and pragmatics, but the cognitive dynamics leading to the content interpretation of brand-new tropes as used by the writer or language user parallels the process leading to the interpretation of user's meaning/intentions. John Hollander (quoted in Bloom et al. [1979: 10]), a leading authority on lyrical form, calls tropes "turns that occur between the meanings of intention and the significances o f linguistic utterances."
218
Beatriz Penas Ibanez
From the point of view of one of these language users, the reader, his competent reading of a narrative, his having a good grasp of its narrativity, depends on his competence for trope-conscious multi-leveled reading, a heightened literacy by which to complement the ad litteram one. The processing of narrativity includes understanding pragmatic factors cooccurring with formal ones: for instance, narrative tropicality, i.e. that a narrative (from the writer's perspective) may stand for something else than a narrative (to readers), or that something else than a narrative (from the writer's perspective) may stand for a narrative (to readers); or again, the perspectival and social-hierarchical interplay of the voices articulated within and without the text as we have considered them above. The processing of the various ways of narrativity therefore depends on integrating tropicality (the external polyphony of real-world meaning and intention at the writing and reading ends) and the internal polyphonic interplay of voices within the narrated world (their intended meanings/intentions). This integration results in a critical-intertextual reading that allows placing a text's narrativity within the gradient from the fully standard to the non-standard (the non-standard ranging from the stylishly deviant to the substandard kinds), both in reference to the text-external and the textinternal ways in which the voices peopling a narrative and the voices and agencies in the social world interact to build a meaningful text and discourse on narrative which operates within a complex communicative context of situations and cultures. Understanding narrative means differentiating among and integrating its various meanings as intended by each and all of these different voices, each voice the source and origin of a particular subjective experience of narrative and its articulation both in a particular real world and in a possible one. 16 Considering narrativity to be not only a textual trigger, but also part of a social semiotics of action and viewing it as a result of the user's intertextual competence 17 aimed at integrating a hierarchy of voices and their 16
Or, as Umberto Eco (1978: 76) says: "The comprehension o f a text—which constitutes not only a semantic problem but also a pragmatic one—is strictly dependent upon a series of cultural codes that constitute a semiotic typology of cultures." For my views on the intertextual dimension of discourse, see Penas Ibäfiez (1996). The coinage o f the term "intertextuality" is traceable to Kristeva (1969 [1966]), although it derives from Bakhtin's concepts of dialogism and heteroglossia. To quote Bakhtin: "The dialogical [i.e. intertextual] orientation is obviously a characteristic of all discourse. It is the natural aim of all living discourse. Discourse comes upon the discourse of the other on all the roads [genres] that lead to its object, and it cannot but enter into intense and lively interaction with it" (Bakhtin [1981: 92]).
A Pragma-stylistic Contribution to the Study of Narrativity
219
corresponding worlds, more or less explicitly represented by textual wording in a variety of ways open to narrative, would result in greater conceptual clarity. The standard kind of narrative (codified as a prototypical textual form and its corresponding content structure) need not be more than that: a codified standard variety of narrativity. Classical narratology 18 has served to perfection the cultural function of codifying the agreed or default norms of usage of narrative for a wide range of Western institutional purposes, among which literary purposes are to be highlighted due to the weight and importance of the studies focused on literary narrative. Narratives—both standard and non-standard, including the substandard—can be viewed as proceeding along a continuum or range of conventionality, not of narrativity. They are all equally narrative, but within a gradient of narrative conventionality and purpose: they are not all equally well formed narratives in relation to the agreed norms of narrative (re)presentation contemplated at a particular time in the Western tradition. A tradition contingent on a succession of temporarily valid parameters measuring narrative well-formedness and deviation from it (as a result sometimes of artistic troping, and other times of illiteracy). They are contingent parameters which have been used to fix and implement an elaborated form of narrative, especially by classical narratology, as essential to all narrativity.
2. Non-standard Narrativities: Deviational Surface-to-bottom Narrativity. The Hemingway "Iceberg" Narrative Text As regards their breaking of the expected semantic link between form and content, parables, allegories and exempla are non-standard tropical forms of narrative. They play on a mismatch between user's meaning and text meaning: the text's meaning, or primary literal narrative, is displaced and resignified, standing for a more meaningful narrative than the primary one. Harold Bloom (1979) calls this phenomenon, which consists of the breaking of the expected meaning of a form or of the expected form of a meaning, "tropicality" with reference to tropes like metaphor, metonymy and irony. For the present discussion, we shall consider calling this kind of phenomenon "tropical narrativity" and see it as one distinctive variety 18
I follow here Ansgar Nünning's (2003) suggestions about the most adequate usages of terminology related to the different approaches to narratology and narratological studies.
220
Beatriz Penas Ibäflez
of non-standard narrativity produced by a deviational use of content in relation to the standard semantics of form in Western narrative. I have worked on the semantics of Hemingway's ironic narrative elsewhere at full length. 19 For present purposes, however, I will focus on how narrativity can stop being standard to become tropical (ironic and parabolic), so that the narrative as a whole is designed to say one (acceptable) thing and yet manage to convey, silently, another (perhaps less easily acceptable) truth. The micropragmatic complexities of the Hemingway text arise from the author's masterly use of symbolism, which can be described as the play of the said or narrated versus the unsaid but inferable from the text. For instance, in "The Snows of Kilimanjaro" (1936) Hemingway catalyses the symbol of snow and whiteness so as to raise a wide range of interpretative possibilities. Symbols are both cultural and textual in nature. "The Snows of Kilimanjaro" uses the symbol of snow and the mythic associations that Western culture and, more specifically, the Christian literary tradition, attaches to the idea of 'snow' as 'whiteness' and of whiteness as 'purity' and 'candid truth'. But the text of "The Snows of Kilimanjaro" also supports the opposite view, namely that 'whiteness' may also be associated with deceit, as in the 'white lies' that Harry tells his wife so as to spare her the painful truth. 20 The narration unveils deep-running tensions in the writer's relationship with his rich wife, for there are conflicts that remain for the most part repressed in their usually polite conversational exchanges. Nevertheless, the root of conflict lies in and is strengthened by those very polite omissions that seem to safeguard their life together. For instance, their childlessness as a couple is never a conversational topic raised between them, underscored by the fact that the narration makes it known that Helen has children by a former husband of whom she dreams at night. It is only in the tension that builds between them after Harry falls ill and develops gangrene that conflict threatens to break out openly. Harry is barren, both as a childless man and as an unproductive writer. He also fails at attempting to cultivate fertile communication with the 19
See Penas Ibäflez (2003), (2004a). '"I'm [Harry] crazy as a coot and being as cruel to you as I can be. Don't pay any attention, darling to what I say. I love you, really. You know I love you. I've never loved anyone else the way I love you.' "He slipped into the familiar lie he made his bread and butter by." (Hemingway 1991 [1936]: 43; hereafter cited in the text as "Snows")
A Pragma-stylistic Contribution to the Study o f Narrativity
221
black African men that earn their living at the Safari camp. He can only manage a master/servant exchange based on daily routine and stereotyped verbal formulas of the type "Yes, Bwana." The only seemingly plausible relationship of male bonding occurs with Old Buddy Compton, but only as part of a dream. Harry's social and creative staleness as well as his biological barrenness construct his identity as the modern white American male writer, a would-be, a kind of nobody. His lack of authenticity stems from his excessive "whiteness": a failed man of failed character, Harry suffers from the blankness of the writer's imagination whenever he attempts to speak of America and Americans in terms of pure "whiteness." A purely white America has never existed, an America without redness and blackness being nothing but a child of the imagination. Analogously, imagining a purely white American literary tradition is deluding and belittling, for America has become America thanks to the fertilizing effects of miscegenation. Pollution entails fertilization, and miscegenation the strengthening of old genetic material. Contamination becomes blood purification. The destruction of mythical cultural purity has caused the assimilation of new knowledge and general growth in both the social and the literary realms. White traditions underlie the positive symbolic values of light, clarity and candid whiteness as the precondition of visibility, transparency, and truth. A standard reading of the story may stop at this point, but a resisting reader may interpret "Snows" differently and question interpretations that fail to take account of the full breadth of these symbolic values, bringing out the negative rather than the positive symbolic values of "whiteness" and the positive values of "darkness" in the story. These inversions of the expected meanings that subvert cultural stereotypes are made possible thanks to Hemingway's strategic deployment of verbal and structural irony. Irony allows for divergent but complementary interpretations of a narrative, although it is the reader who is challenged to pragmatically accommodate the textual ironies and the reversals of culturally expected meanings. This is inferential work and a challenge to Hemingway's reader, who has to navigate among the ironies created by the said and the unsaid in relation to a recurring sign and symbol: snow. The main function of this kind of ironic non-closure is that the text manages to expose the politics of exclusion of certain linguistic and cultural dualities—black vs. white, good vs. evil, even truth vs. the illusion of truth—as ideological
222
Beatriz Penas Ibäfiez
and therefore non-universal but relative to the dominant (and in this case "white") system of values. Thus, Hemingway the artist, in his creative and also deviant use of narrativity, manages to destabilize the idees regues conventionally attached to dual categorizations that contribute so much to stigmatizing or demonizing one of the polar opposites with the consequent perpetuation of patterns of violence21 and domination prevalent in complex contemporary societies. In "Snows," Hemingway exposes the constructedness of narrative textand word-meaning as well as the social constructedness of the positive symbolic values of "snow" (mainly whiteness, innocence, goodness, and truth) attached to it in Western cultures. In "Plato's Pharmacy" (1981 [1972]), Derrida reflects on the dual capacity of the Greek word pharmakon to mean either 'medicine' or 'poison' so as to underline the ironic potential of the linguistic sign and its essentially plastic nature. The same potential can be said to dwell in Hemingway's title, "The Snows of Kilimanjaro." According to Jane Miller, "Even a straightforward title invites conjecture about its possible use beyond a conventional, obvious meaning."22 At first sight, "The Snows of Kilimanjaro" is one of those straightforward titles whose apparent simplicity is designed to trigger the reader's curiosity about the relationship between title and ensuing text. On an analogous line of thought, Robert Wallace writes about the key role of titles and their various textual functions: titles may indicate a text's subject or theme, set the scene, address a person or object, highlight a crucial image or present necessary information withheld from the body of the text.23 "The Snows of Kilimanjaro" as title both does and does not serve all these functions. It includes a placename which at first sight seems to designate the geographical location of the story, although this proves not to be the case, as the action in fact takes place in the Safari camp on the plains, far from the mountains and far from Kilimanjaro. In other words, the title really only names an absence. The mountain is only an object of desire, and the snows of Kilimanjaro are never to be reached by the characters that people the fiction. It is only in his agony that Harry hallucinates that he and his buddy, Compton, are flying together to see the snow-white top of Kilimanjaro. 21
22 23
On the issue o f patterns of violence in Hemingway's worlds o f fact and fiction, see Penas Ibäfiez (1994). Miller (1992: 55). Wallace (1991: 335).
A Pragma-stylistic Contribution to the Study of Narrativity
223
What the fictional voice in this passage tells the reader is not to be understood as really occurring; on the contrary, the typography in italics is designed to signal to the reader not to believe literally what (s)he is reading and not to set this part of the action on the same ontological level as the rest. Thus, the white snows of Kilimanjaro are only talked about, a desire forming part of Harry's agonic dream. This removes Kilimanjaro two steps from the factual: a mere dream within the fiction, it is doubly fictional. The snows of Kilimanjaro conjured by the title are progressively "de-realized" in the reading of the story as Kilimanjaro fails to become a place where things happen. In letting the title carry an unfulfilled promise, Hemingway plays with the reference of the predicates in the title's noun phrase, thus telling a kind of joking lie to the reader. By emptying certain nouns of their immediate referents, he creates blanks, a blank name for a blank place: all-white Kilimanjaro can be talked about, it is an alternative narrative, yet it is not strictly part of the primary one. From a possibleworlds perspective, the word Kilimanjaro does refer, but it refers to an alternative possible world (dreamed for, desired) distinct from the actual text world represented in the main narrative. The symbolism of Mount Kilimanjaro has been dealt with extensively, usually in terms of the opposition mountain/plain. 24 Along a different interpretive path, I find Robert W. Lewis's treatment of the subject especially illuminating. He records that the western summit of the twin-peaked Kilimanjaro is called "Kibo," meaning "the bright," as opposed to the eastern peak, "Mawenzi," meaning "the dark." 25 This information is highly germane to the present analysis in that it discloses that Kilimanjaro does not exist as all of one piece. Kilimanjaro is not only snowy or white because it is not one, but two—Kibo and Mawenzi—and snow is only found on Kibo. 26 Kibo's whiteness overpowers Mawenzi's darkness and 24
25
For a discussion of the symbolism of the mountain versus the plain, we are especially indebted to Carlos Baker (1972, chapter VIII). Baker calls attention to the pervasive presence of the mountain/plain contrast in Green Hills of Africa (1935) as well as in "The Snows of Kilimanjaro." He sees in Hemingway a tendency that is already present in "Fiesta" (1926) and in A Farewell to Arms (1929) to associate mountains with the idea of Home. He also refers to C. C. Walcutt (1949), for whom the mountain is the symbol for Truth, and to E. W. Tedlock (1949), for whom it symbolizes integrity. Lewis (1966: 78). Curiously enough, the name Kibo recurs in the African story embedded in The Garden of Eden. Written about ten years after "Snows" (Reynolds 2000: 43), but published only in 1986, The Garden of Eden also introduces the play of black darkness and blond brightness in the Bournes's frequent hair bleaching which accompany their small mu-
224
Beatriz Penas Ibaflez
thus deceives onlookers, for whom Kilimanjaro's darkness remains invisible. Consequently, Kilimanjaro names the colored complexity of the whole mountain, but when the adjective "white" is used in the title of the story only in association with the mountain's snows, it becomes a kind of "white lie," or at least an oversimplification for a more complex reality. 27 Analogously, the name "America" can be used in such a way as to name and make visible only the white side of what is in fact a more complex reality. This exclusion, like many others in Hemingway's work, is managed by manipulating the narrative and exploiting its metonymic potential. Hemingway, more than other fiction writers, uses the part to represent the whole, with the result that the textual economy of his works operates through understatement, partial representation, subtle use of focalization, rigorous selection of detail, and strategic omission of topic. He was therefore highly sensitive to discursive manipulations of the type mentioned above, which use the whole to represent the part rather than vice versa. From this emerges the idea that strict understandings of dualities or dichotomies result in false either/or conceptualizations of what is a nuanced world. From Hemingway's complex perspective, there is both a bright and a dark side to all human knowledge and experience of the world, and the narratives we tell ourselves about the world we live in should not falsify the fact. Dissimulating the dark side only falsifies human experience. But art is different; its narratives are licensed to play with dissimulation in order to convey truth ironically, as underscored by the title, "The Snows of Kilimanjaro," from which darkness is absent. After the title, the symbol of snow is used again at other strategic points in the story. In the epigraph, the word again resounds ironically, but this time in association with a different kind of whiteness: the whiteness of the shroud, the whiteness of death. The epigraph reads: tual treasons. The Garden of Eden can be considered postmodern, which is not the case with "Snows," but they share a preoccupation with literary truth that provides them with thematic unity and places them, as well as the other narratives written in the 1930s, within an aesthetic continuum leading Hemingway's narrative writing toward postmodern metafictionality. It should be noted here that Hemingway's first title for the story, "The Happy Ending," which was discarded only at the very last moment before publication (Smith 1989: 3 5 1 - 5 2 ) , was equivalent to the late title "The Snows of Kilimanjaro" in its onesidedness. "The Happy Ending" as a title only referred to one of the two endings of the story, and in doing so was as tricky a little lie on Hemingway's part as the title finally selected. This fact cannot be dismissed as insignificant. It illustrates the semantic complexity o f surface-to-bottom narrativity in Hemingway's works.
A Pragma-stylistic Contribution to the Study of Narrativity
225
Kilimanjaro is a snow-covered mountain 19,710 feet high, and is said to be the highest mountain in Africa. Its western summit is called the Masai 'Ngaje Ngni' the House of God. Close to the western summit there is the dried and frozen carcass of a leogard. No one has explained what the leopard was seeking at that altitude. ("Snows" 39)
While the word "snows" in the title may lead the reader to construct an illusion of reference and truth that is not sustained by the overall narrative, "snow," now in the epigraph, is mentioned in connection with the dead frozen leopard. The epigraph thus associates snow and whiteness with death, dryness and rigidity, as represented by the frozen dead leopard. It also links death and the dangers of entertaining too lofty ideals, illusions, or rather delusions represented by the leopard's implausible quest and final doom at that altitude of the snow-covered white western summit of Kilimanjaro. Truth is insistently thematized and questioned in the story, a story whose male protagonist, Harry, is portrayed as a failed writer, somebody who could have been a good writer, but who one day married money and stopped working. The reader is made to see Harry as a pretender, somebody who thinks he is still able to write and make a living for himself when the truth is that, after years of wasting time, he has blunted his talent. But he can afford to go to Africa on a safari. We can say then that "Snows" is about faking, false illusion, delusion, pretence, falsity, deceit, and lying—all of them different ways and degrees of distorting the truth. "Snows" develops as its central motif the dialectical movement linking truth to the different ways of evading it. The development of this dialectics is achieved directly by means of Harry's characterization as a liar and indirectly by means of Harry's characterization as a sick man. Here, Hemingway introduces a second symbol: gangrene. Gangrene operates as an anesthetic. The person who suffers from gangrene stops feeling pain in the affected part and eventually dies. This initial effect of gangrene is seemingly positive: after all, who wants to feel pain? But pain in its various degrees also has a purpose, for it is a biological warning of danger to health and even of the threat of death. Harry's deceptive lack of pain, so characteristic of gangrene, is the terminal lie that Harry tells himself. A lie that literally kills him much as his earlier lies had already killed his capacity for art and a fulfilling life.
28
On the textual sources of the epigraph containing the story of the leopard that was found frozen midway up the summit of Kilimanjaro, see Lewis (1966).
226
Beatriz Penas Ibäflez
From this perspective, the story becomes a parable whose teaching is that a lucid mind does not let itself be deceived by false appearances. In other words, clear thinking is characterized by a methodic search for essentials and a disregard of the accessory. Lucidity in spotting the root of problems may cause pain when it calls for taking painful measures, e.g. cutting off a gangrened limb. Conversely, twisted thinking may tempt us to elude painful truths, it may dull perceptual sensitivity and moral sensibility, or it may prevent us from gaining full awareness of the issues at stake in our lives. Awareness is the artist's essential condition before he can start to write his narratives. Ironically, in Hemingway's story awareness comes too late in Harry's case and it does not last. He dies unawares. Harry is a victim of two types of malady, one physical, the other moral. 29 He has been suffering from moral corruption for years and now suffers from physical rot affecting one of his legs. The first malady, whose symptoms were laziness and a tendency to deceive himself and others, caused the loss of Harry's talent for story-telling, and now gangrene is causing Harry to lose his life. In both cases, the cause of Harry's destruction is Harry himself, for his fear of truth blunts his response to clues that should have warned him against faking in his relationships, his narrative craft and his very life. The idea of self-destruction caused by self-deception is symbolically confirmed by a hyena who, its shadow crossing the African plain every night, makes "a strange, human almost crying sound" ("Snows" 56), a sound which Helen can hear the night Harry dies. In popular belief, the hyena smile/laughter, like crocodile tears, has come to represent hypocrisy or false behavior on the basis of an obvious lack of adequacy in the act of laughing and the occasion, which here is marked by the vicinity of Harry's death. Besides the hypocrisy associated with the hyena, there are other less immediate meanings that emerge not so much from the text itself as from intertextual contrast. For instance, in Green Hills of Africa (1935), the hyena is characterized as "self-eating devourer of the dead." 30 Green Hills represents the hyena as a negative ouroboros, a circular creature who feeds on the dead and finally on itself, thereby causing his own death. From this perspective, the hyena, false and self-devouring, serves as an emblem of Harry, the self-destructive writer, the hypocrite who, 29
30
Or, as Grebstein (1973: 21) puts it: "Harry's disablement by his diseased leg in 'The S n o w s o f Kilimanjaro' signifies also his corrupt moral condition, that is, the rotting away o f his artistic integrity and physical vitality through abuse and disuse." Hemingway (1966 [1935]: 39).
A Pragma-stylistic Contribution to the Study of Narrativity
227
rather than on truth, feeds on rotten lies, his own invented lies, one who does not know from his own experience what experience is like, and therefore cannot write truly and well.31 Hemingway's narrative evokes contrasting images. The image of whiteness figuratively associated with "snow" contrasts with the image of rot created by the symbols of gangrene and the hyena. It is not coincidental that hyenas eat rot and that 'rot' in American slang means 'nonsense, something not to be believed'. That a lie can sometimes be called "a rotten lie," and also "a white lie," and that both hyenas and liars feed on rot suggests that they end up feeding on themselves. Here, Hemingway seems to be posing a riddle, a cryptic modernist "conceit" for his readers to solve. To find the answer, his readers need to know other Hemingway texts. For instance, 'snow' is also slang for cocaine, a consciousness-altering drug. Within this pharmacological field of meaning, although not explicitly meaning cocaine, we find in A Farewell to Arms (1929) one use of the word 'snow' applied to an anesthetic used in operations during World War I. This is the passage: "[The Doctor] used a local anesthetic called something or other 'snow', which froze the tissue and avoided pain until the probe, the scalpel or the forceps got below the frozen portion." 32 This intertextual citation establishes "freeze" (as before in the case of the frozen leopard) as the missing link—missing from the text, but apparent from the intertexts—connecting the symbols of snow and gangrene with anesthesia. The intertextual net of meanings explains why, in "Snows," Harry's gangrene (anesthetic rot) is functionally equivalent to snow (anesthetic white lies). It can also help to explain why the leopard in the opening epigraph is frozen and trapped in snow to become, like the hyena, an animal emblem for Harry's "white" failed way of being a modern American story writer.33 The functional equivalence between gangrene 31
32
In his writings, Hemingway exemplified the principle of adequacy between life and art that characterizes his theory and practice of true writing: "Good writing is true writing. If a man is making a story up it will be true in proportion to the amount of knowledge of life he has and how conscientious he is, so that when he makes something up it is as it would truly be. If he doesn't know how many people work in their minds and actions his luck may save him for a while, or he may write fantasy. But if he continues to write about what he does not know about he will find himself faking. After he fakes a few times he cannot write honestly any more." (Hemingway [1967: 229]; my emphasis). See also Hemingway (1967: 194-200). Hemingway (1957 [1929]: 98) (my emphasis). The deep functional equivalence of both animals is made apparent by the fact that both hyenas and leopards are famously "spotted," i.e. non-white. Still, they both contribute
228
Beatriz Penas Ibäfiez
(rot) and snow (anesthetic lie) in "Snows" is explainable ("grammatical" in Riffaterre's [1984] sense) only after we take the intertextual detour which opens up to scrutiny the additional meaning of 'snow' as 'anesthetic pharmakon', as reported in an earlier narrative, A Farewell to Arms. The Hemingway "snows"-"iceberg" narrative is designed to carry the bulk of its weight below surface, something he could have done in a variety of ways. In this case, the submerged content—through the reader's recall— links up with the surface of other texts, postulating a reader whose intertextual competence will make this recognition possible. Recognition is not the same as reading, it is rereading "in absentia." When a narrative relies so essentially on intertextual recognition, its meaning becomes, more than textual, fundamentally intertextual. Different narratives open up and intersect to create a complex new narrativity: one-eighth told and easily formalizable, seven-eighths untold and elusive of formal description—an intertextual net of additional superimposed meanings that functions without formally breaking the initial form of each of the intertexts. This is what I have called non-standard narrativity of the tropical kind, meaning that a parable, like a metaphor, is a tropical way of signifying beyond the said and literal. As we shall see below, Nabokov's narrativity also relies on intertextual loopings, although of a different kind. Hemingway's narrative treatment of snow as ironic, the white pharmakon which represents at one and the same time truth and lies, must thus be tested by means of intertextual contrast. Snow—the snows of Kilimanjaro—a symbol associated in a certain context (whiteness in the white world) with the myth of truth and innocence, must also be associated in a different context (whiteness in Africa) with the contrary ideas of rot and deceit. 34 It is perhaps for the plurality of symbolic meanings of snow that Hemingway uses the plural form "snows" in the title. In any case, it is out of the need to communicate widely 35 what he saw as asymmetries of
to a "spotted" representation of Harry's white masculinity, just as the leopard in its frozenness belies life and the hyena in its false smile belies truth. Westbrook (1966: esp. 93), has suggested a possible connection between rot and Kilimanjaro. According to him, there exists a contrast "between the temporal and the permanent, between rot and Kilimanjaro." This interpretation is compatible with the House of God/Eternity associations that Baker acknowledges as more pertinent in his work. See also note 24. Hemingway was aware that his work could scandalize a certain kind of reader. The reception of Death in the Afternoon is adequately explicit evidence of that. His textual
A Pragma-stylistic Contribution to the Study o f Narrativity
229
power between the white world and other, less powerful communities that the author wrote a non-standard narrative about an individual white man, a short story whose narrativity could lead to one or two readings: either a standard reading focused on a chain of circumstances and events happening to a specifically positioned fictional character or a troped reading. In the second type of non-standard reading, the ironies found in literal interpretation are resolved within a coherent interpretation critical of the ideology of domination and of the values prevalent in modern first-world Western societies. 36 It is in response to Τ. E. Hulme's (1924) dissatisfaction with the conventional means and ways of narrating that Hemingway, in Death in the Afternoon (1932), like Pound, in "Small Magazines" (1930), recommends stripping narrative of all but the essentials. However, Hemingway's statement on narrative economy is more radical than Pound's when it comes to the idea that good narratives are like icebergs: "If a writer of prose knows enough about what he is writing about he may omit things that he knows and the reader, if the writer is writing truly enough, will have a feeling of those things as strongly as though the writer had stated them. The dignity of movement of an iceberg is due to only one-eighth of it being above water. A writer who omits things because he does not know them only makes hollow places in his writing." 37 In Death in the Afternoon, Hemingway clarifies what he is trying to achieve in his innovative texts: he wants to rid his narratives of verbosity so that linguistic distortion is minimized and so that the images of his memory and imagination can be recovered through the affinity of experience and knowledge between reader and writer, what Hulme called "Sympathy with reader as brother, as unexpressed author." 38 Hemingway's assumption is that "Nothing could happen to me that had not happened to all men before me." 39 Intersubjectivity, the human bond, allows designs also carry implicit evidence, for they are clear examples o f h o w to manage to say the unspeakable. The "openness" o f Hemingway's non-standard narrative to a plurality o f interpretations or a multiplicity of readings, as I see it, is in agreement with Umberto Eco's ( 1 9 6 2 ) notion of the parallel writing and reading protocols accompanying an "open work." Through several literary examples, for instance Joyce's Ulysses or the Nouveau Roman, Eco explains h o w the open contemporary narrative structures appear to match the wealth o f possibilities inherent in the changed condition o f the present world. 37 38 39
Hemingway (1932: 192). Hulme (1998 [1924]: 52). Quoted in Baker (1961: 39).
230
Beatriz Penas Ibäfiez
empathy, anticipation and participation in the feelings and things that are not explicitly described but which the reader is able to recover inferentially from the pregnant gaps the writer has carefully left in the text. In a way, Hemingway is striving to bring his literature closer to the bullfight, which is a visual-image text and a wordless art. He is asking his readers to collaborate and accept a writing which challenges them to read between the lines, to rely less on the uttered word and more on the images of memory and imagination that experience leaves printed in the mind: "Hence the sudden joy these produce in the reader when he remembers a halfforgotten impression. 'Howtrue!'" 40 From this perspective, Hemingway's "iceberg-text" theory opens a revolutionary perspective on reading and writing. It is a reconsideration of the centrality of words to narrative and literature, an art built on words. I think this is a new perspective that can be summarized in the following three points: 1) The iceberg-text theory radically reframes Pound's imagist preoccupation with the quantitative aspect of the textual economy of an ideally laconic text; Hemingway transforms Pound's imagist theory of poetry into a theory of narrative. 2) It also reframes Hulme's preoccupation with the qualitative aspect of texts whose capacity to narrate a lived experience freshly and capture it linguistically relies on novelty and is set against the traditional generic conventions. 3) The iceberg-text theory reaches radically beyond Pound and Hulme in that it questions the role of language, the necessity of words themselves in the construction of narrative fiction. Hemingway was not a narratologist and certainly never heard the term "narrativity." Even so, his "iceberg" theory is a full theory of narrativity and enjoys specialist status. Not only does Hemingway transform a theory of poetry into a new theory of narrative, but he also, without noticeable cross-generic resistance, adopts his theoretical principles in his narrative practice. This explains the troped character of his way of narrating a story. In his Key West Years during the 1930s, when he wrote and published Death in the Afternoon, Hemingway's re-conceptualization of literary narrativity led him to new non-standard ways of writing—and reading—a type of prose fiction that can be analyzed using my concept of "surface-to40
Hulme (1998 [1924]: 52).
A Pragma-stylistic Contribution to the Study o f Narrativity
231
bottom narrativity" and related to a certain linguistic phobia in Hemingway and his modernist mentors. If the bullfighter tells a story with gestural kinetic signs rather than verbal signs, Hemingway the artist understands that he can do something similar by writing iceberg-texts in which what is not verbalized but suggested or conjured by what is said is more important than what is explicitly told in words. Consequently, much interpretative work is left for the reader to do. But the author's control and economy of narrative design is so strict and his selection of relevant detail so accurate that interpretative randomness is minimized. The interpretation of Hemingway's metonymic narratives may vary from parochial adherence to what is literally told to a more global cultural interpretation which takes into account everything that can be inferred from what little is told, but is clearly hinted at, thanks to a meticulously crafted whole. Under these stringent textual conditions, reading may fail to achieve closure, but will not be likely to run astray. It is in Death in the Afternoon that we find the first written statement of Hemingway's iceberg-text theory of narrative. The iceberg-text comprises both visibility and readability, the invisibility of seven-eighths of the ice submerged under water paralleling the impossibility of actually reading "most of the matter" because it is not linguistically articulated on the surface of the text. The unsaid, the invisible but hinted at, the part omitted from the linguistic surface of the text is a narrative truer to experience than the actually said. In other words, like invisibility, silence does not preclude the existence of what remains unsaid. It is the onlooker and the reader's imagination that is activated by an incomplete linguistic narrative form. The submerged portion of the iceberg cannot be seen, but it exists and can be imagined or inferred. The disproportion between the visible tip of the iceberg and the mass below the water line is what makes Hemingway's narratives non-standard. Following his iceberg metaphor, Death in the Afternoon instructs Hemingway's readers in his writing strategies so that they can trust the existence of a mass of meaning, an alternative narrative, that cannot be located in the visible part of narrative text. It is a narrative without a "standard" narrative text sustaining it. It is a narrative to be, a narrative in search of a reader to actualize it in the process of reading. In other words, Hemingway's iceberg-theory of writing is also a theory of reading, and as a theory of reading it encourages readers to re-create and expand the narrative text through committed interpretation. Hemingway seeks to engage readers in a creative kind of
232
Beatriz Penas Ibanez
reading that matches his own creativity as a writer.41 It is a goal he shares with other modernists, but the means he adopts to work it out are radical, exposing the illusory nature of commonsense notions of what makes a good story, or even a story at all.
3. Non-standard Narrativities: Deviational Center-to-margins Textual Space. Nabokov's Narrativity of Displacement There are other ways of deviating from the standard well-formed narrative. A case in point is Nabokov's prose fiction, fashioned out of "(inter)textual openings"—triggers of narrativity constituting textual moments irrelevant to the overt narrative which, through recurrent intrusion, become relevant and more important than might be expected from their marginality. These textual openings manage to connect within a parallel mental (not verbalized) narrative for those readers able to understand their pertinence. They are pertinent within another narrative, more relevant to the subject discussed, the real topic, which has been displaced from the central or topical position to occupy a marginal textual space. Displacement does not result in omission from the surface text, as is the case with Hemingway's surface-to-bottom narratives. The Nabokovian autobiographical narrative tells all it must tell and does so in full detail: it develops on the textual surface, it is there for the eye to see but, like Poe's purloined letter, it is left in an unexpected place, and that simple fact makes it invisible. The narrative design does not hide the most significant narrative from the text by taking it away or by omission of relevant detail. What it does is mislead readers, sending them off in the "wrong" direction. The reader's interpretative effort is also required, as with Hemingway's iceberg texts, but in a different way. The inferences teased out by the Nabokovian reader need not be creative so as to supplement a given reading with an alternative narrative: rather than add what has not been narrated, they restore the narrated to its place in the main narrative. This is so because certain details necessary for understanding Nabokov's biographical narrative are scattered about, explaining the life narratives of secondary characters, for instance, or appearing in unexpected marginal contexts. Their 41
Decades later narratology would accommodate some of these notions. For Roland Barthes (1977 [1971]) a "writerly text" is the "difficult," because unexpected, text that, in order to be understood at all, requires its reader to take the position o f a writer in recreating it through a daring reading.
A Pragma-stylistic Contribution to the Study of Narrativity
233
apparently digressive character can easily lead readers astray, particularly if their attention wavers and the relevancy of detail is overlooked, but their lack of relevance may arouse the reader's suspicion and initiate a search for their true meaning. Such is Nabokov's narrativity of displacement, as can be found in his autobiography Speak, Memory (1966) and in the rest of his fiction. By narrativity of displacement I mean the tendency of a text to displace experiences and situations lying at the core of its form-content structure (an autobiographical narrative, in the case of Speak, Memory) to the margins. This occurs in such a way as to match another, earlier, displacement of those most significant experiences from the heart of memory to its margins only to find them there as mnemonic residue, recoverable some day within a non-standard verbal narrative text. In developing this highly stylized type of narrative, Nabokov does as other fellow modernists: he deviates from standard ways of narrating. In his 1939 essay on Baudelaire's motifs, Walter Benjamin reviews the different explanations that earlier and contemporary authors (Baudelaire, Proust, Freud) gave of the workings of memory and of various types of memory in relation to past experience and its narrative articulation within a verbal text. Benjamin is interested in explaining the kind of narrative text that can appeal to the moderns, commenting on Proust's distinction between voluntary and involuntary kinds of memory and connecting it to Freud's theory of conscious memory, or memory per se, versus unconscious memory, or reminiscence. According to Benjamin, it is the involuntary (mentally unprocessed, verbally unnarrated) reminiscence that gives its (healing) power to the written narrative text that springs from that unprocessed past experience. Nabokov's written narrative will not seem purposefully aware of where it stems from, what its real topic is, nor will it verbalize this topic. This is not an exception which occurs only in relation to traumatic or shameful events in his autobiography, for it is found in his fiction as well, and with respect to fictional events. An overt narrative articulation of the significant event will not be given, and thus the narrative will be about consciously remembered experience {Erlebnis), i.e. about a less significant experience that has displaced the essential one by occupying the major textual space. A second consequence of this displacement is that Nabokov's readers are required to do extra inferential work on the basis of the small intertextual openings, places of repetition where the writing preserves the trace of the raw experience ( E r f a h r u n g ) and of the situation in which it arose.
234
Beatriz Penas Ibäfiez
These intertextual openings, small and marginal to the text, the overt narration of voluntary memory, provide the reader with a clue to little unrecorded items which are not explicitly narrated. What the conscious narrative recording takes for its topic, an original past lived experience, is not original because it has been subjected to the prior double narrative processing of a proto-narrative that made sense of the lived experience and was converted into voluntary memory, the overt topic of the written narrative. From this modern cognitive stance on narrative processing, modernist thought on narrativity and modernist narrative writing have developed a proclivity for what Gerald Prince calls "antinarrative." 42 In my view, this kind of "antinarrative" narrativity is at the basis of Nabokov's non-standard topic-displacing narratives. In his memoirs, Nabokov sought to imitate the workings of memory. As I have noted elsewhere, 43 the modernity of Nabokov's narrativity also owes much to his early acquaintance with movies, which heightened his awareness of the signic (visual and verbal) narrative quality of remembered experience. The intermedial cinematic quality of Nabokov's written narrative results from the author's interest in the visual image in general, and more especially in its cinematographic-narrative articulations. Nabokov's interest in the visual image and its optical reproductions and syntax also extended to other image-producing devices including the microscope, the telescope, the less well-known stereoscope, the camera obscura or magic lantern, to mention only a few. The images and visualnarrative techniques peculiar to each of these optical devices reveal the intermedial and, more generally, the highly intertextual nature of Nabokov's fiction. A striking example of this can be found in the magic lantern, which provides chapter 8 of Speak, Memory with its serial structure. The chapter is organized as a series of magic-lantern slides portraying the Nabokov brothers' private tutors. Underneath this playful design of the narrative runs an implicit desire to manipulate standard narrativity so that the narrative becomes less cohesive, less consistent and more fragmentary than is generally the case. In other words, the narrative design veers towards 42
43
Prince mentions some distinctions between narrative and non-narrative, or between both and antinarrative. Antinarrative is not in opposition to narrative, either as entity or as quality. In order to distinguish between narrative and antinarrative, Prince uses the example of Robbe-Grillet's La Jalousie (1957) as antinarrative by definition, for it lacks "cohesion, consistency and even closure" (Prince [2003: 6]). Penas Ibaftez (2004b).
A Pragma-stylistic Contribution to the Study of Narrativity
235
Prince's (2003) notion of antinarrative. The quite comical series of portraits of the brothers' tutors is foregrounded against a darker background of anxiety. The fifth and central slide corresponds to Lenski, Nabokov's best-liked private tutor, "a very pure, very decent human being" who "conceived the awful idea of showing, in alternate Sundays, Educational Magic-Lantern Projections at our St. Petersburg home." 44 The first afternoon session was organized so that the Nabokov brothers were accompanied by other members of the family, cousins, friends and schoolmates, but also by the servants' children. Among the latter, only one is explicitly mentioned: the son of his father's valet. The boy's presence is conspicuous because they seat him on Vladimir's right (while on his left they place a girl cousin) and because he is dressed in a sailor suit like the young family members. There are other striking facts about the boy, whom Nabokov recalls as being an absolutely motionless boy in a sailor's suit; he bore a striking resemblance to the tsarevich, and by a still more striking coincidence suffered from the same tragic disease—hemophilia—so several times a year a Court carriage would bring a famous physician to our house and wait and wait in the slow, slanting snow, and if one chose the largest of those grayish flakes and kept one's eye upon it as it came down (past the oriel casement through which one peered), one could make out its rather coarse, irregular shape and also its oscillation in flight, making one feel dull and dizzy, dizzy and dull. (SM 163-64; my emphasis)
These so-called "coincidences" 45 are registered by young Nabokov's eye and more or less consciously remembered. In Speak, Memory the narrative of displacement, written by the adult Nabokov, betrays their significance to him at the time he was an inexperienced boy and artist in the making, and now at the time of craftily writing when the textual stylistic design keeps showing the most significant recurrences in a displaced intermedial manner, as if recurrence were random patternless coincidence and thus meaningless. Nabokov says "during this and the following, still more crowded, still more awful Sunday afternoon sessions, I was haunted by the reverberations of certain family tales I had heard." (SM 165) 44
Nabokov (1989 [1967]: 162); hereafter cited in the text as SM. The idea of textual "coincidence" in Speak, Memory is similar to Baudelaire's notion of "correspondence," as Benjamin explains it. Correspondences are intertextual temporal loops, moments of reminiscence which, as in landmark festivities, make us find the experience of la vie anterieure, un texte antirieur—the past perceived from an inspiring analogy, textual or existential, met at another time. There is no consolation for past experience felt from a present time when having that same experience is impossible (Benjamin 1969 [1939]: 184).
236
Beatriz Penas Ibafiez
Among the family narratives he might have heard, he mentions only one: his rich maternal grandfather used to hunt for the best-looking boys among the best, though poor, scholars "to muster companions for his two sons" (SM 166). This story apparently explains Nabokov's anxiety at the presence o f the handsome Tsarevitch-looking son o f the Polish valet. But I think that Nabokov is consciously "antinarrating" here by "inconsistently" leaving out of his overtly autobiographical narrative the recollection of the most interesting family tale of all. Nabokov's biographers record that Dmitri Nabokov and Maria von Korff, Nabokov's paternal grandparents, married out o f interest rather than love. 46 The couple's sixth child, Nabokov's father, was said to be the offspring o f an adulterous relation between Maria von K o r f f and the Czar. I f this rumor is true, the tsarevitch-face and hemophilia o f the Polish valet's son would be an experiential clue (a repeated inter-generational trait is equivalent to a repeated intertextual trace because each betrays a family resemblance, either between genetically related people or between semantically or formally related texts) confirming a double story o f illegitimate fathers and sons that was known in the family but not openly spoken o f or written about. Nabokov's father would be the natural son of the Czar and the natural father o f his own valet's son. (After all, arrangements such as marrying o f f a pregnant servant to one's valet have not been infrequent in aristocratic families. 47 ) In that case, Nabokov and the Polish valet's son would be half-brothers, both illegitimate to a certain extent. According to the family tale, Nabokov's father would be as illegitimate within the Czar's family as the poor Polish valet's boy was illegitimate within Nabokov's family. As a result, Vladimir Nabokov, the legitimate son of his father, would inherit his father's illegitimacy, a second-degree kind of illegitimacy in this case, that would minimize the dif-
46
Field (1987: 7-9); Boyd (1993). The biographers' second version of the tale says that Maria von K o r f f had Nabokov's father by the Czar's brother, Grand Duke Konstantin, to whom Maria's husband, Dmitri Nabokov, was Personal Aide when the Grand Duke was Vice Regent o f Poland. The Duke would be the natural father of Nabokov's father; therefore, Grandfather Dmitri Nabokov would occupy the Polish Valet's structural position within his grandparents' story, and the position of the handsome Tsarevich-looking boy would correspond to Nabokov's father. This version seems to fit more perfectly the overall familiar reduplicative pattern of infidelity and illegitimate sons legitimated by marriage within a lower class. Maria von K o r f f would be the servant-lover to her and her husband's master. This detail shows the increased angle of displacement that Nabokov presses on the central issues o f his autobiographical narrative.
A Pragma-stylistic Contribution to the Study o f Narrativity
237
ference between him and his "clearly" illegitimate half-brother, the Polish valet's son. As the reminiscing Nabokov puts it: "In tenacious old families certain facial characteristics keep recurring as indicants and maker's marks [...] inherited in various stages of intensity as more or less distinct copies of the same print." (SM53-54) 4 8 The magic-lantern narrative series of Tutor Portraits includes the trace o f a secret portrait. This absent portrait has no substantial textual form, but it is a narrative retraceable both by the darker spot on the center textual space
that
its
displacement
has
left behind
and
by
the
intertex-
tual/intermedial allusions that remain at the margins of the central narrative as indexical clues o f the true narrative. The portrait's displacement from the narrated text leaves us with a ghost narrative, and a ghost portrait, of an illegitimate brother whose existence is the central but unspeakable story told in the chapter and the real but unconfessed source o f anxiety for the boy Nabokov. This associated narrative would explain why, as a boy, he could not endure the pressure posed by the magic-lantern sessions, a point not clarified in the overt narrative, which only says: "after three more performances, [ . . . ] my mother acceded to my frantic supplications and the whole business was dropped" (SM
166). Years later, the
boy's frantic need to drop the whole business would again condition the choice of narrativity o f the reminiscing elder memoirist, Nabokov, who omits narrating the events concerning the issue of natural fatherhood in connection to the family. Instead, he writes about natural fatherhood in a displaced trivial context connected to his tutor. Lenski's attitude towards his business, which involved exploiting inventions that were the offspring o f somebody else's mind, is described in the following excessive terms: he "talked about them with a warmth and tenderness which hinted at
48
The text refers a number of times to a family trait that becomes an intergenerational mark or trace within a family grand narrative made of overt stories o f legitimate fatherhoods and a hidden series o f unspoken or natural fatherhoods of illegitimate hemophilic sons. This allusion is an intermedial critical metanarrative device. Intermedial, as Ryan (2003) understands it, means transferring a story between the visual and the verbal modes o f narrative, and accordingly there is both a visual and a verbal narratology. M y understanding o f intermediality is similar: the virtual space navigated by Nabokov's readers branches into different routes as they move through the text. The reading protocol inherent to Nabokov's text includes the reader's use o f intermedial loops between visual and verbal narratives. I haven't found in the typologies o f metanarrative devices a suitable slot for this kind of intertextual/intermedial allusion. Being intermedial, the allusion is a functional hybrid cutting across structural types of metanarrative and content-related types of metanarration.
238
Beatriz Penas Ibäfiez
something like a natural fatherhood—an emotional attitude on his part with no facts in support and no fraud in view" (SM 169). Readers, who are implicitly constructed as both onlookers and hearers of Nabokov's displaced narrative on natural fatherhood, are required to counterbalance this non-standard way of narrating without narrating and are asked to displace the focus of their attention from the fully narrated central events to the little marginal detail mentioned as if in passing. If the reader accepts the challenge, Nabokov's narratives become astonishingly funny and sad representations of a passing world. Otherwise, Nabokov's narrative will not be appreciated in full, although there is of course much left to enjoy. The "magician" (Wood's [1994] term for Nabokov) has conjured up a portrait and explained it in a riddle. And, as in all riddles, where something is made known and something left to be imagined, Nabokov's reader gets only the most necessary, albeit displaced, intertextual/intermedial clues. For instance, another of these narrative clues to the Nabokovian riddle can be found in the same chapter when the narrator Nabokov comments: "There is, it would seem, in the dimensional scale of the world a kind of delicate meeting place between imagination and knowledge, a point, arrived at by diminishing large things and enlarging small ones, that is intrinsically artistic" (SM 166-67). According to Nabokov, the narrative of reminiscence is artistic precisely because it fails to remember the past as it was, because it is selective, because it adds and subtracts, displaces and misplaces, and thus memory becomes the point where imagination and knowledge meet. Narrative art is possible only through a doubling process that retraces the "Alice-in-Wonderland" workings of this faculty of the mind. Like the cognitive narratives of memory, literary and autobiographical narratives metaphorically reconstruct in another medium—writing—an earlier significant experience: the altered relation of the subject to others from another time and another place. In other words, for Nabokov, telling a story about the past means going from one system of signs or medium (memorized sensory data and the events and experience associated with life) to another (a narrative in the written mode). This happens much in the same way as other intermedialities work, e.g. a verbal narrative recreation of a film (novelization), a filmic recreation of a written narrative (adaptation), a filmic recreation of somebody's life (biopic), or a written narrative of someone's life (biography). The peculiarity in Nabokov's biographical narrative, what makes it non-standard or "antinarrative," is that the remi-
A Pragma-stylistic Contribution to the Study o f NaiTativity
239
nisced is not narrated, but appears instead in displaced thematic contexts in the form of a small recurrent detail (marginal to the main narrative but central to the recovery of the unsaid by the literary narrative). The significance of the apparently irrelevant marginal detail is gained through a particular tempo of recurrence. This tempo must necessarily be different, smaller or quicker in pace, than in the source experience. It becomes longer and staccato in the case of this displacing narrative in which the main topic or significant experience must remain unnarrated and displaced from the center of the literary medium, but still accessible. But how? How is access to be gained to those untold things of the past that were essential and have lingered as involuntary reminiscences? For Nabokov, reminiscence and its narrative recall is triggered by the poignancy of a sudden experience of coincidence, a recognition of repetition and pattern, or intertextual coincidence: suddenly, something is/reads like it was or did before, except for a little something missing or added, which brings with it a feeling of loss and a longing to recover, even if in a displaced mode and form, what there was and is no more. In Nabokov's case, displacement or absence from the expected narrative place becomes a pattern applying both to his life in exile and to his text's narrative style—or exile from other, more standard ways of narrating. Speak, Memory shows that there was a pattern in the family's annual migration from city to country, from their St. Petersburg winter lodgings to Vyra, Nabokov's mother's estate and summer house. Vyra was a sanctuary, a kind of wonderland to Nabokov, not only because it meant holidays and warmth (only the family came there; the rest of the household— the Polish valet's family included—stayed in the city and was therefore absent from the summer group), but because it was beautiful and loveable and because it meant for Nabokov the discovery of the secret correspondences between the "narratibility" of life (i.e. what makes a life worth narrating or telling) and narrative art. In chapter eleven of Speak, Memory, Nabokov recounts the summer of 1914 in Vyra when he was fifteen years old and wrote for the first time. Caught in a thunderstorm, the boy took refuge in a pavilion built of wood and red, green and blue stained glass. Alone and stuck in the pavilion like a butterfly in its glass case—a papilio in his pavilion; Nabokov (SM 216) reminds the reader of the shared etymology of both words—the young Vladimir sees the rainbow duplicate in full glory and magic on the land outside the colors he sees inside the pavilion as projected by the stained glass. It is this perfected natural version of the artificial beauty surround-
240
Beatriz Penas Ibänez
ing him that clings to Nabokov's mind and makes his heart swell. It is then that he sees a leaf unbend after having being relieved of its rainwater burden by sheer weight of a raindrop falling and causing its tip to dip. Nabokov narrates in a slower tempo the momentary experience from the past as follows: Tip, leaf, dip, relief—the instant it all took to happen seemed to me not so much a fraction of time as a fissure in it, a missed heartbeat, which was refunded at once by a patter of rhymes: I say "patter" intentionally, for when a gust of wind did come, the trees would briskly start to drip all together in as crude an imitation of the recent downpour as the stanza I was already muttering resembled the shock of wonder I had experienced when for a moment heart and leaf had been one. (SM 217)
The experience of mimetic inversion, visual in the case of the crystal colors inside and the rainbow colors outside the pavilion, or auditory like the release of water on the leaf or his own heart's parallel release in the language of a poem, the patter of rain versus the patter of water falling from the trees blown by the wind, and finally the patter of his stanza—all these simultaneous signs draw the same shape in their different modalities (media) and form together a perfect instant, an unforgettable epiphanic experience. Nabokov's narrative emphasizes the coincidence between his heart's missing a heartbeat and time missing a beat. A disruption, or a fissure in time, as he says, a moment of communion with nature (a oneness of heart and leaf) and a synchronization of all that happens in one moment of time accompanies the momentous experience—"The poet feels everything that happens in one point of time," Nabokov says (SM 218)— the original ahistorical core experience that later will find a place in his narrative. The poetic nature of this experience relates to narrativity through the expansion of an epiphanic moment into a narrative of the historical sequence. In Nabokov's narrative of displacement, this expansion entails placing the narrated epiphanic moment at a marginal point in the narrative. It is a small passage about one among other childhood discoveries which does not make a claim to special status. The pavilion scene, according to Shrayer, can be described as a textual opening in Speak, Memory: "a privileged textual zone that positions the reader in such a way as to ensure its exclusive status in the reader's memory of a given text." 49 But I find a complexity in the pavilion scene that can only be captured by approaching it as an intertextual-intermedial opening in relation both to the reader and to the author as reader of his 49
Shrayer (1999: 68).
A Pragma-stylistic Contribution to the Study of Narrativity
241
past, narratively recalling (as narrator) an experience that has lingered because it was so important. An experience which, precisely because of its significance, has developed into a privileged locus of meaning recurring in narrative form through intertextual association between marginalized recurrent narrative clues belonging both to Nabokov's autobiographical text and to his recurrent oneiric texts. As he says: I dream of my pavilion at least twice a year. As a rule it appears in my dreams quite independently of their subject matter [...]. It hangs around so to speak with the unobtrusiveness of an artist's signature. I find it clinging to a corner of the dream canvas [...]. It is as it was in my boyhood [...]. Just as it was, or perhaps a little more perfect. (W215)
With regard to us, the readers of Nabokov's narrative text, the scene is an intertextual opening in the sense that it is memorable for all it discloses about the artist's life story and narrative project, and more literally memorable because it recurs in other texts by the author. Intertextual recurrence stresses the role of memory and the experience of repetition in the writer's narrative of displacement as well as in the reader's understanding of it when recognizing a textual space as familiar—both in its similarity to and its difference from another text already read and remembered. In the experience of what can be called an "intertextual" rather than simply a "textual" opening, a loop occurs between two narrative times, two moments of narrative understanding, and past and present are brought together in an "intertextual reading" through the reconstructive workings of memory rather than in a reading of the immediate textual evidence (syntax, deixis, free indirect speech, etc.) of the text. By intertextuality, I mean a dialogic relation (of textual similarity or contrast, among other possibilities) between two or more texts rather than simply allusion. 50 Nabokov's autobiography narrates isolated occurrences from different times in the autobiographer's past that are not associated by the narration, but which are there to be associated intertextually by the reader following the narrator's clue about the existence of certain recurrent patterns to the traced by an observant eye. It is not that the autobiographer just talks about his past in what is known as "subsequent" narration. The reader does not find a narrative made up for him, but on the contrary an extremely elusive narrator, apparently digressive and beating about the bush. There is too much detail and too little guidance apart from the intertextual "correspondences" (in Nabokov's terms) that stimulate an "intertextual reading," associating 50
See Penaslbänez (1996).
242
Beatriz Penas Ibäflez
what has just been read and something read before, leading one to articulate a fuller narrative that integrates what the narrator says and what he cannot say in the same connection because he does not know it well himself. Garcia Landa (2004) has described this characteristic of Nabokov's narrators as a "poetics of subliminal awareness." But in recall, as in rereading, the coming together of past and present does not occur without difference, since it is clear from the author Nabokov's awareness that there is no possibility of either remembering or narrating something as it was. The past can only be retrieved, as he says in his memoirs, "perhaps a little more perfect." In the case of the pavilion scene, it is love and happiness that motivate the will and need to remember as well as the perfection of the experience itself: for Nabokov, "Nothing is sweeter than to ponder those first thrills. They belong to the harmonious world of a perfect childhood and, as such, possess a naturally plastic form in one's memory which can be set down [in narrative writing, Nabokov means] with hardly any effort" (SM 24-25). The pavilion scene is just a detail in a life narrative, but Nabokov's memory treasured it, gave it written narrative form in Speak, Memory and made it readable so that "the delicate union did take place, with the magic precision of a poet's word meeting halfway his, or a reader's, recollection" ( S M 2 1 \ ) . The element of plasticity which, according to Nabokov, imbues our memories of a perfect childhood makes these special memories particularly easy to adapt to a written-language narrative. In saying this, Nabokov seems to imply that love and desire lie at the heart of narrative art. Timelessness, too. The transience of a moment in an individual life is transcended by the relative permanence and time expansion of the narrative that tells it. Thus, in Speak, Memory the artist says I confess I do not believe in time. I like to fold my magic carpet, after use, in such a way as to superimpose one part of the pattern upon another. Let visitors trip. And the highest enjoyment of timelessness—in a landscape selected at random—is when I stand among rare butterflies and their food plants. This is ecstasy, and behind the ecstasy is something else, which is hard to explain. It is like a momentary vacuum into which rushes all that I love. A sense of oneness with sun and stone. A thrill of gratitude to whom it may concern—to the contrapuntal genius of human fate or to tender ghosts humoring a lucky mortal. (SM 139)
Nabokov repeatedly elaborates on the theme of transformation or metamorphosis to refer to displaced transferrals from medium to medium, from significant past experience into recurring dreams and memories
A Pragma-stylistic Contribution to the Study of Narrativity
243
whose tactile, visual and acoustic cinema-like images the artist must in turn reinterpret into an aesthetic form articulated primarily into the written-English medium. Nabokov's awareness of the metamorphic migrant quality of his self is the source of his powerful metaphors and the trigger of the non-standard narrativity of his text. The growing of the self into another self leaves behind a plurality of selves comparable to the repeatedly shed shirts of a snake. The growing of a form of literary narrative out of the forms of everyday language is like a pupa becoming a butterfly, a metamorphosis from plain use into stylization, the achievement of more meaningfulness out of non-standard narrative forms. It was through his ample experience of life and work as a translator, fictionist and autobiographer that Vladimir Nabokov became a master in the arrangement and displacement of signs, the signs of Russian that he translated into the signs of English as well as the signs of life and experience that he translated into signs of memory, voluntary and involuntary, before he could translate them into signs of art. His art was narrative, the art of the word, as literature is sometimes called, but his words are designed to convey the visual and auditory images, the remembered sights and sounds, which were most significant in Nabokov's life—most significant because they were perceived as parts within a whole, patterned in an ethical and aesthetic way. For that reason, Nabokov writes: "The spiral is a spiritualized circle. In the spiral form, the circle, uncoiled, unwound, has ceased to be vicious; it has been set free. [...] A colored spiral in a small ball of glass, this is how I see my own life" (SM 275). Again an intertextual opening, this narrated colored spiral in a small ball of glass repeats the remembered and narrated image of a boy, that small papilio, inside a stained glass pavilion, learning to look at himself self reflexively as the stained glass windows surrounding him refracted their colors on him and converted him, or some of his most significant experiences, into the object/subject of his own narrative. 4. Conclusions If untold (unnarrated) narratives can be conveyed by means of a stylistically complex narrative design and if any stretch of textual form can be understood to convey a narrative in a certain situational context, then narrativity (like literariness) need not be understood in the singular as an abstract intrinsic structural quality or as a typical differential property of a specific textual form, what Sturgess calls "the work's capacity for being a
244
Beatriz Penas Ibäfiez
[literary] narrative." 51 Narrativity is, rather, a (con)textual interpretative effect derived as much from a way of reading a text 52 as from a way of writing a text. Placing narrativity both within and beyond the narrative text entails the parallel acknowledgement of an essential identity between text and context, which translates as an essential identity between canonical standard well-formed kinds of narrative text and canonical prestigious contexts. Under the sociopragmatic light, narrativity can be understood as proceeding along different sociosemiotic representational processes regulating the signic/formal material employed in particular forms of reading and writing narrative: (1) Selection on several levels: a) "Language/medium" level, or primary selection of the kind of signic matter (visual written, auditory phonetic signs of a natural language, visual filmic image, musical notes or any others) in which narrativity is to be articulated; b) "Situational Style" level, or secondary selection of specifically proper ways of elaborating text according to a situational need. With regard to narrativity, selection is made from a system of specific generic forms or cultural conventions (the realist novel, the news report, etc. as distinctive types of text among others) according to the speech plan (telling) of the speaker/writer; c) "Individual Style" level. The individual writing-plan, in the case of literary narrativity, is in turn conditioned by a literary system or tradition within which the writer is bound to create his narratives while retaining a measure of freedom to choose either to comply with or to deviate from that tradition in some—never all—respects. Hemingway and Nabokov offer important examples of non-standard stylistically deviational narrativity.
51
Sturgess (1990: 6). The same volume (Sturgess [1990: 16-22]), reports Prince's (1982) insistence on degrees and kinds o f narrativity: in effect, the two types I discuss here might be said to go in the same direction. I do not think so, though. One narrative can be compared to another in Prince's sense, but a text being more or less narrative is an issue only when a strictly formal (transformational or otherwise) conception of narrative is subscribed to. In my concept of narrativity (a text- and context-dependent one), the relevant dimension to be measured is not narrativity per se but the degree of conventionality o f a certain type o f narrativity at a certain time.
52
Scholes (1982: 60).
A Pragma-stylistic Contribution to the Study of Narrativity
245
It is on the level of second-degree kinds of selection (b) and (c) that secondary (literary) foregrounding operates in the terms set out by the Prague School. Of course, primary (linguistic utterance) foregrounding may take place as well, right from level a), when the different voices within the narrative are orchestrated against a social background. Intertextuality cuts across all levels in what I have called the intertextual dimension of discourse, following Bakhtin's theory of language. 53 (2) Elaboration. The next step in the process, after selection, is the elaboration of a diversity of narratives, non-verbal or verbal: oral and written, literary or not, fictional or non-fictional, short or long, high or low (folk), restricted or elaborated, and so on. We see this happen within the specific situation of a literary tradition by the formation of schools and generations of writers following their own rules of composition so as to renew the modes of narrativity. Hemingway and Nabokov belong to a modernist generation of writers bent on the elaboration of deviational narratives to be understood as such against a standard belonging to earlier nineteenth-century conventional well-formed narratives. Still, these deviational narrativities are also meant to become classic in their due term. (3) Codification. On the basis both of sustained elaboration within a tradition and of the corresponding critical reading and comparative observation of an elaborate diverse set of narrative texts, theory can grow and focus either on formulating the typology of different kinds of narrative text or on the scholarly codification of "the" narrative norms which, disregarding their being peculiar to one variety of educated narrativity, come to be regarded as the proper standard by which to measure all narrativity. Here, in my opinion, lies the reason why classical narratology has relied so much on abstract theoretical aspects like narrative grammar, characteristic and -emic features of narrative, and normative narrative components and narrative levels. Codification of the well-formed narrative text entails reduction of internal variability and temporal stability of the selected standard variety of narrative. Thus, the standard—in fact, only the more prestigious variety of narrative in a particular situation and time—comes to be regarded as the intrinsically (atemporal and normal) form of narrative. This happens much in the same way as the educated community's norm comes to stand for a country's language, leading to a confusion between the more prestigious variety of a language and la langue.
53
Cf. P e n a s l b ä f i e z ( 1 9 9 6 ) .
246
Beatriz Penas Ibanez
(4) Implementation. The perception of a form of narrativity as intrinsic to narrative and the various ways (creative and theoretical or primary and secondary) of dealing with a specific kind of narrative leads to the general implementation of a certain kind of narrativity and a certain discourse on narrativity. But it also leads specifically to the implementation of standard narrativity both in creative and critical discourses and in the related focusing agencies. Le Page and Tabouret-Keller identify a phenomenon they call "focusing." 5 4 A focused linguistic community is one in which there is a strong sense of norms. There are four key focusing "agencies": close daily interaction in the community; mechanisms of an educating system; a sense of group loyalty, perhaps caused by perception of a common threat; and the presence of powerful models, such as the opinion of leading critics and canonical authors. These "agencies" draw attention to their discourses, they are discourses of power, call for in-group identification and exclusion, and outline the proper or acceptable ways of using language in their own contexts. An example of focused linguistic (sub)community would be the community of writers, critics and academics reflecting on narrativity from their respective positions and generating narratives, narratologies and criticism. Narrativity is identified as existing against a traditional background of a canonical body of writings and authors as defined by influential critics and theoreticians, usually sacrificing narrative diversity and diverse narrativities—narrativities generated differently and producing different effects connected to different ways of writing and interpretation—on the altar of the inherited canon. The fact is that the history of narrative writing teaches critics, lay readers and authors the value of stylistic deviation. We have grown accustomed to seeing some kinds of deviational narrativity as a sign of creative power. 55 But this positive sociosymbolic value of deviation in the realm of non-standard literary narrativity does not extend to all deviant narrativities. We can think of the apparent awkwardness of children's narratives, the naivety of the everyday life narratives of uneducated working class adults, the exoticism of the narratives of minority group
54
Le Page and Tabouret-Keller's (1985: 187) proposal applies to the general phenomenon of linguistic focusing which, in my sociopragmatic model, is applicable to the more specific case of the linguistic community of readers and writers of narrative. Bloom's concept of creative power and authorial strength as bound to the anxiety of influence is pertinent here.
A Pragma-stylistic Contribution to the Study of Narrativity
247
members. The added qualification in each case points to a stigmatization of non-literate substandard narrativities.56 Most importantly, the kinds of narrativity mentioned above can be integrated as stemming from and contributing to the language user's 57 experiential continuum: in other words, his experience of life as lived among others and also of his life and the lives of others as heard of, spoken and written about in the linguistic (or non-linguistic) narratives of the self and others within which (literary or aesthetic) narrativity is articulated. Nabokov's autobiographical narratives and Hemingway's parables are good exponents of experiential narrativity, but they are also narratologies in disguise that explain the individual experiential motivation of particular styles of narrativity.
References Baker, Carlos 1972 Hemingway: The Writer as Artist (Princeton: Princeton UP). Bakhtin, Mikhail M. 1981 The Dialogic Imagination, edited by Michael Holquist, translated by Caryl Emerson and Michael Holquist (Austin: University of Texas). Barthes, Roland 1977 [1971] "From Work to Text," in R. B.: Image Music Text, translated by Stephen Heath, 155-64 (New York: Hill and Wang).
Benjamin, Walter 1969
[1939] "On Some Motifs in Baudelaire," in W. B.: Illuminations, translated by Harry Zohn, introduction Hannah Arendt, 155-200 (New York: Schocken Books). Bloom, Harold, et al. 1979 Deconstruction and Criticism (London: Routledge and Kegan Paul). Boyd, Brian 1993 [1991] Vladimir Nabokov: The American Years (London: Vintage). Derrida, Jacques 1981 [1972] "Plato's Pharmacy," in J. D.: Disseminations, translated by Barbara Johnson, 61-172 (London: Athlone). Eco, Umberto 1962 Opera aperta (Milan: Bompiani). 1978 "Semiotics: A Discipline or an Interdisciplinary Method?" in Sight, Sound and Sense, edited by Thomas Sebeok, 73-110 (Bloomington: Indiana UP). 56
Cf. Goffman (1963). Scholes gives the following definition of narrativity: "I should like to employ the word 'narrativity' to refer to the process by which a perceiver actively constructs a story from the fictional data provided by any narrative medium." (1982: 60).
248
Beatriz Penas Ibäfiez
Field, Andrew 1987 The Life and Art of Vladimir Nabokov (London: Macdonald Queen Anne Press). Fishman, Joshua A. 1970 Sociolinguistics: A Brief Introduction (Rowley, Mass: Newbury House). Fowler, Roger 1981 Literature as Social Discourse: The Practice of Linguistic Criticism (Bloomington: Indiana UP). Garcia Landa, Josd Angel 2004 "The Poetics of Subliminal Awareness: Re-reading Intention and Narrative Structure in Nabokov's 'Christmas Story'," in European Journal of English Studies 8: 27-48. Genette, Gerard 1976 [1966] "Boundaries of Narrative," in New Literary History 8: 1-15. Goffman, Erving 1963 Stigma: Notes on the Management of Spoiled Identity (New York: Simon & Schuster). Grebstein, Sheldon N. 1973 Hemingway's Craft(Carbondale: Southern Illinois UP). Green, G. M. 1989 Pragmatics and Natural Language Understanding (Hillsdale, NJ: Lawrence Erlbaum). Grice, Herbert P. 1957 "Meaning," in Philosophical Review 64: 377-88. 1975 "Logic and Conversation," in The Logic of Grammar, edited by Donald Davidson and Gilbert Harman, 64-75 (Encino, CA: Dickenson). 1981 "Presupposition and Conversational Implicature," in Radical Pragmatics, edited by Peter Cole, 183-98 (New York: Academic Press). Gumperz, John J. 1992 "Contextualization and Understanding," in Rethinking Context: Language as an Interactive Phenomenon, edited by Alessandro Duranti and Charles Goodwin, 229-52 (Cambridge: Cambridge UP). Hemingway, Emest M. 1932 Death in the Afternoon (New York: Scribner). 1957 [1929] A Farewell to Arms (New York: Scribner). 1966 [1935] Green Hills of Africa (Harmondsworth: Penguin). 1967 By-Line: Emest Hemingway, edited by William White (Harmondsworth: Penguin). 1986 The Garden of Eden (New York: Scribner). 1991 [1936] "The Snows of Kilimanjaro," in Ε. Μ. H.: The Complete Short Stories, 39-56 (New York: Scribner). Hulme, Τ. E. 1998 [1924] "Notes on Language and Style," in Τ. Ε. H.: Selected Writings, edited by Patrick McGuiness, 37-58 (Manchester: Carcanet Press Ltd). Kindt, Tom / Müller, Hans-Harald 2003 "Narrative Theory and/or/as Theory of Interpretation," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt
A Pragma-stylistic Contribution to the Study of Narrativity
249
and Hans-Harald Müller. Narratologia 1: 205-20 (Berlin and New York: Walter de Gruyter). Kristeva, Julia 1969 [1966] "Le mot, le dialogue et le roman," in J. K.: Semeiotike. Recherches pour une semanalyse. Essais. Coll. "Tel Quel," 143-73 (Paris: Editions du Seuil). Le Page, Robert B. / Tabouret-Keller, Andree 1985 Acts of Identity: Creole-Based Approaches to Language and Ethnicity (Cambridge: Cambridge UP). Lewis, Robert W. 1966 "Vivienne de Watteville: Hemingway's Companion on Kilimanjaro," in Texas Quarterly 9: 75-88. Mey, Jacob L. 2000 When Voices Clash: Studies in Literary Pragmatics (Berlin and New York: Mouton de Gruyter). Miller, Jane 1992 Working Time: Essays on Poetry, Culture, and Travel (Ann Arbor: University of Michigan). Miner, Earl 1976 "The Objective Fallacy," mPTL 1: 11-31. Nabokov, Vladimir 1989 [1967] Speak, Memory. An Autobiography Revisited (New York: Vintage). Nünning, Ansgar 2003 "Narratology or Narratologies? Taking Stock of Recent Developments, Critique and Modest Proposals for Future Usages of the Term," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 239-75 (Berlin and New York: Walter de Gruyter). Osgood, Charles E. 1960 "Some Effects of Motivation on Style of Encoding," in Style in Language, edited by Thomas A. Sebeok, 293-306 (Cambridge, Mass: MIT Press). Penas Ibäflez, Beatriz 1994 "Patterns of Violence in Hemingway's Worlds of Fact and Fiction," in Amor, odio y violencia en la literatura norteamericana, edited by Jose Antonio Gurpegui, 234-46 (Alcalä and Madrid: Alcalä de Henares UP). 1996 "Intertextuality, a dialogical relation: Μ. M. Bakhtin's Contribution to Pragmatics," in The Intertextual Dimension of Discourse: Pragmalinguisticcognitivehermeneutic Approaches, edited by Beatriz Penas Ibäfiez and prefaced by Jef Verschueren, 179-90 (Zaragoza: Servicio de Publicaciones de la Universidad de Zaragoza). 1998 "Language and Experience in Cognition and Understanding: A Preface to The Pragmatics of Understanding and Misunderstanding," in The Pragmatics of Understanding and Misunderstanding, edited by Beatriz Penas Ibäfiez, 7 - 1 6 (Zaragoza: Servicio de Publicaciones de la Universidad de Zaragoza). 2003 "Hemingway's Ethics of Writing: The Ironic Semantics of 'Whiteness' in 'The Snows of Kilimanjaro'," in NDQ (North Dakota Quarterly) Hemingway: Life and Art (Monograph) Fall 2003: 94-118.
250 2004a
Beatriz Penas Ibänez
'"Very Sad but very Fine': Death in the Afternoon's Imagist Interpretation of the Bullfight-Text," in A Companion to Ernest Hemingway's Death in the Afternoon, edited by Miriam B. Mandel, 143-64 (Rochester, NY: Camden). 2004b "Signs of Memory, Signs of Writing: Nabokov's Narrative Integration of Self, Language and World," In Memory, Imagination and Desire in Contemporary Anglo-American Literature and Film, edited by Constanza del Rio-Alvaro and Luis M. Garcia-Mainar, 81-93 (Heidelberg: Winter). Pier, John 2003 "On the Semiotic Parameters of Narrative: A Critique of Story and Discourse," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 73-97 (Berlin and New York: Walter de Gruyter). Pound, Ezra 1930 "Small Magazines," in The English Journal 19.9: 689-704. Prince, Gerald 1982 Narratology: The Form and Functioning of Meaning. Janua linguarum, Series Maior 108 (Berlin, etc.: Mouton). 2003 "Surveying Narratology," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 1-16 (Berlin and New York: Walter de Gruyter). Reddy, M. J. 1979 "The conduit metaphor: a case of frame conflict in our language about language," in Metaphor and Thought, edited by Paul Ortony, 284-324 (Cambridge: Cambridge UP). Reynolds, Michael 2000 "Ernest Hemingway, 1899-1961: A Brief Biography," in A Historical Guide to Ernest Hemingway, edited by Linda Wagner-Martin, 15-50 (New York: Oxford UP). Riffaterre, Michael 1984 Semiotics of Poetry. Advances in Semiotics (Bloomington: Indiana UP). Ryan, Marie-Laure 2003 "Narrative Cartography: Toward a Visual Narratology," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 333-64 (Berlin and New York: Walter de Gruyter). Schmid, Wolf 2003 "Narrativity and Eventfulness," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 17-33 (Berlin and New York: Walter de Gruyter). Scholes, Robert 1982 Semiotics and Interpretation (New Haven, CN: Yale UP). Sell, Roger 2004 FILLM Noumea 2003. Emergent Literatures and Globalisation: theory, society and politics (Paris: In Press Editions). Shrayer, Maxim D. 1999 The World of Nabokov 's Stories (Austin: University of Texas Press).
A Pragma-stylistic Contribution to the Study of Narrativity
251
Smith, Paul 1989 A Reader's Guide to the Short Stories of Ernest Hemingway (Boston: G. K. May). Sturgess, Philip J. M. 1992 Narrativity: Theory and Practice (Oxford: Clarendon Press). Tedlock, Emest W. 1949 "Hemingway's 'The Snows of Kilimanjaro'," in Explicator 8 (October): Item 7. Wallace, Robert. 1991 Writing Poems (New York: Harper Collins). Walcutt, Charles C. 1949 "Hemingway's 'The Snows of Kilimanjaro'," in Explicator 7 (April): Item 43. Westbrook, Max 1966 "The Stewardship of Ernest Hemingway," in The Texas Quarterly 9: 89-101. Wood, Michael 1994 The Magician's Doubts: Nabokov and the Risks of Fiction (London: Chatto and Windus).
DAVID RUDRUM (Huddersfield)
Narrativity and Performativity: From Cervantes to Star Trek In making a systematic study of almost anything, there is normally a methodological obligation, going back at least to Aristotle, to differentiate between what the object of study is, and what it does. Such a move has served most disciplines very well. Yet this distinction has sometimes been hard to preserve in narratology, where the term 'narrativity' has been traditionally used to describe both what makes a narrative a narrative, and the nature of how narratives work. Nor is this ambiguity necessarily a problem, since, after Wittgenstein, there is a sound case for exempting linguistic phenomena from this distinction and for keeping what language is and what it does in the same frame. Some of the consequences of this insight for the concept of narrativity are explored in this paper. In what follows, my argument will be divided into two halves. The first sets out to interrogate the ways in which the concept of narrativity has been articulated historically and to diagnose some of the problems with traditional views of the subject. The second half seeks to formulate the grounds for a new understanding of narrativity based on the concept of performative language which offers, I will argue, a more satisfactory way of thinking the relationship between what narrative is and what it does. The argument that narrativity is a performative force amounts to a reorientation of the way we think narrativity, for it aligns narrativity not so much with formal, structural, or textual features, but with particular sets of conventions and expectations located in the linguistic community, and with the specific roles that narratives play in our lives. But before making this argument, it will be worthwhile first to explore some of the difficulties that have arisen from traditional views of narrativity as an intrinsic property—indeed, as the intrinsic property—of narrative itself.
254
David Rudrum
1. Narrativity: A Problematic Concept The concept of narrativity, when it was first mooted some three decades ago, was ambitious and theoretically complex. Gerald Prince, the narratologist who introduced the term to the English-speaking world, defines it as 'The set of properties characterizing narrative and distinguishing it from nonnarrative; the formal and contextual features making a (narrative) text more or less narrative, as it were'. 1 In this brief description, Prince not only outlines the nature of the concept, but gives several hints as to how the concept of narrativity actually works. These hints imply: 1) narrativity is the property that makes a narrative a narrative; therefore 2) the presence of narrativity distinguishes what is narrative from what isn't; and furthermore 3) the more narrativity a text has, the more of a narrative it will be. In crude terms, such a traditional conception of narrativity might well be called the 'Philosophers' Stone' theory of narrativity. It conceptualises narrativity as a property whose presence can, so to speak, 'leaven' a piece of text: in the presence of narrativity, a piece of discourse is transformed from the leaden status of mere text into a narrative. Thinking of narrativity in this way implies that it is the essential ingredient of narrative: 'Just Add Narrativity', might read the narratologist's instructions to the storyteller. To characterise the concept of narrativity in the way I have just set out is of course to caricature it, and to caricature a very traditional articulation of it at that. Prince himself has honed the sophistication of the concept by differentiating between distinct (though related) aspects of narrativity, such as 'narrativehood', 'narrativeness', and 'narratability'. 2 Indeed, it is telling that none of the contributors to the present volume endorses this now dated 'Philosophers' Stone' notion of narrativity. Nevertheless, the point remains that the history of narrativity as a concept originated in an attempt to posit an essential ingredient of narrative: that which is unique to narratives, that which makes a narrative a narrative, that which distinguishes narrative from other texts. And the fundamental problem with this view is that it approximates the idea of an essence of narrative. Indeed, the view that narrativity constitutes an essence of narrative is explicitly endorsed by Philip Sturgess, who claims that narrativity is not only 'the semantic essence of a narrative', but also 'the essential quality or qualities of any particular narrative system'; narrativity, for Sturgess, is 'that which
1
Prince (2003: 65). See Prince (1999) as well as his essay in this volume.
Narrativity and Performativity: From Cervantes to Star Trek
255
must be deemed essential by simple definition, namely the work's capacity for being a narrative, its narrativity'. 3 It is worth asking what such an essence of narrative, such a sine qua non, might be. Sturgess plausibly suggests that it can be described in logical terms: specifically, as a 'logic of narrativity' which is identified with 'higher-level causality'. 4 He claims that 'the theory of a logic of narrativity is a theory of a priori causality in a precise and "essential" sense', not least because 'causality is a necessary feature of narratives'; hence, 'causation may be thought of as a logic of narrativity'. 5 Whilst there is much that is persuasive about this view, it should be kept in mind that causality is a vital ingredient in many other kinds of text which we probably would not consider narrative (such as certain kinds of philosophical or logical reasoning, mathematical equations, and rhetorical strategies). Identifying causality as the essence of narrative is therefore a question-begging move, although not without great heuristic value. 6 Gerald Prince's work offers one of the earliest articulations of the concept of narrativity, dating back to 1982. Though now somewhat dated, as Prince's contribution to this volume demonstrates, this work is of immense historical importance in the evolution of the concept, as it is one of the first attempts to articulate the essential ingredients of narrative. Summarising his views in that piece, Prince has described them thus: the narrativity o f a text d e p e n d s on the extent to w h i c h that text constitutes a d o u b l y oriented a u t o n o m o u s w h o l e (with a w e l l - d e f i n e d and interacting b e g i n n i n g , middle, and end) w h i c h i n v o l v e s s o m e kind o f conflict, [ . . . ] w h i c h is m a d e up o f discrete, particular, positive, and temporally distinct actions having l o g i c a l l y unpredictable antece-
3
Sturgess ( 1 9 9 2 : 1 0 , 6 ) .
4
Ibid.: 32, 52.
5
Ibid.: 31. It has correctly b e e n pointed out to m e that Sturgess is describing causation a s 'a logic o f narrativity', not 'the l o g i c ' . T h e possibility therefore remains that a different kind o f logic o f causality w o u l d obtain for other, non-narrative texts (mathematical equations, philosophical deductions, etc.). M o r e o v e r , Sturgess implies here that causation m a y not be a sufficient condition for narrativity—but it is nevertheless a necessary one. A l though Sturgess refines his position b y distinguishing b e t w e e n a number o f principles according to w h i c h h i s l o g i c o f narrativity unfolds, he nevertheless insists that this logic is a single, unitary logic, a position that, as I am about to argue, has problematic implications. M y thanks g o to Jose A n g e l Garcia Landa and John Pier for their h e l p with these observations.
256
David Rudrum
dents or consequences, and which avoids inordinate amounts of commentary about them, their representation, or the latter's context.
There is, however, one key point missing from this summary, and that is the extent to which Prince's 1982 piece also emphasised the process of reading or 'processing' of the text as a narrative, which adds another dimension to the concept of narrativity. This aspect of his argument has been very influential. As Prince has it: In general, when we read a text as a narrative, we try and process it as a series o f assertions about events the occurrence o f which is not in doubt. The easier such processing proves to be, the more readily a text suggests it and lends itself to it, the more narrativity that text will have.
But bringing the question of 'processing' into the picture is a problematic move, for it risks confusing narrativity with readability, and this effectively locates narrativity in the eye of the beholder: 'Narrativity depends on the receiver', as Prince put it.9 This does not square with his view that locates narrativity firmly in the textual or structural properties and features described above. Indeed, if we follow Prince's suggestion that narrativity inheres in the kind of text that a reader finds 'interesting', 10 then it is unclear how one could subsequently proceed to describe the nature of narrativity in any formal or textual terms, since de gustibus non disputandum est.11 While there is clearly much of interest and indeed of merit in these views, recent scholarship has tended, with good reason, to be more wary of a conception of narrativity that sets itself out as the essence of narrative. To advance claims about the essence of narrative is to attempt to pin down, delimit, and restrict the almost boundless variety of functions, forms, and roles that narratives can and do take on in the cultural practices of everyday life. Furthermore, the idea of an essence of narrative encourages the artificial segregation of narrative discourse from other kinds of text to which it is intimately wedded (such as description, argumentation,
7 8 9 10
Prince (1999: 45). Prince (1982: 150). Ibid.: 160. Ibid.: 153, 160. Prince has since differentiated 'narrativeness', 'narratability' and 'narrativehood' from narrativity itself, a move that distinguishes the reading (and evaluating) of narrative from the property of narrativity, thereby resolving this apparent contradiction.
Narrativity and Performativity: From Cervantes to Star Trek
257
dialogue, etc.) 1 2 Such an approach is highly restrictive—indeed, at times it has even struck some narratologists as prescriptive. In his Gender and Narrativity, Barry Rutland claims that the traditional conception of narrativity entails 'the construction of normative worlds that legislate reality in terms of values'. 13 If there is to be any value in the concept of narrativity, then that value does not lie in trying to reduce the narrativity of narrative to something that can be given once and for all, independently of circumstances: it lies in helping us to understand the many different ways in which narratives work in our literature, culture, and everyday lives. That, I maintain, is what is on offer in a performative view of narrative. 2. A n E x a m p l e from Don
Quixote
It can readily be seen from the preceding discussion that, as Sturgess puts it, 'the idea [of narrativity] itself is often invoked in vague or even contradictory ways'. 14 In order to illustrate how problematic these traditional conceptions of narrativity can be, the best strategy is to apply them to a text. The example I have in mind is taken from Cervantes' Don Quixote (1605/15). In it, Sancho Panza tells a story to Don Quixote, and the ensuing tale is one of the strangest embedded narratives in a novel that is simply packed with strange embedded narratives. This particular example is the abortive story of the goatherd Lope Ruiz, who abandons his former lover, the shepherdess Torralba, and flees to Portugal. Torralba pursues Lope Ruiz to the edge of the Guadiana river, which he must cross in order to escape, taking his flock with him. As Sancho tells it: '[H]e saw a fisherman close beside a boat, which was so small that it could only hold one man and one goat. But, all the same, he hailed him and arranged for him to take himself and his three hundred goats across. The fisherman got into the boat and took one goat over, came back and fetched another, and came back once more and took another. Keep an account of the goats which the fisherman is taking over, your worship, for if you lose count of one the story will end, and it won't be possible for me to tell you another word of it. I'll continue now and mention that the landing-place on the other side was very muddy and slippery, which delayed the fisherman a good deal in
12
13 14
Setting out an essence of narrative akin to advancing a definition of narrative. For an argument against such a move, see Rudrum (2005). Rutland (1997: 3) Sturgess (1992: 3).
258
David Rudrum
his journeys backwards and forwards. But, all the same, he came back for another goat, and another, and another.' 'Take it that they are all across', said Don Quixote, 'and do not go on coming and going like that, or you will never get them all over in a year.' 'How many have you got over so far?' asked Sancho. 'How the devil should I know?' replied Don Quixote. 'There, now, didn't I tell you to keep a good count? Well, there's an end of the story. God knows there's no going on with it now.' 'How can that be?' replied Don Quixote. 'Is it so essential to the tale to know exactly how many goats have crossed that if you are one out in the number you cannot go on?' 'No, sir, not at all,' answered Sancho. 'But, when I asked your worship to tell me how many goats had got across and you replied that you didn't know, at that very moment everything I had left to say went clean out of my head, though there were some good and amusing things coming, I promise you.' 'So,' said Don Quixote, 'the story is finished, then?' 'As sure as my mother is,' said Sancho.
In terms of the traditional view of narrativity, it is very hard to know what to make of Sancho's story, and this confusion may well be related to how we understand that concept. If we assume, after Prince, that narrativity is a structural property related to the dynamics of the text's 'general narrative configuration', 16 then there is good reason to suppose that Sancho's story should achieve an optimal level of narrativity: there is, after all, a straightforward timeline with the narrator refusing to bow to his audience's request to introduce even a simple prolepsis. There is, furthermore, a one-to-one correspondence between the events as they happen and the events as they are narrated: this is a strictly chronological, singulative narrative. Nothing about it suggests any kind of formal complexity—in fact, narratives rarely get simpler or more straightforward than this—and one would therefore assume it should have a direct, simple, 'readerly' appeal and, accordingly, a high level of narrativity, at least in structural terms. Clearly, though, it doesn't: an impatient Don Quixote tries to hurry Sancho's story along, and Sancho doesn't even get to complete his tale.
15
16
Cervantes (1950: 154). I am grateful to Cascardi (2003) for drawing my attention to this idiosyncratic example of abortive narration. Cascardi's discussion emphasises the relationship this piece establishes between the two senses of contar: to count (as in the related word cuenta) and to narrate (as in cuento). His point—that counting is fundamental to recounting—could well contribute a great deal to our understanding of narrativity, but there is not room to explore that possibility here. Prince (1999: 48).
Narrativity and Performativity: From Cervantes to Star Trek
259
That the narrator of this embedded narrative cannot even hold the narratee's attention, and that the story is interrupted and never properly concluded, are surely indications that this particular text lacks narrativity. And there are as many features of the text we could point to to demonstrate this lack of narrativity as there are to demonstrate its abundance: there does not appear to be any satisfactory resolution to the story, either in terms of form or of content, and the tally of goats assumes a greater importance than the fate of two lovers, which is rather counterintuitive to anyone familiar with Propp's or Greimas's accounts of the makeup of narratives, let alone to anyone who likes a good love story. On this view, Sancho's story should surely have minimal narrativity. 17 Furthermore, if we apply the traditional reasoning that the more narrativity a text possesses, the more of a narrative it is, then the converse of this principle would lead us to conclude that this text is at best not much of a narrative. One could even invoke the terms that narratologists have borrowed from Aristotle's Poetics and claim that without the obligatory beginning, middle, and end (and this tale seems to lack any intelligible end), this story is not even a narrative at all. The problem with this view is that Sancho's story clearly does have an end—it simply stops—but this end is not one that is expected, or amenable to description, within the context of the logic of this narrative of conflict between two lovers. To look at the problem this way is to follow, roughly, the position of Sturgess and its emphasis on the logic of the story. Clearly, there is a sudden volta in the logic of the story at the moment of Don Quixote's interruption: prior to this moment, Sancho tells an easy-to-follow, logical story; after this point, he abandons this logic and the story comes to an end. The story, then, has a certain kind of narrativity up to this point and then, as it were, discards it. Attractive as this simple solution is, though, it will not quite do: why does Don Quixote interrupt such a transparently logical story? At its most ambitious, Sturgess's theory of narrativity claims to be able to account even for the interruptions to any given narrative in terms of their narrativity. He emphatically states 'Narrativity determines not only 17
Consider the following assertion from Prince (1982: 153): "Narratives with a high degree of narrativity will not merely describe change and its results but fundamental changes and results. They will take us from the origin to the conclusion, from 'Once upon a time' to 'They lived happily ever after', from the onset o f heterogeneity and difference back to homogeneity and indifference." On these criteria, Sancho's tale would have a very low degree of narrativity indeed.
260
David Rudrum
the chronology of a novel's story, but equally every interruption of that chronology, and every variation in the mode of representation of that story'In short, 'narrativity is defined as that which enables the narrative to proceed at any and every stage', 19 including disjunctive moments such as Don Quixote's interruption. The basis for these claims lies in Sturgess's notion that narrativity provides a 'higher level' logic of causal continuity that drives the story forward. As he puts it: 'Every sequence in a narrative will tell a story, namely the story of its narrativity, but the ways in which this story is discussable in terms of the story being told will be many and various'. 20 To put the point more simply: at a certain juncture in the text, Sancho interrupts the logic of the narrative to tell a story of his own, and this story involves a rather different logic of narrativity; but Don Quixote, who does not appreciate this different logic, interrupts it in order to revert back to a more conventional logic of narrativity, similar to that which governs the framing narrative. Viewed this way, this episode provides a good example of the kind of narrativity that Sturgess describes as 'a logic sometimes related directly but far more often related deviously and deceptively to that logic discernible in the conduct of a novel's represented characters, and of the world they inhabit.' 21 This is a very plausible and attractive view and accounts quite persuasively for the sudden volta in the narrativity of the above extract: there is, as it were, a double logic of narrativity in evidence here. Unfortunately, however, this view contradicts a major aspect of Sturgess's position: he spends no less than two chapters of his book refuting the very idea of a 'double logic' of narrativity. 22 Sturgess in some ways anticipates this contradiction, and his suggested resolution is that in instances where multiple narrativities are evident, the logic of narrativity is 'determined by the author'. 23 He goes on—and this is particularly applicable to the example 18 19
20 21
22
Sturgess (1992: 22) (emphasis in the original). Ibid.: 152. Ibid.: 22 (emphasis in the original). Ibid.: 29. These chapters provide a critique of the idea o f a 'double logic' in narrative texts as propounded by deconstructive criticism (see Chapter Three) and Marxist criticism (see Chapter Four). Ibid.: 141. Indeed, Sturgess's theory places a huge amount of weight on the author. His whole discussion is 'conducted with a view to explaining how a biological being—an author—produces a narrative' (ibid.: 163). There is, I would suggest, a simple reason why this emphasis on the author is necessary to his position. His theory is predicated on
Narrativity and Performativity: From Cervantes to Star Trek
261
at hand—stating: 'This becomes especially clear in novels wherein there are two or more distinct narrators [...]. Neither narrator is controlling the manifestations of the other; both are subject to the workings of narrativity'. 24 But rather than forestall an apparent contradiction in Sturgess's position, this emphasis on the author begs another question. Granted, it seems fairly evident that the logic of narrativity employed by the framing narrator of Don Quixote is a very different logic of narrativity from that employed by Sancho, which is in turn different from that preferred by Don Quixote. But who is to say which of these logics of narrativity, if any, is that of Miguel de Cervantes himself? It may seem reasonable enough to suggest that all of them are to some extent identifiable with their creator, under whose auspices they can be subsumed. But that does not forestall the possibility of two (or more) contradictory logics of narrativity operating within the same narrative. If we follow Sturgess's view that 'to trace narrativity [...] is to trace the logic according to which the text elaborates itself, 2 5 then the clearest explanation for Sancho's abortive embedded narrative is that there is indeed a double logic of narrativity in operation here, which interrupts and undermines Sancho's story. This conclusion may not be congenial to Sturgess's theory, but his solution—to appeal to the greater figure of the author—is unhelpful, since there is just no knowing whether Cervantes intended the logic of his tale to be unitary or double, coherent or self-contradictory, or indeed whether he even succeeded in his intentions in the first place. 26
24 25 26
causality and on the causal nature of narrative: 'It is the logic o f narrativity which controls this unbroken sequence o f causal links from a work's beginning to its end' (ibid.: 141). And yet he argues that 'narrativity as such is not self-initiating' (ibid.: 22). For his theory to be coherent, then, he needs some sort o f agency to set this causal train in motion, some sort o f 'first cause', as it were, onto which this chain can be linked. Without it, his vision of the logic of narrativity falls prey to an ontological flaw. The author is therefore required by his theory in order for this initiating, causal agency to be provided. Whether or not this emphasis on the author constitutes a weakness in Sturgess's position is the subject for another discussion. For Sturgess's discussions of the author, see ibid.: 5 9 - 6 7 and 141. Ibid. Ibid.: 118. It has been suggested to me by John Pier that Don Quixote and Sancho Panza may have different ideas about narrative 'point' here, rather than narrativity itself. This offers a very neat and workable solution to the problem, as long as Sancho's discourse is indeed read as a narrative, rather than, say, a practical joke at Don Quixote's expense. In the latter case, the status o f its narrativity becomes less clear, as I suggest in the conclusion.
262
David Rudrum
There remains one last avenue to explore in trying to make sense of the narrativity of this extract in terms of the traditional theories of that concept, and that is to consider the suggestion that Prince, and occasionally Sturgess, make that narrativity lies in the eye of the beholder. Prince outlines this point as follows: The narrativity of a text depends upon the extent to which that text fulfills a receiver's desire by representing oriented temporal wholes, involving some sort of conflict, made up of discrete, specific and positive events, and meaningful in terms of a human project and a humanized universe.
Prince qualifies his position by saying that 'Events that are equally discrete, specific and positive do not necessarily yield the same degree of narrativity', and that 'an event which is individualized will contribute more to narrativity than one which is not'. 28 This might offer a clue that would account for Sancho's abortive narrative: the crossings and recrossings of the fisherman's boat, not to mention the number of goats he transports, are events that are not in the least individualised and that are scarcely specific, or positive, or even perhaps discrete. Furthermore, the part of the story involving conflict—between the shepherdess and the goatherd—is abrogated in favour of a number-counting narrative involving no conflict whatsoever. Hence, it plainly strikes Don Quixote (who, as the embedded narratee of this embedded narrative, stands in as a proxy or figure for the reader) as a pointless and irrelevant part of the story. If Sancho's rattling off the comings and goings of a seemingly endless stream of goats is lacking in narrativity, then that may well be because, as Prince has put it more recently, 'narrativity is said to depend on the extent to which a text involves hierarchical organization as opposed to a mere temporal concatenation of events (some of the latter should be of greater moment than others)'. 29 This is clearly applicable to Sancho's tally of goats. To summarise, 'events that are not viewed as relevant to the middle (or the beginning, or the end), events that cannot be analyzed as meaningfully related to the change presented are narratively inert, threaten narrative coherence and impair narrativity'. 30
27
28 29 30
Prince (1982: 160). This assertion obviously involves a questionable assumption—not shared by readers of, say, Beckett or Kafka—that receivers desire to read about discrete, specific, positive events involving some sort of conflict. Ibid.: 150, 149. Prince (1999: 49). Prince (1982: 154).
Narrativity and Performativity: From Cervantes to Star Trek
263
This is, once again, a persuasive account of the narrativity of the narrative in question. But it is open to a fairly obvious criticism: if narrativity is to be located in the reader of the text—and there are a multitude of readers, all historically and contextually situated—then it is hard to see how anyone could assert with any confidence which textual features exude narrativity and which do not. Surely this must be as mutable as the readers themselves. Sturgess makes a similar point: the role of the reader in detecting a narrative's logic of narrativity is crucial and this, when the reader is both pluralized and contextualized by his historical moment of reading, at once qualifies any idea o f a 'pure and essential meaning' intrinsic to the work. Different readers will identify and use different principles at different times, and will use them in different w a y s according to their particular interpretative emphases.
So, perhaps regrettably, it appears that appealing to the reader, and to certain textual properties that activate the reader's sense of narrativity, is not a move that can be made with any definitive certainty. It offers a clue—and a very useful one—but nothing more. 32 To recap, then, the theories of narrativity we have been discussing thus far have failed to provide a convincing account of the narrativity of the curious interrupted story that Sancho Panza tells to Don Quixote. If we follow the traditional suggestion that narrativity be understood in formal and structural terms, then the simplicity of the story should suggest it scores highly in terms of narrativity, when clearly it does not. Indeed, this approach even has trouble deciding whether the story can be deemed a narrative at all, let alone how much narrativity it possesses. If we adopt Sturgess's theory of the logic of narrativity, we end up describing the story in terms of a contradictory double logic, a viewpoint which Sturgess himself goes to great lengths to resist. And if we approach the question of narrativity from the viewpoint of the reader—in some ways the most persuasive of these approaches—then we are locating it outside of the text and seeking an explanation for Sancho's bizarre story, and Don Quixote's interruption of it, elsewhere. In sum, then, none of these traditional positions can adequately describe the narrativity of Sancho's tale, or the response it draws from Don Quixote.
31
Sturgess (1992: 53) Indeed, in his later work, Prince (1999) has pointed out the indeterminacy o f this orientation, and his contribution to this volume locates the reader/receiver in the realm o f 'narratability' rather than narrativity per se, no doubt in order to avoid this indeterminacy.
264
David Rudrum
3. Plurality and Performativity Perhaps the above analysis is simply asking too much of a theory of narrativity. Perhaps it is unreasonable to expect a theory to be able to account for what is, after all, an isolated fragment of a broader, more complex narrative whole, especially by laying down once and for all a simple recipe or formula that constitutes the nature of narrativity as such. Surely the beginning of wisdom, when dealing with narratives, is the recognition that there exist multiple narrativities, multiple forms, genres, and above all, conventions that pluralise the nature of the subject. As one of the earliest commentators on the subject put it: 'Whatever the constraints of a particular culture or a particular period, plurality is in the nature of narrativity'. 33 A theory of multiple narrativities is far more likely to provide more insights into more narratives than a theory that seeks to map out the concept as a generalised universal. Indeed, both Prince and Sturgess have advocated the importance of plurality in the notion of narrativity. In Prince's recent writings, he has stressed the multiplicity of the concept: Rather than (or along with) degrees of narrativity, one may distinguish between modes of narrativity (Ryan), such as the simple narrativity of fairy tales or urban legends [...], the complex narrativity of Balzac, Dickens, or Dumas [...], the figural narrativity of the lyric, historiographic, or philosophic texts [...], and the instrumental narrativity of sermons and debates.
To distinguish between these kinds of narrativity is a useful heuristic proposal, but it qualifies (even negates) the attempt to specify which textual features optimise or compromise narrativity, since these features must surely vary with, or even be specific to, the genre of the text under discussion. Consider the following three narratives: Tolstoy's War and Peace, Tchaikovsky's 1812 Overture, and a history text book on Napoleon's Russian campaign. All three of these 'texts' are narratives, and furthermore, all three narrate more or less the same set of historical events, but in different forms, genres, and media. It therefore makes sense to argue that each possesses its own kind of narrativity. But is there a narrativity according to which we could examine them all, let alone compare them? How could we ever determine, in any meaningful sense, whether Tchai33 34
K e r m o d e ( 1 9 8 3 : 111). Prince (2003: 65).
Narrativity and Performativity: From Cervantes to Star Trek
265
kovsky's musical narration of Napoleon's defeat contained more or less narrativity than a historical account of those events, or than Tolstoy's fictional handling of them? Would it even be meaningful to try? And if there is no common ground on which to compare the narrativity of Tchaikovsky's music and Tolstoy's fiction, then what does that suggest about the nature of narrativity? In other words, pluralising the concept of narrativity—a fundamental and necessary methodological move— fragments the unity of the concept, and accordingly complicates and pluralises any analysis that uses the concept. Sturgess's work repeatedly points out the existence of 'differing degrees of narrativity' as well as 'differing kinds of narrativity', 35 drawing attention to what he calls 'the most important point about our conceptual procedure, namely the fact that the concept itself functions according to different principles. A logic of narrativity is by no means monolithic or homogeneous in its mode of operation, or in the way we utilize it'. 36 Whilst agreeing that the point is indeed the most important one in our conceptual procedure, it should be pointed out that Sturgess's reasoning is flawed here, for he argues: 'Since every narrative will possess its own form of narrativity, then every narrative should be examinable in the light of this concept'. 37 But this is a non sequitur. if every narrative possesses its own form of narrativity, then it is by no means certain that there is a set of properties common to all these innumerable narrativities in light of which they may be examined. 38 The fact is that the notion of a plurality of narrativities simply does not square with the traditional 'Philosopher's Stone' notion that narrativity is the essence of narrative. How, then, to rethink the concept of narrativity? A good start would be to follow some underdeveloped hints in Prince and Sturgess. At one point, Sturgess describes 'a larger authorial, cultural, and indeed verisimilar context' which 'situates the logic of narrativity outside the text, in the logic of social life and formations—these may include literary formations—which influence or condition the way in which the text is produced.' However, he immediately qualifies this, stating: 'It should be noted that my emphasis here falls on the logic of narrativity being outside 35 36 37
38
See Sturgess (1992: 9, 15, 19 passim). Ibid.: 33. Ibid.: 28. For an alternative argument, suggesting instead a Wittgensteinian 'family resemblance' understanding o f narrative, see Rudrum (2006), as well as my unpublished Ph.D. thesis, 'Wittgenstein and the Theory of Narrative' (University of London, 2001).
266
David Rudrum
the text, since narrativity itself must always be thought of as a property of narratives'. 39 Prince's early work made a similar move: 'Rather than being definable in terms of the constituent features of a given narrative, the point of that narrative is a function of its context'. 40 It is interesting that both Prince and Sturgess appeal to context in this way, but it is even more interesting that they nevertheless want to keep narrativity itself independent of context: for Sturgess, only the logic of narrativity is contextual, while for Prince, context determines the point of a narrative rather than its narrativity. Indeed, Prince's essay in the present volume follows a similar trajectory: he affords a much greater role to context, but categorises most context-bound aspects of narrativity as 'narratability', thereby preserving the acontextual nature of narrativity proper. I want to go farther than this and claim that narrativity itself is indeed a function of context. Narrativity is firmly embedded in the linguistic, social, and cultural practices of the community. The job of deciding whether something is a narrative or not and whether it approximates a standard of narrativity or not is a job that is carried out by consensus and, above all, by convention. Conventions are what enable us to narrate to each other meaningfully in the first place. Here I am using the term 'convention' as R. M. Berry uses it when asking the question, 'What is a narrative convention?'. 41 That is, as a shared practice that is absolutely fundamental to any understanding of narrative in the first place. Admittedly, the term 'convention' is commonplace in traditional literary criticism, as Berry points out. Expressions like 'conventions of realism', the 'convention' of omniscient narration, 'conventional' character types and so forth frequently crop up in work on narrative, often with violent rhetoric. Such conventions are 'imposed on' texts, 'constraining' them, or else are 'violated', 'flouted', 'punctured', 'transgressed', or 'dismantled'. A struggle is taking place. Convention in this sense is something conformist and not a little conservative, so one strives against it more often than not. Convention in Berry's sense, however, involves the agreement needed to tell something in the first place. These are 'conventions without which we cannot even disagree meaningfully'. 42 A struggle here, though possible, does not merely toy with readerly expectations: it threatens narrative's ability to communicate. Struggles against these con39 40 41 42
Sturgess (1992: 53-54). Prince (1982: 159). See Berry (1995). Ibid.: 22.
Narrativity and Performativity: From Cervantes to Star Trek
267
ventions are therefore rare in everyday narrative, because they facilitate rather than constrain narratives. 'Such conventions are necessary if a practice or concept is to be possible at all. To depart from them is not to perform an act incompetently or innovatively, but to fail to perform it, and if they are "dismantled", the activity stops altogether'. 43 It is a comparatively easy thing, for example, to envision the conventions of the detective plot being parodied, flouted, spoofed, etc., but less easy to envision this with the conventions of narrative itself because as a rule, we have no recognisable alternative set of conventions: if a text flouted a community's conventions for narrative, it would, in the eyes of that community, lack narrativity altogether. A 'real life' example: Shirley Brice Heath's comparative study of language use, Ways With Words,44 centres on two rural, small town working class communities in the southeastern United States. The one, Trackton, is a black community, the other, Roadville, a white one. The findings of Heath's research into the narrative practices of these communities' children have been summarised as follows: In Roadville stories stick to the truth and are factual. They maintain strict chronicity, end with a summary statement or moral, and serve the function of maintaining values and reaffirming group membership. Any fictionalised account is a lie. Trackton stories, on the other hand, are hardly ever serious. The best stories are 'junk', and the best storytellers are those w h o can 'talk the best junk', i.e. make the most wildly exaggerated comparisons and tell outlandish fictional narratives.
What would happen if a member of one community tried to narrate to a member of the other? 'For Roadville, Trackton's stories would be lies; for Trackton, Roadville's stories would not even count as stories'. 46 The Trackton/Roadville scenario involves a signal instance of an absence of a common narrative convention: the Tracktonian and the Roadviller do not share many common narrative conventions at all, and so for each community, the narratives of the other community lack narrativity 43
44 45
Ibid.: 21. I am indebted to Toolan (2001) for bringing this work to my attention. Romaine (1985: 102-03). Heath (1983: 189). It has been suggested to me on several occasions that the difference between Tracktonian and Roadvillean stories is a difference o f 'point' or 'tellability' rather than o f narrativity itself. However, I find myself resisting this suggestion because of this statement of Heath's. If she is right, then not only are there two understandings of the conditions for narrativity in play here, but each contradicts the other. This in turn suggests there must be a very large degree o f plurality and mutability in the concept o f narrativity.
268
David Rudrum
and hence are not considered valid narratives. On this analysis, it can readily be seen that narrativity lies in a shared set of cultural, social, and linguistic conventions, and conversely, that an absence of such shared conventions impairs narrativity and leads to a breakdown in narrative communication. It is hard to see how two such communities could narrate to each other at any length without arriving at a fundamental misunderstanding, perhaps even a conflict. (On that note, it may be easier to envisage occasions when Roadvillers might have lynched Tracktonians for telling lies than to imagine them analytically comparing narrative conventions.) And yet these conflicts and misunderstandings—say, a Tracktonian's vain attempt to entertain a humourless Roadviller or a Roadviller's attempt to admonish a frivolous Tracktonian—are not simply reducible to banalities about different senses of humour or cultural relativity. Rather, these two kinds of narratives seek to do different things. This insight in turn offers a clue into how to develop the convention- and communitybased concept of narrativity: we can evaluate narrativity in terms of how well (or how badly) we succeed in 'doing things with narratives'. To suggest that our narratives seek to do things, and, indeed, that we seek to do things with our narratives, is to invoke terms used by J. L. Austin in describing what he calls the performative. It is this performativity, which is intimately wedded to the conventions that surround narrative, that determines how effective narratives are in their task as narratives, and thus their narrativity. Communities and their conventions are the deciding factors in determining whether a narrative is a narrative or not, and no doubt they may also play a role in determining whether it is an effective narrative or not, but the question of how effective the narrative is is a matter of whether or not it succeeds in its task as a narrative. That is, a successful narrative can be conceptualised as a successful performative. In suggesting this, I am effectively identifying narrativity with Austin's 'perlocutionary effects' and thus deliberately going to the opposite extreme of 'essentialist' theories of narrativity. This assertion has proven to be controversial in some quarters, 47 even allowing for a slightly looser sense of the term 'performative' than Austin sets out in How To Do Things With Words. But narratives carry many of the hallmarks of performative utterances. It is, for example, characteristic of performative utterances that rather than being described as true or false, they are either effective or ineffective. And much the same can be said of 47
See Ryan (2006) responding to Rudrum (2005).
Narrativity and Performativity: From Cervantes to Star Trek
269
narratives: it is rarely pertinent for a narratologist to investigate the truth or falsehood of a narrative, but much more pertinent to ask how that narrative actually works. In this loose sense if nothing more, it is meaningful to claim that narratives are indeed performative, and that their performativity is directly linked to their narrativity. And this point can be driven home by pointing out that, as with performative speech acts, it is more meaningful (to narratology, at least) to ask whether a narrative is effective or not than whether it is true or not.48 Using speech act theory to rethink narrativity is not a particularly new undertaking, for indeed it was first made by Mary Louise Pratt thirty years ago. Pratt takes Labov's analysis of 'natural narratives' as a starting point, stories 'whose relevance is tellability and whose point is to display experience'. 49 Here, 'point' and 'tellability' are tantamount to criteria for narrativity. Pratt argues, crucially, that tellability is first and foremost a use of language,50 and that narratives produce effects (an unsuccessful narrative elicits a response like 'So what?', whereas a successful narrative elicits a response along the lines of 'He did?'). To this extent, then, Pratt equates narrativity with performativity in a convincing manner. Yet it is possible—and, I maintain, useful—to pursue the relationship between narrativity and performativity further still. For Pratt confines the effect of performative narrativity, and hence of narrative performativity, to one specific act of reception: a text has narrativity if it successfully persuades its audience to contemplate, evaluate, or interpret it: the success of this perlocutionary effect is the hallmark of its narrativity.51 This is indeed one use we make of language, and indeed of narrative, but there are many others: narratives perform a variety of tasks in a variety of cultures according to different sets of conventions. The conventions of Labov's street narratives in Harlem, for instance, are different from the conventions of narrative in a church confession booth, where a priest's response is rarely 'So what?'. Both are clearly rule-bound narratives, with contextually and textually
48
There are, o f course, numerous exceptions to this: slander, libel, perjury, or even historical rewriting such as holocaust denial, are all cases where the truth or falsity o f a narrative is indeed highly pertinent. But these are special cases where a true/false criterion is one o f the conventions for narrative success, and failure to be truthful would be tantamount to what Austin calls 'misfires' and 'abuses'.
49
Pratt (1977: 152). Ibid.: 138. Ibid.: 136, 140.
50 51
270
David Rudrum
marked conventions that must be adhered to, yet each has a distinct and different set of perlocutionary effects. The principal weakness of Pratt's approach is that it curtails the full extent of performativity's role in narrativity. Surprisingly, Pratt considers narratives to belong to the category of 'world-describing' utterances, rather than 'world-changing' utterances. This is in keeping with traditional definitions of narrative, which emphasise the representation of events as the sole function of narrative. But this would surely suggest that narratives are simply constative instead of performative in the force they carry: 52 after all, Austin's preliminary isolation of the performative defines performatives as utterances that do things, where the things that are done with our words amount to something more than a 'world-describing' representation or statement. And narratives can indeed 'do things', they have multiple uses. They can, for example, entertain us, inform us, instruct us, scare us, seduce us, humiliate us, warn us, initiate us, exhort us, persuade us, testify for or against us, and so forth. Furthermore, we judge them as successes or failures based on the extent to which they perform these tasks effectively. Pratt, in common with many other narratologists, underestimates the extent of the uses to which we put narrative, what narratives can achieve. Narratives can and do have a world-changing force: for example, in the case of a boy publicly reading a (part of a) narrative at his Bar-Mitzvah, his narrating that narrative is part of what transforms him from boy to man. Narratives told in courtrooms are accepted as evidence, with clear practical consequences for (or against) the accused. Narratives are acts, many of them world-changing, and even if narratives normally succeed in their tasks, they can also fail: many of Labov's Harlem-style narratives would be ruled inadmissible evidence in court, owing to features such as their exaggeration—which is precisely what makes them effective as 'natural' narratives. That the very same narrative can have two such radically different forces is instructive, for it implies that the narrativity of a narrative is specific to a given set of conventions, and that this effectiveness can be graded according to those conventions. If we rethink the concept of narrativity along performative lines, we have in effect a tangible yardstick by which to assess how narratives measure up to their roles as narratives. Narrativity, then, could more profitably be conceived as a mat-
52
On this point, see Rudrum (2005).
Narrativity and Performativity: From Cervantes to Star Trek
271
ter of the relative success or failure of a narrative in carrying out the role assigned to it by conventions. A high degree of narrativity will be achieved when a narrative succeeds in performing a certain task or action, and a low degree of narrativity will typically be associated with a failure in performing that task. In both cases, the task/action performed is embedded in context, convention, and community. To simplify: a bedtime story will be considered a good bedtime story if it gets a child to sleep, while a ghost story will be considered a good ghost story if it scares that child into sleeplessness. Similarly, The Da Vinci Code (2003) may (or may not) succeed in keeping us turning the pages to find out what happens next, while La Nausee (1938) may (or may not) succeed in provoking philosophical reflection. The criterion by which we measure their narrativity is their success in performing the actions expected of them. The expectations and actions themselves, and hence the criteria that constitute success, are embedded in the conventions, contexts, and communities surrounding the narrative. These features have recently been viewed as distinct from narrativity by some narratologists, who have bracketed them off as 'narratability', 'tellability', or 'point'. But I find this distinction somewhat artificial, since it seems to me that convention, community, plurality, and performativity are in the nature of narrativity itself. To put the point simply: narrativity is generated at the point where textual features and contextual conventions are in harmony, when what a text 'is' accords with what it 'does'. Perhaps the obvious objection here is that if the criterion for narrativity is success, then how do we deal with narrative failure? To answer this, here is another example, this time from television. In 'Firstborn', an episode of Star Trek: The Next Generation,53 we encounter Klingon ritualised storytelling. During the festival of Kot'baval, Klingons celebrate the deeds of legendary warriors and heroes in public performances of ritualised narratives. Worf, a Klingon Starfleet officer, is concerned about his son Alexander's initiation into Klingon ways (he is being raised in the Starship Enterprise's very un-Klingon kindergarten) and so tries to interest him in the customs of Kot'baval. Alexander's response to these narratives, however, is anything but according to ritual. At first, he protests he has heard these stories before. To make them more interesting, he then
53
'Firstborn', Star Trek: The Next Generation, episode #173, written by Mark Kalbfeld and Rene Echevarria, directed by Jonathan West, first aired in the United States in 1994.
272
David Rudrum
suggests alternative happy endings involving reconciliation between rival warriors rather than the bloody vengeance that Klingon honour demands. No doubt his deeply unwarlike morals to these stories come from the stories he has encountered at kindergarten, and these stories are presumably read just once, explaining his obliviousness to (and boredom with) ritualised narrative. Wherever his narrative conventions come from, they are not those of his Klingon father. In this example, Worf tells stories to Alexander, but Alexander's response signals a breakdown in communication, since these stories do not meet his narrative conventions. To put it another way, there is a clear-cut set of conventions surrounding these ritualised stories to which Alexander fails to respond—hence, the narratives fail in their task as ritual narratives. For Alexander at least, their performativity collapses, and they lack narrativity. Now, according to certain traditional views, a text that is lacking in narrativity is not a narrative, and, furthermore, since the arbiter of narrativity is the receiver—Alexander, in this case—we would be forced to conclude, according to this view, that Worf s stories are simply not narratives (or not 'much o f a narrative), since, in the eyes of their receiver, they possess little or no narrativity. On the contrary, though, they are clearly recognisable as narratives, even to Alexander ('I've heard these stories before', he says). So a convention-based view of narrativity as performative is a good way of separating the definitional and the evaluative aspects of the concept of narrativity: Worf and his son share enough narrative conventions to recognise each others' stories as stories when they hear them, agreeing to a point on what constitutes narrativity itself. But beyond that point, Worf s ritualised narratives are intended to provoke a very specific kind of response in Alexander—i.e., to perform a certain task. They fail to perform this task, and so the stories Worf tells his son fail. Besides demonstrating the link between narrativity and performativity, these encounters at the interface of rival conventions interrogate traditional assumptions about narrative and its tasks, raising not simple questions about narrating events in different ways, but rather questioning what narratives are capable of doing. Relatedly, the second advantage of the view I am sketching here is that it enables us better to understand and describe both failures and successes of storytelling, by looking at the action of storytelling itself, and the conventions surrounding it. Depending on the conventions surrounding them, narratives can, like performatives, fall prey to what Austin called misfires, infelicities, abuses, and all manner of other glitches, but when they sue-
Narrativity and Performativity: From Cervantes to Star Trek
273
ceed, they can exert a kind of power that narratology has found difficult to describe (as in the case of, say, legal testimony, church confessions, incident reports on insurance claim forms, etc.)· Performativity extends the concept of narrativity by enabling it to describe effects of this kind— indeed, by incorporating these effects into the concept of narrativity. Additionally, perhaps the single biggest advantage accruing from a performative view of narrativity is that it makes more room for an understanding of plurality. Different communities use narratives in different ways, performing a variety of different tasks and actions, some of which may be specific to the conventions of that community. Don Quixote and Sancho Panza, Trackton and Roadville, or Worf and Alexander all have different narrative conventions—they expect their narratives to do very different things—and consequently they judge the success or failure of narratives very differently. Thinking narrativity in performative terms enables us better to understand the huge variety of different roles that narratives carry out in the cultural practices of our lives, and in doing so it gives us the means to describe their successes and failures.
4. Conclusion: Crossing the Guadiana River Let us return, by way of a conclusion, to the troublesome embedded narrative that Sancho Panza tells Don Quixote. Traditional theories of narrativity had some difficulty determining whether Sancho's story achieved a high or a low degree of narrativity, and even whether it constituted a narrative at all. It will be instructive to see how a performative theory of narrativity approaches this problematic story. We need to decide, I think, how to view Sancho's story as an action. Is he trying to tell a story to while away the time as he waits for daybreak, or is he playing a kind of practical joke on Don Quixote? We do not, of course, have to decide on one particular interpretation at the expense of the other, since narratives can perform a variety of tasks at once, but identifying the possible goals of the narrative qua speech act is a good start. If we go for the former, storytelling explanation, then there is good reason to suppose that, like Worf and Alexander or Trackton and Roadville, Don Quixote's preferred stories of knight errantry simply involve a different set of narrative conventions from Sancho Panza's. Don Quixote, growing bored with the story from the start, orders Sancho to: 'Tell it consequentially, like an intelligent man, or else be quiet.'
274
David Rudrum
'The way I'm telling it,' replied Sancho, 'is the way all stories are told in my country, and I don't know any other way o f telling it. It isn't fair for your worship to ask me to get new habits.'
This suggests a discrepancy between what conventions Don Quixote and Sancho consider to be a valid way of telling a narrative, which would in turn account for the communication breakdown between them, and the abortive end to the story: they simply expect different things from a narrative, or rather, they expect narratives to do different things. However, there is also a suggestion that Sancho is setting Don Quixote up here, and that the whole episode is a practical joke at his expense. Here is how Sancho introduces his tale: Ί will endeavour to tell you a story and, if I manage to tell it without interruption, it'll be the best story in the world. Pay attention, your worship, for I'm going to begin'. 55 If, as these hints suggest, Sancho is expecting his story to exceed Don Quixote's attention span and expecting Don Quixote to interrupt him, then the whole episode might be better understood not in terms of narrativity at all, but as a hoax or prank instead. The reader appears to be left with a choice: either Sancho's tale is a narrative that fails for want of a shared narrative convention, or else it is a 'hoax narrative', i.e. not a narrative at all, but a practical joke masquerading as a narrative. These two interpretations carry very different performative forces, and Sancho's explanation offers very little help here: Ί know that so far as my story goes there is nothing more to say, for it just ends where the error begins in counting the goats that cross over', 56 he says, unanswerably. Happily, though, we do not have to choose between these two interpretations. Narratives can do many things at once, and can do different things to different people. It is not a contradiction to suggest that Don Quixote and Sancho Panza deploy narrativity according to different sets of narrative conventions, and it can only deepen our understanding of the text if we evaluate the success of this narrative according to these multiple sets of conventions: for Don Quixote, the story fails as a story, and lacks narrativity, whereas for Sancho, the story is nothing but a joke—it has a kind of mock- or hoax-narrativity. It is a story in the same way that the child who says Ί bet you a trillion dollars...' is making a bet, i.e. using conventional formulas to a similar, but not entirely serious, 54
56
Cervantes (1950: 153). Ibid.: 152 (my emphasis). Ibid.: 155.
Narrativity and Performativity: From Cervantes to Star Trek
275
tional formulae to a similar, but not entirely serious, effect. Finally, for the reader of the novel as a whole, these two narrativities come together in that this episode is the narrative of a clash between them: the story of this abortive story achieves the kind of narrativity we associate with amusing anecdotes, funny stories, and jokes of this kind, but it probably lacks the kind of narrativity we associate with reading for the plot and for 'what happened next'. To define narrativity in terms of conventions, and to evaluate narrativity in terms of performativity, is a complicated undertaking, enabling one to read and interpret narrative success and failure at any number of different levels. There is not room to explore the full implications of this move here. As a methodological departure, though, it avoids the vagueness of locating narrativity in the eye of the beholder and the prescriptive narrowness of locating it solely in intrinsic textual or formal features. Rethinking narrativity as an action successfully achieved (or not), in terms of a set of conventions shared and defined by a community, relocates this difficult and troublesome concept to the context from which narratives originate in the first place, shifting the emphasis from structure, form, or logic, to the more pragmatic question of use. It could well be that putting narrativity and performativity in the same frame together is a move that will raise as many questions as it answers, for both speech act theorists and narratologists alike. But where these questions take us, I think, is beyond narratology and into the philosophy of language. And that in turn suggests it is a move well worth the making.
References Austin, J. L. 1975 How To Do Things With Words. 2nd ed. (1st ed. 1962), edited by J. O. Urmson and Marina Sbisä. (Oxford: Oxford UP). Berry, R. M. 1995 "What is a Narrative Convention?", in Narrative 3: 18-32. Cascardi, Anthony J. 2003 "The Grammar of Telling: The Example of Don Quixote," in Ordinary Language Criticism: Literary Thinking after Cavell after Wittgenstein, edited by Kenneth Dauber and Walter Jost, 135-48 (Evanston: Northwestern UP). Cervantes, Miguel de 1950 [1605/15] The Adventures of Don Quixote, translated by J. M. Cohen (Harmondsworth: Penguin Classics).
276
David Rudrum
Heath, Shirley Brice 1983 Ways With Words: Language, Life, and Work in Communities and Classrooms (Cambridge: University of Cambridge Press). Kermode, Frank 1983 Essays on Fiction, 1971-82 (London: Routledge). Pratt, Mary Louise 1977 Toward a Speech Act Theory of Literary Discourse (Bloomington: Indiana UP). Prince, Gerald 1982 Narratology: The Form and Functioning of Narrative. Janua linguarum. Series maior 108 (Berlin, etc.: Mouton). 1999 "Revisiting Narrativity," in Transcending Boundaries: Narratology in Context, edited by Walter Grünzweig and Andreas Solbach, 43-51 (Tubingen: Gunter Narr Verlag). 2003 Dictionary of Narratology. Revised ed. (1st ed. 1987) (Lincoln: University of Nebraska Press). Romaine, Suzanne 1985 "Grammar and Style in Children's Narratives," in Linguistics 23: 83-104. Rudrum, David 2005 "From Narrative Representation to Narrative Use: Toward the Limits of Definition," in Narrative 13: 195-204. 2006 "On the Very Idea of a Definition of Narrative," in Narrative 14: 197-204. Rutland, Barry 1997 Gender and Narrativity (Ontario: Carleton UP). Ryan, Marie-Laure 2006 "Semantics, Pragmatics, and Narrativity: A Response to David Rudrum," in Narrative 14: 188-96. Sturgess, Philip J. M. 1992 Narrativity: Theory and Practice (Oxford: Clarendon Press). Toolan, Michael 2001 Narrative: A Critical Linguistic Introduction. 2nd ed. (1st ed. 1988) (London: Routledge).
JUKKA TYRKKÖ (Helsinki)
'Kaleidoscope' Narratives and the Act of Reading A good book may have three openings entirely dissimilar and interrelated only in the prescience o f the author, or for that matter one hundred times as many endings. Flann O'Brien, At
Swim-Two-Birds
(1939)
1. Introduction How to approach a narrative with no beginning, no end, and no way of knowing how much of it remains to be read? 1 How does our experience of reading change if we not only read a narrative, but in doing so contribute to its formation? These questions, which at the time Flann O'Brien wrote At Swim-Two-Birds would have been little more than exercises for the imagination, have become acutely important over the past twenty years as electronic hypertexts have become a medium for fiction writing. Hypertextual narratives, composed of textual fragments and interconnecting links, immerse readers into the process of reading by asking them to navigate alternative paths through the text. By turning the passive experience of reception into an activity and a part of the storytelling, fragmented texts place novel demands on narrativity in the constitution of narrative. With these points in mind, I intend to look at two issues. First, I will briefly examine the history and common characteristics of both printed
1
Work for this article was supported by the Academy o f Finland Centre of Excellence funding for the Research Unit for Variation, Contacts and Change in English at the Department o f English, University of Helsinki.
278
Jukka Tyrkkö
and electronic fragmented narratives and argue that hypertextual narratives, rather than being a new phenomenon separate from previous forms of literature, follow in a stylistic continuum. 2 Second, I will discuss the participatory nature of reading inherent to these kaleidoscope narratives and argue that the experience of active participation has a profound effect on the concept of narrative. I will also elaborate on J. Hillis Miller's statement that "hypertext brings into the open the way the generation of meaning in the act of reading is a speech act, not a passive cognitive reception." 3 The narrative effects of the experience of participatory reading will be demonstrated by examples from several fragmented narratives. Although many of the concepts are applicable to various other forms of narration (films, comic books, computer games) 4 and to narratives developed collaboratively by groups of authors, the discussion here is restricted to self-contained fictional narratives by a single author.
2. Kaleidoscope Narratives: From Episodic Structure to Hypertexts Kaleidoscope narrative5 is a term by which I wish to draw attention to the way fragmentary narratives—both printed and electronic—make use of the possibilities of multilinearity, the allowing of readers to actively affect sequences of textual passages in order to create practically endless new combinations from a limited number of elements. The fragments6 dis-
2
3
My discussion will reiterate some of the points brought up by Espen Aarseth, MarieLaure Ryan and other theorists. However, unlike Aarseth (1997: 85), who suggests the concept of ergodicity as an alternative to narrativity, I consider works o f hypertextual fiction as narratives. Miller (1995: 38). Many o f these fall under the concept of distributed narratives studied by Jill Walker. Her ongoing study can best be followed on her website at http://jilltxt.net/txt/distributednarratives.htnil and http://jilltxt.net/?p=931, both available online 7 th November, 2005. The analogy between hypertext and a kaleidoscope has been previously made by Zhang (1998: 128) and Ryan (2001: 219). I will use the term segment when discussing similar chunks o f text in a linear text, characterized by the lack of functional and explicitly marked links. Barthes' (1974 [1970]) concept o f lexia is close to my use o f segment. Landow (1992: 25) introduced the term into hypertext theory as a name for hypertextual fragments, but this usage seems problematic to me, considering the fact that Barthes' lexia are defined by the reader and are specific to a particular interpretation of the text. The hypertextual frag-
'Kaleidoscope' Narratives and the Act of Reading
279
cussed here are neither unfinished nor incomplete texts; rather, they are the short, internally coherent building blocks of a multilinear narrative which typically narrate a single event and are connected to other fragments by links, that is to say, words or word groups designated for this purpose by the narrator and described by Douglas as "decision-points." 7 Links serve the narrative both functionally as interactive devices between the reader and the text and conceptually as narrative devices which imply something about the nature or motivation of connection between two fragments. In this latter regard, both hyperlinks and links in print narratives bear a close resemblance to key words in linear texts 8 by having the power to draw attention to certain issues and topics, regardless of whether or not the link is chosen by the reader. My starting point in discussing fragmentary narratives is to examine briefly the developments that paved the way for fragmented texts by facilitating ever-increasing segmentation and shifts in perspective. 9 The first inklings of narrative fragmentation were seen in the innovatory works of authors like Cervantes and Sterne, both of whom engaged readers by means of metanarratives giving instructions to skip certain "irrelevant" chapters or to reread some previous ones. While these narratives were segmented rather than truly fragmentary, the innovative devices employed in these texts challenged readers by offering them opportunities for participation and thereby disturbing the leisurely reception of the text. The major shift in narrative fiction preceding and contributing to the emergence of true fragmentation as a narrative technique began with Romanticism, expanded with late nineteenth-century Realism and came into full bloom in twentieth-century Postmodernism as the literary community, and society at large, started growing increasingly appreciative of multivo-
ment is defined by the narrator through the organizing o f the text and is consequently as much a structural unit as it is a narrative one. 7 8
Douglas (2001: 54). Cf. Toolan (2004: 226). I will discuss the differences between print and electronic literature only as they pertain to connections between text fragments and to questions o f linearity and multilinearity. Questions related to other differences, ranging from textual stability to the experiential differences between reading from flatscreen monitors and codices, will be left aside. For discussion on these issues, see Hayles (2003), Miall (1998), and in particular Miller (1995). I will also not discuss the narrative aspects o f computer games, M U D s , and various forms o f computer mediated communication. For discussion, see Aarseth (1997), (1999); Douglas (2001); Ryan (2005b).
280
Jukka Tyrkkö
cality and alternative points of view. As authors discarded and replaced the patriarchal heroes 10 and picaros of established literature and started giving voice to women, children, ethnic and finally to other minorities, earlier narratives were rewritten from new points of view which were more reflective of new sensitivities. The decreasing reverence for the author also allowed a new approach to allusion. Moreover, theories of intertextuality have enabled us to see what was earlier regarded as "sources," "imitation" and the like in a significantly new light. J. M. Coetzee gave a voice to a woman unjustly omitted from Defoe's story of Robinson Crusoe in Foe (1986); Don DeLillo delved into the minds of the principal players of the Kennedy murder in Libra (1989); and Kathy Acker made Don Quixote a donna in Don Quixote: Which Was a Dream (1986). 11 These techniques of narrative fiction became increasingly evident on the level of textual structure as well: perspectives were not simply new and different, they were increasingly juxtaposed within the same narrative. The multiple voices in T. S. Eliot's The Waste Land (1922), the stream of consciousness episodes in James Joyce's Ulysses (1922), and later the short stories of Jorge Luis Borges all inspired other authors to experiment with new techniques. Borges in particular, concerned with the relationship between chaos and order throughout much of his writing,12 is
10
There were, o f course, early exceptions like Daniel Defoe's Moll Flanders (1722) and Samuel Richardson's Pamela (1740). To give an even earlier example, one might argue that although Shakespearean female characters (e.g. Beatrice in Much Ado About Nothing [1598]) are not main protagonists, they often claim center stage. Nevertheless, women—and certainly children and various racial and sexual minorities—had a marginal role in earlier literature and, sadly, even today. For examples o f hypertextual transformations in literature, see Gerard Genette's Palimpsests (1997 [1982]). Note that Genette uses the term hypertext without reference to electronic literature; in Palimpsests, the term refers to a rewriting o f a previous literary text, as Henry Fielding's Shamela (1741) was of Richardson's Pamela (1740). The short story "The Book of Sand" (1977) is an illustrative example. Frequently cited by hypertext theorists (and recently transformed into a hypertext by Maximus Clarke), this story about an endless book is linear in structure, but alludes to fragmentation and the uncontrollability of the infinite text. Borges' frequent references to labyrinths, dictionaries and rearrangements o f organization have made his works particularly attractive for hypertext enthusiasts. Stuart Moulthrop's Victory Garden, one of the bestknown hypertext novels to date, is a hypertextual homage to Borges' "The Garden of Forking Paths" (1941). Marc Saporta's Composition No. 1 is another example of a
'Kaleidoscope' Narratives and the Act of Reading
281
frequently cited by modern hypertext authors as a predecessor and an inspiration. Flann O'Brien's At Swim-Two-Birds13 (already mentioned in the introduction) impresses the modern reader as a precursor to modern hypertexts, despite its sequential organization. The story is a mixture of overlapping frames of all description: biographical notes, letters of correspondence between the narrator and his friends, an array of extracts from the three separate stories imagined and perhaps written by the narrator, comments in dictionary style offering sarcastic explanations for various things observed by the narrator, and so on. Throughout the twentieth century, similar structures were explored by authors with the result that narratives became increasingly challenging to follow as voices blended, merged and multiplied. From Doris Lessing's The Golden Notebook (1962) and Vladimir Nabokov's Pale Fire (1962) to Joseph McElroy's Women and Men (1987) and Ben Okri's The Famished Road (1991), postmodern works have tested the limits of structure and organization in search of new ways of telling stories. While the majority of postmodern texts were—and continue to be—linear, they have grown more and more segmented. A small number of truly fragmented print texts, such as Marc Saporta's Composition No. 1 (1961) and Julio Cortäzar's Hopscotch (1963), 14 have also appeared over the years. Hypertextual fiction started emerging in the late 1980s. Authors of electronic hyperfiction, most of whom were new media enthusiasts, took guidelines and inspiration from the visions of early pioneers like Vannevar Bush and Ted Nelson 15 and began experimenting with the possibility of using links to connect parts of a narrative in multiple ways instead of just one. Almost from the very beginning, enthusiastic voices started
14
fragmented print text repeatedly discussed by hypertext theorists (e.g. Douglas [2001]; Aarseth [1997]; Bolter [1991]). The 1998 edition (Dalkey Archive Press) comes with an introduction by William H. Gass, who makes frequent use of highly segmented narration in his own work: In the Heart of the Heart of the Country (1968); Willie Master's Lonesome Wife (1968); etc. Composition No. 1 is a fragmentary novel, published as a stack of unnumbered sheets of paper. Hopscotch, although in traditional codex form, nevertheless offers alternative orders of reading. The publication of Vannevar Bush's visionary article "As We May Think" in Atlantic Monthly in 1945 is widely held to have been the seminal moment in the history of hypertext. Theodore "Ted" Nelson is one of the pioneers of modern electronic hypertext.
282
Jukka Tyrkkö
claiming that hypertext was the advent of a new era in textuality 16 and that the possibility of readerly participation would finally make the deconstructionist ideals of Barthes' writerly text both tangible and real. However, as discussed by critical scholars like Espen Aarseth (1997), David Miall (1998), and Jay David Bolter (1991), most if not all of the major thematic and stylistic features found in hypertexts have antecedents in the ergodic17 works of earlier authors. Described by Douglas as mosaic narratives,18 the fragmentation and multiplicity of perspectives found in the works of writers like Julio Cortazar and Lawrence Durrell foreshadow many of the techniques we now find in electronic hypertexts and experiential print texts. Robert Coover, perhaps the first prominent literary figure to have commented favorably on hypertextual literature,19 has similarly discussed the debt hypertext authors owe to innovators like Cervantes, Sterne, Joyce, and Calvino. All in all, the continuum extending across media from the birth of the novel to modern hypertexts shows that narrative fragmentation is not a phenomenon separate from literary tradition, but instead follows as an almost predictable step in the development of both media and style. Which is not to say that there is nothing new about
16
18
19
More than one theorist and critic has referred to hypertexts as the true coming of the writerly or open text discussed by deconstructionists and semioticians like Barthes, Derrida, Kristeva and Eco. Space does not permit a detailed discussion here, but a problem common to many of these paradigms is that they typically confuse form with content and text with interpretation. Aarseth's (1997: 76-86) brief discussion of this topic is particularly valuable. In an effort to emphasize the similarities between these interactive texts, Aarseth (1997: 2) coined the term ergodic literature to describe literary works which require "nontrivial effort to allow the reader to traverse the text." While the term kaleidoscope narrative refers to essentially the same genre, I focus on the effect that immersive readerly experience has on narrativity; Aarseth discusses ergodicity as an alternative to it. Douglas (2001: 55-62).
Coover's two articles on hypertext literature, published in the New York Times Book Review (in 1992 and in 1993), are considered by many to have been the introduction of hypertext to the wider literary community. Coover's own works, from Spanking the Maid (1981) to The Adventures of Lucky Pierre (2002), although linear in structure, often make use of radical segmentation and unreliable causality. His most segmented works, short stories such as "Babysitter" and "The Magic Poker" from the collection Pricksongs and Descants (1984), play with retelling and repetition as allusions to the metatextual dimensions of writing and reading. Coover is involved with Hypertext Hotel, a collaborative hyperfiction project hosted by Brown University.
'Kaleidoscope' Narratives and the Act of Reading
283
them. Quite the contrary, some of the narrative effects achieved through hypertextual interaction make such texts highly distinct from the majority of linear print narratives and help define hypertext as a literary genre.20 3. Linear and Multilinear Narratives One of the fundamental theoretical paradigms of narratology has been the notion of examining narrative as a double-layered system, consisting of a sequential arrangement of events as they are presented in the text (sjuzet) and of the causal and chronological order of those events in the text world (fabula).21 Reading a narrative transports readers into this imaginary world which is, unless otherwise indicated or implied, similar to the real world in terms of space, time, the laws of physics, and so on.22 The idea of an underlying world of events is reflected in the way narration is usually organized into episodes. Episodic structure is a standard feature of narrative texts, and one might assume that it is well covered by narrative theories. However, as it has occasionally been pointed out, most narrative theorists have focused on questions such as point of view and theme at the expense of investigating how the episodic structure is in fact perceived by readers.23 A particularly important contribution to remedying this situation
20
Douglas (2001: 37-62) provides a comprehensive list of the various ways in which interactive narratives tend to be different from print narratives. These differences include lack of singular beginnings and ends, reading through active choices, networks of narrative segments, multiple coherent readings, and less determinate use of language. The terminology derives from Russian Formalism. Though not exactly equivalent, roughly similar concepts are described by French Structuralists with the terms recit and histoire and in English language narratology by the terms discourse and story. Cf. Gerrig (1993). Naturally, both different genres and individual works within genres exhibit varying degrees of such verisimilitude. Very generally, realistic stories tend to differ from the 'real' world the least, while fantastical and magical stories may be set in alien worlds and can even distort the laws of physics. E.g. Miall (2004: 111). Miall (ibid.: 112) defines episode as a "number of sentences, usually demarcated by a coherence in the temporal or spatial setting or both. The most signal feature of the episode is that it offers a thematically distinctive topic requiring a shift in the reader's understanding." Episodes are defined by Bolter (1991: 122) as one of the two required elements of electronic writing (along with links). He also makes the point that reading the episodes themselves is an act of conventional reading and that it is the selection of a new episode, by means of link selection, which adds something new to reading as an activity.
284
Jukka Tyrkkö
is Catherine Emmott's (1997) use of cognitive frames in modeling readerly processing of narrative texts, an area which she (like Miall) considers sadly lacking in research. Emmott's term contextual frame denotes a "mental store of information about the current context" that carries "facts which are 'episodic' within the fictional world." 24 Frames are established, modified, stored in the memory and recalled as indicated by the narrative. Transition between episodes requires cognitive effort and is contingent upon a number of cognitive and cultural factors which affect the success of negotiating coherence. 25 If, as I would claim, the search for coherence is one of the things utmost in a reader's mind (we do in fact read in order to understand), it is reasonable to assume that the processes by which comprehension is achieved are also of great importance to the experience of reading. Peter Brooks refers to this experience when talking about the "anticipation of retrospection": one of the motives we have when reading is the enjoyment of knowing that once we get to the end we will understand "the larger picture." 26 In order to achieve this goal, we need to be able to process episodic structures by recognizing episode boundaries, establishing new contextual frames as required and juggling numerous settings, all the while keeping track of topical shifts and other plot features. Our experience of reading is affected by how readily we comprehend the textual structure and find the text coherent. Crucial to this process is the concept of gaps, discussed by Wolfgang Iser as spaces of indeterminacy "formed and modified by the imbalance inherent in dyadic relationships." 27 Of particular importance to the discussion of kaleidoscopic
24
Emmott (1997: 121) See also Zwaan/Radvansky (1998). For discussion on some o f the text linguistic and discourse analytical aspects of coherence negotiation in hypertextual linking, see Tyrkkö ( 2 0 0 4 ) . Brooks (1984: 23). The related issue of closure is another hot issue in hypertext literature. Hypertextual structure makes it possible to write stories which do not end in the traditional sense. See Douglas (1994) and Miles (2002). Iser (1984: 167). It should be noted that coherence can be discussed on two levels: as experienced momentarily during reading and retrospectively. A narrative may seem incoherent during reading, but such temporary incoherence is expected to clear up in the end, i.e. any incoherence is expected to have been narratively motivated. The text may thus become coherent retrospectively if the reader is able to generate an explanation for the momentary incoherences experienced during reading. Of course, this may sometimes involve having to backtrack on previous explanations, or, in Iser's terms, alter one's previous projections upon the text.
'Kaleidoscope' Narratives and the Act of Reading
285
narration is Iser's view of the cooperation needed between participants to interaction: while all texts leave some things to the reader to fill in and complete, interactive failure occurs "if the reader's projections superimpose themselves unimpeded upon the text." 28 One similarity between the different theoretical models of narrative, so obvious it has rarely needed to be emphasized until recently, is the fact that narration is by default expected to be unilinear. 29 While events may be recounted chronologically, in order of causation, by spatial arrangement or in practically any other order imaginable, in the end there is one and only one sequence in which they are presented to the reader. In unilinear episodic texts, the order in which events and episodes are encountered is not subject to reorganization by readerly interaction, but only through interpretation. 30 As the episodes unfold, the reader processes and interprets them on the basis of his or her background, biases, interests, and moods. The reader's wandering viewpoint, as Iser calls it, guarantees an endless multitude of readings and experiences, as the reader's focus changes and the same text is reinterpreted again and again. 31 By contrast to conventional episodic texts, kaleidoscope narratives are structurally multilinear. Instead of being read as a predetermined linear plot, a kaleidoscope narrative offers alternative paths of access to events or episodes, leaving the construction of the plot up to the choices of the
29
30
31
Unilinearity is a term frequently used in hypertext theory in reference to conventional sequentially read texts. The implication is that the parts of the text are always read in the same order. Naturally, readers of such texts can—and often do—read the texts out of sequence, particularly after the first reading. Interaction refers here to an activity performed by the reader on an interactive text which affects the organization o f the text as read, e.g. choosing to follow a particular hypertextual link or skipping to a particular section o f the text as prompted by an encyclopedic reference. A s a result, sections of the text are read in an order consequent to the choices made. For discussion of interactivity and digital texts, see Ryan (2005a). Interpretation, by contrast, refers to mental processing o f the text. Iser (1984: 108-09). A s discussed by Eco in The Limits of Interpretation (1990: 3 6 37), textual interpretation is not a free-for-all where any interpretation whatsoever would be as reasonable as another. Arguing against "the cancer o f uncontrolled interpretation," Eco claims that interpretation must begin with a zero-degree meaning agreed upon by "every member o f a community of healthy native speakers" (ibid.: 36). Similarly, while competent readers may disagree about the preferred interpretation of a text, they will generally agree upon which interpretations are reasonable.
286
Jukka Tyrkkö
reader. 32 While the number and content of the actual textual fragments remain the same, the narratives constructed can be quite different from one another. 33 The double logic of organizing principles discussed by Jonathan Culler (1999 [1980]) is subject to another twist: do we see each of the alternative sjuzets as describing different points-of-view in the same underlying fabula, or does each sjuzet describe a new fabulal In a sequential print narrative, multiple points of view are still incorporated into the same narrative, and thus it may be argued that the fabula is a constant. In kaleidoscope narratives, by contrast, each reading experience necessarily anticipates and recalls altogether different readings of the same text. These alternative sjuzets resemble each other, but instead of being different points of view within the narrative they belong to different readings. Are they, then, descriptive and narrative of the same or of different fabulasl Kaleidoscope narratives would appear to lend themselves well to Culler's suggestion of identifying fabula "not as the reality reported by discourse but as its product." 34 The double logic is intact, but does multilinear interaction perhaps in some way tip the scales in favor of the primacy of the sjuzet? 4. Multilinear Literacy: Reading Links—and Loving it The transition from segmented to fragmented narration takes place at the point where the reader's active participation becomes an issue. The principle of searching for coherence and purpose is still present, but now the task of forming the narrative clearly lies with the reader. The ergodic
32
34
"Choices" is intentionally in the plural. The reader does not simply choose a plot, but makes a continuous string of choices which constantly shape the direction of the narration. In most kaleidoscope narratives, the reader has only partial foresight into the effect his or her choices will have on the narrative. This was demonstrated in an experiment conducted by Douglas (2001: 73-77). A group of her students read Moulthrop's Victory Garden and then gave feedback on the reading experience. The results showed that the majority read for the structure instead of the plot, "struggling to establish where the particular places they read belonged inside the framework of the hypertext." (ibid.: 77). Aarseth (1997: 87) rightly sees this kind of exploratory reading as potentially detrimental to the poetic and narrative elements of the text; however, this does not mean that readerly exploration cannot be successfully incorporated as an element of the narration, as discussed in section five. Culler (1999 [1980]: 95).
'Kaleidoscope' Narratives and the Act of Reading
287
process of interaction with fragmentary textual space heightens readerly awareness of textual contours and of potential relations between events.35 While the world depicted by the text still remains beyond the immediate reality of the reader, the act of narrating now becomes—or so it is experienced—something to be negotiated between the author and the reader. The resemblance of this communicative situation to that of oral storytelling is quite striking: while the reader in actual fact interacts with the narrative through only carefully pre-assigned links, the effect is similar to the interaction between a storyteller and his listener, each link representing an extra-textual "wink" to the reader: should I tell you more about this?36 To construct a coherent text world, the reader needs to explore fragments as if asking questions from the narrator. In this search for connections, previously read passages often end up being reread or attached to several narrative lines. Individual fragments may remain unconnected until the right one suddenly comes along, explaining the causalities involved. This dynamic of aporias and epiphanies, of getting lost and suddenly finding the way again, is fundamental to kaleidoscope narratives.37 The first difference between reading linear and multilinear texts is thus the way in which the reader actively participates in creating the plot.38 By
35
36
37
Joyce (1995) uses the term contour to describe the way readers experience hypertextual structures on the basis of the level of connectedness between fragments and the role each of them plays in the overall textual space. See Rosenberg (1996) for a discussion of the relationship between text fragments, links and episodes. Cf. Bolter (1991: 59, 107-19). Toolan (2004: 224-25) discusses a similar topic in linear narratives. He suggests that every proposition in the text can "prompt one or more of the standard array of Who, What, When, Where, Why, How questions" and that "every predicate used in the composition of a narrative carries with it a license for the teller to say more, and for the reader to find more." In multilinear texts, linking not only endorses certain key words (as Toolan suggests happens with all narratives), but also increases the functionality of associated readerly questions that call for an answer. I have discussed the pragmatics of hypertexts from the perspective of question-answer models and the Gricean cooperative principle in Tyrkkö (2006). Aarseth (1997: 90-92). Douglas (1993: 8) points out that in larger hypertexts, the complexity of the network of fragments means that readers are likely to come up with readings which the author could never have imagined. Eco's (1989: 19) description of the "open work" appears to encompass this possibility: "at the end of the interpretative dialogue, [the work] may have been assembled by an outside party in a particular way that [the author] could not have foreseen." However, Eco's argument that the resulting work still belongs to the author is based on the premise that the interpretative possibilities "had already been ra-
288
Jukka Tyrkkö
making the choice to follow one link or another, the reader commits himself to pursuing one particular path in preference to others. Although readers may have tentative expectations, triggered by the word or words of the hyperlink or keywords in a print text, regarding the narrative outcomes of the choices available, each act of selecting a link is followed by its own tiny denouement when the new fragment reveals the true nature of the connection. The wandering viewpoint is thus transformed into the wandering viewer, someone who not only focuses his or her interpretive faculties on one particular part of the narrative at a time, but in fact through active choice directs the physical reality of the narration itself. As Harpold points out, one of the realities of hypertextual reading is that "at any moment, you might (not) know where you are going." 39 A previously established contextual frame may be activated or a new one created. The reader will then have to process the new fragment into the spatio-temporal continuum of the narrative and fill in any gaps in an attempt to fulfill the teleological demands of purposefulness and meaning. Much evidence suggests that this is not an easy task. Studies conducted by Douglas (1994) and Miall/Dobson (2001) have demonstrated the difficulties experienced by readers when encountering multilinear stories which refuse to entertain passive readers and instead require constant participation and heightened powers of inference-making. Indeed, more than a few arguments have been leveled at claims made by advocates of hyperfiction that hypertextual (i.e. multilinear) reading is more intense or immersive than the reading of linear print narratives. Critics often point out that having constantly to step away from the text to evaluate both the possible narrative directions suggested by links and the actual directions emerging after links is disorienting and, in short, anything but immersive. 40 The argument goes that hypertext reduces narration to the individual moment or the "perpetual present," 41 and in doing so undermines the coherence of longer stretches of discourse and consequently the experi-
tionally organized, oriented and endowed with specifications for proper development." Since kaleidoscope narratives offer not only alternative interpretations, but also alternative sequences of narrative fragments, it can be argued that some readings have a potential for creating narratives which the author did not intentionally organize or anticipate. 39 40 41
Harpold (1991: 134). E.g. Aarseth (1994: 6 7 - 7 1 ) . Tuman (1992: 118).
'Kaleidoscope' Narratives and the Act of Reading
289
ence of deep textual meaning. 42 According to these critics, hypertexts are neither more liberating nor more thought-provoking than linear texts: hypertexts are seen as limiting readerly interpretation by force-feeding readers with predetermined meanings, instead of allowing them the freedom to interpret the text as they will. I shall attempt to address these points one by one. To begin with the claim that hypertextual reading is interruptive and thus overly focused on the present moment, the problem as I see it has more to do with the difficulties of adjusting to a new medium than with problems inherent to it. If the reader encounters every link as an intrusive obstacle to the flow of the text, it is to be expected that reading may become labored and disjointed. 43 However, once the reader manages to get past the initial sense of unfamiliarity (which may well be unpleasant at first), linking can become an immersive experience in which the act of reading becomes a part of the narration. Kaleidoscopic narrativity relies on the reader's willingness to participate and be immersed in the search for reasons behind connections and linkings. Hayles remarks that electronic textuality has made it "inescapably clear that navigational functionalities are not merely a way to access the work but part of the work's signifying structure." 44 Fluent literacy is marked by the apparent transparency of the "mechanical" act of reading and the consequent ability to focus on what is read, rather than on how the reading itself is accomplished. In much the same way, hypertextual fluency means that a reader is able to incorporate transitions between fragments into the act of reading—instead of seeing the one as disrupting the other. The skill of maintaining a sense of immersion in the text while simultaneously processing transitions between fragments and contemplating the textual significances of linkings is not acquired overnight. The concept of hypertextual literacy may inform some aspects of the discussion of narrativity as well. It has been suggested, for example, that hypertexts like Joyce's afternoon, a story (1987) are not really narratives at all, although such texts "willingly generate
42
44
Charney (1993). That the difficulty of reading hypertextual narratives should become an issue is curious, considering the difficulty involved with reading many experimental works of fiction. Perhaps the question concerns the fact that the difficulties are of a new kind, rather than the degree of effort required of the reader. Hayles (2003: 264).
290
Jukka Tyrkkö
narratives when experienced." 45 Putting aside the somewhat contrived separation between narrative and "narrative when experienced," I would claim that finding the narrative is a crucial part of hypertextual literacy: the reader must actively look for and construct narratives out of episodes, just as he constructs episodes out of fragments. 46 As Harpold has argued, reading a hypertext "is guided by a determination to make all the chance encounters of the reading meaningful" 47 —a meaningfulness, I would add, generated in part out of the indeterminacy encountered at each hyperlink. If, as the traditional view suggests, narrativity requires causal connections between events, then hypertexts do indeed tend to defer the emergence of narrative, thereby reducing the degree of narrativity. As Toolan observes: "perceiving non-random connectedness in a sequence of events is the prerogative of the addressee: it is idle for anyone else (e.g. the teller) to insist that here is a narrative if the addressee just doesn't see it as one." 48 On the other hand, Ryan's definition of narrativity as the quality of possessing "narrative potential" seems particularly suitable to kaleidoscope narratives: their narrativity is often almost entirely contingent upon the act of reading which embraces that potential and turns it into a narrative. 49 On this basis, readers of hypertexts appear to be (and need to be) willing to give the text the benefit of the doubt and to suspend their need for coherence until enough fragments have been read for them to start coming together to form narrative sequences. Hypertextual links thus function not only as verbal connectors and as functional sites of interaction, but also as de facto narrative connections between fragments of the text. They involve narrative surprise, i.e. new developments which are "unforeseen but, upon reflection, foreseeable." 50 Whether a reader understands the narrative significance of a particular linking or not, he will assume that there is some basis for connecting one fragment to another. The fact that a
45
Aarseth (1997: 94). Ryan (2005b: 523) describes the adaptation of subject matter and the role of the reader to the hypertextual mechanism as "thinking with the medium." It seems clear that the ability o f the author is critical here: successful hypertext fiction makes use of the medium rather than exploit it for mere superficial purposes. Hypertextual literacy becomes meaningful only if the medium itself facilitates it.
47
Harpold (1991: 134). Toolan (2001: 7). Ryan (2004: 417). Toolan (2004: 222).
48 49 50
'Kaleidoscope' Narratives and the Act o f Reading
291
link has been assigned to a particular lexical item is a narratively significant fact: links are not extraneous to narration, but form an integral part of it, for it is the narrator who offers the reader the possibility of making links. Claims to the effect that hypertext limits readerly freedom of interpretation seem to result from this underestimation of the role of linking as a narrative device. While interaction with hypertextual links is a matter of surface level functionality (albeit involving a level of metatextual contemplation), textual interpretation is a process inspired by this interaction. It has been claimed that hypertexts "distract the reader's attention from that self-reflective, formative dimension of reading," particularly since "the author's links to a new perspective are unlikely to coincide with what the reader might have been led to consider unaided." 51 How this would limit readerly reflection is somewhat puzzling: I would have thought the opposite was true, for surely being confronted with a text which does not fulfill expectations and instead forces us to consider new perspectives is more engaging to the imagination than a text where everything coincides with our own pre-established conceptions. 52 The fact that a link occasionally appears to allow us to "explain" a particular point should not be given any more credence than anything we read in linear narratives: after all, why should we overlook narrative possibilities resulting from the unreliability of a narrator or from the deliberate misguiding of the reader simply because the narrative structure happens to be multilinear? Extending Adorno's idea that "the title is the microcosm of the work" 53 to hypertexts, we could say that the hypertextual link is a title by which one fragment is represented in another. Consequently, the title acts as an incentive for the reader to approach the corresponding fragment in a certain way. While this may at first seem to support Miall's view that links are a potent "invitation to follow another's pattern of suggestion rather than work on clarifying one's own more intuitive promptings," 54 the dynamic has to be ex-
51 52
53 54
Miall (1999: 5). Emmott (1997: 34) mentions that one o f the functions of our awareness of story structure is to guide us into understanding the plot-relevance of otherwise trivial facts. While the hypertextual link (when made apparent by underlining, etc.) draws attention to a particular person, item, action or location mentioned in the text, the reader should not feel compelled to consider only those lexical items significant. Adorno (1992 [1958]: 4). Miall (1999: 5).
292
Jukka Tyrkkö
amined from a broader structural perspective. The same fragment can be, and frequently is, represented with a completely different hypertextual link word elsewhere in the text, guiding the reader to a different interpretation of the fragment's topic and meaning. Realizing that one has arrived at a familiar fragment from a new direction and by choosing a different link excites our interpretive faculties, rather than constrains them. 55 Hypertextual narratives are almost invariably predisposed to this kind of rereading: events are revisited and given new narrative functions depending on the preceding sequence. To my mind, reading such a text actually highlights narrativity by making the reader aware of the fact that the current reading is only one of many possibilities. Furthermore, by actively participating in choosing (at least to some degree) what will be told next, the reader is immersively positioned within the narrative rather than outside of it, and if one subscribes to Culler's approach discussed above, she also becomes partly responsible for generating the fabula as well as the sjuzhet. This point develops into the second fundamental difference between linear and multilinear texts: the experience of reading. The shattering of linearity in fragmented narratives is not only a theoretical concept, but also has implications for the way reading is experienced. With kaleidoscope narratives, the reader is constantly reminded of missed opportunities, discarded links and potential plots which have been set aside, at least temporarily. Iser's suggestion that "the interpreter's task should be to elucidate the potential meanings of the text, and not to restrict himself to just one" 56 would appear to describe the situation quite accurately. Individual readings of a multilinear text are always experienced with an awareness of other potential readings, thus emphasizing the multitude of possible interpretations of events in the text world. With a kaleidoscope narrative, the reader is sure to realize this early on: not only is the text potentially as multi-interpretable as any other, but the kaleidoscopic text as a "physical" entity escapes definition even more than the linear text. It
55
56
Douglas's (1991: 118) account o f her experience o f reading Michael Joyce's afternoon describes this phenomenon vividly: "In afternoon, you can trek across a single place four times, as I did, and discover that it possesses four radically different meanings each time. It wasn't until I had encountered the place more than twice I realized that the words themselves had actually stayed the same, although their meaning had been radically altered." Iser (1984: 22).
'Kaleidoscope' Narratives and the Act of Reading
293
has been noted that the features of multilinear texts has led some literary theorists to mistakenly consider such texts to be ambiguous in meaning. 57 This raises new questions particularly characteristic of fragmented narratives: have I read everything there is to read here? What did I miss? References to characters and locations not encountered at earlier points in the text occur frequently, as do moments of not understanding how the current fragment may be connected to previous and/or later fragments. Moments of sudden recognition, as when two previously separate narrative fragments finally connect and the story comes together, are equally common. While such experiences are not restricted to kaleidoscope narratives, they are more fundamental to them than to other types of narrative. In kaleidoscope narratives, ambiguity is a not so much a feature of narration as it is ofthat which is narrated: the multilinearity of form underlines a claim that this story cannot (or should not) be reduced to only one telling. The most fundamental feature of kaleidoscope narratives may be that they incorporate readerly participation into the production or performance of narrative more explicitly than unilinear texts. Miller's comment on hypertext, 58 quoted in the introduction, suggests that hypertextual interaction and multilinearity underscore the fundamental idea that reading is not only a receptive act, but also a participatory one. Miller does not claim that hypertextual reading alone constitutes a speech act,59 suggesting rather that the links and multiple plotlines of hypertexts make the process of generating meaning more apparent. In a similar vein, Iser has stated that "the meaning of a literary text is not a definable entity but, if anything, a dynamic happening." 60 Perhaps, then, hypertext does have a particular role to play in drawing attention to the potential of not only hypertextual link words to generate meaning, but of other words as well. 61 Hypertextual fictions inspire readers to imagine possible narrative paths behind all words, not just the linked ones. To the degree that this experience
57 58
60
Aarseth (1999: 3). Miller (1995: 38). For discussion o f speech acts in literary theory, see Petrey (1990); Sell (2000: 4 8 - 6 4 ) ; Miller (2001). Iser (1984: 22). The generation of text specific connotations is a feature particularly o f content words (i.e. nouns, adjectives and lexical verbs). Functional words (i.e. pronouns, auxiliary verbs, prepositions, etc.) generally resist the acquisition of new or text specific meanings, reflected by the fact that they are rarely used as hypertextual links.
294
Jukka Tyrkkö
carries over to the reading of linear texts, hypertexts may inspire readers to approach all texts with a renewed sensitivity to their narrative potentials and to the untold stories behind them. As Miller suggests, hypertexts may open our eyes "to see earlier works of literature in a new way, as already proto-hypertexts that invite or allow many different pathways of read-
5. Encyclopedic Narratives: Working to Find the Truth The fragmented narrative style is exemplified in print by the comparatively rare genre of encyclopedic narrative,63 a type of kaleidoscope narrative. Richard Horn's Encyclopedia (1969) and Milorad Pavic's Dictionary of the Khazars: A Lexicon Novel in 100,000 Words (1988a), (1988b) are two such encyclopedic narratives employing a multilinear structure reminiscent of an encyclopedia. Instead of organizing events or episodes into one predetermined plot structure, encyclopedic narratives present them as separate narrative chunks to be explored in any order. The restrictions of the codex naturally force the author to arrange the fragments in a set order; consequently, both Horn and Pavic instruct readers that the order in which the fragments appear is not the only order in which they can be read. The possibility of multiple readings, fundamental to kaleidoscope narratives, is not only a possibility, but in many ways the whole point. Encyclopedia presents itself as a "hand encyclopedia" of the events concerning a group of people living in the northeastern part United States. A short pseudo-editorial prologue discusses the changing role of encyclopedias and how modern times (the work was published in 1969) require a new writing style reflective of the needs of modern readers. The reader is to take note of the "governing principle" and make use of the crossreferences provided in the text. Arranged in traditional encyclopedic fash-
62
Miller (1995: 38). A generally accepted term does not exist for this genre, no doubt due to the small number of texts to which it would apply. The term encyclopedic narrative is used here in reference to narrative architecture which, in content and form, resembles an encyclopedia. In addition to the two texts mentioned above, features of encyclopedic narration can be seen in William Gass's In the Heart of the Heart of the Country and Mark Danielewski's House of Leaves (2001). For discussion of encyclopedic textual structure, see Hoey (2001: 72-92); for discussion of encyclopedias in relation to hypertext, see Bolter (1991: 117).
'Kaleidoscope' Narratives and the Act of Reading
295
ion by alphabetized entries of personal names, place names, and various key words and phrases, Encyclopedia offers no further instructions for reading, thus leaving the reader to his or her own devices as to how to approach the story: should one begin at the beginning or perhaps jump to one of the other entries? Whatever the decision, reading begins in medias res with an awareness of the story world contained in the encyclopedia. The first entry, "Abortion," introduces a character called Sadie Massey and the circumstances surrounding her self-induced abortion. The entry also gives a cross-reference to someone called Tom Jones, inviting the reader to either follow up on him (with only the name to go by) or to choose some other entry at random.64 The experience of reading creates a sense of engagement and involvement, especially when the number of cross-references increases and the story begins to unfold. While the story itself is fairly predictable in its portrayal of late sixties artsy circles with the requisite ingredients of sex, drugs and general angst, the fragmentary structure conjures up the experience of slowly getting to know a new group of people through snippets of hearsay and rumor. The intentional discord between the scholarly form of the structure and the disorganized lifestyle of the protagonists creates a sense of detachment and disillusionment about their ordeals, a sense of otherness which strips them of all dignity and transforms them into specimens of their time in history. Dictionary of the Khazars (hereafter Khazars) is set up with a pseudoeditorial frame narrative describing the history65 of a nomadic tribe, the Khazars. The reader is told by the faux-editor that the dictionary is a reprint of a long-lost collection of texts, compiled by scholars of the three major monotheistic religions in order to solve the "Khazar question." An extensive prologue discusses the history and associated legends of the "Dictionary of the Khazars," the main body of the novel. The dictionary
64
This linking serves as a good example of the narrative possibilities of linking. At this point in the reading, the reader does not know whether the link "Tom Jones" refers to the popular Welsh singer, the character in Henry Fielding's novel (The History of Tom Jones, a Foundling [1749]) or perhaps to some other Tom Jones. The decision to follow up on this link thus involves both ambiguity about the next narrative fragment and, at the same time, the reasonable expectation that a link called "Tom Jones" will lead to a fragment which one way or another addresses this ambiguity. On the other hand, not choosing the link means that the question is left unanswered. The Khazars were a real tribal people living around the Caspian Sea during the latter part of the first millennium. Much of the narrative in Paviö's novel is based on historical facts and legends.
296
Jukka Tyrkkö
itself is divided into three sections (according to the religious authorities responsible for respective sections) and two appendices. The alphabetically arranged entries can be either exclusive to a particular section or shared by two or three sections, in which case the definitions reflect the different information gathered on the same topic by their respective authors. The narrative fragments, each of which depicts an event from the history of the Khazar people or the dictionary itself, form a textual tapestry of initially unconnected morsels of knowledge which the reader has to piece together. This fragmentation is further highlighted by the fact that the Dictionary of the Khazars (the book, not the dictionary section within it) is available in two versions—identified as "male" and "female"— which are identical, apart from a single paragraph in one entry. The text box on the title page informing the reader about the existence of these two versions concludes with a crucial statement: "The choice is yours." In Khazars, the participatory act of reading is an important narrative device in itself. The reader is made aware of this from the start through the device of the faux-editor's prologue, which discusses the various ways in which the dictionary can or should be read. Together with the familiar textual organization of an encyclopedia, this guidance has the effect of evoking past reading experiences of encyclopedias. Important to the experiential aspect of reading, the non-linear structure transforms the premise of searching for knowledge from an intellectual concept into the actual experience of doing so. The reading of Khazars becomes an instance of textual archeology in which the reader has to "dig" for new pieces of historical knowledge, and by doing so is allowed to share in the experience of discovery and seeing the pieces of a puzzle fit together. The wealth of names, locations, and events overwhelms the reader and might even lead to note taking, turning the reader into a "student" of the Khazar legend. The cross-references between the fragments, which signal points of intersection between the three textual traditions, serve the function of hypertextual links by acting as metatextual requests for readerly activity. Instead of merely reading along, the reader has to make a decision on whether to continue with the current fragment or to break off and move on to a new one. The reader is also sensitized to the process of information gathering: knowing that a certain entry can be found in each of the three sections, the reader cannot help but realize the unreliability of the information in each one. Even if one decides to read each entry in all three versions of the dictionary, the order in which this is done may affect the impression one gets of the Khazar question.
'Kaleidoscope' Narratives and the Act of Reading
297
Could Khazars or Encyclopedia have been written in linear narrative? The answer is yes, but not with the same effect. The experiential aspect introduced by the need for readerly activity and decision making—the need for interaction—transforms the act of reading from passive reception to into an act which is closely akin to narration. The sequence of textual fragments is what it is because the reader decided to pursue one particular path, rather than another. Having to do certain things while reading, the reader becomes personally involved in finding out, in connecting the dots and, most importantly, in realizing the extent of his or her lack of recognition. By making the experience of information gathering a performative task, the author can build a narrative which highlights the reader's incomplete knowledge of the story world more substantially than in unilinear texts. The reader is made aware of the possibility that he or she doesn't know something, not because of the narrator, but because of his or her own earlier decisions. While the printed text is set and the number of available paths limited, that may turn out to be somewhat secondary to the immediate immersive experience of interaction. A text offering even one extra dimension of freedom is quite different from a text which offers none. The performative form affects narrativity in a crucial way by suggesting that this story can only be told truthfully if told many times and, in a sense, simultaneously. While this is impossible (we can only receive narration one moment and one event at a time), multilinearity suggests that narration does not play favorites and that the choice is left to the reader to engage the narrative in this or that particular sequence, thus rendering the reader partly responsible for the narrative as it unfolds. In this way, the multilinear medium of hypertext fictions entails alternativity, increasing the possibility of alternative choices in the act of narrating.
6. Hyperfiction: Telling Stories One Link at a Time Electronic hypertexts carry fragmentation and linking to another level by removing the interruption experienced when flipping through pages in kaleidoscopic print texts. Compared to the relatively few fragmented print narratives, hyperfiction narratives are a growing literary genre with rapidly developing narrative conventions and genre features of their own. The most distinguishing feature of hypertexts is the link, not only functionally and structurally, but also because of the effect it has on narration. In hyperfiction, links are the very reason why the author chooses to write
298
Jukka Tyrkkö
in hypertext. Accordingly, hypertextual narratives are generally designed to make full use of the narrative possibilities of linking: instead of relying solely on lexical cohesion between link words and subsequent fragments, hyperfiction authors frequently introduce collocative and co-referential ties, references to themes and concepts, and at times even intentionally incoherent connection to achieve specific narrative ends. 66 While linking does not negate the role of the traditional elements of storytelling, narration undergoes a dramatic change as transitions between episodes are no longer incidental to the narrative, but rather a plot-forming element of storytelling. In the first of the two hyperfictions discussed below, Geoff Ryman's 25367 (1997) fragments are linked with proper nouns. In the second, Robert Arellano's 68 Sunshine 69 (1996), linking is accomplished with proper and common nouns, adjectives, numerals as well as full and partial sentences. Both strategies are specific to the narrative effect desired. The irony of hypertext fictions is that while structure is perhaps their most consistently fascinating feature, it is also the very reason why they are so difficult to analyze in the traditional narratological way. 69 A close reading of a story with literally thousands of potential plots would be possible only by choosing one particular reading and then retracing the choices made at each transition between fragments. A more fruitful approach is to examine the text as a structure, focusing on the close correlation between form and content produced by the reader's choice of a particular link. This approach can be exemplified with 253, a story constructed around the passenger list of a train in the London underground Bakerloo line. Each textual fragment corresponds to a 253-word description of one of the 253 passengers, each description linked hypertextually
66
My forthcoming Ph.D. dissertation discusses the negotiation of coherence in narrative hypertexts from the text linguistic and discourse analytical perspectives. 253 has also been published as a "print remix," transforming (or perhaps translating) it into an encyclopedic narrative not unlike Khazars or Encyclopedia. Mark Danielewski's House of Leaves is another example of a hypertext novel also published later in print. The print version of House of Leaves is somewhat similar to PaviC's Khazars in that it has been published in a number of dissimilar editions. Sunshine 69 is alternatively attributed to either Robert Arellano or Bobby Rabyd, his pseudonym. Like so many other hypertext authors, Arellano is also a notable literary and new media theorist. Coover has famously suggested that the only way to successfully critique a hypertext is by doing it in the same genre. See Landow (1994: 36).
'Kaleidoscope' Narratives and the Act of Reading
299
to the description of one or more of the other passengers on the train and whose lives, through some coincidence, intersect with that of the current protagonist. The reader is invited to begin exploring this human network from any point. Depending on the path followed, various themes may arise as central. Regardless of what that theme may be, the structural organization itself provides an imitation of life's unmanageable complexity, brought to the surface by the experience of moving from one character to another. To analyze a text like 253 using some traditionally dichotomous theoretical set-up would be pointless, seeing as how the main theme of the text is communicated as much through the process of structuring the text as it is by determining the sense of a given word. At first, narration in the traditional sense hardly exists at all, since the reader accesses narrative elements in sequences unique to each reading. 70 By following paths connecting characters and events, the reader slowly begins to discern causal patterns and narratives start to emerge. 253 can be meaningfully described only by taking into account the experiential dimension, where the reader's participation in moving through the passenger list is central to defining what the narrative is about. Interestingly, the narrativity of this text can emerge only if reader continues to read and to structure the text by successively choosing one link or another. The coincidentally crossing trajectories of the various characters' lives become narrative only when the reader stumbles through the textual space on her own, discovering the results of her interpretive choices. The effect of verisimilitude achieved with this narrative strategy can be explained as the reversal of coherence in traditional narrative. Hayden White, for example, argues that "real life can never be truthfully represented as having the kind of formal coherence met with in the conventional, well-made or fabulistic story." 71 While the decision to employ only proper nouns in 253 as links means that the reader will always find the linking itself coherent, the fact that the contents of the subsequent fragment are otherwise unexpected reflects real life by foregrounding its incoherent, uncontrollable, and non-causal aspects. A story like 253, with its macrostructure in a perpetual state of mutation, may perhaps be seen as more candidly referential to real life than the well-constructed fabulistic story. As readers, we recognize our experi-
70
71
Each fragment of a fragmented narrative may of course narrate something. The central issue here is whether or not successive fragments create a narrative. Cf. Liestol (1994). White (1987: ix).
300
Jukka Tyrkkö
ence of the world around us as fragmented and difficult to control and may find a reflection of this in reading a narrative which constantly implies underlying stories, but never tells them in simple and explicit ways. Sunshine 69 exemplifies an even more varied and arguably more typical form of hypertextual fragmentation. Set in the late 1960s in Northern California, Sunshine 69 depicts the lives of a group of characters living under the shadow of the Vietnam War and involved in the burgeoning drug culture spearheaded by Timothy Leary (who is one of the characters). As with 253, the reader can enter the text from any one of the multiple entry points, 72 although with this novel the number of alternative beginnings is limited. The number of links per page is high, sometimes to the extent that virtually every word on the page is a link. The fragments introducing the eight principal characters are cases in point: the short description of a character is followed by a list of the contents of his or her pockets, each item a link to an episode from the respective character's life. Narrative paths split, twist, and multiply, only to converge at some crucial points: a pocket watch, found in the possession of every major character, links to a fragment depicting an hour-by-hour timeline on the crucial day of December 6 th in Altamont, California. The passage of time is an important theme of Sunshine 69 and is reflected in transitions between fragments that frequently correspond to temporal frame changes. Both the linking strategies employed and the constant challenges to cognitive framing create a textual organization which can be used to narrate the difficulty of controlling one's destiny through its very structure. Every decision made by the reader has an effect, but he or she can rarely if ever know for certain what that effect will turn out to be. 73 The linked multilinear storyline achieves this narrative end perhaps more strikingly than any other approach could.
72
See Douglas (2001: 3 9 ^ 2 ) . The effect of lacking control over the reading can be heightened even further by other hypertextual possibilities. In Stuart Moulthrop's Hegirascope 2, the reading of each fragment is limited by time. If the allotted reading time expires, the fragment changes on its own.
'Kaleidoscope' Narratives and the Act of Reading
301
7. Conclusion Fragmented literary forms present a set of challenges to narrative theories developed for the discussion and description of traditional linear narratives. Two of the most important of these challenges would appear to be the effect readerly participation has on narration and the implications of multilinearity for how a narrative is experienced as a representation of an underlying story world. Both of these issues refer to the key topic of reading as an immersive experience. 74 While many traditional narrative theories have, to some extent, managed to do away with the reader and instead focus on analyzing narratives as objects an sich, kaleidoscopic fragmented narratives demand the presence of a reader. The text exists simultaneously, both as a unique sequence created by the reader's choices and as the potential of all possible readings. Like a kaleidoscope, the fragmented text gives pleasure not only through the fleeting experience of a beautiful arrangement, but also through the knowledge that new arrangements, perhaps equally attractive, are available and in some sense exist simultaneously with present reality. Multilinear narratives have the singular quality of highlighting the possibility of multiple interpretations and alternative readings, while at the same time they make the reader aware of the need for activity, showing interest, and getting involved.
References Primary sources Arellano, Robert 1996 Sunshine 69. Available May 11, 2005 online at http://www.sunshine69.com Borges, Jorge Luis. 1977 [1975] The Book of Sand, translated by Norman Thomas Di Giovanni (New York: E. P. Dutton). Cortäzar, Julio 1966 [1963] Hopscotch, translated by Gregory Rabassa (New York: Random House). Horn, Richard 1969 Encyclopedia (New York: Grove Press). Joyce, Michael 1987 afternoon, a story (Cambridge, Mass: Eastgate Systems, Inc.).
74
See Ryan (2001).
302
Jukka Tyrkkö
Moulthrop, Stuart 1995 Victory Garden (Cambridge, Mass: Eastgate Systems, Inc.). O'Brien, Flann 1998 1939] At Swim-Two-Birds (Normal, IL: Dalkey Archive Press). PaviC, Milorad 1988a [1984] Dictionary of the Khazars: A Lexicon Novel in 100,000 Words. Male Version, translated by Christina Pribiöeviö-Zoric (New York: Alfred A. Knopf). 1988b [1984] Dictionary of the Khazars: A Lexicon Novel in 100,000 Words. Female Version, translated by Christina Pribi6evi6-Zori0 (New York: Alfred A. Knopf). Ryman, Geoff 1997 253. Online May 11,2005. Available online at: http://www.ryman-novel.com/info/about.htm 1998 253. Print remix (New York: St. Martin's Press). Saporta, Marc 1962
Composition No. 1 (New York: Simon and Schuster).
Secondary sources Aarseth, Espen 1994 "Nonlinearity and Literary Theory," in Hyper/Text/Theory, edited by George P. Landow, 51-86 (Baltimore: Johns Hopkins UP). 1997 Cybertext: Perspectives on Ergodic Literature (Baltimore: The Johns Hopkins UP). 1999 "Aporia and Epiphany in Doom and The Speaking Clock. The Temporality of Ergodic Art," in Cyberspace Textuality: Computer Technology and Literary Theory, edited by Marie-Laure Ryan, 31-42 (Indiana: Indiana UP). Adorno, Theodor W. 1992 [1958] "Titles," translated by Shierry Weber Nicholsen, in Notes to Literature. Vol. 2: 3-11 (New York: Columbia UP). Bolter, Jay David 1991 Writing Space. The Computer, Hyperspace, and the History of Writing (Hillsdale, NJ: Lawrence Erlbaum Associates). Barthes, Roland 1974 [1970] S/Z, translated by Richard Miller (New York: Hill and Wang). Brooks, Peter 1984 Reading for the Plot: Design and Intention in Narrative (New York: Random House). Bush, Vannevar 1945 "As We May Think," in Atlantic Monthly 176.1: 101-08. Chamey, Davida 1993 "The impact of hypertext on processes of reading and writing," in Literacy and Computers, edited by S. Hilligoss and C. Seife, 238-62 (New York: Modern Language Association). Coover, Robert 1992 "The End of Books," in New York Times Book Review (June 21): 1. 1993 "Hyperfiction: Novels for Computer," in New York Times Book Review (August 29): 1.
'Kaleidoscope' Narratives and the Act of Reading
303
Culler, Jonathan 1999 [1980] "Fabula and Sjuzhet in the Analysis of Narrative: Some American Discussions," in Narratology, edited by Susana Onega and Jos6 Angel Garcia Landa, 93-102 (London and New York: Longman). Douglas, Jane Yellowlees 1991 "Understanding the Act of Reading: The WOE Beginner's Guide to Dissection," in Writing on The Edge 2.2: 112-25. 1993 "Where the Senses Become a Stage and Reading is Direction: Performing the Texts of Virtual Reality and Interactive Fiction," in Drama Review 37.4: 18-37. 1994 "How do I Stop this Thing? Closure and Indeterminacy in Interactive Narratives," in Hyper/Text/Theory, edited by George P. Landow, 159-88 (Baltimore: The Johns Hopkins UP). 2001 The End of Books—Or Books without End? Reading Interactive Narratives (Michigan: University of Michigan Press). Eco, Umberto 1989 [1962] The Open Work, translated by Anna Cancogni (Cambridge: Cambridge UP). 1990 The Limits of Interpretation. Advances in Semiotics (Bloomington and Indianapolis: Indiana UP). Emmott, Catherine 1997 Narrative Comprehension: A Discourse Perspective (Oxford: Clarendon Press). Genette, Gerard 1997 [1982] Palimpsests: Literature in the Second Degree, translated by Channa Newman and Claude Doubinsky. Frontiers of Narrative (Lincoln: Nebraska UP). Gerrig, Richard J. 1993 Experiencing Narrative Worlds: On the Psychological Activities of Reading (New Haven and London: Yale UP). Harpold, Terence 1991 "The Contingencies of the Hypertext Link," in Writing on the Edge 2.2: 126-39. Hayles, N. Katherine 2003 "Translating media: Why we should rethink textuality," in The Yale Journal of Criticism 16.2: 263-90. Hoey, Michael 2001 Textual Interaction: An Introduction to Written Discourse Analysis (London and New York: Routledge). Iser, Wolfgang 1984 The Act of Reading: A Theory of Aesthetic Response (Baltimore and London: The Johns Hopkins UP). Joyce, Michael 1997 "Nonce upon some times: Rereading hypertext fiction," in Modern Fiction Studies 43.3: 579-97. 1995 Of Two Minds: Hypertext Pedagogy and Poetics (Ann Arbor: University of Michigan Press). Landow, George P. (ed.) 1994 Hyper/Text/Theory (Baltimore: Johns Hopkins UP).
304
Jukka Tyrkkö
Landow, George P. 1992 Hypertext: The Convergence of Contemporary Critical Theory and Technology (Baltimore and London: The Johns Hopkins UP). 1994 "What's a critic to do," in Hyper/Text/Theory, edited by George P. Landow, 1 48 (Baltimore: The Johns Hopkins UP). Liestel, Gunnar 1994 "Wittgenstein, Genette, and the Reader's Narrative in Hypertext," in Hyper/Text/Theory, edited by George P. Landow, 87-120 (Baltimore: The Johns Hopkins UP). Miall, David 1998 "The hypertextual moment," in English Studies in Canada 24: 157-74. 1999 "Miller: What's new about electronic media?" Available online May 11, 2005 at http://www.arts.ualberta.ca/~dmiall/hyperead/miller_c.htm 2004 "Episode structure in literary narratives," in Journal of Literary Semantics 33: 111-29. Miall, David / Dobson, Teresa 2001 "Reading hypertext and the experience of literature," in Journal of Digital Information, Volume 2 issue 3. Article No. 46, 2001-08-13. Available online May 11,2005 at http://jodi.ecs.soton.ac.uk/Articles/v02/i03/Miall Miles, Adrian 2002 "Hypertext structure as the event of connection," in Journal of Digital Information, vol. 2 issue 3. Article nr. 100, 2002-03-12. Available online May 11, 2005 at http://jodi.ecs.soton.ac.uk/Articles/v02/i03/Miles Miller, J. Hillis 1995 "The ethics of hypertext," in Diacritics 25.3: 27-39. 2001 Speech Acts in Literature (Stanford CA: Stanford UP). Petrey, Sandy 1990 Speech Acts and Literary Theory (London: Routledge). Rosenberg, Jim 1996 "The structure of hypertext activity," in Hypertext '96: The Seventh ACM Conference on Hypertext Washington DC March 16-20 1996, 22-30 (New York: Association for Computing Machinery). Available online February 1, 2006 at http://www.cs.unc.edu/~barman/HT96/P17/SHA_out.html Ryan, Marie-Laure 2001 Narrative as Virtual Reality: Immersion and Interactivity in Literature and Electronic Media. Parallax (Baltimore: The Johns Hopkins UP). 2004 "Multivariant Narratives," in A Companion to Digital Humanities, edited by Susan Schreibman, Ray Siemens and John Unsworth, 413-30 (Maiden, Mass: Blackwell). 2005a "Peeling the Onion: Layers of Interactivity in Digital Narrative Texts," Paper presented at the Conference "Interactivity of Digital Texts," Münster, Germany, May 2005. Available online September 1, 2005 at http://lamar.colostate.edu/%7Epwruan/onion.htm 2005b "Narrative and Digitality: Learning to Think with the Medium," in A Companion to Narrative Theory, edited by James Phelan and Peter J. Rabinowitz, 515-28 (Maiden, Mass: Blackwell).
'Kaleidoscope' Narratives and the Act of Reading
305
Sell, Roger D. 2000 Literature as Communication (Amsterdam and Philadelphia: John Benjamins). Toolan, Michael 2001 Narratology: A Critical Linguistic Introduction. 2 nd ed. (l s l ed. 1988) (London: Routledge). 2004 "Graded Expectations: On the Textual and Structural Shaping of Readers' Narrative Experience," in The Dynamics of Narrative Form: Studies in AngloAmerican Narratology, edited by John Pier. Narratologia 4: 215-38 (Berlin and New York: Walter de Gruyter). Tuman, Myron C. 1992 Word Perfect: Literacy in the Computer Age (Pittsburgh, PA: University of Pittsburgh Press). Tyrkkö, Jukka 2004 "Negotiating coherence in hypertextual linking," in Belgian Journal of English Language and Literatures. The Linguistics / Literature Interface: 133-53. 2006 "Reading as dialogue: Interacting with the electronic text," in Dialogic Language Use /Dimension du dialogisme /Dialogischer Sprachgebrauch, edited by Irma Taavitsainen, Juhani Härmä and Jarmo Korhonen. Memoires de la Societe Neophilologique LXVI, 123-46 (Helsinki: Societi Nöophilologique). White, Hayden 1987 The Content of the Form: Narrative Discourse and Historical Representation (Baltimore: The Johns Hopkins UP). Zhang, Benji 1998 "Hypertext and postmodern textuality," in CUHK Journal of Humanities 2: 11835. Available online May 11,2005 at http://sunzil.lib.hku.hk/hkjo/view/9/900028.pdf Zwaan, Rolf A. / Radvansky, Gabriel A. 1998 "Situation models in language comprehension and memory," in Psychological Bulletin 123: 162-85.
MICHAEL TOOLAN (Birmingham)
The Language of Guidance How does a narrative guide us to the heart of the matter? Can some of the means of such guidance be discussed in linguistic terms, in ways that take us beyond motif-spotting and a kind of lexical 'stamp-collecting'? My paper explores these issues, and our interest as readers in narrative threads and texture, closure, disclosure, nondisclosure and the implicit, looking particularly at the short story and a short story. That narrative form is sometimes regarded, in its best exemplars, as devoid of 'padding', a form where every sentence counts. This is why, in my view, it particularly merits attention in any efforts at narrative stylistics. In guiding us to the heart of its matter (a heart that we experience, and not merely 'process' or read), the narrative text must chart a way between the rock of predictability and the hard place of impenetrability. For the literarily competent consumer, both of these kinds of writing, the too obvious or easy and the too obscure or difficult, approach the condition of unreadability. They are unreadable in the sense that 'nothing is to be gained, learned or experienced' from processing the utterly predictable narrative discourse that is not already known (and known to be known): no benefit from that cost. (That they are not yet read, and yet judged to be entirely predictable, is one of the ways they are distinguishable from twice- or η-told conversational tales, worth the re-telling.) Nor is there much to be learned or experienced from attempting to process the entirely unpredictable narrative discourse (where the reader is inclined to say that the textual progression is either random or impenetrable, and which of these is more the case is a matter of indifference). Those are the extremes, of course, on a notional continuum; what is harder to discuss in an orderly way are the intermediate cases, those we most admire, which combine low or moderate predictability with elements of relative difficulty. More questions arise: does reading for the plot remain foundational, or can we read against the grain, treating the setting as foundational, ranking the experiential above the eventful? Or in doing the latter is one uncon-
308
Michael Toolan
sciously changing the genre within which one is reading the text, so that the text is no longer read as archetypal narrative but some generic kin, such as travelogue or personal diary? Here I use 'setting' as shorthand for a cluster of phenomena in a text which depict a state and do not strongly project movement, change, or development. So setting is a cover term for 'the whole situation': representations of conditions, circumstances, persons, places or things with diminished overt attention to time or its passing. Focus on setting or situation cannot be entirely devoid of the 'and then' which, among other things, will threaten the validity of one or more of its own propositions (these are always questions of degree); but we can say that kinetic energy is reduced, and that the discourse does not prominently entail or implicate imminent changes of state (actions and events). To put all this another way and bluntly: do short stories generally have 'less plot' than novels, or is this another invalid question? Conventional wisdom says that they do, pointing to the multiplicity of events that unfold in a novel compared with the relatively few events in a story (Ulysses vs. 'Eveline', say). Or the story is said typically to capture one or two scenes, while the novel portrays a whole pageant. Or, as Mary Louise Pratt puts it in the first of eight Propositions contrasting story and novel: The novel tells a life, the short story tells a fragment of a life} Still, a few moments' reflection will remind us that defining 'event' is no straightforward matter, that biting into one small madeleine or crossing a London street can count as a momentous event while the deaths of thousands can be cast as scarcely an event at all. At some point, common ground, 2 decency and convention seem to guide our reception of these matters. We tend finally to see that a bun is just a bun, even if it triggers many memories in the taster, or is his first food for two weeks. But the fact remains that trying to contrast novel and story by counting the events in each is as doomed as comparing two rooms by counting the colours in them. 3 What about the counting of culturally-important events, or at least events of a certain significance such as births, deaths, and sexual unions? Again, I think this will fail, as these can turn out to be quite numerous in the short form, scarce even in the longer form (Anita Brookner novels come to mind for some reason). What about the counting of events that are culturally impor1 2
Pratt (1994: 99) (emphasis in the original). Werth (1999: esp. 117-55). Werth defines common ground as "the totality of information which the speaker(s) and hearer(s) have agreed to accept as relevant for their discourse" (ibid.: 119). On this point, see Prince (1999) and his article in the present volume.
The Language of Guidance
309
tant and which also occupy a central place—informing the theme, perhaps—in a given narrative? Here I think there is more secure ground for suggesting that, typically, short stories have just one or two central, plotcrucial and thematically-significant joinings or removals, while novels typically have several (not without an implicit ranking: the numerous marriages in Middlemarch [1871-72] vary in importance for the novel as a whole). But this is a typicality rather than a necessity. As Pratt emphasises concerning the novel/story proportionality: "It is neither a logical nor an empirical necessity, but rather a fact of literary history, that the short story has developed along lines in part determined by the novel." 4 Or so (switching the verb aspect), I would argue, in the increasingly distant past; more recently its development has continued along lines determined in part by film narratives (an alternative way of spending a hundred minutes or so), and today it is adapting anew in response to the digital revolution generally, but in light particularly of renewed interest in the expansion of spoken story recordings to be consumed via the ear rather than the eye. A short and straightforward discussion by David Lodge of some fundamentals of narrative is a very useful starting point. His remarks come in a long review of Peter Brooks's Reading for the Plot (1984): At the heart of all narrative there is a paradox, namely that the "end" of a story is both its goal and that which terminates the pleasure it yields. A story begins by arousing an expectation o f its end (whether this is neatly closed or ambiguously open is less important than the "sense o f an ending" itself), and it is the lure o f this projected ending, continuously defined and modified by the narrative questions raised and (partially) resolved in the text, that draws us through its pages. We want to know "what happens." Yet, in another sense we do not want to know—if w e did, the plot summary would serve our purpose better than the text. We do not even want to be able to predict with certainty what will happen (hence the importance o f peripety in all narrative o f any sophistication). The postponement of satisfaction, in short, is essential to our enjoyment. 5
Crucial ideas here are that we wish not to predict with certainty and that part of our satisfaction lies in the postponement of satisfaction. These in turn prompt the following suggestions: that we are often happy to make uncertain predictions, speculations, and imaginings of outcomes; and that 4
Pratt (1994: 100). Lodge (1988 [1985]: 197). Earlier in this essay Lodge declares: "To discuss a novel w e have first to describe or recall its plot, however crudely and summarily. This suggests that plot (rather than say 'character' or 'theme' or 'atmosphere') is the basic principle of cohesion in any novel, a truth rather neglected by modern Anglo-American criticism, including my own, until fairly recently." (ibid.: 195). But arguably neither Lodge's antecedent nor consequent claims here are strictly true.
310
Michael Toolan
how 'postponement' is achieved merits close study. But perhaps postponement is not a wholly satisfactory label for a narratological topic where none of several overlapping terms seems entirely satisfactory. Postponement is ordinarily a 'not doing or undergoing now, as planned' and seems to apply most clearly to the delayed disclosure of crucial plot events (similarly, suspense and delay are important narrative principles, but also have specific plot-oriented relevance). In the case of everyday postponements (of a concert or a meeting, say), there is a clear sense of regrettable re-arrangement (postponements are never in themselves 'fortunate'), and this because they leave a gap, a dead space in one's life or schedule, where the event should have taken place. The endless postponements in Chancery in Bleak House (1853) are English literature's best example of such mortal delay. But there are no comparable gaps or dead spaces in narratives (Tristram Shandy's [1759-67] page tricks notwithstanding): the telling doesn't lapse into silence for pages or hours; the discourse continues. So even in the case of plot-oriented 'postponement' the process is as much one of prolonging as postponing; in any case we need to understand how these effects of prolonging and postponing are achieved, and to do that we need to look at how in fact the text does continue. Some things routinely get postponed (meetings, medical procedures, concerts, weddings, funerals), while other important narrative events are in a sense 'unpostponable' (illness, accidents, falling in love)—the distinction being between planned futures and unplanned futures. This distinction must inform our reading of narrative discourse: a character's disclosed circumstances can make their subsequent narrated illness plausible, but not expectable in the way that rehearsals lead to a concert, or death to a funeral. These things contribute to narrative pleasure, narrative's meeting of some of our emotional and intellectual needs and desires. But something other than postponement (or deferral, or peripeteia, or other established term) is also involved: the reader's consciousness that they are still in medias res, that the end (even where not felt to have been postponed) is not yet come: the 'not-yet-ness' of the currently-experienced text. One standard form of plot postponement provides a good example of the more conventional notion: the alternation of plot-lines or protagonists so that the reader is kept busy with one while the advancement of the other is notionally halted. By contrast, the sense of prolongation I'm alluding to arises when there is no such flight from the presented protagonist and plot, but rather a showing and telling more about them, though a
The Language of Guidance
311
more that seems not to carry the reader further on, to the inescapable end, but carries them deeper into the situation, with its multiplying range of possibilities. What is it that the reader is partially guided through or conducted along so that some developments are expectable and their delayed disclosure suspenseful? Other developments might have been foreseen but were not, so that their occurrence is a surprise, a revelation, and then an anagnorisis. And others are simply congruent, fitting but by no means necessary. It is narrative texture—analogous to a river's currents and eddies and sudden surges, unforeseen depths and shallows, sweeping a raft of readers on or drawing them to a bank, but always moving and carrying them. In an older metaphor, narrative texture is a complex wrapping or weaving of strands of meaning ('lexicalised ideation' in Halliday's terms), evoked intermittently but in patterned ways so as to create coherence and continuity.6 With the help of techniques borrowed from corpus linguistics, this paper attempts to track some of the textual or thematic 'strands' set out in the narrative (for our 'guidance') and thus to make some part of our experience of a story's narrativity more accessible to inspection. In view of the semiotic richness (the excess) of even a few lines of text, their ability to presage numerous continuations, how are coherence and thematic prominence achieved? In a sense, we already know how these things are achieved—not simply because so many narrative theorists have produced persuasive expositions on these matters, but also because taking in such effects of continuity and coherence and emphasis are what all proficient readers do. All I am attempting to do is develop a textlinguistic way of describing and tracking what I believe readers do. And the approach I am exploring is quite dependent on corpus analytic resources. In an earlier article, 7 1 wrote about Alice Munro's long story 'The Love of a Good Woman' (1996), making some preliminary attempts to use readers' responses and a simple matrix method so as to explore the story's gaps and surprises systematically. I am still fascinated by this
6
It is instructive to note how hard it is to break away from metaphors that treat narrative composition and reading as a directed, road-like, linear progress—even as various authors have resisted that figure, including Henry James, o f course, with his 'house of fiction', and Alice Munro, who has commented: "[...] I don't take up a story and follow it as if it were a road, taking me somewhere, with views and neat diversions along the way. I go into it, and move back and forth and settle here and there, and stay in it for a while. It's more like a house." (Munro [1982]: 224).
7
Toolan (2004a).
312
Michael Toolan
story, and still analysing it. In that earlier study, I reported readers' responses when they were shown the story's brief prologue-like opening paragraphs and asked certain questions about it. I particularly wanted to see what readers given this opening as 'narrative prime' would identify as expected continuations and the endpoint of the story; thereafter I wanted to see if purely statistical lexical prominences, in the 'Prologue' or the remainder of the text, might in any way match up with those expectations. Here is that Prologue: For the last couple of decades, there has been a museum in Walley, dedicated to preserving photos and butter churns and horse harnesses and an old dentist's chair and a cumbersome apple peeler and such curiosities as the pretty little porcelain-and-glass insulators that were used on telegraph poles. Also there is a red box, which has the letters "D. M. Willens, Optometrist," printed on it, and a note beside it, saying, "This box of optometrist's instruments though not very old has considerable local significance, since it belonged to Mr. D. M. Willens, who drowned in the Peregrine River, 1951. It escaped the catastrophe and was found, presumably by the anonymous donor, who dispatched it to be a feature of our collection." The ophthalmoscope could make you think of a snowman. The top part, that is—the part that's fastened onto the hollow handle. A large disk, with a smaller disk on top. In the large disk a hole to look through, as the various lenses are moved. The handle is heavy because the batteries are still inside. If you took the batteries out and put in the rod that is provided, with a disk on either end, you could plug in an electric cord. But it might have been necessary to use the instrument in places where there wasn't any electricity. The retinoscope looks more complicated. Underneath the round forehead clamp is something like an e l f s head, with a round flat face and a pointed metal cap. This is tilted at a forty-five-degree angle to a slim column, and out of the top of the column a tiny light is supposed to shine. The flat face is made of glass and is a dark sort of mirror. Everything is black, but that is only paint. In some places where the optometrist's hand must have rubbed most often, the paint has disappeared and you can see a patch of shiny silver metal. 8
Readers were then asked about their expectations concerning the story that would follow. Here is one of the questions they were asked: Are there any words or sentences in the text so far that you feel are especially important in creating your expectations about the story's continuation?
Readers' responses showed individual differences of emphasis, but nearly all expected the subsequent story to explain the 'catastrophe', to
g
Munro (1999 [1996]: 3-4); page numbers cited hereafter in the text.
The Language of Guidance
313
tell how Mr Willens came to drown in 1951, 9 However, they are only one way, or one set of ways, of reading the story's opening. I only mention them here, since an interestingly different reading of the prefatory section comes from the Munro literary critic, Catherine Sheldrick Ross, who writes: [That red box is] a powerful element of disruption and disorder in the story, linked with the invasive gaze, illicit sex, and violent death. Everyone associated with the red box eventually is silenced—the optometrist who carried it with him on his professional, and maybe not-so-professional, house-calls; Mrs Quinn, who on her deathbed told a story about it, which may be a malicious lie; and the anonymous donor who dispatched it to the museum, where it has been on display for several decades. It is tempting to see this box as the mysterious centre around which the whole story has been built. 10
Another reader has drawn my attention to the name of the optometrist, Willens, which 'contains' within it the word lens, this arguably encapsulating the core role of lenses and seeing in the story. Counter to this, however, is the fact, noted by Ross, that Willens was the name Munro gave "to a similarly opportunistic character in a very early short story, 'Story for Sunday', published in April 1951 in the University of Western Ontario's literary magazine Folio."" In any event, what I find most striking of all is that neither the redness of the instruments box nor the name Willens were what seemed to attract my attention, as narrationally 'resonant', as key initiations to be expanded upon later, in the course of many readings of that opening page. For Ross, on the other hand, the red box is just the first of a number of disparate physical items by means of which Munro can instil in the reader the sense that she has "included too many things to be held together in the same frame. Some overlooked thing can be expected to emerge from the shadows to overturn any achieved pat„12 tern. 9 10 11
12
See Toolan (2004b) for fuller discussion; there, I also provide a synopsis of the story. Ross (2002: 767). Ibid.: 788. Ibid.: 767. This seems to me a telling insight into Munro's technique, also remarked on by Jonathan Franzen (2004): "The moments [Munro]'s pursuing now aren't moments of realization; they're moments of fateful, irrevocable, dramatic action. And what this means for the reader is you can't even begin to guess at a story's meaning until you've followed every twist: it's always the last page or two that switches all lights on." It should be noted that some critics have regarded Franzen's praise here for Munro as suspicious, disingenuous, and too fulsome to be honest—indeed, as a strategic suppression of Franzen's underlying antipathy towards modern North American and 'difficult'
314
Michael Toolan
These things are textual matter, Barthesian indices perhaps, interpretable as motifs, and some of those that Ross includes are the drowned man's 'waving' hand, the bouquets of forsythia, the ugly brown porch paint, Rupert's little hatchet, and the secretive motion of the rowboat. Again, which to include and which to set aside is an interesting question: why might one pay less attention to Captain Tervitt's hearing aid, or Jimmy Box's father's disability, or Enid and Rupert's crossword puzzlesolving, for example (none of which are commented on in the Ross article)? We have come to know that such scatterings of things, images, narrative threads, are not accidental. The seeming scatter of disparate things has nothing random or inconsequential about it, as the reader half-knows all along, since what we are reading is a short story, the fictional form which pre-eminently requires, as long recognized and reiterated in my opening paragraph, that there be no slack, no padding, no line—certainly no paragraph—that does not earn its place. Unlike the novel. The idea that longer literary works have less chance of being 'flawless' and more likely to have disconnections and fragmentedness is mooted by Philip Hobsbaum, and linked to his concept of 'availability'—which he might have alternatively termed 'finite availability': "just as all of his experience is not available even to the most gifted creative writer, so all of the writer's work is not available to even the most interested reader." 13 The idea is taken up by Wolfgang Iser, who suggests that the more limited a work's availability is—the 'graspability' of the whole text—the more 'stringent' or forced will be the 'consistency' of the interpretation that the critic imposes. 14 Before leaving the red box, let us note the obvious in Ross's quoted remarks above, where she reminds us that the red box is "linked with the invasive gaze, illicit sex, and violent death." That is very much the holistic, conspectual and retrospective summative judgement that a reader can make after—perhaps quite long after—the act of reading. It is clearly not a judgement, an association, which can be made at the time of reading the prefatory section (where we can only link it with the museum, a drowned fiction. See, especially, Franzen (2002), and in rebuttal, Marcus (2005). Risking consignment to the ranks of the uncomprehending naive, I would venture that in his 2004 review, at least, Franzen should be taken at face value in his salute to Munro, as one literary artist to another. 13 14
Hobsbaum (1970: 47). Iser (1978: 16-17).
The Language of Guidance
315
man, an anonymous donation, and home visits). If Ross's formulation is the kind of complex thematic knowledge that is the goal and endpoint of reading, my interest is in the precise means and stages by which the 'innocent' initial linkages or implications of the red box are developed into the dark and violent ones she identifies in her conspectual comment. It is surely a truism that we read for the plot. Not only for the plot, and probably not chiefly for the plot; but, in part, we read for the plot. Put very simply, I believe this means that we identify a character or characters, at one place and time, and 'follow' or compute their movement forward certainly to new times, and usually to new places, in the course of which they become different and more than they had been at our first encounter with them. Whether much more than this is required, in reading for the plot, is open to question. But this suggests a kind of link between this sense of tracking a character becoming different and the role of repetition in narrative and in narrativity. It emerges if we reflect on narrativity's opposite: scenic description. In the latter (a descriptive guide to the city of Zaragoza, for example), there is a kind of requirement for the discourse to move on topologically, to shift topic, to talk about a particular building and then move on to another, without return (returns to previouslydescribed items only happening optionally and for pragmatic and crossreferencing purposes). The discursive procedure is something like the restless, saccadic scanning of a large painting by the viewer, who cannot take everything in or focus on everything at once, but—even before they infer any 'narrative' or sequence implied by the artist's selection and disposition of figures and representations in the canvas—must repeatedly shift focus and attention in the process of viewing the painting. By contrast, an iterative return is required in narratives and underpins narrativity, reporting again and again the same character, who will be different at least to the extent that time has passed and usually different or changed due to the emergence of new causes and effects during that passage of time. Munro's story is a long one (25,000 words, about seventy-five pages) and contains many characters and episodes, some of them mutually linked only in a loose manner (e.g. the slender connections between the three pre-teen boys to whom Section 1 is devoted and the two women— Jeanette, the dying wife of Rupert, and Enid, the home nurse who is tending to her—on whom the remainder of the story, Sections 2 to 4, is centred). At the narrative's disturbing core is the short Section 3 in which Jeanette tells of the optometrist Willens's habitual sexual assaults on her during his home visits to her, and Rupert's catching Willens in the act,
316
Michael Toolan
killing him, and conspiring with Jeanette to pass the death off as an accidental drowning. Part of the fascinating complexity of the story stems from what Carrington has called its many-voiced narration: its characternarrators and character-narratees "withhold, postpone, or reveal information; prohibit and/or prevent the disclosure of information; and validate or invalidate each other's interpretations." 15 The final section is a remarkable one, full of irrealis hypotheticality, of what Enid thinks might happen and what she would do in those circumstances. After Jeanette's death, Enid appears to have resolved upon confronting Rupert (whom she has hopes of one day marrying) with Jeanette's deathbed confession. At least, she devises an elaborate and melodramatic scheme in which she will confront him with Jeanette's account during a boating excursion on the local river so that, if the shame and punishment of going to the police is too much for him, he can drown her. But even as she embarks upon this plot, she begins to have second thoughts, and starts to realize how "collaborating in a silence" or a secret can bring benefits and "keep the world habitable" (76). The story ends before we discover whether Enid actually confronts Rupert or not. No mid-river confrontation is narrated, and judging by the story's brief proleptic Prologue, perhaps none took place. The most blatant plotting in the story then is Enid's, not Munro's, and in fact we never know how much of it is enacted. But we do try to guess, of course, having formed some expectations. For example, we may clutch at textual scraps like the following, narrated from Enid's point of view, when Rupert goes into a woodshed to find her some boots to wear on the walk to the lake: A house like this, lived in by one family for so long a time, and neglected for the past several years, would have plenty of bins, drawers, shelves, suitcases, trunks, crawl spaces full of things that it would be up to Enid to sort out, saving and labelling some, restoring some to use, sending others by the boxload to the dump. (70)
That may warrant us in surmising that at some point after the boating excursion, Enid has found the optometry instruments in the shed, left there by Jeanette, and has despatched them to the museum. The impersonal museum caption could hardly be worded as it is if it had become public knowledge that Mr Willens had not died by drowning, so that presumably Rupert did not go to the police. Beyond this, however, we approach secrecy: Enid may or may not have confronted Rupert over Willens's death; he may or may not have admitted a part in it (he may have been entirely 15
Carrington (1994: 161).
The Language of Guidance
317
innocent, of course); and he and Enid may or may not have married. These episodes are part of a different narrative, not the one told, and so are not available to be experienced—like a death that has not happened. I am interested in seeing whether corpus analysis can help with the kinds of challenge posed by the progression from the Prologue to Section 1 of this story (and by all subsequent progressions). How can a researcher go about suggesting relevant connection between that opening passage about Mr Willens's optometry instruments and everything that follows, including the section immediately following (in which three boys go out to the river for their 'polar bear' dip, only to find Mr Willens's car semisubmerged in the water with the optometrist's body inside)? I believe a start can be made by using the Wordsmith Tools software program to identify the keywords in the story (i.e. those words that are disproportionately frequent—or rare—by comparison with a suitable 'control' frequency list). There is no necessarily direct link between a text's keywords and its themes, structure, or point (however these may be defined), but given that literary word-choice is (a) often crucial to the particularity of effect, and (b) carefully made, it seems reasonable to expect a link which is at least indirect. A beginning was made on the identification of keywords by comparing lexical frequencies in the Munro story with frequencies in a small comparator corpus which I prepared for purposes of close comparison. The comparator corpus comprises approximately 256,000 words of twentiethcentury fiction (stories and novels by Katherine Mansfield, Edna Ferber, Ellen Glasgow, and Ann Beattie). A stoplist of grammatical items was applied before drawing up wordlists for both corpora (i.e. frequencies of the, in, of, had, etc. were ignored). Having also made comparisons with frequencies in other corpora, after excluding proper names and place names, the chief keywords confirmed in the story turn out to be: car, children, water, river, and riverbank. Let us proceed with water, which is here seven times more frequent than in Cobuild's US Books corpus, and more than four times more frequent than in the specially-prepared Mansfield/Glasgow/Ferber/Beattie corpus. As it happens, water appears no fewer than five times in the opening paragraphs of Section 1, that is, in the text that immediately follows the Prologue quoted earlier. Nearly all of these uses refer to the boys finding the car in the water (and then, on closer inspection, Mr Willens's body inside): car tracks to the water's edge...they would jump into the water ...pale-blue shine to the water. Is water important in guiding the reader
318
Michael Toolan
here, given the 'induction' that the Prologue (which does not use the word water) has provided? Such a question invites fuller consideration of the reader-processing background. 16 The obvious and crucial fact is that, by some means, readers 'take in' a text without memorizing it, without retaining every word: sense-making necessarily involves a kind of map-making and schematizing by which reading becomes bearable. Paradoxically, reading is crucially bound up with a process of discarding, of subordinating the individual trees so that the forest can in time be apprehended. Again, my interest is in what it is in the texture and the processing of a written narrative that guides the reader along the way. For that reason, the metaphor of trees and forest is not entirely satisfactory, any more than it would be to characterize the reading of literary narrative as the feeling of various parts of an elephant in order finally to be able to conclude that the animal is an elephant: those figures are much too oriented to a resolution or an outcome, an endpoint. Every commentator on narrative (from Chatman to Lodge to Fludernik to Brooks to Labov), whatever their differences of modelling, agrees that it is the process and not the conclusion or resolution that is central to the reader's experience of narrativity. Narrativity reflects our sense, as we read a narrative, not that there will certainly be an outcome, but only that matters as presented may 'come out', that change may occur. A focus on Wordsmith-identified keywords is no more than a focus on the most disproportionately repeated words of a text (with, here, the decision to exclude grammatical items, proper names, etc.). But how much will evidence of 'noticeable' or 'heavy' use of a word like water or car in this story tell us? Among the things it does not cover are development, variation and change, where a word used at one point is in effect displaced by use of a cognate or variant term a little later on. Counting repetitions alone emphatically neglects such change and modulation, so important in the process and progression of narrative texts especially (it might even be regarded as perverse). So besides repetitions we have to consider all the associations, collocates, semantically-related terms that are arguably activated, or at least made more accessible to activation, by the use of a particular 'starting' term. Take again water: besides all its collocates, its semantic associates must include drought, moist, earth, air, fire, land, and many more lexical items which are neither collocates nor synonyms. The twenty most frequent collocates as recorded in the Cobuild collocates CD 16
This is sketched out in Toolan (2004b).
The Language of Guidance
319
(which operates a sixteen-word stoplist: the, and, of, in, etc.) are, in order of frequency: hot cold under food drinking no like boiling down through electricity supply some off supplies running warm fresh air deep. However, dwelling on disproportionately frequent words and their collocates and semantic associates amounts to focussing on the 'what is said' rather than the unsaid but implicit, derivable by implicature on the basis of what is said and the assumed context of utterance. For the latter purposes, some version of Gricean implicature must be applied to our reading of narrative, and it must often be the case that such implicatures, in combination with the stated message, are what guide the reader. It remains unclear, however, whether corpus analytic methods can assist in highlighting such implicit cues. Consultation of the Cobuild Collocations CD reveals that water and river are among the top twenty collocates of drowned, as node word, in the Bank of English corpus. Thus, upon encountering the word drowned, anyone who is reasonably fluent in English has some kind of tacit expectation that the words water and river may appear in the vicinity, in prior or subsequent text. A strong collocational relationship is not quite an entailment, of course, but it does represent a kind of lexical-prosodic predisposition, so that finding the word water soon after one has read the word drowned is no surprise at all, but strongly foreseeable. Of course, it is important that the priming word in the prefatory section is drowned, which predicts water later; an early mention of water on its own does not, by contrast, make subsequent appearance of drowning particularly expectable. What is reader-guidingly significant in the story's Prologue, and how can one justify selection of certain words or propositions as 'most significant'? An answer to that question, it seems, must invoke a wide gamut of cultural, psychological, and genre-based criteria, some of which have been explored by psychologists under 'narrative interest'. In the present case, I would argue that if we embark on a selective account, to ourselves, of the most important words, scenes, events in that opening passage, the demands of narrativity prompt us particularly to attend to significant events. Few events are reported or implied in the Prologue; but as Labov (1972) and others noted long ago, nothing as an event of human significance beats death or danger of death. On the above bases, then, arguably the most significant event in the opening passage from 'The Love of a Good Woman' is the following: Mr D. M. Willens [who] drowned in the Peregrine River.
320
Michael Toolan
This phrase was singled out by most of my twenty informants as a likely ground for subsequent narrative elaboration. As an event-report, it contrasts with much of the rest of the Prologue, formed mainly of reported states. These are all detachedly reported, at least until the final, more evaluative paragraph which memorably begins: Everything is black, but that is only paint. This sentence, for me, remains the most memorable sentence in the Prologue for reasons having to do with its apparent generalizability and potential figurative interpretation, but also to do with its slight informality and succinctness after the dry formality of the preceding paragraphs describing the instruments, the perhaps deliberately-laboured descriptions of the instruments, complete with doubtfully apposite analogies (snowman, elf s head). Is this the language of a teacher, perhaps, taking her pupils around the museum? At any rate, the discourse effects a register-change with Everything is black, but that is only paint, so I would be inclined to take that as a key evaluative guide along with Willens drowning in the Peregrine River as a key referential guide (to use the Labovian terms) for what follows. What are our scripts, mental models or schemas for a D. M. Willens drowning in the Peregrine River? What words, phrases, even full sentences, does this event or episode licence and make expectable, or likely, or relevant, at some point in the remainder of the story-text? One of the words that 'Willens drowned in the Peregrine River' makes collocationally predictable, as noted earlier, is water. Other words that could be claimed to be semantic associates include death, accident, swim, current, suicide, plus phrases like out of his depth, overcome, weak(ened), unconscious, and so on. Even so, none of the latter words or phrases, as it happens, occurs prominently in the story that follows: only water does. So in a story that challenges our processing and connection-making faculties by apparently offering few or no links between its Prologue and its long first section (which dwells at great length on the boys who go for a swim and come upon Willens's submerged body and makes no mention of the optometry box), the word water creates an implicit thematic-cum-textual spine, or thread, through the whole story. And with water, of course, we return to the story keyword independently identified earlier. Towards the end of the story, water is frequent in two paragraphs of Enid's speculative narrative about confronting Rupert mid-river, as are several words having close semantic association with water, such as river and swim. Again, such claimed association is a matter of judgement rather than objectively provable, and where one chooses to 'draw the line' be-
321
The Language of Guidance
tween 'closely associated' and not closely associated is debatable (e.g. is the word boat that occurs in the passage a semantic associate of water!). In any case, lexico-semantic clustering around water, river, swim, sink, and shore is so prominent that it is reasonable to argue that it has to be prominent, too, in a reader's experience at this point—their grasp of what is centrally evaluated and tellable, what the story's chief preoccupation is at this point. In a first effort to break free from the totalizing perspective and to be more sensitive to the incremental development of the text, 'The Love of a Good Woman' was subdivided into five parts: the brief introductory passage and the four main sections of the story. Then, wordlists were drawn up for each of the story's five parts, and on the basis of comparison with the MFGB corpus, keywords (occurring four times or more) for each section were identified. Once proper names are excluded (along with the 100+ grammatical items on the stoplist), one or two interesting patterns emerge (see Figure 1). Prologue
disk Section 1, Jutland:
car, boys, forsythia, mother, flats, eggs, office, lumps, stores, father, block phoned, police, lake, river
bridge,
Section 2, Heart Failure:
mother, nursing, promise, children, people, pills, wanted, sometimes, senior, notebook, kidneys, nurse, hospital Section 3, Mistake:
car, cloth, glug, got, box, leg, just, floor, sucking Section 4, Lies:
boat, children, water, river, lies Figure 1: Keywords by story section (frequency of four occurrences or more in each section)
These keywords reflect some principal topics of attention in the story, if not its narrative progression. Take the word mother, for instance. Mothers are quite prominent in the first long section, even though it is ostensibly about the three boys and their finding of Mr Willens and his car in the river (a key narrative 'event', the delayed disclosure of which becomes a story in itself). All three of the boys' mothers are alluded to at some
322
Michael Toolan
length, suggesting that in each case they are the most important figure in the boys' lives. This story-length section can be seen in dialogue with Munro's stories of thirty years earlier, and could be titled 'Lives of Boys and Mothers'. In Section 2, Heart Failure, a fourth mother looms quite large, though we really only see her indirectly; this is Edith's mother, with whom Edith lives and on whom she is still in some respects materially and psychologically dependent. But mother as a term and as a role drops away in the latter half of the story when, particularly in light of the Jeanette Quinn's disturbing revelations, Edith ceases to be mothered and devises her own strange plan of action. She also, as it happens, becomes something of a mother herself, to Jeanette's two little girls. In short, an assumed attention, by author and in turn by reader, on water and mother and so on is the kind of 'aboutness' that keywords can highlight. If, however, this story's central events and global interpretation are felt to concern "rape," "manslaughter," "complicity," and so on, then those are forms of 'aboutness' untouched by keywords methodology, as may be 'in' and experienced during the reading of the text. As words, however, none of those three occurs. A further expansion can be effected by adding to the above array all lexical items, independently identified as being among the commonest collocates of the keywords, which are in addition relevant to the storytext (e.g. the collocate itself occurs in the text). The results of this expansion are presented in Figure 2 below. For this purpose, I consulted the Collins Cobuild Collocations CD for the top twenty collocates of each of the identified keywords, where the latter were themselves sufficiently core vocabulary to have their collocates recorded on the CD. With reference to Section 1, for example, among car's top twenty collocates, one {police) is particularly relevant and actually occurs several times in the storytext, and is thus duly recorded. None of the top recorded collocates of boys occurs in the story, while forsythia is itself too 'non-core' to have its collocates (if any) recorded on the CD.
323
The Language of Guidance
disk car (police), boys, forsythia, eggs, lumps, block, phoned (police, home, office, mother, hospital), lake (water, around, river), river (bank, water), water, cars, front (car), nobody, nursing (hospital), promise, children (school, home, wife), people, pills, wanted (said, know, just, go, people), sometimes (people), kidneys, senior, notebook, nurse, hospital, ha, miserable car, cloth, glug, got (know, just, people), box, leg, just, (know, people), sucking, around
floor,
boat, children, water, river (bank, water), lies, jail, punished, cows, swim Figure 2: A keywords map for 'The Love of a Good Woman' without place or personal names and with the commonest collocates occurring in the storytext also noted in parentheses
It can be argued that a table such as that in Figure 2, which suppresses personal name and place name keywords (in addition to excluding grammatical items), thereby leans more towards theme than plot. A further expansion of the keyword base can be achieved by including in the 'map' not only the commonest collocates of keywords occurring textually, but also any words occurring in the text that are plausibly regarded as semantically related to one of the keywords. Moreover, 'soundings' of smaller segments of text and of their keywords (smaller than the 6,000-word sections studied here) can be undertaken to achieve a more fine-grained analysis. It can also be noted that in the table above, nouns predominate over verbs, even though narrative events are normally conveyed by verbs, so that the corpus analysis may need to be weighted in some way more towards identification of 'key verbs'. According to Stubbs, who has conducted a corpus analysis of Heart of Darkness, "Verbs are often a better candidate [.sc. than nouns] for stylistically relevant words." 17 Even a focus on verbs, however, is only an intermediate step where one's interests are in progression or in narrative (not just stylistic) relevance: we need verbs (and nouns) embedded in postulated core propositions or narrative sentences of a kind that might 'drive' a narrative and guide our expectations concerning that which is passing and to come. 18 By all these various means, it is hoped, the most textually prominent and distinctive lexical networks of the narrative—both the lexis used and 17
Stubbs (2005: 11); see also Herman (2005). Cf. Werth (1999: 190ff.) on what he calls 'function-advancing propositions'.
324
Michael Toolan
the propositions it gives rise to, or for which we might claim the reader has been primed—may be made more noticeable. The corpus stylistician goes about this task fearing he may be stumbling towards the blindingly, circularly obvious. The aim is to uncover that lexical structuring that most distinctively makes this story its unique self, partly unlike other stories, and in doing so, to describe the foregrounding that an attentive reader may have become oriented to (without particularly remarking on that orienting). It is precisely because these corpus analytic methods are not those of the ordinary reader—the corpus analytic tool does not 'read' as an acculturated human reader does—that I am confident that the procedure is neither blind nor circular. What does the corpus analytic approach miss or fail to capture? So much! So much of the texture and the mechanisms that generate expectation as we read a story as yet defy 'capture' by semi-automated search and the tools of computation. One such resource of texture and narrativity is the intratextual link, particularly links which involve a faint echo, an oblique analogy—the repetition that scarcely counts as a repetition at all, since it seems to be largely new material. Here are two brief examples, both of which involve late lexical and narrative 'echoes' of just a few words, or the scene those words project, from the opening paragraphs (the Prologue). The relevant sentences are those in the short final paragraph of the Prologue: Everything is black, but that is only paint. In some places where the optometrist's hand must have rubbed most often, the paint has disappeared and you can see a patch of shiny silver metal. (4)
(i) Seventy pages later, Enid is worrying about the truthfulness of what Jeanette has said, thinking about how lies and deceptions can develop in our tellings of events: Look how elaborate dreams are, layer over layer in them, so that the part you can remember and put into words is just the bit you can scratch off the top. (74)
Is there a connection between these two images, some kind of allusion in the latter to the former (notwithstanding the fact that the story's introductory paragraphs date, in story-time, to a point several if not many years after the time of Enid's thought in Section IV)? No simple text-search will highlight a link between these brief passages, and the lexis of the former {hand, rub, paint, disappear, patch, etc.) is not repeated in the latter (layer, part, scratch o f f , top), although I am asserting a semantic relation. In order for the semantic or thematic relation to be made explicit, it needs
The Language of Guidance
325
to be worked out (by the active reader, indeed). Working out such relations is of the order of calculation of Gricean implicatures (the first maxim involved here being that of relation or relevance) at least, if not of a mathematical equation. It involves intermediate terms present in neither of the actual extracts, as for example surface (cf. places rubbed in extract one, the part scratched in extract two). In fact, perhaps surface, face, layer, cover, top form a thematic network in the text. The poetic underspecification of the image in the second extract, where perhaps Enid is alluding to the skin or scab on a partially healed wound (and perhaps not), adds to the uncertainty of the interpretive process. (ii) The final paragraphs of the story are the following: When she and Rupert went underneath the roof of summer leaves it was dusk, it was almost night. You had to watch that you didn't trip over roots that swelled up out of the path, or hit your head on the dangling, surprisingly tough-stemmed vines. Then a flash of water came through the black branches. The lit-up water near the opposite bank of the river, the trees over there still decked out in light. On this side—they were going down the bank now, through the willow—the water was tea-colored but clear. And the boat waiting, riding in the shadows, just the same. "The oars are hid," said Rupert. He went into the willows to locate them. In a moment she lost sight of him. She went closer to the water's edge, where her boots sank into the mud a little and held her. If she tried to, she could still hear Rupert's movements in the bushes. But if she concentrated on the motion of the boat, a slight and secretive motion, she could feel as if everything for a long way around had gone quiet. (78)
What I notice in particular comes in the first three lines, where in the surrounding dark "a flash of water came through the black branches" and lit up the water. In my view, this is an intratextual echo of the earlier image of the black material with the shiny silver showing through: enough fellow readers have agreed that there is some such oblique connection for me to claim that seeing such a link is a process of construal, or plausible interpretation, and not a pure imagining of a connection. Thus, it is an element in the story's texture, present there now and for all time (questions about how 'fully consciously' Munro planned and intended such a textural connection are entirely unhelpful: in construing that link, I and other readers have already given our answer to that question, as best we can, namely that she has written what she has written, and it was not—we can assume—automatic writing). Even so, the association is not glaringly straightforward, and there lies part of its strength. Its interest to reader and analyst is almost in proportion to the extent that it eludes standard corpusanalytic 'capture'. Thus, for instance, there is no lexical common ground
326
Michael Toolan
between the cited passage and the image contained in opening paragraph, other than the image in the word black. Munro's story ends then, seeming to breach one of the stated preferences of our naturalized narratives: that we find out what finally happened. Her stories are a long way from Beckett's meditations (I would hesitate to call some of his later writings, Stirrings Still [1988] and so forth, narratives) and even are less admired by some critics precisely because of their alleged orthodoxies of characterization and setting. As adumbrated earlier, however, Munro seems entirely aware that it is the experience of narrativity, and not the completeness or conclusion of anything so expectable as a plot, that her readers particularly value. So the fading of solution into secrecy, at her stories' closings, is no kind of lack. As Fludernik has argued, the way modern literary narratives speak to readers' experientiality is central to the store we set by them. At the same time, I am quite sympathetic to the emphasis on event and eventfulness discussed by Peter Hiihn in this volume, who refers to the Hamburg Narratology Research Group's proposals concerning eventfulness, which involves a functional approach to narrativity. The Hamburg group emphasize that events in narratives are presented as 'real' and complete within the narrative world and have some degree of the following important properties: "relevance," "unpredictability," "persistence," "irreversibility" and "non-iterativity." 19 And as Peter Hiihn has argued, marked eventfulness surely relates to point and tellability. Matters soon turn complex, however, in the artistic narratives of film and literature. So looking again at the opening to Munro's 'Love of a Good Woman', it is noticeable that the most prominent 'event' adumbrated here is not what it initially seems. That is, we are given to understand here that Mr Willens drowned in the river in 1951; however, the following narrative will provide us with a wealth of circumstantial and reported evidence to suggest that he did not drown, but that he was killed or murdered (another uncertainty) and that his body was subsequently dumped in the river to cover up the crime. Actually, we cannot be absolutely sure that he was murdered, as the testimony comes from a dying woman, allegedly the victim of Willens's sexual assaults, and her confessions were made only to her caregiver, who explicitly wonders whether her revelations are lies. A good deal of what Prince (in this volume) calls the 'positiveness' of happenings in standard narrativity drains away. By 19
Cf. Schmid (2003).
The Language of Guidance
327
the end of the story, the reader trawling for events has landed a curious catch—maybe this didn't happen, maybe that didn't happen, and so on— but has come to a deep insight into Enid and her situation. To repeat my main point here, it is simply that the 'reality' of the drowning event, so prominently announced, without qualification, in the story's Prologue, turns out to be a chimera as a darker narrative is gradually disclosed. Yes, Mr Willens's body is found in the river and he is undoubtedly dead, so many of the other criteria of eventfulness (unpredictability, irreversibility, and so on) remain importantly true. Even so, he was not drowned but killed, and for reasons of relevance that grow stronger (like Faulkner's cables of continuity and inheritance, the memorable figure he uses in Go Down, Moses [1942]) as the narrative unfolds. I have focussed heavily on the fascinatingly enigmatic opening paragraphs of 'The Love of a Good Woman', puzzling over how corpus stylistics might help me understand the workings of Munro's narrative craft in leading us from there onwards through the Viennese tunnels of her story. But I think there is much (more) to be said also in favour of considering story-endings, their structures and foregroundings and resolutions. Steps have been taken in this direction by Susan Lohafer (2003) in a pioneering enquiry into readers' sense of the generic types to which stories may be suspected of conforming, and the way such judgements may prompt readers to identify particular potential 'pre-closure' endpoints for stories. 20 Among other things, we can ask: given this ending, what kind of textual weaving might one expect to have built up to it? * I would like to express my thanks to John Pier and Jose Angel Garcia Landa, editors of this volume, for their helpful comments on this paper. References Carrington, Ildikö de Papp 1994 "Talking dirty: Alice Munro's 'Open Secrets' and John Steinbeck's O f Mice and Men'," in Studies in Short Fiction 31:159-70. Fludernik, Monika 1996 Towards a 'Natural' Narratology (London and New York: Routledge). Franzen, Jonathan 2002 "Mr. Difficult: William Gaddis and the Problem of Hard-to-Read Books," in New Yorker, September 30. 20
This study is reviewed in Toolan (2005b).
328 2004
Michael Toolan
"Alice's Wonderland." Review of Runaway by Alice Munro, in New York Times Magazine Book Reviews, November 14, Section 7, page 1. Hobsbaum, Philip 1970 A Theory of Communication (London: Macmillan). Iser, Wolfgang 1978 The Act of Reading (Baltimore: The Johns Hopkins UP). Labov, William 1972 Language in the Inner City (Philadelphia: University of Pennsylvania Press). Lodge, David 1988 [1985] "The Human Nature of Narrative," in D.L.: Write On: Occasional Essays 1965-1985, 194-99 (London: Penguin). Lohafer, Susan 2003 Reading for Storyness: Preclosure Theory, Empirical Poetics, and Culture in the Short Story (Baltimore and London: The Johns Hopkins UP). Marcus, Ben 2005 "Why Experimental Fiction Threatens to Destroy Publishing, Jonathan Franzen, and Life as We Know It," in Harper's Magazine, October, 331: 39-52. Munro, Alice 1999 [1996] "The Love of a Good Woman," in The Love of a Good Woman: Stories, 3-78 (New York: Alfred Knopf). 1982 "What is Real?," in Making It New: Contemporary Canadian Stories, edited by John Metcalf, 223-26 (Toronto: Methuen). 2004 Runaway (New York: Alfred Knopf). Pratt, Mary Louise 1994 "The Short Story: The Long and the Short of It," in The New Short Story Theories, edited by Charles E. May, 91-113 (Columbus: Ohio State UP). Prince, Gerald 1999 "Revisiting Narrativity," in Grenzüberschreitungen: Narratologie im Kontext. Transcending Boundaries: Narratology in Context, edited by Walter Grünzweig and Andreas Solbach, 43-51 (Tübingen: Gunter Narr Verlag). Ross, Catherine Sheldrick 2002 '"Too Many Things': Reading Alice Munro's 'The Love of a Good Woman'," in University of Toronto Quarterly 71.3: 766-90. Schmid, Wolf 2003 "Narrativity and Eventfulness," in What is Narratology? Questions and Answers Regarding the Status of a Theory, edited by Tom Kindt and Hans-Harald Müller. Narratologia 1: 17-33 (Berlin and New York: Walter de Gruyter). Stubbs, Michael 2005 "Conrad in the computer: examples of qualitative stylistic method", in Language and Literature 14.1: 5-24 Toolan, Michael 2004a "Graded Expectations: On the Textual and Structural Shaping of Readers' Narrative Experience," in The Dynamics of Narrative Form. Studies in AngloAmerican Narratology, edited by John Pier. Narratologia 4: 215-37 (Berlin and New York: Walter de Gruyter).
The Language of Guidance
329
2004b "Values are Descriptions; or, from Literature to Linguistics and back again by way of Keywords," in Belgian Journal of English Language and Literatures (BELL New Series 2), 11-30. 2005 "Is there a genericness that shapes our ends?" Review article on Susan Lohafer's Reading for Storyness, in Journal of Literary Semantics 34.1: 61-68. Werth, Paul 1999 Text Worlds: Representing Conceptual Space in Discourse (London: Longman).
ANSGAR NÜNNING, ROY SOMMER (Gießen, Wuppertal)
Diegetic and Mimetic Narrativity: Some further Steps towards a Narratology of Drama1 1. Introduction In her seminal monograph Towards a 'Natural' Narratology (1996), Monika Fludernik not only provides a radical reconceptualisation and reorientation of narrative theory, she also makes some very original and perceptive observations on drama, which is not generally accepted as a narrative genre in most narratological discussions. Though there have recently been some path-breaking attempts at outlining a narratology of drama, only few narratologists would probably go as far as Fludernik, who calls drama 'the most important narrative genre whose narrativity needs to be documented'. 2 She goes on to discuss, albeit only briefly, a number of convincing arguments that serve 'to emphasize drama's basically narrative nature', 3 e.g. the introduction of different kinds of dramatic narrator figures, the narrative quality of many stage directions, as well as recent cross- and transgeneric developments in drama and the novel which underscore the narrativity of drama. Convincing and suggestive though Fludernik's arguments are, they do not fully resolve the question of what precisely we mean when we argue for or against the narrativity of drama. Taking the various definitions of narrativity that have been proffered by narratologists as our point of departure, we should like to argue that there are different kinds and degrees of narrativity and that taking these different forms into account can throw new light on the disputed narrativity of drama, opening up new horizons for the project of a transgeneric
1
2 3
We would like to thank John Pier and Jose Angel Garcia Landa for their very helpful comments on an earlier draft o f this article. Fludernik (1996: 348). Ibid.: 351.
332
Ansgar Nünning, Roy Sommer
and transmedial narratology. 4 This article attempts to show that the representation of a series of events and the eventfulness of the story are by no means the only typical elements of drama which can be classified as 'narrative'. Proceeding from the theoretical foundations of transmedial narratology as recently delineated by Marie-Laure Ryan (2005), we pursue three main goals: first, to suggest that one should distinguish between diegetic and mimetic kinds of narrativity as well as between different degrees of narrativity; second, to provide an overview of the main diegetic narrative elements in drama including, e.g. audience address, messenger reports, and metalepsis as well as modern narrator figures, all of which support the view that drama by no means lacks a communicative level of narrative transmission; third, to outline some further steps that might be taken towards a fully-fledged narratology of drama by surveying the various contributions that have recently been made (e.g. by Monika Fludemik, Manfred Jahn, Brian Richardson) and by highlighting some areas for further research. Storytelling can be regarded as the default case of dramatic or filmic entertainment: the history of drama is also a history of representing narratives and of communicating them to live audiences. But is there really no difference between the audio-visual representation of action on a theatre stage or movie screen, on the one hand, and the account of an action as given by a fictional narrator in a novel, on the other? Can the specific communicative situations typical of drama and the novel 5 as well as pragmatic aspects (e.g. contexts of reception) be neglected in favour of a transgeneric and transmedial narcological approach to storytelling? The answer to these questions largely depends on the underlying definitions of narrative and narrativity. These matters will be discussed in the following section (section 2). This is followed by a short differentiation of kinds and degrees of narrativity (section 3), and by a survey of diegetic elements in drama (section 4). The remaining sections outline the latest efforts to develop a fully-fledged 'narratology of drama' which pay special attention to narratorial discourse (section 5)—a key question within the ongoing debate concerning drama and narrative—and areas for further research (section 6).
4 5
Cf. Nünning/Nünning (2002a); Herman (2004); Ryan (2004), (2005). Pfister (1988 [1977]).
Diegetic and Mimetic Narrativity
333
2. On Defining 'Narrative' and 'Narrativity', or: Drama as a Narrative Genre? Narratological research concerned with the intricate question of narrativity has so far focussed mainly on the distinction between narratives and other text-types (e.g. argument, description),6 reconstructing 'recent concepts of narrative and the narratives of narrative' 7 and providing very sophisticated models of narrative modalities and generic categorization. 8 In contrast, the questions of different kinds of narrativity and of the historical changes of the forms and functions of narratives in allegedly non-narrative text-types (e.g. drama) have received relatively scant attention. One of the main reasons for the comparative neglect of these two issues is that most of the scholarly work on narrativity has focussed on defining this key concept and on gauging the relation that pertains between narrative and description and between narrative and drama. Though some scholars have convincingly argued that a 'clear distinction between narration and description is of course untenable', 9 many theorists have been preoccupied by the attempt to establish such clear borderlines, not just between narration and description, but also between narrative and drama. Since we will focus on other issues that, comparatively speaking, have been neglected, a quotation from the entry in the splendid Routledge Encyclopedia of Narrative Theory may suffice to recall how the key term of the present volume is generally defined: Narrativity designates the quality o f being *narrative, the set o f properties characterising narratives and distinguishing them from non-narratives. [ . . . ] It also designates the set of optional features that make narratives more prototypical^ narrative-like, more immediately identified, processed, and interpreted as narratives.
This, of course, only shifts the burden of definition onto the concept of narrative, which is itself notoriously difficult to pin down. From a narratological point of view, narrative has been defined as 'the representation of an event or a series of events' , u as the recounting of at least two real or Active events, neither of which logically presupposes or entails the other
6 7 8 9 10 11
Cf. Chatman(1990). Richardson (2000b). Cf. Fludemik (2000); Ryan (1992). Cobley (1986: 397). Prince (2005: 387). Abbott (2002: 12).
334
Ansgar Nünning, Roy Sommer
(Gerald Prince, Shlomith Rimmon-Kenan), as a series of statements that deal with a causally related sequence of events that concern human or human-like beings (Dorrit Cohn), as a representation of a series of causal events or situations not limited to human agents or anthropomorphic entities (Brian Richardson), or as verbal (as opposed to visual or performative) transmissions of narrative content (Gerard Genette). In contrast to the story-oriented definitions of narratives and narrativity championed by Prince, Genette, and others, 12 Stanzel insists that narrativity resides in the mediacy of events through narrative discourse, i.e. in narrative mediation rather than in the narrated story. Monika Fludernik (1996) has proposed yet another, radically deviant position, defining narrativity, from a cognitive point of view, as a mode of naturalising a text or performance in the reception process, i.e. as a function of human experientiality. According to Fludernik, narrativity does not consist in a set of properties that characterise narratives, but can rather be conceptualised as a sort of measure of how readily a given text can be processed as a story. Since her understanding of narrative and narrativity centres on an anthropomorphic kind of experientiality, it can readily embrace drama as a narrative genre: the fact that plays always feature characters on stage guarantees that they project consciousness, experience, speech, and stories. As these competing definitions serve to show, answers to the question about the specificity of narrative can be either text-oriented, defining narrativity as 'the set of properties characterizing narrative and distinguishing it from nonnarrative', 13 or reception-oriented, defining narrativity in cognitive terms as 'a function of narrative texts' which 'centres on experientiality of an anthropomorphic nature'. 14 The latter reconceptualisation of narrativity both allows for a more inclusive concept of narrative which is not restricted to prose and epic verse, and avoids an exclusive focus on aspects of plot. Thus, narrative can neither be reduced to formal properties of narrative discourse, i.e. narratorial mediation in the traditional sense, nor to its representation in a special medium, the written fictional text, but
12
These definitions of 'narrative' or rdcit are reminiscent of the Aristotelian tradition linking drama and epic as narrative genres in the sense of 'genres with a plot (imüthos)' and allow for the inclusion o f any genre with a plot or action sequence into the narratological debate.
13
Prince (2003 [1987]: 65). Fludernik (1996: 26).
14
Diegetic and Mimetic Narrativity
335
also includes drama, film, and narrative poems and performances, provided they inspire interpretation through narrative frames. For the purpose of developing a transgeneric and transmedial narratology, what is needed is an understanding of narrative that does not posit 'the occurrence of the speech act of telling a story by an agent called a narrator' 15 as a necessary condition. In her wide-ranging outline of the theoretical foundations of transmedial narratology, Marie-Laure Ryan has recently proposed to define narrative as a 'cognitive template' characterized by the following three constitutive features: 1)
2)
3)
Narrative involves the construction of the mental image of a world populated with individuated agents (characters) and objects (spatial dimension). This world must undergo not fully predictable changes of state that are caused by non-habitual physical events: either accidents (happenings) or deliberate actions by intelligent agents (temporal dimension). In addition to being linked to physical states by causal relations, the physical events must be associated with mental states and events (goals, plans, emotions). The network of connections gives events coherence, motivation, closure and intelligibility, and turns them into a plot (logical, mental and formal dimension). 6
But even if one does not go as far as Fludernik's radical reconceptualisation of narrative and narrativity, or as Ryan's cognitive redefinition of narrative, there are still convincing reasons for including drama among the narrative genres, not only on account of its plot, but thanks also to the importance that narratives have in many plays. It goes without saying that a definition of narrativity which focuses on plot must include drama among the narrative genres—the more so since the level of the represented world of the characters and the story usually corresponds to that in novels, short stories, and other narrative genres. But, as Monika Fludernik's article in this collection convincingly shows, even definitions of narrativity that do not focus on plot can be applied to drama without any problems. The main reason for this is that narrative and storytelling played a central role both in Classical drama 17 and in the medieval mystery plays (i.e. the annual Cycles performed in major towns such as York, Chester,
15 16
17
Ryan (2005: 2). Ibid.: 4. Cf. de Jong's (2004) analysis of narration and focalization in the Iliad.
336
Ansgar Nünning, Roy Sommer
Wakefield, and Coventry 18 ), and thus since the very beginnings of English drama. Long before the emergence of the novel as the dominant narrative genre, narrative experiments were a central feature of stage plays. Diegetic elements catered to the need for effective exposition and helped to overcome limitations of time and setting imposed by the pressure of normative poetics, by social and political conventions and by the restrictions of stage design. The medieval mystery plays, aiming to show, in the course of one to three days, the whole history of the universe from the creation of Heaven and Earth to Doomsday, made extensive use of narrative. Narrator figures such as the gossips in the staging of "Noah's Ark", a pageant from the Chester cycle, anticipate the functions of the onstage audience (grex) in Ben Jonson's comedies. As Barbara Hardy (1997) demonstrates, Shakespeare's plays also rely heavily on storytelling, making narrative theatrical and stressing the performative quality of the act of narration. More recently, both the emergence of narrative stage genres such as the memory play and the continuing development of narrative visual media such as film, television, and computer games have focussed attention on drama as a medium for fictional storytelling. Therefore, it seems high time to give up cherished normative dichotomies between fiction and drama according to which drama 'is without narratorial mediation', 19 whereas mediacy is seen as the defining characteristic of fiction which distinguishes narrative art from drama. 20 What we should like to explore instead is the broad range of diegetic narrative elements that we find in drama, both on the extradiegetic level of narrative transmission and on diegetic as well as hypo- or metadiegetic levels. Before we do so, however, we will first propose a distinction between different kinds as well as degrees of narrativity.
18
19 20
It is contended, however, in which towns the Wakefield Cycle and the ludus coventrice plays (also known as 'N-Town plays') were actually performed. Elam (1980: 119). Stanzel (1984: 4).
Diegetic and Mimetic Narrativity
337
3. D i e g e t i c a n d M i m e t i c Narrativity: Distinguishing b e t w e e n d i f f e r e n t K i n d s a n d D e g r e e s o f N a r r a t i v i t y The conventional generic distinction between dramatic narratives which imitate speech and actions (mimesis) and epic narratives which recount actions and events (diegesis, poiesis) goes back to Plato's Politeia and also forms the core of Aristotle's Poetics. In Aristotle's work, epic narrative is differentiated from drama with its two formal types, tragedy and comedy, a distinction whose heuristic and didactic value is still widely acknowledged in contemporary literary theory. Of course, the traditional juxtaposition of tragedy vs. comedy as the two main forms of drama is one of ideal types which represent opposite ends of a scale. This scale also accommodates a variety of hybrid forms such as tragicomedy, black humour, farce, theatre of the absurd, and so on. Therefore, the generic differentiation of tragedy and comedy has always presented all sorts of theoretical difficulties. Both tragic and comic drama are clearly plot-oriented genres mainly concerned with telling morally and didactically motivated or sometimes merely amusing stories—stories of the rise and fall of prototypical heroes and their antagonists. These functional characters form and inhabit complex and meaningful fictional worlds. From the point of view of plot-oriented narratologies such as the approach chosen by Thomas Pavel (1985), the functional roles played by characters, conceived of as actants (as suggested by Greimas), as well as fictional events, teleological developments, and semanticised closures not only allow for a systematic analysis of the narrative structure of dramatic texts, but also for a narrative grammar of drama. What plot-oriented narratological approaches to drama that focus on the dramatic structure of plays tend to overlook, however, is that plays do not just represent narratives (i.e. a series of events), they also stage narratives in that, more often than not, they make storytelling, i.e. the act of telling narratives, theatrical. In other words, plays not only represent series of events, they also represent 'acts of narration', 2 1 with characters serving as intradiegetic storytellers. The fact that approaches to drama have not been much concerned with the representation of acts of narration
21
Hardy (1997: 19).
338
Ansgar Nünning, Roy Sommer
has helped to foster the prevailing tradition of drawing clear-cut distinctions between drama and narrative. 22 Despite the hitherto predominant tradition of separating drama and narrative, the last two decades have seen numerous attempts to bridge the gap between the two. The close relationship between dramatic (mimetic) and narrative (diegetic) storytelling is also emphasised by the transgeneric use of critical terminology: key terms and core concepts like scene, character, protagonist, antagonist, dramatic irony, suspense, comic relief, point of view, perspective structure, monologue, dialogue, exposition, subplot, poetic justice, ending, and closure, to name but a few, are used in the analysis of both novels and plays. This tradition of 'exporting narratological concepts' 23 clearly suggests that there are overlaps and interplays between different modes of storytelling 24 and that there are many instances of plays and novels which blur the boundaries and cross the borders between narrative theory and drama criticism. Given the broad range of diegetic narrative elements in drama discussed below (see section 4), it seems sensible to suggest that one should distinguish between diegetic and mimetic kinds of narrativity. Mimetic narrativity could be defined as the representation of a temporal and/or causal sequence of events, with the degree of narrativity hinging upon the degree of eventfulness. Diegetic narrativity, on the other hand, refers to verbal, as opposed to visual or performative, transmission of narrative content, to the representation of a speech act of telling a story by an agent called a narrator. Whereas diegetic narrativity presupposes the presence of a speaker, a proposition, a communicative situation, and an addressee or a recipient role, mimetic narrativity does not. Similarly, while diegetic narrativity presupposes an underlying 'communicational paradigm', 25 mimetic narrativity does not. The cognitive parameters and concepts introduced by natural narratology provide another way of distinguishing between mimetic and diegetic narrativity. Mimetic narrativity focuses on the projection of a sequence of 22
23 24 25
The dramatic representation of acts of narration cannot be reduced to saying out loud what is written on the page, as the actress/actor has at her/his disposal a large array of paralinguistic, gestural and kinetic resources that stage instructions can specify only in part. As Elam and others have shown, the move from 'page' to 'stage' involves the adjunction of semiotic channels that are absent from the written text. Wolf (2005). Cf. Hardy (1991); Ryan (1992). Fludernik (1996: 340).
Diegetic and Mimetic Narrativity
339
events, highlighting the kinds of aesthetic illusion that Werner Wolf christened 'illusion of action' and 'illusion of characters'. 26 Diegetic narrativity, by contrast, serves to create the illusion of a teller, a personalized voice serving as a narrator. By suggesting the presence of a speaker or narrator, diegetic narrativity foregrounds the act of narration rather than the narrated storyworld. In doing so, it does not have to disturb the aesthetic illusion as such; rather, by accentuating the act of narration, it can serve to create a different type of illusion, thus triggering a different strategy of naturalization, viz. what has felicitously called 'the frame of storytelling'. 27 Mimetic narrativity foregrounds 'the story frame' rather than 'the telling frame'. 2 8 In order to avoid a possible misunderstanding, one should add that diegetic narrativity is not confined to the kinds of plays that feature 'generative narrators': '[H]e [the generative narrator] generates a fictional world (hence my name for this practice) in a manner similar to that of an omniscient narrator.' 29 In drama and in narrative fiction, diegetic narrativity is not restricted to such narrators who tell, and generate, stories on an extradiegetic level of communication, but can occur, rather, on various levels of a dramatic text: many prologues and choric narrations would be typical examples of extradiegetic narratives, while the stories told by characters in both novels and plays represent intradiegetic narratives which can feature a high degree of what we have called diegetic narrativity. In addition to avoiding the dangers involved in reductive definitions of narrativity that privilege one of the possible dimensions of narratives at the expense of the others (eventfulness, the presence of a narrator, or experientiality), the proposed distinction opens up productive new horizons for a transgeneric, and, one might add, transmedial narratology. 30 It allows us to account for the fact that different genres typically display different kinds of narrativity. On the one hand, mimetic narrativity tends to prevail in drama, while diegetic narrativity typically dominates in both oral storytelling and many novels. On the other hand, there are many exceptions to the rule. Memory plays like Peter Shaffer's Amadeus (1979), for instance, are a paradigm example of a subgenre that is characterized by a high degree of diegetic narrativity, whereas its degree of mimetic narrativity is 26 27 28 29 30
Wolf (1993: 97). Fludernik (1996: 341). Ibid.: 339. Richardson (2001: 685). Cf. Ryan (2005).
340
Ansgar Nünning, Roy Sommer
usually fairly low. 31 As far as fiction is concerned, Laurence Sterne's Tristram Shandy (1759-67) serves to show that a novel can also have a high degree of diegetic narrativity and very little (if any) mimetic narrativity, while Virginia W o o l f s novels, as well as a host of other modernist novels focusing on the representation of consciousness, display quite a low degree of both diegetic and mimetic narrativity, privileging the projection of experientiality instead. The proposed distinction between diegetic narrativity and mimetic narrativity is not only helpful in order to overcome the traditional dichotomy between narrative and dramatic art, it also provides a rough yardstick that allows one to determine the respective portions of mimetic and diegetic narrative features that a given play or novel displays. Barbara Hardy (1997) has alerted us to the fact that narratives are as prominent in Shakespeare's plays as characterization and imagery. 32 And recent work on the genre that has come to be known as the memory play has served to show that the latter relies much more heavily on diegetic narrativity than on mimetic narrativity, with the stage manager or dramatic narrator fulfilling more or less the same functions as a homodiegetic or heterodiegetic narrator in novels. 33 The next section will give a brief overview of the broad range of diegetic narrative elements in drama, showing just how widespread diegetic narrativity actually is in plays.
4. Diegetic Narrative Elements in Drama Functional characters and the representation of sequences of events are by no means the only regular elements of drama which can be classified as 'narrative', as our preliminary survey of diegetic narrative techniques has demonstrated. The latter include various forms of metalepsis, i.e. transgressions of the boundaries between diegetic levels by characters or narrators, 34 direct audience address by a narrator character, prologue, epilogue, asides, soliloquies, and parabasis (i.e. a song performed by the classical Chorus in the Old comedy, addressing members of the audience), choric speeches, and messenger reports in Greek drama as well as modern narrator figures such as the stage manager in Thornton Wilder's Our Town 31 32 33 34
Cf. Nünning (1994). See also Hardy (1981); Jewkes (1984); Wilson (1995); Costigan (1996). Cf. Bronkhorst (1980); Nünning (1994). Cf. Pier/Schaeffer (2005); Wolf (2005).
Diegetic and Mimetic Narrativity
341
(1938), verbal summaries of offstage action, the play within the play, mise en abyme, narratives embedded within dramatic action, all kinds of metanarrative comment, 35 stage directions, choric figures and narrating characters. This list of diegetic elements in drama can be expanded by transgeneric narrative strategies and story-telling techniques which can be used by both playwrights and novelists, such as montage techniques, scenic narration, and reversals of chronology. As these examples show, narration in drama is not restricted to mimesis or mimetic narrativity, i.e. imitation of an action or of a sequence of events in the Aristotelian sense or, in narratological terminology, the 'showing' or representation of a series of events. It can also make use of the diegetic mode of narration, i.e. of a wide range of techniques for the narratorial mediation, or 'telling', of stories. 36 Traditionally, diegetic elements in drama have either been labelled as 'un-dramatic' exceptions to the (mimetic) rule or else closely linked to specific rhetorical functions such as exposition or communicating events that cannot be shown because of temporal or spatial and practical restrictions. According to this view, narrative is primarily used in drama in order to provide backstory, explain chronological gaps or represent armies, landscapes, seasons, etc. which cannot be shown. Another way of naturalising dramatic narrative was opened up by Brecht, who explicitly linked the diegetic strategies used in his plays to specific ideological effects and purposes such as defamiliarisation or anti-illusionist theatre. In general, however, narrative theory has stuck to the assumption that the events in narrative texts are recounted by a narrator, while dramatic presentation relies on a mimetic portrayal of fictional events, normally lacking any form of narratorial mediation. Even a very brief look at a couple of examples, however, can serve to show that mimetic narrativity and diegetic narrativity have always figured prominently in many plays. The history of English drama alone features a great diversity of dramatic storytellers, Shakespeare's plays representing many cases in point. 37 Prominent examples of plays that display a particularly high degree of diegetic narrativity include, e.g. Othello, As You Like It, The Tempest, Henry V, and Pericles, the latter featuring a highly selfconscious extradiegetic narrator, John Gower, who fulfils all of the functions typically associated with an authorial narrator in narrative fiction, 35 36
37
Cf. Nünning (2004). For the narratological differentiation between showing and telling, see, for instance, Genette (1988 [1983]). See Nünning/Sommer (2006).
342
Ansgar Nünning, Roy Sommer
even including metanarrative functions oriented towards the act of narration, i.e. utterances of the generative narrator which dominantly relate to the level of discourse. John Go wer's generative and metanarrative functions become particularly obvious at the beginning of scene 18, when he appeals directly to the audience, self-consciously explaining his own style of narration in a long metanarrative comment: Thus time we waste, and long leagues make we short, Sail seas in cockles, have and wish but for't, Making to take imagination From bourn to bourn, region to region. By you being pardoned, we commit no crime To use one language in each sev'ral clime Where our scene seems to live. I do beseech you To learn of me, who stand i'th' gaps to teach you The stages of our story [...](18.1 - 9 )
Another paradigm example that typically displays a very high degree of diegetic narrativity is the memory play (cf. section 3). Like the dramatic monologue, with which it has much in common, the memory play is one of those hybrid genres that cut across established generic categories of poetry, drama, and narrative. With their limitation to a single speaker who usually reveals key episodes of his or her life, the dramatic monologue and the memory play combine poetic diction with dramatic presentation and story-telling elements. Moreover, the memory play, once again like the dramatic monologues of nineteenth-century English literature, provides ample evidence of the use of unreliable narration. Just as there are many noteworthy examples of unreliable narration in Victorian poetry, the most famous of which are probably Browning's "My Last Duchess" (1842) and Tennyson's Maud: Α Monodrama (1855), many memory plays display almost all of the features of unreliable narration that are typically associated with this type of narration: 38 they involve first-person speakers whose disturbed perceptions, egotistic personalities, and problematic value-systems lead the reader to question the accuracy of their accounts. Despite the significance of the reliability of dramatic characters, however, the study of both unreliable narration and point of view or focalization in drama has received hardly any attention to date. In the only available article on the subject, Brian Richardson has convincingly shown that the deployment of narrato-
38
Cf. Nünning (1998).
Diegetic and Mimetic Narrativity
343
rial mediation and the appearance of unreliability in plays call 'for the kind of analysis of point of view usually reserved for modern fiction'.39 Such memory plays as Tom Stoppard's Travesties (1974) and Peter Shaffer's Amadeus (1979), which feature Henry Carr and Antonio Salieri respectively as narrators, demonstrate that post-war English playwrights make very subtle use of unreliable narration. In the stage directions of his play, Stoppard explicitly draws attention to Carr's unreliability, which results from the old man's poor memory and his reactionary prejudices: 'the scene (and most of the play) is under the erratic control of Old Carr's memory, which is not notably reliable, and also of his various prejudices and delusions'. 40 The main reasons for Salieri's unreliability are his limited knowledge, the high degree of his emotional involvement, and his problematic value-system. 41 In Amadeus, dramatic irony results primarily from the tension between what the audience sees and what Salieri describes, while Travesties contains a wide range of textual clues to Carr's unreliability. Other examples of plays which violate naturalistic stage conventions by relying on narrators who turn out to be unreliable would be Samuel Beckett's Krapp's Last Tape (1958) and Harold Pinter's Landscape (1968), the latter being composed of alternating and independent acts of narration spoken by two characters. Like many contemporary English novels, these memory plays call into question conventional notions of unreliable narration because they question 'both "reliable" and "unreliable" narration and the distinctions we make between them'. 42 The fact that many recent English novels and plays challenge realist notions of truth and objectivity seems to confirm Wall's view that we perhaps 'need to re-think entirely our notion that unreliable narrators give an inaccurate version of events and that our task is to figure out "what really happened'". 43 It could be argued that the unreliable first-person speakers in plays like Beckett's Play (1963) or Stoppard's Travesties, just like those in novels like Julian Barnes' Talking It Over (1991) or Kazuo Ishiguro's The Remains of the Day (1989), are ultimately not at all unreliable: the stories they tell may not provide objective renderings of events, but they do depict, in a very truthful way, the illusions and self-deceptions of the narrators themselves. 39 40 41 42 43
Richardson ( 1 9 8 8 : 1 9 4 ) . Stoppard (1975: 27). Cf. Nünning (1994). Wall (1994: 23). Ibid.: 37.
344
Ansgar Nünning, Roy Sommer
In the light of the recent insights into the prominent role that narratives and storytelling have played in the history of drama, not only in Shakespeare's plays, but also in the epic theatre, the memory play, and, more recently still, in contemporary Irish plays,44 it is hardly surprising that a number of theorists, notably Manfred Jahn (2001), have also made proposals for a 'narratology of drama'. Such a narratology could not only provide a theoretical and terminological framework for the analysis of both mimetic and diegetic storytelling in drama and in film, but also prepare the ground for a systematic transgeneric and transmedial narrative theory. 5. Towards a Narratological Analysis of Drama Genette's influential approach to narrative did not prove helpful for narratologists interested in drama, film, or painting. Narrative, Genette insists, has to be viewed as a specific way of representing events, rather than as a quality of texts belonging to different genres. Its specific characteristic feature is the verbal transmission of fictional stories. If one follows Genette, neither the dramatic representation of stories on the stage nor their transmission by what he terms extranarrative media (film, comic strip, etc.) can be termed 'narrative', and neither are phenomena that are genuinely interesting to narratology or, more specifically, to discourseoriented narratology. Genette's definition differs markedly from the more recent attempts to develop a transgeneric and intermedial approach to narratology 45 which, during the mid-twentieth century, focussed on storytelling regardless of the medium or mode in which it became manifest. In order to account for this, Genette suggests a division of narratology into two approaches: a thematic approach, concentrating on the story as the narrative content, and a formal, discourse-oriented approach concerned with the analysis of narrative representation. In Genette's system, thematic approaches to storytelling might well include drama, film, and painting, whereas the formal analysis of narrative is restricted to narrative fiction 46
44
Cf. Wehrmann (2004).
45
Cf. Ryan (2004), (2005) and (1992: 368), where she states that the definition of narrativity does not require a narrator. Cf. Genette (1988 [1983]: 16).
46
Diegetic and Mimetic Narrativity
345
Chatman favours a more liberal point of view. He argues that all fictional texts devoted to storytelling share narrative features such as a temporal structure, a set of characters, and a setting. In accordance with the Aristotelian tradition, structural similarities between plays and novels are regarded as more important than the different ways in which these stories are told, or shown. Thus, the difference between diegetic and mimetic storytelling is secondary to the distinction between narrative and nonnarrative text types. Equally important is Chatman's theoretical assumption of the existence of narrating instances in seemingly narratorless narratives, e.g. novels without an overt narrator, including films and plays. This systematic inclusion of narrators as structural elements, ranging from overt narrators to mere arranger of functions, into a theory of narrative paves the way for a post-Genettian, plot- and discourse-oriented theory of narrative, as recently outlined by Jahn (2001), Sommer (2005), and Nünning/Sommer (2002), (2006). Following a discussion of the limitations imposed on narrative research by Genette's restrictive definition of narrative and of his arguments regarding the communicative structure of dramatic texts, Jahn (2001) highlights some crucial areas such as the diegetic status of stage directions, which constitute narrative pauses and thus can be ascribed to a narratorial discourse organising the temporal structure of the story in a play; the possibility of multiple levels of communication within dramatic texts resembling the multi-levelled narrative structures of novels with embedded narratorial acts; and functional genre correspondences such as crossover techniques of dramatisation and epicalisation. He then proceeds to modify Chatman's well-known taxonomy of text-types by suggesting three useful additions. First, he introduces a 'playscript mode' that can be found not only in plays, but also in novels, just as plays and performances can make use of epic narrative modes. Second, Jahn replaces Chatman's subdivisions 'diegetic' and 'mimetic' with the more pragmatic categories 'written/printed' vs. 'performed' Third, he suggests that a transmedial narratology of drama should include separate categories for both the scripts and the performances of plays, films, and operas, which allows for a precise conceptualisation of the interactions between these various forms of representation within the overall model of genres and modes of narration. Jahn's modifications of Chatman's model together with Richardson's essays, which address various facets of narrativity in drama such as categories of order, frequency, duration, point of view, narrative voice, and narration, constitute a theoretical framework that can bring the proposed
346
Ansgar Nünning, Roy Sommer
narratology of drama into relation with other systematic approaches, such as communicational drama theory 47 and theatre semiotics. 48 A narratology of drama is one important building block in a genuinely transgeneric and inter- or transmedial theory of narrative, as outlined in the essays in Nünning/Nünning (2002a), Ryan (2004), and Meister (2005). Another central issue to be addressed in this context concerns the functions of narrative in drama. 49 In addition to its four primary functions, i.e. exposition, suggestion, compression, and address, narrative contributes an important mode to the playwright's creative tools, facilitates interdiscursive experimentation, and encourages self-reflexivity by the disruption of dramatic action. Historically, dramatic narrative has been used by playwrights seeking to observe the principles of verisimilitude and decorum. In Renaissance drama, for instance, violent action had to be narrated rather than shown on stage in order to maintain public order. A systematic approach to the history of drama and narrative would also have to take into account generic issues, as the same narrative elements can fulfil quite different functions in comedy and tragedy.
6. Areas for Further Research for a Narratology of Drama As we have tried to show, the distinction between different kinds and degrees of narrativity and critical attention to the wide range of diegetic forms and functions of narrative found in plays can open up productive fields of enquiry for the on-going project of developing a narratology of drama. Despite the productiveness of the critical industry, the questions surrounding the use of narration and narratives in drama still provide a number of very fertile areas of investigation. There are at least six important issues which have yet to be adequately explored in the frameworks of a 'transgeneric' or 'transmedial narratology' and in which the emerging narratology of drama could initiate further research. (1) Following Jahn's modifications of Chatman's model of text types and narrative modes, the hitherto prevalent generic distinction between mimetic and diegetic narrative needs to be extended by an integrative model which allows for an analysis of diegesis, or diegetic narrativity, in plays, movies, cartoons, etc. as well as for an investigation of mimetic 47 48 49
Cf.Pfister (1988 [1977]). Cf. Elam (1980). Cf. Hart (1991).
Diegetic and Mimetic Narrativity
347
modes of narrativity in novels. As we have tried to show above, there are many diegetic narrative elements and forms in plays, just as there are many mimetic features in modernist novels, for instance, which tend to 'show' rather than 'tell' the story-worlds they represent. (2) This model should be correlated with the communicative model of fictional storytelling 50 in order to provide the broad theoretical framework which is required to keep up with recent developments in fictional storytelling. Though some attempts have been made to systematize the manifold forms and complex functions of dramatic storytelling by drawing on a narratological framework and analytical toolkit, the approaches subsumed under the umbrella of a narratology of drama have yet to develop a fully-fledged theoretical framework. The same holds true for other fields of a transgeneric and transmedial narratology. Whereas significant progress has been made in the narratological analysis of poetry, most notably by Peter Hiihn, Jörg Schönert, and Jens Kiefer, 51 the innovative uses of narrative and kinds of narrativity developed in other genres and media have yet to be adequately mapped. For instance, computer games continuously explore new ways of accessing fictional worlds which are neither epic nor dramatic in the traditional sense, but combine mimetic and diegetic narrative in a variety of ways. Moreover, narrative comprehension in the theatre 52 has yet to be adequately conceptualized. (3) Taking up Monika Fludernik's cue that drama is arguably 'the most important narrative genre whose narrativity needs to be documented', 53 a fully developed narratology of drama would also be well advised to explore the different ways in which novels and plays tend to use narratives. Barbara Hardy is certainly right when she points out that 'drama need not apologize when it is narrative but handles narrative in special ways to make it theatrical', 54 drawing attention to 'the theatrical power of narrative, its capacity to change events, its control and compounding, its passion and its immediacy'. 55 (4) Another field of research for a narratology of drama is the systematic analysis of the various functions of diegesis and mimesis: diegetic 50 51
52 53 54 55
See Pfister (1988 [1977]). Hiihn (2004), (2005); Hühn/Schönert (2002); Hühn/Kiefer (2005); see also MüllerZettelman/Rubik (2005). Cf. Garner (1989). Fludemik (1996: 348). Hardy (1997:29). Ibid.: 60.
348
Ansgar Nünning, Roy Sommer
features such as a narrator figure are not only used differently in novels and drama, but also in different dramatic genres like comedy and serious forms of drama. What Hardy calls 'the performance of telling' 56 already serves to change not only the forms, but also the functions of narrative and narrativity in drama, foregrounding the act of storytelling by making it theatrical and thus privileging diegetic narrativity. As we have tried to show elsewhere by outlining some of the characteristic features and functions of Shakespeare's dramatic and theatrical uses of narrative and narration, 57 much more work needs to be done if we want to come to terms with 'the theatrical power of narrative, its capacity to change events'. 58 (5) Closely related to this generic approach is a diachronic survey of the occurrence and functions of diegetic elements in dramatic narrative, preparing the ground for both a revised and more finely nuanced theory and history of narrative and an alternative history of drama. Despite some decades of innovative and often pathbreaking narratological research, we still know very little of the history of transgeneric and transmedial forms of literary storytelling, and not much more about e.g. the close interrelation between narrative, pictorial, and theatrical arts in eighteenth- and nineteenth-century England. 59 In short: the history of the role of narratives, narration and storytelling in drama has yet to be written. The same holds true for the use of particular narrative techniques in novels and plays, point of view, 60 unreliable narration, 61 and multiperspectivism being but three of many cases in point: Characteristic o f Shakespeare's reflexive narration is the repetition, variation or revision o f an event the audience has already seen acted. Multivocalism was not invented by James Joyce, and Shakespeare too tells the same story from different points of view, at different times, in different moods.
(6) The narratological distinction between mimetic and diegetic storytelling in drama is the first step towards a diachronic analysis of the cultural functions of dramatic storytelling, linking narratology with concepts of performativity. Historical and cultural research in this area might focus on the ways in which different forms of narrativity relate to the distinction 56 57 58 59 60 61 62
Ibid.: 62. Nünning/Sommer (2006). Hardy ( 1 9 9 7 : 6 0 ) . See, however, Meisel (1983). Cf. Floyd (1977); Richardson (1988). See the essays in Nünning (1998). Hardy ( 1 9 9 7 : 2 3 ) .
Diegetic and Mimetic Narrativity
349
of dramatic text vs. performance text, on semiotic aspects of staging diegetic and mimetic narrative, and on the changing role of drama as a medium of cultural storytelling. (7) Finally, one of the most interesting aspects of a narratology of drama is that it reveals blind spots of novel-centred narrative poetics. As the majority of existing narratological models were developed on the basis of and for the analysis of fiction, some areas such as the nature and potential of performance time have been neglected. 'Metatemporal' dramatic narratives with contradictory story times such as Shakespeare's Midsummer Night's Dream63 as well as radical antinomies/asymmetries between text time and stage time and implicit or explicit meta-theatrical reflections on the conventions of temporal representation, as in Tom Stoppard's Travesties, pose new challenges to narrative theory and require a systematic evaluation and modification of existing models of temporal structure. We should like to leave the last word to Brian Richardson, however, who has not only done some of the pioneering work in the recently burgeoning field of a narratology of drama, but also mapped out some of the exciting areas that a transgeneric and transmedial narratology interested in different kinds of narratives and narrativity would only continue to ignore at its own peril: [S]everal important works on cinematic narration have appeared. Comparable investigations o f the theater, however, are still extremely rare. This is an unfortunate state o f critical affairs for two reasons: it is important to acknowledge the rich tradition o f narration in drama, and by doing so it will allow us to identify certain blind spots in theories of point o f view based too narrowly on post-Jamesian novels. Analysis of narration on stage can also lead to interesting questions about other aspects of traditional narrative theory such as the relations between mimesis and diegesis, consciousness and representation, and even the author and the text. 6 4
References Abbott, H. Porter 2002 The Cambridge Introduction to Narrative (Cambridge: Cambridge UP). Bronkhorst, Martin 1980 "Der Erzähler im Drama: Versionen des Memory Play bei Fry, Shaffer, Stoppard und Beckett," in AAA 5: 2 2 5 - 4 0 .
63 64
Cf. Richardson (1987). Richardson (2001: 694).
350
Ansgar Nünning, Roy Sommer
Chatman, Seymour 1978 Story and Discourse. Narrative Structure in Fiction and Film (Ithaca and London: Cornell UP). 1990 Coming to Terms: The Rhetoric of Narrative in Fiction and Film (Ithaca and London: Cornell UP). Cobley, Evelyn 1986 "Description in Realist Discourse: The War Novel," in Style 20.3: 395-410. Costigan, Edward 1996 "Aspects of Narrative in Some Plays by Shakespeare," in English Studies IIA: 323-42. de Jong, Irene J.F. 1991 Narrative in Drama: The Art of the Euripidean Messenger-Speech (Leiden: Brill). 2004 [1987] Narrators and Focalizers: The Presentation of the Story in the Iliad (London: Bristol Classical Press). Doty, Kathleen 1989 "Dialogue, Deixis, and Narration in a Dramatic Adaptation," in Poetica 31: 4 2 59. Elam, Keir 1980 The Semiotics of Theatre and Drama (London and New York: Methuen). Floyd, Virginia Hastings 1977 "Point of View in Modem Drama," in Studies in Interpretation, vol. 2, edited by Esther M. Doyle and Virginia Hastings Floyd, 13-27 (Amsterdam: Rodopi). Fludemik, Monika 1996 Towards a 'Natural' Narratology (London and New York: Routledge). 2000 "Genres, Text Types, or Discourse Modes? Narrative Modalities and Generic Categorization," in Style 34.2: 274-92. Freeman, Roger 1997 "Narrative and Anti-Narrative: Televisual Representation and Non-Causal Linearity in Contemporary Drama," in Journal of Dramatic Theory and Criticism 12.1:39-55. Garner, Stanton B. 1989 The Absent Voice. Narrative Comprehension in the Theater (Urbana: University of Illinois Press). Genette, Girard 1988 [1983] Narrative Discourse Revisited, translated by Jane E. Lewin, Foreword by Jonathan Culler (Ithaca and London: Cornell UP). Hardy, Barbara 1981 "Shakespeare's Dramatic Narrative," in Papers from the First Nordic Conference for English Studies, Oslo, 17-19 September, 1980, edited by Stig Johansson and Björn Tysdahl, 1-17 (Oslo: Institute of English Studies, University of Oslo). 1997 Shakespeare's Storytellers. Dramatic Narration (London and Chester Springs, PA: Peter Owen). Hart, Jonathan 1991 "Narrative, Narrative Theory, Drama: The Renaissance," in Canadian Review of Comparative Literature 18.2-3: 117-65.
Diegetic and Mimetic Narrativity
351
Herman, David 2002 Story Logic. Problems and Possibilities of Narrative. Frontiers of Narrative (Lincoln and London: University of Nebraska Press). 2004 "Towards a Transmedial Narratology," in Narrative across Media. The Languages of Storytelling, edited by Marie-Laure Ryan. Frontiers of Narrative, 4 7 75 (Lincoln and London: University of Nebraska Press). Herman, David / Jahn, Manfred / Ryan, Marie-Laure (eds.) 2005 Routledge Encyclopedia of Narrative Theory (London and New York: Routledge). Hühn, Peter 2004 "Transgeneric Narratology: Applications to Lyric Poetry," in The Dynamics of Narrative Form. Studies in Anglo-American Narratology, edited by John Pier. Narratologia 4: 139-58 (Berlin and New York: Walter de Gruyter). 2005 "Plotting the Lyric: Forms of Narration in Poetry," in Theory into Poetry. New Approaches to the Lyric, edited by Eva Müller-Zettelmann and Margarete Rubik, 147-72 (Amsterdam and New York: Rodopi). Hühn, Peter / Kiefer, Jens 2005 The Narrato logical Analysis of Lyric Poetry. Studies in English Poetry from the 16th to the 2(fh Century, translated by Alastair Matthews. Narratologia 7 (Berlin and New York: Walter de Gruyter). Hühn, Peter / Schönert, Jörg 2002 "Zur narratologischen Analyse von Lyrik," in Poetica 34.3-4: 287-305. Jahn, Manfred 2001 "Narrative Voice and Agency in Drama: Aspects of a Narratology of Drama," in New Literary History 32.3: 659-79. Jewkes, W. T. 1984 '"To Tell my Story': The Function of Framed Narrative and Drama in Hamlet," in Shakespearian Tragedy, edited by Malcolm Bradbury and David Palmer, 3 1 46 (New York: Holmes & Meier). Lovrod, Marie 1994 "The Rise of Metadrama and the Fall of the Omniscient Observer," in Modern Drama 37.3: 497-508. Meisel, Martin 1983 Realizations. Narrative, Pictorial, and Theatrical Arts in Nineteenth-Century England (Princeton, NJ: Princeton UP). Meister, Jan Christoph (ed.) 2005 Narratology beyond Literary Criticism. Mediality, Disciplinarity. Narratologia 6 (Berlin and New York: Walter de Gruyter). Morrison, Kristin 1983 Canters and Chronicles. The Use of Narrative in the Plays of Samuel Beckett and Harold Pinter (Chicago and London: University of Chicago Press). Müller-Zettelmann, Eva / Rubik, Margarete (eds.) 2005 Theory into Poetry. New Approaches to the Lyric (Amsterdam and New York: Rodopi).
352
Ansgar Nünning, Roy Sommer
Nünning, Ansgar 1994 "Be my Confessors! Formen und Funktionen epischer Kommunikationsstrukturen in Peter Shaffers Amadeus" in Forum Modernes Theater 9.2: 141-60. 2004 "On Metanarrative: Towards a Definition, a Typology, and an Outline of the Functions of Metanarrative Commentary," in The Dynamics of Narrative Form. Studies in Anglo-American Narratology, edited by John Pier. Narratologia 4: 11-57 (Berlin and New York: Walter de Gruyter). Nünning, Ansgar (ed.) 1998 Unreliable Narration: Studien zur Theorie und Praxis unglaubwürdigen Erzählens in der englischsprachigen Erzählliteratur (Trier: WVT). Nünning, Ansgar / Nünning, Vera (eds.) 2000 Multiperspektivisches Erzählen. Zur Theorie und Geschichte der Perspektivenstruktur im englischen Roman des 18. bis 20. Jahrhunderts (Trier: WVT). 2002a Erzähltheorie transgenerisch, intermedial, interdisziplinär. WVT-Handbücher zum Literaturwissenshaftlichen Studium 5 (Trier: WVT). Nünning, Vera / Nünning, Ansgar 2002b "Produktive Grenzüberschreitungen: Transgenerische, intermediale und interdisziplinäre Ansätze in der Erzähltheorie," in Nünning/Nünning (2002a: 1-22). Nünning, Ansgar / Sommer, Roy 2002 "Drama und Narratologie: Die Entwicklung erzähltheoretischer Modelle und Kategorien für die Dramenanalyse," in Nünning/Nünning (2002a: 105-28). 2006 "Die performative Kraft des Erzählens: Formen und Funktionen des Erzählens in Shakespeares Dramen," in Shakespeare Jahrbuch 142, edited by Ina Schabert, 124-41 (Bochum: Verlag und Druckkontor Kamp). Pavel, Thomas 1985 The Poetics of Plot. The Case of English Renaissance Drama (Manchester: Manchester UP). Pfister, Manfred 1988 [1977] The Theory and Analysis of Drama, translated by J. Halliday (Cambridge: Cambridge UP). Pier, John / Schaeffer, Jean-Marie (eds.) 2005 Metalepses. Entorses au pacte de la representation. Recherches d'histoire et de sciences sociales 108 (Paris: Editions de l'EHESS). Prince, Gerald 1999 "Revisiting Narrativity," in Grenzüberschreitungen: Narratologie im Kontext / Transcending Boundaries: Narratology in Context, edited by Walter Grünzweig and Andreas Solbach, 43-51 (Tübingen: Narr). 2003 A Dictionary of Narratology. Revised ed. (1 st ed. 1987) (Lincoln and London: University of Nebraska Press). 2005 "Narrativity," in Routledge Encyclopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 387-88 (London and New York: Routledge). Richardson, Brian 1987 '"Time is out of Joint': Narrative Models and the Temporality of the Drama," in Poetics Today 8.2: 299-309.
Diegetic and Mimetic Narrativity 1988
353
"Point of View in Drama: Diegetic Monologue, Unreliable Narrators, and the Author's Voice on Stage," in Comparative Drama 22.3: 193-214. 2000a "Narrative Poetics and Postmodern Transgression: Theorizing the Collapse of Time, Voice, and Frame," in Narrative 8.1: 23-42. 2000b "Recent Concepts of Narrative and the Narratives of Narrative Theory," in Style 34.2: 168-75. 2001 "Voice and Narration in Postmodern Drama," in New Literary History 32.3: 681-94. Ryan, Marie-Laure 1992 "The Modes of Narrativity and Their Visual Metaphors," in Style 26.3: 368-87. 1997 "Interactive Drama: Narrative in a Highly Interactive Environment," in Modern Fiction Studies43.3: 677-707. 2005 "On the Theoretical Foundations of Transmedial Narratology," in Narratology beyond Criticism. Mediality, Interdisciplinarity, edited by Jan Christoph Meister. Narratologia 6: 1-23 (Berlin and New York: Walter de Gruyter). Ryan, Marie-Laure (ed.) 2004 Narrative across Media. The Languages of Storytelling. Frontiers of Narrative (Lincoln and London: University of Nebraska Press). Shakespeare, William 1997 Pericles, Prince of Tyre, in The Norton Shakespeare. Based on the Oxford Edition, edited by Stephen Greenblatt et al., 2719-83 (New York and London: Norton). Sommer, Roy 2000 "Funktionsgeschichten: Überlegungen zur Verwendung des Funktionsbegriffs in der Literaturwissenschaft und Anregungen zu seiner terminologischen Differenzierung," in Literaturwissenschaftliches Jahrbuch 4 1 : 3 1 9 - 4 1 . 2005 "Narrative and Drama," in Routledge Encylopedia of Narrative Theory, edited by David Herman, Manfred Jahn and Marie-Laure Ryan, 119-24 (London and New York: Routledge). Stanzel, Franz K. 1984 [1979] A Theory of Narrative, translated by Charlotte Goedsche (Cambridge: Cambridge UP). Stoppard, Tom 1975 Travesties (London and Boston: Faber and Faber). Tschauder, Gerhard 1991 "Wer 'erzählt' das Drama? Versuch einer Typologie des Nebentexts," in Sprache und Literatur in Wissenschaft und Unterricht 68: 50-67. Wall, Kathleen 1994 "The Remains of the Day and its Challenges to Theories of Unreliable Narration," in Journal of Narrative Technique 24.1: 18-42. Wilson, Rawdon 1995 Shakespearean Narrative (Newark, NJ and London: University of Delaware Press). Wehrmann, Jürgen 2004 "Irish Tradition or Postdramatic Innovation? Storytelling in Contemporary Irish Plays," in Zeitschrift für Anglistik und Amerikanistik 52.3: 243-56.
354
Ansgar Nünning, Roy Sommer
Wolf, Werner 1993 Ästhetische Illusion und Illusionsdurchbrechung in der Erzählkunst: Theorie und Geschichte mit Schwerpunkt auf illusionsstörendem englischen Erzählen (Tübingen: Niemeyer). 2002 "Das Problem der Narrativität in Literatur, bildender Kunst und Musik: Ein Beitrag zu einer intermedialen Erzähltheorie," in Erzähltheorie transgenerisch, intermedial, interdisziplinär, edited by Vera Nünning and Ansgar Nünning, 2 3 104 (Trier: WVT). 2003 "Narrative and Narrativity: A Narratological Reconceptualization and Its Applicability to the Visual Arts," in Word & Image 19.3: 180-97. 2005 "Metalepsis as a Transgeneric and Transmedial Phenomenon. A Case Study of the Possibilities of 'Exporting' Narratological Concepts," in Narratology beyond Criticism. Mediality, Interdisciplinarity, edited by Jan Christoph Meister. Narratologia 6: 83-107 (Berlin and New York: Walter de Gruyter).
MONIKA FLUDERNIK (Freiburg-im-Breisgau)
Narrative and Drama 1. Introduction It is only in recent times that narratology, i.e. the study of narrative inspired by linguistic structuralism that developed in the 1950s and 1960s, has started to focus on drama as a narrative genre—quite oddly so, since some of the key concepts of narratology derive from Aristotle's Poetics, whose prime example for the elucidation of plot was the Athenian stage. Modern narratology, by contrast, emerged from the analysis of the novel, which has been narrative's most popular form for three centuries. By focusing on the novel, narratology inevitably ended up highlighting the function of the narrator as a basic element of the narrative text. Whether in Goethe's triad of Epik, Dramatik, Lyrik or in later theoretical proposals by F. K. Stanzel and Gerard Genette, drama was treated as the other of narrative, that genre in which action was supposedly directly enacted on stage (rather than represented in the words of a narrator) and which lacked the figure of a narrator persona. Narratology, therefore, paradoxically turned a blind eye to the similarities shared by novels and plays in terms of plot construction, similarities commented on by a whole line of critics from Corneille and Dryden to Brecht and Henry James or Percy Lubbock. In recent years, this dichotomy between narrative (i.e. exclusively verbal narration) and drama has come under attack from the narratological side. For instance, Brian Richardson criticizes the reduction of drama to the exclusively mimetic mode: "[...] major theorists of both narrative discourse and the semiotics of theater generally agree that drama is exclusively a mimetic genre, while fiction combines mimesis and diegesis."' Richardson then goes on to blast this consensus. A landmark in the realignment of drama and narrative is Manfred Pfister's The Theory and Analysis of Drama (1988; originally published in German as Das Drama 1
Richardson (1988: 193).
356
Monika Fludernik
in 1977). Not only did Pfister employ narratology's conceptual toolbox to describe plot structure and temporal rearrangements in presentation (successfully using the methods of Lämmert and Genette for the analysis of novels), but he additionally introduced the terms perspective and perspective structure into the analysis of drama, thereby suggesting that issues of point of view are relevant to both fiction and drama. Since then, the inclusion of drama among narrative genres has been advocated most forcefully by Mieke Bal (1997 [1985]), who was the first narratologist 2 to recognize that a definition of narrative on the basis of plot inevitably required the inclusion of drama alongside film, ballet, cartoons and other forms of narrative outside the prototypical scenario of a narrator's telling a story orally or in writing. A second wave of narcological theorising trying to expand the field of the discipline has emphasised mediality and transmedial adaptation as an important aspect in narrative studies. Film was discovered as a narrative genre early on, both on the narratological side 3 and by film theoreticians. 4 Recently, the screen play has emerged as a genre in its own right, and, at least in Germany, the radio play has attracted considerable attention as a separate genre.5 More recently still, analyses of transgeneric and transmedial forms of narrative in Ansgar and Vera Nünning's volume Erzähltheorie transgenerisch, intermedial, interdisziplinär (2002a) and in Marie-Laure Ryan's Narrative across Media (2004b) include even more 'marginal' genres as possible candidates for narrativity: the lyric;6 painting and music; 7 hypertexts or interactive cinema.8 From this perspective, it
2
3 4
5
6
In fact, as Marie-Laure Ryan (2004a: 1) has pointed out, Claude Bremond already in 1964 recognised the transmedial transposability o f narratives. Cf. Bal (1997 [1985]); Chatman (1978), (1990). Cf. Metz (1974); Bordwell (1985), (1989); Kozloff (1988); Branigan (1992). I am here concentrating on the story since 1966; Russian Formalism and film theory (Eisenstein) never opposed drama and narrative in this manner and saw the connections between film and narrative. Cf. Korte/Schneider (2000) on screenplays; on radio plays, cf. Frank (1981); Hannes (1990); Maurach (1995). See also W o l f (1998), (2003); Fludemik (2000); Müller-Zettelmann (2002), (2005); Hühn (2004); Hühn/Kiefer (2005).
7
Wolf (2002); Ryan (2004b: 139-94; 267-328).
8
For theoretical discussions of transmedial narratology, see the introductory sections to Nünning/Nünning (2002a) and to Ryan (2004b); also, Nünning/Nünning (2002b); Ryan (2004a); Herman (2004).
Narrative and Drama
357
appears a much less radical move to consider drama as a narrative genre, i.e. an object of narratological study. Among narratologists who have started to take drama seriously as narrative, one needs to mention first and foremost Brian Richardson who, over the years, has produced a series of articles about the narrative aspects of (mostly, but not exclusively, postmodernist) drama. 9 Richardson has drawn particular attention to dramatic uses of a narrator figure, as in Thornton Wilder's Our Town (1938)—compare also Peter Shaffer's Amadeus (1979)—and to metadramatic 'frame narrator' plays like The Taming of the Shrew (1593). He has also provided extensive analyses of Beckett's dramatic art as narrative. 10 Since Brian Richardson's contribution, Ansgar Nünning and Roy Sommer have turned to the issue of narrativity in drama. In their 2002 article and in their contribution to this volume, they emphasise not only that one can establish narrativity for drama, but that dramatic elements are common in narrative as well." They note that self-reflexive dramatic strategies, which are discussed under the label metadrama, 12 frequently correlate with the introduction of narrators and narration on stage, 13 typical of epic theatre. Moreover, plays have a level of communication framing the dramatic action and including prologues, epilogues, chorus addresses to the audience, as well as asides. Quoting Richardson, 14 Nünning and Sommer move towards a schema of narrative levels and techniques in drama such as Richardson's "generative narrator": a narrator figure or chorus fulfilling the functions of an authorial narrator 'generating' the plot. Referring to Manfred Jahn's (2001) essay on voice and narration in drama, Nünning/Sommer conclude with a proposal for a research programme that assesses the utility of narratological categories for the analysis of drama. Such a study should discuss whether concepts like story and discourse, events, existents, focalisation or voice, narrators, narratees or narrative situations could also be employed in the study of dramatic texts. In the article included in this volume, Nünning and Sommer go one step further by proposing a "narrative grammar of drama." In addition, they distinguish between diegetic and mimetic narrativity as two ends of a 9 10 11 12 13 14
Cf. Richardson (1987), (1988), (1997), (2001). E.g. Richardson (1988: 202-04). Nünning/Sommer (2002: 108). Cf. Hornby (1986). See Nünning/Sommer (2002: 112-13), referring to Vieweg-Marks (1989). Richardson (2001: 685).
358
Monika Fludernik
scale on which various narrative genres and subgenres can be placed, with novels prototypically located at the diegetic and plays at the mimetic end. What I would like to do in this paper is to return to the narratological definition of narrativity and to the position of drama as a narrative genre within that framework. This paper extends my earlier remarks on the narratological analysis of drama provided in Towards a 'Natural' Narratology.15 There, I had noted the importance of narrator figures on stage and the narrativity of stage directions and commented on the similarity in postmodernist transgressions of mimesis between novels and plays. In these pages, I want to consider the status of drama in the framework of my experiential definition of narrativity. I will go on to contrast narrative aspects of plays with dramatic strategies in fiction and will then conclude with some remarks on performativity versus textuality.
2. How Narrative is Drama? Starting out from the traditional definitions of narrative—the narration of a series of events (Prince, Genette); the conjunction of story and discourse (Chatman); mediacy of events through narrative discourse (Stanzel)—one will tend to include drama among the narrative genres on account of its plot, but exclude it from narrative because of the missing narrator/narration function. Neither argument is, however, really convincing. On the one hand, like modernist and postmodernist fiction, modernist and postmodernist drama has decisively moved away from a dominant plot element (just think of O'Neill's The Emperor Jones [1920] or Beckett's Krapp's Last Tape [1958]).16 Conversely, it could be argued that the absence of a narrator persona or an act of narration does not inevitably disqualify drama from the narrative genre. After all, in film, plot mediation may be located in the image sequences and soundtrack, which would suggest that the enactment of plot on stage in performance is a type of mediacy as well. 15
Fludernik (1996: 3 4 9 - 5 2 ) .
16
Compare Richardson's dictum: "I wish to suggest that the idea of narrativity presupposed by most poetics is fundamentally misleading, if not completely mistaken. The assumption is that there exists some foundational essence common to all narratives, factual and fictitious, natural and literary, which the function of poetics is to determine and clarify [sic]. In particular, the notion of a single, univocal and extractable story is common [...]" (Richardson [1987: 306]). Prince—notably in this volume—and Herman, among others (e.g. Ryan [1992]), argue for types and degrees o f narrativity.
Narrative and Drama
359
The comparison with film is instructive. In film theory the representation of the fictional world (action, characters, space) can be described as a 'discourse' of pictures (shots) and soundtrack, with angles, shot sequences (and rearrangements), lighting, music and so forth, inflecting a mimetic representation in the direction of point of view, selectional mechanisms and evaluative stances. In other words, what we see in film is a complex mosaic of technical choices, a filmic discourse that resembles narrative discourse in novels with its anachronies, focalisations, narratorial commentary (e.g. voice-over), etc. So strong is this analogy that film critics have even invented a 'cinematic narrator' 17 to account for the overall meaning of juxtaposition, camera angles, sound, light and colour effects. 18 Hence, in analogy with Jahn, one could argue that drama is narrative because there exists a 'dramatic composition device' that regulates anachronies, selection and juxtaposition of scenes, performance factors such as staging and acting, lighting and mise-en-scene in general. Dramatic 'discourse' would then be equivalent to the medium of performance and enactment. 19 Conversely, a definition of narrativity that does not focus on plot, but on fictional worlds 20 and/or experientiality, 21 can likewise absorb drama without any problems. The memory play, the monologue play and the dream play are prime examples of plotless drama that nevertheless satisfy the definitional criteria of narrativity in these frameworks. In fact, drama is rarely confronted with the problems of postmodernist fiction in which several linguistic strategies militate against narrativity. Thus, in drama we rarely have incomprehensible language or plots in which the identity of the characters dissolves. Moreover, juxtapositional postmodernist techniques (compare novels that consist of obituaries, newspaper reports, lists etc.) are more clearly disruptive of narration and plot constitution than any visual experiments that I am aware of in the dramatic canon. Nor do we find speech in which the narrative discourse syntactically and semantically disintegrates. Even so, there are some postmodernist plays in which
17
18
20 21
Chatman (1990). Manfred Jahn proposes to call the instance responsible for such rearrangements the "filmic composition device (FCD)." For a discussion o f this issue, see Jahn ( 2 0 0 3 ) and Alber (2007). If one considers the playscript as a separate level, performance becomes only an enactment o f that text. But one would then also have to treat film as equally double-layered. Cf. Ryan (1991). Cf. Fludemik (1996).
360
Monika Fludernik
the chronology of events, the connection between characters and scenes and even the characters' discourse resist easy interpretation. (Compare, for instance, Caryl Churchill's The Skriker [1994]22 or Sarah Kane's 4.48: Psychosis [2000].) From a narcological point of view, it is therefore easier to tackle postmodernist drama than postmodernist fiction. All drama, in fact, needs to have characters on stage, and from this minimal requirement narrativity is immediately assured, if one defines narrativity as I do in Towards a 'Natural' NarratologyP A character on stage guarantees consciousness and usually speech; by dramatic convention, he or she is additionally located in a space-time frame that resembles human experience of space and time: the clock is ticking, time moves forward as the dramatic figure stands on stage, and this staging of the space-time continuum provides the concreteness of dramatic space which narratologists have traditionally found a necessary condition for narrativity.24 Drama as a narrative genre in fact demonstrates convincingly why academic history—dealing with nations and other more abstract entities— is not inherently a narrative genre. A fictional world, or action per se, does not provide sufficient criteria to define narrativity: embodied consciousness and human figures are indispensable constituents of narrativity. As Elam suggests, though, the human actor functions not only as an iconic sign, but may also take a diagrammatic or metaphoric role, 25 as in allegorical plays, for instance. In what way do characters on stage by their very presence evoke a fictional world? Let me take two examples from Beckett's minimalist dramatic art. Not I (1972) features a mouth from which a stream of words is pouring forth and an auditor downstage left reacting minimally to the woman's discourse. The mouth—synecdoche for the woman—constitutes a kind of narrator persona, a monologist, and therefore could be argued to instantiate a dramatic narrator figure. The auditor, on the other hand, does not narrate and reacts to the voice merely by tiny gestures. Hence, the auditor, by his mere presence, endows the play with a scenario of a specific time and place (the monologue as such seems to be indefinite). Likewise, in Krapp's Last Tape (1958), the character of Krapp in and by itself (rather than his story) 'makes' the world of the narrative. The dissociation of figure and narrative already observable in Not I also occurs in A 22 23 24 25
Cf. Fludernik (1996: 2 9 6 - 9 9 ) . Fludernik (1996: 3 5 1 - 5 2 ) . E.g. Prince (1982: 149). Elam (1980: 24).
Narrative and Drama
361
Piece of Monologue (1979), in which the 'speaker', described by the stage directions as having white hair and wearing a white nightgown and white socks, utters a soliloquy that is narrative but fails to cohere referentially, logically and situationally. Again, the situation of an old man in a nightgown is the primary visual impact which invokes a scenario that the audience will have to try and correlate with the figure's discourse, but will inevitably fail to do so. Having argued that drama is clearly a narrative genre for more than one reason, I would now like to propose a model for the narratological analysis of drama—a model that will also have some repercussions on the standard narrative typologies. I think we can take it for granted that the level of the represented world in drama closely corresponds to that in fiction or film. (Even though, unlike written or even spoken narratives, several dramatic actions can be portrayed on the stage simultaneously.) The main questions regarding the narratological analysis of drama therefore do not touch on the characters and event structure (the story level) or on the temporal dimensions of drama (anachronies, for instance, can easily be analysed with the help of the narratological toolbox, as in Pfister [1988]); they touch instead on the level of what I call the 'discourse' level of drama. In analogy with novels and films, the discourse level refers to the level of mediation. In a view that sees drama as exclusively a performative genre, it is what one gets to see and hear on stage that counts. As with film, there is one element located on the story level in Chatman's narratological model which departs significantly from the treatment it receives in fiction: setting. In drama, as in film, the setting is a visual 'given' and thus need not be evoked by means of description. 26 On closer analysis, however, this exception to the rule turns out to be less central or significant than might at first appear. After all, characters, too, are referred to linguistically (like places) in fiction, and they may or may not receive extensive descriptive elaboration. It is therefore not necessary to treat the lack of 'description' for settings, or the replacement of description by visual depiction, as an aspect of drama's irremediable alterity. Description can be treated as placeable on a scale of explicit to implicit in which the visual medium tends to be more explicit. Another difference between plays and novels lies in the fact that drama can present simultaneity in iconic fashion, as in Arthur Miller's Death of a
26
Compare Chatman's (1999) proposals about the different story components for film and fiction.
362
Monika Fludemik
Salesman (1949) when we see the parents in the living room on one side of the stage and the two boys in their bedroom on the other. Narrative in prose fiction has to support simultaneity by means of temporal juxtaposition. However, even in drama, characters in two settings which are simultaneously visible on stage do not usually act or converse at the same time; visual simultaneity is juxtaposed with temporal sequence. Drama, moreover, has difficulty summarizing condensed action. Whereas in fiction, a paragraph can take us from the hero's birth through his schooldays to the day on which he meets his future wife at the theatre, similar time jumps tend to be avoided in plays or require choruses, messenger reports and the like. Summary is therefore frequently replaced by ellipsis in drama: we simply jump to the next scene five years later and find out about interim developments through the dialogue of the characters. A further distinction between media can be noted in the uses of sound in film and drama, possibly echoing the distinction made between diegetic and non-diegetic sound in film theory. In films, music and speech can be part of the fictional world (characters talk with one another [dialogue], they sing or play records), but there is also speech and music outside the diegetic level (in Genette's terminology): voice-over narration, music that interprets the action on screen (such as the well-known suspense-raising tunes marking imminent attack in thrillers) and music that serves the purpose of entertainment, 'decoration' or, perhaps more basically, to avoid acoustic vacuity. In drama, on the whole, characters talk on stage and there may be diegetic music or noise, but there is little non-diegetic sound. The film genre departs from the conventions of the stage by providing a continuous soundtrack. In film and in the radio play, it is silence that is significant; in drama, it is sound. (Of course, these distinctions are being undermined by the now frequent influence of film techniques on staging.) One may approach a definition of the 'discourse' of drama by distinguishing those elements that constitute what would traditionally be called the 'plot' and those elements of dramatic narration that belong to the performance level of the staging. The playscript takes an intermediate position between these two levels. The plot is a level that allows realisation in several media (film, written narrative, plays, ballets, cartoons); the playscript—like the filmscript—already incorporates features that take account of the performance, indicating visual and acoustic orchestration. From the perspective of the audience, the playscript is only the underlying blueprint for the play as performative experience. For them, the 'discourse' of drama is equivalent to the performance, and it is a feature of
Narrative and Drama
363
drama that it does not have one discourse but that each performance has its discourse from which the audience extrapolates the plot and fictional world. In contrast to this radical reception-oriented view, one could also reduce the performative level to those features that differ exclusively in any one performance (staging, acting) and compare these features with the narrational level of Genette. In this schema, the selection of scenes, the temporal rearrangements and the mise en scene, in so far as they are provided in the stage directions, form a part of the discourse level, which can then be read off the playscript. By contrast, the visual qualities of the staging, the director's choice of props and costumes, the inclusion of music and of superimposed visual elements as well as the actors' interpretation of the characters and plot are equivalent to the narrational level, and this narrational level is in fact a performative level. If one reads playscripts, rather than seeing plays performed on stage, one only has access to the discourse level, i.e. the 'dramatic text'. This serves as a model for its medial realization in what is then called the 'performance text' 27 (which is accessible to study only if filmed). In reading a play, we imaginatively 'stage' it in our minds. Although this is a metaphorical way of expressing it, this reading process is different from that of reading fiction because— owing to the explicit staging information in the stage directions—it involves more visualisation than does novel reading. Having discussed two alternative models—the dramatic discourse as equivalent to what Elam calls the performance text and the dramatic discourse as equivalent to the playscript—it needs to be noted that these two models not only locate performance on different levels (on the discourse level and on the narrational level, respectively); they also place the dependence on the medium at different levels. If the discourse of drama is the equivalent of the playscript, this discourse already contains performance-related features typical of the stage; but it can also be analysed as the author's 'narration' that is to be realised in performance thanks to the medial narration effected by actors, director, costume designer, and so on. The discourse would then be both very similar to novelistic discourse and yet different because modelled on the constraints of performative realisation. By contrast, a definition of the discourse as performance, eliding the playscript, preserves a unity of discourse and narrational level (as in the 27
See Elam (1980: 3, 2 0 8 - 0 9 ) . A s Elam points out, the relative priority o f dramatic and performance texts is hotly contested among theatre scholars.
364
Monika Fludemik
novel), i.e. the narrative discourse of the novel and the dramatic discourse are the product of the narration and of what we see as readers and audience. Since anachronies are possible in film, drama and novels, such temporal rearrangements would have to be located on the plot level or a premedial level before (beneath) the dramatic text that is already dependent on the medium. A narratology based on the story and discourse distinction and that then treats film, drama and fiction as three differently mediated discourses, as does Chatman, thus requires a rewriting of the narrative deep structure to arrive at a non-medial discourse level that includes focalisation, analepses and possibly the choice of narrational agents. Another interesting question that can be asked in connection with the narratological model as a whole concerns reverse applications of the revised model to 'non-performative' narrative. Is there perhaps a performative level in fiction as well? Although the communicative level of narration (narrators addressing narratees) has been a hallmark of narratological theorizing, it needed the seminal work of Ansgar Nünnnig on metanarrative 28 to introduce the notion of performance into narratology. In fact, Nünning suggests that in some texts, the narrational performance of the narrator may outstrip the reader's interest in the story and provide for a primary aesthetic illusion.29 My own work on medieval narrative has opened up the performance factor even more decisively. In medieval romances, saints' legends and epics, the narrator figure is generally a bard performing a traditional tale. This bard constitutes himself not as the writer of the tale (he is not responsible for the text at all), but as an enactor of the tale qua narrative performer. Although such a performance is clearly less vivid than dramatic stagings, it displays significant parallels with drama. For one thing, the text of a medieval romance can then be seen as a kind of 'screen play' which will reach its fullest realization in performance. The antics of the bard (addressing the audience, calling for a cup of ale, etc.) may be regarded as a kind of equivalent of the stage directions of drama. Of course, the parallel ceases at this point because dramatic performance is so much more complex and varied than the bard's performance. What this introduction of a performative level does allow one to do, however, is to distinguish between a performer and a 'normal' narrator who is responsible for his text. This distinction can be seen as two ends of a scale, just like Nünning's and Sommer's diegetic and mimetic
28 29
Nünning (2001), (2004). Nünning (2001: 131).
365
Narrative and Drama
narrativity (in this volume). The medieval bard figure is one instance of an authorial substitute who is distinct from the author and yet not identical with the narrator function of the text. Such a constellation raises interesting questions regarding the author/narrator distinction in narratology, 30 but it also opens the way towards a general reassessment of the persons involved in the current narrative communication model. This model could be revised to include performance as one of the optional levels of narrative: author
reading public
performer
(actors)
audience
narratee
narrator Character