Strange Voices in Narrative Fiction 9783110268645, 9783110268577

Citation preview

Strange Voices in Narrative Fiction

Narratologia Contributions to Narrative Theory

Edited by Fotis Jannidis, Matı´as Martı´nez, John Pier Wolf Schmid (executive editor) Editorial Board Catherine Emmott, Monika Fludernik ´ ngel Garcı´a Landa, Peter Hühn, Manfred Jahn Jose´ A Andreas Kablitz, Uri Margolin, Jan Christoph Meister Ansgar Nünning, Marie-Laure Ryan Jean-Marie Schaeffer, Michael Scheffel Sabine Schlickers, Jörg Schönert

30

De Gruyter

Strange Voices in Narrative Fiction Edited by Per Krogh Hansen Stefan Iversen Henrik Skov Nielsen Rolf Reitan

De Gruyter

ISBN 978-3-11-026857-7 e-ISBN 978-3-11-026864-5 ISSN 1612-8427 Library of Congress Cataloging-in-Publication Data Strange voices in narrative fiction / edited by Per Krogh Hansen ... [et al.]. p. cm. ⫺ (Narratologia : contributions to narrative theory, ISSN 1612-8427 ; 30) Includes index. ISBN 978-3-11-026857-7 (alk. paper) 1. Fiction ⫺ History and criticism ⫺ Theory, etc. 2. Narration (Rhetoric) I. Hansen, Per Krogh. PN3331.S925 2011 808.3193⫺dc23 2011035615

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. ” 2011 Walter de Gruyter GmbH & Co. KG, Berlin/Boston Printing: Hubert & Co. GmbH & Co. KG, Göttingen ⬁ Printed on acid-free paper Printed in Germany www.degruyter.com



Contents Introduction ....................................................................................................

1

SYLVIE PATRON Homonymy, Polysemy and Synonymy: Reflections on the Notion of Voice ...................................................................................... 13 RIKKE ANDERSEN KRAGLUND ‘Alternate Strains are to the Muses dear: The Oddness of Genette’s Voice in Narrative Discourse .............................................................................. 37 HENRIK SKOV NIELSEN Fictional Voices? Strange Voices? Unnatural Voices? ............................. 55 PEER F. BUNDGAARD Significant Deviations: Strange Uses of Voice are one among other Means of Meaning Making ................................................................ 83 LARS-ÅKE SKALIN How Strange Are the “Strange Voices” of Fiction? ................................. 101 STEFAN IVERSEN States of Exception: Decoupling, Metarepresentation, and Strange Voices in Narrative Fiction ............................................................ 127 ROLF REITAN Theorizing Second-Person Narratives: A Backwater Project? ............... 147 MARINA GRISHAKOVA Toward a Typology of Virtual Narrative Voices ...................................... 175 MARIA MÄKELÄ Masters of Interiority: Figural voices as discursive appropriators and as loopholes in narrative communication ........................................... 191

VI POUL BEHRENDT & PER KROGH HANSEN The Fifth Mode of Representation: Ambiguous Voices in Unreliable Third Person Narration ............................................................. 219 BRIAN RICHARDSON Unnatural Voices in Ulysses ........................................................................... 253 Index ................................................................................................................ 265



Introduction How does narratology relate to, or not relate to, narrative strangeness, that is to defamiliarization, non-naturalizable aspects, non- and anti-mimetic textual strategies, and narrational ambiguity?1 To Russian Formalism and Czech Structuralism aspects like these were among the main concerns insofar as attention was paid to the differential formal qualities of literature, its ‘literariness’, and this with a special focus on narrative literature. By investigations of the strangely deviating formal aspects of narratives as compared to everyday reporting, these two ‘pre-narratological’ schools of criticism concentrated on the strangeness, or estrangement, both linguistic and cognitive, of artistic prose. When narratology became a distinct sub-discipline in the early 1970s, it inherited the formalistic stance, even if interest in estrangement and literariness perhaps lost some of its edge. Moreover, narratology from its very beginnings incorporated a model that would seem to be utterly foreign to formalism: a communicative model of literary narrative. Literary narratives were thus considered as simulations of natural, oral acts of communication. An internal global narrator or voice was supposed to be a constituent of all narrative, and readers were assumed to process artful fictional narratives in the same way and according to rules employed in processing everyday reports. This approach, however, has had some problems with accounting for the strangeness and anomalies of Modern and Post Modern writing, and as many sceptics have shown, not even classical realism would seem to conform to the standards set by oral or ‘natural’ storytelling. Thus, a certain urge to confront narratology with the difficult task of reconsidering a most basic premise in its theoretical and analytical endeavors has for some time been undeniable. In the last ten years or so, Scandinavian, or rather Nordic (including Finnish and Estonian), narratologists have been among the most active and insistent critics of the communicative model. They share a marked 1

The editors would like to express their gratitude to one of the anonymous readers of the book manuscript. The reader’s report contained a sharp and elaborated contextualization of the common grounds for the contributions to this volume. We have drawn heavily on the report in the first paragraphs of this introduction.

2

Introduction

scepticism towards the idea of using “natural” narratives as some kind of genetic model for understanding and interpreting all kinds of narratives, and for all of them the distinction of fiction is important.2 In this anthology, for the most part written by Nordic scholars, we present a collection of articles that deal with strange narratives, narratives of the strange, or more generally with the strangeness of fiction, and even with some strange aspects of narratology. Moreover, they all deal with some aspects of strange voices. This attempted focus on ‘voice’ may perhaps need an explanation. Prior to the publication of Gérard Genette’s Discours du récit in 1972, the word “voice” in literary criticism served mainly as an unsystematic, but handy, umbrella term for observations on an author’s style or rhetoric. Since Genette’s proposal to distinguish between mood and voice, voice has also acquired a more focused meaning in narrative theory in terms of answers to the question “who speaks?” In Genette’s theory, the voice of a narrative will always have to be the voice of a (fictive) narrator, and this voice is both “the generating instance of narrative discourse” (Narrative Discourse, 213) and “the final instant governing [it]” (157). Whenever a voice is narrating, a narrative exists, and vice versa: there is always a voice narrating the narrative, and that voice is the voice of the narrator. The term voice thus would seem to combine the stabilizing function of an always-necessary narrator with the seductive mimetic intuitivity of someone talking (to us). Certainly, for Genette, a voice like Proust’s, for instance, may mock any conventional vocal rules; the idea of really strange voices, however, is still foreign to his theory. Although being a rather simplified version of Émile Benveniste’s theory of enunciation, Genette’s voice concept is usually located within the framework of French structuralism. Hence, several general criticisms of theoretical and methodological assumptions of—or about—structuralism have reappeared on specific levels in a very lively debate concerning narrative voice. Nevertheless, the idea that the fundamental narrative voice is a narrator’s voice is shared by a wide range of scholars from Franz K. Stanzel to Gerald Prince to James Phelan; it rests upon the conception that narratives can and should be analyzed as acts of communication. MarieLaure Ryan’s 1993 definition of the term narrator in The Encyclopedia of Contemporary Literary Theory might still be representative in 2010: “[i]n a narrative text, the narrator is the speaking ‘voice’ which takes responsibility for the act of narration”. What Lars-Åke Skalin aptly calls the “stand2

Cf. Henrik Skov Nielsen 2004; Per Krogh Hansen and Marianne Wolff Lundholt 2005; LarsÅke Skalin 2005, Pekka Tammi and Hannu Tommola 2006, Marina Grishakova 2006, Marianne Wolff Lundholt 2008, and (in Danish) Rolf Reitan 2008.

Introduction

3

ard theory” of narrative transmission has as one of its key elements the belief that in a narrative somebody—the narrator—is speaking to someone—the narratee. Critiques of the “standard theory”, for the most part emanating from “strange” linguistic or philosophical positions, have had a hard time being heard. Ann Banfield, alluding in 1982 to both Käte Hamburger and Benveniste, famously stated that in a fictional narrative “No one speaks […], nor addresses anyone”—narrative fictions are not communicative texts. (Banfield 1982: 97.) Genette’s crushing response to Banfield in Narrative Discourse Revisited is well known, and mainstream narratology has followed in Genette’s steps—mostly by its silence. Andrew Gibson arrived at a similar conclusion, albeit from another direction, in that he questions the validity of the concept of voice by drawing upon tenets from deconstruction and post-structuralism. Stressing the purely metaphorical character of the term voice—“there are in fact no narrative voices and no voices in narrative” (Gibson 2001: 640)—he attempts to debunk the ideas of coherence, identity over time and stable hierarchies that he takes to be manifest in the concept of voice as envisioned by Genette. Gibson did not fare better than Banfield among mainstream narratologists. Pragmatic and/or cognitive approaches have been slightly more successful. Injecting doses of constructivism in their reconceptualizations of voice, they emphasize the role played by the reader in locating and figuring the voices of a text. Monika Fludernik, who ridicules the generalized postulates of the standard theory, talks about “the linguistically generated illusion of a voice factor”—the reader, rather than the narrator in the text, is the creator of voice effects. (Fludernik 1996: 344, italics added.) While she might in principle agree with Gibson that the concept of voice is metaphorical and risks blocking insights into more peculiar qualities of textual phenomena, she would seem to argue in favour of using the term: even though the constructivist and pragmatic approach “may not quite meet the standards of rigorous philosophical enquiry […] it works!” (Fludernik 2001: 710.) As a working term in narratology, voice continues to attract attention and stimulate discussion. This is so even though the term poses rather substantial challenges to anyone who wants to use it in a rigorous fashion. See for example two recent and rather different responses to this challenge: Liesbeth Korthals Altes’s “Voice, Irony and Ethos” (2006), and Richard Walsh’s chapter on voice in The Rhetoric of Fictionality (2007), and, in this anthology, Sylvie Patron’s “Reflections on Voice” and Rolf Reitan’s “Second Person Narrative: A Backwater Project?” Altes’ article was published in Stimme(n) im Text: Narratologische Positionsbestimmungen (Narratologia 10 (2006)), a rich anthology presenting several

4

Introduction

highly interesting papers on the challenges raised by the notion of voice in literary fiction. Faced with the polyvalent nature of the term and the need to account for some strange effects of vocal ironies, Altes argues in favour of a synthetic solution: a “constructivist angle is required, because it takes into account the dynamic processing of a text by readers”, while “the linguistic-rhetorical perspective is not simply to be dismissed, as readers form hypotheses about the intentionality of a text on the basis of its linguistic, stylistic and generic, as well as paratextual features […]” (166). Walsh, who finds the idea of voice as intuitively appealing as “story”— “commonsensical at first sight”, but “unexpectedly elusive on closer inspection”—proposes an alternative, transmedial perspective on the background of which he insists on differentiating between three diverging and partly overlapping metaphorical meanings of the term: voice as “instance” (a representational act), voice as “idiom” (an object of representation), and voice as “interpellation” (a representational subject position) (87). The current debate among Nordic scholars on the subject of strange voices in written, mainly literary narrative is international in nature and does not confine itself to work done in this region. Rather it draws upon and engages with work done in the growing international narrative community. We have chosen to include two of these voices from abroad: Brian Richardson, whose contribution to the issue of strange voices have been particularly important, and Sylvie Patron, whose defence of “older” theoretical positions like Hamburger’s, Kuroda’s, or Banfield’s have been an inspiration to many. What all contributors to this issue have in common is something that separates them from what arguably may be considered the dominant view of the connection between voice and narrative. Most of them take special interest either in strange voices in the literary text or in the strangeness of the voice of the literary text and they do so by questioning the widespread notion that literary narratives always abide by the same rules as real life narratives. All of them shed light on some of the analytical and theoretical consequences of working with narrative voices without necessarily presupposing that narratives are acts of communication—acts by which, as Phelan puts it, “somebody [is] telling somebody else on some occasion and for some purpose(s) that something happened” (Phelan 2005: 18). And all of them probably would share the view that while Phelan’s elaboration of “standard theory” offers a most valuable widening of the field of narratological studies, it still runs the risk of leaving a range of phenomena, crucial to the literary narrative, in the dark—for instance the fact that literary narratives excel in the construction of and playing with the strangeness of the written, narrating voice.

Introduction

5

Short Presentations of the Articles in this Issue It can be extremely difficult to distinguish between homonymy and polysemy when the meanings compared are neither very distant nor very close. Sylvie Patron, in “Reflections on Voice”, investigates the phenomenon of homonymy (one word, different meanings) in the case of narrative voice, i.e. voice in narrative theory, and of the synonyms of the word “voice” in the different narrative theories. The first part concerns voice in “communicational” theories of narrative. Here Patron analyzes the relations of homonymy, polysemy and synonymy surrounding the notion of voice in the theories of Genette, Chatman and Stanzel. She finds three notions of voice in Genette’s work that are different in nature and in origin, and argues that the relation between these notions is merely one of homonymy. She then compares the notion, or notions, of voice in Genettian and post-Genettian narratology with the narrative theories of Chatman (for whom voice is synonymous with verbal expression and/or narratorial expressivity) and Stanzel, who seldom uses the notion of voice, replacing it with that of mediacy, which he defines as the characteristic trait of the narrative genre as opposed to other literary genres. The second part of the article concerns Kuroda’s and Banfield’s critique of communicational theories of narrative in the case of fictional narrative. Kuroda does not employ the notion of voice, however, an implicit critique of the notion can be seen in his critique of John R. Ross’s performative analysis, according to which every sentence is derived from an underlying structure containing a performative verb in the first person. Banfield, for her part, demonstrates the inadequacy of what is known as the “dual voice theory” in a certain form of free indirect discourse (represented speech and thought in her own terminology). For Banfield, “As long as a third-person subjectivity is represented, no speaking voice can be realized”. (Banfield does not regard the author, who is responsible for the representation of speech and thought, as a “speaking voice”.) In conclusion, Patron questions whether it is advisable to retain the notion of voice in narrative theory, since eradicating homonymy and, incidentally, polysemy and multiple synonyms, in her view has to be the first logical and even ethical requirement of any theoretical language. In “Alternative Strains Are to the Muses Dear’: The Oddness of Genette’s Voice in Narrative Discourse” Rikke Kragelund Andersen analyzes The Seducer [Forføreren], a novel by Norwegian author Jan Kjærstad, in order to demonstrate how Genette’s focus on textual movements works in practice, and thus showing the necessity of Genette’s inclusion of time and level in the understanding of voice (a point often misrepresented in the reception of Genette). The text is a short study of Genette’s conception of

6

Introduction

narrative voice, which is shown to be far more open-ended, “low structuralist”, and less “categorical” than assumed among most critics of Genette’s narratology. Intervening in the current discussion of how one should define the field of unnatural narratives, Henrik Skov Nielsen in “Fictional Voices? Strange Voices? Unnatural Voices?” attempts an answer with a specific view to strange fictional voices. “What are the implications of treating all narratives, including fictional ones, as if they were real-life instances of narration?” Observing that some peculiar narrative forms such as first person present tense narration, you-narration, and paraleptic first person narration challenge common sense understandings of narrative and fictional voices, and drawing on discussions with Fludernik, Tammi, Phelan, and others, he attempts to establish the relationship between the specific strangeness of each of these narrative forms and the general conditions of fictional narratives. A close reading of Edgar Allan Poe’s “The Tell-Tale Heart”, which is “an especially interesting and challenging unnatural narrative that leaves almost all real world parameters behind”, concludes these discussions. It presents the reader to voices that are strange in that they conflate text and paratext, here and there, then and now, I and you. Finally, Skov Nielsen asks what consequences, if any, these peculiar forms and narratives have for our understanding of fictional narratives in general. In “Significant Deviations: Strange Uses of Voice are one among other Means of Meaning Making”, Peer Bundgaard’s approach rests on the tacit claim that there is no need to make any ontological fuss about the narrator, with whom “voice” is often identified, if by narrator we understand some substantial entity whose mode of being is essentially different from that of the formal narratorial devices of presentation. The narrator is neither a necessity nor an impossibility, but a semiotic and aesthetic resource at the disposal of any given author who may choose or not choose to make use of it in whatever way he may find convenient for his purposes. Thus, Bundgaard aims to shed some light on the semiotic function served by what is commonly called “voice” in narrative art: How do authors use voice, i.e. the narrating instance, in order to trigger specific meaning effects? Starting by framing the issue of voice and its semiotic function: viz. relative to what we mean by “natural”, “unnatural” or even “strange”, when we use such terms to qualify voice, he next uses a sample of examples from both first person and third person narratives to “flatfootedly”, as he says, show how meaning making obtains by virtue of specific uses of voice. Finally, the conclusion of the paper is that while it is certainly possible to determine the semiotic function of voice, whether strange or straight, it is so only on empirical, case-to-case grounds: there is no predetermined meaning effect linked to given variations in or applica-

Introduction

7

tions of voice (such as change from hetero- to homodiegetic narrator, from authorial to figural mode of narration, or free indirect speech). Voice is a meaning-shaping device that applies to all narrative situations. The semiotic result of its application is therefore also relative to the specific whole in which it is used. Finally, Bundgaard makes a general claim as to this relation: an important subset of meaning effects in narrative art can be characterized in terms of significant deviations from a given standard. In “How Strange Are the Strange Voices of Fiction?” Lars-Åke Skalin proposes an “Aristotelian” counter-model to standard narrotological theories and their common assumption of a general retrospective position of narration. Skalin argues that while the telling of natural narratives can be seen as motivated by the teller’s relation to something past, fiction generally refrains from such motivation. Fictional stories, progressing “blindly”, must violate the kind of expectations we have in meeting natural narratives. In the perspective of Skalin’s proposed poetics, standard narratological approaches commit “a narrative fallacy” in their dealing with literary fiction. Skalin demonstrates his points by exemplary close readings of selected passages from Dickens and Twain. In “States of Exception: Decoupling, Metarepresentation and Strange Voices in Narrative Fiction”, Stefan Iversen discusses narratives that defiantly challenges the possibility of answering the question “Who speaks?” The very strange voices of these texts emanate from a narrative that simultaneously evokes the mutually exclusive reading practices connected to fiction and non-fiction. In an attempt to describe how such exceptional “narrative situations” work, Iversen brings the concept of narrative voice in dialogue with the concepts of decoupling, metarepresentation and scope syntax as they are developed and used in evolutionary psychology. Investigating Leda Cosmides and John Tooby’s theories of metarepresentation and scope syntax, he discusses some important ramifications they might have for the distinction between fiction and non-fiction and in particular the possible uses of these theories when treating voices in fiction. It turns out that the treated amalgamation of theoretical approaches has potential consequences, not only for dealing with this specific type of strange narrative voice, but also for an understanding of what fiction might be and how it might function. Demonstrating his points by an aspectualised reading of the Danish text Selvmordsaktionen [The Suicide Mission] (2005), written by clausbeck-nielsen.net, Iversen—in a critique of both Cosmides and Tooby and of Zunshine (2006)—concludes that what narratology and theory of fiction might, and perhaps should, learn from Cosmides and Tooby’s work, is not so much that fiction can act as a training ground for real life competences, but rather that fiction—a certain state of excep-

8

Introduction

tion—is a crucial component of what enables real life competences to function in the first place. In “Second Person Narration: A Backwater Project?”, Rolf Reitan proposes a closer look at three very different perspectives on second person narratives. Brian Richardson (1991), Irene Kacandes (1994), and Monika Fludernik (1993, 1994b) have for some time been classical references, but according to Reitan they have never been seriously discussed. The article starts by discussing Kacandes’ concept of ‘radical narrative apostrophe’, and next analyses the three authors’ very different typological proposals. In particular, Fludernik’s famous typological “Diagram 1” becomes submitted to a lengthy close reading. Borrowing Richardson’s idea of a Standard Form second person narrative, the author returns to Butor’s La Modification to investigate the question of address (a pivotal question in Fludernik’s articles), which leads to a strict definition of a prototypical “genre” of Standard Form narratives. Passing through landscapes of fiction, apostrophe and postmodernism, the article addresses some tricky questions concerning self-address and Uri Margolin’s analytic schemes. At last, by way of proposing a much needed subdivision of the Standard Form, the strange “narrating voice” in La modification is discussed: not a narratorial voice, but a readerly voice created in the author’s writing. In “Toward a Typology of Virtual Narrative Voices Voices”, Marina Grishakova, for whom Mikhail Bakhtin is a major inspiration, offers some preliminary observations on the phenomenon of virtual voice, i.e. hypothetical discourse that might have been, but allegedly is not, pronounced by the purported speaker, being instead attributable to a different textual consciousness (“originator-consciousness”). The author develops the concept of virtual voice by using Gerald Prince’s idea of the ‘disnarrated’ and Marie-Laure Ryan’s idea of ‘virtual narrative’. Narrative voice manifests itself on the level of discourse: it functions as a means of perspectivization and interpretation (cf. Aczel’s definition of reading as an experience of overhearing voices). Virtual voice is not an unrealized or discarded possibility of plot development, but a realized discursive effect whose value is purely interpretive. For example, narration by the dead or ghostly narrator as a representation of either “natural” (everyday) or “unnatural” (otherworldly) events—all components of the textual world and thus part of the game of “make-believe”—is not meant to raise the (rather trivial) question of whether ghosts are able to speak or whether represented events indeed took place, but rather to shed new light on these events and occurrences of human life, to reveal new perspectives on life and death in general—or to pose questions concerning the nature of fiction. Otherwise, the narrator’s voice establishes a certain interpretive transworld relationship between the fictional actual world and the reader’s actual world. The phe-

Introduction

9

nomenon of virtual voices discloses a discrepancy between the mimetic and diegetic aspects of fictional storytelling: due to the existence of “virtual” counterparts of the representation, the storyworld opens indirect access to worlds much more extensive than the world it represents, which is its immediate communicative context (cf. MukaĜovský 1970). In “Masters of Interiority: Figural Voices as Discursive Appropriators and as Loopholes in Narrative Communication”, Maria Mäkelä addresses the peculiarities of figural voice in consciousness representation and the disruptive effect that this voice has on our narratological readings of minds and of narrative transmission. Instead of taking natural narratives as her starting point, she bases her arguments on a diachronic reconsideration of figural voices in literary fiction. First, she argues that the seeds of the “unnaturalness” of figural voice are already planted in the discourse of the early modern novel and especially in epistolary narration. The psychologically “natural” (mimetic) reading of fictional minds—favoured by both classical and cognitive narratology—is shown to foreground the “unrestrained” expression of thought and emotion via figural voice. Yet the conventions of consciousness representation bear in themselves the traces of mediacy: conventional frames of verbalization, intentional structure, and communicative features. She continues by tracing the evolution of this charged relationship between mediacy and immediacy in the third person narrative context and in instances of focalization and stylistically unmarked free indirect discourse. The key argument of this article concerns both narrative as well as thematic conventions. Characters who master their own interiority by appropriating narratorial conventions form a recurrent motive through the early modern, modern, and the late modern novel. This literary convention is—and at the same time, peculiarly, is not—in contradiction to the hierarchization and the naturalization of literary discourse prevalent in narratology. The “narrativizing focalizers” and the like also issue a threat to the currently much favoured rhetorical approaches to narrative fiction, since, more than often, fictional minds highlight the nature of consciousness representation as incommunicable communication. Finally, encouraged by the epistolary digression, The Princess of Clèves, Madame Bovary and Coetzee’s Disgrace, she sets forth a narratological take on consciousness representation as a derivative of, not speech, but writing (with a faint nod towards Bakhtin and Derrida). In “The Fifth Mode of Representation: Ambiguous Voices in Unreliable Third Person Narration”, Poul Behrendt and Per Krogh Hansen suggest that the problem of ‘who speaks, and who sees?’ requires more complex answers than usually given. Their aim is to show how two different parameters by which FID, and the related modes of representation of thought, are claimed to be analysed, do not necessarily support each other.

10

Introduction

The authors refer to these levels as the ‘level of discourse’ (or the ‘discursive level’) and the ‘level of narration’ (or the ‘narrative level’). The former is based on the grammatical parameters that usually determine the wellknown modes of representation (that is FID, ID and DD) as definite kinds of discourse: person, inquit, tense and syntactical markers; the latter is based on parameters of voice distributions and characteristics more or less colliding with the grammatically defined parameters of discourse. Behrendt and Krogh Hansen’s study encompasses two seemingly separate aspects of narrative unreliability—that of factual and that of ideological unreliability. Factual unreliability, usually regarded as unimaginable in third person narration, is examined in a reading of Isak Dinesen’s subtle and seductive play with FID in her classical short story “Sorrow-Acre”. Ideological unreliability is discussed in an investigation of ambiguous discourse in Henry James’ much-discussed third person narrated short story “The Liar”. A closer look at Wayne C. Booth’s controversial transposition of this story to a first person narration ‘in disguise’ serves to reveal the deficiencies of an analysis of unreliability carried through without any consideration of the flexible parameters of FID. Thus, in both Dinesen’s and James’ texts a double voicing is uncovered, which as yet has passed unnoticed, also by supporters of the dual voice hypothesis. In “Unnatural Voices in Ulysses: Joyce’s Postmodern Modes of Narration”, Brian Richardson draws on the work of a number of Joyce scholars to provide an inventory of the more unnatural or impossible kinds of narration in Ulysses. He points out how Joyce’s “subjectification” of traditional means of objective narration leads to a collapse of the opposition between objective third-person narration and subjective first-person discourse, arguing that this multiplicity of narrative practices that would seem to call for a narrator and an arranger or better, a supernarrator, is challenged by antithetical or incommensurate styles of narration that cannot be reasonably contained within a single figure, voice, or style. Moreover, as Ulysses progresses, Richardson finds ever more radical, unnatural, and impossible kinds of narration: Joyce both establishes and violates independent, autonomous sources of narration as he draws on and then moves away from a realist or humanist poetics. Virtually every binary opposition that mimetic narratives presuppose would seem to be undermined in this process. And repeatedly Joyce dissolves the notion of a single, human-like narrator who is telling a story. Richardson argues that for this text, we must replace the narrator concept with a broader, more fluent category, as the narrator—and even a single implied author—are supplanted by a depersonalized, deindividualized idea of narration. Concerning the larger issues of narratology, Richardson proposes to use Joyce’s example to corroborate a relatively new position, which would seem to be

Introduction

11

gaining ground in narrative theory: one can no longer presume the existence of a real human or a human-like narrator standing behind and above the discourse; in many works, there is only the discourse itself. Citing Monika Fludernik, Marie-Laure Ryan and his own Unnatural Voices, he defends the need for a theoretical model of narration that can embrace the full spectrum of narrative voices, from the most compellingly human-like to the most radically decentred and posthuman. The anatomy of Joycean strategies traced in this article confirms the importance of this position and firmly situates Joyce as a preeminent author of unnatural narratives. Per Krogh Hansen, Stefan Iversen, Henrik Skov Nielsen and Rolf Reitan



SYLVIE PATRON (Paris)

Homonymy, Polysemy and Synonymy: Reflections on the Notion of Voice [E]very concept should be understood as a stenogram for the questions it makes accessible. Thus it becomes clear that concepts with the same name may in fact be totally different, because they summarize different groups of questions; inversely, concepts with different names may be strictly equivalent, because it becomes clear that the questions they summarize are in fact the same.1

Introduction This article proposes an investigation of the phenomenon of homonymy (one word, different meanings) in the case of narrative voice, or of voice in narrative theory. It is an accepted fact however that it can be difficult to distinguish between homonymy and polysemy when the meanings compared are neither very distant nor very close. I shall also address the synonyms of the word “voice” in the different narrative theories considered. The article is divided into two parts. The first concerns voice in “communicational” theories of narrative (to use the term employed by S.-Y. Kuroda2). Here I analyze the relations of homonymy, polysemy and synonymy surrounding the notion of voice in the theories of Gérard Genette, Seymour Chatman and Franz K. Stanzel.3 I show that there are three different notions of voice in Genette’s work which are different in nature and in origin: voice as a category of narrative analysis (forming a system with the two other categories of time and mood, and including the sub-categories of time of narrating4, level and person); voice in the chapter on “Mood”, or voice 1 2 3 4

Milner 1995: 17-18. (Translation by Susan Nicholls). See Kuroda 1979b: 205-207 et passim. See Genette 1983, Genette 1988, Chatman 1980 and Stanzel 1984. [Translators note: All references to translated narratological theories in this article use the published English translations. In order to avoid potential confusion, please note that the

14

Sylvie Patron

as opposed to mood (reduced to the meaning of person); voice as a synonym of narrative enunciation or narration (confused with discursive enunciation, in other words with enunciation presupposing a speaker and an addressee, or a narrator and a narratee). In my view, the relation between these three notions is merely one of homonymy. I then compare the notion, or notions, of voice in Genettian and post-Genettian narratology with the narrative theories of Chatman (for whom voice is synonymous with verbal expression and/or narratorial expressivity) and Stanzel (who seldom uses the notion of voice, replacing it with that of mediacy, which he defines as the characteristic trait of the narrative genre as opposed to other literary genres). The second part of the article concerns S.-Y. Kuroda and Ann Banfield’s critique of communicational theories of narrative in the case of fictional narrative.5 Kuroda does not employ the notion of voice, however an implicit critique of the notion can be seen in his critique of John R. Ross’s performative analysis, according to which every sentence is derived from an underlying structure containing a performative verb in the first person. Banfield, for her part, demonstrates the inadequacy of what is known as the “dual voice theory” in a certain form of free indirect discourse (represented speech and thought in her own terminology). For Banfield, “As long as a third-person subjectivity is represented, no speaking voice can be realized”.6 It should be understood that, for Banfield, the author, who is responsible for the representation of speech and thought, is not considered to be a “speaking voice”. In conclusion, I question whether it is advisable to retain the notion of voice in narrative theory, since eradicating homonymy and, incidentally, polysemy and multiple synonyms is in my view the first requirement of the logic and even the ethics of any theoretical language. Voice in Communicational Theories of Narrative I will make the same distinction here as in previous work7 between communicational theories of narrative, where communication between a narrator and a narratee, be they real or fictional, is constitutive of the definition of narrative, and “non-communicational” theories, which could also

French narration occurs as both “narration” and “narrating” in English, énonciation as both “enunciation” and “enunciating”. Other synonymous translations are indicated in the body of the text.] 5 See Kuroda 1979b: 205-216, and Banfield 1982, specifically 183-223. 6 Banfield 1991: 26. 7 See Patron 2009.

Homonymy, Polysemy, Synonymy

15

be termed “poetic”8 theories of fictional narrative, where fictional narrative, or a certain type of fictional narrative, and communication are two mutually exclusive categories. Communicational theories of narrative are based on a postulate concerning the function of language without developing any linguistic analysis; the non-communicational or poetic theories of fictional narrative considered here are the product of linguists (Kuroda, Banfield) and are founded on a close analysis of the distinctive linguistic features of fictional narrative. The first four sections of this first part of the article will deal with voice in Genettian narratology, a quintessential communicational theory of narrative. 1. Voice is, first, the name given by Genette in Narrative Discourse to a category of narrative analysis. It is a term lifted from the grammar of verbs and used metaphorically: I walk, Pierre has come are for me minimal forms of narrative, and inversely the Odyssey or the Recherche is only, in a certain way, an amplification (in the rhetorical sense) of statements such as Ulysses comes home to Ithaca, or Marcel becomes a writer. This perhaps authorizes us to organize, or at any rate to formulate, the problems of analysing narrative discourse according to categories borrowed from the grammar of verbs, categories that I will reduce here to three basic classes of determinations: those dealing with temporal relations between narrative and story, which I will arrange under the heading of tense; those dealing with modalities (forms and degrees) of narrative “representation”, and thus with the mood of the narrative; and finally, those dealing with the way in which the narrating itself is implicated in the narrative, narrating in the sense which I have defined it, that is, the narrative situation or its instance, and along with that its two protagonists: the narrator and his audience, real or implied.9

As Genette himself emphasizes, the category of voice might just as easily have been called “person”, were it not for the psychological connotations of the word “person” and the fact that Genette saved the term for designating a sub-category of voice.10 It might also be considered that mode, as Genette defines it, following Littré (“Name given to the different forms of the verb that are used to affirm more or less the thing in question, and to express [...] the different points of view from which the life or the action is 8

“Poetic” should here be understood as “centred on the work of the author as poietes”. The term was coined by Kuroda (see 1979b: 130, 140). It follows certain propositions in Hamburger (see 1993 : 10-13) and was adopted by Banfield (see for example 2003: 479). In my view, several articles in the present issue (see especially Nielsen, Reitan, and Skalin) are variants of or move close to poetic theories of narrative fiction. They are nevertheless distinct by the fact that their reflexions on the foundations of narrative theory do not emanate “from a linguistic point of view”. 9 Genette 1983: 30-31. 10 See ibid. 31.

16

Sylvie Patron

looked at”11) would have been far more apt than voice and, a fortiori, than person for designating “a relation with the subject (and more generally with the instance) of the enunciating”.12 Voice is defined extremely vaguely as “the mode of action [of the narrative] considered for its relation to the subject [...] who carries out or submits to the action, but also [...] who reports it, and, if need be, all those people who participate, even though passively, in this narrating activity”.13 It is divided into three sub-categories respectively called the “time of the narrating”, the “narrative level” (thereafter simply “level”) and “person”, with quotation marks (or “relationship” as an occasional variant). Time is thus reintroduced as a “vocal” or “vocalic” determination.14 As for embedded or framed narrative (data of level, according to Genette), it could just as easily be considered under the category of mood (or more precisely, of distance) to the extent that it is a question of a form of direct discourse, which tells a story of a certain length. According to Richard Aczel, who sums up the position of the majority of narratologists on this subject, “[…] nothing—except perhaps confusion—is gained by rechristening person, time, and level as ‘voice’”.15 2. There is a second notion of voice in Genette, in his chapter on “Mood”, where voice appears in opposition to mood (or more precisely, to perspective or focalization). On point of view in fictional narrative, Genette writes: […] most of the theoretical works on this subject (which are mainly classifications) suffer from a regrettable confusion between what I call here mood and voice, a confusion between the question who is the character whose point of view orients the narrative perspective? and the very different question who is the narrator?—or, more simply, the question who sees? and the question who speaks?16

Here, the word “voice” designates what the chapter on “Voice” describes using the category of person. The question “who speaks?” does indeed require the answer of a “[n]arrator as a character in the story” or a “[n]arrator not a character in the story”17; “who speaks?” is therefore the 11 12 13 14 15 16 17

Ibid. 31, n. 9. The elision is Genette’s. Ibid. 31-32. Ibid. 213. My replacement “of the verb” with “of the narrative” in the quotation from Joseph Vendryès; my linking of the quote to the commentary that follows. The adjectives “vocal” and “vocalic” appear in Genette (1988: 78, 112, 113) solely in relation to voice 2. Aczel 1998: 468. To my knowledge, only Jean-Marie Schaeffer (see 1999: 722-727) uses voice in the sense of Genette’s voice 1. Genette 1983: 186. Ibid. See also ibid. 244-245, 248, where these two types of narrative, and therefore narrators, are named, respectively, “homodiegetic” and “heterodiegetic”.

Homonymy, Polysemy, Synonymy

17

equivalent of “is the person speaking a character in the story or not?”. It is noteworthy that there is no true parallel between the answer to the question “who sees?” and the reply to the question “who speaks?”. If, for Genette, the answer to the question “who sees?” might be “nobody” (nonfocalized narrative or narrative with zero focalization), the answer to the question “who speaks?” is nevertheless always “somebody” (there are no “non-vocalized”18 narratives or narratives with “zero vocalization”). The reduction of the notion of voice to a question of person, that is, to the homodiegetic or heterodiegetic nature of the narrator, is confirmed in Narrative Discourse Revisited, when Genette refers to “‘vocal’ selection”, “vocalic positions” or “vocalic choice”, and to “transvocalization” to designate the choice of homodiegeticity or heterodiegeticity, and the transition from one to the other in certain real or imaginary experiences.19 Voice is also reduced to person by Shlomith Rimmon-Kenan, who distinguishes between narrative levels on one hand, and voices on the other (“Narration: levels and voices”20), and who defines voice as “the narrator’s position vis à vis the story, in the spirit of Genette”.21 In general, this second notion of voice has met with far more success than the preceding one. As Monika Fludernik writes, “in the reception of Genette’s work […], the formula of ‘who sees’ (focalization) versus ‘who speaks’ (person, voice) […] has become the hub on which narratology is supposed to turn”.22 3. There is a third notion of voice in Genette, which is synonymous with narrative enunciation or what he calls “narration” (“narrating” in the English translation). At the end of his chapter on “Mood”, Genette writes: Again and again we have seen this subversion of mood tied to the activity, or rather the presence, of the narrator himself, the disturbing intervention of the narrative source—of the narrating in the narrative. It is this last instance—that of voice—which we must now look at for its own sake, after having met it so often without wanting to.23

In the first sentence of this quote, substituting the word “voice” for the word “narrating” would result in a statement with exactly the same meaning as the original. The same is true in other contexts, if not in all of them (for example: “Of course, these signs of the organization of the narrative 18 19 20 21 22 23

The verb “to vocalize” is used in Genette 1988: 112. See ibid. 78, 109, 112, 113. The adjective “vocal” is not used in quotation marks in Genette 2007: 352. Rimmon-Kenan 2002: 87-106. Rimmon-Kenan 1989: 159. Fludernik 2001: 620. Genette 1983: 211.

18

Sylvie Patron

are in themselves marks of the instance of narrating, which we will meet again as such in the chapter on voice”; “It seems that poetics is experiencing a comparable difficulty in approaching the generating instance of narrative discourse, an instance for which we have reserved the parallel term narrating”24; “[...] the main point of Narrative Discourse, beginning with its title, reflects the assumption that there is an enunciating instance — the narrating — with its narrator, and its narratee, fictive or not, represented or not, silent or chatty, but always present in what is indeed for me, I fear, an act of communication”25). We are therefore justified in speaking of synonymy, or at least of near synonymy, between voice and narrative instance or narrating. In the reception of Genette’s work, this synonymy has occasionally been pointed out. Rimmon-Kenan writes, for example, that: “Indeed, the category of narration (or ‘narrating’ in English translation) becomes one aspect of the récit (i.e. ‘voice’), so that the ternary model is turned in practice into a binary one.”26 Dan Shen goes one step further: “Although Genette stresses the importance of ‘narrating’ and expresses his regret at the fact that ‘until now the theory of narrative has been so little concerned with the problems of narrative enunciating’, his trichotomous distinction becomes a dichotomy between story and discourse in practice, as Rimmon-Kenan observes, ‘Indeed, the category of narration [etc.]’.”27 It is worth specifying that, for Genette, narrative enunciation or narration corresponds to what Émile Benveniste calls “énonciation de discours” (“utterance of discourse” in the English translation), which he associates with the use of tenses of discourse, notably the perfect tense, when it is a question of telling a story.28 For Genette, all narration is a form of “utterance assuming a speaker and a hearer”; it is a genre “in which someone addresses himself to someone, proclaims himself as a speaker, and organizes what he says in the category of person”29—whatever the verbal tenses used.30 However, if voice is synonymous with narrative enunciation or narration, in the sense of discursive enunciation or narration, emanating from a narrator who, in the case of fictional narrative, is different from the author, then all of the data referred to in all of the chapters, and not simply those in the chapter on “Voice”, can be considered as data pertaining to 24 25 26 27 28 29 30

Ibid. 78, n. 107, 213. Genette 1988: 101. Rimmon-Kenan 1989: 159. Shen 2001: 123. See Benveniste 1971: 211 and 308, n. 10. Ibid. 209. See Genette 1982: 138-143, and 1988, 99. For a different reading of Benveniste, see Kuroda 1979b: 211-212, 217-218, and 228, and Banfield 1982: 141-180. See also Patron 2011.

Homonymy, Polysemy, Synonymy

19

voice and therefore associated to the activity of a narrator. This is quite clearly the case in the footnote quoted previously: “Of course, these signs of the organization of the narrative [the marks of the end of the prolepsis] are in themselves marks of the instance of narrating, which we will meet again as such in the chapter on voice”31, or in the following remark from the chapter “Voice” (in the section concerning the functions of the narrator): “[...] the narrator can refer to [the narrative text] in a discourse that is to some extent metalinguistic (metanarrative, in this case) to mark its articulations, connections, interrelationships, in short, its internal organization [...].”32 It can also be seen in numerous passages from the chapters on “Order”, “Duration”, “Mood” and “Voice” (for example: “[...] the narrator, having evoked the quarrel between Achilles and Agamemnon that he proclaims as the starting point of his narrative [...] goes back about ten days to reveal the cause of the quarrel in some 140 retrospective lines [...]”; “[...] the narrator, forsaking the course of the story [...] makes it his business, in his own name and solely for the information of his reader, to describe a scene that at this point in the story no one, strictly speaking, is looking at”; “[...] thoughts and feelings are no different from speech, except when the narrator undertakes to condense them into events and to relate them as such”; “Here we are typically in external focalization, because of the narrator’s marked ignorance with respect to the hero’s real thoughts”; “The real question is whether or not the narrator can use the first person to designate one of his characters”33). As Nils Soelberg wrote in his review of Narrative Discourse Revisited, “it is the narrator who selects, temporalizes, modalizes, vocalizes and valorizes in and through his narrating.”34 Viewed from this perspective, everything is subordinated to voice, which means that all the “signs of the organization”, but also all the data of the organization of the narrative can be analyzed within a unified vision in the light of enunciative parameters (subject and situation of enunciation, in the sense of a subjectivity and a situation of enunciation differing from those of the author). 4. At this point, we are within our rights to wonder whether “voice 1”, “voice 2” and “voice 3” are three homonyms, or whether they correspond to three meanings of one polysemic word. According to their objectives and theoretical choices, commentators tend either to emphasize the unity of the meanings (polysemic tendency) or to see the differences as incommensurable (homonymic tendency). Genette quite naturally illustrates the 31 32 33 34

Genette 1983: 78, n. 107. Ibid. 255. Ibid. 36, 100, 171, 194, 244. See also Genette 1988: 35, 58, 73, 75, 81-82. Soelberg 1984: 126. (Translation by Susan Nicholls).

20

Sylvie Patron

first tendency in his introduction and in the first section of his chapter on “Voice” (to which the page numbers listed under “Voice” in the index refer): “[The last determinations are] those dealing with the way in which the narrating itself is implicated in the narrative, narrating in the sense which I have defined it, that is, the narrative situation or its instance, and along with that its two protagonists: the narrator and his audience, real or implied”;35 “[...] for us voice [...] will refer to a relation with the subject (and more generally with the instance) of the enunciating [...]”;36 “[There is] a tight web of connections among the narrating act, its protagonists, its spatio-temporal determinations, its relationship to the other narrating situations involved in the same narrative, etc. [...] we will look successively at elements of definition whose actual functioning is simultaneous: we will attach these elements, for the most part, to the categories of time of narrating, narrative level, and ‘person’ (that is, relations between the narrator—plus, should the occasion arise, his or their narratee[s]—and the story he tells).”37 In the same way, Fludernik emphasizes the links between voice 2 and voice 3 when she writes that: “The term voice in narratology has been coined in connection with the question ‘who speaks?’ (Genette), usually in distinction from the narrative categories of perspective or point of view (Genette’s Mood), which correlate with ‘who sees?’. An analysis of ‘who speaks?’ is patently predicated on a communicative model of narration in which the words of the text have to be uttered, i.e. enunciated, by a narrative instance, either the narrator or a character”.38 Aczel, on the other hand, illustrates the homonymic tendency (“[…] nothing—except perhaps confusion—is gained by rechristening person, time, and level as ‘voice’”39). For my part, I would tend to emphasize the differences rather than the similarities between voice 1, voice 2 and voice 3. Differences, firstly, in terms of their origin: a grammatical metaphor for voice 1 (the term of comparison being voice in the sense of grammatical voice, a homonym but not a synonym of the human voice); an interpretation of the difference between fictional narratives in the first person and in the third person for voice 2; Benveniste’s enunciative linguistics, including a revision of the opposition he established between historical enunciation and the discursive enunciation, for voice 3. Differences, too, in their nature: a formulation, and a relatively arbitrary one at that, of certain determinations of narrative for voice 1; the summary of a series of theoretical questions for voice 2 and voice 3. Voice 2 sums up the question of the primacy 35 36 37 38 39

Genette 1983: 31. Ibid. 31-32. Ibid. 31, 31-32, and 215. Fludernik 1993: 325. Aczel 1998: 468.

Homonymy, Polysemy, Synonymy

21

of “diegetic” (homodiegetic or heterodiegetic) in preference to linguistic factors in Genette’s work, and the question of the independence of person, in the diegetic sense, and of focalization in fictional narrative. Voice 3 sums up the question of the application of interlocutory linguistics to all fictional narrative.40 5. In Story and Discourse, Chatman does not use the notion of voice in the sense of Genette’s voice 1. He does not resort to grammatical metaphors to formulate the problems of analyzing fictional narrative (composed of story and discourse, in his own terminology). He considers narrative time under the category of discourse-time in the chapter “Story: events” (in the section devoted to time and plot).41 Neither person, nor narrative level, is assigned a separate section.42 The relations between Chatman’s voice and Genette’s voice 2 and 3 are rather more complex. Chatman first explains the notion of voice via a distinction: “Point of View and Its Relation to Narrative Voice” (the page numbers under “Voice” in the index refer to this section). Point of view is defined as “the physical place or ideological situation or practical life-orientation to which narrative events stand in relation”, and voice as “the speech or other overt means through which events and existents are communicated to the audience”.43 Other synonyms are proposed for the word “voice”: “expression”, “means of communication” (or, more exactly, “the medium through which perception, conception, and everything else are communicated”), or yet again, “verbaliz[ation]”.44 When Chatman writes that “[p]oint of view does not mean expression; it only means the perspective in terms of which the expression is made. The perspective and the expression

40

41 42

43 44

Taking advantage of rereading my article, I should add that Liesbeth Korthals Altes (see 2006: 168) speaks of what I term Genette’s voice 1 and voice 2 in terms of a “slide” from one to the other (and from there to “a psychological notion of personality”, a formulation which I find slightly problematic given Genette’s anti-psychological standpoint). However, she does not consider what I term voice 3 in Genette and makes no mention of the opposition between Benveniste’s historical enunciation and discursive enunciation (see ibid. 168 on Genette and 169-170, on enunciative linguistics, called “discourse linguistics”). I would like to thank the editors of Strange Voices for mentioning this article to me. Chatman 1980: 63. See also ibid. 80-84 (section devoted to the way temporal distinctions are manifested). Embedded or framed narrative is mentioned in the chapter on “Discourse: Covert versus Overt Narrators”, in the section devoted to the narratee (ibid. 254-256, 258). Person is briefly alluded to in the chapter on “Discourse: Non-narrated Stories”, in the section devoted to point of view and its relation to narrative voice (ibid. 155). Ibid. 153. Ibid. 154, 156. It can be noted that, in his introduction, Chatman also defines discourse as “the expression, the means by which the content is communicated” (ibid. 19).

22

Sylvie Patron

need not be lodged in the same person”45, he is close to Genette and his opposition between the questions “who is the character whose point of view orients the narrative perspective?” and “who is the narrator?”46 (as long, of course, as we overlook the fact that “who is the narrator?” means “is the narrator homodiegetic, or heterodiegetic?). Like Genette, Chatman uses the verb “to voice” (“limited third person point of view voiced by a covert narrator”, “limited third person point of view voiced by an overt narrator”47) in contrast to the verbs “to see” or “to perceive”, which are reserved for the character who is the source of point of view.48 Chatman, however, distances himself from Genette in his use of the terms “homodiegetic” and “heterodiegetic”. In fact, for Chatman, the adjective “homodiegetic” can only be applied to characters (including the “I as character” of first-person fictional narratives): “It makes sense to say that the character is literally perceiving something within the world of the work (‘homodiegetically’, as Genette would say)”.49 The narrator, for his part, is always heterodiegetic (including the “I as narrator” of first-person fictional narratives): “But what the narrator reports from his perspective is almost always outside the story (heterodiegetic), even if only retrospective, that is, temporally distant. Typically, he is looking back at his own earlier perception-as-acharacter. But that looking-back is a conception, not a perception. The completely external narrator presents an even more purely conceptual view. He never was in the world of the work: discourse-time is not a later extension of story-time”.50 This position is explained by the substantializing of the levels of story and discourse which, to begin with, were purely operative51 (Chatman also writes that “point of view is in the story [...], but voice is always outside, in the discourse”52). For Chatman, voice (by implication, the voice of the narrator) also contrasts with the absence of voice that characterizes the implied author: “Unlike the narrator, the implied author can tell us nothing. He, or better, it has no voice, no direct means of communicating. It instructs us silently, through the design of the whole, with all the voices, by all the means it has chosen to let us learn.”53 It may appear surprising that this instance with no voice, that is, with no means of communicating, is shown as an ad45 46 47 48 49 50 51 52 53

Ibid. 153. Chatman’s emphasis. Genette 1983: 186. Chatman 1980: 154. Ibid. 155. Ibid. Ibid. See ibid. 19-26, 31 and especially 37. Ibid. 154. Ibid. 148.

Homonymy, Polysemy, Synonymy

23

dresser in the diagram of narrative communication situations on p. 151.54 Should the implied author be seen as addressing the reader, or the implicit reader, using the narrator as intermediary? If so, what happens when there is no narrator (in “nonnarrated” stories, as Chatman terms them)55)? Chatman provides no answers to these questions. There is a second notion of voice in Chatman’s work which, strangely, includes the absence of voice or of “audibility” of the voice: “the concept of narrator’s voice (including its ‘absence’)”; “the concept of narrator’s voice—including the case where one is ‘not’ (or minimally) present”; “[...] the separate narrating voice may or may not make itself heard.”56 The voice (or audibility of the voice, or audibility of the narrator, or narratorpresence, or narrator-prominence—all these expressions are synonymous) is seen as a spectrum of possibilities going from non-audibility to maximum audibility via the adjunction of characteristic traits of audibility: “[...] the more identifying features, the stronger our sense of a narrator’s presence. The ‘non’- or minimally narrated story is simply one in which no or very few such features occur.”57 The characteristic traits of audibility are listed as follows: use of indirect discourse and free indirect discourse; presence of phenomena of presupposition; “limited omniscience” (points of view and shifting of point of view); descriptions; narrative summaries; reports of what characters did not think or say; interpretative commentaries.58 As is apparent, these traits are quite heterogeneous59; they involve concerns, which are rhetorical or stylistic (the opposition between direct and indirect discourse which are in fact called direct and indirect “styles”), pragmatic in the sense of pragmatic linguistics (presupposition), related to “classical” narrative theory (point of view, the opposition between scene and summary, etc.).60 Chatman does not include a properly linguistic dimension in his spectrum: the presence of first-person pronouns only appears at the end: “Commentary, since it is gratuitous, conveys the overt

54

55 56 57 58 59 60

This observation has already been made by Rimmon-Kenan: see 2002: 89. On Chatman’s diagram of the narrative communication situation, see also Brenkman 2000: 290-291, 292, 295. Chatman 1980: 34, 166-195. Ibid. 147, 151, 154. Ibid. 196. See also ibid.: 166-194, for the definition and description of “nonnarrated stories” (“written records”, “speech records”, “interior monologue”, etc.). Aczel 1998: 470 reproaches Chatman for not having included style, defined as “narratorial idiom”, in his list of characteristic traits of voice. The addition of style, however it is defined, would only extend their heterogeneity. For a non-expressive definition of style, see Philippe (2005). See Patron 2009: 68-76 for a critical presentation of each of these characteristics.

24

Sylvie Patron

narrator’s voice more distinctly than any feature short of explicit selfmention.”61 6. Stanzel, in his A Theory of Narrative, does not use the notion of voice in the sense of Genette’s voice 1 either. He does not resort to grammatical metaphors for the formulation of the problems of analyzing fictional narrative discourse. He addresses the problem of narrative time in chapter 2, “Zero grades of mediacy: synopsis, chapter heading, outline”62, and touches briefly on it, under the category of “narrative distance”, in Chapter 4, “The opposition person”, in the sections on the “embodiment” of the narrator in first-person fictional narrative and its consequences for the interpretation of narrative.63 It should be noted that Stanzel uses the term “person” in a similar sense to Genette (“identity and non-identity of the realms of the narrator and the fictional characters”).64 However, for Stanzel, person forms a system with perspective and mode, which constitute the two other aspects or elements of the narrative situation. There is no relation, in his view, between person, time and narrative level, nor does he devote separate chapters or sections to the problem of narrative level. Stanzel begins the first chapter of his work with the affirmation that “[w]henever a piece of news is conveyed, whenever something is reported, there is a mediator—the voice of a narrator is audible”, only to add immediately afterward: “I term this phenomenon ‘mediacy’ (Mittelbarkeit). Mediacy is the generic characteristic which distinguishes narration from other forms of literary art.”65 What Stanzel says about mediacy can therefore be applied to voice, or to the audibility of voice: it has “two manifestations”: the “overt mediacy of narration and that covert or dissimulated mediacy which produces the illusion of immediacy in the reader”.66 Here we can recognize Chatman’s voice 2, only minus any “absence of voice”, which corresponds to narratives which Stanzel does not consider, or does not consider from this perspective, and minus the list of characteristics traits of audibility. In the absence of such a list, statements affirming that the “voice of a narrator” is “still” or “no longer audible” in such and such a narrative, or “more audible” in a particular sentence of a narrative than 61

62 63 64 65 66

Chatman 1980: 228. In his conclusion, Chatman wonders what narrative theory might be able to do “with recent views of pronominal structure and the whole subject of deixis” (ibid. 264). Stanzel 1984: 23-28, 32, 40-45. Ibid. 93-94, 95-97. Ibid. 79. Ibid. 4. Ibid. 141. “Covert or dissimulated mediacy” corresponds to the figural narrative situation, that is, to “point-of-view narratives”.

Homonymy, Polysemy, Synonymy

25

in another, seem even more arbitrary in Stanzel’s text than they do in Chatman’s.67 Voice and Kuroda, and Banfield’s Critique of Communicational Theories of Fictional Narrative 7. In his articles on narrative theory, Kuroda does not use the notion of voice, however, a critique of the notion is implied in his critique of John R. Ross’s performative analysis applied to sentences of point-of-view narratives.68 Kuroda begins by establishing a system of relations between: – the communicational theory of linguistic performance: according to this theory, every linguistic performance is an act of communication, thus implying an addressor, an addressee, a message and a code common to both addressor and addressee; – the communicational theory of linguistic competence: this theory includes the communicational theory of linguistic performance; it was represented, in its day, by Ross’s performative theory, also termed “performative analysis”; according to this theory, every sentence is derived from an underlying structure containing a performative verb in the first person; – the communicational theory of narrative: narrative being a form of linguistic performance, the communicational theory of narrative must also be a form or an application of the communicational theory of linguistic performance. Kuroda quotes the following sentence from Roland Barthes’ “Introduction to the Structural Analysis of Narrative”: “[...] narrative as object is the point of a communication: there is a donor of the narrative and a receiver of the narrative. In linguistic communication, je and tu (I and you) are absolutely presupposed by one another; similarly there can be no narrative without a narrator and a listener (or reader).”69 For Kuroda, this quote is an example of a statement of the most explicit and frank kind, according to which a theory of narration having recourse to the no67

See ibid. 9, 58-59, 112, 198. As far as posterity is concerned for Genette, Chatman and Stanzel, I would subscribe to the assertion made by Korthals Altes (2006: 167) according to which “[v]oice in narratology has been made to mean everything and its opposite”. However I am surprised that after making such an assertion, she herself endeavours to rehabilitate the idea, which competes with much better constructed linguistic or pragmatic formulations (for example “linguistic subjectivity”, “represented speech and thought” [Banfield], “irony as echoic mention” [Sperber and Wilson]). 68 See Kuroda 1979b: 205-210. 69 Barthes 1973: 110. In another passage of his article, Barthes introduces the term “narrataire” (“narratee”) to designate the addressee of the narrator. Kuroda does not adopt this term although he occasionally mentions an “imaginary addressee” differing from the reader (see Kuroda, 1979b: 223).

26

Sylvie Patron

tion of narrator must have its theoretical basis in the communicational theory of linguistic performance.70 Narrative theory having recourse to the notion of narrator (or the “narrator theory of narration”71) thus implies that every sentence of the narrative should be considered as a message communicated by the narrator to the addressee (narratee or reader). This hypothesis is compatible with performative analysis. It would suffice to say that the narrator is the referent of the “I”, the subject of the performative verb underlying each sentence of the story: I (the narrator) assert, tell, etc. to you (the narratee) that S.

This is precisely the claim made by Genette regarding what, in his view, constitutes the invariant element of the narrative situation: “to wit, the presence (explicit or implicit) of the ‘person’ of the narrator”: “This presence is invariant because the narrator can be in the narrative (like every subject of an enunciating in his enunciated statement) only in the ‘first person’. [...] Insofar as the narrator can at any instant intervene as such in the narrative, every narrating is, by definition, to all intents and purposes presented in the first person [...].”72 A problem is raised, however, by sentences from third-person fictional narratives which are supposed to represent the point of view of one or several characters (that is, the thoughts, feelings or perceptions of the character or characters). Obviously, the representation of the profound structure conforming to performative analysis—with a first-person subject and second-person indirect object of a performative verb such as “assert”, “tell”, etc.—does not suit this kind of sentence if the first-person subject is to be interpreted referentially as the source of point of view (that is, of thoughts, feelings or perceptions): *I (referentially a character of the story) thought, felt, perceived, etc. to you (the narratee) that S.

This is the claim made, in a different way, by Chatman and Stanzel about the character who is the source of point of view: “It is simply a mistake to argue that Lenehan is in any sense the ‘narrator’ of ‘Two Gallants’. When he speculates, reminisces, or whatever, he is not telling a story to anybody, not even himself”73; “By contrast [with the teller-character], a reflector70

See ibid. 206. Other examples of open, explicit communicationalist statements may be found in Genette (see 1983: 260, and 1988, 101) and Rimmon-Kenan (see 2002: 88-89). 71 Kuroda 1979b: 206. 72 Genette 1983: 243-245. Other non-theorized expressions of the correspondence between narratological analysis and performative analysis can be found in the work of Gerald Prince (see 1982: 7-8 and 16-17) and Mieke Bal (see 2004: 21-22). For a theorized analysis of the correspondence, see Marie-Laure Ryan 1981. 73 Chatman 1980: 198.

Homonymy, Polysemy, Synonymy

27

character reflects, that is, he mirrors events of the outer world in his consciousness, perceives, feels, registers, but always silently, because he never ‘narrates’, that is, he does not verbalize his perceptions, thoughts and feelings in an attempt to communicate them.”74 Kuroda also considers the possibility of combining performative analysis with an analysis through direct or indirect speech, which would lead to representations of the following type: I (the narrator) assert, tell, etc. to you (the reader): John thought, felt, etc.: “S”. I (the narrator) assert, tell, etc. to you (the reader): John thought, felt, etc. that S.

The question boils down to knowing how the narrator can make assertions concerning the propositional content of inner acts such as thinking, feeling, etc. As Kuroda writes, “[g]rammatically speaking, we do not know the grounds on which the narrator makes his assertions about the inner acts of characters. But we must, so long as we follow the performative analysis, accept that he does make such assertions. One could just assume that he directly perceives the mental states of his characters. Thus we are led to the notion of an omniscient narrator.”75 For Kuroda, the need to admit, first, the theoretical existence of an omniscient narrator, and second, that the omniscient narrator speaks using a peculiar syntax of his own, is too high a price to pay for accounting for facts which could easily be explained in another way.76 8. In Unspeakable Sentences: Narration and Representation in the Language of Fiction, Banfield demonstrates the inadequacy of what is known as the “dual voice theory” in a certain form of free indirect discourse, in the third person and in the past tense (represented speech and thought in her own terminology). Her demonstration involves three stages.77 First stage: Banfield examines the similarities and differences between the forms of direct and indirect discourse and establishes the impossibility of switching from one to the other by any plausible grammatical transformation (this point is based mainly on the existence, in direct discourse, of elements and constructions which become inacceptable in indirect discourse). Second stage: 74 75 76

Stanzel 1984: 144. Kuroda 1979b: 210. See Kuroda 1979a: 190-199, 1979b: 216-223, and 1979c: 10-11. See also Patron 2009: 177202 for an overview of Kuroda’s narrative theory. 77 This passage summarizes part of the chapter on Banfield in Patron 2009: 203-218. My presentation differs markedly from Aczel’s (1998: 484-488), which does not take Banfield’s linguistic argument into account, and which uses substantialized narratological concepts (voice, focalization). It also differs from that by Korthals Altes 2006: 173, which does not take Banfield’s linguistic argumentation into account either, and adopts Fludernik’s (1993) assertions on pragmatics “surpassing” linguistics without critical appraisal.

28

Sylvie Patron

Banfield replaces the transformational hypothesis with a syntagmatic one, that is with a series of different rewriting rules for direct or indirect discourse, introducing a new initial symbol, E (for “expression”, in the narrow sense of “expression of subjectivity”), to replace S as the initial symbol of the basic rules. Unlike S, E is not recursive. It is associated with a principle of interpretation, which Banfield formulates in the following way: 1 E/1 I: For every expression (E), there is a unique referent of I (the SPEAKER), to whom all expressive elements are attributed, and a unique referent of you (the ADDRESSEE/HEARER).78

This principle explains why the referents of the pronouns “I” and “you” can be different in the two propositions of a sentence in direct discourse, but must remain the same in the two propositions of a sentence in indirect discourse. It also explains why the “style”, that is, the series of expressive elements and constructions, as well as the language or dialect can be different in the two propositions of a sentence of direct discourse, but must be of the same kind in the two propositions of a sentence in indirect discourse. The expressive elements and constructions are defined in a strict manner on exclusively syntactic grounds. It is a matter, firstly, of non-embeddable elements and constructions appearing in the second proposition of direct discourse, but not in that of indirect discourse (interjections, exclamatory sentences and exclamatory constructions with no verb, incomplete sentences); secondly, of embeddable elements which, when they appear in the second proposition of indirect discourse, are always attributed to the speaker of the whole, that is, to the person quoting and not to the person quoted (first- and second-person personal pronouns, deictic adverbs of time and place, “qualitative nouns” such as “idiot” in “that idiot of a doctor”, evaluative adjectives like “poor”, terms expressing relationship, such as “Mummy”). In a third stage, Banfield observes that represented speech and thought are distinct from forms of direct discourse and indirect discourse, and cannot be derived from the underlying structures of the two forms of reported speech either. Like the sentences of direct discourse, the sentences of represented speech and thought are E expressions: they have the same syntactical properties, notably that of not being able to be embedded in other sentences; they can contain interjections, exclamatory sentences and constructions, incomplete sentences, etc. This is evident in the following example, taken from Mrs Dalloway: “No, no, no! He was not in love with her any more!”79 In sentences of this sort, the previously asserted relation between the first person and the expression of 78 79

Banfield 1982: 57. Woolf 1996: 85.

Homonymy, Polysemy, Synonymy

29

subjectivity no longer holds. It is therefore necessary to reformulate the principle “1 E/1 I”, by decomposing it into two principles, of which only one relates the first person and the expression of subjectivity. 1 E/1 SELF: For every node E, there is at most one referent, called the “subject of consciousness” or SELF, to whom all expressive elements are attributed. That is, all realizations of SELF in an E are coreferential. Priority of SPEAKER. If there is an I, I is co-referential with the SELF. In the absence of an I, a third person pronoun may be interpreted as SELF.80 Following the first principle, the personal pronouns “he” or “she” can take on the role that ordinary discourse normally reserves for the pronoun “I”, which is the role of source of subjectivity. This is the case, for example, in “No, no, no! He was not in love with her any more!”. Following the second principle, the presence of a speaker who refers to him or herself as “I” necessarily implies that of a subject of consciousness co-referential with “I”; but in the other sentences of represented speech and thought, those which possess a subject of consciousness referred to as “he” or “she”, the first person is excluded. This can be verified using simple tests. In the sentence, or sentences, “No, no, no! He was not in love with her any more!”, the exclamation is attributed to “he”, referring to Peter Walsh. But if we add an “I”, which would produce something along the lines of: “No, no, no! He was not in love with her any more, nor with me!”, then the exclamation must be attributed, not to Peter Walsh, but to the referent of the first-person pronoun; there is no longer any trace of a subject of consciousness referred to as a third-person pronoun. Banfield concludes: “[…] linguistic argumentation has led to some surprising conclusions for literary theory. Since no first person may appear in represented speech and thought except one interpretable as the E’s SELF and since that first person must also appear in any parenthetical attached to the represented E, this means that represented Es cannot be simultaneously attributed to a covert or ‘effaced’ narrator. Rather than being narrated, consciousness in this style is represented unmediated by any judging point of view.”81 This approach to free indirect discourse may appear counter-intuitive to the extent that it is at odds with the usual way in which theoreticians and critics account for the same sentences or passages. According to the usual approach, free indirect discourse is founded on the mixing or fusion 80

Banfield 1982: 93. The term “subject of consciousness” comes from Kuroda (see Kuroda 1979b: 208, and Banfield 1973: 30, n. 22). On the level of the narrative or of the passage of the narrative, the subject of consciousness who is co-referential with a third-person pronoun is the equivalent of the focal character, the reflector-character or the character who is the source of the point of view in the theories of Genette, Stanzel and Chatman. 81 Banfield 1982: 97.

30

Sylvie Patron

of two voices: the voice of the author or narrator, and the voice of the character. Apart from the epistemological vagaries surrounding the notion of voice, the dual voice approach has the following drawbacks: – it places all the elements attributed to the narrator, i.e. the third person and the past (for example, the “he” and the past in “No, no, no! He was not in love with her any more!”), as well as the expressive elements defined by Banfield on a strict syntactical basis (for example, the interjection “No, no, no!” and the exclamatory modality), on the same level. To evaluate whether this treatment is justified, it is worth considering the consequences, the most general one being that the domain of expressivity or of the expression of subjectivity becomes unlimited. It includes, notably, all personal pronouns and all verb tense forms in the language. On the other hand, the notion of expression of subjectivity is rendered useless; in particular, it can no longer be used to explain the syntactical differences between direct discourse and indirect discourse; – it ignores the principle of the priority of the first person as subject of consciousness, as well as the test allowing the justification of this principle (“No, no, no! He was not in love with her any more!”, i. e. “No, no, no! He was not in love with her any more! thought Peter” vs. “No, no! He was not in love with her any more, nor with me!”, i. e. “No, no! He was not in love with her any more, nor with me! thought I”); – it ignores, therefore, the dissymmetry between first-person free indirect discourse sentences (with a first-person subject of consciousness coreferential with the “I” of the tag, if there is one) and third-person free indirect discourse sentences (with a third-person subject of consciousness co-referential with the “he” or “she” of the tag, if there is one, the presence of which would exclude that of a speaker referred to as “I”). The usual approach, via the dual voice, can also be criticized for: – having a vague, informal view of the duality or mix of voices; – in particular, for not seeing that when a third-person subject of consciousness is found in one or several sentences of a first-person fictional narrative, on the syntactical level it is a matter of sentences of third-person free indirect discourse (for example: “His life had been confused and disordered since then, but if he could once return to a certain starting place and go over it all slowly, he could find out what thing was.…”82). Nothing in fact prevents an author from using sentences of third-person free indirect discourse in a fictional narrative which can be characterized, globally, as a first-person fictional narrative.83 82 83

Scott Fitzgerald (1991: 86). This remark can be extended to whole passages of narratives. In my view, narrative theory has no need of the notion of “impersonal voice” (see Henrik Skov Nielsen 2004, specifically 138-143, and also in the present issue) for an explanation of such passages. It is worth com-

Homonymy, Polysemy, Synonymy

31

It may of course be thought that the role of syntax in the analysis of free indirect style is more limited than Banfield claims, and that certain properties of sentences in free indirect style are the remit of semantic or pragmatic studies (which Banfield in fact recognizes herself in the case of irony84). This does not mean however that Banfield’s syntactic analyzes can be ignored, any more than the linguistic and literary conclusions she draws from them. 9. It can be seen that, in her critique of the dual voice theory, Banfield uses the notion of voice in the sense of the presence of a speaker who is referred to, or could be referred to, as “I” and to whom all the expressive elements of the sentences are attributed. This is the meaning to which most of the occurrences of the word “voice” refer in Unspeakable Sentences85 (leaving aside those occurrences where the word is mentioned— rather than used—as in the “dual voice theory”86). This meaning is encountered in “L’écriture et le non-dit”, an article written several years after Unspeakable Sentences, in which Banfield replies to the critiques of the French linguist Oswald Ducrot. The following are significant examples: “[...] the crucial issue does not turn on whether an utterance is restricted to one or more than one ‘voice’ but on whether every utterance must have a ‘voice’”87 (in this quotation, the first occurrence is a mention of Ducrot, while the second could be paraphrased by “a voice according to the strict definition of the word”); “As long as a third person subjectivity is represented, no speaking voice can be realized”88; “That the author’s role in writing, in composing, need not, cannot, receive a linguistic representation unified around the notion of a voice does not, however, seem to have occurred to Ducrot, and he can conceive of him only as in some way ‘speaking’ in the text, even if he never says ‘I’, unless it is to represent a fictional persona”.89 It can also be perceived that the author, who is responsible for the third-person representation of subjectivity or for the creation of a narrator in a first-person fictional narrative, is never considered as a voice, or as possessing a voice: “All parts of the text are composed by the author, but their relation to their creator is different from their textual relation to any

84 85 86 87 88 89

paring the use of this notion taken from Maurice Blanchot in Nielsen (see 2004: 139-140 et passim) and Banfield (see 1985: 8-13). See Banfield 1982: 220-223. See also Fludernik 1993: 350-356, 359, Gibson 1996: 149-151, and Korthals Altes 2006: 177-181 et passim. See Banfield 1982: 10, 184, 188, 222, 250, 279. See ibid. 69, 185, 189, 211. Banfield 1991: 23. Ibid. 25. Ibid. 27.

32

Sylvie Patron

fictional subject of consciousness or speaker. The text speaks, not the author in it. He has written it, which [...] is a very different act from speaking”90; “[...] an author who is not directly embodied in a first person, as a speaker is in his speech, may manipulate language in accordance with the possibilities inherent in it—and at the same time respecting the limits it imposes, in order to represent a fictional subjectivity—but he does not speak in it. He writes, rather, and in writing disappears”.91 A connection can be made between Banfield’s author and Chatman’s implied author (as long as we overlook the fact that Chatman’s implied author is considered as an agent of his “diagram of the narrative communication situation”). There is a second, more restrictive notion of voice in Banfield’s work, which is voice in the sense of the presence of a speaker who is referred to as “I”, and to whom all the expressive elements of the sentences are attributed within a situation of communication, be it real or fictional. This is the meaning to which the following occurrences of the word “voice” refer in Unspeakable Sentences: “What seems to be excluded from represented Es is the rendering of the speaking voice”92 (meaning: the rendering of the speaking voice by indications of pronunciation); “The language of narration has no accent, because its narrator has no voice”93 (meaning: the language of narration in third-person and “classic” first-person fictional narratives; “[...] the only place for representations of the speaking voice within a narrative text is really in some sense outside of it. Its boundaries are set by a graphic notation with no phonetic realization—inverted commas [...]”.94 Voice can only be used in this sense for characters for which the author creates speeches by means of direct discourse and for the narrator of the type of first-person fictional narrative which Banfield calls “skaz” or “skaz-narrative”.95 10. A final remark: Banfield is right to say, in the article already cited, that “the choice between the two competing models [Ducrot’s and her own] will depend on argumentation appealing to the evidence”, but that “the issues have taken on an ideological coloring” and that “it suffices to name one position ‘unitary’ and the other ‘polyphonic’ for it to appear that the sides are chosen”.96 This promotion of polyphony rather than what could, 90 91 92 93 94 95

Banfield 1982: 211. Banfield 1991: 27. Banfield 1982: 116. Ibid. 178-179. Ibid. 250. See ibid. 171-179 et passim. I pointed out in Patron 2009: 222-226 that the attribution of certain first-person fictional narratives as skaz-narratives using Banfield’s criteria was relatively problematic. 96 Banfield 1991: 28.

Homonymy, Polysemy, Synonymy

33

by antithesis, be termed “monophony” appears in all recent discussions of the notion of voice, whatever the theoretical allegiances of the participants. Marc Blanchard criticizes the “reverence to an individual subject who remains unitary and whole”97 in Genettian narratology. Andrew ", who quotes Blanchard abundantly, provides the following commentary on Genette’s claim concerning narration which, in his view, is always in the first person: “The argument massively if somewhat abstractly consolidates the idea of person as a unifying, homogenizing, singular presence within a narrative text. Genette’s case reverts to a theological conception of the text in which all separate narrative instances, including instances of voice, are governed by a single authoritative, ‘higher’ voice”98 (Gibson on the contrary gives a lot of space to Banfield’s theory). Aczel criticizes the narratological approach to voice as “a unified speaker position” and pleads for a notion of voice as “a composite configuration of quoted speech styles”99; he then aims to show “how the essentially Bakhtinian notion of voice as composite and quotational [...] can provide a fruitful response to the critical anxieties of deconstruction concerning voice”100 (via deconstruction, Aczel is essentially targeting Gibson). Lastly, Korthals Altes, who refers back to both Gibson and Aczel, also concludes unsurprisingly that “[t]he attention for voice and its slipperiness should be part of the training in literary—and more general, cultural—competence, as the capacity to understand complex, polyphonic and dialogic language use, and how individuals and interpretive communities react to it”.101 Conclusion The phenomenon of homonymy has long been an object of concern for grammarians and logicians alike, who consider it a failure of languages particularly in technical languages such as those of science or philosophy. Homonymy and, incidentally, polysemy, are opposed to the ideal of a well-made language which would associate one form to one meaning; they lead to a penury of designations and thus generate confusion. Synonymy is not much better, as it introduces unnecessary redundancy and thus also runs counter to logical requirements. Homonymy between the different uses of the word “voice” for Genette, polysemy for Chatman and in a different way Banfield, multiple synonyms, an absence of strict antonyms: 97 Blanchard 1992: 64. 98 Gibson 1996: 145. 99 Aczel 2001: 598 (about Aczel 1998). See 100 Aczel 2001: 598. 101 Korthals Altes 2006: 191.

also ibid. 493-495.

34

Sylvie Patron

all the data listed in this article should, I believe, lead to abandoning the notion of voice in narrative theory. Without any doubt, the possibility of non-contradictory discourse and full comprehension between theoreticians is founded on the univocal meaning of words and sentences. Translated by Susan Nicholls

References Aczel, Richard 1998 “Hearing Voices in Narrative Texts”, New Literary History, 29.3, 467-500. 2001 “Understanding as Over-hearing: Towards a Dialogics of Voice”, Voice and Human Experience, New Literary History 32.3, 597-617. Bal, Mieke 2004 [1985] Narratology: Introduction to the Theory of Narrative, Toronto, Buffalo, London: University of Toronto Press. Banfield, Ann 1973 “Narrative Style and the Grammar of Direct and Indirect Speech”, Foundations of Language 10, 1-39. 1982 Unspeakable Sentences: Narration and Representation in the Language of Fiction, Boston and London: Routledge and Kegan Paul. 1985 “Ecriture, Narration and the Grammar of French”, in: Narrative: From Malory to Motion Pictures, Stratford-Upon-Avon Studies, Second Series, edited by Jeremy Hawthorn, London: Edward Arnold, 1-24. 1991 “L’écriture et le non-dit”, Diacritics 21.4, 21-31. 2003 [1992] “Literary Pragmatics, in: The International Encyclopedia of Linguistics, vol. 2, edited by William Bright, New York and Oxford: Oxford UP, 475-480. Barthes, Roland 1973 [1966] “Introduction to the Structural Analysis of Narratives”, in: Text: Essays Selected and Translated by Stephen Heath, New York: Hill & Wang, 79-124; also at http://phoenixandturtle.net/excerptmill/barthes3.htm. Benveniste, Émile 1971 [1959, 1966] “The Correlations of Tense in the French Verb”, in: Problems in General Linguistics, translated by Mary E. Meek, Coral Gables: Miami University Press, 205215. Blanchard, Marc 1992 “His Master’s Voice”, in After Genette: Current Directions in Narrative Analysis, edited by Carl R. Kropf and R. Barton Palmer, Studies in the Literary Imagination 25.1, 61-78. Brenkman, John 2000 “On Voice”, Novel: A Forum on Fiction 33.3, 281-306. Chatman, Seymour 1980 [1978] Story and Discourse: Narrative Structure in Fiction and Film, Ithaca and London: Cornell University Press; reprint Cornell Paperbacks. Fludernik, Monika 1993 The Fictions of Language and the Languages of Fiction: The Linguistic Representation of Speech and Consciousness, London and New York: Routledge. 2001 “New Wine in Old Bottles? Voice, Focalization, and New Writing”, Voice and Human Experience, New Literary History 32.3, 619-638.

Homonymy, Polysemy, Synonymy

35

Genette, Gérard 1982 [1966] “Frontiers of Narrative”, in Figures of Literary Discourse, translated by Alan Sheridan, New York: Columbia University Press, 127-143. 1983 [1972, 1980] Narrative Discourse: An Essay in Method, translated by Jane E. Lewin, reprint Ithaca: Cornell UP, Cornell Paperbacks. 2007 [1983] Nouveau discours du récit, Paris: Le Seuil, reprint “Points”. 1988 [1983] Narrative Discourse Revisited, translated by Jane E. Lewin, Ithaca: Cornell UP, Cornell Paperbacks. Gibson, Andrew 1996 Towards a Postmodern Theory of Narrative, Edinburgh: Edinburgh University Press; “Postmodern Theory”. Hamburger, Käte 1993 [1957/1968] The Logic of Literature, translated by Marilynn J. Rose, Bloomington and Indianapolis: Indiana University Press. Korthals Altes, Liesbeth 2006 “Voice, Irony and Ethos: the Paradoxical Elusiveness of Michel Houellebecq’s Polemic Writing in Les particules élémentaires”, in: Stimme(n) im Text: Narratologische Positionsbestimmungen, edited by Andreas Blödorn, Daniela Langer and Michael Scheffel, Berlin and New York: Walter de Gruyter, “Narratologia”, 165-193. Kuroda, S.-Y. 1979a [1973] “Where Epistemology, Style and Grammar Meet: A Case Study from Japanese”, in: The (W)hole of the Doughnut: Syntax and its Boundaries, Ghent: E. Story-Scientia, “Studies in Generative Linguistic Analysis”, 187-203. 1979b [1975, 1976] “Reflections on the Foundations of Narrative Theory—from a Linguistic Point of View”, in: The (W)hole of the Doughnut: Syntax and its Boundaries, Ghent: E. Story-Scientia, “Studies in Generative Linguistic Analysis”, 205-231. 1979c “Some Thoughts on the Foundations of the Theory of Language Use”, Linguistics and Philosophy 3.1, 1-17. Milner, Jean-Claude 1995 [1989] Introduction à une science du langage, Paris: Le Seuil, “Points”. Nielsen, Henrik Skov 2004 “The Impersonal Voice in First-Person Narrative Fiction”, Narrative 12.2, 133-150. Patron, Sylvie 2009 Le Narrateur: Introduction à la théorie narrative, Paris: Armand Colin, “U”. 2011 “Homonymie chez Genette: l’opposition entre histoire et discours dans les théories communicationnelles et non communicationnelles du récit”, in : Relire Benveniste: Actualité des recherches sur l’énonciation, edited by Émilie Brunet and Rudolf Mahrer, Louvain-la-Neuve: Academia-Bruylant (forthcoming). Philippe, Gilles 2005 “Le style est-il une catégorie énonciative ?”, in: De la langue au style, edited by JeanMichel Gouvard, Lyon: Presses universitaires de Lyon, 145-156. Prince, Gerald 1982 Narratology: The Form and Functioning of Narrative, Berlin, New York, Amsterdam: Mouton Publishers, “Janua Linguarum. Series Maior”. Rimmon-Kenan, Shlomith 2002 [1983] Narrative Fiction: Contemporay Poetics, London and New York: Routledge. 1989 “How the Model Neglects the Medium: Linguistics, Language, and the Crisis of Narratology”, The Journal of Narrative Technique 19.1, 157-166. Ryan, Marie-Laure 1981 “When ‘je’ is ‘un autre’: Fiction, Quotation, and the Performative Analysis”, Poetics Today 2.2, 127-155. Schaeffer, Jean-Marie

36 1999

Sylvie Patron

[1995] “Temps, mode et voix dans le récit”, in: Nouveau dictionnaire encyclopédique des sciences du langage, edited by Oswald Ducrot and Jean-Marie Schaeffer, Paris: Le Seuil, “Points”, 710-727. Scott Fitzgerald, Francis 1991 [1925] The Great Gatsby, Cambridge: Cambridge UP. Shen, Dan 2001 “Narrative, Reality, and Narrator as Construct: Reflections on Genette’s Narrating”, Narrative 9.2, 123-129. Soelberg, Nils 1984 “La narratologie: pour quoi faire?”, Revue romane 19.1, 117-129. Stanzel, Franz K. 1984 [1979, 1982] A Theory of Narrative, translated by Charlotte Goedsche, Cambridge: Cambridge UP. Woolf, Virginia 1996 [1925] Mrs Dalloway, London: Penguin Popular Classics.



RIKKE ANDERSEN KRAGLUND (Aarhus)

‘Alternate Strains are to the Muses Dear’: The Oddness of Genette’s Voice in Narrative Discourse In the reception of Gérard Genette’s Narrative Discourse there has been a special interest in Genette’s “typology of narrative forms” and its “binary oppositions.”1 This emphasis on the “technology of narrative” has contributed to the portrayal of Genette’s work as a deductive method more interested in firm distinctions and fixed hierarchies than in the literary texts themselves. Shlomith Rimmon-Kenan describes some of the perceived problems in classical narratology in the second edition of Narrative Fiction (2001): Under the influence of structuralism, narratology excluded theme, interpretation, the reader […] the referent(s), ideology, space, and even […] language itself.” (Rimmon-Kenan 2001: 138)

Actually, we find most of these exclusions not in Narrative Discourse but rather in the reception of the book. As Monika Fludernik notes in her article “New Wine in Old Bottles? Voice, Focalization, and New Writing”: “[…] the concept voice covers a number of phenomena relating to the narrating instance (when, where, and who). The issue of voice in Genette, therefore, cannot be simply reduced to the question “who speaks,” or to the subcategory person.” Yet, Fludernik finds Genette’s placing of the categories “time of narrating” and “narrative level” under voice “particularly odd” because “the categories order, frequency, and duration were treated in the temporal realm (under temps in Chapters 1-3 [33-160])” (2001: 620). Here it sounds as if the categories in Narrative Discourse are expected to be impermeable, an assumption Fludernik later implies in relation to differences between Genette and F. K. Stanzel: “Suffice it to say 1

Cf. the section “Structuralist Narratology: The Rage for Binary Oppositions, Categorization, and Typology” in Fludernik (2008): “From Structuralism to the Present” in A Companion to Narrative Theory: “Even more prominently Gérard Genette’s typology of narrative forms (1980) sports a panoply of binary oppositions: homodiegetic vs. heterodiegetic (the character is or is not a character in the tale); extradiegetic vs. intradiegetic (the narrator’s act of narration is situated inside or outside the story); focalization interne vs. externe, and so forth.” (38)

38

Rikke Andersen Kraglund

that Genette privileges clarity of distinction […], whereas Stanzel has been more interested in historical and prototypical combinations of voice and mood in actual texts.” (620-621) This supposition is also known from Dorrit Cohn’s criticism of Genette’s categorical separations—two decades before the quoted Fludernik passages: Genette’s categories (Mode, Voice, and their Subdivisions), by contrast [to Stanzel], tend to stand apart, cogent in themselves, but without systematic integration—with the grammatical trope providing, if anything, additional inducement for treating separately (as grammar books are wont to) rather than for correlating them. (Cohn 1981: 159-160)

Obviously, the adverse topic of structuralism has been one main tenet in the German and Anglo-American reception of Narrative Discourse. To criticize Genette equals lamenting his Structuralist heritage, his alleged preferences for dichotomies, synchronic descriptions of systems, and his disregard for diachronic dimensions. A pattern of more or less disguised stock responses has dominated the field, making it difficult to even imagine a less prejudiced reading of Genette. One unavoidable, it would seem, result of this has been an unbelievable tradition of shallow readings and seminal misunderstandings of Narrative Discourse.2 In what follows, I want to illustrate that Genette’s method is not as 3 inflexible as the reception might suggest. I want to show that Genette’s readings in Narrative Discourse emerge from a complex interplay of different categories. I shall combine theoretical and interpretive inquiries in an attempt to comprehend Genette’s inclusion of time and level in his understanding of voice. Thus, I want to re-assess Genette’s method by using his visions of interpretive practice with a close reading of The Seducer (1993), a novel by the Norwegian author Jan Kjærstad. Other novels could have been chosen, but this novel has a special interest with regard to the difference between understanding Genette’s concept of voice as cogent in itself or perhaps as a bricolage and even a merging of categories.

2 3

For a detailed analysis of some crucial misreadings of Narrative Discourse, see Reitan (2008). Cf. Phelan’s and Rabinowitz’s description in the introduction to A Companion to Narrative Theory: “On the one hand, we have the search for a stable landing, a theoretical bedrock of the fundamental and unchanging principles on which narratives are build. This approach is often associated with what is called structuralist (or classical) narratology, and especially after the rise of post-structuralism, it is often viewed as old-fashioned, even quaint—and it is often believed to yank the life out of the works it considers.” (Phelan and Rabinowitz 2008: 1)

‘Alternate Strains are to the Muses Dear’

39

Borderline Situations Narrative Discourse is famous for its complex technical vocabulary, which has played an important role for students of literature over the past forty years. In addition to the treatment of narrative theory, Narrative Discourse also contains a close reading of Proust’s A la recherché du temps perdu.4 The systematic approach and high level of generalisation is often dismissed in order to show the uniqueness of the novel and “the strangeness of the text.”5 Genette often points to the oddness and lawlessness of Proust’s novel. He shows how the work playfully supersedes conventions. Genette is more text-oriented and more empirically oriented than he is thought to be. He often allows A la recherché du temps perdu determining his method. In this sense, his method seems “odd” from the perspective of the conventional critical reception of narratology as a system that “would not seek to explain what individual works mean but would attempt to make explicit the system of figures and conventions that enables works to have the form and meaning they do”, which has a “main focus on closed systems and static products”, “preference for (reductive) binarisms and graded scales” and “focus on universalist features of all narratives.”6 Genette investigates the techniques in Recherche and shows their relation to classic conventions, and he does not seem much interested in techniques that do not exist in Recherche. In this way, Genette’s readings are not an attempt to show all the possibilities of order, duration, frequency, mood and voice. This restriction and lack of universalistic approach has been criticised in the reception—also by the later Genette. In Narrative Discourse Revisited he describes his focus on one single text in Narrative Discourse as problematic: The systematic recourse to Proustian examples was obviously responsible for certain distortions: an excessive insistence, for example, on matters of time (order, duration, frequency), which take up considerably more than half the book, or too scant a notice of phenomena of mood whose role in the Recherche is obviously minor […]. (1988: 12)

In my opinion, some of the highest merits of Narrative Discourse are due to its readings of a single novel. Genette chooses to focus mainly on a single text and is thus able to reveal the dynamics and changes of voice throughout the text. Trying to make his typology open for the outcome of the interpretation of Proust’s novel, Genette stresses that his method is not 4

In Narrative Discourse Revisited Genette mentions the lack of interest in his reading of Proust “the strictly Proustological aspect of that earlier work has hardly been challenged.” (Genette 1988: 12) 5 Genette (1980: 8) (foreword by Jonathan Culler). 6 Ansgar Nünning situates structuralist (’classical’) narratology as a binary opposition to (’postclassical’) narratologies (Nünning 2000, 358). Cf. Rimmon-Kenan (2001: 142).

40

Rikke Andersen Kraglund

intended to capture all narratives at all times. All the way through Narrative Discourse he attempts to show the “monstrous” in Proust’s work. The subject of his book is often highlighted as something unique: […] the Recherche, more than all other works, must not be considered closed; and therefore it is always legitimate and sometimes necessary to appeal to one or another of its variants. (21) […] the Recherche du temps perdu uses prolepsis to an extent unequalled in the whole history of narrative” (68) […] he thus made clear, more than anyone had done before him and better than they had, narrative’s capacity for temporal autonomy. (85)

Genette’s valorization of Proust’s transgression of classical norms is often misunderstood.7 Deviations are valued as something positive. Genette’s interest in the strangeness of texts is even more obvious in Narrative Discourse Revisited: “Every day Nature and Culture breed thousands of “monsters” that are as fit as fiddles.“ (129) Narrative Discourse Revisited closes with an opening for narrative possibilities: What is certain is that poetics in general and narratology in particular, must not limit itself to accounting for existing forms or themes. It must also explore the field of what is possible or even impossible without pausing too long at that frontier, the mapping out of which is not its job. Until now, critics have done no more than interpret literature. Transforming it is now the task at hand. That is certainly not the business of theoreticians alone; their role is no doubt negligible. Still, what would theory be worth if it were not also good for inventing practice?” (Genette (1983) 1988, 157)

However, Genette’s method has since been challenged, extended and broadened through the study of new literary works. Brian Richardson’s Unnatural Voices has shown new possibilities through its considerations of experimental voices such as second-person narration and multi-person narration. Richardson draws attention to non-mimetic acts of narration and is placing the unnatural in the centre of his investigation, while Genette all the time shows Proust’s transgressions as deviating from the classic and as estrangement. They place non-mimetic narrative forms in different ways. In the “Foreword” to Narrative Discourse, Jonathan Culler underestimates Genette’s interest in the strangeness of fiction: “[…] the categories for the description of narrative discourse are in fact based on what we may for convenience call a model of the real world.” (Genette (1983: 12) But Genette points in Narrative Discourse several times to the strangeness of narrative discourse and voice. He points for instance to the “inconceiv7

Genette also describes this misunderstanding in Narrative Discourse Revisited: “But to my mind, everything nudging Proustian narrative in the direction of interference and repetition was an element of transgression of classical norms and therefore a factor of valorization.” (1988: 2728)

‘Alternate Strains are to the Muses Dear’

41

able” (150), the “irreducible unrealism” (150) the “floating” (151), the “indeterminate” (151), “The Game with Time” (155), the “unlikelihood” (205), the “physical impossibility” (208), “polymodality” (210), “the antithesis of realism”, “the concurrence of theoretically incompatible focalizations” (211) “metalepses” (234) and voices that “blend and merge” (253). Often he even tries to find a new terminology that is not restricted “by the real world”: This term [personnages] is used here for lack of a more neutral or more extensive term which would not unduly connote, as this one does, the “humanness” of the narrative agent, even though in fiction nothing prevents us from entrusting that role to an animal […] or indeed to an “inanimate” object. (244)

Narrative Discourse extends beyond its well-known conceptions of narrative forms. It gives a powerful reading of an individual work that shows Genette’s rich responsiveness as a reader. In Narrative Discourse we find Genette’s ‘odd’ attraction to details that disconfirm and test his concepts in order to show the uniqueness of a single work. In the chapter “Distance?” in Narrative Discourse Revisited, he shows his points with typology and numerous parentheses as “(I know)”, “(since, in one way or another, that is always what is at issue)”, “(for it is none, or I am no judge of these matters)” and “This objection is not stupid (and for good reason)” (1988: 49). In the chapter “Level” Genette mentions that all that is needed to convert an extradiegetic narration into an embedded narration is a sentence of presentation. Afterwards we get the following quotation: “In a Parisian drawing room, three men were chatting in front of the fireplace. All of a sudden one of them said, “My dear Marcel, you must have led a fascinating life. Would you tell us about it?” “With pleasure,” answered Marcel, “but I advise you to sit down, for it may well take some Time.” While his listeners were making themselves comfortable in easy chairs, Marcel cleared his throat and began: “For a long time I used to go to bed early,” etc.

This quotation is accompanied with a note: This unpublished incipit, a mediocre pastiche of Maupaussant, does not appear in the notebooks preserved at the Bibliothèque Nationale, N.A.FR. 16, 640-16,702 [the Proust manuscripts]. I owe my knowledge of it solely to a collector in Olivet in the department of Loiret, whose wish to remain anonymous I respect. (1988: 95)

There is another story in Narrative Discourse which doesn’t fit the focus on categorical separations that has been the main attraction in the reception,

42

Rikke Andersen Kraglund

but offers a much more open-minded approach that explicate the interrelationships among its key components.8 Interplay The preface to Narrative Discourse reflects its author’s attempt at developing a method that refuses to choose between two “incompatible systems of defence”: It seems to me impossible to treat the Recherche du temps perdu as a mere example of what is supposedly narrative in general, or novelistic narrative, or narrative in autobiographical form, or narrative of God knows what other class, species, or variety. The specificity of Proustian narrative taken as a whole is irreducible, and any extrapolation would be a mistake in method; the Recherche illustrates only itself. But, on the other hand, that specificity is not undecomposable, and each of its analyzable features lends itself to some connection, comparison, or putting into perspective. (22-23)

Of course, there is often a contrast between an author’s expressed intentions and their realization in the text. But Genette’s claim that “The specific subject of this book is the narrative in A la recherché du temps perdu” (21) is not incorrect. In many ways, the method of Narrative Discourse is inspired by close readings of the Recherche and reflects its composition. In the chapter “Order” in Narrative Discourse Genette sketches Proust’s narrative composition in this way: […] the Recherche du temps perdu is launched with a vast movement of coming-andgoing from one key, strategically dominant position, obviously position 5 (insomnias) and its variant 5’ […]—positions of the “intermediary subject,” who is insomniac or beneficiary of the miracle of involuntary memory. His recollections control the whole of the narrative, giving point 5-5’ the sort of indispensable transfer point or—if one may say so—of a dispatching narrative: In order to pass from Combray I to Combray II, from Combray II to Un amour de Swann, from Un amour de Swann to Balbec, it is always necessary to come back to that position, which is central even though excentric […]. (45)

Genette’s comment on the method in Narrative Discourse resonates with this characterisation: […] perhaps the real relationship between “theoretical” dryness and critical meticulousness is one of refreshing rotation and mutual entertainment. May the

8

Genette’s humorous use of notes, parentheses and digressions are often overlooked. See for instance the play between the text and the note in Narrative Discourse Revisited: “My purpose is not so much to associate myself with this countervaluation (I appreciate Flaubert, James, and Hemingway as much as Fielding, Sterne, and Thomas Mann)”, and then, in a note: “To be entirely honest, that isn’t true.” (1988: 45)

‘Alternate Strains are to the Muses Dear’

43

reader also find in that relationship a sort of periodic diversion, like the insomniac turning over and over in search of a better position: amant alterna Camenae. (23)9

The composition of Narrative Discourse has the same accumulative structure, whereby previously mentioned categories are taken into accounts when new ones are introduced. Order, Duration, Frequency, Mood, and Voice are not held apart by a Chinese wall. For instance on page 70: “To the extent that they bring the narrating instance itself directly into play, these anticipations in the present constitute not only data of narrative temporality but also data of voice: we will meet them later under that heading” (70). Genette returns several times to the same episodes in Recherche in a continued search for a better understanding of the novel, emphasising the importance of understanding narratological categories as relationships: […] analysis of narrative discourse as I understand it constantly implies the study of relationships: On the one hand the relationship between a discourse and the events that it recounts (narrative in its second meaning), on the other hand the relationship between the same discourse and the act that produces it [...]. I propose, without insisting on the obvious reasons for my choice of terms, to use the word story for the signified or narrative content (even if this content turns out, in a given case, to be low in dramatic intensity or fullness of incident, to use the word narrative for the signifier, statement, discourse or narrative text itself, and to use the word narrating for producing narrative action and, by extension, the whole of the real or fictional situation in which the action takes place. (1983: 26-27) We will be careful, however, not to hypostatize these terms, not to convert into substance what is each time merely a matter of relationships. (32)

The focus on the typology and categorization of Narrative Discourse doesn’t give a satisfactory and sufficient picture of Genette’s way of thinking. In the following pages I will analyze the novel The Seducer [Forføreren] by Jan Kjærstad in order to show the importance of using Genette’s inclusion of time and level in the understanding of voice and to demonstrate how Ge10 nette’s focus on the movements in a text work in practice. My analysis will be a study of his conception of narrative voice, which is more open than that of the reception: “the way in which the narrating itself is implicated in the narrative.” (1983: 31)

9

“Alternate strains are to the Muses dear.” (Virgil (1952: Eclogue III.59. James Rhoades’ translation.) 10 Forføreren [The Seducer] was published in Norway in 1993 and translated in 2003 by Barbara J. Haveland. In this article, I use the edition from The Overlook Press from 2006. Forføreren was published in Germany in 1999 as Der Verführer by Kiepenheuer & Witsch Verlag.

44

Rikke Andersen Kraglund

The Seducer The Seducer opens with the “Publisher’s Foreword” (a we-narration) which we recognize as a fictional preface when the author of the novel is described as “anonymous.” As Genette explains in Paratexts: thresholds of interpretation (originally published in French as Seuils), the essential function of fictional prefaces is “to effect a fictional attribution.” (1997: 279) In The Seducer this fictional attribution is supplemented by a “simulation of the serious preface” (279), which in this case is a comment from the fictitious publisher explaining that the novel “the reader now holds in his or her hand” has won their competition for the best biographical novel. The “Publisher’s Foreword” is established as a kind of frame narrative in which we are given the first hints about the heterocosm or storyworld. Events are implied that the actual readers have no knowledge of but which become a thrilling mystery when we realize that the foreword is addressed to fictional readers who would recognize that “a number of names which appear in this novel correspond to those of real people” and who know “the extraordinary and much publicized events which form the basis for the novel—and, even more so, the grim sequel to said events of which, by the way, no mention is made in the novel.” At once, the reader’s attention is drawn to the narrating and the time of the narrating by the enigmatic character. “The way in which the narrating itself is implicated in the narrative” plays a significant role from the first pages. We can identify the narrating as subsequent (also by the use of past tense), but we don’t know how much time has passed since the events of the fictive novel. The temporal position is indefinite. The first chapter, “The big bang”, begins with the sentences: Let me tell you another story. Although I do not know whether that is possible, not after all that has been written and said, but at any rate let me try. I have balked at it for long enough, I admit. I have put it off and put it off. But I have to do it. (Kjærstad 2006: 1)

Once more, the “you” that is addressed knows details about the story that the actual readers don’t know. The narration is in present tense and there is obviously a narrator or a narrating I who has a different story to tell about the protagonist Jonas Wergeland: “Something about those sides of his character which have never come to public attention and which should serve to shed considerable new light on the man: Jonas as the Norwegian Tuareg, Jonas Wergeland as a disciple of the Kama Sutra, as champion of the Comoro Islands and, not least, as lifesaver.” (1) These fragments points forward to stories that have not yet been told. In the first chapter of Narrative Discourse, “Order”, Genette develops a terminology for studying the temporal order of a narrative by comparing “the order in which

‘Alternate Strains are to the Muses Dear’

45

events or temporal sections are arranged in the narrative discourse with the order of succession these temporal segments have in the story, to the extent that story order is explicitly indicated by the narrative itself.” (1980: 35) “Anachrony” is the umbrella term for “the various types of discordance between the two orderings of story and narrative.” (36) One of the most common anachronies, according to Genette, is “beginning in medias res, followed by an expository return to an earlier period of time”, which is one of the “formal topoi of epic.” (1980, 36) The Seducer also marks this classical beginning: But to begin in medias res, as they say, or in what I prefer to call ‘the big white path’, representing as it does a stretch of terrain of which Jonas Wergeland—all of his fantastic journeys notwithstanding—was totally ignorant, and which he would spend the remainder of his life endeavouring to chart. (Kjærstad (2006: 1)

This is retrospection, or analepsis, as Genette calls it. The narrator is looking back on Jonas Wergeland’s life and trying to reawaken the stories which led to the protagonist’s destiny. Within this overall analepsis there are several prolepses which draw attention to stories that have not yet been told. According to Genette the use of prolepsis conflicts with the idea of suspense: Anticipation, or temporal prolepsis, is clearly much less frequent than the inverse figure, at least in the Western narrative tradition—although each of the three great early epics, the Illiad, the Odyssey, and the Aeneid, begins with a sort of anticipatory summary that to a certain extent justifies the formula Todorov applied to Homeric narrative: “plot of destination.” The concern with narrative suspense that is characteristic of the “classical” conception of the novel (“classical” in the broad sense, and whose center of gravity is, rather, in the nineteenth century) does not easily come to terms with such a practice. (1983: 67)

Nonetheless, the use of prolepsis in The Seducer does create suspense. Genette neglects an important difference between suspense and surprise—in A Dictionary of Literary Terms described thus: Most great art relies more heavily on suspense than on surprise. One can rarely reread works depending on surprise; the surprise gone, the interest is gone. Suspense is usually achieved in part by foreshadowing—hints of what is to come […]. Suspense is […] related to tragic irony. The tragic character moves closer and closer to his doom, and though he may be surprised by it, we are not; we are held by suspense. (Barnet et al. 1960: 83-84)

The pronounced use of prolepsis accentuates the narrator’s role as the composer of the story. There is an obvious nonlinearity in the sequence of chapters in The Seducer; the setting alters abruptly.11 Although episodes 11

Jonas Wergeland is a TV producer and the composition of “The Seducer” is in many ways reminiscent of the television viewer’s “flipping through the channels” that David Harvey has describes as a condition of postmodernity: “The contemporary television viewer’s routine

46

Rikke Andersen Kraglund

from Jonas’ life are not connected chronologically, the causal connections between the episodes are strong, a causality whereby future episodes in Jonas Wergeland’s life can affect the past. Laws of nature or dogmas of realism do not restrict The Seducer. For instance, Jonas Wergeland is at the age of fifteen warned about a catastrophe by a melody he played in the chapter before—the catastrophe would take place many years later: It may seem hard to believe, in the normal way of looking at things, I mean, but I beg you please to believe me when I say that Jonas Wergeland received a warning in the form of a stanza of organ music, a little phrase from ’Leid milde ljos’ which seemed to Jonas to be carried across the water and resonate off the hull of the boat. (2006: 120)

The narrator seems to grant the protagonist knowledge of the future (a future that has been presented in earlier chapters). When Jonas is six years old his cousin pushes him in the water and Jonas almost gets cut by the propeller of a nearby boat: The first face he sees is Veronika’s, a sight which prompts him to blurt out a weak and incredulous ‘Jesus Christ, Veronika’. And only those who have read this far will guess that what Jonas is actually saying, even though he does not have the words with which to verbalize this perception, is: ‘So this is how you thank me for saving your life on the Zambezi?” (2006: 185)

This description points forward to an event which has not yet happened in the story and backward to an event that has already been described in an earlier chapter in the discourse: “When later is earlier, and earlier later, defining the direction of movement becomes a delicate task.” (Genette 12 1983: 83). Here it seems appropriate to recall Henri Bergson’s description of perception: Your perception, however instantaneous, consists then in an incalculable multitude of remembered elements; and in truth, every perception is already memory. Practically, we perceive only in the past, the pure present being the invisible progress of the past gnawing into the future. (Bergson 1919: 194)

In The Seducer, not only does the past gnaw into the future; the future gnaws into the past. When chronological sequencing does not order the narration, it seems “intentionally removed from the domain of automatized perception.” (Cf. Shklovsky 1990: 12) The narrator’s intention in the

procedure of “flipping through the channels” juxtaposes radically disjunct times and spaces in visual simultaneity, creating a “perception of history as an endless reserve of equal events.” (Harvey 1990: 61). 12 These passages also points to The Seducer as fiction where “the narrative act initiating (inventing) both the story and its narrative, which are then completely indissociable.” (Genette 1988: 15)

‘Alternate Strains are to the Muses Dear’

47

sequencing of stories becomes essential and the narrator often comments 13 on the ideas behind the structure. The study of narrative temporality in The Seducer is also a study of voice. In The Seducer, events are often connected both by their thematic resemblance and by the use of a frame. The first chapter describes how Jonas Wergeland returns one evening to find his wife Margrete dead on the living room floor. This experience is outlined in a past tense thirdperson narrative by an omniscient narrator who knows the thoughts of the protagonist: Jonas Wergeland stood with his hand on the door-handle and was filled all at once with a desire to stay just there for a long, long while, had no wish to enter the room, stood there knowing, like someone who has stepped on a mine, that he would be blown sky-high the moment he raised he foot. But he had to. […]. And I might as well reveal right here and now that here lies the heart of my story: Jonas Wergeland, standing in a room with a dead woman, caught in the colossal psychological big bang that gave birth to the universe which, in the following account, I intend to explore. For those who do not know, I ought perhaps to add that the woman on the floor was none other than his wife. (2006: 4-5)

This room becomes the essential frame of the narrative and a transfer point where stories from different times and places in the protagonist’s life are hinted at or evaluated. One of the best-known examples of a frame narrative is Arabian Nights, in which Scheherazade every night tells a story to King Sharyar in order to delay her death sentence. In Theory of Prose, Victor Shklovsky calls this technique “deceleration”: “[...] a whole series of fables is told with the purpose of dragging out time and forestalling a hasty decision.” (Shklovsky 1990: 42) After one thousand and one nights the King is changed by the stories and Scheherazade is reprieved. The Seducer has similarities with this tale; like Scheherazade, its narrator tells stories in order to help and change Jonas Wergeland. Just as the frame gives the embedded chapters motivation and coherence, the embedded narratives colour and alter our understanding of the frame. The narrator returns to the scene with the dead wife eight times. Each return brings new angles to the story and an increasingly intimate tone, as if the narration itself also works to transform the narrator’s own project. To begin with, there is a marked difference between the tone of the narrator in the frame and in 13

There is of course a delicate relationship between the intentions of the novel “The Seducer” by the anonymous author’s and the intentions of The Seducer by Jan Kjærstad. Some elements of the composition are obvious only “known” to Jan Kjærstad. The Seducer is followed by two other novels about Jonas Wergeland, which are told by different narrators. Together the three books are connected through the use of a serial form. Seen in context Jonas Wergeland for instance meets 23 women, travel to 23 countries and makes 23 TV programmes, which are connected in different ways.

48

Rikke Andersen Kraglund

the embedded narratives.14 Where the scene in the first chapter is a thirdperson narrative in past tense with a distanced and often ironic voice, the return to the scene is foregrounded as a second-person narrative in present tense with an identifying tone:15 and you look at those circles, circles within circles, and you try to see the connection [...] and you look at the circles and see that they are spinning, like a wheel, you think, and you are growing dizzy, everything is moving too fast, round and round, and you realize that you are shaking; if only you knew how I wish I could be there, how I wish I could comfort you, hold your hand, help you to pick up the thread of the story, the thread you have lost, show you that all things are spokes in the same wheel [...] (2006: 465)

Brian Richardson’s compelling work Unnatural Voices emphasises how the choice of the “you” form can radically alter the tone of a work. The “you” form immediately creates intimacy and defamiliarises. According to Richardson, the form is irreducible to the traditional distinction between first and third person and Genette’s dichotomy of homodiegetic and heterodiegetic narrative.16 The “you” form is characterized by “ambiguity and fluctuation” (22) which means that the reader has to sharpen his awareness of how voice shifts throughout the work. Awareness of the dynamic of the “you” form is essential to understanding The Seducer. In the Seducer the “you” is not only the protagonist, he is also the narratee in the frame story. The book “The Seducer” in The Seducer has been written for Jonas Wergeland in order to help him recall the trauma of the death of his wife. The narrator is thus telling the protagonist what he already knows. Even though Jonas is the only witness, the narrator retells his story so that it becomes a sort of joint experience shared by the narrator and Jonas. Jonas has lost his memory and has to experience the event once more in order to be able to tell about it. Thus, the second-person narration functions as 14

There is also a variation of tempo between the frame and the embedded narratives. There is a deceleration with a repetition of the same movement patterns and a focus on details “you flick through the row of CD’s, more or less at random and pull out a CD, and you lift it out of its holder, you study the disc, seeing how it shines, like a miniature sun you think, or no, it strikes you that it looks like a wheel shot with rainbows, you think and you lay it in the tray in the CD player and pick up the remote control, select the track you want and fall into a more abstract reflection on the feel of the tiny rubber bottom on the remote control” (Kjærstad 2006: 70). “[…] needless to say, a detailed narrative, in “scene” tempo, gives the reader a greater impression of presence than does a quick and distant summary.” (Genette 1988: 46) 15 These returns differ from the embedded chapters by not having a title. Instead the embedding scenes are marked by paragraphs breaks. In Norwegian the frame is also marked by an uncial with a big O. O being the first letter in the word “og” which means “and” as to mark that the scenes are not moving forward but have a repeating structure like a circles “O”. 16 Genette distinction between hetero- and homodiegetic narration is not as rigid as it might seem in the typology of Narrative Discourse. In Narrative Discourse Revisited Genette wants to “concede the borderline” to gradualism: “[…] we must indeed admit the possibility (and must note the existence) of mixed or ambiguous borderline situations […]” (1988: 104).

‘Alternate Strains are to the Muses Dear’

49

an imperative and an instructional mode to help Jonas understand and recall the incident in the future. The voice is very intimate and empathetic and it has an affective appeal. In addressing this narrative to Jonas, the narrator is saying: “I believe this to be your story, your version. I sympathize with you.” The embedded narratives are filled with stories about losses Jonas has learned to live with through the help of narratives from someone close to him. The narrator seems to have the same intention with the book “The Seducer”: helping Jonas by composing Jonas’ life into a meaningful story. The second-person narration does not only give information about the “you”. Implicit in this form of address is an “aura of the speaking subject”, as H. Porter Abbott explains in the Routledge Encyclopedia of Narrative Theory. (Abbott 2007: 342) We get a feeling of the person behind the voice. In the Seducer, we haven’t much information about the identity of the narrator and how the narrator knows all the stories about Jonas. At first, the narrator creates the impression of being an omniscient narrator who has access to the protagonist’s thoughts and feelings, and who looks down on the characters from above: “I see things differently. I am capable of taking the broader view, of seeing Norway from above, with all the necessary detachment.” (Kjærstad 2006: 63) The narrator has a godlike access, which the novel emphasises with humoristic references to stories about the Christian God: I ought to have introduced myself, I know, but I am very much afraid that this would only lead to misunderstanding. For some, this tale would thus be lent too much; it would lose all credibility in the eyes of others. My own popularity is, after all, plummeting, and—this much I can say—I am now so much persona non grata that a lot of people have declared me to be dead. I must, therefore, choose my words with care. I am who I am. More than that I cannot say. […] (2006: 62)17

Traditionally, an omniscient narrator would appear as an extra- and heterodiegetic narrator, a kind of disembodied and depersonalised voice with superior knowledge. At first, the narrator of The Seducer seems to function on an ontological level distinct from that inhabited by the protagonist and the other characters. But it soon becomes obvious that this calm, Olympian detachment is an illusion, and that the authoritative voice is fallible: […] there are no words and no metaphors to cover this: every boy’s first fingertip contact with a girl’s vulva. There is a limit to my omniscience and this is it, so I will have to leave them to drift onwards, Jonas and Margrete, hand in hand […] (154)

17

There is for instance a paraphrase of Nietzsche and Exodus 3,14.

50

Rikke Andersen Kraglund

In several cases the narrator reveals knowledge which is limited but at the same time more detailed than a situated narrator realistically would be able to acquire. Paradoxically, the narrator seems to be an omniscient narrator with limited knowledge. This narrator is obviously emotionally involved and it is hard to believe his claim to be objective. “The Seducer” is a persuasion and a seduction not only of Jonas but also of its readers. In The Seducer the relation between the narrator and narratee is in many ways the main story. At the end of the book, however, it is still a somewhat of a mystery. The story is not brought up to the time of the telling. We do not know what has happened between the end of the narrator’s “The Seducer” and the frame with the “Publishers’ foreword”. We do not have access to the whole context. We know from the beginning that the narrator wants to tell another story about Jonas Wergeland. But the actual readers do not know the other stories. The fictitious paratext, “The Publishers foreword,” points as an explicit ellipsis and a narrative void or gap to “the grim sequel to said events of which, by the way, no mention is made in the novel”. The big mystery of The Seducer is why Margrete is lying dead on the floor. A detail, announced by a nonchalant “by the way”, implies another important mystery. What happens after Jonas Wergeland finds Margrete? Why doesn’t the narrator describe “the grim sequel”? The narrator points to the narration so many times that the narrator himself becomes the focus of attention. In The Seducer we are constantly being reminded of “The Seducer” as a story from somebody to someone. However, who the narrator is, what the nature of this apparently authorial figure is, and what the relation between narrator and narratee, remains enigmatic. This uncertainty about the narrating situation is of great importance in the novel. Trilogy When The Seducer was first published in 1993, Jan Kjærstad withheld the information that the novel would be the first narrative in a trilogy about the protagonist Jonas Wergeland.18 It was followed by The Conqueror [Erobreren] in 1996 and The Discoverer [Oppdageren] in 1999,19 which tell some of the same stories about Jonas Wergeland from different angles and with new narrators. When the three novels are read in succession, there are several clues in the frame story which point to The Discoverer as the conclu18

The same thing happened after The Conqueror, where people didn’t knew that it would be followed by The Discoverer. 19 For The Discoverer Jan Kjærstad won the prestigious Nordic Prize for Literature in 2001.

‘Alternate Strains are to the Muses Dear’

51

sion of a traditional story about finding the meaning of life and making the right decisions. The trilogy seems to contain a narrative progression, since the second novel comments on the first, and the third novel comments on the first and second; textual richness increases in “proportion to the number of individual narratives included.” (O’ Neill 2007: 368) However, the narratives of The Seducer and The Conqueror cannot be definitely contained in the retrospectives of The Discoverer; the books are full of contradictory tales that refuse to conform to a single angle. The narrators give partially conflicting accounts of episodes in the protagonist’s life. Some of these inconsistencies can be explained by the choice of different narrators and might be seen to thematize how the protagonist is constructed by the narrators’ discourses and how our identity is defined by our relations with others. The story of a life can always be transformed to a totally different story when told from a different angle. You cannot write a “true” or “final” biography or even autobiography (the protagonist is one of the narrators in The Discoverer). Description implies selection and simplification, and subjectivity is a fundamental premise. The narrators have personal interests and purposes in relation to the narratives. The Conqueror is clearly a retelling of Jonas Wergeland’s life seen from a darker perspective (whereas the narrator in The Seducer has divine traits, the mysterious narrator in The Conqueror has similarities with the devil). We now get an explanation of “the grim sequel” to Margrete’s death—the time span between the end of the narrator’s tale to “The Publishers foreword” in The Seducer. Jonas is in jail arrested for the murder of his wife. From The Conqueror’s perspective, the narrative in The Seducer appears as a defence of Jonas Wergeland delivered with an extreme emotional appeal that borders on kitsch. The Conqueror also reveals the identity of the narrator in The Seducer. The narrator is a visitor to the prison, Kamala Varna, who has heard the stories from Jonas Wergeland. This serves to explain some of her knowledge but not all of it. The narrator still has a touch of omniscience; for example, she can narrate events that Jonas never told anyone about. In many ways, The Conqueror could be read as a denarration of The Seducer, each novel functions as a correction of the other. Some stories are just modified; others are negated. The last novel, The Discoverer, provides an emotionally satisfying ending. The protagonist has been in jail for the murder of his wife. This ‘murder’ is now explained as a suicide. The protagonist has chosen to take the punishment for this suicide because he feels guilty for not helping his wife in time. But although the previous two novels’ description of the protagonist as a murderer is here denarrated, it still continues to influence the reader’s perception. In some ways the three novels follow each other in succession; in other ways they stand in an equal relationship and offer three different versions of the story. The nov-

52

Rikke Andersen Kraglund

els call each other’s description into question; fundamentally, there is a gap between their different versions, which can’t be resolved in the text. The framing of the other narrators cannot reveal the “true” story. None of them has greater authority than the others and we cannot really judge between them. These novels contain several contradictory events; there is no way to resolve these contradictions by imposing some sort of ontological hierarchy on the trilogy. In Brian Richardson’s terms, we are unable to falsify their versions. “Unlike any nonfictional account […] once these words are written in a narrative fiction, they constitute some of the unalterable facts of its world. There is no way of falsifying these descriptions by drawing on material outside the text, […] the narrator’s word is definitive”, but only “as long as the narrator does not contradict them.” (Richardson 2006: 91-92) There are indications within the text that show contradictions in the three novels, but we look in vain for any final solution. In the Wergeland trilogy the three novels are three fictional possibilities. There are three discourses without a story to provide the common denominator of the trilogy. The three narrators give three multiple, contradictory versions. Besides these different narratorial angles, there are also some chronological conjunctions—as if Jonas Wergeland has lived several parallel lives; he has been different places and has known different people at the same time. This multiversional structure points to fictional narratives’ capacity to set their own boundaries. The Wergeland trilogy breaks with the dogma of realism and points to the potential of fiction. As Ursula Heise mentions in her description of time structures in postmodernist novels, Postmodernist novels project into the narrative present and past an experience of time which normally is only available for the future: time dividing and subdividing, bifurcating and branching off continuously into multiple possibilities and alternatives.” (Heise 1997: 55)

As numerous examples above will have shown, the reading of voice in the Wergeland trilogy is related to an understanding of how the three novels play with time and level, and with voices. This play cannot be reduced to a question of “who speaks” or to questions about “person”; narrative voice is nothing if not temporally related to discourse and story. Narratology may still have much to gain by approaching Narrative Discourse with a renewed focus on Genette’s attention to the dynamics and interaction of narrative devices and techniques throughout one narrative. In my analysis of Kjærstad’s trilogy—a narrative consisting of several interrelated narratives with different narrators and only one protagonist—I have consistently used Genettian concepts of time, level, and voice. However, Genette’s cherished “axiom” of one overarching extradiegetic position of enuncia-

‘Alternate Strains are to the Muses Dear’

53

tion has been of no use in the analysis; it might be seen as “deconstructed” by the narrative itself. ‘Alternate voices are to the Muses dear.’ References Abbott, H. Porter 2007 “Narration”, Routledge Encyclopedia of Narrative Theory, eds. David Herman, Manfred Jahn, and Marie-Laure Ryan, London and New York: Routledge. Barnet, Sylvan, Morton Berman, and William Burto 1960 A Dictionary of Literary Terms, Boston, Mass.: Little, Brown and Company. Bergson, Henri 1919 [1908] Matter and Memory [Matière et Mémoire], trans. Nancy Margaret Paul and W. Scott Palmer, London: George Allen & Unwin ltd. Cohn, Dorrit 1981 “The Encirclement of Narrative: On Franz Stanzel’s Theorie des Erzählens”, Poetic Today 2.2, 157-82. Fludernik, Monika 2001 “New Wine in Old Bottles? Voice, Focalization and New Writing”, New Literary History 32.3, 619-638. (2008) [2005] “Histories of Narrative Theory (II): From Structuralism to the Present”, in: A Companion to Narrative Theory, Oxford: Blackwell. Genette, Gérard 1983 Narrative Discourse. An Essay in Method. [Discours du récit, 1972], trans. Jane E. Lewin, Ithaca, New York: Cornell University Press. 1988 Narrative Discourse Revisited [Nouveau discours du récit, 1980], trans. Jane E. Lewin, Ithaca, New York: Cornell University Press. 1997 Paratexts: Tresholds of Interpretation. [Seuils, 1987], trans. Jane E. Lewin. Cambridge: Cambridge University Press. Harvey, David 1990 The Condition of Postmodernity, Oxford: Blackwell. Kjærstad, Jan 2006 The Seducer [Forføreren, 1993], trans. Barbara J. Haveland, Woodstock & New York: The Overlook Press. 2007 The Conqueror [Erobreren, 1996], trans. Barbara J. Haveland, London: Arcadia Books. 2009 The Discoverer [Oppdageren, 1999], trans. Barbara J. Haveland, London: Arcadia Books. Nünning, Ansgar 2000 “Towards a Cultural and Historical Narratology: A Survey of Diachronic Approaches, Concepts, and Research Projects”, in: Bernhard Reitz and Sigrid Rieuwerts (eds.), Anglistentag 1999 in Mainz. O’ Neill, Patrick 2007 “Narrative Situations”, Routledge Encyclopedia of Narrative Theory, eds. David Herman, Manfred Jahn, and Marie-Laure Ryan, London and New York: Routledge. Phelan, James and Peter J. Rabinowitz 2008 [2008] A Companion to Narrative Theory, Oxford: Blackwell Reitan, Rolf 2008 Fortællerfiktionen. Kritik af den rene narratologi. Aarhus: Aarhus Universitetsforlag. Richardson, Brian 2006 Unnatural Voices. Extreme Narration in Mordern and Contemporary Fiction, Ohio: The Ohio State University Press. Rimmon-Kenan, Shlomith 2001 Narrative Fiction, 2nd edition, London and New York: Routledge.

54

Rikke Andersen Kraglund

Shklovsky, Victor 1990 Theory of Prose. [O teorii prozy, 1929], trans. Benjamin Sher, Illinois: Dalkey Archieve Press. Virgil 1952 The Poems of Virgil, trans. James Rhoades, Chicago: Encyclopedia Britannica.



HENRIK SKOV NIELSEN (Aarhus)

Fictional Voices? Strange Voices? Unnatural Voices?1 Introduction This article is a small contribution to the current research concerning unnatural narratives and unnatural narratology2. Members of the unnatural narratology group have described some general ideas behind the unnatural approach in earlier articles.3 It is an ongoing discussion, however, how to determine what we mean when we talk about unnatural narratology and to define exactly what unnatural narratives might be. I wish here to ask some of these questions with a specific view to strange fictional voices, and also to say a few words about the background against which the unnatural is posited, i.e. the natural, and to ask, in a rather down to earth manner, what difference it makes if one treats all narratives, including fictional ones, as if they were real-life instances of narrating. The argument will go from the general in the form of a short presentation of parts of the relevant landscape of narrative theory and some of its urgent questions to some specific challenges to common sense understandings of narratives and fictional voices. Such peculiar narrative forms as first person, present tense narration, you-narration, and paraleptic first person narration posit these challenges. I wish to ask in each case what the relationship might be between the specific strangeness of these cases and 1

I wish to thank Stefan Iversen, Per Krogh Hansen, and Rolf Reitan for their suggestions for improvements to this article. A special thanks goes to David Herman with whom I have been discussing many of the ideas here presented. It is a sign of his unlimited academic generosity that he has been extremely encouraging and welcoming in his responses even though he has very different takes on many of the main ideas. 2 After the ISSN Conference on Narrative in Washington in 2007, a small group incidentally gathered in the lobby. The group included Brian Richardson, Jan Alber, Maria Mäkelä, and me. We realized that the conference in Washington seemed to have yielded a rather strong resistance to otherwise predominant naturalizing paradigms. Consequently, we invited everyone interested to join a group on what we tentatively called “unnatural narratology.” Four members of the group have recently published the joint article “Unnatural Narratives, Unnatural Narratology: Beyond Mimetic Models” in Narrative, May 2010. 3 See www.unnaturalnarratology.com.

56

Henrik Skov Nielsen

the general conditions of fictional narratives. Next I will propose a brief reading of Edgar Allan Poe’s “The Tell-Tale Heart” which seems to me an especially interesting and challenging unnatural narrative that leaves almost all real world parameters behind. It presents the reader to voices that are strange in that they conflate text and paratext, here and there, then and now, I and you. Finally, I will ask what consequences, if any, these peculiar forms and narratives have for our understanding of fictional narratives in general. What’s so Unnatural about Unnatural Narratives Anyway? First of all: Why talk about unnatural narratives or unnatural narratology at all? Is it not the case that there is something unnatural in the sense of artificial about any and all kinds of storytelling? When we fashion stories in real life interactions we do so in incredibly complex ways. We use all kinds of techniques and tricks to suit them for our different purposes including exaggerating, fictionalizing, telling backwards, leaving out & filling in details, telling in another’s voice, using free indirect discourse etc., etc. As soon as new tricks are invented for one genre or purpose, they seem to be stolen and transferred to other genres and situations, frequently including transfers between fictional and non-fictional genres4. Or the other way round—coming from the opposite assumption to a very similar conclusion: Is it not the case that all narratives, including the most experimental novels and anti-mimetic post-modern narratives, are rock bottom natural? In the long history of human beings, the imaginative powers have always resulted in narratives that only depict the existing but also explore the virtual, the strange, the unlikely, and the impossible? Isn’t the anti-mimetic impulse a very natural human impulse? And isn’t it the case that many of the “transgressive” techniques celebrated by highbrow theorists are often employed in even the most low brow every-day communicative situations? Thus, we should not lose sight of the complexity also of non-fictional narratives and we should be careful to not posit firm dichotomies between natural and unnatural narratives when in fact the relation is one of dynamic exchange and continuity5. I therefore willingly admit that I feel uneasy about assuming that certain kinds of narratives are natural, and doubly so by assuming that others are unnatural. It is not per se a distinction I wish to invest in myself. 4 5

Cf. Meir Sternberg’s “Proteus-principle” (Sternberg 1982). See also Phelan 2005: 68. Thanks to David Herman for stressing this.

Fictional Voices? Strange Voices? Unnatural Voices?

57

Even so—and I guess that at this point most readers will have felt a “but” coming—I think there are at least two reasons to talk—for the momentary lack of a better term—about unnatural narratives. The first one is that some theoretical assumptions associated with the (explicit or implicit) view that all narratives are (or are genetically linked to) natural narratives have gained tremendous influence over the last years, and that I believe that these theories tend to miss important aspects about important narratives. The second reason is closely connected and consists in the fact that even if it may be true that all narratives are natural from a certain view point and/or all are unnatural from another point of view, then it does not follow from this that all narratives are alike and operate in accordance with exactly the same rules and principles. By natural narratives, then, I here simply refer to narratives that have been designated as such by influential narrative theorists. For instance Monika Fludernik, who in Towards a ‘Natural’ Narratology most prominently connected the term “natural” with narrative theory. Here, she describes the term as follows: Natural narrative is a term that has come to define “naturally occurring” storytelling […]. What will be called natural narrative in this book includes, mainly, spontaneous conversational storytelling […].6

The point here, then, is not to discuss whether conversational narratives are more natural than, say, literary or visual narratives. Rather I wish to just acknowledge the fact that the opinion and the term exist and have been extremely influential, and to examine some of the consequences (many of which I believe are actually more or less incongruent with Fludernik’s theories) it has had. In two closely related articles (“Against Narrative” and “Against ‘against’ Narrative”) Pekka Tammi delivers a sober and thought-provoking criticism of easy assumptions that narratives are everywhere and a natural part of what we all do all the time. Tammi argues that these assumptions have led to an oversight of the differences between different kinds of narration and to a privileging of the natural. In the latter article, Tammi describes his enterprise in terms of “strangeness”: […] strangeness is indeed what the present paper is all about: strangeness, that is, in opposition to the natural—the strangeness of narrative fiction as well as the strangeness (in some instances) of narrative theory itself. (Tammi 2008: 37)

Tammi goes on to say about the privileging of the natural: Narratives, it is said, are everywhere in our lives. Consequently, adherents of the post-classical approach have often tended to opt for the applicability of natural, 6

Fludernik 1996: 13.

58

Henrik Skov Nielsen

real-life parameters for all varieties of narrative presentation, literary and nonliterary, fictional and non-fictional alike. […] the broad definition may also result in a somewhat diluted or, frankly, wishy-washy view of distinctively literary fiction. (Tammi 2008: 37-38)

Fludernik, on her part, is often very alert to transgressions of and experiments with real-life parameters. She is also a very attentive reader of fiction exploring non-natural occurring storytelling situations: Rather than privileging naturally occurring storytelling situations, Natural Narratology, by contrast, attempts to show how in the historical development of narratorial forms natural base frames are again and again being extended. […] [O]nce an originally non-natural storytelling situation has become widely disseminated in fictional texts, it acquires a second-level “naturalness”7 from habituality, creating a cognitive frame […] which readers subconsciously deploy in their textual processing. Even more paradoxically, fiction as a genre comes to represent precisely those impossible naturalized forms and to create readerly expectations along those lines (Fludernik 2003: 255)

While it is instructive to see explicitly stressed that there exists such a thing as “originally non-natural storytelling situation”, the question is whether the reader will always try to naturalize anything—and if so, if it can always be done successfully. This connects back to the overarching argument in Tammi: That there is a price to pay for the anchoring of narrative understanding in real-life experience: And now we also see—which is evidently what I have been driving at all along – that such a theory comes with a cost. For what happens when this way of narrativizing texts does not work out just like that? (Tammi 2008: 40)

It is from this point of view that I wish to briefly examine some specific narrative forms as they occur in fictional narratives in order to ask in each case how they differ from real-life instances of narrating and what this tells us about fictional narratives in general. The four subjects that I touch upon in the next sections are first person, present tense narration, younarration, paraleptic first person narration, and the distinction between narrated I and narrating I. Each of them could surely deserve an article or 7

In Nielsen, “Unnatural Narratology, Impersonal Voices, Real Authors, and Non-Communicative Narration” (in Alber and Heinze, forthcoming), I argue for the usefulness of distinguishing in this context between naturalization and conventionalization. Without this distinction, any conventionalization of new techniques or forms would amount to naturalization. I would like to stress instead the unnaturalness also of conventional forms, like, say, the use of omniscience in traditional works of realism and the use of present tense first person narration in much recent fiction. Fludernik herself stresses that: “Non-natural textual constellations refer to text-types that are naturally non-occurring. For this reason new generic options (such as reflector-mode narrative) do not in the process of narrativization become natural, although they become naturalized […]”. (1996: 330). I agree with the point but would instead prefer to say that they become conventionalized.

Fictional Voices? Strange Voices? Unnatural Voices?

59

even a book of their own, and fortunately some of them are treated more thoroughly elsewhere in this volume,8 but my specific aim in the context of this article is not to engage in an exhaustive investigation of each. Rather, I want to point out how they all cue the reader to interpret in ways that differs from the interpretation of real world acts of narration, like these of conversational storytelling. First Person, Present Tense Narration In chapter 6 of The Distinction of Fiction Cohn describes a “mounting trend in modernist first-person fiction to cast a distinctively narrative (not monologic) discourse in the present tense from first to last” (1999: 97). Cohn rejects both the historical present and the interior monologue as satisfactory explanations of the phenomenon, and takes as her main example a passage from Coetzee’s Waiting for the Barbarians (1980), containing the words that form the title of her chapter 6, “I doze and wake.” Cohn comments on this as follows: But the introspective instance that most strongly resists the interior monologue reading is no doubt the one that reads: “I doze and wake, drifting from one formless dream to another.” Here semantic incongruence combines with the formal feature that most forcefully counteracts the impression of an unrolling mental quotation in this passage as a whole: the pace of its discourse is not consistently synchronized with the pace of the events it conveys […]. (103)

In short: In real life you can’t truthfully report that you sleep while you sleep. But in fiction examples abound where there is a flagrant incongruity between what is narrated and the act of narration itself. But incongruity between doing and telling needs not be that obvious for it to exist. Take, for example, the following straightforward example from Bret Easton Ellis’ Glamorama: “See you, baby.” I hand her a French tulip I just happen to be holding and start pulling away from the curb. “Oh Victor,” she calls out, handing Scooter the French tulip. “I got the job! I got the contract.” “Great, baby. I gotta run. What job you crazy chick?” “Guess?.” “Matsuda? Gap?” I grin, limousines honking behind me. “Baby, listen, see you tomorrow night.” “No. Guess?.” “Baby, I already did. You're mind-tripping me.”9

8 9

Cf. Rolf Reitan’s article on second person narration in this issue. Ellis 1999: 19.

60

Henrik Skov Nielsen

The example is fairly typical and representative for the thousands of first person narratives resulting from the trend that Cohn identified and which was indeed mounting when Cohn wrote her book. It is not immediately apparent that there is something strange about the voice here in Ellis. But notice how there is in fact a clear difference between two levels of words and the ways in which they can and cannot be ascribed to a character narrator. There is clearly a character who starts out by saying “See you, baby.” These words are situated in a communicative situation and uttered by the character (Victor) to a female acquaintance. But at no point is there a narrator situated anywhere before, during or after the events, who says: “I hand her a French tulip.” No communicative situation seems imaginable in which a narrator will narrate these words to a narratee. And never will Victor say, think, or mumble to himself or anyone else “I hand her a French tulip”. There is no context and no occasion for telling them. The techniques used in the quote dissociate the words from the narrator’s account. The present tense here is clearly not the historical present or simply an interior monologue, but rather corresponds to what Cohn calls “fictional present,” (1999: 106). The words of the narrative in Glamorama are unnatural in the sense that they are not modelled on natural narrative, i.e., everyday conversational storytelling. It may be that no great harm is done in saying that “Victor is the narrator in Glamorama”; it might be seen an economical, albeit imprecise alternative to “Victor is the person referred to by the pronoun “I” in this homodiegetic narrative”. Importantly, however, we are led astray if we take full consequence of imposing real world narrative situations on the narrative and begin to wonder when and to whom and for what reasons Victor would narrate the story—whether or not we imagine him doing this simultaneously with being so busy doing the things described. For several reasons the technique used in Glamorama is not stream of consciousness (among these: it includes dialogue, and it does not appear to represent thoughts going through Victor’s head). Nonetheless, it is possible to extend an argument pertaining to stream of consciousness to the use of first person present tense narration here. Phelan’s famous definition of narrative goes: First, narrative itself can be fruitfully understood as a rhetorical act: somebody telling somebody else on some occasion and for some purpose(s) that something happened. (Phelan 2005: 18)

As Phelan himself and many other acknowledges, this is not the case in stream of consciousness. Stream of consciousness is therefore often considered a non-communicative rupture in the narrative since no one is tell-

Fictional Voices? Strange Voices? Unnatural Voices?

61

ing anything to anyone.10 But then the same type of problem arises in first person, present tense narratives where it seems equally implausible that the protagonist is narrating to anyone as he goes along.11 And the challenge posed by first person present tense narratives like Glamorama is not quite remedied by changing a word or two like in, say, “[…] that something is happening” or “[…] that something happens”, since it is not so much a matter of the time of the telling as a matter of no one telling to no one at all—just as in stream of consciousness. Fludernik writes brilliantly on the form in Towards a ‘Natural’ Narratology: Nor can the present tense here [in simultaneous narration] be equated with the use of the present in conversational narratives, since, as we have seen, it has no truck with the historical present tense. (1996: 252) Although not usually recognized as an experimental technique, present tense narrative therefore significantly departs from the traditional schemata of the realworld temporality of storytelling. (256)

Especially thought provoking is a conclusion that draws together present tense narration, imperative enunciation and second person narration:

10

Cf. Cohn: “In contrast to the lack of attention narratologists have given to the narrative functioning of the present tense in first-person novels, they have given considerable attention to a first-person fictional form where the present functions as a non-narrative tense: the autonomous interior monologue […].” (Cohn 1999: 99) 11 Although I see some difference between the two types, I don’t want to fully buy into the distinction Cohn establishes: “Unlike the autonomous monologue—where the non-narrative quality of the silent, self-addressed discourse annuls a priori all semblance of mediated presentation—simultaneous narration does imply a narrative situation, but one that defies all manners of picturing it on verisimilar lines”. (105) To me it seems most important what the two related forms have in common: that the words are not told by anyone or to anyone. It is certainly true, I think, as David Herman suggested to me, that “an author has performed an act of narrative communication part of whose design cues readers to ascribe mental states and dispositions to the character at issue”, and that “the author is telling a story about a character whose mind has the features evoked via the interior monologue,” and I see why one would wonder: “doesn't the author's encompassing act of narrative communication make the label “non-communicative rupture’ problematic in this context?” I would like to add, therefore, that while there is a global communication from author to reader, this description applies to any written narrative whether natural or unnatural, mimetic or nonmimetic, fictional or non-fictional, it hardly captures the specificity of the mentioned passage in the “fictional present” and the consequences of using techniques of fictionality and unnatural narration. To do this, I believe we have to disentangle the words from a narrator. The author violates the limits of narratorial communication, but also of real world discourse in which “I” inevitably refers to the speaker. Not so here, in so far as the protagonist is not speaking to anyone. It is a moment of fictional invention (whether the narrative is globally a fiction or not), not a moment of report by the character-narrator. In my view, attributing the words to the author is correct but only in the sense that he is producing a fictionalized passage that is not reducible to naturally occurring oral discourse.

62

Henrik Skov Nielsen

Like present tense narrative with its unrealistic simultaneity between speaking and doing, the narrative imperative contravenes schemata of real-life storytelling by foregrounding the process of invention […] by employing the imperative and the narrative present tense such second-person fiction additionally foregrounds the act of invention and illustrates how telling generates the story in the first place, rather than representing and reproducing in narrative shape a sequence of events that is prior to this act of linguistic creation. (262)

The point that present tense narration presents a challenge to standard accounts of narrative is thus not a new one; although it has often been overlooked how strange it is even in simple passages like the one from Glamorama. The question is what to conclude from that and how to interpret these narratives. Certainly no law prevents us from trying to imagine comparable real-world situations as Fludernik tells us that we do: “When readers read narrative texts, they project real-life parameters into the reading process and, if at all possible, treat the text as a real-life instance of narrating.” (2001: 623) There may be some truth to the claim that this is what many readers tend to do,12 but that does not oblige us to repeat the projection at a methodological level. And do we really, methodologically, theoretically and didactically want to go there? Is it even possible? Do we want to ask our students to imagine situations in which a person could in fact be telling the story as he goes along, maybe using some kind of dictaphone, or invite them to imagine him after the events telling the story (while for some reason using the fictional present) to some or other narratee—the imagined presence of whom has left no trace in the narration at all? Do we want, then, to question the accurateness of his memory—certainly a pertinent reservation in non-fictional cases? All of these are rhetorical questions: clearly, we do not. However, the quotes nicely points to a rather diffuse ambiguity in the relation between natural and unnatural understandings, and to a difficulty in translating theoretical findings and distinctions into viable reading strategies.13 12

Although I believe it is not at all a self-evident fact that readers do such a thing. In fact, the willingness of most readers not to question page long renderings of monologue in character narration and their being perfectly comfortable with narration in zero focalization seems to suggest otherwise. Sure, these techniques have been conventionalized. But in my opinion, that fact does not amount to saying that readers treat them as “real-life instance[s] of narrating”. 13 A large complex of problems is associated with this connection between theory and practice. Fludernik has a strong point in showing that we use storytelling scenarios with which we are familiar as templates for making sense of literary narratives, and then deviate from those templates if the text appears to depart from or actively resist them. This process of deviation in turn causes the templates to be modified and re-added to the repertoire of resources for sense making in new, altered forms. But exactly what happens, and how do “deviations” turn into viable reading strategies? These are problems still to be explored. Sometimes the same

Fictional Voices? Strange Voices? Unnatural Voices?

63

Per Krogh Hansen’s strong article on exactly the same topic, titled “First person, present tense. Authorial presence and unreliable narration in simultaneous narration,” seems to me to hesitate in a similar way between choosing a naturalizing reading or not. Discussing a Danish novel and Dennis Cooper’s My Loose Thread, Hansen nicely frames the problem. He also provides very useful distinctions in his discussion of cases that could possibly be considered a form of historical present and cases that cannot. In his discussion of My Loose Thread Hansen writes tellingly: Nonetheless, the most outstanding quality of the narration is that it is very difficult to naturalize. (Hansen 2008: 321)

In the rest of the article he goes on to discuss the phenomenon in terms of unreliability, and takes as his main example another text by Bret Easton Ellis, Less than Zero. Hansen suggests that […] by the end it points out one direction for the act of naturalization. (329) […] one could see the FPPT [first person present tense] as an act of Freudian repetition, and thereby ascribe to Clay a traumatic trait. (331)

A reading like this is not unpersuasive, and it can probably not be disproved. When it comes to the general consequences, Hansen suggests that naturalization will sometimes succeed and sometimes not, and that a text without any temporal distance will not be a narrative: One might go as far as saying that at least a minimal temporal distance is necessary for structuring incidents in a narrative pattern. (325-26)

This, then, is the point on which I disagree. In my view, we need not impose real-world necessities on fictional narratives. In Glamorama, Less than Zero and other first person present tense narratives, we need not imagine situations that would allow for non-fictional narratives in the same form. We should prefer—in light of the above—the rather simple and economical explanation that this is one of many instances in which a fictional narrative foregrounds its own inventive powers and resistance to real-world descriptions and real world necessities (cf. Fludernik 1996: 262 on “foregrounding the process of invention”). This then has immediate consequences for the interpretation since we are allowed to trust narration that couldn’t possibly be reliable real world narration. Matt DelConte has a good point about this resistance to real world description in his article “A kind of problems emerges when we turn reading strategy into viable theory. Take, for example the Cyclops episode in Ulysses. In actual practice hardly anyone would ascribe to the character narrator the kind of memory, style and command of several languages that we find in this chapter. We probably wouldn’t admire Joyce as much if we thought that a less than admirable drunkard could extemporally produce the very narrative. But how do we theorize as character narration a reading practice that actually does not ascribe this chapter’s equilibristic movements to the character narrator?

64

Henrik Skov Nielsen

Further Study of Present Tense Narration: The Absentee Narratee and Four-Wall Present Tense in Coetzee's Waiting for the Barbarians and Disgrace”: There exist very few examples of simultaneous present tense narration, and there seem to be two main reasons for its rarity. Firstly, it is an unnatural form of narration, something that doesn’t have a clear, real-world analogue. (429)

In my view he also correctly concludes that it makes no sense to speak about a narratee here: [T]he scene I open this essay with from Richard Ford’s Independence Day explicitly recognizes the absence of a narratee: “only there’s no one. No one here or anywhere to say this to (217). (431)

DelConte, however, seems to arrive at the right conclusion for the wrong reason and is therefore not going far enough. He says: [W]e encounter a number of narratives (perhaps most noticeably epistolary narratives) whose narratee is not present at the time of the narrating. We also frequently encounter narratives that provide so little detail about the narrative occasion that we cannot determine the presence or absence of a narratee, in which case the default is to assume there is a narratee with the narrator; that is, we can easily imagine that the narrator is talking to someone even if that someone remains uncharacterized. But simultaneous narratives do in fact supply enough detail of the narrator’s situation (specifically that his location coincides with the story-level scenes) for us to recognize that the narratee is clearly not present at the time of narrating […]. (431)

DelConte is right in that there is no narratee. The reason, however, is not that the protagonist is in a situation at the time of narration that does not allow for a narratee. This situation just as much disallows an act of narration as it does the presence or absence of a narratee. Rather than a present narrator and an absent narratee, there is no narration to or from anyone at the level of the storyworld. Another brilliant article on the topic is Phelan’s “Present Tense Narration, Mimesis, the Narrative Norm, and the Positioning of the Reader in Waiting for the Barbarians” (1994). Phelan very convincingly demonstrates how norms and standards can change in fictional contexts: In the case of present tense narration, the fictionality of the text is foregrounded, but this foregrounding does not impede our mimetic engagement with the characters. (229-30)

He concludes (revising a mimetically based suggestion by Fleischmann): Nonfictional narratives refer to specific real experiences that occurred in the past and are accordingly reported in a tense of the PAST. Fictional narratives refer to imagined experiences presented as if they were real, sometimes through imaginary instances of narration. (230; emphasis in the original)

Fictional Voices? Strange Voices? Unnatural Voices?

65

Likewise, I am highly sympathetic to his subtle description of how he does not mean to rule out, say, the use of the simultaneous present in nonfictional narrative but rather wants to emphasize that such a technique will achieve its effect in part by its deviation from the norm—and to suggest that the effect is likely to be different from the effect of the present tense in fictional narrative. (230) One point of slight disagreement, though, is when Phelan goes on to say: It will be apparent, I think, that there is neither any plausible “occasion of narration” here nor any violation of narrativity. This narrator is doing the impossible—living and telling at the same time. Furthermore, his discourse locates us in the genre of narrative: his subjectivity is obvious. (233-34)

Right, there is no “occasion of narration” and this is a slightly paradoxical situation, but to describe this as the narrator “doing the impossible— living and telling at the same time” runs the risk of placing the paradox and the impossibility at the story world level as if this was a story about a character capable of the impossible. I would prefer to say (and I think this is maybe also what Phelan means) that we have to just accept this act of narration even though it is an impossible real world narrative act. Still, Phelan usefully stresses that the subjectivity of the narration is the subjectivity of the character, a fact that troublingly places the narration in the proximity of and at a distance from the character at the same time. My own response to this problem is to compare it to reflector or internally focalized narratives in which there is a very similar sense of the character’s subjectivity and point of view and often even idioms and mistakes—all of which does not make the character a narrator. Obviously, there is much work to be done on this complex of problems. You-narration Another mounting trend in contemporary fiction is that of narration in the second person, you-narration. Rolf Reitan provides a thorough review of the field in this volume and my specific purpose here is not to repeat a study like that. Instead, I want—in the specific context of the understanding of unnatural narratives as narratives that cue the reader to interpret in ways that differs from the interpretation of real world acts of narration— to point out how problems connected to you-narration are in many respects similar to those connected with first person present tense narration. In Unnatural Voices (2006), Brian Richardson makes a comprehensive list of second-person narratives and does a really good job in defining and delimiting the field so that it doesn’t just include any narrative employing

66

Henrik Skov Nielsen

the second person pronoun, since this is used also in several standard situations in which an author unambiguously addresses his reader, and in apostrophes. Richardson aptly writes: We may define second person narrative as any narration other than an apostrophe that designates its protagonist by a second person pronoun. (19)

Notice that Richardson does not talk about addressing the protagonist. He continues: It is important to note that second person narration is an artificial mode that does not normally occur in natural narrative […]. (19)

I think Richardson is surely right, but would want to add a few words on why that is. We talk about and to each other using “you” all the time. Isn’t it the case, then, that second person narratives are the most natural things in the world? To answer this we have to remember, firstly, that using the “you” as a disguised form of “I” or “everybody”,14 as in “you just get so mad in these kind of situations, don’t you?” does not count as you-narration since it does not specifically designate the protagonist but rather the speaker as part of an imagined community. Secondly, the curious thing about most fictional second person narratives (with Butor’s La modification as a prominent and classical example) is that although the protagonist is designated by “you” throughout these narratives, nothing at all suggests that he/she feels in any way addressed. He is not hearing voices, does not feel he is being spoken to, and he does not respond to the narrative.15 In short: nothing except the very use of the second person pronoun suggests that he is being addressed16. So if, in natural linguistics, the first person pronoun designates “the speaker”, third person “the one spoken about”, and second person “the one spoken to”, then it seems that in many fictional second person narratives the pronoun loses this functionality. The protagonist is referred to and designated, but not addressed by the second person pronoun. He is just as ignorant of being the centre of a narrative as are the protagonists in third person narratives. Outside fiction, then, in, say, conversational narratives, the referent of “you” is inevitably addressed, and obviously not created by the pronoun.

14 15

This is sometimes referred to as “generalized you”. At least in what Richardson call standard cases and what Reitan shows to be more or less the only “real fictional, second person” narratives. For the point I am making here it is not important, if it holds true for some or all fictional, second person narratives. For a much more comprehensive account of second person narration and its effects on the reader, cf. Reitan’s article in the current issue. 16 Cf. Reitan in the current issue: “Summing up: [...] Only category C [Narrative you referring to protagonist, but not used as address] covers proper second person narratives [...]”.

Fictional Voices? Strange Voices? Unnatural Voices?

67

In most fictional second person narratives, the referent of “you” is inevitably created, and obviously not addressed by the pronoun17. Again, this difference is so fundamental that it would be theoretically and didactically unsound to teach ourselves and our students to put these kinds of narratives into communicational models based on real-life storytelling situations. This time, I will refrain from exemplifying the futility, since the usefulness of examples is not to suggest that any of the theorists mentioned would want to go there, but rather to demonstrate that in certain narratives real-life models run into problems that we still need to think about how to address. Common to the two first examples is the employment of a technique that sits awkward with communicational models of narrative—narratives without sender and without receiver, and without occasion and situatedness,18 but narratives all the same. Access to Other People’s Thoughts in First Person Narratives As Tammi also points out, it is perfectly possible in a range of fictional narratives to narrate in the first person while providing access to a range of minds beside that of the protagonist. Of the sub-subjects mentioned in this article this is the only one I have written extensively on in earlier articles,19 so I will not go into great detail again here. Instead I just wish to mention that from the earliest examples of first person narratives like The Golden Ass over classic novels like Moby Dick to postmodern fiction like Glamorama examples abound. Recently Rüdiger Heinze published a wellargued article, “Violations of Mimetic Epistemology in First-Person Narrative Fiction” (2008), where he mentions a whole range of very interesting examples of first-person narratives displaying knowledge that their first person narrators could not possibly possess. Keeping in mind all the examples (and let us here include the infamous recounting of the thoughts of the dying Bergotte alone in his bed in Proust’s first person narrative and Phelan’s scrutiny of the impossible narration in parts of The Great Gatsby) where first person narratives relate to the reader what the first person protagonist need not, will not, or cannot narrate, we can compare to 17

This seems to me completely in line with Reitan’s conclusion in the current issue: “What is it, this “voice”, if not a benevolent and, literally, an “authorial voice”; some version of the creator’s voice talking (no, not talking to) his creation, or more accurately, writing his creation, flaunting its fictionality […]“. 18 Except, that is, in the global sense of being a narrative from an author to her readers. The author is not, however, using the first person pronoun to refer to herself or the second person pronoun to address the protagonist—no one is. 19 Cf. Nielsen 2004, and Nielsen, forthcoming in Alber and Heinze.

68

Henrik Skov Nielsen

the situation in second person narration and say that just as nothing except the very use of the second person pronoun suggests that the person designated be the pronoun is being addressed there, nothing except the very use of the first person pronoun suggests that the person designated by the first person pronoun is narrating here. It even holds equally true for several first person narratives (including Glamorama) as for second person narratives that the protagonist is just as ignorant of being the centre of a narrative as are the protagonists in third person narratives. From whom, then, do the words emanate, and what do we do with them, and: how do we interpret narratives like this? The works and words discussed are unnatural in the sense that they designate and refer to a character as “I” without emanating from that character. The narrating “voice” does not emanate from the character but invents and creates a world, including the first person and his knowledge or lack of knowledge. Narrated I/Narrating I Time to raise an objection now: If the narrative act is disconnected from the pronoun normally signalling a speaker, then how about the instances in which this speaker is explicitly present as a narrating I? How about all the typical stuff; the non-experimental, realist narratives? How about the very normal cases in which—far from being transgressed—the limits in knowledge and memory of the character narrator is explicitly thematized? And, finally, how about the good old distinction about narrating I and narrated I (erzählendesIch/erzähltesIch)? These kinds of narratives and this distinction would seem to suggest that the standard case is that of a very close bond between protagonist and narrator. Let’s quote two examples—both of which are opening lines of a novel. I, Tiberius Claudius Drusus Nero Germanicus etc. etc. [...] is about to write the strange story of my life. (Graves 1961: 1) Not everybody knows how I killed old Phillip Mathers smashing his jaw in with my spade; but first it is better to speak of the friendship with John Divney because it was he who first knocked old Mathers down by giving him a great blow in the neck with a special bicycle-pump which he manufactured himself out of a hollow iron bar. (O’Brien 1993: 7)

In both examples, the act of narration is explicitly thematized, though most obviously in the first, and in both a “narrating I” is reflecting upon how to retrospectively tell about earlier events. The question, though, is what to conclude from this explicit split of the protagonist into “I, then” and “I, now”. First of all let us notice that, contrary to appearances, noth-

Fictional Voices? Strange Voices? Unnatural Voices?

69

ing in this split makes the narrative’s enunciator manifest for us. The distinction between narrated and narrating I works quite well for didactical purposes. It is often useful to ask the students to reflect upon the importance between the then and the now and upon the difference in knowledge between the two situations. The interpretation of sentences like “little did I know then, that this would later change my life….” hinges crucially on the understanding that the “I” can take up two very different positions in the same narrative. And yet, though at times a useful one, the feeling that a narrative can provide us access in this way to the time of the narrative act, is an illusion. There are only three options for the so-called “narrating I”: 1. Completely un-distanced from the narrated I (like in Glamorama) 2. Non-thematized and unlocated in space and time (Camus’ The Stranger and several of Hammetts novels would be examples of this case)20 3. Explicitly thematized and described in the text (cf. Graves and O’Brien above) For examples in the first category the assumption of simultaneity between living and telling leads to implausibilities, as we saw. For examples in the second category, speculations about the situation of the telling are at best useless and at worst misleading since there is often not only no identifiable, but even no imaginable, point in time and space in which the narrative act is situated. That holds true for a large number of first-person as well as third person narratives. For examples in the third category, finally, the paradox lies in the fact that in the very moment where the narrating I is explicitly identified and located in time and space, it becomes, inevitably, a narrated I. It is thus not only a telling person, but also a told one. What the third category presents, then, is not really a narrated and narrating I, but rather two (or more), temporally distant versions of the narrated I. In conclusion: The narrating I is either: a) absent and useless, or b) a narrated I; neither of the two cases guarantee to the reader that the narrative has to remain within the limits of real-life storytelling and the limits of the knowledge of the protagonist.

20

At least we are very hard pressed to think that Marlowe in his old days suddenly decides to tell stories from his life—including page-long exact renderings of dialogues from whiskeyfilled nights.

70

Henrik Skov Nielsen

Strange Voices in Poe’s “The Tell-Tale Heart” My intention in this section is to use Poe’s story to exemplify some of the general theoretical points above with a bit more analytical details. The overarching argument is that interpretations of the text need to take into account the ways in which it differs from real life narrative and the ways in which its voices are strange and differ from non-fictional, human voices. I divide this section into discussions of the conflation of voices in:  text and paratext  here and there, then and now  you and I Text and Paratext (Textual Conflation of Voices) On page 29 of Pioneer, where the original version of Poe’s text begins,21 is a short poem by “W. W. Story,” which serves as a prelude of sorts. This W. W. Story writes, just before Poe’s story, a poem that seems to correspond almost too nicely with what follows. The good Story shares his initials with Poe’s most famous double-ganger, William Wilson, but it must be William Wetmore Story, even though his full name is not given. The last stanza in his poem, called “Longing”, reads as follows: What is the worth of human art, If the weak tongue can never speak That which lies heavy on the heart, Even though the heavy heart should break. (Story 1843, 29)

Whether the reason for placing Story’s poem before Poe’s story was editorial, humoristic, ironic, arbitrary, or a combination of these, is hardly pos-

21

All quotations from “The Tell-Tale Heart” are from the original edition in Pioneer. The version found in most anthologies and, for instance, in The Complete Tales & Poems of Edgar Allan Poe, is Griswold’s version, which is unreliable in many ways. Some of the differences between the versions will be addressed below. Earlier considerations of this text include among the most important those of Rajan 1998, who undertakes a feministic reading, claiming that the narrator in The Tell-Tale Heart is a woman; Pritchard 2003, who discusses the strong sexual connotations in the description of the murder in the bedroom; Zimmerman 2001, who writes about Poe’s rhetoric and “oratory”, both generally and specifically in respect to the narrative, and Witherington 1985, who sees the reader as voyeur. Finally, Pillai 1997 should be mentioned as offering the best and most detailed close reading. Pillai’s article makes some of the same points about the role of the heart in the narrative, and about the self-annihilation that takes place. I will also consider this issue in the present article, but my perspective is quite different from his, which focuses on the reader and the historiographical aspect. In addition, Pillai could have made an even better case if he had used a more relevant version of the text (cf. below on the various versions).

Fictional Voices? Strange Voices? Unnatural Voices?

71

sible to say today, but the final stanzas nevertheless herald a narrative in which we in fact encounter words that “come straight from the heart” and that “the weak tongue can never speak.” No critics have, as far as I know, commented on the way in which Poe’s text appears in the original version. Its layout and position seems however to be of great importance. The above-mentioned page 29 of Pioneer where “The Tell-Tale Heart” begins has a curious disposition, in part because it is difficult to say altogether where the text actually begins. At the very top of the page is written THE TELL-TALE HEART in capital letters, which on the face of it merely seems to indicate that the narrative begins somewhere on this page, since what follows next is apparently not “The Tell-Tale Heart,” but rather, as mentioned above, the poem “Longing” by Story, which is indeed about the heart and contains the word “heart” in both the first and last stanzas. After the last stanza in Story’s poem, the title of Poe’s narrative is repeated again, and only one ripple mark and 2 centimetres separate the two occurrences of “heart” in the last two stanzas of Story’s poem from HEART in the title of the narrative “THE TELL-TALE HEART.” Just prior to Poe’s text, the impersonal, non-prosaic voice of poetry has thus described what the heart can bear but the tongue cannot say. Not until the title of the narrative has been repeated, is the source of the tale finally stated, the next words, “BY EDGAR A. POE”, establishing the name of the author in capital letters. However, even now the prose does not begin. The narrative about the heart, or the heart’s narrative, has either long since begun or not yet arrived at its first word, because although Poe’s name has signed the words, another short poem and an italicized signature follow like a preamble: Art is long and Time is fleeting, And our hearts, though stout and brave, Still like muffled drums, are beating Funeral marches to the grave. Longfellow. (Poe 1843: 29)

We still have not reached the start of the prose, since between Poe’s name and the first word of the first-person is Longfellow’s poem and signature. On the other hand, “The Tell-Tale Heart” must have begun at the repetition of the title (even though, oddly enough, including Longfellow’s poem is without precedence22 (the same goes for Story’s)), and the narrative thus begins by saying what someone else says, by borrowing its words from someone else—from Longfellow’s poem. Longfellow’s poem refers precisely to the importance of the heart. The heart beats, but this activity by the most vital of all organs is not described as a life-sustaining function, 22

See, for instance, Poe 1958 and 2001.

72

Henrik Skov Nielsen

but rather as a funeral march. The rhythm of the heart beats to the tempo of death. Heartbeats. “True!” After “the grave” follow the first words of the narrative: True! —nervous—very, very dreadfully nervous I had been, and am, but why will you say that I am mad? (Poe 1843: 29; Poe’s italics)

It is true. It is true that I was, and am, nervous. From the very beginning, from and in the very first sentence, the “I” is a Cartesian subject reflecting its existence, and it reflects through a double anticipation of and response to others’ speech: “I am nervous,” “true,” but “I am mad”—“you will say that”. About me. Text and paratext and the relationship between the two are all about the difficulty of knowing who speaks when someone says “I”. Here and There, Then and Now (Spatial and Temporal Conflation of Voices) The narrative ends as follows: Was it possible they heard not? Almighty God! —no, no! They heard! —they suspected! —they knew! —they were making a mockery of my horror! —this I thought, and this I think. But anything better than this agony! Anything was more tolerable than this derision! I could bear those hypocritical smiles no longer! I felt that I must scream or die! —and now—again! —hark! louder! louder! louder! louder! —“Villains!” I shrieked, “dissemble no more! I admit the deed! —tear up the planks! —here, here! —it is the beating of his hideous heart!”(1843: 31; italics in the original).

The last page of the text alternates between the past and the present tense, and the deictic present is used as well: “No doubt I now grew very pale […]—and now—again! —hark!” (ibid.; italics in the original). The two levels that should be made up of the time of narration and the time of the narrated tend to merge. There is thus no situatedness outside the narrative; no narrating “I” now safely outside the narrated then. Exposure and repetition converge and become indistinguishable so that it becomes impossible to decide whether the last words “now—again” and “here, here!” belong to the now of the telling or the then of the told. In addition to this temporal convergence, an extremely insistent rhythm emerges toward the end of the narrative. The affect of time of the narrated comes to completely negate the studied calm of the time of narration (“observe […] how calmly I can tell you the whole story”). In the end, the heart completely takes over the narrative and we hear more than anything else the heart's beating, which, upon reconsidering the text, we are clearly told in the very last words: “here, here! —it is the beating of his hideous heart!”.

Fictional Voices? Strange Voices? Unnatural Voices?

73

Here, at the end, the narrative and the heartbeat converge. At the boundary lines of language and logic, rhythm and the unspeakable take over: it is the heart that is narrating here. The sound of the heartbeat that tells the heart's story is in a sense invisible, but still present throughout the text as a sound. At this point, we hear it like the ‘I’, but we cannot find it by looking for it, as the policemen do. The very last words of the text create and hide this sound: “here, here! —it is the beating of his hideous heart!” Looking “here”, we see no heart beating, but when we listen to it, it seems to shout at us: “hear, hear!” Symptomatically, rather than returning to the frame “The Tell-Tale Heart” ends with a moment of horror and revelation, and more importantly, what is told—in the form of reappraisals, doubts, fear and nervousness—is directly reflected in the narrative. Poe's story begins with the following words: “[…] and observe how calmly I can tell you the whole story,” but when the narrative arrives at the upsetting events it is itself upset, faltering and breathless, and even the deictic present tense is used. If the narrative and the narration thus seem to converge, and if what is told seems to be playing out here and now, it becomes difficult for us to conceive of a narrative as narrating something that has occurred. What happened then and there is indistinguishable from what happens now and here23. You and I (Conflation of Personal Voices) The text frequently suggests that it is not the old man whom the firstperson pronoun wishes to erase, but rather “his eye.” Take, for example, the following passages: I loved the old man. He had never wronged me. He had never given me insult. For his gold I had no desire. I think it was his eye! —yes it was this! He had the eye of a vulture […]. I found the eye always closed; and so it was impossible to do the work; for it was not the old man who vexed me, but his Evil Eye. [...] The old man was stone dead. His eye would trouble me no more. (1843: 29-30; italics in the original) 23

I acknowledge, as David Herman pointed out to me, that one could interpret this “fusion, or confusion, of time-frames [as] the hallmark of traumatic or traumatized experience, in the context of which the present, narrating self can find no overarching frame in which to make sense of past events and integrate them into a model of the self as continuously unfolding in time”. It could be argued, though, that the narrator actually shows few signs of remorse and does not necessarily seem traumatized, but rather seems satisfied and proud by his act. Secondly, one could argue that the natural explanation does not quite do justice to the elements that contradict the natural storytelling situation in which the act of narration at the time of telling comprises a unity of “I, here, now.” As I hope to show here, the text complicates each of the dichotomies between I/him, here/there and now/then.

74

Henrik Skov Nielsen

In most of the current versions that follow Griswold, the final sentence of the first quotation reads as follows: “One of his eyes resembled that of a vulture […]” (Poe 2001: 199), which changes the structure of the sentence so that it becomes comparative and changes the singular “eye” to the plural “eyes.” The meaning of the sentence in the original version diverges semantically as well as phonetically from Griswold’s version. Semantically, in the original version, the reader is cued to thinking of the narrator as being observed by the eye of a vulture, which, as any reader of pulp fiction knows, is on the lookout for dying or dead creatures. Phonetically, it is significant that Poe sticks to the singular form throughout the tale. It is thus quite obvious that it is not the old man whom the ‘I’ wishes to kill, but rather his eye, “his Evil Eye.” For it is not only the eye, “the Evil eye,” but also “the evil I” that is destroyed by the murder (see also Pillai 1997, and Robinson 1965). It is precisely from the point when the first-person believes that he is finally free from his doppelganger and has liberated himself from “the eye,” that the heart takes over. Henceforth the first-person no longer is master of his own speech. Furthermore, this point is doubly exposed as the point in the apartment where the first-person hears the heart telling tales about him, and he, against his will, must hear himself being exposed and as a point in the tale, which—just as the heart’s speech is to be rendered—seems even more clearly to be overtaken by the heart, not in the sense that it is literally speaking in the narrated world, but in the sense that it manifests its own tale-telling revelation of the protagonist as a murderer against his will at the level of narration. Concluding Remarks on “The Tell-Tale Heart” “The Tell-Tale Heart” seems to me to take the inventional aspect of fiction to the extreme by contradicting almost all elements of a natural storytelling situation in which the act of narration at the time of telling comprises a unity of “I, here, now.” In fact, the text seems strategically to undermine all these three concepts since now is “now – again” and thus both then and now, and since similarly, “here” is elusive in its homophone appeal to both sight and hearing and in its ambiguous reference to the situation of the told as well as the telling, and since, finally, the play on ‘I/eye’ ambiguously refer to both the speaker and his antagonist. In this text, which is so obsessively interested in sight and hearing, it seems unlikely to be a coincidence that the use of homophones is connected, respectively, to exactly sight and hearing.

Fictional Voices? Strange Voices? Unnatural Voices?

75

Theoretical Implications of Unnatural Narratives: From Unnatural Fiction to Fiction as Unnatural Returning to Tammi, he argues for the importance of the study of the exceptions to the rule: Therefore, the way to go for literary narratology, I argue, is squarely in the opposite direction from that taken in the wake of the narrative turn—away from the model-building and standard definitions, towards studying, rather, the subversive and strange, previously, untheorized or insufficiently theorized cases: the glorious exceptions to the rules […] (Tammi 2006: 29)

I am completely sympathetic to that, obviously, but still I find it important to also try to return to the rules and to examine what the exception in turn tells us about the rule itself. Some of the conclusions above were:  In first person present tense narration nothing except the present tense itself suggests that someone is telling while acting.  In second person narration nothing except the pronoun suggests that the protagonist is being addressed.  In first person narratives that transgress the limits of the knowledge of the protagonist nothing except the pronoun suggests that the protagonist is a narrator. What these conclusions have in common is most explicitly that the grammatical functions of verbs and pronouns do not work exactly the same way in these narratives as in non-fictional, conversational one. One could ask, then, tentatively: If stream of consciousness and first person present tense narration (in the specific form of the fictional present) is non-narrative in the sense of not telling something to somebody, then what does this tell us about fictional narration in the past tense? Would that be a fundamentally different case? Similarly: if the second person pronoun loses its normal grammatical function of addressing in second person fiction, what does this tell us about first and third person pronouns in fictional narratives? Is natural linguistics fully applicable here? Finally: Even if it is a relatively widespread view of fiction that the “work of fiction creates the world to which it refers by referring to it” (Nielsen 2004: 145, building on Cohn 1999: 13), what are the consequences—in the standard cases and in the exceptional ones? Is the protagonist (along with all the other things in the fictional universe) always created in the reference? Or does this happen only in some special cases? The consequences of a consistently inventional view on fiction are still almost entirely unexplored. Especially so when it comes to first-and second person

76

Henrik Skov Nielsen

narratives, and when it comes to aligning this view with rhetorical and communicational models of narratives. I will end this article by pointing to a few possible paths for future research, and by asking some questions about the potential recursivity between the exception and the rule. Perspectives Let us return to an early question in this paper: Given that we accept the view that some narratives are natural; do we also have to accept that all narratives operate according to the rules and principles of these natural narratives? Early in Towards a ‘Natural' Narratology, Fludernik makes the premises clear: In recent years new developments in linguistics have introduced the term ‘natural’ to designate aspects of language which appear to be regulated or motivated by cognitive parameters based on man’s experience of embodiedness in a real-world context” (Fludernik 1996: 17; cf. also Tammi 2008: 39).

The problem, as Tammi points out, is not so much to acknowledge that such aspects of language exist and are important. The problems arise when the specificity of these aspects of language are generalized and applied to all narratives. The use of language in much fiction strongly resists descriptions that would locate it in the experience of a specific embodiedness in a real-world context. One way to go from this observation (in my view a very useful one), would be to acknowledge that narratives can contain language and narration that cannot be explained solely by taking recourse to natural storytelling situations—and then make distinctions based on that acknowledgment. This would seem to be what Jim Phelan does when distinguishing between narrator functions and disclosure functions: […] communication in character narration occurs along at least two tracks—the narrator-narratee track, and the narrator-authorial audience track. Along the narrator-narratee track, the narrator acts as a reporter, interpreter and evaluator of the narrated for the narratee, and those actions are constrained by the narrative situation (a character narrator, for example, cannot enter the consciousness of another character); let us call these actions “narrator functions.” Along the narrator-authorial audience track, the narrator unwittingly reports information of all kinds to the authorial audience (the narrator does not know that an authorial audience exists); let us call this reporting “disclosure functions.” (Phelan 2005: 12)

Having identified and coined disclosure functions that designate narration which may transgress conversational or rhetorical situations by being unnecessary, implausible or even impossible at the level of the character nar-

Fictional Voices? Strange Voices? Unnatural Voices?

77

rator, the next question to ask is this: what happens when disclosure functions clash with narrator functions? In his article on first person present tense narration, Per Krogh Hansen presents an illuminating comment on my idea of an impersonal voice in fiction: This is a plausible explanation in so far as it concerns the description of the narrative situation and the problem of naturalization of FPPT narration. Yet, it could also be argued that an example such as this actually gives the reader the opportunity to naturalize the text by referring to the authorial framework of the text. [It] could then be subsumed under the heading of the author and understood as conventions of fiction in the form of what Phelan describes as disclosure functions. Hence, instead of describing these aspects as an “impersonal voice of fiction”, we might as well consider them as an indirect communication from the author. (Hansen 2008: 326)

While I am completely sympathetic to the idea of talking about the real author,24 I tend to think that this is a case in which one cannot both have one’s cake and eat it too. By this I mean that we cannot at the same time assume that the narrative emanates from a character narrator and that an author adds information for his own purposes. If on the other hand we agree that the narrative does not emanate from the character, it makes sense in many contexts to talk about the author as the enunciator rather than an impersonal voice. To put the same point a bit differently, Hansen’s suggestion seems in line with Phelan’s own stressing that “disclosure functions typically trump narrator functions.” (Phelan 2006: 323). But what exactly does “trump” mean here? Since the two functions are often mutually exclusive (when disclosure functions disclose what the character narrator cannot, will not, or need not narrate, this is incompatible with what is related by the narrator via narrator functions); “trump” seems in these case to take on the meaning of “explode” rather than “supplement”. In other cases disclosure functions work more discreetly and less incompatible alongside narrator functions. Whether they explode or supplement narrator functions probably will have to be decided on a case-to-case basis. In a larger context unnatural narration seems to me to make it preferable to add another track of communication to make it compatible with a rhetorical approach. Phelan suggested to me that “in unnatural narration, the disclosure functions proceed not along the narrator-authorial audience track but the author-authorial audience track as the author, in the interest of disclosure, violates the limits of narratorial communication”. I would happily subscribe to that description. Compared to the description quoted 24

Cf. Nielsen, “Natural Authors, Unnatural Narratives,” forthcoming in Alber and Fludernik (eds.): Post-Classical Narratology.

78

Henrik Skov Nielsen

above with the two tracks (the narrator-narratee track and the narratorauthorial audience track), this seems to me a very welcome addition. I much prefer the description that the author violates the limits of narratorial communication over the description that the narrator unwittingly reports information, since I believe that there is no report at the local level and at the level of the character-narrator. In this formulation, then, Phelan’s model is close to mine—or rather mine to his. This model reinforces the idea that only the author, and not the narrator, would be required in order to explain the specific phenomena discussed. In unnatural narration, then, the author violates the limits of narratorial communication, but also of real world discourse. In my view unnatural narratology has much to gain from the rhetorical approach and its insistence on purposes, means and ends. In each narrative there are reasons and causes for the choice of unnatural techniques and diversions from real-life storytelling situations. And vice versa: The only indisputable communication in narrative fiction is from author to reader. The rhetorical approach is (among other things) examining the means and ends and techniques by which an author realizes or fails to realize his or her intentions.25 Unnatural narratives do not necessarily negate or mock natural conditions. It is just as much the other way round: Certain conditions impose the experience of embodiedness in a real-world context. The embodied and the unembodied narratives are equally exceptional, as are also the fictional and the non-fictional. So much seems potentially to be gained by stressing the ways in which a fiction author is free to choose techniques and strategies of fictionality that go beyond the specific limitations and conditions imposed by certain natural storytelling situations. If the study of literature benefits tremendously from insights from narrative theory, narrative theory will benefit from reflecting on the way real worlds, minds, and acts of narration are subject to exceptional restrictions compared to all possible worlds, minds, and acts of narration. Thus, the idea of unnatural narratives is perfectly compatible with rhetorical models like that of Phelan and not less with that of Richard Walsh, who draws on Sperber and Wilson’s reading of Grice to argue that readers will optimize the relevance of many narratives by assuming that the narrative is fictional (Walsh 2007: 13-37). Anything can be naturalized but only sometimes will the result be useful or successful. Unnatural narratives, then, cue readers to engage in decoding activities that differ from those he would use in, say, face to face communication. Similarly, the reader will often (consciously or uncon25

Cf. Phelan 2004: 630 seq.

Fictional Voices? Strange Voices? Unnatural Voices?

79

sciously) apply rules of interpretation different from those applied to natural, conversational narratives in order to maximize relevance and understanding. Nothing prevents a reader from dismissing fiction as the obvious lies of an author, and nothing prevents a reader from applying the specific restraints of natural narratives also to unnatural narratives, but in both cases relevance will be minimized. If unnatural narratives cue readers to employ a set of decoding strategies that she wouldn’t apply to oral storytelling, then that makes a significant difference for the interpretation of innumerable specific cases. Returning to the specific example of Glamorama it is—in addition to being told in the first person present tense—in some respects a classical doppelgangernarrative. The protagonist and first person narrator Victor Ward apparently has a double, and gradually this double takes over his identity. In the end one Victor—and everything seems to indicate that he is the one we have followed throughout most of the book—dies in Italy while the other Victor, his double, enjoys life in New York. The really odd and unnatural thing about Glamorama and its act of narration, however, is that not only does the double overtake the identity of the first-person narrator on the thematic level and in the narrated universe; he even becomes the enunciator of the pronoun “I”. The double becomes the narrating narrator and thus takes over part of the narration. This phenomenon certainly seems to correspond to no manner of real-world, natural discourse. But the understanding of just the basic events and the storyline in Glamorama hinges crucially on understanding this pronominal takeover – and this understanding, in turn, is an interpretational consequences that we arrive at if and only if we do not naturalize the enunciation. My ultimate claim is that facing unnatural narratives readers will often not be able to optimize relevance and understanding by applying rules of interpretation that normally are applied to everyday, conversational narratives and real-world reports. Instead, the reader has the option to try to maximize relevance by applying a qualitatively different set of interpretational rules. For example, the reader can strategically assume that it actually makes sense to trust narrative details which the first person narrator cannot possibly know.26 Exactly the same option is available in everyday storytelling, but here relevance will hardly be maximized. The assumption that the narrator is lying or guessing will usually be the better option. Similarly, the reader has the option to assume that the protagonist in first person narratives is designated by the pronoun “I” while its enunciator is not, and that the protagonist in second person narratives is designated by the pronoun “you”, but as a rule not addressed by it. Here the op26

Which means assuming that they come to her as what Phelan calls “authoritative”.

80

Henrik Skov Nielsen

posite option exists too: the reader can assume that the protagonist is deaf or that he refuses to respond or react even though he is consistently addressed throughout. Again, this will hardly maximize relevance, and I believe it makes sense to say that the narrative cues the reader to apply other strategies (for example, if nothing else suggests that the protagonist is deaf and if he does in fact respond in completely normal ways when addressed by other characters). The reader can choose to assume that a heart can voice a narrative, or that a protagonist can be invented in the present tense without necessitating the conclusion that he is telling and acting at the same time. In each case interpretation will be different from interpretations of real life storytelling. Finally, though in each case the options will be agonistic, they may nonetheless be negotiable if guided by the principle that interpretations must be judged by how they optimize relevance and understanding. Unnatural narratology, therefore, is not a hermeneutics, but an argument competing with other arguments. It cannot be ruled out a priori on the assumption that all narratives work “naturally”, or that readers always read in accordance with that assumption. References Alber, Iversen, Nielsen and Richardson 2010 “Unnatural Narratives. Unnatural Narratology. Beyond Mimetic Models., Narrative 18.2 (May), 113-136. Alber, Jan and Rüdiger Heinze (eds.) [In progress] Unnatural Narratology, Berlin and New York: de Gruyter. Cohn, Dorrit 1999 The Distinction of Fiction, The Johns Hopkins University Press, Baltimore. DelConte, Matt 2007 “A Further Study of Present Tense Narration: The Absentee Narratee and Four-Wall Present Tense in Coetzee's Waiting for the Barbarians and Disgrace”, Journal of Narrative Theory 37.3 (Fall), 427-446. Ellis, Bret Easton 1999 Glamorama, New York: Alfred A. Knopf Fludernik, Monika 1996 Towards a ‘Natural’ Narratology. London and New York: Routledge. 2001 “New Wine in Old Bottles? Voice, Focalization, and New Writing”, New Literary History 32, 619-38. 2003 “Natural Narratology and Cognitive Parameters”, in: Narrative Theory and the Cognitive Sciences, David Herman (ed.), 243-67. Stanford, CA: CSLI Publications. Graves, Robert 1961 [1934] I, Claudius. New York: Random House. Hansen, Per Krogh 2008 “First person, present tense. Authorial presence and unreliable narration in simultaneous narration”, in: Elke D’Hoker & Gunther Martens: Narrative Unreliability in the Twentieth-Century First-Person Novel, Berlin: De Gruyter, 317-338.

Fictional Voices? Strange Voices? Unnatural Voices?

81

Heinze, Rüdiger 2008 “Violations of Mimetic Epistemology in First-Person Narrative Fiction,” Narrative 16.3, 279-97. Nielsen, Henrik Skov 2004 “The Impersonal Voice in First-Person Narrative Fiction”, Narrative 12.2, 133-50. [Forthcoming] “Natural Authors, Unnatural Narratives”, in: Postclassical Narratology: New Essays, edited by Jan Alber and Monika Fludernik, Columbus: Ohio State University Press. [In progress] “Unnatural Narratology, Impersonal Voices, Real Authors, and Non-Communicative Narration”, in: Unnatural Narratology, edited by Jan Alber and Rüdiger Heinze, Berlin and New York: de Gruyter. Phelan, James 1994 “Present Tense Narration, Mimesis, the Narrative Norm, and the Positioning of the Reader in Waiting for the Barbarians”, in: James Phelan and Peter J. Rabinowitz (eds.), Understanding Narrative,Columbus: Ohio State University Press, 222-245. 2004 “Rhetorical Literary Ethics and Lyric Narrative: Robert Frost’s ‘Home Burial’”, Poetics Today 25.4 (Winter), 627-651. 2005 Living to Tell about It: A Rhetoric and Ethics of Character Narration, Ithaca, NY: Cornell University Press. Phelan, James with Robert Scholes and Robert Kellogg 2006 The Nature of Narrative, 2nd Edition, New York: Oxford University Press. Pillai, Johann 1997 “Death and its Moments: The End of the Reader in History”, MLN 112.5, 836-875. Poe, Edgar Allan 1843 “The Tell-Tale Heart”, The Pioneer (January), 29-31. 1958 [1946] The Complete Poems and Stories of Edgar Allan Poe, New York: Alfred A. Knopf. 2001 The Complete Tales and Poems of Edgar Allan Poe, New Jersey: Castle Books. Pritchard, Hollie 2003 “Poe’s ‘THE TELL-TALE HEART”, Explicator 61.3, 144-148. Rajan, Gita 1998 “A Feminist Rereading of Poe’s ‘The Tell-Tale Heart’”, Papers on Language and Literature 24.3, 283-300. Richardson, Brian 2006 Unnatural Voices: Extreme Narration in Modern and Contemporary Fiction, Columbus: Ohio State University Press. Richardson, E. Arthur 1965 “Poe’s ‘The Tell-Tale Heart’”, Nineteenth-Century Fiction 19.4, 369-378. Sternberg, Meir 1982 “Proteus in Quotation-Land: Mimesis and the Forms of Reported Discourse”, Poetics Today 3.2, 107-56. Story, W. W. 1843 “Longing”, The Pioneer. A Literary and Critical Magazine, 1 (January). Tammi, Pekka 2006 “Against Narrative (‘A Boring Story’)”, Partial Answers 4.2, 19-40. 2008 “Against ‘Against’ Narrative”, in: Narrativity, Fictionality, and Literariness: The Narrative Turn and the Study of Literary Fiction, Lars-Åke Skalin (ed.), Örebro: Örebro University Press, 37-55. Walsh, Richard 2007 The Rhetoric of Fictionality: Narrative Theory and the Idea of Fiction, Columbus, OH: Ohio State University Press. Witherington, Paul 1985 “The Accomplice in ‘The Tell-Tale Heart’”, Studies in Short Fiction 22.4, 471-475.

82

Henrik Skov Nielsen

Zimmerman, Brett 2001 “Frantic Forensic Oratory: Poe’s ‘The Tell-Tale Heart’”, Style 35.1, 34-49.



PEER F. BUNDGAARD (Aarhus)

Significant Deviations: Strange Uses of Voice are One among other Means of Meaning Making1

But the voice is his natural disguise Paul Simon

Preamble The purpose of this paper is to shed light on the semiotic function served by what is commonly called “voice” in narrative art. Its scope is therefore quite restricted: I aim at characterizing voice, the narrator or the narrating instance, as one among other meaning making devices in narrative art. In a nutshell: In what ways do authors use voice to produce this and that meaning effect? Given its scope and explicit goal, the present approach rests on a tacit claim which may be worthwhile unwrapping: I guess that there is no need to make such an ontological fuss about the narrator, with whom “voice” is often identified, if by narrator we understand some substantial entity whose mode of being is essentially different from that of the formal narratorial devices of presentation. In the present account, the narrator is neither a necessity nor an impossibility. It is just a real possibility, that is to say, a semiotic and aesthetic resource at the disposal of any given author who may choose or not choose to make use of it and whatever way he may find convenient for his purposes. Just as, in visual art, there exist paintings where perspective plays no role, and paintings where it does play a role, and just as there exist paintings with incongruous, compressed or double perspective, there exist narratives where the narrator is no real deal, others where voice indeed is an audible issue, and yet others with 1

I am indebted to Tom E. Griffith for most valuable comments, critique and expressions of incomprehension. I would also like to thank the editors Per Krogh Hansen, Stefan Iversen, Henrik Skov Nielsen and Rolf Reitan for their very helpful suggestions.

84

Peer F. Bundgaard

incongruous, double voices, etc. And just as no sensible art historian or art ontologist would claim that a given presentational device (such as perspective or the golden section) follows or does not follow with necessity from the visual art format, literary, scholars should also refrain from deciding beforehand whether this and that presentational device, say, voice, is substantially embedded in or excluded from narration as such. In the following, I will therefore use “narrator”, “the narrating instance” or “voice” interchangeably. Since I make a point out of not engaging in a discussion about the ontological or logical status of the narrator irrespective of its semiotic function, I suppose it wouldn’t make much sense if I started lining up the positions in the bifurcation that defines the parting of (say, Hamburgerian and Genettian) ways in narrative theory as regards the status of the narrator. However, since my claim that I am not too preoccupied with that discussion might be considered a way of taking sides anyway, I probably ought to state my case in a more explicit way. My stance is pretty well represented by what Paisley Livingston calls the “minimal thesis.” It champions the idea that: [...] for some narratives, the content of the narrative is indeterminate with regard to questions as to how the events are knowable to some (implicit) audience. In some cases, no implicit presenter or narrator need be imagined as part of the utterance’s contents, and indeed the imaginative hypothesis that the events are not narrated is to be preferred (Livingston 2001, 363).

This said, the main concern of the paper is, I repeat, to capture some of the semiotic functions voice can take on, which means that I intend to show how authors can exploit the narrating instance in order to trigger certain meaning effects. To do so, I will initially, in rather general terms, frame the issue of voice and its semiotic function as I consider it: viz. relative to what we mean by “natural”, “unnatural” or even “strange”, when we use such terms to qualify voice. Next, a sample of examples from both first person and third person narratives will flatfootedly show how meaning making obtains by virtue of specific uses of voice. Finally, the conclusion of the paper is that while it is certainly possible to determine the semiotic function of voice, whether strange or straight, it is so only on empirical, case to case grounds: there is no predetermined meaning effect linked to given variations in or applications of voice (such as change from hetero- to homodiegetic narrator, from authorial to figural mode of narration, or free indirect speech). Voice is a meaning-shaping device that applies to all narrative situations. The semiotic result of its application is therefore also relative to the specific whole in which it is used. I do, however, make a general claim as to this relation: an important subset of

Significant Deviations

85

meaning effects in narrative art can be characterized in terms of significant deviations from a given standard. If It’s Strange, it Means Something How does voice and meaning combine? A fundamental claim in the present framework is that some of the meaning effects obtained by means of specific uses of voice in narratives can be accounted for in terms of the “strangeness” of a given use of voice. The (empirically attested) underlying idea is here that any deviation from a standard is, for that very reason, significant. What is strange, or appears as strange, epitomizes such a “significant deviation.” The notion of strangeness necessarily presupposes the notion of some sort of baseline normality, that is to say some norm or standard X that prescribes a horizon of expectations. It is impossible to talk and think about strangeness without framing it as “strange relative to X”. What is strange is then strange as a deviation from that standard or an exception from the set of expectations entailed by it. Now, two aspects of strangeness are important in this context. I. What is strange is thus also significant qua deviation. This is a cognitive, experiential empirical fact; humans are, among other things, probabilistic machines: we count on the expected and navigate in our Umwelt relying on regularities. Inversely, unexpected, improbable, non-standard phenomena are for the very same reason salient, i. e., phenomenologically significant. The non-generic is a natural source of meaningfulness (cf. Petitot 2004, 2009; Bundgaard, 2009, 2011); our attention is, for example, immediately attracted toward “singularities”, rare phenomena and unexpected events in our surroundings. The intrinsic, natural meaningfulness of the strange qua deviant is furthermore something that can and pervasively has been exploited in art (unlikely or improbable symmetries in paintings and sculpture, interruptions of natural scalar progressions in music, assonances and alliterations in literary art, etc.). The simple claim here is that experimentations with voice in narrative art can be considered as significant deviation in this sense: they are correlated to a meaning effect, and can thus serve determinate semiotic functions. II. There exist different types of norms, and it is impossible to define a priori criteria for strangeness, which would make it possible to predetermine or predict the character of strange phenomena. There exists of course a prototypical zero-level, external standard for what normal speech is in everyday life. It is relative to that standard that certain manners of narrating (or talking or writing or voicing) may appear as pretty strange (the first person diarist in Peter Weiss’ The Shadow of the Body of the Coach-

86

Peer F. Bundgaard

man is for example very strange relative to this standard). Importantly, however, standards can also be set up within a text, at many different levels; and significant deviations can thus be created relative to these internal or local standards. A significant deviation thus obtains if, e.g., a text has been internally focalized in one protagonist the whole way through, only to change locus of focalization at the very end (or at a few chosen moments during the narration). With respect to voice, we may then have deviations, or loci of meaningful strangeness, as it were, wherever the narratorial situation changes within one and the same text (changes from first to third person narration, changes from authorial teller-like narrating voices to figural reflector-like voices, or simply changes in style, etc.). In short, sources of possible strangeness abound; what counts is the way they are exploited with the purpose of triggering meaning effects. This is what I will try to illustrate in the following. Voices Don’t Voice in the Void Voice should of course not be considered only relative to its enunciation, as it were, i.e., its manner of diction. Whatever the nature, style and grain of the talking in a narrative may be, it is never assessed in its own terms and in its own right. It voices relative to something: namely, relative to the story world and the protagonists’ experiences of their situation in that world to which it gives access. Voice, considered under this aspect, is thus not a simply a question of manner of talking (say, in a Portnoy-way, a Holden Caulfield-way or a Jake Barnes-way)—it also, and probably primarily, concerns the position of the voicing instance with respect to the experienced world it represents. This means that voice in narratives may be appraised, and significant deviations may obtain, along two axes: 1. The what-axis: what it gives access to, to what extent it has and gives access to the story world, and how it is positioned with respect to it.2 2

It may seem so, but I am actually not mixing up Genettian mood and voice here (although I of do course consider them as tightly correlated when considering voice as essentially bound up with the issue of (the kind of) access it gives to the story world). When I talk about the “position” of the voicing instance with respect to the world of experience and its “access” to it, I do not do so in terms of “focalization” (external, internal, zero) nor in terms of “homo-“ or “heterodiegesis”. I use this term in one much more literal way: where is the voicing instance (or the narrator) actually situated, in time and space, with respect to what is related, does this position have any semiotic import on narration, does this position change in the course of narration, and does this have any semiotic consequences? I’ll give examples that illustrate these cases in a short while.

Significant Deviations

87

2. The how-axis: how it gives access to it, and whether or not the style of voicing undergoes changes in the course of narration. The norms and standards I mentioned above are distributed along these axes and it is thus on this background that voice can be considered more or less natural, more or less unnatural, or more or less strange. In the following, I shall primarily examine aspects of significant uses of voice along the what-axis, but before doing so, I’d like to give an example of a conspicuous and significant transformation of voice along the how-axis. Strangeness along the How-Axis Even if Hemingway is usually recruited as the king of external focalization, he ought to be considered something like a registered trademark for the figural (Stanzel 1984) or internally focalized (Genette 1980, 1988) mode of narration (pace “The Killers” and “Hills Like White Elephants”); in cases like Hemingway’s, and as far as voice is concerned, this implies that—however recognizable the author’s style is, and no matter how many reader-oriented information the author (the narrator or the narrating instance) distributes—telling becomes spatiotemporally coextensive with the character’s actual experience. It would, indeed, be imprecise to qualify such a style of telling as simply showing, since what is displayed is not merely the scenery experienced by the character, but also the character’s very seeing, experiencing the scenery. As Hamburger (1957) remarkably noticed, this sort of discourse is the only third person type of speech, which directly, immediately gives access to first person givenness (i. e., to what it is to experience in a first person manner).3 Here’s, then, an epito3

This indeed qualifies this kind of narrative voice—that is to say, the most naturally sounding of all voices, the one that seamlessly places the reader directly in the character’s shoes—as one of the strangest, i.e. outstanding, of all voices: it is a well-known fact that consciousness has a first person ontology (Husserl 1900-1901), i.e. it feels like something to experience what one experiences, and this “feeling-like” character cannot be directly accessed from the outside by another person (but of course pretty faithfully represented or understood indirectly); so much the less can it be directly represented or conveyed in speech—except, of course, in this specific kind of fictional telling. Wallace Chafe (1994, 255) makes the same point in his admirable comments on Hemingway’s “Big Two-Hearted Mountain” when he notices that, in a context of internal focalization, descriptive sentences (such as “Nick sat down”, “Nick leaned back”, etc.) also render the protagonist’s immediate experience of performing these actions. From this follows, however, that this property of third person fiction (a voice which gives access to first person consciousness) is not a logical property of discourse, derivable from the components of a sentence (say, it’s “epic preterite”), but a semiotic property of the context in which such descriptive sentences appear, i.e. derivable from au-

88

Peer F. Bundgaard

me of this: the opening lines of For Whom the Bell Tolls (cf. Bundgaard, 2010; see also Jahn, 1997, 445 seq.): He lay flat on the brown, pine-needled floor of the forest, his chin on his folded arms, and high overhead the wind blew in the tops of the pine trees. The mountainside sloped gently where he lay; but below it was steep and he could see the dark of the oiled road winding through the pass. There was a stream alongside the road and far down he saw a mill beside the stream and the falling water of the dam, white in the summer sunlight. “Is that the mill?” he asked “Yes.”

Now, it is puzzling that this trademark voice-mood (or telling-perspective) correlation is submitted to quite some variation only a few pages further on where Robert Jordan witnesses a heated discussion between two Castillan Republicans: The old man turned toward him suddenly and spoke rapidly and furiously in a dialect that Robert Jordan could just follow. It was like reading Quevedo. Anselmo was speaking old Castilian and it went something like this, “Art thou a brute? Yes. Art thou a beast? Yes, many times. Hast thou a brain? Nay. None. Now we come for something of consummate importance and thee, with thy dwelling place to be undisturbed, puts thy fox-hole before the interests of humanity […] I this and that in the this and that of thy father. I this and that and that in thy this. Pick up that bag.” (15)

Obviously, the passage conveys what it feels like, or what it may have felt like, for Robert Jordan to hear a dialect containing such archaic pronominal and verbal forms. Yet, even if a sentence “It was like reading Quevedo” still could be part of Jordan’s immediate (and thus pretty wellschooled) consciousness, the following sentence “and it went something like this” is clearly not part of his ongoing experiences, but a statement made by a displaced teller, constructing a pseudo-Shakespearian English analogue in order to convey to the reader an idea about what kind of talk Robert Jordan was listening to: In short, the origin of the deictic “this” (in “it went something like this”) is evidently not Jordan’s here and now, but the teller narrator’s here and now.4 And the “nays”, the “thous”, the “arts” are not part of Jordan’s consciousness, but words used by the narrator to evoke the right representation in the reader. It is by no means easy to define the meaning effect of such a variation of voice along the how-axis. Yet, considering the fact that it wouldn’t have thor’s meaning intentions: you cannot tell from “Nick sat down”, on purely linguistic grounds, whether it is externally or internally focalized. 4 So much more than Jordan actually seems to master the local tongue as is clear from the following passage: “Robert Jordan drank half the cup of wine but the thickness still came in his throat when he spoke to the girl. ‘How art thou called?’ he asked” (Hemingway, ibid., 26).

Significant Deviations

89

been too complicated for Hemingway to account for the existence of a peculiar dialect while keeping internal focalization, the patent shift in voice probably does reframe the story, and thus reading, in the sense that it recalls the told or mediated character of the narration, hereby suggesting that while narration is generally focalized in Robert Jordan from the very beginning to the very last line, the narrative focus exceeds Jordan as an individual: focus is on the civil war at large, the scene of a tragedy, Shakespearean or not. After this short example of strange, conspicuous and therefore significant deviation in voice along the how-axis, I now turn to meaning effects linked to the relation between the telling instance and what it has and gives access to. Strangeness along the What-Axis There is a, trivial, sense in which all voice or telling in narrative art (with due exceptions) is strange or specious. In contrast to what is the case in everyday communication and oral telling there seems to be a constitutive gap between what a teller gives access to and what he has access to. Whenever a narrator, even or first and foremost a first person narrator, sets out to tell what actually happened, what he actually experienced, and not simply his present recollection of his past experiences, then we have a strange or specious voice to the extent it possesses (or pretends to possess) what Chafe (1994) called “unconstrained remembering”. Consider the opening of Heart of Darkness: The Nellie, a cruising yawl, swung to her anchor without a flutter of the sails, and was at rest. The flood had made, the wind was nearly calm, and being bound down the river, the only thing for it was to come to and wait for the turn of the tide. The sea-reach of the Thames stretched before us like the beginning of an interminable waterway. In the offing the sea and the sky were welded together without a joint, and in the luminous space the tanned sails of the barges drifting up with the tide seemed to stand still in red clusters of canvas sharply peaked with gleams of varnished sprits.

Now, there is an obvious gap between what the representing self (voice) can have access to through memory and the detailed, rich experiences of the represented self it gives access to; i.e., to continue with Chafe’s terminology, there is a conflicting mix, pervasive in narrative art, between the displacement of the representing, narrating “I”, and the immediacy and fullfledged character of its past experiences conveyed by descriptions such as “the tanned sails of the barges […] seemed to stand still in red clusters of canvas sharply peaked with gleams of varnished sprits”. As suggested, this

90

Peer F. Bundgaard

situation obtains for any piece of narrative which does not aim to convey a character’s or a first person narrator’s painstaking reconstruction of past experiences. It is a strangeness of voice inherent in telling. But as all conditions it can be circumvented in different ways with the aim of obtaining given meaning effects. In the present case, the immediacy of the experiencing represented self seems deliberately enhanced on behalf of the displacement of the armchairing representing self, thus endowing the scenery with a flavor of absolute presence (and the past tense with a scent of epic preterite): in short, the outcome and showdown of the story is by no means part of the reader’s temporal horizon. If the above is a correct reading, it naturally follows that, according to the way voice (or the representing, narrating self) is positioned relative to the experiencing or represented self, first person narration can be of a more or less authorial or more or less figural sort, or its narrator can be more or less teller- or reflector-like (as Stanzel 1984, 1990 claims). If this is so, Stanzel seems right in championing the continuous relation between third person and first person narration in the vicinity of the reflectormode pole. Cohn’s “roadblock” between these types of narration (Cohn 1981) could and should be removed whenever the temporal horizon of the story is not pre-established, or in a nut-shell: when the end of the story (as it is supposed to be known by a first person narrator telling in the past tense) is not actually a part of its temporal horizon and thus of the reader’s experience of the gestalted world. Topological Strangeness (Relative to Mode of Narration) I shall now go into a little more detail with this point. Let’s return to the axes according to which the naturalness or the strangeness of voice are likely to be accessed, and more precisely to the what-axis. I here follow Stanzel’s (1984) claim to the effect that there exists a whole subset of narratives in which the author, in order to trigger certain meaning effects rigorously correlated to the mode of narration, operates a deviation from the initial narratorial situation. This transformation can be instantiated and thus qualified in different ways, for example relative to voice itself: the transformation then concerns the phenomenological character of the telling instance which may go from being a clearly incarnated narrating instance (conveying to the reader, from the distance, post festum, the highlights of some past chain of events), to being a pure experiencing instance without any other temporal or epistemological horizon than the one crystallized in the narrative here-and-now; or it may concern the place from which the telling is done, which to begin with may be well-defined (retrospective

Significant Deviations

91

armchair teller) and then, by way of reduction to pure experiencing (in a reflector mode), may become spatially undecidable, as it were, a-spatialized, non-localizable. I’ll call this transformation from the authorial first person narrator toward the purely experiencing voice of a derelict (first person) character an instance of topological strangeness: it concerns the relation of the narrating instance to the events it, at a first stage, gives access to, and, progressively or ultimately, are caught up by. Here are a couple of examples: [A1] Robert Cohn was once middle-weight boxing champion of Princeton. Do not think that I am very much impressed by that as a boxing title, but it meant a lot to Cohn. He cared nothing for boxing, in fact, he disliked it, but he learned it painfully and thoroughly to counteract the feeling of inferiority and shyness he had felt on being treated as a Jew at Princeton. (Hemingway, And the Sun Also Rises, opening lines.) [A2] We sat close against each other. I put my arm round her and she rested against me comfortably. It was very hot and bright and the houses looked sharply white. We turned out on the Gran Vía. ‘Oh, Jake,’ Brett said, ‘we could have had such a damned good time together.’ Ahead was a mounted policemen in khaki directing traffic. He raised his baton. The car slowed suddenly, pressing Brett against me. ‘Yes,’ I said. ‘Isn’t it pretty to think so.’ (The Sun Also Rises, closing lines.) [B1] TRUE! —nervous—very, very dreadfully nervous I had been and am; but why will you say that I am mad? The disease had sharpened my senses—not destroyed—not dulled them. Above all was the sense of hearing acute. I heard all things in the heaven and in the earth. I heard many things in hell. How, then, am I mad? Hearken! and observe how healthily—how calmly I can tell you the whole story. (Poe, The Tell-Tale Heart, opening lines.) [B2] Oh God! what could I do? I foamed—I raved—I swore! I swung the chair upon which I had been sitting, and grated it upon the boards, but the noise arose over all and continually increased. It grew louder—louder—louder! And still the men chatted pleasantly, and smiled. Was it possible they heard not? Almighty God! —no, no! They heard! —they suspected! —they knew! —they were making a mockery of my horror! —this I thought, and this I think. But anything was better than this agony! Anything was more tolerable than this derision! I could bear those hypocritical smiles no longer! I felt that I must scream or die! and now again! —hark! louder! louder! louder! louder! “Villains!” I shrieked, “dissemble no more! I admit the deed! —tear up the planks! here, here! —It is the beating of his hideous heart!” (The Tell-Tale Heart, closing lines)

Both excerpts clearly illustrate the above-mentioned change. In both cases the initial set-up involves a spatiotemporally distanced, thus knowing teller narrator, and in both cases, we strand in the all-consuming present (even the grammatical present) of a reflectorial no-man’s-land, without any de-

92

Peer F. Bundgaard

terminate temporal horizon, abandoned by whatever epistemic instance who might know the course of things from the distance.5 Moreover, and crucially, in both cases, this modulation of the teller voice, this induction of topological strangeness with its spatial displacement of voice toward an undecidable location, is a meaning carrying formal counterpart to the process of dereliction undergone by the protagonists (each in his way, each for his own reasons).6 Topological strangeness due to the unclear position of the narrator relative to his own story is an artful means of meaning making in the sense that it produces a strong correlation between story world and mode of presentation. Now, it is important to notice that the artfulness of the device, and the semiotic efficiency of the correlation between mode of presentation and represented world, is not necessarily experienced as “artful” by the reader. Strangeness is not necessarily tantamount to estrangement. In cases like the above probably even far from that. Arguably, the meaning effect is simply one of intensification by way of re-focalization: the teller narrator’s initial words belong to a report, an account, whereas the subsequent words and descriptions are fully integrated moments of an on-line experience in the reflector-mode. Such intensified internal re-focalization is a common (and efficient!) means of meaning making in literature. I have commented on it in some more detail in Bundgaard, 2010, on the grounds of Chafe’s enlightening example of the same phenomenon (Chafe 1994, p. 257).7 Below is an example of this from Roth’s American Pastoral. The novel is in itself a powerful illustration of the narratorial changes discussed above: it sets out as a 5

This is of course so much more alarmingly the case in Hemingway, since, initially, the narrator pretends to tell, not his own story, but Robert Cohn’s story. The distance covered by the transition is, consequently, quite considerable: from a quasi heterodiegetic teller narrator, telling someone else’s past story, to a pure reflector confined in his presents experiences. 6 Henrik Skov Nielsen, in “Fictional Voices? Strange Voices? Unnatural Voices?” (in the present issue), also observes this convergence of what he calls the “now of the telling” (and its “studied calm”) and the “then of the told”, or rather the absorption of the former in the latter. 7 The following passage from Hemingway’s “Big Two-Hearted River” describes the culminating point of the narration (Nick’s landing a big trout). Chafe acutely notices that the climax at the representational level is formally correlated to an intensification of focalization from [a] psychonarration, via [b] indirect free thought to [c] quasi direct thought (a progression which is unprecedented in the story): “He thought of the trout somewhere on the bottom, holding himself steady over the gravel, far down below the light, under the logs, with the hook in his jaw, Nick knew the trout’s teeth would cut through the snell of the hook. The hook would imbed itself in his jaw. [b] He’d bet the trout was angry. Anything that size would be angry. That was a trout. He had been solidly hooked. Solid as a rock. He felt like a rock, too, before he started off. By God, he was a big one. [c] By God, he was the biggest one I ever heard of (Hemingway, “Big-Two-Hearted River”; Chafe 1994, 257; my letters in bold).

Significant Deviations

93

straight Zuckerman first person narration (relating Zuckerman’s memories of the High School icon Seymour “The Swede” Levov) and drifts toward an internally focalized third person narration with no Zuckerman, but a lot of Levov. In the text passage, the Swede even overtakes the first person narrator modus—at a particularly critical moment in his life and of the narration—and we thus have the case just mentioned: a re-focalization, and a displacement of voice, which are the formal counterpart of the existential and mental state of the protagonist. Below, “The Swede” is here initially externally focalized, introducing a young woman (which reminds him of his bomb-killing and escaped daughter) to the different sections of his hand glove factory; progressively, focalization is remodalized as internal, ending up with full-blown direct thought: This is the silking, that's a story in itself, but this is what she's going to do first . . . This is called a pique machine, it sews the finest stitch, called pique, requires far more skill than the other stitches . . . This is called a polishing machine and that is called a stretcher and you are called honey and I am called Daddy and this is called living and the other is called dying and this is called madness and this is called mourning and this is called hell, pure hell, and you have to have strong ties to be able to stick it out, this is called trying-to-go-on-as-though-nothing-hashappened and this is called paying-the-full-price-but-in-God's-name-for-what, this is called wanting-to-bedead-and-wanting-to-end-her-and-to-kill-her-and-tosave-her-from-whatever-she-is-going-through-wherever-on-earth-she-may-be-atthis-moment, this unbridled outpouring is called blotting-out-everything and it does not work, I am half insane, the shattering force of that bomb is too great . . . And then they were back at his office again, waiting for Rita's gloves to come from the finishing department, and he was repeating to her a favorite observation of his father's, one that his father had read somewhere and always used to impress visitors, and he heard himself repeating it, word for word, as his own. If only he could get her to stay and not go, if he could keep on talking about gloves to her, about gloves, about skins, about his horrible riddle, implore her, beg her, Don't leave me alone with this horrible riddle... (Roth, American Pastoral, Part II, chapter 4).

Although I mainly mention this as a case of intensified internal refocalization, it of course also comports surface voice effects: the “I” is indeed remodalized; for the first time in the story it does not refer to Zuckerman. This, in turn, has deeper voice effects relative to topological strangeness: who is in charge of telling? Obviously not Zuckerman, who can have no access to Levov’s inner life. Again: the artfulness of devices like the above is probably formally evident, and for the more or less trained reader rather conspicuous, but they need not at all have estranging consequences for the reader. On the contrary, such formally unnatural transformations of voice are natural correlates to the increasing intensity of the represented experiences or conditions of experience. To that extent they may indeed have quite natural or

94

Peer F. Bundgaard

“enhanced empathizing” meaning effects for readers in general (including the trained ones).8 Topological Strangeness II (Relative to Location of Voice) There exist also types of strangeness along the what-axis – i.e., the axis which concerns the narrator’s access to what is told – which are inescapably strange (or at least are so to some degree). The strangeness in such cases does not necessarily or primarily pertain to any particular “grain” of voice, or to any more or less manifest redefinition of narrator mode as the ones just mentioned. On the contrary, voice may be normal, and the mode of narration stable and constant. The source of strangeness is rather the position itself, its location and the nature of its access to the story world. Thus, in Paul Auster’s Travels In the Scriptorium, it is pedantically and repeatedly made clear to the reader that whatever is reported about an old man’s doings in an apartment is accessed through a camera (and recorded by a microphone): The old man sits on the edge of the narrow bed, palms spread out on his knees, head down, staring at the floor. He has no idea that a camera is planted in the ceiling directly above him. The shutter clicks silently once every second, producing eighty-six thousand four hundred still photos with each revolution of the earth. […] Who is he? What is he doing here? When did he arrive and how long will he remain? With any luck, time will tell us all. For the moment our only task is to study the pictures as attentively as we can and refrain from drawing any premature conclusions. (Travels In the Scriptorium, opening lines)

8

First person narrations (say, some of Raymond Carver’s short stories or Paul Auster’s Man in the Dark) in the present tense are also good examples of formally strange, but experientially rather unproblematic narration: here, the narratorial situation, the localization of the voicing instance, remains evidently undecidable through the whole narrative; you can’t tell the situation while you are living it. Now, do readers necessarily make sense of the text relative to this impossibility; is it something the reader necessarily “hears” or “senses” and which constrains his meaning making; or is it something, which, albeit formally strange, may have a natural meaning effect as the one discussed above: it intensifies the presence of the presentified world. I gather that the conclusion which should be drawn from this is that there is no strangeness in and per se—that is, there exists no “strange” or “deviant” form of presentation (of telling) which in virtue of this formal strangeness necessarily shapes a strange meaning: formal strangeness may very well be correlated to natural meaning effects. This, of course, does not at all imply that formal strangeness—or example of the topological sort— cannot be permeated with semantic strangeness, that is, be continuously felt as strange, very strange.

Significant Deviations

95

Arguably, this counts as vintage external focalization, yet only a couple of paragraphs further on does the miracle or the art of storytelling take over. Though restricted to the pictures and sounds delivered by the recording devices, the reporting voice, or rather the first person narrator, has no difficulty whatsoever in accessing the old man’s, Mr. Blank’s sensations and thoughts related to inner mental states: He is dressed in blue-and-yellow striped cotton pajamas, and his feet are encased in a pair of black leather slippers. It is unclear to him exactly where he is. In the room, yes, but in what building is the room located? In a house? In a hospital? In a prison? He can’t remember how long he has been here or the nature of the circumstances that precipitated his removal to this place.” (In the Scriptorium, fourth paragraph).

The result is a highly artful and consistently maintained duality between two parallelly running focalizations: the external for all those things which pertain to visual or auditory sense data, and the internal through which the inner, psychological manifestations of his mental states are manifested (the mens data, as it were). All this supported by one and the same reporting master mind voice, which, even though equipped with a machine not only capable of grasping the manifestation of physical life, but also the salient phenomena of inner life (sensations, desires, thoughts, interrogations, etc.), has no further access to any depths of any soul. In short: the narrator seems incapable of integrating or blending these parallel tracks of description into a consistent Gestalt. It is a remarkable fact that Travels in the Scriptorium is virtually devoid of psychonarration. Narration in the mode of inner focalization is here rigorously analogous to narration in the mode of external focalization: it captures the signs of inner life, without gestalting any full-fledged character, just as the external focalization simply captures the adumbrations or aspects of Mr. Blank’s physical appearance. The duality is extreme in the sense that narration in the mode of inner focalization does not complement the observational facts conveyed in external focalization, it does not “deepen” the character. Rather, it reports the concrete facts of a series of mental states without conveying any global representation of the character’s psychology or inner life. To this extent, Mr. Blank is blank twice rather than once: his physical make-up is limited to the pictures captured by the camera, his inner life and “character” are so many mens data captured by the thought recorder. As becomes clear at the end of the story, this double, external and internal, intrusion in Mr. Blank, and the ensuing double indeterminacy are perhaps the most subtle punishment fictional characters plotting for payback—as is the case here—can submit their author to.

96

Peer F. Bundgaard

As already suggested, here the deviant character of voice is not due to any change of style or any change of (say, authorial vs. figural) position within one and the same narrative, but rather to the very position of the narrating instance (and the conditions under which its telling takes place) relative to what it gives access to. So in the Auster case, we have a voice restricted to recording machines which has nonetheless access to the psychologically real signs of a person’s inner life, yet without, and this is all as strange, having complete access to the person’s character as such. Now, although almost too much of a parade example, Travels in the Scriptorium may be a token of a general subtype of topological strangeness along the what-axis. It could be considered as pertaining to the class of narratives with only partially knowing narrators or epistemically limited narrating instances. Such narrative instances—pace Sternberg (1978)—do not withhold any information (for the sake of suspense or whatever tricky reasons), they do have access to their characters’ inner life as well as to all relevant facts and events in the story world, but only, and critically so, within the limits of their characters’ own epistemic horizon. They know everything the characters know, each of their intuitions, fears, expectations, memories, etc. They even know the full apparent plot, its temporal order, its beginning and end, everything that possibly may be recorded by any external observer, but they don’t know anything which exceeds the epistemic universe of the characters. A case in point here is Ian McEwan’s On Chesil Beach and its minute, painfully thorough description of a marriage never consummated, the history of the young just married and the aftermath of their disastrous wedding night. Everything is there, all the bits and pieces are conveyed to the reader: the groom’s frustration, the bride’s disgust by even the faintest thought of bodily contact, the past circumstances, their thoughts, etc.– everything but the most essential: the reason why the bride feels the profound repulsion of any sexual intimacy. In other words: the whole cause of this local tragedy, reported by an overtly telling narrating instance, is unaccessible to the latter. Now, what kind of strangeness is this? Well, statistically speaking it is obviously an empirical strangeness: conspicuous, constantly intervening, extradiegetic, and omniscient narrators usually know the whole business all the way through: not only the geographical, temporal, historical, and psychological surface of things (place, socio-cultural conditions, chronology, past and present thoughts, emotions and sensations), but the real driving force of the plot. However, this strangeness would be specious if the knowledge was merely withheld (it would be a piece of Genettian paralipsis.) Another real possibility is that the strangeness is topological: the form of narration, or the epistemic situation of the narrator, is a perfect

Significant Deviations

97

counterpart, and thus a semiotic correlate, to the epistemic situation of the protagonists–he knows what they know and no more. It would be a semiotic correlate in the sense that this form of telling, this truncated access to the story world, has crucial meaning effects on the reader, who, as a result, must reenact every scene, every gesture, everything which surfaces inside and outside the protagonists exactly like the protagonists themselves do it: that is, recording, taking note of every event, however fatal, without knowing why it occurs in the first place and what the outcome will be.9 Concluding Remarks In the preceding sections I have intended to shed light on a meaning making device in narrative art which is a function of the position of the narrator or the narrating instance relative to what he or it has or gives access to. Since the semiotic effects of this relation to the story world are contingent upon determinate deviations from some given standard, I have proposed the expression topological strangeness (of voice) to capture this phenomenon. It has been shown that topological strangeness comes in different types depending on the kind of significant deviations obtained. One such type was illustrated with examples from Hemingway, Poe and Roth: it concerns “within” text transformations of voice relative to shifts in the narratorial situation (from armchair “teller” to “reflector” first person narrator in Poe and Hemingway, or as in Roth from first person with internal perspective (Zuckerman on Zuckerman’s relation to Levov), via first person with external perspective (Zuckerman on Levov) to third person figural narration (Levov tout court). In all these cases, specific meaning effects were identified. Another type of topological strangeness concerns the “external” relation between the narrating instance and the story world. It was illustrated with more or less paradox examples from Auster and McEwan where we encounter narrating instances which have access to the story world only through recording devices, but which nevertheless can capture the surface phenomena of the protagonist’s inner life (Auster), or which are omnisci9

As Frans Gregersen (2010), probably among others, has noticed, the story offers a hint at an explanation of Florence’s revulsion: a possible incestuous relation with her father. When at the age of twelve, she was sailing alone with him to France: “waiting, shivering in the narrow bunk ... her mind was a blank, she felt she was in disgrace”. And while her father is said to be undressing, “her only task was to keep her eyes closed and to think of a tune she liked” (which makes a fine link to her future career as a musician). Crucially, however, this suggestion only exists as a reminiscence in Florence, one of the surface forms of her internal life, yet another mens datum, which is only accessible as such for the narrator, with no further depth, nor mystery.

98

Peer F. Bundgaard

ent, but don’t possess any knowledge which exceeds the sum of the protagonists’ insights (McEwan). Again, this deviant use of voice was shown to hook up with given meaning effects. Aesthetic objects (whether musical, pictorial, plastic or literary) are expressive, intentional objects: they manifest aesthetic meaning intentions. Meaning intentions can be shaped or meaning effects can be triggered in different ways: as regards the content level, they can be so relative to the subject matters or the motifs represented by the object—we can call these the representational meaning-making devices (or Vorstellungs-devices); formally, they can be so relative to the means each aesthetic genre disposes of to present its stuff (means, which characterizes it as a genre)—we can call these the presentational meaning-making devices (or Darstellungs-devices). Thus “perspective” in art can be used in plentiful (and deviant) ways in order to obtain given meaning effects (Bundgaard 2002, 2004, 2009, 2011), the same goes for shape, color, stroke and composition. The claim on which the present work is founded is then, simply, one to the effect that this also holds true for narrative art: any given author disposes of presentational devices such as (distribution of) focalization, person, plot structure (temporal and causal structure), speed (or granularity/density), and … voice. Each of these can be activated in any given way, strange or straight, or be submitted to changes within one and the same narrative and thus fulfill various semiotic functions. One of the major tasks of narratology, as it is envisaged here, is, of course, to establish a list of formal means of meaning-making in literary art, at large, but also to track down and suitably characterize the concrete ways authors use them (or don’t use them) in order to give literary shape to what they have in mind. If we do so, my claim would be that a crossover higher order category of meaningmaking devices will be laid bare: the use of deviations from a given standard which for that very reason appears as significant. Such significant deviations obtain within all narratological categories (as well as all types of aesthetic meaning-making devices in general), and as we have seen also relative to voice. The strangeness of voices might thus benefit from being reframed as significant deviations in the order of telling. These deviations are then the phenomena that should be submitted to scrutiny and systematic characterization. But this is never an easy task.

Significant Deviations

99

References Auster, Paul 2007 Travels in the Scriptorium. 2008 Man in the Dark. Bundgaard, Peer F. 2002 “Presentation and representation in art – Ontic and gestaltic constraints on aesthetic experience”, Vision 7.1-2, 187-204. 2004 Kunst-semiotiske beskrivelser af æstetisk betydning og oplevelse, Copenhagen: Haase & Søn. 2009 “Toward a cognitive semiotics of the visual artwork – Elements of a grammar of aesthetic intuition”, Cognitive Semiotics 5, 42-65. 2010 “Means of Meaning Making in Literary Art–focalization, mode of narration, and granularity”, Acta Linguistica Hafniensia 42.1, 64-84. 2011 “The Grammar of Aesthetic Intuition – On Ernst Cassirer’s concept of symbolic form in the visual arts”, Synthese 179.1 43-57 (doi:10.1007/s11229-009-9631-8). Chafe, Wallace 1994 Discourse, Consciousness, and Time, Chicago: The University of Chicago Press. Cohn, Dorit 1981 “The Encirclement of Narrative: On Franz Stanzel's Theorie des Erzählens”, Poetics Today 2.2, 157-182. Conrad, Joseph 1902 Heart of Darkness. Genette, Gérard 1980 Narrative Discourse: An Essay in Method, Ithaca, N.Y.: Cornell University Press. 1988 Narrative Discourse Revisited, Ithaca, N.Y.: Cornell University Press. Gregersen, Frans 2010 “How nice to be good at it. The aesthetics of everyday language and the optimization of it in literature”, Acta Linguistica Hafniensia 42.1, 159-176. Hamburger, Käte 1994 [1957] Die Logik der Dichtung, Stuttgart: Klett-Cotta. (Engl. tr. The Logic of Literature, Bloomington and Indianapolis: Indiana University Press, 1973.) Hemingway, Ernest 1926 The Sun Also Rises. 1975 [1940] For Whom the Bell Tolls, Hammonsworth: Penguin Books. Husserl, Edmund 1900-1901 Logische Untersuchungen, Tübingen: Max Niemayer Verlag. Jahn, Manfred 1997 “Frames, inferences, and the reading of third-person narratives: towards a cognitive narratology”, Poetics Today, 18.4, 441-468. Livingston, Paisley 2005 “Narrative”, in: Berys Gaut and Dominic McIver Lopes, The Routledge Companion to Aesthetics, 2nd Edition, 359-370. McEwan, Ian 2007 On Chesil Beach. Petitot, Jean 2004 Morphologie et esthétique. Paris: Maisonneuve et Larose. 2009 Non-generic viewpoints as a method of composition in Renaissance paintings. Cognitive Semiotics 5, 7-41. Poe, Edgar Allan 1843 The Tell-Tale Heart. Roth, Philip 1969 Portnoy’s Complaint. 1997 American Pastoral.

100

Peer F. Bundgaard

Searle, John 1975 New Literary History 6.2, 319-332. Stanzel, Franz K. 1984 A Theory of Narrative, Cambridge: Cambridge University Press. 1990 “A Low-Structuralist at Bay? Further Thoughts on A Theory of Narrative”, Poetics Today 11.4, 805-816. Sternberg, Meir 1978 Expositional Modes and Temporal Ordering in Fiction, Baltimore and London: The Johns Hopkins University Press.



LARS-ÅKE SKALIN (Örebro)

How Strange Are the “Strange Voices” of Fiction? The phenomenon of “strange voices” in literary fiction has received some attention from narrative theorists lately, as we shall see below. What could be regarded as “strange” are voices that diverge from the norms expected to guide narration in general, which is, of course, the reason why the matter is of interest to narrative theory. However, narrative theory is not a homogeneous field of study with a non-disputed set of core concepts; quite the contrary, every concept that has been claimed to be fundamental to theoretical analysis has been disputed. In my discussion, I take as point of departure a strategy preferred by at least some theorists: I regard the act-aspect of narrative as primary to the aspect of formal structure. This aspect is the basis, for example, of the traditional view on fictional content that it must be mediated by some source (cf. Franz K. Stanzel’s central term “mediacy”1), that is, it must be mediated by an act of communication. Gérard Genette also observes that the most fundamental level of narrative should reasonably be its producing act, the narration, although narrative theory does not seem to have paid much regard to this.2 However, definitions of what narrative is as seen from this perspective have been proposed, for instance by Barbara Herrnstein Smith who found that narrative discourse is an act that consists of “someone telling someone else that something happened”.3 A slightly extended version is James Phelan’s rhetorical definition of narrative discourse as: “the act of somebody telling somebody else on a particular occasion for some purpose that something happened”.4 These definitions try to state something assumed to be fundamental to all those acts we are willing to identify as “narrating”. They have two parts: one says what one is doing when narrating, the other what the narration is about. The central defining phrase for the act, “telling … that”, suggests that it should be understood as a referential act, 1

Stanzel 1984, A Theory of Narrative. Genette 1980: 26. Smith 1981: 228. 4 Phelan (2005: 217). 2 3

102

Lars-Åke Skalin

motivated mainly by an informative purpose, where a teller mediates knowledge about events and circumstances to a receiver. As a result, listeners to, or readers of, narratives are supposed to regard themselves as informed about true facts and circumstances. These definitions do not say anything about fiction. However, a very common opinion among theorists has been that the formula is valid also with regard to fictional narrative but here the receiver has to deal with the text as presented in a certain key, or as introduced by an operator: as if informing about a world in which the reader is positioned. I will call this model the “internal” approach5 and later on contrast it with an “external” approach. The key, as if, should make narrative theory economical: all narrative readings could be regarded as observing the same basic communicational schema. In his often cited essay “The Logical Status of Fictional Discourse”, John Searle affirms: “The utterance acts in fiction are indistinguishable from the utterance acts of serious discourse, and it is for that reason that there is no textual property that will identify a stretch of discourse as a work of fiction”.6 Perhaps Searle is right in denying that there are textual properties that will tell us in whether to take a presented discourse as a work of fiction or non-fiction. Actually, he sets textual form in relation to the utterance acts in fiction. This, I would say, puts the question in a different light because even if, as Searle assumes, the author of fiction is just “pretending” to perform speech-acts corresponding to the forms of the sentences of the literary text (most of them having the form of declarative sentences), there must be an act behind his “pretending” that communicates what the reader is actually making sense of, this real act still using the very sentences presented by the text. The “pretending” is thus assumed to be the vehicle, the form, by which this serious act is performed and as such the real object for a theory of literature. If the reader is not able to go behind the act suggested by the linguistic form only (such as the affirmative sentences) he or she will possibly only find “strange voices” unable to communicate a comprehensible message. Theorists have noticed, however, that in the matter of fiction there occasionally appear elements which seem to violate the expected communicational pattern: some logical absurdity, or a transgression of assumed ontological levels, or deviation from what is taken as the running code (metalepsis, and also paralipsis and paralepsis), or plainly totally unrealistic scenarios. Some theorists have expressed the opinion that we should take such observations seriously and ask what possible consequences for the 5

Fiction reading implies a “recentering” into a possible world. Being in it the interpreter makes sense of it according to the “principle of minimal departure”. See Ryan 1991: 48-60, and also Walton 1990: 144-161 on the “reality principle”. 6 Searle 1974/75: 327.

How Strange Are the “Strange Voices” of Fiction?

103

overall narrative theory such infringements might have. This is done for instance by Dorrit Cohn in the essays collected in The Distinction of Fiction,7 and even more extensively by Monika Fludernik in the concluding chapters of Towards a ‘Natural’ Narratology and in several later essays.8 Brian Richardson, who in his Unnatural Voices: Extreme Narration in Modern and Contemporary Fiction9 presents a systematic, book-length survey and analysis of the problem of narrative “strangeness”, ends with the following methodological reflections on narrative theory: While it may be true that it is necessary to start doing theory with the established practices of nonfictional, natural, and conventional narration, this is only the beginning. One must be prepared to go far beyond if one is to have any hope of effectively describing the practice of Conrad, Joyce, Woolf, Faulkner, and Nabokov, to say nothing of Beckett and the more radical innovators associated with the nouveau roman, magic realism, and postmodernism. Narrative fiction is dynamic, mutable, subversive, and, above all, dialectical in its perpetual reconfiguring of adjacent genres and established practices. In its more innovative forms, its convention is to alter convention, its essence is to elude a fixed essence, and its nature is to seek out the unnatural. (140)

Common to these theorists, however, is that they suggest supplementary improvements of the current narrative theory. Richardson is quite explicit on this issue. Commenting on a declaration made by J. M. Coetzee with regard to his anti-realist literary method, Richardson says: The very language Coetzee uses to describe his divergence from the model of psychological realism suggests that his practice is not merely different, but consciously opposed to this poetics, and indeed that it may not be fully comprehensible except in reference to the mimetic strictures it so palpably violates. This is why I argue in this book not for a different poetics but for an additional one; that is, for an anti-mimetic poetics that supplements existing mimetic theories. (138)10

What standard theory has achieved so far is mostly regarded as good progress: it has answered important questions and offered a battery of useful concepts and technical terms. Its achievements, however, must be completed with new perspectives for the theory to make further progress. It is an undisputable fact that much narrative literature privileges “narrative realism” creating “mimetic illusion” in which it is also successful. But we cannot be blind to the fact that a great deal of literature, especially modern literature, does not comply with this convention but makes use of “anti7 8 9 10

Cohn 1999. Fludernik 1996 (chapters 6 and 7); see also Fludernik 2001; 2003a; 2003b; 2003c. Richardson 2006. Richardson’s use of the term “mimetic” here is common in narratological discussions. It is synonymous with for instance “realistic” or “common sense” and compatible with what I have called the “internal” approach. In my own discussion, I will use “mimetic” in a manner supposed to be in line with Aristotle’s use, which is contrasted with the “internal” view.

104

Lars-Åke Skalin

illusionistic strategies”. It is in this latter category we are expected to find instances of “strange voices”. This seems to be the view of the reflective critics. Consequently, their proposals do not imply a radical re-evaluation of standard narrative paradigm; they seem content to admit that a theory aspiring to completeness would be more complex than the standard variant. Accordingly, to already known, described, and defined narrative building-elements, we should add new ones. My approach to the issue is much more radical. It will not only detect “strange voices” but also attempt to silence them. This attempt starts by asking whether the “strangeness” is created by the art of fiction or by a particular theory on that art. It is argued that if we take standard narrative theory for granted, we will find much more of “strange voices” in literature than is assumed. Even traditional narrative fiction will appear “unnatural”. This thesis has consequences for core concepts of narratology such as “narrative”, “narrator”, and “story”, as well as for the chosen strategies for theorizing on narrative. Taken as a definition of “what all and only”11 narrative acts have in common the formula “someone telling someone else that something happened” might very well turn out to be empty. I doubt that it could lead us to a comprehension of what it is to narrate or to tell a story, because in order to construe the term “telling” in all its syntactical combinations— “someone telling someone else”, “someone telling …that”, “someone telling …that something happened”—we must relate the phrases to their functions in different communicative practices (or, with the Wittgensteinian term, “language-games”, corresponding to certain “forms of life”) where they operate to produce meaning. The formula will work only if a particular practice is implied. In my presentation of the issue, I made the interpretation that Herrnstein Smith as well as Phelan had in mind the practice of ordinary informative telling about what has gone on in our own lives as well as in the lives of others, that is, a paradigm of so-called natural narrative. And the formula may very well function when referring to manifestations of that practice or set of practices. Yet, it could be the case that if the wrong practice is presupposed to be governing the sense-making of a certain type of communicative act, a theoretical attempt to analyze and explicate them would only find “strange voices”. This is what I suggested in my reference to John Searle’s declaration above, and also what I want to demonstrate in my essay with respect to examples of narrative fiction that could not be regarded as marginal cases but as quite traditional ones. From the perspective of what I suggest as a counter-model to the standard theory, the problem of strange voices gets reversed. It is suggest11

Cf. Gerald Prince, “Narratology”, in Prince 1989.

How Strange Are the “Strange Voices” of Fiction?

105

ed that what from the standard narrative paradigm seemed strange, but perhaps marginal, in fiction could be the normal sense-making operation within that practice. I try to demonstrate this thesis by exhibiting a couple of peculiarities of narrative fiction as compared with natural narratives. One very important peculiarity is the logic of the presentational perspective. Most narrative theorists have affirmed that a narrative issues from a narrator in a teller situation. This sounds very reasonable. The telling of natural narratives could be seen as motivated among other things by the aspect of the told: it is something past in relation to a teller in such a situation.12 It constitutes the actual sense-governing locus from which the details of the presented chain of events and circumstances are realized as having relevance for what has been intended by the sender and accepted by the receiver as being the topic of the narrative. The direction of the narrative going out from this position is a backwards direction. It is also from this same position that it is possible to judge whether the telling was really about something which could be recognized as something a narrative can be about, within the practice acknowledged. However, functionally, if not always formally, literary fiction refrains from such motivation, thus seemingly violating the conversational maxim of relevance obligatory for natural narratives. This is because narrative fiction typically displays a direction forwards without a previous backward movement. In natural narratives such backward operation motivates the subsequent forward moving narrative elements as having been chosen from a locus before them, that is, from a position from where the subject of the presentation is stated. The elements of the progressive movement are then supposed to be the answers to questions implied by the subject. Narrative fiction, however, does not typically rely on this kind of communicative operation. Yet, how should one explain the presentation of a progressive direction seemingly without a subject? Well, this is what an Aristotelian poetics tries to do. According to such a theory, the purpose of narrative fiction is to present a mimesis praxeos kai bion, the imitation of action and life, as Aristotle saw manifested in tragedies, comedies, and heroic epic. However, with natural narrative as the norm of sense-making, mimetic narrative (that is, in the Aristotelian sense of mimesis), which moves “blindly” forwards, will appear as “strange”. Iris Murdoch illustrates this peculiarity of narrative fiction in an introductory passage entitled “Bradley Pearson’s Foreword” to her novel The Black Prince. The protagonist of the homodiegetic novel states it as his intention to tell his story according to the 12

This is the point of departure for Käte Hamburger in her construction of a “logic of literature” as deviating from common rules of temporal presentation. See Hamburger, The Logic of Literature (1993: 67).

106

Lars-Åke Skalin

“modern technique of narration”. Accordingly, he will present his younger self as moving forward, “aware of the past, unaware of what is to come”.13 This feature of “blind” progression, which must violate the kind of expectations we have in meeting natural narratives, becomes conspicuous in autodiegetic narratives with internal focalization. Still, it is my ambition to demonstrate that its “strangeness” is a natural as well as a necessary quality in the practice we associate with the notion of narrative fiction generally. This peculiarity also informs novels normally regarded as being told in a traditional manner and independent of grammatical person.14 If this reasoning is valid, it might be summed up as the laying bare of a theoretical conflict. On the one hand, we have an approach to literary narrative which understands “narrative” from the paradigm of natural narrative and therefore deems it is possible to have a theory applicable to “all and only narratives”. On the other hand, we have an approach from an aesthetic point of departure with an Aristotelian paradigm of narrative fiction as mimesis, where mimesis is a practice that differs from the practice of natural narrating. The first approach could be described as that of narratology, the second as that of poetics in a stricter sense than the sense usually intended by narratologists. From the point of view of such a poetics the narratological approach might be accused of having committed a “narrative fallacy” in its dealing with literary fiction.15 Another peculiarity of literary fiction related to the former one has to do with the discourse aspect, that is, of the text aspect, and it will cause trouble for the assumption of the distinction between story and discourse. It has been touched upon by aesthetic analyses that have used the notion of “spatiality”, as does for example Joseph Frank.16 It is the anti-Lessing assumption that the work of literary fiction does not produce its aesthetic effects only by the temporal dimension—literature articulates signs in time 13 14

Murdoch 2003: 3. I have discussed the idea of such a project and its formula in Skalin 2009. Already at this place I will make it clear that this opinion is both in compliance and in conflict with the literary theory of Käte Hamburger as argued in The Logic of Literature. As is well known, Hamburger distinguishes between, on the one hand, novels and short stories in the third-person, which are defined as fictional narration (fiktionales Erzählen) a notion which could be identified with the interpretation I have made here of the Aristotelian mimesis praxeos kai bion, and, on the other hand, novels and short stories in the first-person, defined as fictive speech (fingierte Aussage). In my opinion, stories in the first-person do also display the properties typical of mimesis as contrasted with natural narratives. It is also evident that Hamburger got problems herself with the strict separation and had to make certain modifications of it. 15 This is a theme of a doctorial thesis (written in Swedish) by Ulrika Göransson 2009. She takes her point of departure in a reflection by Michael McKeon in his foreword to The Theory of the Novel. McKeon observes the tendency in the theory of the novel to seek answers to its questions in narrative theory, a trend that he deplores. See McKeon 2000: xiv-xv. 16 Frank 1991, The Idea of Spatial Form.

How Strange Are the “Strange Voices” of Fiction?

107

while painting uses forms and colour in space. Instead, there is also the possibility to “unroll” the discourse in space and see it as a canvas with a painting, as it were, where elements not directly associated with the represented action of the work are operating, producing meaning within the work as a whole. The effects of this aspect might have something in common with what Roman Jakobson called “the principle of equivalence”.17 It refers to the ascription of parallel significance to logically diverse elements, although, according to common sense, such an operation would be impossible. My purpose is now to demonstrate these theses by reference to some literary examples. Limited space makes it necessary that this demonstration be but a sketch. * Chapter Four of Charles Dickens’ A Tale of Two Cities18 presents the following motif: A bank man, Mr. Lorry, is trying to tell a young woman, Miss Manette, about the fate of her father whom she has never seen. The delicate message he has to deliver, often interrupted by his interlocutor, is constructed as a particular act of speech. It is indicated by quotation marks like all the other representations of acts of direct speech in this novel. And as readers we have no difficulty in realizing what kind of thing in the real world this motif represents. It is an informative oral narrative, “telling someone else that something happened”: “Like Monsieur Manette, your father, the gentleman was of repute in Paris. I had the honour of knowing him there. Our relations were business relations, but confidential. I was at that time in our French House, and had been—oh! twenty years.” (31) “I speak, miss, of twenty years ago. He married—an English lady—and I was one of the trustees. […] To go on—” (31) “Miss Manette, your mother took this course with you. And when she died—I believe broken-hearted—having never slackened her unavailing search for your father, she left you, at two years old, to grow to be blooming, beautiful, and happy, without the dark cloud upon you of living in uncertainty whether your father soon wore his heart out in prison, or wasted there through many lingering years.” (33)

However, given that this kind of linguistic act is clear, what about the textual instances surrounding this fiction of a narrative, that is, the sequences without quotation mark? How should we, for example, understand expressions like the following ones:

17 18

Roman Jakobson 1960. Dickens 1994.

108

Lars-Åke Skalin

After this odd description of his daily routine of employment, Mr. Lorry flattened his flaxen wig upon his head with both hands (which was most unnecessary, for nothing could be flatter than its shining surface was before), and resumed to his former attitude; (32) Mr. Lorry took the hesitating little hand that confidently advanced to take his, and he put it with some ceremony to his lips; (31) As he said the words he looked down, with admiring pity, on the flowing golden hair; as if he pictured to himself that it might have been already tinged with gray; (33-34) He felt his wrist held closer, and he stopped. The expression in the forehead, which had so particularly attracted his notice, and which was now immovable, had deepened into one of pain and horror. (34)

And further, how should we understand the marks of an interrupted utterance from Mr. Lorry such as: “There has been no new discovery, of money, or of any other property; but—” (34)? What I would like to draw attention to is the disparate character of the details given around Mr. Lorry’s own account. Standard narrative theory has not hesitated how to explain the matter. Here the text of the novel contains the semiotic signs of a primary narrative by a primary anonymous teller—the extradiegetic narrator—whose discourse is quoting the discourse of the intradiegetic narrator, Mr. Lorry. However, as I have now described the theory from the act-aspect (“someone telling…”) I cannot but conclude that it would be problematic to regard the assumed primary narrative as analogous in nature to Mr. Lorry’s. It seems, for example, quite to the point to say that Mr. Lorry gives “an account of” the sad circumstances about Miss Manette’s father; to “narrate” the way he does is just this—to “give an account of” events and circumstances. But presuming that the text surrounding Mr. Lorry’s narrative is also a narrative, I would certainly hesitate about using the phrase “an account of” because I cannot imagine how I should continue after the preposition “of”; what subject could I name that would do justice to all the disparate elements of the actual text? In comparison with Mr. Lorry’s account, the assumed voice giving this “account” must appear as a “strange” voice, an “unnatural” one. Assuming the paradigm “someone telling someone else that…”, what I find missing is a motivation from the narrator assumed to have produced this, for example: “You see, I couldn’t but giggle when I saw him kiss her hand so ceremoniously and incessantly flattening that already flat wig of his” or “Mr. Lorry told me about his feelings in fulfilling his delicate mission to inform Miss Manette. He mentioned that when he felt how that poor girl held his wrist closer, and when he saw the expression in her forehead he had to stop his report for a while.” This would be narration in the natural way because it is presented as information about something relevant to a topic—here Mr. Lor-

How Strange Are the “Strange Voices” of Fiction?

109

ry’s ticks and feelings are in focus. Of course, if all the very disparate details of the framing text had their specific motivations the narrator and his product as a whole would appear rather eccentric but, still in isolation, this text might possibly be identified as a type of natural narrative.19 The theoretically interesting question is whether my intuition of “strangeness” in reading the actual framing text according to the paradigm of a natural narrative would reveal that a different practice is actualized here and, if so, whether we are presented with the representation of a different practice or if we are in fact dealing with a real practice in operation, which, in that case, could not be an informational paradigm. (I will argue for the latter option later on.) Theorists have actually found that there is something strange in utterances of the type exemplified by “He felt his wrist held closer, and he stopped” and have tended to analyze the “unnatural” in epistemological terms: it is fiction because no real teller can know and therefore state what is going on in the mind of others.20 However, my objection has nothing to do with what is supposed stated (some unrealistic matter) or how it was possible to state it (unrealistic powers or sources not accounted for)—what is subject to discussion is the sense-producing language-game which is assumed to operate in this case. The quoted expressions appear as not being formulated from a situation where you can propose a topic for a narrative to be launched. The disparate character of the presented details of the discourse framing Mr. Lorry’s creates the impression (to travesty Paul Ricoeur) of “life in quest of narrative”21 more than of a fully articulated narrative whole. And that impression has to do with a sense of the narrative direction I have referred to above as a peculiarity of fiction as mimesis of action and life. The assumed narrative does not seem to have its solid base in a teller situation from where a topic is suggested. Therefore, the presented details pose questions rather than suggesting answers to questions implied by a given topic—as should be expected by a natural 19

The title of Dickens’ novel announces as its theme the tale of two cities. It starts with a meditation over the historical situation to be dealt with. Then the next chapter has as its motif the details of how a coach with exhausted horses struggles to get up a hill on its way to Dover. If taken to be a narrative about something, the reader will be at a loss guessing what will come out of the mist and mire of that hill. There are now and then textual “signs of a narrator”, for example, “In those days, travellers…”, but that narrator never announces the reasons why he gives us all those details about the struggle to get up the hill. His topic remains a secret. 20 Dorrit Cohn has consistently argued from this epistemological strategy. See her The Distinction of Fiction. A narrator with the power of knowing directly what goes on in the minds of others must be a “signpost” of fiction. The subsequent problem is then how to conceive of fiction in this case. I will return later on to that problem and its relation to the question of the concept of a narrator and the meaning of “narration” and “story”. 21 Ricoeur 1991.

110

Lars-Åke Skalin

narrative. It is significant that one of the six elements, which, according to the model of a full narrative proposed by Labov and Waletzky, is an “abstract” which answers to the question what the narration to come will be about and what will be its point.22 When this is done, the narration can start its course: going back, fetching from the past those elements regarded as relevant to the given topic and, after that, moving forwards on a line prepared by that same topic, satisfying some possible expectations. The text passages embedding Mr. Lorry’s account come forth simply as a forwardgoing progression, and so they come across, as topic-less. Literary theorists relying on Aristotle’s conception of mimesis, on the other hand, would not find anything surprising in this peculiarity of literary fiction. Käte Hamburger, for instance, finds that the property constituting fictional narration, which is mimesis,23 is its absence of a teller situation.24 But such a theory would not try to explain narrative fiction as governed by the same principles as natural narratives. Such a theory would not take fiction as narrated experience from the vantage point of a teller situation but rather as the mimesis of life as ongoing unorganized experiences. These motifs, in turn, will be elements of a meaningful structure in the end but not as complying with the principles of a natural narrative. However, in trying to integrate the mimetically rendered elements into a theory of a narrator, standard narrative theory has made this agent into just a “reporter of facts” as they appear in the unfolding of time. If so, they seem to be “reported” for the sake of their having happened only— not for the sake of their relevance to a presented topic. In other words, the product of such an objective “reporter” would not signal a teller situation, from which is chosen what is relevant to a theme.25 We shall see in the discussion of homodiegetic stories that theorists do actually slide between different conceptions of this purported narrator. *

22

William Labov and Joshua Waletzky 1967. The six elements are: abstract, orientation, complicating action, evaluation, result or resolution, and coda. See also Labov 1972: 363. 23 Cf. the title of Chapter 3 of The Logic of Literature: “The Fictional or Mimetic Genre”. 24 The Logic of Literature, 67-81; 136-142. 25 In fact, “to report” and “reporter” are inadequate terms for what is referred to in this context because they designate a communicative act by an agent in a teller situation. It would be more adequate to think of a machine automatically registering “facts” from a position cotemporal with the happenings. Curiously enough, Mr. Lorry refers to himself to such an automaton: “don’t heed me any more than if I was a speaking machine—truly, I am not much else”. (31) Interpreted as a meta-element in the hands of the author, one might perhaps say that he functions as an exposition-producing machine at this part of the story.

How Strange Are the “Strange Voices” of Fiction?

111

Suppose that we object that natural narratives are not necessarily so unlike literary narratives as that of Mr. Lorry. In Chapter Fifteen of Dickens’ novel, we find another embedded story which does not sound like his at all. The context is this: In Paris a group of conspirators under the leadership of a Monsieur Defarge are making plans for the revolution that was to come some years later. Every person joining the group is spoken to by the name of Jacques plus a number, a name associated with peasant insurrections in the Middle Ages. Now Defarge has called a simple man from the country side, a mender of roads, as a witness to tell about an occurrence in his village. The marquis who owns this province, a monster of cynicism and cruelty, has been murdered by someone who left his name, Jacques, written on a piece of paper round the knife that killed the marquis. Defarge knows that the killer has been found but wants to hear the details of his arrest. As we know, Dickens has a special liking for making his characters “characters”, equipping them with individual attributes, ticks, and obsessions, perhaps a residue from his early career as the creator of “sketches”. Mr. Lorry’s act of narration is accompanied by incessant attempts to cover his personal emotional engagement in what he is telling. The road-maker has the desire to stand out as a real performer with the power to affect the listeners by his way of presenting his narrative. It is full of details evocative of an ominous atmosphere, and he intensifies his talk by gesturing and miming. He is also explicitly referred to as “the narrator” (171). “I saw him then, messieurs,” began the mender of roads, “a year ago this running summer, underneath the carriage of the Marquis, hanging by the chain. Behold the manner of it. I leaving my work on the road, the sung going to bed, the carriage of the Marquis slowly ascending the hill, he hanging by the chain—like this.” Again the mender of roads went through the whole performance; in which he ought to have been perfect by that time, seeing that it had been the infallible resource and indispensable entertainment of his village during a whole year. Jacques One struck in, and asked if he had ever seen the man before? (169) [After some interruptions the teller is asked to go on.] “Good!” said the mender of roads, with an air of mystery. “The tall man [i.e., the man hidden under the marquis’ carriage] is lost, and he is sought—how many month? Nine, ten, eleven?” “No matter the number,” said Defarge. “He is well hidden, at last he is unluckily found. Go on!” (169-170)26

26

Actually, one can get the feeling that Dickens has intended this part as a kind of parody of his own literary narration. Cf. the following expression, “He described it as if he were there, and it was evident that he saw it vividly” (170), with Dickens’ commentary in the foreword to the first edition: “When I was acting with my children and friends, in Mr. Wilkie Collins’s drama of The Frozen Deep, I first conceived the main idea of this story. A strong desire was

112

Lars-Åke Skalin

In this style the narrator goes on. Yet this is a natural informative narrative, and is taken as such by his listeners who do not pay much regard to his performance skills but urge him to go on in his witness report, asking questions whose purpose is not to get a more vivid feeling of the atmosphere but simply to get a clearer picture of what was the case. Despite its rhetorical excesses—the narrator’s propensity for “dramatizing” his experiences by turning to the present tense, for instance—his speech is still perfectly recognizable as a report of particular events and circumstances witnessed by this man. Even the most atmosphere-rendering details make sense in the informative context because they are motivated partly as a consequence of the impact the creepy situation has made on the witness, partly trough the narrator’s wish to show off before an audience. In the terms of sociolinguists like Labov and Waletzky or a theorist of narrative pragmatics like Mary Louise Pratt, this narrator strives to make his discourse match the extreme “tellability” quality of its content.27 His listeners may wish that his talk be less mingled by such superfluities—but these would be a matter of style (it is obviously contrasted with Mr. Lorry’s style), not making his narrative into anything radically different than a report. It has a consistent topic, and there is a point in every statement he utters. Mr. Lorry’s narrative is about the fate of Miss Manette’s father and the road-mender tells about his first encounter with the ghostlike man who was to be the murderer of a tyrannous Marquis, and then about the arrest of this man and, finally, about his subsequent fate. That his narrative is supposed to present these themes is evident from his listeners’ reactions. They have opinions about the relevance of certain elements of the discourse and would be glad if he left out what in their opinion are redundant parts. Yet, as to the text framing the road-mender’s narrative we must ask the same question as we did to the one surrounding Mr. Lorry’s. What is its topic, and what is its point? What is it about? The stuff it displays seems to precede any articulation, which could order the flux into a fixed category such as “an experience of this or that kind”, that is, as something that requires an articulating retrospective station. No framing narration seems upon me then to embody it in my own person, and I traced out in my fancy the state of mind of which it would necessitate the presentation to an observant spectator, with particular care and interest. As the idea became familiar to me, it gradually shaped itself into its present form. Throughout its execution, it has had complete possession of me; I have so far verified what is done and suffered in these pages, as that I have certainly done and suffered it all myself.” (ix) 27 Labov and Waletzky 1967, Labov 1972, Pratt 1977. I have discussed the topic of “tellable” natural narratives in relation to narrative fiction in Skalin 2009.

How Strange Are the “Strange Voices” of Fiction?

113

to motivate what is presented, saying, for instance: “It was evident, you see, that these four Jacques did want him to stick to the facts that were of interest to them, so as soon as he became too personal or rhetorical they told him to go on in his reporting only these facts and not his personal impressions.” This kind of motivation would be relevant to a narrative alleging to report for instance the circumstances during a witness hearing. But what motivation should be given to sentences like the following: Defarge and the three glanced darkly at one another. The looks of all of them were dark, repressed, and revengeful, as they listened to the countryman’s story; the manner of all of them, while it was secret, was authoritative too. They had the air of a rough tribunal; Jacques One and Two sitting on the old pallet-bed, each with his chin resting on his hand, and his eyes intent on the road-mender; Jacques Three, equally intent, on one knee behind them, with his agitated hand always gliding over the network of fine nerves about his mouth and nose; Defarge standing between them and the narrator, whom he had stationed in the light to the window, by turns looking from him to them, and from them to him. “Go on Jacques,” said Defarge. (171)

One might perhaps object that we could imagine a framing narrator reporting how he has observed the situation of how these men react to what they hear from their witness. It is his impression that the men looked as if forming a tribunal. Still, the details of the “agitated hand gliding over the network of fine nerves…” etc would certainly demand its special motivation for it to be accepted in a natural narrative. And even if such a motivation was given, there is still the problem with the passing on from the description to the line: “Go on Jacques, said Defarge” which, as it stands, would sound very strange in a natural narrative. There it would have been motivated as an answer to a possible question posed by the teller/receiver situation: “The fellow’s listeners, you know, were eager to get to what really happened so now and then Monsieur Defarge urged him on saying ‘Go on Jacques’”. This would be face-to-face narrative and might, in a written report, be: “At several moments there was a stop in the witness narration but since Monsieur Defarge was eager to get to the end one could now and then hear his voice: ‘Go on, Jacques!’” These variants would be purported quotations from a teller situation deemed worthy of being reported by the teller. But as the phrase is used in Dickens’ novel, with its abrupt turn from description to direct speech, we will not associate it with the representation of a natural narrative. It is mimesis in the Aristotelian sense and, accordingly, not to be explained by the formula “someone telling someone else that…” *

114

Lars-Åke Skalin

The discussion above, suggesting a distinction between (natural) narration and Aristotelian mimesis, brings to the fore the old controversial issue about the necessary existence of a narrator. I find it not necessary to refer to any of the partakers of that debate, since it is so well known. Yet my own discussion will mark a position on the issue. Some narratologists have advocated the view that all narratives, whether in the telling or in the showing mode, must have a narrator; this is a purely logical conclusion. Others have raised objection to this, arguing that it is not necessary to ascribe a narrator to fiction in the strict showing mode; fiction in the authorial mode, on the other hand, has a narrator. So has fiction in the first-person. My own opinion is that if fiction is taken as mimesis in the Aristotelian sense, which was also the sense Käte Hamburger ascribed to thirdperson literary narrative, then it cannot have a narrator if that notion is taken in the telling act-aspect as described above. Hamburger found, on linguistic criteria, that first-person stories were not mimesis and therefore they have a teller, although “narrator” was not a term she used for that role (“narrator” is to her theory what the “mimetic poet” is to Aristotle).28 However, I disagree (like many others) with Hamburger’s analysis of firstperson literature and find that this mode too is mimesis.29 In my discussion above I have tried to demonstrate that certain text passages in heterodiegetic fiction can be understood as the representation of an act of natural narration while other passages appear as “strange” if understood from that paradigm: the elements presented do not seem to answer to the relevance requirement. Yet, the lack of that property, which is obligatory for natural narratives, becomes even more conspicuous in first-person narratives because here the pronoun will suggest a speaker. The reading of such narratives as fiction seems to require that we do not take the “I” as a pronoun referring to a retrospective teller position. In literary criticism one has often made the distinction between the “narrating I” and the “experiencing I”. The latter term is, of course, a logical anomaly but it contains the very problem at issue. I would like to demonstrate the “unnatural voice”, typical of much first-person fiction, by an episode in Mark Twain’s Adventures of Huckleberry Finn (Chapter xi).30 Huck and Jim have got the idea that Huck shall dress like a girl, go into a very little town, with which Huck is well acquainted from before, and try to get information about their situation without having his identity revealed. Huck finds a house with a woman who, he as28 29

Hamburger 1993: 139-140. I argue for the similarity between fiction in first- and third-person already in Skalin 1991: 176-180. 30 Twain 1995.

How Strange Are the “Strange Voices” of Fiction?

115

sumes, must be newly moved in. She will not know him, he thinks, so he knocks at the door. At that place we are transferred into a new chapter (xi) which opens like this: “Come in,” says the woman, and I did. She says: “Take a cheer.” I done it. She looked me all over with her little shiny eyes, and says: “What might your name be?” “Sara Williams.” “Where ‘bouts do you live? In this neighbourhood?” “No’m. In Hookerville, seven mile below. I’ve walked all the way and I’m all tired out.” “Hungry, too, I reckon. I find you something.” (75)

The conversation dwindles along in drawn-out dialogue parts mixed with parts displaying the emotional state of Huck, who has soon come to feel very uncomfortable. The reason for his feelings is what he hears about the woman’s suspicion—that the runaway Jim might be hiding on Jackson’s Island—but also his growing suspicion that she suspects that he isn’t the one he has given himself out to be. And the woman does suspect that the person in her kitchen isn’t a girl at all. She makes Huck the subject of some tests which will reveal whether he is boy or girl. Among other things she asks him to try to kill the rats that now and then peep out from a rathole by throwing a lump of lead at them. She watches him throw the lump, fetches it back, and then drops it into his lap. Huck catches it by clapping his legs together. By his way of throwing and by his way of catching the lump by clapping his legs together instead of spreading them to catch the object in the cloth of the dress, he is unmasked. And then the dialogue goes on with new lies from Huck’s side and after a long speech from the woman where she exposes the tricks she has played on him and makes clear in a good-humoured way that she hasn’t been convinced by the other names he has given either, but that she will always be ready to send him help, should he need it. After her last word there is this: I went up the bank about fifty yards, and then I doubled on my tracks and slipped back to where my canoe was, a good piece below the house. I jumped in, and was off in a hurry. (80)

And with this the episode has come to its end. Adventures of Huckleberry Finn is, as we know, a homodiegetic novel presented as told by Huck himself. According to standard narrative theory the novel’s text as a whole should be understood as the representation of a natural narrative, Huck’s memoirs, as it were. But if we as readers are supposed to make sense of this text as memoirs we would certainly think it is “a tale told by an idiot”. The reason is exactly the same as that which we found when looking at the heterodiegetic extradiegetic parts of Dick-

116

Lars-Åke Skalin

ens’ novel framing Mr. Lorry’s and the road-mender’s narratives, namely that we cannot possibly identify this as narrating, if narrating is “someone telling someone else that something happened”. There is no topic here organizing the speeches, the actions and the feelings of the experiencing I as a whole presenting what the purported narrator’s discourse is about: what we have is just a stream of life presented moment by moment as in drama. If taken as the representation of a natural narrative one would expect some kind of abstract from the supposed narrator, for example, “I’ll tell you about that rather creepy experience I had with that woman. I planned to get information from her without revealing who I was and I got it too, so in a way I succeeded in taking her in, but gee, how this little clearsighted woman took stupid me in! Listen, this is what happened!” Such an abstract can only be given from a retrospective position, as we have observed above, when the narrator is able to sort out memories to be composed into a narrative with a consistent theme, for example, how the deceiver is deceived. But this would mean the organization of the elements into meaningful hierarchies with regard to their relative importance. Here, the elements are just posited side by side as of equal importance. Why does not the supposed narrator tell us, for example, to take specific care when he comes to the episode with the rat-hunting and the lump of lead, saying something like: “Now, this is the way she got me, how she framed me…”? But instead we are given his reflection about what would have happened if the rat had stayed a little longer. So, again, what we get is only step by step what happens and what the experiencer feels and thinks at the precise moment. When the dialogue is over we just follow Huck away from this house and into the canoe without having been given a reflection about what kind of experience he had just gone through, a reflection which would merge the disparate elements into a whole, that is, into a narrative with a point. When Huck later talks with Jim about this episode he refers to it as: “the time I had jabbering with that woman” (82). And so it would certainly appear to a reader who takes the text passage as a narration by Huck himself in retrospective. However, Jim finds that “she was a smart one” (82), as the reader has already observed. Yet, Huck does not tell an assumed receiver of his “memoirs” that he is fooled by the woman’s tricks; as readers of the novel, we see the theme rather than having it stated by a narrating act of an assumed homodiegetic narrator. *

How Strange Are the “Strange Voices” of Fiction?

117

I have tried to demonstrate a problem that seems to open itself when explicating narrative fiction as a representation of natural narration. That it should be understood that way is argued by, for instance, Barbara Herrnstein Smith31 and, more recently, by Ansgar Nünning in his essay “Mimesis des Erzählens”.32 There have admittedly been objections to this suggestion,33 and I sincerely doubt that readers process Adventures of Huckleberry Finn according to a schema that says “Take this as if Huck Finn told about his adventures”. And I don’t think that we tend to read A Tale of Two Cities as if it were a “chronicle” or a “history”, as the text itself sometimes says (“this chronicle”, “this history”, 15). If we did, we would most possibly find that also Dickens’ tale was “told by an idiot”, which we do not. So giving up the suggested paradigm for making sense of such texts, how should we regard the voices not to find them “strange”? In fact, I think there is not much of a problem here, we just avoid the “narrative fallacy” and ask anew what we have before our eyes. Why not opt for “an aesthetic composition” by Mark Twain or Charles Dickens, respectively? Then it might be objected: “But they are not telling about these occurrences!” No, they don’t, but why stick to the idea that these novels are narratives if the only acceptable meaning of this term is “telling about something”? After all, the Russian formalists took the view that such works are aesthetic compositions constructed by motifs, and if we take them as motifs and not as fictional propositions their source cannot be anyone but the artist. Motifs are constituent elements of works of narrative fiction. True, they have fictional ingredients but they are functional elements of novels and short stories, which are real, although intentional, things. Motifs are not signs referring to a fictional world as the object of interpretation. The fiction ingredient is just a vehicle for the construction of the real object, the work. Accordingly, the thesis from the aesthetic view advocated here is that we appreciate works, such as novels and so on, and not worlds even if they are fictional. This would be a one-level approach opposing the multi-level assumption of standard narratology. * One concept suggested by Gérard Genette in Narrative Discourse has to do with the question of where narrative fiction may display “strangeness”.34 31 32 33

Herrnstein Smith 1970/71: 259-281, 274. Nünning 2001: 13-48. Fludernik 2003c: 1-39, 30-39—especially the argument presented on p. 37. See also Wolf 2004. 34 Genette 1980: 234-237.

118

Lars-Åke Skalin

The concept is metalepsis defined as the mingling of two distinct diegetic levels. An authorial narrator can, for example, make a move that seems to be an invitation to his readers into the assumed fictional world. Monika Fludernik discusses this device. She suggests that it has important theoretical implications and argues that many seemingly ontological transgressions of boundaries should be understood as “metaphorical” rather than “real”. In fact, this assumed ontology exists only in the minds of the readers when applying a “realistic and pragmatic narratology”.35 With this, she is referring to what I have called the narrative paradigm together with the immersion-into-a-fictional-world theory of fiction reading, that is, the internal approach. Fludernik seems to be willing to accept the idea that this may be the only relevant way of regarding traditional narrative literature; the problem appears when we are dealing with literature not based on the story/discourse distinction. I agree with Fludernik that metalepsis has something to tell us about narrative theory but will draw further conclusions from her suggestion about its metaphorical nature. Fludernik presents an informative discussion of metalepsis in scene shifts. She gives examples of how markers of this function, originally with traces of oral literature, have changed through the history of narrative literature from the Middle Ages to modern times. I present here two of the examples of such markers of shifts given by her: Mrs. Tow-wouse […] recovered the usual serenity of her temper, in which we will leave her, to open to the reader the steps which led to a catastrophe. (Fielding, Joseph Andrews) Leaving it to pursue its journey at the pleasure of the conductor aforementioned […] this narrative may embrace the opportunity of ascertaining the condition of Sir Mulberry Hawk. (Dickens, Nicholas Nickleby)

Interpreted from the multi-level approach, the author (or the authorial narrator in Fludernik’s terms) and the reader on their logical and ontological levels seem to be transposed to the level of the fictional entities, Mrs. Tow-wouse and Sir Mulberry Hawk. This would produce an effect of strangeness while being simultaneously a “transgression” of boundaries taken for granted by the theory. But, as Fludernik observes, the function of the device is to produce a scene-shift, which, according to her definition, “is a move from one setting and a set of characters to a different setting and set of characters”. Here, I would like to point to the fact that “setting”, “character”, and “set of characters”, “scene”, “scene-shift”, as well as “plot”, are not represented (that is, fictional) entities but vehicles in the act of the artist’s composing his work. So in my suggested one-level 35

Fludernik 2003a: 389.

How Strange Are the “Strange Voices” of Fiction?

119

approach the reader takes the sentences the only way they can be understood, as the artist’s instruction: “now I’m going to make a shift of scene in my composition”—which would confirm Fludernik’s intuition about the metaphorical status of such expressions. In other words, there would be no point in a literal understanding; the “transgression” is there because of its stylistic value giving a playful touch to the expression, while “regarded as an imaginative transfer into the impossible” (see Fludernik 2003a: 393). But as the real function will still be to produce an actual shift of scene in the novel, understood as the author’s text, such logical and ontological intricacies would hardly be a relevant object of narrative theorizing. However, reflecting on how these presumed narrators converse with their readers in a light-hearted style about what is regarded to be the narrative content of their stories, we can see that they are treating this content as if simultaneous to the narrating of it (Genette 1980: 235, discussing an example from Balzac; cf. Fludernik 2003a: 386). These narrators are not referring to events and circumstances in the past but are commenting on something, which seems to be ongoing, as on a stage. But, this concept of narration and narrator will not be in congruence with the standard definition, “someone telling that something happened”. It seems to be much more in conformity with the one-level model of the aesthetic approach I have advocated. Instead of being a “teller of facts” the act of this so-called narrator will rather call to mind the performance by the puppeteer maese Pedro and his assistant in Cervantes’ Don Quijote de la Mancha. (I will come back to this comparison.) Fludernik’s historical survey of narrative devices is informative because it shows that there are constants making use of shifting means. Look again at the quotation from Dickens’ Nicholas Nickleby: “Leaving it to pursue its journey at the pleasure of the conductor aforementioned […] this narrative may embrace the opportunity of ascertaining the condition of Sir Mulberry Hawk […]” (my emphasis). I suggest that by “this narrative” Dickens as the artist means “this novel”. Similarly, when in A Tale of Two Cities it is said “this chronicle” and “this history”, the expressions should be taken as referring to the work in our hands. And why should it not be so, regarding what we can learn from a diachronic study of functions and forms? As Fludernik observes, there are signs of earlier oral formulas in English medieval literature. One could also point out traces of such formulas, for example in the genre of the Icelandic saga, such as “Ok er hann nú ór sagunni [and now he is out of this saga]”. When later on in the history of prose fiction we find “narrators” of the non-omniscient type, attributed with phrases such as: “this is what I know of …etc.” this should be comprehended just a formal variant of the functional constant “at this place in the composition the following motif

120

Lars-Åke Skalin

should be introduced to produce the effects it is constructed for”. The consequence of such a thesis for the conception of a narrator is obvious: even the most “intrusive” commentaries could be understood as just variant forms for functional constants. And the performer of these is the artist by means of his text. As Plato obviously took for granted in the Republic book III: 392-394: every sentence is uttered by the author although in different functions. When the slogan “exit author” appeared among modern novelists, suggesting that the traditional “telling” mode should be replaced by “showing”, its proponents knew what they were talking about—not the fiction of a narrator but the author as the composer of art. The showing mode was not to be understood as a different sense-making paradigm, because the basic functions necessary to produce a work of narrative fiction were constant. It was all a discussion about what forms and devices could be and should be used to perform the functions intended by the author. With what could one replace, for example, “the dialogue which the reader will find in the next chapter” (Joseph Andrews, end of Chapter xvi)? The most economical measure seemed to be just to take such things away from the list of possible means. And that readers should think that stories in the first-person should be read as true accounts of events, only with the operator as-if, was out of the question. The modern authors knew that the mode of showing was as available for first-person as for third-person stories. The discussion so far has assumed that it would be possible to subordinate a presumed level of represented natural narration informing about a presumed fictional world under a voice which is not a fiction but the voice of an agent that says “this saga” or “this narrative”, meaning “this novel”. In other words, this voice is referring to what is actually present for a reader or listener; it is an act of real reference, not an invitation to them to imagine a fictional world and try to make sense of it to the best of their ability. This reasoning reverses the order of the discourse and the fiction assumed to be its content: the raison d’être of the discourse of the saga or of the novel is not to bring us to what is not real, that is, to a fictional level, but the other way around. This turn would make the concept of metalepsis invalid by taking away the fiction regarded as a level or dimension independent of the discourse. The assumed fictional elements do not constitute an independent world more than, for instance, metaphoric elements do. As a theory of reading narrative fiction the aesthetic model says that our reading is neither of narrative (in the standard sense), nor of a fictional world but of an aesthetic composition, a language-game of its own, presupposing its own distinctive form of life. In fact, Cervantes has in his maese Pedro episode given a very informative contribution to the theoretical issue in demonstrating a tempt-

How Strange Are the “Strange Voices” of Fiction?

121

ing but still erroneous approach to fictional stories by making a parody of its application. Don Quixote has really taken the internal view literally when he attacks the puppets in order to save the unfortunate Melisandra. But why shouldn’t he, if appreciating fiction is to immerse oneself into an assumed world and momentarily understand it as real? Yet, what Cervantes seems to demonstrate is rather that aesthetic appreciation is of the performance as such, not of an independent world of the imagination. This means that the aesthetic object is the laid out discourse, which has fictions as parts of the constituent elements. This would render the observation Genette makes on one type of metalepsis true, namely the one that has the story (in the sense of content) be just a function of the discourse as in “Virgil has Dido die” or in Diderot’s metafictional jestings in Jacques le fataliste (Genette 1980: 234; cf. Fludernik 2003a: 384, 388). The suggested course of reasoning takes it for granted that this is the practice we share in reading fiction—even should we be tempted to make a theory that proposes something else. But narrative theories, as well as aesthetic theories, are by necessity circular in nature. If we were to explicate literature by means of a concept, “aesthetic illusion” for instance, and asked for the meaning of that concept, ultimately we would have to take refuge in something like “the way you appreciate narrative fiction, a practice I know you master”. There is no valid definiens independent of the practice. * Referring back to practice is the only way, as I see it, to solve the problem of strange-appearing voices in literary fiction. According to the aesthetic, or “external” approach argued here, the relevant sense-governing practice cannot be analyzed as a layered procedure as implied by standard narrative theory. That model proposes several sense-governing strata that make the text “double-“ or “triple-voiced”, as it were. When the aesthetic approach ascribes meaning production to the artist, this will exclude the idea that there are other acts communicating meaning, for instance by represented agents. Represented communicators do not communicate with aesthetic appreciators, that is, with us. In acquiescence with the proposed counter-model, I would like to point out one more characteristic of literary works that must appear very strange to what is presumed to be a natural narrative yet perfectly reasonable to what is presumed to be literary art. It is the presence of elements outside the structures of the story in its more restricted dynamic sense— what can be ascribed the property of having “beginning, middle, and end”. I am referring to what might be called the “spatial” aspect of a work

122

Lars-Åke Skalin

which is quite anti-Lessing in nature. This aspect draws attention to the relation between certain elements in a work of fiction, not in its temporal dimension, which is the traditional plot, but so to speak in how they relate in the discourse seen as a plain surface such as we see a painting on a canvas. Therefore, I call it the “canvas-aspect”. Since the painting on the canvas is something real this metaphor will emphasize the thesis that appreciating literary fiction as art is paying heed to something real and not fictive. In a section of Narrative as Rhetoric (1966), James Phelan says with regard to narrators’ self-consciousness: One of the conventions of homodiegetic narration is that unselfconscious narration is the unmarked case: that is, we take the homodiegetic narrator as unselfconscious unless we are given reason to do otherwise. Thus, we assume that the homodiegetic narrator is not the source of such things as foreshadowing, patterns of imaginary, parallelism of incidents, the lyricism of a particular style—unless we have some signal that calls our attention to the narrator’s self-consciousness. (81)

If the source of such elements is not the narrator, it must be the author, or “implied” author, as Phelan prefers to name it. In his discussion about metanarrative Ansgar Nünning picks out mise en abyme as an example of what could not reasonably be attributed to a fictional narrator.36 It is not to be referred to as an element of either narration or metanarration. According to the thesis of no more than one level of communication advocated by my external approach, the features to which these theorists call attention should be regarded as equivalents to features they regard as represented, i.e., fictional, narration. Let me give just a few examples of how this “canvas-aspect” can produce meaning. Remember the road-mender’s talk about “long shadows […] like the shadows of a giant” (170) and his illustration of the sound of steps: “tramp, tramp” (ibid.). These are examples of leitmotifs in A Tale of Two Cities where a network of symbols operates beside the plot in a more restricted sense. This symbolic dimension recurs in many of the chapter titles. To take only those referring to shadows and footsteps there are: “The Night Shadows”, “Echoing Footsteps”, “The Substance of the Shadow”, and “The Footsteps Die Out For Ever”. Technically, most of these motifs suggesting shadows and echoes of footsteps are fictions. They are presented as actual happenings in the story, or as something on which the characters reflect; they may be presented as their fantasies in a particular situation, or someone may narrate them, as we saw in the road-mender’s example. But they can also appear on the paratextual level, as we saw from the chapter titles. Still, they are all held together by the principle of equivalence. As foreshadowing devices they function as preparation of the reader by 36

Nünning 2001: 34-35.

How Strange Are the “Strange Voices” of Fiction?

123

being ominous of something that will appear sooner or later in the plot, but together with the plot they contribute to the aesthetic effect of the novel as a whole. There is something similar I would suggest with the verses recited by the heroes of the Icelandic sagas whenever they find it appropriate to do so. Supposedly no reader feels a need for “naturalizing” these extremely unrealistic components to make them fit into the fiction of an action; this might be an indication that the real object of appreciation is the discourse as work and not “something told about”. The described effects are not local and temporal but inform the whole as a single vision which is the “spatial” dimension, where the discourse itself stands out as an aesthetic composition. Of course, if we should just stick to the idea of this discourse as the instrument of fictive information about events and circumstances, elements such as these ones must appear as “strange”. Yet, since they do not, this fact should have something to tell us about the nature of narrative fiction. We have to remember that what on the face of it is presented as the fiction of a piece of action does not necessarily have its function in the plot; it may have its function in the canvas dimension as a whole where the plot is one constituent. Let me finish my discussion by just one more example from Adventures of Huckleberry Finn. Jim and Huck have discovered the dead man in the house drifting on the river and Jim has told Huck not to look at him (chapter ix). Still, Huck is eager to talk about that man (chapter x). Jim refuses blankly, arguing that such talk would bring bad luck. Then the theme for some time becomes bad and good luck and the turning of luck. Huck is a little sore about Jim’s attitude and gets the idea of playing a practical joke on him when he finds a rattlesnake, which he kills (in their discussion they have not been in agreement about whether touching a snakeskin will bring disaster or not). Jim is bitten by the snake’s mate—Huck has forgotten that snakes always take care of each other. However, Jim cures himself by a series of magic rituals but also by drinking lots of “Pap’s whisky”. And “Pap” is referred to in several places on these pages; for example, Huck remembers stories “Pap” has told about what causes bad luck. Then chapter xi follows, with Huck’s confrontation with the woman who outsmarts him. He learns from her that “old Finn” has raised hell and has been seen in the company of “hard-looking strangers”. The next chapter starts with some memories of “Pap’s” attitude as to “borrowing” things from people without asking their permission and then comes the adventure with the robbers and murderers in the steamboat wreck on the river. The wreck is about to sink. Huck unknowingly causes the death of these men by taking their skiff in order to escape since the raft has broken its moorings and is gone.

124

Lars-Åke Skalin

In the very last lines of the novel, Jim reveals to Huck that the dead man in the drifting house was his Pap so Huck’s money is saved. Here are compositional motivations of two different kinds, both of them with a logic totally foreign to common sense. A happy end to the plot requires the “elimination” of Huck’s father—he must be out of the story. Still he is Huck’s “Pap”, and the son has to revenge the foul murder of him. We take the men on the wreck as his killers because of the woman’s information about Finn’s company in chapter xi, which is followed by the drowning of the scoundrels in the following chapter. Huck is unconscious about all this. However bizarre this logic is, it has to do with the composition of the plot. Still, there is no reasonable causality in the connection of these elements—the connection is on the canvas only as patches of certain colours, which are set in relation to each other. Our “inference” as readers is not dependent on causal hypotheses based on probability argumentation. If so, there would be an area of indefiniteness: these men might be “Pap’s” killers or they might not. But there is no such thing because there is no referent independent of the discourse: what we have are patches on the discourse as canvas, as it were, and they are bound to have a relation to each other.37 How shall we make sense of the theme of luck and the frequent references to “Pap” in those chapters? I think that Mark Twain has found it necessary to construct a bridge from the dead man in the house to the ending. The reader should have “Pap” on his or her mind in connection with the drifting-house episode; therefore it must be “Pap’s whisky” that cures the snake-bitten Jim. Some commentators have suggested that Huck’s practical joke on Jim should be seen as a mean action and an indication of how morally immature the protagonist still is in this part of the novel. Well, Jim has refused to tell him about the death of his father! But Huck didn’t know at that moment that the man was his father, someone might object. Didn’t know? Here there is no one to know anything because there are no people here. There are literary characters, patches of form and colour on a canvas being connected to other building patterns. If anybody took this as a narrative that copies accounts of our real world, he or she would certainly meet a most “unnatural” logic. Still, it is the logic of art.

37

Of course, there are no “proofs” of what relations are really there and of how these should be interpreted, because these are not empirical questions but normative ones.

How Strange Are the “Strange Voices” of Fiction?

125

References Cohn, Dorrit 1999 The Distinction of Fiction, Baltimore: The Johns Hopkins University Press. Culler, Jonathan 1977 “Foreword”, in: Tzvetan Todorov, Poetics of Prose (1971), translated from the French by Richard Howard, Ithaca: Cornell University Press. Dickens, Charles, 1994 A Tale of Two Cities [1859], London: Penguin Popular Classics. Fludernik, Monika, 1996 Towards a ‘Natural Narratology’, London: Routledge. 2001 “New Wine in Old Bottles? Voice, Focalization, and New Writing”, New Literary History 32.3, 611-638. 2003a “Scene Shifts, Metalepsis, and the Metaleptic Mode”, Style 37.4, 382-400. 2003b “Chronology, time, tense and experientiality in narrative”, Language and Literature 12.2, 117-134. 2003c “Metanarrative and Metafictional Commentary: From Metadiscursivity to Metanarration and Metafiction”, Poetica 35.1-2, 1-39. Frank, Joseph 1991 The Idea of Spatial Form, New Brunswick: Rutgers University Press. Genette, Gérard 1980 Narrative Discourse: An Essay in Method [Discours du récit, 1972], translated by Jane E. Lewin, Ithaca: Cornell University Press. Göransson, Ulrika 2009 Romankonst och berättarteori: En kritisk granskning med utgångspunkt i Göran Tunströms författarskap [The Art of the Novel and Narrative Theory: A Critical Discussion Based on the Writings of Göran Tunström], Örebro: Örebro University. Hamburger, Käte 1993 The Logic of Literature [Die Logik der Dichtung, 1957], 2nd rev. ed., trans. Marilynn J. Rose, Bloomington: Indiana University Press. Jakobson, Roman 1960 “Closing Statements: Linguistics and Poetics”, in: Thomas A. Sebeok, Style in Language, Cambridge, Massachusetts: MIT Press, 350-377. Labov, William and Joshua Waletzky 1967 “Narrative Analysis: Oral Versions of Personal Experience”, in: June Helms (ed.), Essays on the Verbal and Visual Arts, Seattle: University of Washington Press. Labov, William 1972 Language in the Inner City: Studies in Black English Vernacular, Philadelphia: University of Pennsylvania Press. McKeon, Michael (ed.) 2000 Theory of the Novel: A Historical Approach, Baltimore: The Johns Hopkins University Press. Murdoch, Iris 2003 The Black Prince [1973], New York: Penguin. Nünning, Ansgar 2001 “Mimesis des Erzählens: Prolegomena zu einer Wirkungsästhetik, Typologie und Funktionsgeschichte des Akts des Erzählens und der Metaroman”, in: Jörg Helbig (ed.), Erzählen und Erzähltheorie im 20. Jahrhundert. Festschrift für Wilhelm Füger, Heidelberg, 13-48. Phelan, James 1996 Narrative as Rhetoric: Technique, Audiences, Ethics, Ideology, Columbus: Ohio State University Press.

126 2005

Lars-Åke Skalin

Living to Tell about It: A Rhetoric and Ethics of Character Narration, Ithaca: Cornell University Press. Pratt, Mary Louise 1977 Toward a Speech Act Theory of Literary Discourse, Bloomington: Indiana University Press. Prince, Gerald 1989 A Dictionary of Narratologym revised edition, Lincoln: The University of Nebraska Press. Rabinowitz, Peter J. 1977 “Truth in Fiction: A Reexamination of Audiences”, Critical Inquiry 4.1, 121-141. Richardson, Brian 2006 Unnatural Voices: Extreme Narration in Modern and Contemporary Fiction, Columbus: The Ohio State University Press. Ricoeur, Paul 1991 “Life in Quest of Narrative”, in: David Wood (ed.), On Paul Ricoeur: Narrative and Interpretation, London: Routledge. Ryan, Marie-Laure 1991 Possible Worlds, Artificial Intelligence, and Narrative Theory, Bloomington: Indiana University Press. Searle, John 1974/75 “The Logical Status of Fictional Discourse”, New Literary History 6, 319-332. Skalin, Lars-Åke 1991 Karaktär och perspektiv: Att tolka litterära gestalter i det mimetiska språkspelet [Character and Perspective: Reading Fictional Figures in the Mimetic Language Game], Uppsala/Stockholm: Almquist & Wiksell. 2009 “‘Telling a Story’: Reflections on Fictional and Non-Fictional Narratives”, in: LarsÅke Skalin (ed.), Narrativity, Fictionality, and Literariness: The Narrative Turn and the Study of Literary Fiction, Örebro: Örebro University, 201-260. 2009 “Centres and Borders: On Defining Narrativity and Narratology”, in: Per Krogh Hansen (ed.), Borderliners: Searching the Borders of Narrativity and Narratology, The University of Southern Denmark, 19-75. Smith, Barbara Herrnstein 1970/71 “Poetry as Fiction”, New Literary History 1, 259-281. 1981 “Afterthoughts on Narrative III: Narrative Versions, Narrative Theories”, in: W.J.T. Mitchell (ed.), On Narrative, Chicago: The University of Chicago Press, 209-232. Stanzel, Franz K. 1984 A Theory of Narrative [Theorie des Erzählens (1979), 2nd ed. 1982], translated by Charlotte Goedsche, Cambridge: Cambridge University Press. Twain, Mark 1995 Adventures of Huckleberry Finn: A Case Study in Critical Controversy [1885], edited by Gerald Graff and James Phelan, Boston: Bedford Books of St. Martin’s Press. Walton, Kendall 1990 Mimesis as Make-Believe: On the Foundations of the Representational Arts, Cambridge, Mass.: Harvard University Press. Wolf, Werner 2004 “Aesthetic Illusion as an Effect of Fiction”, Style 38.3, 325-351.



STEFAN IVERSEN (Aarhus)

States of Exception: Decoupling, Metarepresentation, and Strange Voices in Narrative Fiction 1. Introduction Ever since Gérard Genette’s groundbreaking work in Figure III (1972), the concept of voice in narrative fiction has been inextricably linked to the question “Who speaks?” In the work of the early Genette the question is easily answered: the narrator speaks, either as a participator in the narrated world or as someone standing outside it. Originally situated in a structuralist approach to narrative, the framings of the question as well as the possible answers to it have since evolved considerably, the former by adapting to constructivists’ insights from among other gender studies, reader response theory and cognitive narratology,1 and the latter by also focusing on narrative situations that do not easily, if at all, allow for unequivocal judgments when it comes to linking the words of a narrative to an anthropomorphic speaker. This second interest is central to what has become known as unnatural narratology.2 Expanding upon the so-called “no-narrator” theory, Henrik Skov Nielsen has drawn attention to certain cases of first-person fiction that present the reader with information that cannot possibly stem from the character-narrator, and on that ground has argued

1

A recent, convincing attempt to combine some of these post-structuralist conceptualisations of voice is Liesbeth Korthals Altes’ “Voice, Irony and Ethos” (2006). She argues in favour of bringing together a “cognitivist-constructivist” and a “rhetorical-stylistic” (165) approach to the notion of voice in verbal narratives in order to take into account both the dynamic process of the reader’s reconstruction of a voice and the fact that certain stylistic features can be shown to trigger certain reconstructions. 2 See Alber, Iversen, Nielsen and Richardson: “Unnatural Narratives, Unnatural Narratology. Beyond Mimetic Models” (2010) for an outline of this subfield, and Iversen: “Unnatural Minds” (forthcoming) for an account of some of the different positions in the field. See also in this anthology articles by Kraglund, Mäkelä, Nielsen, Reitan and Richardson for different ideas related to the interest in unnatural narratives and to the field of unnatural narratology.

128

Stefan Iversen

in favour of what he calls an impersonal voice in fiction.3 In Unnatural Voices (2006), Brian Richardson catalogues and systematises a large set of what he calls anti-mimetic narratives, mainly but not exclusively drawn from late modernism and postmodernism, all of them heavily invested in mocking, playing with, transgressing or defying the idea of a “who” that is “speaking”. My point of departure in this article is similar to that of Nielsen and Richardson in that I take interest in a type of text that, through identifiable textual cues, defiantly challenges the possibility of answering the question “Who speaks?” and thus resists the anchoring of its narrative in any one voice. What is new in the approach presented here is the choice of text and the choice of theoretical framing of the question of narrative voices and fiction. The very strange voices that form the object of study in the following emanate from a narrative that simultaneously evokes the mutually exclusive reading practices connected to fiction and non-fiction. In an attempt to describe the working of such a narrative situation I will bring the concept of narrative voice into dialogue with the concepts of decoupling, metarepresentation and scope syntax as they are developed and used in evolutionary psychology. This amalgamation of theoretical approaches potentially has consequences not only for dealing with this specific type of strange narrative voice but also for an understanding of what fiction might be and how it might function. The article falls into five parts, of which this introduction is the first. In the second part I want to highlight some of the important ramifications that the theories of metarepresentation and scope syntax, as developed by Leda Cosmides and John Tooby, might have for discussion of the distinctions between fiction and non-fiction. In the third section I will narrow the focus in order to examine the possible uses of bringing these theories to bear on dealing with voices in fiction before, in the fourth part, testing this blend of theories in an aspectualised reading of the text Selvmordsaktionen [The Suicide Mission] (2005),4 written by clausbeck-nielsen.net. In the fifth and final section I will conclude on the work presented in the article. I am not the first narratologist to let the notion of metarepresentations influence theories of narrative fiction, nor even the first to do so by drawing upon the works of Cosmides and Tooby. Their seminal article “Consider the Source: The Evolution of Adaptations for Decoupling and Metarepresentation” plays an important role in the second part of Lisa Zunshine’s Why We Read Fiction (2006), one of the standard works in the 3 4

Cf. Nielsen’s “The Impersonal Voice in First-Person Narrative Fiction” (2004). At the time of this writing the text has not yet been fully translated into English. Quotes are my own translations from the Danish text.

States of Exception

129

wave of cognitive narratology of the last decade. While clearly inspired by and indebted to Zunshine’s approach, my article aims at drawing different, though not only incompatible, conclusions from the works of Cosmides and Tooby. Basing her approach on what is known as Theory of Mind or MindReading,5 Zunshine uses ideas from cognitive science to draw attention to important interactions between readers and fiction, focusing on the latter’s ability to “rely on, manipulate, and titillate our tendency to keep track of who thought, wanted, and felt what and when” (Zunshine 2006: 4). This approach frames fiction as an epistemological training ground that can satisfy as well as create what Zunshine calls “cognitive cravings” (4). Through a series of readings, Zunshine demonstrates how fiction ranging from Clarissa to the modern detective novel may be seen as an “exaggerated literary engagement with our source-monitoring capacity” (5). According to Zunshine, we are able to anchor the sayings of—as well as decode the desires and beliefs of—the characters in fictional narratives by using our Theory of Mind, which in return benefits from these vicarious runs through our folk psychological routines. My approach differs from that of Zunshine in at least three ways. First, I concentrate on a type of narrative that not only raises the question “Who speaks?” but does so in a way that radically challenges any finite answer to the question. Second, these strange, permanently unresolved voices are primarily located not in the story world but at the source of the story world; they emerge at the level of the narrative act rather than at the level of the narrated world. Third, and this is where the discrepancies between the stance taken by Zunshine and traditional cognitive narratology on the one hand and the stance taken in this article on the other are most significant, I will argue that what narratology and theory of fiction can and perhaps ought to learn from the works of Cosmides and Tooby amounts less to the insight that fiction may act as a training ground for real life competences and more to the insight that fiction is a crucial component of what enables real life competences to function in the first place.

5

Since Premack and Woodruff’s classic article “Does the Chimpanzee have a Theory of Mind?” (1978), the idea of humans understanding beliefs and desires of other humans through a Theory of Mind or mind-reading ability has been a cornerstone of much research done in different branches of the cognitive sciences. During the last couple of years the critique of this notion has grown stronger and ever more convincing, thanks to work done by, for example, Daniel Hutto and Shaun Gallagher. See Iversen (forthcoming) for a discussion of the possible impact of this critique on cognitive narratology.

130

Stefan Iversen

2. The Cognitive Niche: Metarepresentations and Scope-Syntax Much of the pioneering insight developed in what has become known as evolutionary psychology, a field founded and led by Cosmides and Tooby, revolves around trying to explain the evolution of what they term “the cognitive niche” of human consciousness. How is it, they ask in their highly influential article “Consider the Source: The Evolution of Adaptations for Decoupling and Metarepresentations”, that humans are capable of using “contingent information for the regulation of improvised behaviour that is successfully tailored to local conditions” (Cosmides and Tooby 2000: 53)? The novelty in their answers, and the main reason these answers are productive for dealing with theories and practices of fiction, including the question of narrative voice, stem from the fact that they highlight not only what characterises the computational abilities of human intelligence but also what it is that ensures that these abilities function. In the approach advocated by Cosmides and Tooby, the human brain is envisioned as a highly sophisticated piece of information-processing equipment whose unique combination of hardware and software makes it capable of constantly improvising by using local and contingent features of the immediate environment. One of the challenges faced by such an inference-based computational system is the problem of inferences being drawn on the basis of information that is either false or is only true under certain local and contingent conditions – what Cosmides and Tooby refer to as “the scope problem”. Given the networked nature of inferencebased information processing, a false piece of information inserted into such a system will very quickly lead to a large number of false inferences, which in turn will be fed into other chains of inferences, and so on. Equally problematic is information with a truth-value heavily limited by scope and conditions. Therefore, according to Cosmides and Tooby, the evolution of “cognitive firewalls” (105) has played a crucial part in the relative success of the human species. Such firewalls have been set up in order to manage “the threat posed by false, unreliable, obsolete, out-of-context, deceptive, or scope-violating representations” (105). In an effort to model how such a system and the mechanisms in place to prevent the system from malfunctioning might work, Cosmides and Tooby suggest distinguishing between fundamentally different types of information inside the architecture of the system. One type consists of information that the system treats as architecturally true, that is, information that is free to be used as a means of drawing inferences; it is “allowed to migrate (or be reproduced) in an unrestricted or scope-free fashion throughout an architecture” (60). Information of this kind makes up

States of Exception

131

what they call our “semantic memory”. Another kind of information is, not surprisingly, tagged as false information. But, and this is where the model holds interesting insights for those working with fiction and voices in fiction, the system is capable of applying a wide range of other tags to information, such as “provisionally true”, “undecided”, “previously true”, “true from a certain perspective”, and so on. This ability to perform a temporarily or permanently bracketing out of the question of truth-value of a piece of information is not an optional feature of the system but in fact crucial to its workings, because it is this bracketing out of pieces of information that enables the system to adapt and to cope with the problems of scope and false inferences. This is due to the fact that information not yet classified as true or false may still be processed by the system’s computational and inferential power: [...] one critical feature is the capacity to carry out inferential operations on sets of inferences that incorporate suppositions or propositions of conditionally unevaluated truth value, while keeping their computational products isolated from other knowledge stores until the truth or utility of the suppositions is decided […]. (59)

A major source of the adaptive strength of human intelligence, according to Cosmides and Tooby, comes precisely from its ability to experiment with the inferential outcome of information whose truth-value is not yet decided or is not decidable at all. A very wide range of competences, considered by most to be fundamental to human understanding and human interaction, are made possible by this ability to draw inferences on the basis of temporarily or permanently quarantined information; or, as Cosmides and Tooby put it, “an architecture that only processes true information is highly limited in what it can infer” (65). Among such competences are our skills for planning future action, for interpreting, evaluating and reacting to information given by others, for detecting intentions and deceptions, and for inventing and using suppositions, pretend play, counterfactuals and fiction. A basic premise behind the distinction between non-tagged, free-floating information and tagged information is the idea, drawn from evolutionary thinking, that a system will use the most cost-effective way of organising its information in order to maximise its chances of survival. Having information that is treated as true circulating as unmarked, rather than marked as true, reflects this idea – too much energy would be wasted if the system was to tag all its information. This is why, Cosmides and Tooby convincingly argue, the default setting of human consciousness is what they term “naïve realism”. How, then, does the system handle this tagging, untagging and retagging of information? It does so mainly by the use of what is known as metarepresentations. While by no means an unfamiliar phenomenon in fields

132

Stefan Iversen

such as psychology, philosophy, logic, theory of art and linguistics, the notion of representations that represent representations (i.e. “John thought that polar bears love ice-cream” or “John said that Polar bears love ice-cream”) has become a central object of interest in several subfields of cognitive studies.6 In “Consider the Source: The Evolution of Adaptations for Decoupling and Metarepresentation”, Cosmides and Tooby opt for a “relaxed [...] sense” (61) of the term when they define it as “representations described by other data structure” (61). Their main contribution to the study of metarepresentations is aimed at explaining how the system attempts to solve the problem of scope raised by the use of contingent information: “because information that is only applicable temporarily or locally begins to be used [by the human species], the success of this computational strategy depends on continually monitoring and re-establishing the boundaries within which each representation remains useful” (58). These tasks of constant evaluation and re-evaluation of information, this permanent questioning of the source of information, is handled by what Cosmides and Tooby call “scope syntax”. Scope syntax refers to a set of “procedures, operators, relationships, and data-handling formats that regulate the migration of information among subcomponents of the human cognitive architecture” (60). This regulation takes place in what is referred to as a “workspace”, containing different data structures or sets of data. This data may come from various sources such as perception, memory or supposition. The datasets in the system “exist in structured, hierarchical relations” (62), where the ground elements are free of tags and thus treated as architecturally true inside the system, such as: (1) Polar bears love ice-cream7 Now consider the same piece of data regulated or subordinated by a metarepresentation: (1) Six months ago John told me that (2) Polar bears love ice cream Taken together, the two pieces of information form what Cosmides and Tooby call a “scope operator”, that is a dataset consisting of two or more 6

Among the important contributions are Leslie’s “Pretense and Representation: The Origins of ‘Theory of Mind’” (1987), Sperber’s anthology Metarepresentations (2000) and Wildgen and van Heusden’s anthology Metarepresentation, Self-Organization and Art (2009). 7 I would like to thank Tim Caudery for providing this example and for helping me finalise this article.

States of Exception

133

pieces of information constructed from the need to “define, regulate, or modify” (63) the relationships between different truth-levels of information. The subordination of the information in the second sentence radically modifies the truth-value of the sentence but does not remove it from the system. Rather, it is kept available for future re-evaluations, but tied to a superordinate piece of information, in this case the architectural truth that John made an utterance. The subordinated or scope-regulated information will always carry its superordinated tag with it when reassessed, since it is this tag that “describe[s] the boundary conditions under which the information is known to be accurate” (63). Such representations that are “bound or interrelated by scope operators” (64) are called “S-representations” in the vocabulary of Cosmides and Tooby. Further super- or subordinations are possible, such as: (1) last night I dreamt that (2) six months ago John told me that (3) Polar bears love ice cream The leftmost piece of information remains unmarked or scope free, while the subordinated pieces of information are scope-limited to different degrees. This logic of framing or scope-regulating information inside the system relies on clear distinctions between the different levels of superand subordination. It is this hierarchical order that helps limit the number of possible allowed inferences and thus solve the scope problem: “the full power of whatever parts of propositional logic are implemented in the human mind can be brought to bear as long as the levels are kept separate for the purpose of inference making” (77). Before zooming in on the possible ramifications this theory might have for working with fiction and voices in fiction, one more concept needs to be mentioned, the concept of decoupling. In Leslie’s classic article “Pretense and Representation: The Origins of “Theory of Mind” (1987), the term “decoupling” is used to designate the shift from what he calls a primary representation (the world as it is) to what he calls a metarepresentation (the world as quoted): the “metarepresentational context decouples the primary expression from its normal inputoutput relations” (Leslie 1987: 417). This decoupling allows for humans to perform, among many other things, pretend play. Cosmides and Tooby further develop this notion. In their vocabulary it characterises the mechanism that allows a computational system to subordinate information using scope syntax. Any type of information that is not stored as architecturally true may be decoupled from the unmarked data in the system while still being enabled for scope-regulated processing. The decoupling ability

134

Stefan Iversen

is what makes it possible to “regulate the migration of information into and out of subordinate datasets, coupling (allowing data to flow) or decoupling them according to the nature of the operator and the arguments it is fed” (63). While coupling and decoupling are highly dynamic mechanisms (different data and datasets may change places in the chain of super- and subordinations as new information, originating from perception, memory, inferences or other sources, enters the work space of the system), their relative success depends on the degree of certainty with which they are temporarily fixed. An impaired decoupling function, i.e. a lacking ability to separate representations from metarepresentations, may cause the system to either store false or fictive propositions as true8 or to limit the system’s access to the massive benefits gained by metarepresentational abilities such as making plans and suppositions as well as understanding desires, beliefs, and intentions of other people.9 3. Decoupling, Fiction and Narrative Voice In this view of how the human species have been able to enter and sustain their cognitive niche, the ability to ask and answer the question “Who speaks?”—the ability to “consider the source”—is crucial. The ability to decouple, to separate representations and metarepresentations without expelling or overwriting either of the two, is what prevents the computational system of human intelligence from breaking down. One important insight gained from this explanation of what distinguishes human intelligence is that fiction and different kinds of suspended disbelief are not optional features or supplements to an already completely operational system, but rather core components of the system itself; they are “design features of a computational architecture designed to solve the problems posed by the many varieties of contingent information exploited by our ancestors” (Cosmides and Tooby 2000: 72). What one might call the naïve realistic approach to information and representation is nested within the human ability to process information as if it were made up, rather than the other way around. The notions of decoupling and scope syntax not only emphasise the fundamental role played by fiction, they also offer an attempt at explaining 8

Roughly speaking this is what Christopher Frith refers to as “the schizophrenic situation” in his seminal book The Cognitive Neuropsychology of Schizophrenia (1992). 9 This is what has been referred to as “the autistic situation” in Baron-Cohen’s Mindblindness (1995). Prior to this, Leslie hypothesises that autistic children are “specifically impaired in their power of decoupling” (Leslie 1987: 424).

States of Exception

135

why and how fiction works. To explain the “why” of fiction, Cosmides and Tooby compare fiction to other types of vicarious experiences such as pretend play, claiming that pretend play is “designed to organize the adaptation in preparation for the performance of its function” (73). This raises the following question: “What could possibly be useful about fictive, counterfactual, or imagined worlds – that is, about false or indeterminate information?” (74). The main function of fiction is not to “guide behaviour” (89) because of the “obviously and blatantly false situations” that are often “extravagantly at variance with ordinary reality” (89). Rather, they suggest that, in contrast to Aristotle’s notion of catharsis, the “fictionally triggered emotion is a re-weighting process of the motivational system” (90). So while we play in order to prepare for the real thing, we read fiction in order to reset or test the ways in which we posit weight on the desires, beliefs and motivations of others and of ourselves. This is the idea of fiction of as an epistemological training ground, picked up and elaborated on by Zunshine. The “how” of fiction is closely tied to Cosmides and Tooby’s notion of decoupling. According to their understanding, s-representations “explain how humans process fictional worlds without confusing their environments and inhabitants with the real world” (108, note 12). They see the whole work of fiction as once and for all decoupled: Representations of [fictional] stories require hierarchical levels so that inferences are restricted in their scope of application to one level at a time. When applied to the content of the story itself, we are able to make all the ordinary implicit inferences necessary to understand the goals, intentions, beliefs and motivations of the characters in the story. When applied superordinately to a representation of the agent who is telling the story, one can make inferences about the real world and the agent. (92)

Most narratologists would recognise this as the distinction between the fictive possible world and the world of the real author. Thanks to an initial decoupling and to the further scope-regulating mechanisms, the beliefs and desires of these two levels (and any additional sublevels that might be subordinated to the fictive frame) are clearly separated. This explanation mirrors most of narratology’s standard models of how fictive narrative works: What Humbert Humbert might desire or believe has no necessary connection to – is situated on another level than – what Vladimir Nabokov might desire or believe. The insistence on hierarchical levels is also what makes part of the ideas of Cosmides and Tooby isomorphic to Genette’s theory of narrative voice. When it comes to anchoring the narrated discourse in the voice of a narrator, Genette leaves no room for deviations. In the following passage from Narrative Discourse Revisited (1988),

136

Stefan Iversen

Genette discusses and rejects Banfield’s no-narrator approach, which deals with the possible lack of anchoring a narrative: […] the main point of Narrative Discourse, beginning with its title, reflects the assumption that there is an enunciating instance—the narrating—with its narrator and its narratee, fictive or not, represented or not, silent or chatty, but always present in what is indeed for me, I fear, an act of communication. (Genette 1988: 101)

To Genette, every narrated word emanates from a source, from the narrator. Genette adds that were he ever to experience a non-anchored narrative, a story not told by a teller, he would “flee as quickly as my legs could carry me” (101). As with Cosmides and Tooby Genette, we cannot help but consider the source, we must consider the source, and there is always a source. Taken together, the why and how of fiction suggested by Cosmides and Tooby offer a very plausible explanation of how most readers deal with most fiction. But the assumptions that the source is always securely locked and that the hierarchical levels of the dataset are always in a fixed, stable position seems potentially reductionist, not only with regard to the existence of narrative fiction that behaves otherwise but also and more importantly with regard to their own theory of how humans have been able to enter and sustain their cognitive niche. To me, there seems to be a discrepancy between the role awarded to fiction in the theory as a whole and the notion of fiction they outline. A key element of the way scope syntax is supposed to function is its dynamic character. While almost always fixed, the source is rarely, if ever, permanently locked. However, in their sketch of what distinguishes fiction and the voices of fiction, Cosmides and Tooby restrain from drawing the full consequences of their own point that most sources, and perhaps all, are able to change status. The history of narrative fiction holds many examples of datasets that have changed representational status over time.10 But what happens when a narrative incorporates or exploits this sourcechanging potential—a potential of all information—as part of its structure, forcing its reader to continuously perform scope-re-evaluations when dealing with the narrative? Such a narrative would not primarily ask of its 10

Such changes are most often hotly disputed, which testifies to the fact that the distinction between whether to store a set of information as scope-free or as fiction matters a great deal to many readers. Zunshine discusses the debate surrounding Defoe’s Robinson Crusoe (Zunshine 2006: 68-70). In the article “Natural Authors, Unnatural Narration” (forthcoming), Nielsen launches the useful distinction between underdetermined and overdetermined as a means of distinguishing between different kinds of fiction/non-fiction transgressions on the basis of paratextual information. His example—the fray surrounding the ontological status of Frey’s A Million Little Pieces—clearly shows that such debates still matter.

States of Exception

137

reader to interact with the desires and beliefs of characters in the fictional world; rather, it would put source-consideration as such on top of the agenda, repeatedly provoking the question “Who speaks?” In the notion of fiction suggested by Cosmides and Tooby (and other cognitive scientists) and advanced by cognitive narratologists such as Zunshine, cases of permanently unresolved voices seem to be relegated to pathology.11 But the world of narrative fiction contains a fair number of texts that revolve around the notion of transgressing hierarchical levels and questioning sources of information, not only sources in the story world (characters) but also the source of the story world (the narrating instance). In the next section of this article I will direct my attention towards the strange voices of such a narrative in an attempt to elucidate some of the consequences of the concept of decoupling and scope syntax developed by Cosmides and Tooby. 4. The Suicide Mission Before we begin the case study of The Suicide Mission, some framing information is required, not only because the text itself is very much concerned with framings and the questioning of framings, but because the text is part of a larger project that insists on not referring to itself as art. Behind, or rather in the middle of, this larger project one finds a Danish performer and inventionalist who was baptised Claus Beck-Nielsen, but who since 2001 has declared his own identity dead. He wrote the biography of Claus Beck-Nielsen, called Claus Beck-Nielsen (1963-2001), and his tombstone is placed in a churchyard in Copenhagen. In the place of his identity stands a person who refers to himself as the nameless one and who spends most of his waking hours as an employee of the company Das Beckwerk (www.dasbeckwerk.com), a production company involved with books, drama, music and happenings. Das Beckwerk describes the passing or disappearing of Claus Beck-Nielsen in the following words: The only real deed that Claus Beck-Nielsen carried out in his lifetime was to marry a woman. And the only real trace he left in the Danish sand was a child, a beautiful little girl named Emma. But as his story is that of the average Danish man around the turn of the century, his marriage failed, he lost his child (to the 11

As mentioned previously, researchers such as Baron-Cohen and Frith have suggested using the distinction between representations and metarepresentations and thus the mechanism of decoupling as a diagnostic tool in dealing with autism and schizophrenia. People with the former condition are unable to grasp suspended representations like fiction while people with the latter condition are unable to grasp the distinction between suspended and nonsuspended representations.

138

Stefan Iversen

woman), and in the autumn of 2001 he "gik bort" as the Danish say, he didn’t just pass away, no, he literally seem to have "walked off". One year later, in 2002 Das Beckwerk was founded to "govern and develop the life&works of Claus Beck-Nielsen". But as Claus Beck-Nielsen wasn’t the Leonardo of his time, but rather just an average wannabe, the mission of Das Beckwerk is not to develop the works of a genius. On the contrary the grand task of Das Beckwerk is to turn the failures of the average man into masterpieces, to turn the casual ideas of the wannabe into visions, to turn the life and story of the average human being into world history! (www.dasbeckwerk.com; English text quoted from the website.)

What immediately strikes one as the most spectacular aspect of clausbecknielsen.net’s The Suicide Mission [Selvmordsaktionen] (2005) is the fact that the events it depicts—two young European men crossing the border from Kuwait to Iraq during the early phase of the American invasion of Iraq in order to introduce democracy to the Iraqi people—have actually taken place. On foot, alone and carrying a box labelled “The Democracy” in their hands, the two Danes Nielsen and Rasmussen set out to literally introduce free debate, choice and dialogue into the war-torn country, not with the help of an army but with the help of mutual respect and constructive doubt. What makes the text interesting in relation to questions of source monitoring, decoupling, and the meaning and function of fiction, is the fact that a narrator situated in what apparently is the year 2026 frames this highly referential story. The very first sentence of the text emphasises the link between the two voices: “‘It was early in the millennium and hot’, Nielsen writes, ‘we were sweating, our ties were too tight and already Democracy felt like a burden’” (clausbeck-nielsen.net 2005: 7). Nielsen speaks out from inside quotation marks, placed by the other voice of the book, a voice located in Bagdad, the “cradle of the state of exception” in the year 22 of a new era. Using the scope syntax notation developed by Cosmides and Tooby, we might sketch the representational dilemma arising from this construction as follows. One dataset is anchored in the narrative of the Danish author and performance artist Claus Beck-Nielsen, who during 2004 in fact travelled to Iraq and wrote about his trip in lengthy articles that appeared in several leading Danish newspapers and magazines as journalistic correspondence. (1) Claus Beck-Nielsen reported that (2) he and Rasmussen entered occupied Iraq on foot. This dataset has Nielsen as its source and it claims to be, and for the most part indeed appears to be, true—that is, scope-free. The other dataset presented to the reader is channelled through a person from the future and clearly an s-representation, bound by a scope operator:

States of Exception

139

(1) In this fiction (2) a narrator from the year 2026 is telling the reader a story. What turns The Suicide Mission into a challenging and interesting case with regard to deciding who is speaking is the fact that this fictive metarepresentation frames the factual voice of Nielsen: (1) In this fiction (2) a narrator from the year 2026 is telling the reader a story about the writings of one Claus Beck-Nielsen. (3) whose scope-free story is then quoted at length. The degree to which this construction is unusual becomes evident when compared to a historical novel. In a traditional historical novel the events and persons depicted are like real events but, and this is the crucial point, every bit of information in the historical novel is scope regulated by the fiction-marker. Because of the source-anchoring, the information given in a historical novel, while most probably true, does not make any truthclaims. In The Suicide Mission we are presented with two competing narrative voices—one factual, one fictitious—whose internal relationships are puzzling, to say the least. The obvious place to look for a quick solution to this conundrum would be the framing of the fiction that frames the nonfiction: is this not all written by an author, and thus all different types of srepresentation? Are we not able to anchor it using the hierarchy below? (1) The author Claus Beck-Nielsen wrote (2) a fiction in which (3) a narrator from the year 2026 is telling the reader a story about the writings of one Claus Beck-Nielsen (4) whose scope-free story is then quoted at length. Yes and no: This is the place where the framing of the text in the project directed by Das Beckwerk becomes important. In the place where one would typically find the name of the author the reader finds an internet address in the form of an URL: www.clausbeck-nielsen.net. I shall return to this scope-regulating frame later; first, the goal is to investigate how the text uses the relationship between the two voices to construct its two main plot lines. The first plot line is told through metarepresentations of Nielsen’s writings in the form of diary entries, recovered notes, emails, blogs, and newspaper articles. The writings centre on Nielsen and Rasmussen’s dan-

140

Stefan Iversen

gerous and tenacious journey into the heart of a non-European culture, a heart eventually found by the Europeans to be if not dark then at least incomprehensible and opaque, a frightening nexus of unstable conflicts, undecodable intentions and values in a state of permanent flux: “Down here every value is changing. There is no order, nothing is stable, no-one is what you think he is.” Iraq anno 2004 is truly a state of exception.12 As the journey progresses so too does the deterioration of the psyches, identities and value systems of Nielsen and Rasmussen, and this plot line becomes a version of the well-known narrative of blindness and horror as the repressed foundations of modern western civilisation. Seeking enlightenment and rationality, the two Europeans end up engulfed and absorbed by the darkness their journey was supposed to have supplanted: “I swear that this will be the last time that I lie here, such humiliation, a skinny, fragile and very white body. A small body suddenly way too hot, a consciousness suddenly way too dizzy and on its way down into the darkness in order to sleep, sleep” (300). And thus, the state of exception expands from being a sociological condition to becoming a psychological condition, a fragile or exceptional state of mind. The second plot line relates to, and is related to us by, the voice from the future. This voice apparently represents one or more members of some sort of archaeological or historical society, situated in a culture radically different from that of Nielsen and Rasmussen. The aim of this second voice seems to be the reconstruction of the life of the mysterious traveller Nielsen by the use of carefully recovered documentation. At first glance, the voice from the future appears to be a sort of negation of the voice of Nielsen. In contrast to the latter’s massive epistemological uncertainty in the overdetermined world of an occupied Iraq, the voice of the future is characterised by considerable epistemological certainty in an underdetermined world of a future civilisation of which we are given very few details. In a tone far more sober and scientific than the ever more shaken and disturbed voice of Nielsen, the voice from the future uses footnotes to explain persons and events that in Nielsen’s presentation are taken to be known by the reader, such as Donald Rumsfeld (referred to by the voice as an “American poet from the 21st Century” (29) and Osama Bin Laden (referred to by the voice as a “media artist” (215)). As suggested by the quotes above, The Suicide Mission reverses the initial scope-regulated hierarchy between the factual and the fictive as well as 12

In “Colonized Thinking” (2008), Henrik Skov Nielsen focuses on some of the ethical and political as well as aesthetic conundrums set in motion by The Suicide Mission’s multilevelled treatment of the idea of a state of exception. He does so by placing the book in a very relevant dialogue with the notions of democracy and autoimmunity, drawing upon work done by Nicholas Royle, Jacques Derrida and Giorgio Agamben.

States of Exception

141

between the voices of the European preacher of democracy through dialogue on the one hand and the Middle Eastern archaeologist from the future on the other. As the two plot lines unfold, the non-fictitious and scope-free statements of Nielsen become the subject of an increasingly radical critique of the values they claim to represent. This critique is carried through by the fictive voice from the future, exercising considerable philological care in dismantling and decomposing the project Nielsen and Rasmussen attempt but eventually fail to complete. Rather than offering a true democratic alternative to the violent intrusion of the Americans, the Europeans turn out to be, according to the future scientist, “the worst possible disguised as the best possible” (359). Their reliance on dialogue, rationality and humanism turns out to be nothing but a veneer, covering their own lack of faith, and therefore, still according to the voice from the future, it was people like them who were first in line when what is referred to as “the decapitations” (358) began, i.e., when the world revolution leading to the current permanent state of exception took off. But the book’s experiment with decoupling mechanisms goes beyond promoting the fictive as the voice of insight and reason while denoting the non-fictive as the voice of self-deceiving dreams. A closer reading shows that the voice from the future is both more like and more different from the voice of Nielsen than one would expect, at least if one wants to describe both voices as homodiegetic narrators. First, the difference. It is an important aspect of the narration done by the voice from the future that it does not limit itself to quoting, explaining and commenting on the writings of Nielsen. At times it is able to penetrate the mind of Nielsen and present the reader with what seems like first-hand insight into the desires and beliefs of Nielsen: In an attempt to get rid of his doubt, Nielsen writes an article and uploads it to the Internet, hoping the newspaper will accept it, will free him of his doubt, will wrap the paper around his doubt and spread it across the nation in which it belongs (89) They had finally, or so he thought, reached the people, or at least the right people (103)

In passages such as these the voice from the future exercises powers that go beyond those of a homodiegetic narrator. It is not simply making educated guesses based on his readings of Nielsen’s texts; it is actually reading the mind of a person he has never met. As the use of free indirect speech in the second quote shows (the unwritten thoughts of Nielsen intermingle with the voice from the future), the voice from the future is privileged not only with regard to hindsight and epistemological supremacy but also with regard to access to the mental life of others, a choice of focalisation typically reserved for heterodiegetic narrators.

142

Stefan Iversen

In contrast to this widening of the difference between the two voices, we find passages in which the voice from the future seems to take on the properties of the voice of Nielsen. As previously mentioned, the tone and rhetoric of the voice from the future appears to be scientific and distanced, while the voice of Nielsen in its heavy use of tentative, ironicpolemical statements and formulations more resembles the prose of modern travel literature. However, this difference breaks down in several key passages of the text such as the following: […] in his temporary spot among the British in the barracks Nielsen turns his back on reality, hides on the Internet and continues to paint [maler videre paa] his picture of a crying child for the newspaper back home on the old continent (154)

This quote is filled with stylistic elements otherwise tied to the voice of Nielsen, most manifestly in the lack of Danish letters (“aa” in paa instead of “å”), a defining trait of Nielsen’s style, stuck as he is with typewriters and computers without the Danish letters æ, ø and å. The blending of the two voices has some important consequences for the reader’s attempt to locate the source of the writing. The experiment carried out with scope syntax cannot be boiled down to a fictive framing of a non-fictive narrative. The access to the mind of Nielsen partly fictionalises him while simultaneously the spill-over of Nielsen’s style onto the fictive narrator’s voice lends an element of factuality to the framing voice. The fictive framing contains the non-fictive, while the factual motif is fictionalised. This, then, points towards a third way in which the work incorporates and comments upon the notion of a state of exception, namely the fact that the text itself becomes such an exception. These experiments with a permanently unresolved narrative voice, oscillating between fiction and non-fiction, do not stop at the border of the text; they continue into or are enforced by the framing of the narrative. Where one would expect to find the name of an author, one finds a part of an internet address: clausbeck-nielsen.net. Upon visiting the web page www.clausbeck-nielsen.net one is confronted with a Flash animation, consisting of two images, one fading into the other. One image shows the earth as seen from space but rotated in an unusual way, at least to this viewer: the earth is centred on the Middle East in a way that makes the American continents almost invisible. The other image shows a photograph of a young man in a white tunic, relaxed but looking intensely at the camera. Most viewers from Denmark would recognise the young man as the artist Claus Beck-Nielsen. Two connections are established between the two images, with the first being the forementioned fade (the young man fades in as the Earth fades out; this pro-

States of Exception

143

cess is then reversed, removing the man) and the second being the use of red circles on both images (in the image of the Earth a red, slowly pulsating circle is placed on top of Iraq; in the image of the young man two such circles are placed in his eyes). What sense is to be made of this source of The Suicide Mission, located where the author would normally be? Not surprisingly, several contradictory readings seem possible. The combination of the satellite image and the red dot in Iraq cues military connotations, implying an operation or perhaps detonation of sorts. The latter reading is further motivated by the pulsating effect of the red dot, which seems to signal the taking place of an event so massive and/or disruptive that it is visible from outer space. Turning to the image of the man, the red dots may be read as traces from the photographer’s flash, thus lending the image an aura of amateurism and unedited reality. But the intense red gaze also evokes demonic beings and intentions, while on the other hand the combination of a buttoned but very loose white tunic and long hair seems to give the figure a Jesuslike quality. This effect is emphasised by the size of the man in white: he is only visible during a short period of time, but during that time he hovers over the entire Earth. As should be evident, many different interpretations can arise from the place where the scope of the information in the text ought to be anchored. In The Suicide Mission the difficulty of framing and regulating the hierarchy of levels of information continues past the threshold of the text, and the doubt regarding the source survives the text by escaping through the very hole normally filled out by the author. 5. Conclusion What can the theory and analysis of strange voices in narrative fiction learn from the concepts of decoupling and scope syntax as developed by and used in evolutionary psychology? This question has been the guiding force behind the work presented in this article, and the time has now come for some final remarks on the particulars and the perspectives of the approach. I have argued that work done by Cosmides and Tooby may be used to explain the effects of a source-questioning narrative such as The Suicide Mission, hovering between fiction and non-fiction as a text that in return may be used to reaffirm the crucial but so far partly untapped insight from this very theory with regard to the nature and function of fiction as not only a possible but also a necessary part of the ability of human intelligence to adapt.

144

Stefan Iversen

The reading showed how The Suicide Mission employs the notion of a state of exception on several different levels. With Phelan we might say that on a mimetic level the state of exception reflects an actual state, a state whose degree and kind of exception varies radically depending on the point of view taken: European, American or Iraqi. On a thematic level the state of exception refers to a state of mind, to the process of disillusion and estrangement experienced by the knights of rationality as the outer chaos surrounding them slowly creeps into and subverts their system of beliefs and values. On a synthetic level the state of exception refers to the state of the arts, or at least to the status of this particular work. Neither fiction nor nonfiction are apparently written by an URL, one may read the text as a systematic impairment of the decoupling function in that it rigorously complicates the reader’s separation of representations from metarepresentations and puts into doubt what pieces of information the system is supposed to store as false, true or fictive. One might and perhaps should add that such a reading gives too much credit to what appears to be the idea behind not only a work like The Suicide Mission but also the paradoxical project of Das Beckwerk. How is it possible, or even remotely sensible, to claim that one’s personal identity is no longer in existence, that it has “passed away” or “walked off”? Seen from the perspective of a literary critic or an art historian, one would be tempted to source-regulate experiments like this by the same measures as one would use for regulating performative avant-garde art. When confronted with such comments, the living entity formerly known as Claus Beck-Nielsen kindly informs questioners that his project should not be treated like a work of art, but as an intervention, as an attempt, if need be by the means of self-inflicted defeats, to move beyond the differences regulating our present life, the most important of those being the idea of identity. How one chooses to relate to this position, if it is a position, is probably a matter of temperament and personal preferences. Seen from the perspective of a narratologist who is eager to learn from cognitive sciences while still insistent on leaving room for what might adhere to different rules than those shaping the assumption of the current studies of cognition, the instability at work in The Suicide Mission is what makes it useful as a corrective to the, to my mind, too static and stable notion of fiction put forward both by Cosmides and Tooby and by the application of their work as carried through by Zunshine in Why We Read Fiction. Cosmides and Tooby claim that “Representations of [fictional] stories require hierarchical levels so that inferences are restricted in their scope of application to one level at a time” (92) and that “one would expect a fiction to remain decoupled […] insofar as source tags are an important part of a self-monitoring system” (92). This reliance on strict hier-

States of Exception

145

archical structures and secure scope regulations as conditions for fiction to function presents a simplified idea of what fiction is capable of, and it actually goes against their own idea of decoupling as a highly dynamic process. Works dealing with metalepsis and other types of scope-regulative short circuits are able to keep the process of source consideration open indefinitely. I want to emphasize that this fact does not run counter to their overall argument regarding the way human adaptation needs the ability to perform couplings and decouplings; rather, it fits very nicely with it. A voice always potentially begs the question “Who speaks?” Narrative fiction exists because of this, and does often exploit or play with it. This is not a mysterious effect but a simple consequence of how our treatment of information functions. The ability of the human mind to posit and treat information in states of exception, decoupled from secure scope-regulations, is not an option but a condition for the workings of human intelligence. References Alber, Jan, Stefan Iversen, Henrik Skov Nielsen and Brian Richardson 2010 “Unnatural Narratives, Unnatural Narratology: Beyond Mimetic Models”, Narrative 18.2, 113-136. Altes, Liesbeth Korthals 2006 “Voice, Irony and Ethos: The Paradoxical Elusiveness of Michel Houellebecq’s Polemic Writing in Les particules élémentaires”, in: Blödorn, Langer and Scheffel (eds), Stimme(n) im Text. Narratologische Positionsbestimmungen, Berlin, New York: Walter de Gruyter, 165-194. Baron-Cohen, Simon 1997 Mindblindness. An Essay on Autism and Theory of Mind, Cambridge, Mass.: MIT Press. clausbeck-nielsen.net 2005 Selvmordsaktionen. Beretningen om forsøget på at indføre demokratiet i Irak i året 2004 [The Suicide Mission. The Story of the Attempt to Introduce Democracy in Iraq in the Year 2004], Copenhagen: Gyldendal. Cosmides, Leda and John Tooby 2000 “Consider the Source: The Evolution of Adaptations for Decoupling and Metarepresentations”, in: Dan Sperber (ed.), Metarepresentations. A Multidisciplinary Perspective, New York: Oxford University Press, 53-116. Frith, Christopher D. 1992 The Cognitive Neuropsychology of Schizophrenia, Hove, Sussex: Lawrence Erlbaum. Genette, Gérard 1972 “Discours de récit”, in: Figures III, Paris: Seuil. 1980 Narrative Discourse. An Essay in Method, Ithaca: Cornell University Press. 1983 Nouveau discours du récit, Paris: Seuil. 1988 Narrative Discourse Revisited, Ithaca: Cornell University Press. Hutto, Daniel D. 2009 “Folk Psychology as Narrative Practice”, Journal of Consciousness Studies 16.6-8, 9-39.

146

Stefan Iversen

Iversen, Stefan [forthcoming] “Unnatural Minds”, in: Alber, Nielsen and Richardson (eds.), A Poetics of Unnatural Narrative. Leslie, A. M. 1987 “Pretense and Representation: The Origins of ‘Theory of Mind’”, Psychological Review 94, 412-426. Leudar, Ivan and Allan Costal 2009 Against Theory of Mind, Basingstoke: Palgrave Macmillan. Nielsen, Henrik Skov 2004 “The Impersonal Voice in First-Person Narrative Fiction”, Narrative 12, 133-50. 2008 “Colonized Thinking”, Oxford Literary Review 28, 117-132. [forthcoming] “Natural Authors, Unnatural Narration”, in: Alber and Fludernik, Postclassical Narratology. Approaches and Analyses, Ohio State University Press. Premack, D. and G. Woodruff. 1978 “Does the Chimpanzee have a Theory of Mind?”, Behavioural and Brain Sciences 1, 515526. Richardson, Brian 2006 Unnatural Voices: Extreme Narration in Modern and Contemporary Fiction, Columbus: Ohio State University Press. Sperber, Dan (ed.) 2000 Metarepresentations. A Multidisciplinary Perspective, New York: Oxford University Press. Zunshine, Lisa 2006 Why We Read Fiction: Theory of Mind and the Novel, Columbus: Ohio State University Press. Wildgen, Wolfgang and Barend van Heusden 2009 Metarepresentation, Self-Organization, and Art, Bern: Peter Lang.



ROLF REITAN (Aarhus)

Theorizing Second-Person Narratives: A Backwater Project? Prototypical second-person narratives like Michel Butor’s La Modification or Georges Perec’s Un homme qui dort are ridden with at least two intertwined paradoxes and two probably interdependent ambiguities. First paradox, and first ambiguity: “you”—a pronoun of address—is used to refer to and perhaps to address a protagonist who obviously does not know he is being referred to and does not hear, if he is being addressed. Second paradox, same ambiguity: the flagrant fictionality of opening in medias res with an immediate reference, and perhaps address, to a “you” that would seem to be created by the same reference and eventual “address”. Second ambiguity: although you are not in any traditional sense addressed as a reader, you may nonetheless feel addressed by the initial “you” (vous, tu), in which case we—perhaps—would have a “double address”: You have put your left foot on the grooved brass sill, and you try in vain with your right shoulder to push the sliding door a little wider open. (Butor 1959: 1)1 As soon as you close your eyes, the adventure of sleep begins. In the familiar half-light of the room, dark volume broken by details, where your memory easily identify the paths you have followed a thousand times, […] (Perec 1990: 133)2

In 1958, Roland Barthes, refusing the opinion that Butor’s use of “you” is “a formal artifice, a smart variation of the novel’s third person that should be credited with the ‘avant-garde’”, enthusiastically exclaimed: “this constant use of vous [ce vouvoiement] appears to me literal: it is that of the creator to the created, named, constituted, created through all its actions by a judge and begetter.” (Barthes 1964: 103) But then, this was before narratology. Some thirty years (and quite a few new or rediscovered second person narratives) later, a period of near silence was succeeded by some slowly growing academic interest, which then erupted, as it were, in a famous issue of Style (1994, 28.3) dedicated to the study of second-person narration and edited by Monika Fludernik. After 1994, however, productive in1 2

Jean Steward’s translation slightly modified. Andrew Leak’s translation slightly modified.

148

Rolf Reitan

terest disappeared again. Today, despite a few attempts at restarting a discussion,3 one might get the impression that all had already been said in the first half of the ‘90’s and the topic therefore not worth pursuing any further. This, however, is inaccurate. The discussion had only just started when it lost its impetus, the reason being, I think, that different positions never was brought to a real confrontation with one another. Authors continued pursuing their own course, at times more or less repeating earlier propositions. Discussions froze, development halted. In this paper, I propose a closer look at three very different perspectives on second-person narratives.4 Brian Richardson (1991), Irene Kacandes (1994), and Monika Fludernik (1993, 1994b) have been classical references for some time, but they have never, as far as I know, been seriously discussed. I shall start by discussing Kacandes’ intriguing concept of ‘radical narrative apostrophe’, and next discuss the three authors’ attempts to find some order in the field, i.e. their, again very different, typological proposals. Especially the analysis of Fludernik’s famous “Diagram 1” will take some time and effort. Then, borrowing Richardson’s idea of a Standard Form of second person narration, I return to Butor’s La modification to investigate the question of address (a pivotal question in Fludernik’s articles), which then leads to a strict definition of a prototypical “genre” of Standard Form narratives. Passing through landscapes of fiction, apostrophe and postmodernism, I address some tricky questions concerning self-address and Uri Margolin’s analytic formulas. At last, by way of proposing a much needed subdivision of the Standard Form, I discuss the strange narrating voice in La modification: not a narratorial voice, but a readerly voice created in the author’s writing. A Rhetorical Approach: Apostrophe Kacandes, proclaiming that “narrating with the second-person pronoun is a rhetorical act”, noted that Barthes’ allusion to Genesis “accurately parallels the bizarre narrative structure of La modification: an absent speaker whose words have the power to create (by addressing) a tangible, complex narratee, but one who cannot talk back”. Butor’s narrative mode in La modification thus exemplifies a remodelled classical apostrophic structure: The block to communication between addresser and addressee in La modification is even more profound than that of traditional apostrophe since the addresser 3 4

See Schofield 1998, DelConte 2003. I am grateful to Henrik Skov Nielsen and Per Krogh Hansen for perceptive and extremely helpful comments on the first draft of this paper, and to Monika Fludernik for reading the second draft and respond to it with courteous and encouraging comments.

Theorizing Second-Person Narratives

149

speaks from an unknown place and does not speak of him/herself. I propose we call it and other second-person texts like it “radical narrative apostrophe.” (9)5

The radicality of this apostrophe, we understand, is due to the “absent speaker”. Had Butor’s speaker here been more than a voice, we would perhaps have something like a simple narrative apostrophe. The figure of apostrophe was “used by rhetoricians to describe the act of an orator turning away (Gk. apo ›away‹ and strophein ›to turn‹) from his normal audience, the judges, to address another: whether his adversary, a specific member of the jury, someone absent or dead, or even an abstract concept or an inanimate object” (2). But Butor is not first addressing his normal audience and then addressing an absent character. From the beginning— in medias res, as it were—he addresses the protagonist, and he does so in a way that would seem to simultaneously address each member of his audience, his readers (and judges). Clearly, the concept of apostrophe somehow will have to be remade to be applicable to cases of radical narrative apostrophe. Kacandes, leaning on communication theory (according to which interchange consists of addresser, message, and addressee, and where addresser and addressee “normally switches roles”) proposes to “explain the structure of apostrophe in terms unavailable to ancient rhetoricians by describing that communicative circuit on which the figure [of apostrophe] tropes.6 Thus, apostrophe is on the one hand ‘short-circuited’ communication; messages do not flow in both directions. On the other hand apostrophe bears two addresses: Overtly, a speaker sends a message to someone or something as if that being or thing could respond but will not. Covertly, an apostrophe is meant to provoke response through its reception in a second(ary) communicative circuit, received by the readers of a poem in the case of lyric or the audience in the case of oratory. That is to say, the convention of apostrophe is to differentiate between explicit addressee and receiver-audience, between the referent of the “you” and the “listener”. (3)

This might be an acceptable reformulation of the classical concept of apostrophe in lyric and oratory and a perfect description of what can be regarded as the constitutional apostrophe in second person narrative, “fiction that employs a pronoun of address in reference to a fictional protagonist”.7 However, two aspects of second person narration now seem to have descended below Kacandes’ horizon: (1) The idea of creating by addres5

Kacandes’ text will here be referred to in an html version; page numbers may therefore deviate from the Style version. 6 Kacandes borrows this expression from Jonathan Culler’s dictum that apostrophe is different from other tropes “in that it makes its point in troping not on the meaning of a word but on the circuit of communication itself”. (“Apostrophe”, in Culler 1992: 135) 7 This is Monika Fludernik’s definition—in Kacandes’ opinion “the clearest definition” of second person narration (1, note 2). I shall later discuss the merits of this clarity.

150

Rolf Reitan

sing has been substituted by the less spectacular idea of sending a message to someone or something as if the addressee could but will not respond— as if Kacandes’ own discourse, in order to better relocate itself within the realm of rhetoric, has decided to shed the touch of poetics borrowed from the early Barthes’ enthusiastic allusion to Genesis. And (2) the idea of aborted communication circuits has no “slot” for the observation that “the pronoun of address” not only invites the reader to listen to and “witness” the events, but also lures him or her to slip into the position of the protagonist (i.e. the double address function)—and even perhaps into the position of the addressing voice itself (after all, the reader will be “performing” this voice as he reads).8 At the end of Kacandes’ text, however, we find the following solution—and a very interesting, slippery consequence: In listening as if one were the direct addressee, the desired other of the narrator, one is positioned to feel the emotional force of the relationship that is created. As secondary listener, one is moved by witnessing the apostrophe’s animation and is persuaded to the narrator’s view of the relationship: it did exist. Furthermore, the narrative mode invites slippage between these two positions, thus eliciting the potential involvement of anyone who will take up this available pronoun, anyone who will consider her/himself addressed by the narrator’s “you” (20).

With Kacandes’ proposal we have a theory that takes care of our first paradox, radical narrative apostrophe being its solution, and it almost takes care of the paradox of fiction, if only to forget it. Moreover, with the concept of ‘slippage’ it even points to the possible problem of double address. However, not all second person narratives use an apostrophic mode of narration. Kacandes proposes “a categorization along a spectrum of reversibility”. Reversibility designates “the ability of an addresser and an addressee actually to change positions, to be in turn speakers and listeners” (1994: 9-10). The list below shows how Kacandes (9-12) imagines a preliminary realization of a typology according to the principle of reversibility. It consists of seven “steps” on a scale between the dialogic and the apostrophic poles: Dialogic pole · Quoted dialogue > Self-address > Epistolary fiction > Reader address > Traditional apostrophe > Radical narrative apostrophe > Generalized “you” (“one”) · Apostrophic pole

The mixture of types or genres of fiction, rhetorical figures, narrative techniques or stylistic devices seems to belong to rather different scales, even if almost everything on the list may occur in second person narratives. A main fault with this list is that it fails to distinguish between types of narra8

Cf. Kacandes 1993. Françoise van Rossum-Guyon (1970) found after several re-readings of La modification that it is the “narrating” voice and not the protagonist that gains a lasting interest.

Theorizing Second-Person Narratives

151

tives and the wide variation of second person elements that may be used in narratives. Whatever consistency a scale like this may have, it is not what we need if we want to be able to find our way in the varied landscape of second person narratives. Both Richardson and Fludernik, however, have attempted to be more helpful. Let us try to follow their lead. Mapping the Area: Typologies Brian Richardson (1991) was one of the very first to present a comprehensive map of the enormously varied landscape of second person narratives. His mapping method is an elegantly executed inductive one, open, pedagogically informative and without any specific theoretical hang-ups. Starting by delimiting the field (“any narration that designates its protagonist by a second person pronoun”9), he first finds “the most common type”, which is also the form “closest to more traditional forms of narration”. It earns the name of The Standard Form: In it, a story is told, usually in the present tense, about a single protagonist who is referred to in the second person; the “you” also designates the narrator10 and the narratee as well, though […] there is frequently some slippage in this unusual triumvirate. This is the form in La Modification, Aura, Un homme qui dort, A Pagan Place [a past tense narrative], and Jay McInerney’s Bright Lights, Big City.” (311)

Next, two other forms are discerned by being spotted as deviations from the standard norm: The Subjunctive Form, renamed The Hypothetical Form in Richardson (2006), developed in narratives like the second person stories in Lorrie Moore’s Self Help”, written in the style of the user’s manual or the self-help guide”. It is characterized by a “consistent use of imperatives, frequent employment of the future tense, and a strong distinction between the narrator and the narratee” (319). The Autotelic Form, paradigmatically represented by Italo Calvino’s If on a winter’s night a traveller, is defined by the narrator’s “direct address to a ‘you’ that is at times the actual reader of the text and whose story is juxtaposed to, and can merge with, the characters of the fiction”. Its “unique and most compelling feature” is “the ever-shifting referent of the ‘you’ that is continually 9

This definition excludes narratives that are not second person, although they use the second person pronoun, as in the common practice (from Fielding to George Eliot) of direct address to the “Dear reader”, or “the monologue addressed to a real or imaginary homodiegetic audience” as in Camus’ La Chute or Hawkes’ Travesty (310). 10 Before Richardson, Darlene Marie Hantzis (1988: 47) wrote: “The second person narrator is present when the ‘you’ constitutes the narrator as well as the actant and the narratee(s) of the text”. However, “you” designating the narrator will normally mean that the narrative is a “self-address”. Later in his article, in a comment on McInerney’s Bright Lights, Big City, Richardson makes a decision: “the protagonist, whom I will call the narrator” (315).

152

Rolf Reitan

addressed”. As in the subjunctive form narrator and narratee are kept “quite distinct”, and like the standard form it is primarily told in the present tense. (321) In her two very rich investigations of second person narrative, Monika Fludernik (1993, 1994a and 1994b)11 proposed to map the area by building on and developing narratological concepts, primarily Stanzel’s and Genette’s, and in the process also attempting a negotiation between these giants’ somewhat incompatible methodologies. “Narrative You” opens with a criticism of previous research for “ignoring the central issue” of the combination of the address and reference function of the pronoun of address. A comment on Kacandes’ approach in Kacandes (1990) suggests some dissatisfaction with a “mere rhetorical or apostrophic” approach (218); Fludernik’s approach is, in a first step, to consider the “structural” partition of the field that ensues from the three possible combinations of address and reference. The protagonist is (A) addressed but not referred to, or (B) both addressed and referred to, or (C) referred to, but not addressed. A. Narrative “you” as address, but not referring to protagonist The address function can be seen either by “an explicit address-you or by means of imperatives”. It requires “the positing of an enunciatory instance who may be explicit (by means of a narratorial I) or implicit (but not merely ‘covert’,” and the addressee is either a generalized you or an extra-diegetic narratee. We find this in a great number of “teller” narratives, which have an address function, and it is particularly frequent in the pseudo-oral skaz type narrative12 (e.g. in the third person present tense sections in Gloria Naylor’s Mama Day which have a persistent allocutory you), the dramatic monologue (e.g. Hawkes’s Travesty), or epistolary narrative (“Narrative You”, 221). However, occurrence of allocutive you in third person narratives, dramatic monologues or epistolary novels—do obviously not qualify these as second person narratives, and Fludernik does not return to this “possibility”.13 Although delivering an instructive lesson, category A turns out to be a “structural” abortion of sorts. (Kacandes actually could have learned something from this.)

11 12

I shall refer to these articles as respectively “Narrative You” and “Test Case”. Pseudo-oral types of narrative (skaz) usually have a teller figure that personally knows the protagonists or lives in the same town or village and knows the stories from hearsay, “thus allowing for a very fuzzy demarcation line between the world of the teller and the world of the told.” (“Test Case”, 474, note 27) 13 Hawkes’ Travesty reoccurs later (238) in a short discussion of dramatic monologue.

Theorizing Second-Person Narratives

153

B. Narrative you as both address and reference to protagonist “Under these circumstances the addressee is an intradiegetic narratee, but not in the well-known ›metaleptic‹ mode [ … ] where the extradiegetic narrator playfully addresses a character”. The most common case in this category is said to be “that of the narrator and narratee sharing both a presence of interlocution on the enunciatory plane and an existential past on the story plane.” As examples Fludernik points to the second person strand in Naylor’s Mama Day and Günter Grass’ Katz und Maus. (“Narrative You”, 222.) She does not, however, pay attention to the fact that characters addressed by the second person pronoun in these novels are— at the time of narration—either (probably) dead or plainly imagined; i.e. that they are (partly) apostrophically produced characters. C. Narrative you referring to protagonist, but not used as address In this category the address and the referential function are disentangled; in reference to the protagonist “you” is not used as address, there is no observable addresser, “although there may be an omniscient authorial narrator divorced from the fictional you”. Here Fludernik finds a couple of novels from Richardson’s Standard Form: Besides this “authorial” type which consistently leans towards the figural mode (cf. Butor’s La Modification and Peter Bowman’s Beach Red), the most common second person text is the one where the fictional you predominates as an experiencing self in what I would like to call—after Stanzel—a reflective mode narrative. […] Recent examples include McInerney’s Bright Lights, Big City (1984) or the Diego strand in Virgil Suarez’s Latin Jazz. (“Narrative You”, 222)

Summing up: Category A, lacking substantial referents for the second person pronoun, does not qualify as a second person narrative category. Category B would seem to contain only homodiegetic narratives in which second person narration occurs apostrophically. Only category C covers proper second person narratives, but would seem to cover too much: it contains two groups of second person narratives—one belonging to the so-called authorial-figural continuum (though with a “consistent leaning toward the figural mode”), while the other is to be found at the very end of this scale. This is the group of reflectoral narratives, which bear no trace of an authorial presence. Actually, the main bulk, “the most common” of second person narratives would seem to belong here. We are left, then, with two categories, but in reality with three “categorical” groups. The reason for this anomaly is however clear. The introduction of Stanzel’s “authorial-figural” scale between the situations of “teller” and “reflectoral” narratives complicates the original structural partition. Fludernik’s next step is to use the results from the structural exercise in an attempt to find “slots” for second person narratives in a revised version of

154

Rolf Reitan

Stanzel’s three-polar typological circle for narrative situations (in which the “poles” are authorial narration, reflectoral narration and I-narrative). This daunting task requires some terminological innovations: (1) A distinction between Narrative with Communicative Level and Narrative with No communicational level, and (2) (within the area of communicative narrative) a partition in homo- and heterocommunicative areas, and (3) within the homocommunicative area a distinction between homodiegetic and homoconative narratives. Using this new terminology, Fludernik generates the awesome “Diagram 1”, reproduced somewhat simplified in the diagram below: (A)

Narrative with a Communicative Level (Teller Mode) Heterocommunicative narrative

Homocommunicative narrative I + you

I + you

You

Homodiegetic Peripheral homo- (i.e. auto)diegetic narrative with you protagonist: Grass, Katz und Maus

(B)

Both narrator and addressee share realms of existence: we narratives; White, Nocturnes for the King of Naples

You

You

he, she

Heterocommunicative you; you only protagonist, not addressee: Butor, La modification

Narrational level existentially divorced from story level: authorial and authorialfigural third person, Fielding’s Tom Jones, Joyce’s Portrait of the Artist as a Young Man

Homoconative Addressee as character, (narrator only explicit or implicit address function): Farley, “House of Ecstasy”

Authorialfigural continuum possible, too: Calvino, If on a Winter’s Night a Traveller

Narrative with No Communicational Level (reflector narrative and neutral narrative) e.g. Ernest J. Gaines, “The Sky is Gray” (first person); Jean Muno, Le Joker; McInerney, Bright Lights, Big City (second person); Faulkner, “Was” (in Go Down Moses), third person.

We immediately see how the diagram, even in this simplified reproduction,14 catches the logic of the two remaining categories B and C from the 14

In Fludernik’s full diagram (“Narrative You”, 225 and “Test Case”, 447) the homocommunicative area stretches out to the left and covers two slots: one for autodiegesis (indexed “I”) and at the extreme left a slot indexed “I + he/she” for “peripheral homodiegetic (first person narrative) (including we narratives of exclusive we), Mann, Doctor Faustus.” In addition, a slot indexed “you + he/she” is inserted at the right end of the homocommunicative scale. It is ex-

Theorizing Second-Person Narratives

155

structural exercise: First, the higher-level distinction between (A) communicative and (B) non-communicative narrative catches the unexplained division of category C in two different groups. Second, the main distinction between narratives using the pronoun of address both as address and as reference to the protagonist, and narratives which use it only as reference, reoccurs in the diagram as the distinction between homo- and heterocommunicative narrative (a heterocommunicative narrator addressing the protagonist of his narrative will inevitably produce some kind of metafictional play and not a “real” address).15 We may also note that the diagram’s four slots for homocommunicative narrative easily can be reduced to two. Katz und Maus and Nocturnes for the King of Naples are both, despite interesting differences I cannot discuss here, “peripheral homodiegetic narratives with you protagonist” in which “narrator and addressee share realms of existence” and a story in the past. Moreover, in both novels the second person parts of the narrative is conducted apostrophically.16 The two “homoconative” texts (Ralph Milne Farley’s horror-and-fun story from 1938, “House of Ecstasy”, and Calvino’s On a Winter’s Night a Traveller) are both of the variety that Richardson separated from his Standard Form and called autotelic narratives: In both texts the “authorial-figural continuum [is not only] possible, too”, it is realized. Fludernik’s slot-text “addressee as character (narrator ex- or implicit address function only)” points to, but does not explain, the essential aspect of these texts (which, in Genette’s terminology, are heterodiegetic narratives): not only do they open with a direct address to the reader, they also—and this is the essential point—explore the ambiguity of the addressing “you”

plained thus: “peripheral you (in relation to a third person protagonist) not found, but possible in principle”. 15 A typical exception to this rule (if rule there is) can be found in Norwegian author Erik Fosnes Hansen (1990) Salme ved reisens slutt (Psalm at the End of the Journey), a heterodiegetic novel with some inlaid apostrophes to several of its different protagonists. (This intarsio-technique being perhaps reminiscent of German expressionism.) 16 Among the many consequences of this, I shall here only observe the following: The “second person sections” in these novels, being of various length (stretching from short sentences up to a couple of pages at most) do not constitute a narrative by themselves; they intermittently accompany or interrupt the homodiegetic narrative, at times (but not often) relaying extra narrative information. These novels are therefore not really “I + you” narratives; rather, they are 1st person narratives with scattered (though thematically important) apostrophes to the protagonists. One main difference between these novels is that in Nocturnes the protagonist is referred to only apostrophically (though with one important exception: the message about the protagonist’s death). This would seem to be very different from Fludernik’s main example in “Test Case”, Oriana Fallaci’s Un uomo, in which a peripheral homodiegetic narrator tells the dead protagonist’s story in a continued second person narration (the whole narrative thus being perhaps “a simple narrative apostrophe”).

156

Rolf Reitan

to “make” the initially addressed reader be the addressed protagonist.17 With this pronominal trick the autotelic text circumvents the “ontological” impossibility of a heterodiegetic narrator addressing the protagonist of his story. (Fludernik does however not show any interest in this aspect. Her focus is on the “homoconative” aspect, i.e. the narrator’s effort to impose a role on the addressee. The narrator thus shares with the addressee a “presence of interlocution on the enunciatory plane”.) A compacted paragraph presents a general explanation of the diagram, emphasizing the use of Stanzel’s scalar continua: It should be observed in the diagram that, as in Stanzel’s typology, the individual categories are arranged in open scales between peripheral and central involvement of narrators and narratees in the stories in which they participate, and between the homo- and heterocommunicative modes of narration. The diagram illustrates a scale of forms between homocommunication and heterocommunication. There are intermediate areas between I and you, between you and he or she, and between I and she or he, so the diagram should be read as connecting up at its right and left margins. (224)

The use of between here obviously points to areas between pure, or better: simple forms conceived as stations on a scalar trajectory. For example: “Between the purely homo- (and auto-)-diegetic and the purely homoconative mode (the addressee but not the narrator participate in the story) one can locate several combinations of which the most common is we-narrative, as in Grass’s Katz und Maus.” (224) In this case, the “impure” intermediate area gets populated with “both-and” narratives, i.e. narratives that in different degrees combine defining characteristics of the two adjacent simpleform narratives. Or, as in another example: “between homocommunication and heterocommunication” indicates the area between purely homodiegetic and purely heterocommunicative narration18—i.e. the area covered by the first four slots in the simplified diagram. Likewise, the italicized pronouns in the last sentence of the quote refer to simple forms, way stations on the most comprehensive scale in the diagram. We see that the 17

A simple demonstration of this autotelic trick device in action is the opening paragraph of Farley’s short story: “This actually happened to you. And when I say “you,” I mean you—now reading these very words. For I know something about you—something deeply personal— something which, however, I am afraid you have forgotten.” (Farley 1988: 3) The narrator then continues by telling “you” his forgotten story in the past tense and with some vague, enigmatic allusions to how he (the narrator) may somehow have been an evil part of the story, in which case the narrator and the protagonist would have had “a shared past”. This, of course, gives the narrative a definite, but undecidable, leaning towards the homodiegetic end of the scale. 18 “Heterocommunicative” lies within the semantic range of Genette’s “heterodiegetic” (but not the other way around, since also the homoconative area is one for heterodiegetic narration). Fludernik prefers -communicative (before -diegetic) because second person narrative “puts the narratee on the agenda”.

Theorizing Second-Person Narratives

157

concept of scale is eminently elastic; any part of a scale can itself be a scale provided it is extended between two simple-form way stations. Another aspect concerning scales is an effect of the category “Narratives with No Communicational Level” being positioned below the “slotted” part of the diagram. Here it is not indexed with “you” since it comprises all kinds of reflectoral and neutral narrative. Fludernik explains why there should not be a separate slot for second person reflectoral narrative: In purely reflectoral texts pronominal distinctions come to be of minor importance and lose their deictic significance. References to the protagonist cover up for the subjectivity of an underlying deictic centre, and—in the absence of a narratorial standpoint—the protagonist’s I can be referred to also as a she, he or you, without such a renaming effecting any considerable change in the reader’s apperception of the story (227).

We see here a precise topical convergence with an effect of Genette’s difference between mode and voice: Reflector mode narration (internal focalization) is possible in homo- as well as heterodiegetic narration; it does not require a specific vocal choice. But choosing a pronoun of address to refer to the protagonist is neither a vocal nor a modal choice. A pronoun of address shares with heterodiegetic narration (so-called third person or he/she narration) the possibility of being used in both authorial and strictly reflectoral narration. Third-person narration must not be confused with authorial narration and is not a specific narrative situation in Stanzel’s typology, and neither can second-person narration be.19 However, having separated reflectoral narration from the slot area, thereby making the term “he/she narration” point to the authorial narrative situation as its only principal reference, Fludernik now seems to play with the idea of grafting second person narration on to the scale of narrative situations as a principally simple-form way station on a par with Inarration and authorial narration—and if simple-form second person is not a narrative situation, it might conceivably figure as one of three principal nodes marking a reborn, as it were, typological circle. In the full diagram, homocommunicative second person narration is situated between homo- and heterodiegetic narration with intermediate forms on both sides, and we should get this series: / I / I + you / you / you + he or she / he, she /

The hypothetical slot indexed as “you + he or she” is needed as an intermediate area between purely second person and the purely heterommuni19

The separation of reflectoral narrative from the slot area also suggests a complex (but not developed) scalarity in three dimensions: the “(authorial-)figurative” scale now moves vertically on the diagrammed scales, and one could easily imagine reflectoral narrative as foot pieces in the slots, though only simple second person narrative, i.e. narratives that are not combinations, will permit that the addressed protagonist be rigorously focalized.

158

Rolf Reitan

cative area. However, the series breaks off at this point. The next slot should have been “he/she”-narrative, but instead we get another you-indexed slot for “heterocommunicative you”, instantiated with Butor’s La Modification, after which follows the deferred “he/she” slot. Instead of one continuous scale we have got two separate scales, one for homocommunicative (address you + referential you) and one for heterocommunicative narrative (you as reference only). This fracture could however easily be mended by moving the heterocommunicative you slot back to the middle position in the series. This would of course force a removal of the “homoconative” slot for trick-narratives of the autotelic type, arguably a small price to pay for a perfectly continuous scale. Fludernik does not consider this possibility, and the reason seems clear: The overriding concern is to preserve the internal continuity of homo- and heterocommunicative scales respectively, even at the cost of giving up the idea of a neat, continuous circle of narrative types.20 Not explicitly admitting this cost, Fludernik makes an excuse: “If the schema here presented seems overly typological, I wish to counter that the proposal is merely an attempt to document the large variety of second person fiction and to present a useful terminology for future discussion” (226).21 We have seen that the homocommunicative area of the diagram—the area in which the pronoun of address combines address and reference— leaves us with only two very special groups: (1) “I and you narratives”: homodiegetic narratives with apostrophic addresses to the protagonist (or at least a main character), and (2) a handful of heterodiegetic narratives in which the otherwise impossible address to the protagonist is made possible by the trick of an initial direct address to the reader. For what reason should we accept the first group as properly second person narratives? Let us take a closer look at Fludernik and Richardson’s very look-alike definitions. Richardson proposes “any narration that designates its protagonist by a second person pronoun”. Fludernik’s definition is more elastic: “fiction 20

One may remember how the strange heterogeneity in Stanzel’s principles for constructing the fabulous typological circle appealed to Genette’s humour, he could not help see the circular form and the preference for the number 3 as emblematic for Pre-Enlightenment forms of sensibility; the circle with its axes and poles and all kinds of narratives sitting on their respective branches made him think of Gothic window roses. Fludernik, in correcting Stanzels principles, but still trying somehow to reproduce the intuitivity and easy malleability of the Typenkreis, has to accept loosing the charm and naïve simplicity of its graphically seductive rhetoric. 21 Diagram 1 is reproduced in its entirety in Fludernik 1966, though with a graphical addition pointing to figural narrative as a vertical scale (represented by a small double-arrowed vertical line on the lower left side, meant to cover the area between “communicative” and “noncommunicative” narrative). The diagram’s function in the book is obviously to graphically represent the applicability of terminology invented in “Narrative You”.

Theorizing Second-Person Narratives

159

that employs a pronoun of address in reference to a fictional protagonist”. This in fact covers the occurrence of the second person pronoun in Katz und Maus—a homodiegetic narrative that only uses the second person in apostrophic addresses and not in the narrative proper (which refers to the protagonist in third person).22 Richardson’s “any narration”, however, points to a use of “you” that abides by the very same rules that govern the use of “I” and “he” or “she” in narrative reference to a character in action. Apostrophes are therefore not accepted as second person narration (though with the possible exception of Un uomo), and both Grass’ and White’s novels will have to be excluded from the field. On the other hand, autotelic narrative, which is covered by Richardson’s definition, deviates from the Standard Form in not following rules common for homo- and heterodiegetic narration and consequentially becomes labelled as a special case (as is the subjunctive form of “self-help” texts). The result of this very fine piece of induction is an astonishingly fresh and very simple map of the area, and one that is not obliged to somehow fit into the experimental straitjacket of preconceived typological schemes. In the last resort, this definitional difference might reflect different methodological orientations. To put it schematically: Both definitions try to catch some essential empirical criterion, in fact, they would seem to use the same criterion; but while Fludernik’s formula probably interprets it “linguistically” and qualifies any narrative part (of some length) as second person in so far as it refers to the protagonist by a pronoun of address— the idea of a “standard form” would work against the grain of this perspective. Richardson in principle prefers a more “generic” interpretation, which only qualifies a narrative as second person in so far as its use of a pronoun of address by itself constitutes a whole story. Fludernik’s approach opens a vast and complex area of “fuzzy”, sliding and overlapping phenomena. Richardson’s approach cuts out, as a first step, a smaller, and “cleaner” area, and one more accessible for investigation.23 For example, when Richardson does not distinguish between homo- and heterocommuni22

In an Introduction to this issue of Style, Fludernik (1994a) pointedly criticizes some main theorist’s different attempts at defining second person narrative. This discussion ends by juxtaposing Uri Margolin’s elaborate definition (in Margolin 1994) to Richardson’s “viewpoint”, which she has “somewhat endorsed” since his “formulation allows for second-person fiction without a narrator’s “I” and without an (extradiegetic) narratee”. (287) Here, in proposing as preliminary definition “narrative whose (main) protagonist is referred to by means of an address pronoun” (which is almost identical with Richardson), Fludernik adds an elasticity clause: “[I] add that second-person texts frequently also have an explicit communicative level on which a narrator (speaker) tells the story of the “you” to (sometimes) the “you” protagonists’ present-day absent or dead, wiser, self.” (288) The amendment declares apostrophes to the protagonist to be legitimate under the proposed definition. 23 This, of course, is my interpretation of Richardson’s approach.

160

Rolf Reitan

cative narratives, it is because he has no immediate need to: the definition of the field principally excludes homodiegetic narratives, and the special case of autotelic narrative (the “homoconative” slots) takes care of what is left of the homocommunicative area. Moreover, while Fludernik’s diagram gets absorbed (as it seems) in the homocommunicative, simultaneity of address and reference field and leaves second person “heterocommunicative” narrative standing alone on only a vertical, figural scale, Richardson’s outline, not bothering about scales at all, singles out a Standard Form, i.e. the larger group of proper second person narratives, for further investigation. I shall try to point out some questions that yet need to be addressed in such an investigation, the question of address being perhaps the most important; it will help us make some useful distinctions among the vast group of standard form second person narratives. The Standard Form: Focalization, and the Question of Address Butor’s prototypical, “standard form” novel La Modification is the prime example of a “heterocommunicative” second person narrative in “Diagram 1”. It is here characterized somewhat hesitatingly as a narrative that “consistently leans towards the figural mode.”24 In “Test Case”, we find this more exhaustive description: La modification [ …] has neither a narrator's experiencing self (i.e., the narrative is heterodiegetic in Genette's terms) nor a ›conative self‹ for the addressee (i.e., there are no distinctly allocutive or exhortatory clauses, the “vous” is not being “talked to”). One does, however, have ample exposure to the “you” protagonist's narrated experiencing self, and there are some muted indications of a narratorial (omniscient) frame in the background. (“Test Case”, 449)

Unfortunately, Fludernik never explains what she means by these ‘muted indications of a narratorial and even omniscient frame’. Why should La Modification not be read as a consistently focalized second person novel? In “Test Case” one may perhaps find at least an indirect answer. Commenting on a quote from the “first Diego section” in Suarez’ Latin Jazz Fludernik explains: From Diego’s impressions of what he sees, we move straight into his thoughts (rendered in free indirect discourse) […]. One is therefore inclined to continue with a reading in internal focalization, taking the flashbacks as Diego's memory [ … ]. Because of the present tense, the ‘you’ narrative loses a good deal of its narratorial and allocutive qualities, backgrounding, that is, both the narrator's func-

24

Since the figural mode is said to be a continuum between teller and reflectoral mode, it is likely that Fludernik’s “leaning” is meant to be towards the reflectoral mode.

Theorizing Second-Person Narratives

161

tion of narration (which usually consists in a telling after the fact) and the effect of address or allocution inherent in the use of you (451).

This does not concern pure reflector-mode but narratives which have to be situated on the figural scale, perhaps with a consistent leaning towards the reflectoral end: The quoted passage from Suarez’ novel “still has some hints of extradiegetic management”, and a passage from McInerney’s Bright Lights, Big City that “clearly concentrates on the protagonist’s flow of experience”, “still preserves a residual perspective of zero focalization”—”owing to the many verbs of consciousness” (451).25 Only a passage from one of Joyce Carol Oates’s wonderful short second person texts is said to have “a fully developed” reflector mode, because here “the narrative disappears entirely behind the thoughts of the protagonist”. Would this observation perhaps be a reason for Fludernik’s spectral hints at narratorial management also in La modification—i.e. a puristic definition of the reflectoral mode or internal focalization that inevitably will leave the greater part of internal focalizations with some residue of “narratorial” activity and push them out onto the figural scale? Although perhaps linguistically consistent, this would not, to say the least, be very practical in narratological analysis. Genette, as one will know, very sensibly decided “to take the term ‘internal focalization’ in a necessarily less strict sense” and use Barthes “rewriting criterion” as a practical definition.26 Likewise, when observing some of Proust’s rigorous focalizations, Genette only understood these as being without any retrospective inference from an older and presumptively wiser narrator’s “enunciatory” position.27 Of course, in Butor’s internally focalized novel the narrative does not entirely disappear behind the protagonist’s thoughts, the novel is a narrative of the protagonist’s thoughts. Moreover, starting with the second paragraph and incessantly repeated throughout the novel we find fragments of what might be read as narratorial comments. At times, they may give the impression of a narrator trying to engage the protagonist in a dialogical exchange: You edge your way in through the narrow opening, then you lift up your suitcase of bottle-green grained leather, the smallish suitcase of a man used to making long journeys, grasping the sticky handle with fingers that are hot from having carried even so light a weight so far, and you feel the muscles and tendons tense not only in your finger joints, the palm of your hand, your wrist and your arm, but in your

25

This surprising remark would seem to presuppose that some narratorial deixis had already been established in the narrative. “Verbs of consciousness” would otherwise indicate internal focalization. 26 See Genette 1980: 193. 27 Cf. Genette 1980: 218-222.

162

Rolf Reitan

shoulder too, all down one side of your back, along the vertebrae from neck to loins.” (Butor 1959: 1; italics mine.)

However, unless one restricts the definition of internal focalization to something very close to and even identical with interior monologue, “the smallish suitcase of a man used to making long journeys” may very well describe some content of a character’s conscious or preconscious mind, and the whole passage can be read as second person internal focalization, as can indeed the entire novel. Nonetheless, many readers of this narrator-less narrative have felt that its narration is a continued address to the protagonist. How do we know, then, whether a pronoun of address is used in its address function? For Fludernik, the use of an addressing “you” inevitably implies, as we have seen, an addresser “who may be explicit (by means of a narratorial I) or implicit” and often is no more than a voice, but in any case has to be posited as an “enunciatory instance”. This well-worn linguistic axiom generates a handful of practical imperatives, like for example ‘look for imperatives and hortatory expressions’. On the other hand, Fludernik does not hesitate when using formulae like “the effect of address or allocution inherent in the use of you”. Is this inherent effect nullified if it cannot be tied to an addressor? If so, why is it that we (or at least many of us), reading texts like La Modification cannot but respond to the second person pronoun’s inherent effect of address or allocution? It is an illusion, of course, no one speaks here—and perhaps it is even caused by a discrete and subtle version of Farley’s and Calvino’s autotelic trick? The pertinent answer is, I think, simple, and a consequence of a simple observation: when, in a narrative, a pronoun of address is used to refer to the protagonist, the protagonist does not know that he is being referred to, he does not hear the reference, unless the narrative is addressed to someone who is being told a story about himself, as for example in “The House of Ecstasy”. If, then, the protagonist is also addressed by the referential you, the address will not be heard either. This is the situation of address in apostrophe, it pervades the narrative in La Modification. We respond to narrative “you” in Butor’s novel as we respond to apostrophe, even though there is no “apostropher”. One-way communication, except for the fact that there is no communication. We are back to Kacandes’ radical narrative apostrophe—and to questions of fiction. Second Person Fictionality and the Present Tense If a general paradox of fictional narratives is that they refer to things and events created by referring to them, then prototypical second person narration gives an extra turn to the screw. It not only refers to, it also ad-

Theorizing Second-Person Narratives

163

dresses a character created by being addressed. Without explicitly talking about the paradox of fiction, Fludernik closes in upon it by emphasizing that second person texts, apart from breaking the frame of “discrete levels within a model of communicative circuits”, also foreground the processual and creative nature of story telling: the “you’s” experiences are explicitly projected from the discourse and are attributed to the “you,” without—in many cases—any evidence relating to the story world. (“Test Case”, 457-458)

The ‘lack of evidence relating to the story world’ probably refers to the manner in which most second person openings introduce the protagonist, i.e. directly, without any prior information. If medias res-openings as a rule signals fictionality, a simultaneous reference and address to the protagonist will accentuate the fictionality signal. Opening a narrative by addressing a fictional protagonist flaunts its fictionality, doubling it up, as it were. Fludernik’s remarks on the specific fictionality of second person narratives are found in the second chapter of “Test Case”, in which the main objective is to show how second person narration undermines constitutional elements of Genette’s and Stanzel’s narratologies. Here, obviously, is the right moment for a comment on Richardson’s Subjunctive (Hypothetical) Form: Some second person fictions even “openly, metafictionally, invent the addressee’s experience” (458)—Calvino’s If on a Winter Night is a notorious example—and “this condition becomes especially obvious for narratives in the “subjunctive” mode […], where the imperative engenders story matter by enunciational fiat”, and quite a few “imperative texts starts out with an initial ›imaginez‹ as a guideline for the reader.” Referring to Lorrie Moore’s stories from Self-Help, Fludernik observes that “these stories are both more openly fictional in comparison with ‘normal’ fiction, and less so”: More so, because “the invention is here openly signalled by recurring linguistic devices, not merely presupposed in the frame (the fact of a text's being a novel for instance)”. And less so, because for the reader involved in the situation as a generalized “you”, “for whom this very predicament might become virtual reality, such projected scenes appear to be less removed from real-life experience and therefore less ‘fictional’.” (458) In these fine observations, a certain scalar linearity perhaps hides an important distinction. Fludernik seems to suggest a progression from narratives that hide their fictionality (traditional first and third person narratives), over second person narratives that flaunts their fictionality by opening in medias res, and to narratives that openly invent the addressee’s experience. The distinction between flaunting and openly inventing is however crucial. If the truth of fiction is that it invents its story while making believe it is real, then stories engendered by enunciational fiat will threaten to destroy this belief. This is actually what happens in the framing story of If

164

Rolf Reitan

on Winter’s Night in which the addressed reader is given the role of an addressed protagonist. The result is not plain fiction but some kind of metafictional joke, even if it works as a frame for the embedded fragments of fictional stories. In a beautifully precise sentence, Fludernik (1995), explaining the preferred use of present tense in second-person fiction, discretely alludes to Käte Hamburgers epic preterit: The present tense serves a crucial function in most second person narratives since it allows an easy shift from the generalizing you (with its obligatory present tense—a gnomic present) to the you-in-reference-to-a-character, in which the present tense acquires the timeless quality of an “epic” or narrative present, loosing its deictic anchoring in the here-and-now of the narratorial process. (Fludernik, 1995: 109)

This would be valid for non-communicative narratives, narration without narrators. Then, Fludernik, always with an eye on possible scales, adds: It is therefore no coincidence that second person texts in the overwhelming majority employ the narrative present tense, even when they have a prominent speaking subject. In fact, the situation of enunciation or narration remains ambiguous in most present tense texts.

Agreed! However, the only present tense second person narrative with a ‘prominent speaking subject’ mentioned in “Test Case” is If on a Winter’s Night a Traveller—an autotelic form. Perhaps, then, any present tense second person narrative with a “prominent speaking subject” will turn out to be more or less atypical or deviant from the standard form?28 In a fine compilation of second-person narratives, Fludernik (1994c) presents an interesting array of (proto-)second-person texts from St. Augustine and forward, showing that these texts, even when containing surprisingly early precursors of internal focalization, are all “authorial”, i.e. they all have a prominent narrator. None of these proto-second person narratives are continually narrated in the present tense. However, Fludernik’s compilation of about 140 “Texts of second-person Narration and Border Cases” would seem to confirm that prominent narrators in 20th century present tense second person narrative as a rule occur only in narratives which deviate from the standard from; in fact, only three present tense narratives with a “prominent address function“ are found on this list. Two of these are not standard form second person narratives: Ilse 28

Fludernik also mentions Calvino’s “Re in ascolto” (“A King Listens”) as another example. Earlier (“Test Case”, 448) this text is mentioned as a possible self-address narrative. (“Re in ascolto” can easily be read as an internally focalized second person narrative. It does not have a prominent narrator; the “speaking subject” may be the protagonist speaking to himself, though initially in the voice of his assumed, internalized royal instructors. Internal focalization does not exclude polyphony!)

Theorizing Second-Person Narratives

165

Aichinger’s “Spiegelgeschichte” (1954) presents a story that somehow moves backward while the narrative moves forward, and John Updike’s “How to Love America and Leave It at the Same Time” (1979) is some kind of subjunctive narrative. The third example, Reginald McKnight’s “Soul Food” (1991), does have a prominent address function, but one which—in spite of the story’s defamiliarization technique (Fludernik 1994a, 461)—invites being “naturalized” as a retrospective self-address. Moreover, McKnight’s story is not a present tense narrative; apart from the “intro” (1 page) with its “presence of enunciation”, it is narrated over 18 pages in the past tense. This perhaps not so strange fact of absence of “prominent narrators” in present tense second person narratives might belong to a group of conventions or “laws” that so-called postmodern narratives seem obliged to play with, break or circumvent. Sunetra Gupta’s fabulous novel The Glassblower’s Breath (1994) is a present tense second person narrative with an approximate29 prototypical opening and with a discretely marked homodiegetic narrator who nonetheless narrates from an omniscient position and in general behaves like a heterodiegetic narrator. The narrative is for the most part internally focalized through the woman protagonist, it revels in the kind of extremely detailed narration we know from Butor or Perec and produces an even more extreme effect of intimacy. At the end of the story it turns out that the narrator (perhaps) is one of the characters, the protagonist’s husband. This would seem to turn the narrative into an elaborate apostrophe to the protagonist, and one made for her to read! But after the “unveiling” of the narrator the narration seamlessly continues using the same pronouns, “you” referring to and addressing the protagonist, “he” used in reference to any male character, the possible-impossible narrator included. At the very end of the story, the narration slides into the past tense, only to revert in the last paragraphs to some kind of present tense when addressing the protagonist. This novel, being neither autotelic nor subjunctive, would seem to move across most intermediate areas in “Diagram 1”. In using a narrator and then suggesting that the narrator might be one of the characters, it breaks, or rather circumvents, the law of the prototypical standard form, while still keeping the effect of radical narrative apostrophe. This “post29

“Approximate”, because, although the novel opens in medias res, it is initially narrated in the past tense, using a past dream to move further back in time, and only shifting to present tense after a few pages. This second-person technique suggests some narrator (who could be the protagonist herself): “That night, you dreamt, that instead of dying fresh as a blade of grass, last year, one winter afternoon, your sister had lingered, until her flesh had blackened so that the moonstone on her finger stood a monstrous white against her skin, […] (Gupta 1994: 5).

166

Rolf Reitan

modernist” strategy is not unlike what we see in some past tense postmodernist narratives, for example Cortazar’s “Graffiti”, Joyce Carol Oates’ “You”, or Josipovici’s Contre-Jour;30 they are all playing with apostrophes and with the borderline between homo- and heterodiegetic second person narration. However, they do not really unsettle or undermine useful narratological concepts and categories. On the contrary, they thrive on them. Recognizing that most narratives discussed in this section are told in the present tense, and refusing the supposition that narratological concepts in general need linguistic support to be acceptable, we may conclude: If an initial “you” can have the effect of flaunting a present tense narrative’s fictionality without destroying it, the logical implication will in these cases be that there cannot be a narrator who opens his narrative with this addressing and referring pronoun. The narrative is “from the creator to the created”; narrators do not create their stories, they are only telling someone something that happened. This is why there are no narrators in La modification and Un homme qui dort, or in Fuentes’ Aura. The pronoun of address in these narratives implies that protagonists are addressed, that there is no narrator addressing them, and finally, and fundamentally, that protagonists do not hear the address. This apostrophic mode of present tense narration, this “radical narrative apostrophe”, defines at least a prototypical standard form second person narration. It remains to be seen whether it is valid for all standard form narratives. Self-Address Richardson, as we saw, decided that the second person narration in McInerney’s Bright Lights, Big City is a form of self-address. How do we then register that a second person narrative is a “self-address narrative”? Let us first look at the case of past tense self-addressing narrative. Margolin (1986/87), talking about deictic transfers, coins a precise formula for this situation: If […] the [I‫׈‬you (tu/vous)] transfer is in the past tense, we are dealing with […] a case of what Bühler has termed ‘[D]eixis am Phantasma.’ In it, the speaker conjures up a past version of himself and talks to and about it from his present identity through the difference in tense and person.” (Margolin 1986/87: 169; quoted after Herman 2002: 354).

Given this neat and precise definition, construed using homodiegetic narrative as model, one can generate a testing procedure: Substitute every occurrence of narrative you with the pronoun I and check if the narrative still consistently narrates the same story as the one in the original narra30

Analyzed respectively in Kacandes 1994, and Fludernik 1993 and 1994b.

Theorizing Second-Person Narratives

167

tive. However, if the test result is positive, the stories identical, we still will not be able to conclude that the you-narrative is a self-address text. Other transfer formulas are possible, he/she‫׈‬you, for example (and a simple definition of heterodiegetic narrative) would give the same kind of positive result. Margolin’s definition obviously does not by itself deliver a procedure for deciding whether a past tense second person narrative actually operates an I‫׈‬you transfer; it can be used as a model for testing the consistency of interpretations only. In an investigation of “contextual anchoring” in Edna O’Brian’s A Pagan Place—“one of the very few second person texts narrated almost entirely in the past tense”31—David Herman seems to be aware of this. He points to “what might be deemed the default interpretation” of the novel—“that is, the assumption that, unless otherwise indicated, you is a pronominal stand-in for an I who figures as the protagonist of a narrative that she addresses to herself”. (Herman 2002: 358) Not showing much interest in the question of how to decide whether a given second person narrative is a self-address narrative, Herman nonetheless concludes the quoted sentence by adding an observation that might be helpful: “ […] a narrative that she addresses to herself and that we eavesdropping readers (quasivoyeuristically) overhear.” We recognize one of Kacandes’ points: the narrative you addresses and refers to the protagonist, the reader is only a secondary audience. Or as Herman puts it: “self-address intertwines the referential and address functions of narrative you but without cuing recipients to actualize that you.” (417, note 23) If I understand this right, it means that in a past tense narrative like A Pagan Place readers will generally32 not be exposed to a double deixis mechanism by the novel’s narrative you, there is no double address continually at work in this novel. If true, this might be

31 32

Richardson 1991: 358. Herman finds several particular passages in the novel that do trigger an “actualization” of the pronoun of address, and not only an actualization but also an address form that draws readers to somehow be participants in the fictional world. He calls the two forms “apostrophic” address to the reader, and “double deixis”. The apostrophic or actualized address is construed as entailing “address that exceeds the frame (or ontological threshold) of a fiction to reach the audience” (341); double deixis occurs when “you functions as a cue for superimposing two or more deictic roles, one internal to he story world [ …] and the other(s) external to that storyworld.” However, as Herman’s examples show (though there is some room for interpretation here) the apostrophic address is about identification, empathy and recognizing topics and situations that might have or possibly may occur in readers’ actual world, while double deixis may cause readers “to find themselves to be oddly nonvirtual participants in discourses from which they are nevertheless spatiotemporally removed” (348). This would seem to be a very precise description of my initial observing a double address in second person narratives like Butor and Perec’s. Different from Herman, however, I do not construe this in terms developed form discourse analysis but in terms of fiction.

168

Rolf Reitan

an effect by which one could decide whether a past tense second person narrative is a self-address. Fludernik, on her part, does not seem interested in this type of deictic transfer. On the contrary, in the second part of “Narrative You”—which discusses “the most important naturalizations of second person narratives and the generic models on which they rely” (230)—self-address figures as one among four models, closely connected with generalized you (which is another of the four models): The model of generalizing you for reflectoral second person fiction is reinforced by one final fictional (and indeed also experiential) model, that of self-address you. When people in their private thoughts argue with themselves, assuming a dialogue between their egos and superegos, or (re-)enact an exchange between themselves and a (possible) interlocutor, they may find themselves addressing their own selves employing a second person pronoun. Instances of this use are observable in interior monologue fiction (e.g. passim in Stephen’s musings in Ulysses), but for this reason has no extensive occurrence in literature. (238)

Like Herman, Fludernik does not ask how one can decide whether a second person narrative is narrated as self-address. But in “Test Case” the question is as if answered by a negative demonstration: Since the narrator, by definition, occupies the deictic position of the ›I‹ and the addressee the deictic position of the ‘you,’ you can only refer to the narrator in passages of self-address in which an ›I‹ splits into two voices that interact dialogically. This is not the case in the Diego passages of Virgil Suarez’s Latin Jazz (1989) where no first-person pronouns occur (except in the quoted dialogue. (“Test Case”, 450)

One of the Suarez passages is then quoted: Getting out of the car and advancing toward the entrance, no longer do you feel the itch in your nose. This stuff's a killer. Check the tie knot on the side mirror. Didn't get cut shaving, a miracle. In a hurry you went from Pilar's to your parents' and found the house empty. Your grandfather must have gone out somewhere, he usually stays at home to catch the news. Certainly he must be excited about the break-in. (Suarez 1990: 37)

The modality here seems to be “plain” reflectoral narration, i.e. internal focalization, therefore no narrator, no vocal choice, no split between a narrator’s I and the you, and therefore no self-address, either. Moreover, the pronoun could here as well have been he, she or I. Now, look at the initial passage of McInerney’s Bright Lights, Big City, Richardson’s preferred choice of self-address narrative, and in “Diagram 1” a prominent choice among reflectoral second person narratives: You are not the kind of guy that would be at a place like this at this time in the morning. But here you are, and you cannot say that the terrain is entirely unfamiliar, although the details are fuzzy. You are at a nightclub talking to a girl with a shaved head. The club is either Heartbreak or the Lizard Lounge. All might come

Theorizing Second-Person Narratives

169

clear if you could just slip into he bathroom and do a little more Bolivian Marching Powder. (McInerney 1992: 1)

In this text, as in Suarez’, there is no split between a narrator’s I and the you being referred to, no self-address. Using classical narratological concepts, Fludernik rules out both the necessity and the possibility of assuming selfaddress in these texts. Nonetheless, the temptation to read narratives like these as narratorial self-address has some reason to it. Their specific use of the pronoun of address makes them significantly different from narratives of the prototypical standard form. In McInerney’s novel “you” is in fact used as address in a few initiating sentences, and here with an allure of “generalized you”, as if someone were addressing everyone and no one, himself included: “You are not the kind of guy that would be at a place like this at this time in the morning” (1), “You see yourself as the kind of guy who appreciates a quiet night at home with a good book” (55), “Your interest in clothing doesn’t normally take you beyond Brooks Brothers and J. Press” (183). After each initiator phrase the narrative will continue using the pronoun in its reference function only. This makes it easy—and natural, as it were—to read, or anyhow to perceive, the pronoun as a “stand-in” for the I of present tense homodiegetic narration: a perfect present tense I‫׈‬you transfer (and not a he‫׈‬you transfer, which is barred by the initiating use of generalized address you). What then causes the temptation to read Bright Lights, Big City as a self-address narrative is simply that the I‫׈‬you transfer will change the reflectoral self-reference (I am not the kind of guy…, etc.) to a self-address. It is the unbelievable effect of this initializing and strategically repeated use of generalizing you that get sensitive people to believe they are perceiving a self-address where there probably is none. This seductive mechanism sets this kind of narrative apart from the prototypical standard form: We probably do not read Bright Lights, Big City as we read an apostrophe, i.e. in spite of its reflectoral modality and lack of narrator it is not a radical narrative apostrophe. Moreover, like the past tense narration in A Pagan Place, McInerney’s present tense narration does not cue the recipients to feel addressed by the pronoun of address—though perhaps with the very exact exception of the initiating use of it as generalized address. Speculations: Double address By “putting” the narratee on the agenda, second person narratives “query narratology’s privileging of the narrator as the locus of the story-discourse distinction, and that already from the ontology of the communicative mo-

170

Rolf Reitan

del itself.”33 (“Test Case”, 461) Moreover, second-person texts “additionally reach out to the reader roles projected by the text and invite active participation and even identification by real readers”. How, then, does this radically involving address to the reader really work (if at all)? Some passages from the third part of “Test Case” perhaps come close to an explanation: [Y]ou (like the present tense) is one of the attention-inducing features of the discussing mode in Weinrich's model34: you always alerts the current listener to pay attention since he or she may be directly called upon to react. (“Test Case”, 469)

This attention-inducing feature is probably what is being referred to by “deictic significance of the original address function of this pronoun” in the next quote: [T]he second-person pronoun in its generic or generalized usage—which constitutes its major opening gambit, successfully steering the reader into the fictional world so oddly represented by the second person—always relies on the submerged deictic significance of the original address function of this pronoun, even where that deictic significance is then channelled into an enhancement of the reflectoral quality of the text. (“Test Case”, 468; my emphasis)

The possible depth of this sentence may easily pass unnoticed: however we respond to the ambiguous you, its original attention-inducing address function may accompany the reading as a latent function, without the reader being aware of it. However, we may expect this latent address function to operate differently in different types of second person narrative. Following Herman we have assumed that there is not a general double address function in the past tense narration in A Pagan Place. We have observed that this is the case in Bright Light, Big City also, in spite of its specific use of generalized you. Obviously, much work will have to be done on individual narratives in order to gain a differentiated perspective on the varieties of double address. Here, I can only suggest some bits and pieces concerning our prototypical second person texts. In these, there are no “opening gambits” in the form of a generalized you, though readers may momentarily feel addressed by an initially ambiguous you and—knowing we are reading fiction—we may enter a game of being the protagonist, accepting the fiction of being the protagonist.35 After a while, we probably will forget it; getting ab33

The probable reference to Genette here is a bit inaccurate; Genette’s story-discourse distinction does not presuppose a narrator “locus”. Rather, Genette builds his theory of narratorial voice on top of the story-discourse distinction. (For a detailed investigation of this, see Reitan 2008.) 34 Cf. Weinreich 2001. 35 Here is probably the main point in my disagreement with Herman: If readers “accept the fiction of being the protagonist”, they will perhaps not confusedly superimpose spatio-tem-

Theorizing Second-Person Narratives

171

sorbed in the protagonist and his or her vicissitudes, we slip back to a “witness” position, like when reading a standard third person novel. But the latent deictic significance of “you”, having been activated once at the beginning, lies ready to be reactivated at any instant. Richardson’s and Kacandes’ idea of slippage may be useful here. The most intriguing aspect in our prototypical cases, however, concerns the way in which textual form keeps the reader in the protagonist’s place—with the result that the address duplicity melts into one fluent address-and-reference function. In La Modification, its consistent internal focalization would seem to be a primary condition for producing this effect, though this can only be a general requisition (you can be a spectator, or witness, a “secondary audience” to what happens on the stage of a fictional character’s mind). Second, there is an extreme limitation of time and space: Everything (the protagonist’s memories and apocalyptic dreams included) enfolds on a train during a voyage from Paris to Rome. Third, there is a strange and very complex factor, which I until further notice shall call “the voice”. It “immediately creates intimacy and defamiliarises”36—in Butor’s novel a narrating voice that incessantly, minutely, slowly, in the sleep-inducing rhythm of shifting shades of light and darkness and wagon wheels’ monotonous rrrom-ta-ta against rails—and absolutely without any digression whatsoever—reports your actions, perceptions, thoughts and dreams at the very moment they happen. The enigmatic origin and status of this “voice” may produce an idea of a guardian angel: patiently it watches you and watches over you as you struggle with your thoughts and dreams, your desolation and your repressed fears—as if, like the reader, it is waiting for something to happen.37 This is not a narratorial voice, it does not belong to someone who knows what is going to happen; it does not occupy a retrospective position of enunciation. It may be considered a voice of consciousness (which can be represented in internal focalization), but it is not a judging, warning, castrating or instructive voice, and above all: it is not a voice heard by the protagonist. It does not know anything that Leon Delmont does not know, but it voices some thoughts that he does not want to know are his own thoughts. The unbelievable suspense of this rather banal conflict drives the novel’s plot. Internal focalization, extreme limitation of time porally different worlds upon one another. Admittedly, it will require a lengthy negotiation between “fiction theory” and Herman’s refined discourse analysis to settle this issue. 36 Cf. Rikke Andersen Kraglund, paraphrasing Richardson on the second person “form”, in “Alternate Strains are to the Muses dear” (in the present issue). 37 For a completely different perception of Butor’s voice in La modification, see Rossum-Guyon 1970: 164. “La perspective narrative”, the third chapter of this book, written before narratology knew itself by name, and perhaps somewhat hesitant and inconclusive, still contains one of the most thorough and instructive discussions of voice in La modification. For a recent, challenging semiotic-pragmatic perspective, see Vega y Vega 2006.

172

Rolf Reitan

and space, a hyper-attentive “voice”, all contribute to keeping the reader in the protagonist’s position.38 It is an effect not achieved so fully in other second person narratives, perhaps with the exception of Perec’s A Man Asleep and Gupta’s The Glassblower’s Breath, even if time and space limitation in that novel is substituted by the tropical density of the protagonist’s immense intellectual sensuality, and even if the attempt to figure its narrative voice as an omniscient blood-and-flesh character gives it a hilarious postmodernist “twist”. Perhaps all this can be summarized as an extreme and absolute focus on the protagonist’s here-and-now. If a “further notice” is needed, here it is: What is it, this “voice”, if not a benevolent and, literally, an “authorial voice”; some version of the creator’s voice talking (no, not talking to) his creation, or more accurately, writing his creation, flaunting its fictionality and writing readers into the protagonist’s position—thus leaving, perhaps, the talking part, and therefore also the addressing, to the more or less unnatural naturalizations that readers’, and some writers’, imagination unavoidably will produce?39 References Barthes, Roland 1958 “Il n’y a pas de d’école Robbe-Grillet”, reprinted in Essais critiques (1964) Paris: Éditions du Seuil, 101-111. Butor, Michel 1980 La modification, Paris: Les Éditions de Minuit. 1959 A Change of Heart, New York: Simon and Schuster. Culler, Jonathan 1992 The Pursuit of Signs, London: Routledge. DelConte, Matt 2003 “Why you can’t speak: second-person narration, voice and a new model for understanding narrative”, Style 37.2, 204-219.

38

Not to forget the fact that this “mimetic” layer in Butor’s novel is saturated with echoes of the whole history of western culture and literature, producing a thematically and spiritually important texture too vast to be explored here, except for noting that it gives an exceptional intertextual depth to its impossible use of the second-person (see, for example Lydon 1980: 100-122). 39 A good friend recommended that I add one or two sentences here to clarify my “relation to the strange/unnatural”. Apart from in general agreeing with Henrik Skov Nielsen (see his “Fictional Voices? Strange Voices? Unnatural voices?” in this issue), I tend to think that one should look for the roots of any narrative unnaturalness in the possibilities & necessities of written language and its primordial difference from the usual suspect of naturalness: conversational story telling. (Plato knew this, of course, as does also the typewriter salesman Leon Delmont towards the end of Butor’s novel.) For the significance of writing in the genealogy of modern consciousness representation, see Maria Mäkelä, “Masters of Interiority” in the present issue.

Theorizing Second-Person Narratives

173

Farley, Ralph Milne 1938 “House of Ecstacy”, in: Marvin Kaye (ed.), 1988, Weird Tales, New York: Doubleday Book & Music Clubs, 3-11. Fludernik, Monika 1993 “Second Person Fiction: Narrative You As Adressee And/Or Protagonist”, Arbeiten aus Anglistik und Amerikanistik 18.2, 217-247. 1994a “Introduction: Second-person narrative and related issues” Style 28.3, 281-311. 1994b “Second-person narrative as a test case for narratology: the limits of realism”, Style 28.3, 445-479. 1994c “Second-Person Narrative: A Bibliography”, Style 28.4, 525-548. 1995 “Pronouns of Address and ‘Odd’ Third Person Forms: The Mechanics of Involvement in Fiction”, in: Keith Green (ed.), New Essays in Deixis. Discourse, Narrative, Literature, Amsterdam-Atlanta: Rodopi, 99-129. 1996 Towards a ‘Natural’ Narratology, London and New York: Routledge. Genette, Gérard 1980 Narrative Discourse. An Essay in Method, New York: Cornell University Press. Gupta, Sunetra 1999 The Glassblower’s Breath, London: Phoenix. Hantzis, Darlene Marie 1988 ‘You are about to begin reading’: The nature and function of second person point of view, Ph.D. dissertation, Lousiana State University, Department of Speech Communication, Theatre, and Communicative Disorders. Herman, David 1994 “Textual You and Double Deixix in Edna O’Brien’s A Pagan Place, Style 28.3, 378-410 (Reused in Herman (2002), chapter 9, “Contextual Anchoring”). 2002 Story Logic. Problems and Possibilities of Narrative, Lincoln and London: University of Nebraska Press. Kacandes, Irene 1990 Narrative Apostrophe: Case Studies in Second Person Fiction. Dissertation, Harvard University, Comparative Literature Department. 1993 “Are You In the Text? The “Literary Performative” in Postmodern Fiction”, Text and Performance Quarterly 13.2, 139-153. 1994 “Narrative Apostrophe: Reading, Rhetoric, Resistance in Michel Butor’s La modification and Julio Cortazar's “Graffiti””, Style 28.3, 329-349. Lydon, Marie 1980 Perpeetum Mobile. A Study of he Novels and Aesthetics of Michel Butor, Edmonton, Alberta: The University of Alberta Pres. Margolin, Uri 1986-87 “Dispersing/Voiding the Subject: A Narratological Perspective”, Théories du Texte 5/6, 181-210. 1994 “Narrative ‘You’ Revisited”, Language and Style 23.4, 1-21. McInerney, Jay 1992 Bright Lights, Big City, London: Bloomsbury Classics. Perec, Georges 1980 Un homme qui dort, Paris: Les Éditions de Minuit. 1990 Things: A Story of the Sixties / A Man Asleep, Boston: David R. Godine. Reitan, Rolf 2008 Fortællerfiktionen. Kritik af den rene narratologi, Århus: Aarhus Universitetsforlag. Richardson, Brian 1991 “The Poetics and Politics of Second Person Narrative”, Genre 24.3, 309-30. 2006 Unnatural Voices. Extreme Narration in Modern and Contemporary Fiction, Columbus: The Ohio State University.

174

Rolf Reitan

Rossum-Guyon, Françoise van 1970 Critique du roman. Essai sur “La Modification” de Michel Butor, Paris: Gallimard. Schofield, Dennis 1998 The Second Person: A Point of View? The Function of the Second-Person Pronoun in Narrative Prose Fiction, Geelong, Australia: Deakin University. Steinberg, Günter 1972 Zur erlebten Rede in Michel Butors “La Modification”, Vox Romanica 31, 334-364. Suarez, Virgil 1989 Latin Jazz, New York: William Morrow and Company. Vega y Vega, Jorge Juan 2006 “L’Éthos romanesque chez Michel Butor. Rhétorique, linguistique et littérature”, Orbis Litterarum 61.5, 341-360. Weinreich, Harald 2001 Tempus. Besprochene und erzählte Welt, Munich: C.H. Beck.



MARINA GRISHAKOVA (Tartu)

Toward a Typology of Virtual Narrative Voices A New Pragmatics of Voice1 Theoretical concepts are historically stipulated forms of meaning: principles of structuring and selection adopted by research communities determine the quality and quantity of information they are able to filter. In the humanities, concepts are often borrowed from the vernacular language and preserve a memory of their previous states. For instance, the concept of “voice” has never fully lost its “aural” quality, in the very same way that “focalization” has never been free from visual connotations—despite Genette’s effort to provide strictly linguistic foundations for narratological terminology and to purify the latter of vernacular meanings. Moreover, the two concepts coined to separate the functions of „speaking” and „seeing” (or „perceiving”), narration and focalization, proved to be inextricably entangled (see Jahn 1996, Phelan 2001, Herman 2009, Grishakova 2002). Recently, the qualitative aspects of “voice” have attracted close scholarly attention: voice was defined as a distinctive expressive and idiomatic quality of speech (Aczel 1998) or a trace of “individualized perception” (Richardson 2006, 8). Insofar as “voice” becomes connected with „perception”,2 the strict separation of narration and focalization, or, in Greimas’ terminology, the performative and cognitive functions of discourse, becomes questionable. It was the pioneering work of Mikhail Bakhtin that established the qualitative aspects of voice—its distinctive idiomatic traits and expressive capacities; its dialogic, hybrid, quotational quality and polygenetic origins—in the focus of scholarly discussion. In Bakhtin, voice emerges as a slippery phenomenon with blurred borders: it is a site of interaction of other voices whose traces and reverberation it comprises—a word with “a 1 2

This contribution has been supported by the Estonian Science Foundation (Grant ETF 8874). The viewpoint “is established linguistically, with cognition and perception adequacy interacting to abet or refute the linguistic evidence” (Fludernik 1998: 327).

176

Marina Grishakova

sideward glance at someone else’s word” (Bakhtin 1984, 199). By drawing on the Bakhtinian theory, Richard Aczel highlighted the interpretative value of voice as a readerly construction. Likewise, Monika Fludernik argues that the phenomenon of dual voice in free indirect discourse is constructed by the reader from the textual clues.3 The special issue of New Literary History on voices, including papers by Aczel, Jahn, Fludernik, Richardson and other scholars, is just one landmark of what might be called the Bakhtinian (anti-essentialist, contextualist, pragmatic) turn in the humanities, particularly in sociolinguistics and narratology. Increased interest in the distributed forms of authorship and complex forms of narrative identity (Georgakopoulou 2007, de Peuter 1998, Page and Thomas (forthcoming), Hydén 2009, Grishakova (forthcoming)), multi-person and multi-voiced narratives (Richardson 2006, Nielsen 2003), multiple protagonist films (Israel 2010, Tröhler 2010), and hybrid speech forms (Fludernik 1998) is a characteristic trait of this period. Seemingly essentialist entities and categories are re-defined as contextual configurations: Identity […] emerges as a result of the interaction of two or more narrative agents, every agent’s intentions being modified, completed or suppressed by other agents, – it emerges from the difference between the possible (the agent’s intentions and goals) and the actual (the agent’s intentions and goals modified by other agents) elements of the situation: by capturing this difference, identity imposes temporary control over the situation (Grishakova, forthcoming).

The shift of focus from the textual form to the situation of storytelling and of the reader’s (listener’s) activity leads to the critical revision of the code model of communication based on the principle of a “shared message”, and its substitution for the inferential and interactional models. From the viewpoint of the inferential theory, fiction is a part of “serious” everyday communication, a product of a special use of language whose poetic effect “results from accessing a large array of very weak implicatures” (Sperber and Wilson 1995: 224; see also Pilkington 2000) to be approved or discarded in the process of reading (listening) and drawing inferences. It is exactly the principle of inference rather than truth or understanding that governs communication, particularly fictional. From the perspective of interactional theory, based on Erving Goffman’s concept of “interaction order”, information exchange and partici3

“The dual voice hypothesis is [...] a convincing model of narrative speech and thought representation because the concept of a textual agon of voices corresponds with the reader’s intuitive perception of discourse in general [...], and with the reader’s intuitions about the effects of such voice. Readers do in fact construct a narrator’s (or author’s) voice as a default value and, given sufficient linguistic evidence, experience an evocation of figural voices on that background” (Fludernik 1998: 350).

Toward a Typology of Virtual Narrative Voices

177

pant’s intention are of secondary importance; communication takes place and is co-implemented by participants independently of their capacity to emit information or display intentions. The interactional model “suggests that what is communicated is inherently situated, and often situated (in an interpretive sense) in different ways for different people. Situated interpretations are made possible by contextualization cues” (Schiffrin 1994: 401). This kind of immediate experience of togetherness or co-presence (insofar as all language is infused or populated by other people’s discourse) is also an important part of Bakhtin’s theory. Virtual Voices The concept of voice and its functions in both fictional and nonfictional storytelling have yet to be explored in the light of the new pragmatics of communication. In what follows I intend to offer some preliminary observations on the phenomenon of virtual voice, i.e. hypothetical discourse which might have been, but allegedly is not, pronounced by the purported speaker, being instead attributable to a different textual consciousness (“originator-consciousness”). According to Ryan, the dynamics of the “actual” and the “possible” (“virtual”) is essential for understanding the logic of fictional narratives. The dichotomy “actual/non-actual” repeats itself, recursively, within the fictional world: When we read a text of narrative fiction, we take some statements as establishing hard facts for the story world and others as describing what is merely possible or what exists only in the minds of the characters. In other words, a fiction is not just a nonactual possible world; it is a complete modal system centered around its own actual world (Ryan 2006: 646).

As Ryan argues, the traditional conception of story „should be expanded into a model that involves not only the actual but also the virtual” (ibid, 647). Likewise, Manfred Jahn considers “internal” (virtual, mental, implicit) stories as crucial counterparts of the “external” ones (Jahn 2003) and suggests a cyclical model of narrative, which links “internal” and “external” elements together as data structures within the information flow. The relation between the actual and the virtual, between “external” and “internal”, i.e. between the facts taken for granted and the mere hypotheses or figments of imagination, belongs to the domain of the ontology and semantics of fiction. The “virtual” overlaps with certain forms of linguistic and narrative representation, such as “direct” and “indirect discourse”, the “omniscient” or “limited” point of view, but is not reducible to surface manifestations. Rather, the “virtual” is indexically invoked by the “actual”: the semiotic notion of “indexicality”, as appropriated by so-

178

Marina Grishakova

cial anthropology and sociolinguistics, refers to the ability of a linguistic form to indirectly invoke contextual, associated meanings (see e.g. Georgakopoulou 2007: 8-10). The concept of virtual voice should be considered in its relation to the notions of the “disnarrated” (as defined in Prince 1988) and “virtual narrative” (Ryan 1991: 166-169). Whereas Ryan’s “virtual narrative” partially overlaps with Prince’s “disnarrated,” the difference between my concept of the “virtual voice” and Ryan’s “virtual narrative” is functional. Two distinctive traits of the “virtual”, according to Ryan, are: 1) the secondary role of linguistic manifestation: the “virtual” is less dependent on linguistic manifestation (ibid. 168), it is most often implicit or inferred by the reader; 2) the ontological status of a mental act. However, “virtual narrative” as a sub-category of the disnarrated belongs to the level of plot development: “virtual narratives” are unrealized or unrealizable plot versions. In a broad sense, the disnarrated may be defined as counterfactual (what could or might have happened but didn’t), and as such it is part of the (alternative) course of the plot events. The narrative voice manifests itself on the level of discourse: it functions as a means of perspectivization and interpretation (cf. Aczel’s definition of reading as an experience of overhearing voices). The virtual voice is not an unrealized or discarded possibility of plot development, but a realized discursive effect whose value is purely interpretive. For example, narration by the dead or ghostly narrator as a representation of either “natural” (everyday) or “unnatural” (otherworldly) events, which are all, nevertheless, components of the textual actual world and, as such, part of the game of “make-believe”, is not meant to raise the (rather trivial) question of whether ghosts are able to speak or whether represented events indeed took place, but rather to shed new light on these events and occurrences of human life, to reveal new perspectives on life and death in general—or to pose the questions concerning the nature of fiction. Otherwise, the narrator’s voice establishes a certain interpretive transworld relationship between the fictional actual world and the reader’s actual world. The phenomenon of the virtual voice discloses a discrepancy between the mimetic and diegetic aspects of fictional storytelling and challenges the Genettian conception of person-related “speaking voice”: due to the existence of “virtual” counterparts of the representation, the storyworld opens indirect access to worlds much more extensive than the world it represents, which is its immediate communicative context (see Mukaőovský 1970).

Toward a Typology of Virtual Narrative Voices

179

“Natural” Virtual Voices Linguistic research testifies to the presence of virtual voices in the situation of naturally occurring everyday nonfictional storytelling. Following in Bakhtin’s footsteps, Deborah Tannen offers a preliminary outline of the “poetics of everyday communication” in her Talking Voices: Repetition, Dialogue, and Imagery in Conversational Discourse (2007). For Tannen, any representation of speech in oral discourse is constructed rather than reported: being extracted from its actual context (i.e. subjected to “re-keying” or “re-framing”), it acquires a virtual quality. In oral discourse, specific prosodic means, and in written discourse specific graphic4 and linguistic signals reveal the “polyphonic layering of voices” (Tannen 2007: 19). I shall use my own and other scholars’ examples to point to cases of what I call “virtual voice” in the situation of oral communication. 1. First, the case of “speaking for another” or “ventriloquizing” (Tannen), e.g. the forms of speech which seem fictional par excellence; in a similar way, the omniscient narrator “animates” characters in fiction. People often use their pets to communicate with each other in an indirect, mediated way, e.g. by talking to and simultaneously “for” a cat desperately trying to get the attention of a family member who is back home after a prolonged absence: “He’s not looking at you? He’s not paying attention to you, hmm?” Tannen observes a similar behavior in families with pre-verbal or minimally verbal children, e.g. when mother addresses father who’s ignoring their little daughter, craving for his attention, by using a high-pitched, sing-song “baby talk register”: “Can you say, / I was just trying to get some of Daddy’s attention, / and I don’t really feel good, either” (Tannen 2007: 22; see also Tannen 2004). Finally, the function of “speaking for another” is performed by a carer or a family member who helps a disabled person who cannot tell his or her own story: the former authorizes the other person’s speech in situations where the other’s authorial voice is weak or cannot function. This kind of “voice support” is also characteristic of psychotherapeutic relations. LarsChrister Hydén, drawing on Bakhtin, describes it as a co-operation of broken and vicarious voices (Hydén 2009): the vicarious voice supports, supplements or replaces the broken voice of an affected person. The analogous situation is that of dramatic or musical performance, authorized by the actor or a musician, who turns the “silent” text of a drama into a play or a musical score into a performance. Finally, the model of “voice sup4

Italics are the most common means to indicate the “strangeness” or „otherness” of voice, which is, otherwise, not separated from the embedding discourse. Cf. “...somebody would have seen and, too soon, everybody would have known that the McCarthy girl was, you know, a bit funny, like that” (from Ali Smith’s Like (1997); Smith 2005: 159).

180

Marina Grishakova

port” may be used to describe the situation of translation, in which the translator mediates the translated text into his (her) own culture and thus (partially or wholly) appropriates the translated discourse.5 2. Second, representation of implied speech (1st or 2nd person singular) that was not actually pronounced. Tannen’s example is that of a woman who tells her friend how, when she was a child, her father embarrassed her by scolding her in front of her peers for not being obedient enough: “You can’t say, ‘Well Daddy I didn’t HEAR you’” (Tannen 2007: 113). The emphasized “hear” and idiomatic “Daddy” are contextualized traits of the “childish” discourse that introduce the viewpoint of the self as a child and a character in the story. 3. Third, what Tannen calls “choral dialogue” or, more exactly, a generalized summary of hypothetical “collective” speech (1st or 3d person plural), as, for example, in George Gershwin’s lyrics They All Laughed: “They all said we never could be happy”, “They all said we’d never get together”. This generalized speech form is rather usual in oral discourse: “all mothers say to their children...”, “my friends always say to me...”, “all Americans expressed their indignation...”, etc. 4. Fourth, representation or, rather, reconstruction of another person’s thought, or what is called “omniscience” in literary narratology and often considered a distinctive feature of fictional discourse. In Tannen’s terminology, it is the implied “inner speech of others”, for example, a guess on what a batter hitting a pitch in a baseball game must have been thinking: “And he—you could see him just draw back like ‘Man, I’m going to knock this thing to Kingdom Come’” (Tannen 2007: 116). 5. Finally, two or more voices belonging to the same subject, of which he (she) is unaware or which are not liable to his or her control, as in the case of Tourette’s syndrom, multiple personality disorder, “inserted thoughts” and “hearing voices” in schizophrenia (see e.g. Brown and Kushner 2001, Miller 2001 etc.). Virtual Voices in Fiction Fiction considerably intensifies and enhances the effect of “virtual voices”, while using them to build storyworlds, which are simultaneously relatively autonomous and open for interpretative expansion. By opening access to “a large array of very weak implicatures”, as formulated by adepts 5

Thus, in the Russian tradition, The Golden Key, or Adventures of Buratino by Aleksey Tolstoy (a free adaptation of Pinocchio by Carlo Collodi) and Winnie the Pooh by Zakhoder (a free translation of Milne’s classical book) function as original texts.

Toward a Typology of Virtual Narrative Voices

181

of the inferential theory of communication, fictional representation considerably extends the sphere of the implicit and virtual. 1. As for the case of „speaking for another”, recorded in situations of nonfictional everyday storytelling, insofar as the fictional narrator’s and character’s direct discourse is taken as „actual” and as having a constitutive force by virtue of a basic fictional convention,6 any embedded (mediated) thought and speech representation (psycho-narration, free indirect discourse, stream-of-consciousness) may qualify as “virtual voice”, i.e. a narrator’s or character’s verbalization of another’s “potential” or internal speech, which has allegedly not been verbalized by its purportive speaker and remains “on the threshold of verbalization” (Cohn 1978: 103). The fragments of another’s implied, hypothetical discourse are detected or inferred by the reader from contextual cues. It is well known and widely accepted by literary narratologists that linguistic markers are neither reliable nor sufficient for distinguishing between the embedding narrator’s discourse and the embedded character’s discourse (see e.g. Tammi 2003). What seems to be an interesting and most urgent task, is not finding strict formal criteria to distinguish between the components of mixed and hybrid forms of representation, but rather defining their ontological status and function in the whole storyworld economy, their positioning on the borderline of what is accepted as “fact” or “knowledge” in the textual actual world and what belongs in the realm of the possible, hypothetical or imaginary. I shall take the novel The Accidental (2005) by Scottish author Ali Smith as an example of the skillful employment of virtual voices. The novel consists of three parts, “The Beginning”, “Middle” and “The End”, each part comprising sections of third-person narration with extensive embedded 6

For Doležel (1998: 149) only the third-person anonymous narration possesses “authenticating” force in the storyworld. Obviously, the authentication is a matter of context and degree: any type of discourse has a relative authenticating value in the storyworld’s ontology. Doležel’s typology seems to rely on the old Platonic distinction between diegesis and mimesis: the closer the discourse is to the purely mimetic (representational, subjective) type on the scale of opportunities between the purely diegetic and the purely mimetic, the less “authentic” and reliable it is. However, the most purely mimetic type of discourse, direct discourse, separated from the narrator’s discourse by formal markers (i.e. tags, such as “he said”, commas, quotation marks or the graphic form of the dialogue) has the status of an actual event (speech act) in the fictional world, whereas a character’s discourse embedded in the narrator’s discourse has the status of a subjective rendering or “virtual quotation” from the character’s idioms, the narrator coming “more and more to intervene, mediating between the character’s discourse and the reader” (McHale 2004: 197). The “slipping” perspective and “speech interference” (a Bakhtin-Voloshinov’s term), i.e. interaction of the reported utterance and reporting context, accounts for “heightened subjectivity” (ibid. 208) of mixed types of speech and thought representation. For our purpose, it is not the difference between the diegetic and the mimetic that matters, but rather a degree of narrative mediation: by a constitutive fictional convention, the less mediated the discourse the greater its authenticating value.

182

Marina Grishakova

fragments of free indirect discourse by four members of the Smarts family, each section focusing on one family member. The embedded speech retains the tense and person of the narrator’s discourse (present tense in the case of the children, Astrid and Magnus; past tense in the case of the parents, Eve and Michael) and thus qualifies as free indirect discourse. However, the narratorial discursive presence seems to be minimal: in each section, the free indirect discourse manifests not only specific idiomatic traits (colloquial and “smart” Latin words in Astrid’s sections; erudite school vocabulary and fragments of the half-suppressed memory of a schoolmate’s suicide in Magnus’s sections; literary quotations and Romantic clichés in Michael’s sections) but also each character’s perceptions and observations through which, in their turn, other characters’ and outside voices (the voices of TV, newspapers and schoolbooks) are filtered. For instance, one of Eve’s sections has a form of a self-interview, mimicking a genre of “autobiotruefictinterview” of her own successful invention, Eve’s consciousness-splitting between the third-person “me” of the interviewer and the first-person interviewed; Michael’s sections comprise verses as a distinctive feature of professional discourse. Such deep and extensive immersion into the characters’ consciousness along with a minimal narratorial presence in the different sections allows free indirect discourse to function as a vehicle for stream-of-consciousness. Moreover, Astrid’s and Magnus’ sections may be read as samples of simultaneous narration as they contain characters’ immediate reactions to the surrounding reality and reflect this reality in a sufficiently coherent, though indirect and subjectively mediated way. This is probably the reason why certain critics considered Smith’s novel a narration by 5 narratorial voices. The fifth voice belongs to a mysterious first-person narrator, Alhambra, a woman who had been conceived in a cinema café during a film performance and seems to be somehow associated with the cinematic world. Textual cues reveal a similarity between Alhambra and the uninvited guest, Amber, who suddenly shows up in the house the Smarts rent for the summer, intruding into their life, provoking them to communicate with each other and trying to turn the novel into a “polylogue”. Alhambra’s sections enclose other narrations, in the opening and at the end as well as between the three parts of the novel. This enclosement as well as the connection between Alhambra’s cinematic origin and the structure of the novel (Pasolini’s 1968 film Theorema being its main subtext) leads the reader to identify Alhambra with the authorial narrator. Though relegated to the role of a character inside the storyworld, Amber-Alhambra demonstrates an extraordinary degree of intrusion, imbuing protagonists’ lives with her presence, evoking controversial feelings of strong attachment and aversion. Being thrown out of the Smarts’ summerhouse, she performs a presumed

Toward a Typology of Virtual Narrative Voices

183

act of revenge by emptying their urban house of all its possessions. Along with Eve’s escape from the family and her round-the-world trip, which threatens to turn her into another Amber, an uninvited guest of other people’s homes, this tour-de-force produces the allegorical and metafictional ending of the novel. As a result, the sections narrated from the point of view of the characters’ consciousnesses with seemingly minimal narratorial intrusion reveal their mediated, virtual nature: the authorial-narratorial presence, epitomized by the crucial role Amber-Alhambra plays in the novel, manifests itself in the selection and characterization of the characters’ voices as well as the narrator’s ironic-empathic engagement with characters. 2. The next type of virtual voice is hypothetical discourse that has not been pronounced, most often presented in the 1st, 2nd or 3rd person subjunctive form. This hypothetical speech is usually related to hypothetical focalization. For instance, in the episode from The Double by Dostoevsky, a most critical moment of Mr Golyadkin’s life, after his shameful retreat from Klara Olsufyevna’s birthday, when he runs in the St. Petersburg streets in a stormy night and meets his double, the author introduces the figure of a hypothetical “disinterested observer” who might see Golyadkin’s “miserable gait”: If some disinterested outside observer had now glanced, just like that, from the outside, at Mr Golyadkin’s miserable gait, even he would at once have been pierced by all the terrible horror of his misfortunes and would without fail have said that Mr Golyadkin looked now as if he himself wanted to hide somewhere from himself, as if he himself wanted to escape somewhere from himself. Yes, it really was so! We shall say more: Mr Golyadkin now not only wished to escape from himself, but even completely to annihilate himself, not to be, to turn to dust (Dostoevsky 2004: 41).

Why is the “disinterested outside observer” needed in the episode, preparing Golyadkin’s encounter with his double? After all, the “disinterested observer’s” hypothetical discourse is not much different from the nervous, repetitive speech of the Dostoevskian unreliable narrator, which, in its turn, reflects the peculiarities of Golyadkin’s inner world: likewise, the observer’s hypothetical discourse contains repetition, parallelisms and tautology (“terrible horror”) reflecting Golyadkin’s compulsive nervousness and incertitude. What is symptomatic is Dostoevsky’s effort to present the subjectively tinged speech as the seemingly objective report of a disinterested observer. For Dostoevsky, who distrusted the Enlightenment’s belief in rational consciousness, “subjectification” of reality via different subjective perspectives meant a higher degree of objectivity.7 In Dostoevsky, 7

Dostoevsky is often considered a predecessor of modernist “negative” aesthetics, whose “otherness” in respect to the 19th century humanist and positivist culture has been pointed

184

Marina Grishakova

the virtual is presented as the actual, confirmed by the 1st-person plural narratorial voice (“we shall say more”, emphasis is mine. M.G.), the latter providing an alibi for the character’s subjective point of view (“Mr Golyadkin now not only wished to escape from himself, but even completely to annihilate himself, not to be, to turn to dust”). The next two types of virtual voice in both oral and written, factual and fictional discourse are 3) generalized or impersonal voice (“one”- or “we”-forms) and 4) overtly fictive or projected voices whose existence depends on what we termed the “originator consciousness”.8 Richardson’s example is Beckett’s The Unnamable, where [n]arrative voice often refers to other characters and voices which it then discloses to be fictions invented by itself, as apparently independent persons are collapsed into a single narrative voice. On the other hand, the same narrator goes on to claim to have invented the frustrated narrators of other novels written by Beckett; the narrator of The Unnamable seems to be impersonating its author (Richardson 2006: 76).

Beckett’s narratorial voice is often populated by other narratorial potentialities and virtual voices. The narrators of The Unnamable (1953) and the simultaneously written Stories for Nothing seem to be looking for their own selves: narration presents a series of self-projections neither of which is stable or durable.9 The instability of the narratorial voice short-circuits reader’s expectations and epitomizes the approximate and distorting character of any self-identification. The narration may burst into a whole array of multiple, disjointed and fractured voices. Beckett’s narration may serve as a model of autocommunication: the narrator groping for his own self, while trying to avoid both closure and the disintegration of self-identity. A variable degree of presence and embodiment may be ascribed to the fictive voice, while the latter originates in a purely mental, ghostly or a semi-anthropomorphic entity. Martin Amis’ novel Time’s Arrow (1991) is a story of a German war criminal, Odilo Unverdorben aka Hamilton de Sousa aka John Young aka Dr Tod T. Friendly, told from the end to the out by many theorists of modernity. The “objective reality” is constructed by modernists “in a radically subjective manner”: on the surface level it is objectified and presented in greater detail than ever before (naturalism, imagism etc.), but this objectified representation is radically modified and eroded from the inside by subjective perceptions (see Eysteinsson 1990: 43). 8 Cf. Humbert Humbert’s words in V. Nabokov’s Lolita: “Please, reader, imagine me! I shall not exist if you do not imagine me!” 9 Likewise, in the situation of oral everyday communication, narrative identity emerges as a resultant force of the whole situational pattern: it is distributed between the authorial, narratorial and characterial self-projections, or, according to the other, more complicated scheme, between “animator”, “author”, “figure” and “principal” (adopted from Goffman and Schiffrin by Georgakopoulu 2007: 16).

Toward a Typology of Virtual Narrative Voices

185

beginning. The narrator is the disembodied mental ‘self,’ somehow unified with Odilo’s body, yet also separate. Chatman calls the narrator “Soul”: Tod and Soul move in opposite temporal directions. For Tod the discourse moves backwards, from his death to his birth; for Soul, the movement is ‘forward into the past’ [...] The implied reader is presented with two narratives – the narrative of Tod’s life, and the narrative of Soul’s effort to uncover and understand that life (Chatman 2009: 41).

This type of narration is partially modeled upon Nabokov’s novella The Eye (Russian text 1930, English version 1965). In Nabokov’s novella, the first-person narrator commits suicide, yet his mental self lives on and creates the world anew. The self (the “I”) has the function of a spectator, witness, observer—the invisible “presence” in the diegetic world, retaining some traits of bodily presence. In his narratorial capacity, he observes the events from the outside, but also enters the fictional world as either a character or as a ghost (“my disincarnate flitting from room to room,” Nabokov 1965, 69), anxiously looking for the meaning of the events and for the “real” image of his other half that exists under Smurov’s name as a character in the newly created diegetic world. He pretends to be the author of the story who “can accelerate or retard” the motions of characters, “or distribute them in different groups, or arrange them in various patterns” (Nabokov 1965, 100). However, as an intermediary figure between the authorial heterodiegetic narrator and the homodiegetic narratorcharacter, the “I” of Nabokov’s novella lacks the full narratorial power for which he is striving—hence the incompleteness of his existence and the final merging with the character at the end of the story.10 The narrator of the first section („Past”) of Ali Smith’s Hotel World (2001) is the ghost of a dead girl, a hotel chambermaid, who has fallen to death in a hotel food elevator. The ghostly narrator is insistently seeking an answer to the question: how long did the fatal fall last? This “intermediary” moment between life and death epitomizes a peculiar, shifting “inbetween” state of the narrator who either associates herself with the dead girl, using the first person plural form (“we”), when referring to their previous co-existence: “We were a girl, we died young; the opposite of the old, we died it. We had a name and nineteen summers; it says as much on the stone. Hers/mine. She/I” (Smith 2002: 9)—or dissociates herself from the dead girl in her present state—a corpse, with whom she is, nev10

In Nabokov’s late novels, such as Pale Fire, Transparent Things, Ada, there are several more or less fictive and more or less ghostly narrators struggling for their rights to narratorial voice and authority in each novel. In Transparent Things, the leading candidate for narratorship among Nabokov’s anagrammatic alter egos, is the writer R., who is already dead when the narration starts. Therefore voice metamorphoses in the novel are easily explainable: „the dead are good mixers, that’s quite certain” (Nabokov 1989: 93).

186

Marina Grishakova

ertheless, able to communicate, whom she provokes to talk of the last minutes of her/their life: “Actually it is very nice, where they buried her”; “They’ve put up a stone with her name and her dates and an oval photograph on it”; “so many exciting things are happening to her now. Maybe the earnest ticklish mouths of worms; anything” (ibid). Nevertheless, the ghostly narrator is dependent on other people for filling in the gaps in the fragmented plot line and for discovering her own role in the story. 5. The next subcategory of virtual voices is metaleptic virtual voice: an extradiegetic narrator’s imaginary speech addressed to a character, for example, the imaginary dialogue between narrator and character in the beginning of Nabokov’s Transparent Things (1972). The novel’s protagonist’s name is Hugh Person (just “a person”, “somebody”). Being relegated to the role of an anthropomorphic character-like entity within the fictional world, the narrator picks his “hero” by chance from out of the crowd: Here’s the person I want. Hullo, person! Doesn’t hear me. […] Hullo, person! What’s the matter, don’t pull me. I’m not bothering him. Oh, all right. Hullo, person… (last time, in a very small voice) (Nabokov 1989: 1).

6. Finally, alternative voices belonging to the same subject, yet resistant to control and maintaining their separate quality—as symptoms of alternate state of consciousness or a psychiatric disorder. Thus, in Nikolai Gogol’s story The Diary of A Madman (1835), considered one of the oldest accounts of schizophrenia (Altschuler 2001), a split exists between the two aspects of the self: the spontaneous, impulsive and pathetic characterial aspect and the self-confident, reasoning narratorial aspect, the narrator taking a stance of dignity and condescension, contrasting with his alter ego’s poverty and humiliation. Eventually the narratorial aspect predominates and leads the madman to believe that he is the King of Spain. However, in the lunatic asylum, he is again relegated to the role of a character suffering from violence and humiliation. In A School For Fools (1976) by Sasha Sokolov, a Russian postmodernist writer living in the U.S., the schizophrenic character’s monologue presents a dramatization of the narrator’s two selves and includes stream-ofconsciousness monologues of other characters, who also, in their turn, split into double or multiple selves. The status of these monologues does not reveal, however, whether they are, indeed, quoted monologues or rather the narrator’s confabulations, which, from time to time, turn into sequences of disconnected onomatopoetic language. The narrator’s two selves, a loquacious, gregarious one and a reasoning, controlling one, are in a state of permanent dialogue, though they appear to the outside observer a single person—even if their singleness is a matter of agreement (“I thought we agreed once and for all that there is no difference between us”, Sokolov 1988: 21). These two voices, which, in their turn, incorporate

Toward a Typology of Virtual Narrative Voices

187

other voices, are telling the story, being written down by the author (alias the teacher and, as suggested by some critics, a projection of the narrator’s interiorized, reasoning self) who intervenes and comments on the occurrences in the storyworld and the storytelling process. The voices are trying to overcome or suppress their difference, as the difference is punishable: …we lived with our parents in the room where I now live alone, or you live alone, in short—one of us does. But still—who precisely? What’s the difference? I am telling you a very interesting story, and you’re starting to pester me again, and I don’t pester you after all, I thought we agreed once and for all that there is no difference between us, or do you want to go there again? (Sokolov 1988: 21)

The idiomatic traits of the narrator’s voices depend on their idiosyncratic perceptions, e.g. on their inability to perceive linear time. As nothing looks quite complete and fatal in the nonlinear temporality, the narrator intentionally uses the word “former” instead of “deceased”, referring to his “former grandmother” instead of the “deceased grandmother”. The novel ends with a metaleptic tour-de-force, introducing the author as a mere “scriptor”, recording the first-person narrator’s stories: Student so-and-so, allow me, the author, to interrupt your narrative again. The thing is that it’s time to end the book: I’m out of paper. True, if you intend to add two or three more stories from your life, I’ll run to the store and buy several more packages right now. With pleasure, dear author, I’d like to, but you won’t believe it anyway […] Student so-and-so, that is extremely interesting and it strikes me as totally believable, so let’s go get the paper together, and along the way you tell me everything in order and in detail... (ibid. 228).

* This paper, which has sought to single out six categories of “virtual” narrative voice (that is, mediated or embedded, hypothetical, generalized, fictive, metaleptic, and alternative) may contribute to further the remodelling of the traditional conception of the story, initiated by Marie-Laure Ryan and Manfred Jahn, among others—a remodelling that would introduce transitional and virtual states into the storyworld ontology. Fictional narratives exploit the effect of virtuality that is also inherent in nonfictional everyday storytelling. However, fiction considerably intensifies properties of language and appropriates them into its own translinguistic realm, where individual voices function as prominent and powerful “centrifugal forces”, to use Bakhtin’s expression. “Virtual voices” are centrifugal tools that considerably extend the “multiverses” of fiction by linking them with various discourses, genres and contexts, by intruding into the realms of the possible, hypothetical and imaginary, and by questioning the familiar, accepted and stereotypical. In fiction, this is accomplished not by the direct presentation of hypotheses or surmises about the world,

188

Marina Grishakova

but by their mediated construction and consideration inherent in the process of storytelling. In modern experimental fiction, where linguistic markers serve as unstable, floating signifiers of subjectivity, the phenomenon of the “virtual voice” may be considered as a form of fictional auto-communication that articulates the “transitional space” of subjectivity (cf. Sclater 2003) and points at its hidden dimensions beyond the limits of language and the immediately communicated “story”. On the other hand, the significant role of voices in contemporary fiction reflects the shift of interest towards “small stories”, local, idiosyncratic, subjective narratives, elusive phenomena of individual memory and perception, which are difficult to appropriate ideologically and politically —and the distrust of any foundational essences, be they personal or corporative. References Aczel, Richard 1998 “Hearing Voices in Narrative Texts”, New Literary History, 29.3, Theoretical Explorations, 467-500. Altschuler, Eric Lewin 2001 “One of the oldest cases of schizophrenia in Gogol’s Diary of a Madman”, British Medical Journal 323, 1475-1477. Bakhtin, Mikhail 1984 Problems of Dostoevsky’s Poetics, Minneapolis-London: University of Minnesota Press. Brown, Kate E. and Howard I. Kushner 2001 “Eruptive Voices: Coprolalia, Malediction, and the Poetics of Cursing”, New Literary History 32, 537-562. Chatman, Seymour 2009 “Backwards”, Narrative 17.1, 31-55. Cohn, Dorrit 1978 Transparent Minds: Narrative Modes for Presenting Consiousness in Fiction, Princeton: Princeton University Press. Doležel, Lubomír 1998 Heterocosmica. Fiction and Possible Worlds, Baltimore: The Johns Hopkins University Press. Dostoevsky, Fyodor 2004 The Double, London: Hesperus. Eysteinsson, Astradur 1990 The Concept of Modernism, Ithaca: Cornell University Press. Fludernik, Monika 1998 [1993] The Fictions of Language and the Languages of Fiction, London: Routledge. Georgakopoulou, Alexandra 2007 Small Stories, Interaction and Identities, Amsterdam and Philadelphia: John Benjamins Publishing Company. Grishakova, Marina [forthcoming] “The Voices of Madness: Performativity and Narrative Identity”, in: G. Rossholm (ed.), Disputable Concepts in Narrative Theory. Berlin and New York: Peter Lang. 2002 “Towards the Semiotics of the Observer”, Sign Systems Studies, 30.1, 529-553.

Toward a Typology of Virtual Narrative Voices

189

Herman, David 2009 “Beyond Voice and Vision: Cognitive Grammar and Focalization Theory, in: Hühn, P., W. Schmid and J. Schönert (eds.), Point of View, Perspective and Focalization. Modelling Mediation in Narrative, Berlin-New York: Walter de Gruyter, 119-142. Hydén, Lars-Christer 2009 “Broken and Vicarious Voices in Narratives”, in: L. C. Hydén & J. Brockmeier (eds.), Health, Culture and Illness: Broken Narratives, New York: Routledge, 36-53. Israel, Samuel Ben 2010 “Interaction Movies: Multiprotagonist Films and Relationism”, in: M. Grishakova, M.-L. Ryan (eds.), Intermediality and Storytelling, “Narratologia” series 24, Berlin and New York: Walter de Gruyter, 122-146. Jahn, Manfred 1996 Windows of focalization: Deconstructing and reconstructing a narratological concept, Style 30.2, 241-267. 2003 ‘Awake! Open your eyes!’ The Cognitive Logic of External and Internal Stories, in: D. Herman (ed.), Narrative Theory and the Cognitive Sciences, Leland Stanford Junior University: CSLI Publications, 195-213. McHale, Brian 2004 [1978]. “Free Indirect Discourse. A survey of recent accounts”, in: M. Bal (ed.), Narrative Theory. Critical Concepts in Literary and Cultural Studies, London-New York: Routledge, vol. I, 187-296. Miller, James 2001 “The Voice in Tourette Syndrom”, New Literary History 32, 519-536. Mukaőovský, Jan 1970 Aesthetic Function, Norm and Value as Social Facts, Ann Arbor: The University of Michigan Press. Nabokov, Vladimir 1965 The Eye, New York: Phaedra. 1989 Transparent Things, New York: Vintage International. Nielsen, Henrik Skov 2003 “The Impersonal Voice in First-Person Narrative Fiction”, Narrative 12.2, 133-150. Page, Ruth and Bronwen Thomas (eds.) [Forthcoming] New Narratives: Stories and Storytelling in the Digital Age, Lincoln: University of Nebraska Press. Peuter, Jennifer de 1998 “The Dialogics of Narrative Identity”, in: M. M. Bell and M. Gardiner (eds.), Bakhtin and the Human Sciences, London: SAGE, 63-77. Phelan, James 2001 “Why narrators can be focalizers—and why it matters”, in: Van Peer, W. and S. Chatman (eds.), New Perspectives on Narrative Perspective, New York: SUNY Press. Pilkington, Adrian 2000 Poetic Effects. A Relevance Theory Perspective, Amsterdam and Philadelphia: John Benjamins Publishing Company. Prince, Gerald 1988 “The Disnarrated”, Style, 22.1, 1-8. Richardson, Brian 2006 Unnatural Voices. Extreme Narration in Modern and Contemporary Fiction, Columbus: The Ohio State University Press. Ryan, Marie-Laure 1991 Possible Worlds, Artificial Intelligence, and Narrative Theory, Indiana: University of Bloomington and Indianapolis University Press. 2006 “From Parallel Universes to Possible Worlds: Ontological Pluralism in Physics, Narratology, and Narrative”, Poetics Today 27.4, 633-674.

190

Marina Grishakova

Schiffrin, Debora 1994 Approaches to Discourse, London: Wiley-Blackwell. Sclater, Shelley Day 2003 “What is the Subject?”, Narrative Inquiry 13.2, 317-330. Smith, Ali 2002 Hotel World, London: Penguin Books. 2005 Like, London: Virago. Sokolov, Sasha 1988 A School for Fools, New York: Four Walls Eight Windows. Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition, Oxford: Blackwell. Tammi, Pekka 2003 “Risky Business: Probing the Borderlines of FID. Nabokov’s An Affair of Honor as a Test Case”, in: P. Tammi and H. Tommola (eds.), Linguistic and Literary Aspects of Free Indirect Discourse from a Typological Perspective, Tampere: Tampere University Press, 4154. Tannen, Deborah 2004 “Talking the Dog: Framing Pets as Interactional Resources in Family Discourse”, Research on Language and Social Interaction, 37.4, 399–420 2007 [1989] Talking Voices: Repetition, Dialogue, and Imagery in Conversational Discourse, Cambridge and New York: Cambridge University Press. Tröhler, Margrit 2010 “Multiple Protagonist Films. A Transcultural Everyday Practice”, in: J. Eder, F. Jannidis, R. Schneider (eds.), Characters in Fictional Worlds. Understanding Imaginary Beings in Literature, Film, and Other Media, Revisionen. Grundbegriffe der Literaturtheorie 3, Berlin-New York: De Gruyter, 459-477.



MARIA MÄKELÄ (Tampere)

Masters of Interiority Figural Voices as Discursive Appropriators and as Loopholes in Narrative Communication Introduction: Narratology and the Easy Access Fallacy Narratorial voices in literary fiction may be strange—as demonstrated in many articles of the current volume—but never quite in the same manner as figural voices are. Here I mainly refer to the unuttered, internal “voices” of story-internal literary characters that are conveyed to us through multiple means of consciousness representation. The long history of discourse narratology attests to the fact that the figural voice is even more perplexing than the narrative voice. Yet this is strangeness in respect to linguistic and cognitive based theoretical categories, not strangeness before the reading audience—quite the contrary, a lay reader is perfectly comfortable with passages of narrative literature exposing a hidden discursive agency of a story-internal character. This type of narration—third-person past-tense narratorial report intermingled with recurrent outbursts of characters’ language—has been exhausted by popular fiction from hardboiled crime stories to Harlequin romances. Nothing strange here. Yet, such double-voiced or double-intentional narrative discourse has inspired some of the most prominent narratologists (most notably McHale 1978a, Cohn 1978 and Fludernik 1993), as well as evoked some illustrious theoretical controversies (e.g. Banfield 1982 vs. McHale 1983; Miller 1988, Seltzer 1984 & Bender 1987 vs. Cohn 1999 [1995]). Despite of the fact that this mode has been through several narratological redefinitions and contestations, it is still best known by the name of free indirect discourse (FID), which has its origins in the linguistic “speech categories” (cf. Palmer 2004)—long since discarded as insufficient and replaced by contextual (McHale 1978a) and cognitive (Fludernik 1993) approaches. The persistence of the linguistics-originated term perhaps hints at the unsettling effect that this mode has on the relationship between verbalization (language) and discursive agency (“mind”).

192

Maria Mäkelä

The paradigm shift in narratology from classical, linguistic-based model building to the postclassical variety of approaches has not significantly altered the narratological take on figural voice. Are you reading the narrator’s voice or the character’s voice, or both at the same time? A considerable amount of narratological work has been dedicated to isolating the alleged “voices” of narrative agents from the discourse of narrative fiction. In order to put one’s finger on the discursive “self” of the character a diligent student of narrative discourse immediately starts to dissect the discourse and label some parts as narratorial (the informative, the authorial), other parts as figural (the colloquial, the expressive). However deeply rooted in narratology this separation procedure might be, one question remains largely unanswered and even unaddressed in both classical and postclassical studies: whence all the ambivalence between the narratorialobjective and the figural-subjective? Narrative instances that continue to trigger the narratological imagination are those that seem not just to represent alternating voices, but also to amalgamate narratorial report with the unuttered voice of the figural consciousness. Even if we were to ultimately “reestablish” individual figural and narratorial voices from ambiguously voiced discourse, would not these individuated discursive selves bear in themselves the traces of the onetime discursive union? Now I must hasten to add that the narratively and thematically productive effects of this ambivalence have, indeed, been widely studied as narratorial functions. Roy Pascal’s seminal study on FID (The Dual Voice, 1977) considered the mode mainly as a vehicle for either narratorial empathy or irony, and this well-argued interpretation was echoed in several subsequent studies (e.g. Cohn 1978, Aczel 1998, Gunn 2004). Monika Fludernik’s massive study on speech and thought representation (1993) continues in the same vein of regarding ambivalence of discursive agency mainly as narratorial mimicry: in Fludernik’s theory, figural voice is but a “linguistic hallucination” and reducible to the narratorial voice. Yet how about figural functions? Can we assign any discursive intentions to the silent reflector-characters whose inner flow we nonetheless “hear”? Another strand in studies of free indirect (or just fuzzy) discourse, the one that foregrounds the figural component, has been dominated by ideological readings: Kathy Mezei (1996) sees an emancipatory potential in FID as a mode which allows the characters to “have their say,” more supported than restricted by narratorial framing; conversely, the Foucauldian readers of FID such as Mark Seltzer (1984), D. A. Miller (1988) and John Bender (1987) regard the mode as a form of ideological oppression, and as inheriting a biased power relation for the benefit of the authoritative narrator

Masters of Interiority

193

who is able to survey the characters’ mental life (especially inner discourse) with its non-reciprocal gaze. What, all in all, seems to characterize those approaches—whether they foreground narratorial functions or emphasize the (emancipatory or oppressive) transparency of characters’ internal discourse—is to automatically regard narratorial discourse as a representative of narrative mediacy and character’s discourse as an access to the immediate experience or impression. If we were to take at a face value all the hierarchically driven efforts for analyzing narrative discourse during the past four decades, I think we could actually agree with the Foucauldian theorists in that the literary character appears, indeed, as an object of narrative-discursive oppression, as the unfortunate passive creature at the end of the narrative food chain. The roots of this preconception are traceable all the way to Plato-Aristotelian distinction between diegesis and mimesis: to the dichotomous conception that separates telling (narration as mediation) from showing (the experience of immediacy). The only theorist to fully acknowledge the rhetorical potential of figural interiority is the one who committed what according to his successors seems to be a crucial taxonomic blunder: for Booth, a reflector qualifies as an unreliable agent just as any first person narrator—the focalizers of Mrs. Dalloway appear in the same record with Holden Caulfield (see Booth 1991: 493-494, “A Gallery of Unreliable Narrators and Reflectors”). Might it even be that this alleged misunderstanding in the theory of narrative rhetoric in fact reveals the counter-intuitivity of the Genettean distinction between speakers and perceivers? Clearly, Booth is inclined to ascribe communicative features to figural internal discourse, but does not really address the differences between unreliable narration and unreliable focalization; he does not problematize the pseudo-communicative quality of consciousness representation. This productive blunder lives on in Tamar Yacobi’s formulations on “fictional mediation” and unreliability that place reflectors and narrators on the same axis of the “mediation-gap”, the perspectival distance between the fictional mediator and the author (Yacobi 1981 and 1987). Yet also Yacobi—herself an observant critic of theoretical “package-dealing” (see Yacobi 2001)—settles for the evident (“natural”) package-deal as far as the position of the silent reflector/monologist is concerned: inner speech in fiction is, by definition, unconsciously communicative (Yacobi 1987: 338). Yet, one might wonder whether the reader’s sense of being led on by a focalizer would not also insinuate rhetorical intention. We may also be reminded of Franz K. Stanzel’s classical study on narrative mediacy (1984), where also the character’s perceptual and discursive presence has mediating functions: the “figural narrative situation” or “re-

194

Maria Mäkelä

flector-mode narrative” defined by Stanzel are narrative instances where mediacy brought on by the narratorial voice is replaced by mediacy via figural consciousness. However, also Stanzel’s formulations reveal the underlying discordance: Realistic presentation of consciousness seems to require the illusion of immediacy, that is, the apparent suspension of mediacy, more than does presentation of external events. The modern novel especially shows a very pronounced tendency to give the presentation of consciousness the semblance of immediacy, of the unedited and the spontaneous. […] Interior monologue, free indirect style [FID] and figural narrative situation, that is, the forms of the reflector-mode and of internal perspective, suggest immediacy, that is, the illusion of direct insight into the character’s thoughts. (Stanzel 1984: 127)

Why does Stanzel speak of “the illusion of immediacy […] the apparent suspension of mediacy” without specifying the illusory quality of the mimesis of the mind?1 Everything is illusory in fiction; what is it that is specifically illusory in our sense of immediacy when reading fictional minds? Stanzel’s most appreciative follower and interpreter Monika Fludernik also pays abundant attention to the “evocation of figural voice.” Yet both her theoretical assertions and textual analyses are inclined to downplay the discursive agency of a figural consciousness and reduce multi-voicedness to narratorial functions of empathy, irony, and stylistic parody. In her vocabulary of ambiguous discursive agency, both “reflectorization” and “figuralization” are varieties of narratorial mimicry or appropriation of figural (expressive and deictically marked; individual, collective, or impersonal) voice (see Fludernik 1996: 178-221). Indeed, she dispenses with the dualvoice approach on the level of language and narrative technique and considers ambiguity of voices only an interpretive effect (see, e.g. Fludernik 1993: 322-356). Yet one of the cornerstones of Fludernikian natural narratology (1996) is an unlimited access to figural experience via, among other narrative elements, discursive markers of expressivity—qua indicators of experientiality. In cognitive terms, the figural narrative situation defined by Stanzel requires a deictic shift inside the character’s experiential plane: Telling can be dispensed with, readers simply orient themselves to a position within the fictional world […]. Such a reading experience is structured in terms of the natural frame of experiencing, which includes the experiences of perception, sentiment and cognition. Real-life parameters are transcended. Instead of merely observing and guessing at other people’s experiences, frames naturally available only for one’s own experience become accessible for application to a third person. (Fludernik 1996: 48)

1

Alongside with Stanzel, also Fludernik (1996: 48) refers to the reading of fictional minds as the “willing suspension of disbelief”.

Masters of Interiority

195

These “frames […] available […] for one’s own experience” must surely include discursive frames as well. Yet both Fludernik as well as other cognitive narratologists who emphasize the importance of a deictic shift for the readerly immersion in the figural experience tend to treat the discursive shift from narratorial to figural as a mere catalyst. However, the ultimate frame for interpretation is, and remains, discursive. For Fludernik, the natural frames of TELLING (domination of narratorial discourse) and EXPERIENCING (immersion in figural experience) are basic units on the level of reading; she does not, however, consider the possible overlapping between these two frames in cases when the reader is confronted with a figural narrative situation that displays diegetic qualities or appropriates narratorial functions. In the notoriously double-voiced narration of Madame Bovary (1857), we find a sentence illustrating Emma’s restlessness after Léon has left her alone with the stultifying company of her disappointing husband and the rest of the Yonville: (1) Now the bad days of Tostes came back again. This time she thought herself far more unhappy: for she was experienced in sorrow, with the certainty that it would never end. Any woman who had imposed such great sacrifices on herself could well be permitted a few fancies.2 She bought a Gothic prie-dieu, and in one month she spent fourteen francs on lemons for cleaning her nails […] (115; italics mine)

The cited passage opens with an undisputable occurrence of psychonarration (indirect “thought report”; see Cohn 1978: 21–57; Palmer 2004: 75–80), indicated by a linguistic marker of indirectness (“she thought herself...”).3 However, already the second sentence leaves the intentional stance somewhat ambivalent: who is of the opinion that Emma is “experienced in sorrow”? When we reach the third sentence (“Any woman who...”) we are ready to give up insistence on linguistically marked speech categories, since what we are dealing with must be free indirect discourse camouflaged as narratorial judgement.

2 3

In the French original: “Une femme qui s’était imposé de si grands sacrifices pouvait bien passer des fantaisies”. (217) As to Cohn’s term “psycho-narration” and Palmer’s notion of “thought report” it should be noted that both theorists in fact express a distrust of figural verbalizations of experience, which is partly in tune with my own assertions. For Cohn, it is precisely the narratordominated psycho-narration that allows for the most penetrating look “inside” fictional minds since the mode is independent of figural capacities of verbalization and introspection (Cohn 1978: 46, 56, 139–140). One crucial point in Palmer’s critique of the classical speech category approach is the rehabilitation of the indirect modes of consciousness representation that, according to Palmer, are capable of rendering the non-verbalized component of fictional minds: for Palmer, consciousness representation involves “the whole mind” in its “social and physical context.” (Palmer 2004: 76)

196

Maria Mäkelä

Already the classical definitions of FID (by, for example, McHale 1978a and 1983, Pascal 1977, Ginsburg 1982) include such instances of discursive double-intention as in the sentence “Any woman who...”: the utterance does not display any overt, linguistic or expressive markers of the character’s inner discourse but reads as a subjective construction of the storyworld only when considered in its narrative and thematic context (the extralinguistic markers that guide our contextual interpretation of the discursive agencies or voices behind the surface of language). After the breakthrough of ‘Natural’ and cognitive narratology, perennial questions raise their heads: Why give the subjective perspective of a character the persuasiveness of an objective report? Why must the identification of the subject behind the discourse necessarily be based on all sorts of extra evidence? Where is the fictional voice, if not in the utterance itself? I think that it is precisely at the face of such narrative instances as the Bovary example above that the cognitive-narratological approaches fall short, partly because of their insistence on such natural categories as TELLING and EXPERIENCING, partly because of their conviction that as all narratives, also any literary narrative is mainly concerned with qualia, of the “raw feels” of an individual, of the “what is it like to be x” quality of human experience (see esp. Herman 2007: 256–257). In these exhaustive approaches to mind and narrative the peculiarities of figural voice dissolve into “natural” readings: problematic discursive agency is overlooked in favour of establishing a firm experiential plane from which the feeling of being-in-the-fictional-world can emanate. Yet there is nothing raw in the sentence “Une femme qui s’était imposé de si grands sacrifices pouvait bien passer des fantaisies,” a sentence, which read in context is wrought with subjectivity. We do not enter Emma’s consciousness all the way through; instead, we bump into an inner persuasion process, which, being verbalized in narrative discourse, comes disturbingly close to resemble an objective narratorial report—or even an authorial gnomic statement on how things are in the world. The passage seems to require more than an alternating application of the TELLING and the EXPERIENCING frame; we must appreciate literary narrative’s capacity to activate these frames simultaneously—as well as the resultant unreliability of both frames. In the following, I wish to argue that ambivalence of voices should not merely be reduced to either narratorial functions or to a challenge to dig out the truth about a character’s truest motives, intentions or emotions. Ambiguous discourse may also give rise to figural takeovers that cannot be explained away by cognitive or rhetorical approaches. It will be argued that it is precisely the multi-layered communicational and discursive structure of narrative fiction that provokes the violations of that structure; the literary hierarchy of voices has inspired the writers as well as

Masters of Interiority

197

the readers to explore the nexus between mind, language and narrative in ways that are not completely compatible with our everyday cognitive mechanisms of understanding the world and each other. I will try to make my point with a simulated diachronic shift in the history of literary representation of figural voice. I will start with a couple of canonical examples from epistolary narration to highlight the fact that figural voice is much more than a literary extension of the theories and conceptions concerning the “real mind” or human subjectivity in general. I argue that, by having recourse to the epistolary first-person forerunners of figural narrative situations, we are able to trace back the literary evolution of the “strangeness” of figural voice-—a strangeness that, at least partly, originates from a mind that is both writing and written. From epistolary fiction I move on to a very limited corpus of examples displaying ambivalent focalization and stylistically unmarked free indirect discourse in third person contexts. Fictional minds will be considered as displaying a distortedly dualistic angle to literary experientiality: (1) the allegedly represented consciousnesses appear as discursively mastering their own minds, appropriating narratorial functions and thus playing the “willing suspension of mediacy” game on us readers; yet simultaneously (2) such diegetic mind games undermine fundamentally the entire mimetic agency of these figural consciousnesses and their existence as anything else but writing. Epistolary Minds: Mediated Immediacy (2) O my dearest Father and Mother, LET me write and bewail my miserable hard Fate, tho’ I have no Hope that what I write will be convey’d to your Hands! – I have now nothing to do but write, and weep, and fear, and pray; and yet, What can I pray for, when God Almighty, for my Sins, to be sure, vouchsafes not to hear my Prayers; but suffers me to be a Prey to a wicked Violator of all the Laws of God and Man! —But, gracious Heaven, forgive me my Rashness! O let me not sin against thee; for thou best knowest what is fittest for thy poor Handmaid! (98)

Here, to be sure, we have a beautiful exemplar of a fictional figural voice in distress—unmediated by any narratorial framing or temporal distance. Similarity to an unuttered voice of third person consciousness representation is made even more evident by the “eclipse of the confidant,” a conventional rhetorical move in the epistolary novel: the writer’s (the narrator’s) discourse turns inward and the locus of the receiver (the narratee) appears to be empty. As Janet Gurkin Altman notes, such an eclipse is likely to occur whenever the letter-writer is going through an emotional turbulence. (Altman 1982: 57-59) In the above example, the disappearance

198

Maria Mäkelä

of the audience is made literal since the poor Pamela, kidnapped by her tormentor Mr B, will not be able to send the letter to her parents. Moreover, the heroine is convinced that even God himself ignores her prayers. Yet Pamela’s outburst cannot be considered as completely gratuitous, since the feminine master plan, which ultimately dominates over Mr B’s stratagems, relies on the fact that Pamela knows Mr B to be reading her letters. A gap opens up for a non-mimetic reading of consciousness: the immediate distress appears as a discursively mediated fabrication. The pioneering role of epistolary narration in consciousness representation has been only fleetingly considered by postclassical narratologists such as Fludernik (see 1996: 48), Zunshine (see 2006: 86), or Palmer (see 2004: 242-243). Moreover, if mentioned at all, epistolary narration’s contribution to the evolution of fictional minds is considered to be it’s allowing of an “immediate access” to figural consciousness. Such a take on the epistolary clairvoyance echoes the words of the 18th century masters themselves, Samuel Richardson’s famous and triumphant characterization in his preface to Clarissa (“the only natural Opportunity […] of representing with any Grace those lively and delicate Impressions which Things present are known to make upon the Minds of those affected by them”; cit. McKeon 1997: 259) as well as Samuel Johnson’s praise of Richardson’s ability to “dive into the recesses of the human heart” (cit. Watt 1984: 261). A notable exception to this psychologically mimetic approach is offered by Joe Bray, whose study on the epistolary novel draws out the discontinuity between the perceiving self and the perceived self (Bray 2003: 16 passim)4; between the distressed captive Pamela and the Pamela who verbally constructs herself as her Lord’s poor Handmaid. The outcome is a disturbing mixture of hysterical expressivity and persuasive rhetoric. According to Bray, “the impossibility and unseizability of the epistolary present” as well as the “constant interaction between the narrating self and the experiencing self” in the early modern epistolary fictions foreshadow the ambiguous voice and identity games (such as free indirect discourse) between the heterodiegetic narrator and the diegetic character displayed by later, realist and modernist novels (ibid. 19–28). Indeed, we may notice how some modes as well as recurring themes of epistolary narration seem to controvert the entire notion of figural voice as immediacy—let alone as quasi-authentic “experientiality”—or as a simulation of “raw feels.” In Richardson’s Pamela, Mr. B. repeatedly refers to Pamela as an “artful slut”; in a deceitful letter to Pamela’s father, he 4

Also Janet Gurkin Altman’s classical study (1982) dwells on the unnatural aspects of the epistolary mode as well as points out some recurrent epistolary homologies between modes of writing and thematic patterns.

Masters of Interiority

199

makes an observation which many interpreters of the novel find to be quite apt: “with all her pretended Simplicity and Innocence, I never knew so much romantic Invention as she is Mistress of” (Pamela, 93). In fact, most of the literary letter writers are artful sluts. This can be said even of the very trendsetter of epistolary passion, sister Mariana in Guilleragues’ The Portuguese Letters (Les Lettres Portugaises, 1669), whose verbalized anguishes have served as inspirations and examples for letter writers, both fictional and real, for centuries. As Joe Bray remarks, Mariana’s frantically emotional letters to the deceitful lover were read as close simulations of “natural” feminine passion and as models on how to verbalize unpremeditated sensations (Bray 2003: 29-30). Yet, as Bray goes on to claim, the letters display order in their disorder and reflect distance in their expressivity (31-32). And when Mr. B refers to Pamela as a “mighty letter-writer,” he does not only point to the unnatural volume of those letters but to the improbable skill they exhibit in their power to construct and manipulate interiority. Already this narrow evidence suggests that even the early exemplary texts of epistolary fiction foreground themselves the linguistic and narrative mediacy of the immediate. Yet, in the letters of Pamela and sister Mariana, the verbal mannerisms or the discursive façade of an ingénue figure are ultimately all that we have—these sentences form the figural interiorities and the literary experientiality that we readers are so keen on capturing. This paradox leads us to my central assertion: that the syntheticity of a textualized “inner voice” is an elementary part of literary fiction and does not get lost in translation from the first person epistolary form to the focalized third person narration of mainstream psychological realism. The earlier mentioned study by Joe Bray examines carefully the shaky relationship between the experiencing self and the writing self and suggests that this shakiness gets transferred into the ambiguous relationship between the heterodiegetic narrator and the experiencing character. However, I think something elementary gets lost if we—again—consider the problem of verbalization always to come down to the narrator-character controversy; I would claim that the figural voice’s internal controversy lives on, just as well in the third person context as in the hysterical and yet well-composed voices of epistolary heroines. If we take this road (chose to emphasize figural functions over narratorial functions) we may conclude that in many narrative instances, the most productive ambivalence lies between the “voice” of an unmediated experience and the unavoidable sense of premeditation brought on by language, intentional structure and communicativeness. An illustrative example of bringing this ambivalence to an overtly thematic level is Choderlos de Laclos’ Les Liaisons dangereuses (1782), where the notorious seducer Valmont makes an art of verbally fabricated pas-

200

Maria Mäkelä

sions. In letter No. 70, he describes the production of an emotionally charged and seemingly incongruous letter (No. 68) to Madame de Tourvel: (3) I have therefore declined her precious friendship and insisted upon my claim to the title of lover. Since I am under no illusions as to the real importance of securing this title (though it might appear at first to be a mere quibbling about words), I took great pains with my letter and attempted to reproduce in it that disorder which alone can portray feeling. I was, at all events, as unreasonable as I was capable of being: for there is no showing tenderness without talking nonsense. It is for this reason, it seems to me, that women are better writers of love-letters than men. (150)

Valmont’s calculated evocation of a figural voice and the disclosure of his methods form the most crucial metatext for the entire novel: the Richardsonian dive “into the recesses of the heart” proves a mere illusion produced by writerly conventions. Furthermore, Valmont’s perception of (feminine) epistolary passions seems to suggest that there is always a method in textual madness; even in the final letter of the hysterical Madame de Tourvel, dictated from her death bed and addressed to no one particular. Her unrestrained flow of thought speaks to Valmont (“Cruel and malignant man…”), to her husband (“Return and punish an unfaithful wife”) and to no one (“Where are the friends that love me, where are they? […] No one dares come near me”). In Tourvel’s letter, the irrationality of reference completes the eclipse of the (anonymous) confidant. And yet, it is in this letter where we find epistolary expressivity reaching its peak. Here we get to one of the greatest paradoxes of consciousness representation: how can one express anything without an audience? Through this intrinsic turn in narrative communication the epistolary novel ushers in a new kind of figural voice. This voice represents a noncommunicative and yet speaking subject and it comes to be thoroughly conventionalized by later novelistic techniques such as internal focalization, free indirect discourse and stream-of-consciousness narration. In terms of the narratological easy access fallacy, the expressivity of language gives rise to an uncensored figural experience. Yet something gets thrown away with the bathwater when narratologists turn from letters to consciousness representation: the intentionality behind an utterance, the pseudo-communicative nature of figural voice, the artfulness with which the unuttered emotion, experience or thought is formulated. Yet this ambivalence between immediacy and mediacy of expression (between Tourvel and Valmont) is already thematized in epistolary fictions: the experiential logic of epistolary narration is based on double-dealing between intentional structure (“art”) and free expression (“life”). This controversy is the most salient legacy that the epistolary novel of the early modernity has passed on to our days: the notion of a constructed and mediated nature of written experience – not that of an immediate and amorphous one.

Masters of Interiority

201

When reading Laclos or Richardson, we may discern how the evocation of figural immediacy is deconstructed and the entire communicational design seems to loop from the rendering of “true” subjectivity back to the origins of fictional representation—that is, to authorial design. As Altman (1982: 105) notes, a recurrent theme in epistolary novels are letters as selffulfilling prophesies; the extratextual events are often scripted and manoeuvred in letters that, at least for the unsuspecting diegetic reader, first appear as pure expressions of emotion. Here is Pamela’s pen at work again: (4) the unparalleled wickedness, stratagems, and devices, of those who call themselves gentlemen, yet pervert the designs of Providence in giving them ample means to do good, to their own everlasting perdition, and the ruin of poor oppressed innocence! (83)

Indeed, Altman parallels Pamela’s “artfulness” and Mr B’s ambivalent role play as her “reader” with the more explicit masterminding of Laclos’ characters: “B’s consciousness of Pamela’s letters as her “novel,” their novel— which he reads and yet is an agent in—resembles Valmont’s and Merteuil’s awareness of themselves as creators and readers of their own story” (Altman 1982: 105). The above example (4) shows Pamela (again) in distress, and yet at her most authorial, occupying the discursive locus of an 18th century intervening and judging narratorial voice. As a discursive Trojan horse, she appropriates the pre-modern narratorial convention of an exemplum and constructs herself as the proverbial virtuous victim while simultaneously representing one. Consequently, the elements that narratologists lump together under the term “expressivity”5 such as exclamations (“O…!”) and the breathless syntax reappear to us as means to a narrative end. One is led to ask: why would all the “artful sluts” stop discursive manoeuvring when cast into the third person context, even though this modern position might—at a first glance—imply the narrative role of a mere innocent victim (of representation)? Unmarked FID: Downplaying of Discursive Agency and Appropriation of Narratorial Conventions Madame de Lafayette’s La Princesse de Clèves (1678), the allegedly first “psychological novel”, plays the role of an intermediary between epistolary forms and modern consciousness representation. Indeed, the novel itself seems to be a work in progress, representing concretely the historical tran5

On the slackness of the narratological uses of this term, see Jongeneel 2006.

202

Maria Mäkelä

sition from the external to the internal focalization and from first person to third person experientiality. We may find one striking example that displays this change in an ingeniously formed nutshell: a letter, followed by an ancestral occurrence of free indirect discourse. The case is that of an anonymous letter which circles around the court of Henry II and arouses admiration and upheaval; the letter is mistakenly thought to be addressed to the Duke of Nemours by a jealous mistress; and it misleads Madame de Clèves to read it as an evidence of Nemours’ false-heartedness in the affections he has been showing to her. (5) [From the anonymous woman’s letter:] never was grief equal to mine; I thought you had the most violent passion for me, I did not conceal that which I had for you, and at the time that I acknowledged it to you without reserve, I found that you deceived me, that you loved another, and that in all probability I was made a sacrifice to this new mistress. […]I was of opinion that if anything could rekindle that flame, it would be to let you see that mine was extinguished, but to let you see it through an endeavour to conceal it from you, as if I wanted the power to acknowledge it to you: this resolution I adhered to; but what a painful decision and how difficult, once I had seen you again, to put into practice! (59) Madam de Clèves read this letter, and read it over again several times, without knowing at the same time what she had read; she saw only that the Duke de Nemours did not love her as she imagined and that he loved others who were no less deceived by him than she. What a discovery was this for a person in her condition, who had a violent passion, who had just given marks of it to a man whom she judged unworthy of it, and to another whom she used ill for his sake! Never was affliction so cutting as hers […] in short, she thought of everything that could add to her grief and despair. What reflections did she not make on herself, and on the advices her mother had given her! How did she repent, that she had not persisted in her resolution of retiring […]. (60–61)

The content—and more remarkably, the form—of the letter are perplexing enough to give rise to a pioneering narrative mode such as FID. Since this is but an early manifestation of the mode, the extradiegetic narrator’s empathetic voice (or psycho-narration /thought report) dominates—“What a discovery was this…!” —and we may sense how the figural voice emerges only gradually from the narration: “… and to another whom she used ill for his sake!”. It is almost as if we were let to read how FID is born as a result of discursive pairing between the authorial and the figural; springing from mere hypotheses made by the narrator, then growing towards the narrator’s empathizing and imitating exclamations—and gradually evolving into a discourse that actually reflects the form of thoughts as they appear to the experiencing mind. Yet I would also claim that it is a justifiable interpretive move to foreground the figural discursive component from the very beginning. It is the

Masters of Interiority

203

preceding letter that provides us with the most crucial frame: if we compare the letter’s syntax and the expressions to those in the following third person passage, we may notice striking similarities: [The letter:]

[The reaction:]

never was grief equal to mine Jamais douleur n’a été pareille à la mienne.

never was affliction so cutting as hers Jamais affliction n’a été si piquante et si vive.

I thought you had the most violent passion for me, I did not conceal that which I had for you, and at the time that I acknowledged it to you without reserve, I found that you deceived me, that you loved another Je croyais que vous aviez pour moi une passion violente; je ne vous cachais plus celle que j’avais pour vous et, dans le temps que je vous la laissais voir tout entière, j’appris que vous me trompiez, que vous en aimiez une autre this resolution I adhered to; but what a painful decision and how difficult, once I had seen you again, to put into practice! Je m’arrêtai à cette resolution; mais qu’elle me fut difficile à prendre, et qu’en vous revoyant elle me parut impossible à executer!

who had a violent passion, who had just given marks of it to a man whom she judged unworthy of it, and to another whom she used ill for his sake! qui avait une passion violente, qui venait d’en donner des marques à un homme qu’elle en jugeait indigne et à un autre qu’elle maltraitait pour l’amour de lui!

What reflections did she not make on herself, and on the advices her mother had given her! How did she repent, that she had not persisted in her resolution of retiring. Quels retours ne fit-elle point sur elle-même! quelles réflexions sur les conseils que sa mère lui avait donnés! Combien se repentit-elle de ne s’être pas opiniâtrée à se séparer du commerce du monde.

In fact, it seems that Madame de Clèves absorbs the art of the letter-writer to convey her own experience—the alleged uniqueness of it (“Jamais affliction n’a été si piquant et si vive”). Yet, ironically, we as readers see the two women’s anxieties as juxtaposed—and as identical (“Jamais douleur n’a été pareille à la mienne”). Their passions are not unique but schematic, at least when verbalized. The third person consciousness representation imitates the passionate syntax of the letter as well as the exclamation “Combien/Que…!”—a structure that later became a conventional marker of free indirect discourse in French. In this early context, however, the sentences such as “Quelle vue et quelle connaissance…” or “combien se re-

204

Maria Mäkelä

pentit-elle…” remain doubly-voiced in a very peculiar manner because they foreground both the figural expressivity and the hypothetical nature of figuralization. We could actually interpret the narrator as asking, rhetorically: “combien – “, “quelle vue – “.6 Ultimately, the ambivalence between the narratorial and the figural can be seen as culminating in the sentence “never was affliction so cutting as hers.” It offers the reader a seemingly objective report of the unequalled emotional load of the heroine—of a woman throughout the novel characterized as, indeed, a sui generis case both in appearance as in character. However, in its narrative context, the sentence gains in figural expressivity at the expense of objectivity. The letter’s exemplary dame in distress suggests a reading where the incomparability and uniqueness of Madame de Clèves’ shock is the heroine’s own interpretation—just as the following sentences are the result of her misguided conclusions (we learn later that the letter was not addressed to Nemours and all the emotional fuss was unnecessary—a fact that the chronicler-narrator of La Fayette was well aware of all along). Moreover, such wavering between uniqueness and schematicity, between expressivity and authorial design is one of the trademarks of the courtly (literary) traditions in which La Fayette herself lived and from where she found her inspiration. As Mary Jo Muratore has noted in connection of her La Fayette study, the courtly life of the 16th and the 17th century was dominated by exemplaires: the perfection of one’s conduct as well as of one’s inner life was to be accomplished by imitating superior examples (Muratore 1994: 94–95). In La Princesse de Clèves, the virtuous heroine who, even after her husband’s death, rejects Nemours and dies in a monastery, becomes the absolute “exemple inimitable.” Yet, before that it is the letter-writer who sets the literary example for passion and true interiority: the found letter is introduced to the court as a masterpiece both in writing and in sentiment, as “the finest letter that was ever writ”. The art of emotion is reduced to the art of writing, and when transferred (or copied) into the third-person context of Madame de Clèves’ affliction, this art of verbally mastering one’s interiority translates into an uncanny figural voice, which is simultaneously the producer and the result of consciousness representation. 6

Indeed, a post-Lafayettean pattern of emotional letters giving rise to turbulent free indirect discourse can be traced from Goethe’s Elective Affinities (1809) and Austen’s Pride and Prejudice (1813) to Pushkin’s Eugene Onegin (1828-32). Pascal 1977: 39 offers an illustrative example from Goethe (but does not consider the connections between FID and the preceding epistolary expression): “This last expression flowed out of his [Eduard’s] pen, not his heart. Yes, as he saw it on paper he began bitterly to weep. In one way or another he was to renounce the happiness, yes the unhappiness, of loving Ottilie!”

Masters of Interiority

205

We may also be reminded of Madame de Clèves’ successor Emma Bovary and our introductory example (2) where the sentence “Any woman who had imposed such great sacrifices on herself could well be permitted a few fancies” occurs. Might there be something similar going on in La Fayette’s narration, even in the seemingly most objective evaluations such as “What a discovery was this for a person in her condition…!”? Another point of contact can be found in the example (4) where Pamela writes about Mr. B’s “stratagems […] to the ruin of poor oppressed innocence". Just as Pamela takes on an authorial tone in the midst of persecution and crisis, also Madame de Clèves and Emma gain authority for their subjective constructions by applying the narratorial convention of making a character (themselves) an exemplum. All the three heroines are “Mistresses of romantic Invention,” as Richardson’s Mr. B would put it, and derive the language of their experience from literary exempla; yet only one of them is writing. The rub with FID is, as noted by several literary theorists, that more than often it appears as unmarked by linguistic or expressive signals, and yet does not go unnoticed by the readers. As Cohn herself points out, FID is still relatively easy to tell apart from other modes of consciousness representation but often impossible to distinguish from the allegedly reliable narratorial report. (Cohn 1978: 106) Sentences such as “Never was affliction so cutting as hers” or “Any woman who had imposed such sacrifices on herself…” jumble the logic of mimesis and diegesis, of figural and authorial (cf. Ginsburg 1982 and Ron 1981). What has gone largely unnoticed is the fact that narrative literature teems with “artful bitches”— characters in the diegetic fictional worlds—who take advantage of this available discursive loophole in order to establish their own idiosyncratic version as a preferable truth. Roy Pascal’s commentary on what he calls “narrative usurpation” crystallizes the uneasiness that narratologists experience with such disrespect towards the hierarchy of voices. When analyzing consciousness representation in Madame Bovary, Pascal fixes his critical eye on descriptions of Emma’s sensations that are too artistically accomplished to render a realistic image of Emma’s imagination. (Pascal 1977: 103–111) Furthermore, when approaching modernist texts, Pascal recognizes the opposite tendency: to give the figural voice the persuasiveness and apparent authority of narrator’s discourse. Yet Pascal is reluctant to theorize this phenomenon any further in the framework of his “dual voice hypothesis”. (Ibid. 108) As McHale (1978b: 400) notes, Pascal fails to appreciate Flaubert’s resistance to “the interpretive strategy by which we assign impressionism to a character’s vision”. As I have been trying to demonstrate, this narratological disinclination towards thematically potential narrative takeovers is mani-

206

Maria Mäkelä

fest in its reverse meaning as well, as the “easy access fallacy”. Yet the authorization and pseudo-communicativeness of figural voices are not merely stylistic blunders that dilute the effects of psychological realism: they are the paradoxical essence of psychological realism. Such downplaying of discursive agency is thematically and formally titillating precisely because it momentarily deconstructs what is conventionally regarded as the communicational hierarchy of the literary narrative. Furthermore, these loopholes are in a contrapuntal relationship to the evolution of recurrent literary themes: the double standards of virtue (Pamela); the conflict between individual experientiality and shared conventions of narrativization (La Princesse de Clèves); or the anxiety of language and literary communication (Madame Bovary). Thus the impression of a figural consciousness “writing their own novel” should not be regarded merely as a discursive side effect; on the contrary, the figure of the pseudo-hysterical “artful slut” or that of the “(over)narrativizing focalizer” are recurrent literary tropes produced in the interplay between narrative techniques and thematic foregrounding. Every time one addresses overlapping discourses in a novel one name should—and will—come up: the indie-theorist Bakhtin is closely connected to the arguments I am defending in this article. For Bakhtin, as well, novelistic techniques are deeply rooted in writing—and silent reading which leaves the construction of voice(s) to the reader (Bakhtin 1981: 3; see also Lock 2001: 7-77).7 It has been suggested that Bakhtin is, in fact, more “graphocentric” than Derrida (Robert Cunliffe, cited in Lock 2001: 71). Yet what prevents me from simply embracing Bakhtin’s dialogism and accepting a sentence such as “Any woman who had imposed such great sacrifices on herself could well be permitted a few fancies” as a mere locus of an eternal heteroglossia, of an unresolved play of subjectivities (cf. Bakhtin 1981: 324-325; “another speech in another’s language”; see also Grishakova in this volume) is that the Bakhtinian angle to doublevoicedness ignores the interplay between the “freedom” of free indirect discourse and the suggested hierarchy of narrative agents that contravenes it. In my view, the most significant novelistic dialogue is conducted between the possibility and the impossibility of dialogue between narrative agents on different diegetic levels; it is the hierarchical backdrop of novelistic communication against which the discursive freedom—or the unencumbered nature of the written word—is played. Consequently, as much as I sympathize with Lars-Åke Skalin’s arguments in the present volume 7

The Bakhtinian notions about the novel as an emphatically written genre tie in with the early theories of free indirect discourse as “unspeakable” and as such (exclusively) a novelistic mode (see Bally 1912 and 1914, Vološinov 1986: 136-159, see also Lock 2001: 80-83).

Masters of Interiority

207

on the peculiarity of fictional (non-)communication, I cannot fully embrace his notion of third-person literary narratives as elementarily tellerless: the uncanny intentionality and rhetorical design of figural voices is created in a dynamic dialogue with the constructed narratorial position. Furthermore, the thematic potential of narrative usurpation and exploitation attests to the readerly intuition of an underlying structure of voices. To interpret Emma Bovary or the Princess of Clèves as “artful sluts” reaching for narrative authority is to presuppose an “upper” and a “lower” discursive agency (the authority of the narratorial report and the controversiality of local truths). In short, we cannot say that literary discourse or literary interpretation is a play if we do not recognize the bricks with which one plays; and the alleged hierarchical structure, to my mind, is one of the most fundamental bricks. At this point it may be warranted to turn to a late modern text which obviously flaunts the figural discursive takeover, bringing heterodiegetic narratorial functions to the verge of omission—and we should turn to this narrative precisely in order to reflect on the interpretive effects it shares with the allegedly more authorial “traditional” novels and their conventional use of free indirect discourse. J.M. Coetzee’s Disgrace (1999) begins with an iterative framing of the protagonist’s habitualities as a 52-year old divorcé: (6) For a man of his age, fifty-two, divorced, he has, to his mind, solved the problem of sex rather well. […] Soraya is tall and slim, with long black hair and dark, liquid eyes. Technically he is old enough to be her father; but then, technically, one can be a father at twelve. He has been on her books for over a year; he finds her entirely satisfactory. (1)

In all its innovativeness, Coetzee’s novel is indebted to the long tradition of those ambiguous figural voices that appear as if “narrated” inside someone’s head and yet being caught in the middle of their experiential confusion.8 The contradictory effect is amplified by the use of present tense and unpredictable changes between singular and iterative narration. Thus the literary frame of epistolary tension between experiencing and transcribing (or fabulating) lives on even more forcefully than in conventional FID narration. The Coetzeean present-tense-third-person narrative situation allows for exquisite changes from temporarily unanchored interpretation and value judgements (“He is all for double lives, triple lives, lives lived in compartments” (6)) to reports on singulative, temporarily marked sequences (“Then one Saturday morning everything changes” (6)). 8

As in the double temporality reflected in a sentence describing a filmed dance performance, Lurie is showing to his young student mistress: “the instant of the present and the past of that instant, evanescent, caught in the same space”. (15)

208

Maria Mäkelä

Yet many thought acts of the protagonist David Lurie are left lingering between iterative and singulative. Especially his (reoccurring?) moments of insight (“His needs turn out to be quite light, after all, light and fleeting, like those of a butterfly” (5)) are a constant source of ambiguity: is he really a master of his existentially lingering way of life, or is his confusion only shallowly covered by his eloquence? Temporal changes and third person reference suggest extradiegetic manoeuvring behind the scenes of Disgrace: in fact, at a first blush, the narrative situation resembles conventional psycho-narration. In the same conventional vein, the present tense reinforces the protagonist’s role as the receiver of unmediated sensations and not as a retrospective organizer of experience. By definition, then, we should have the extradiegetic narrator to verbalize, organize and summarize the thoughts of the focal character in a manner typical of analytic psycho-narration—even if occasionally resorting to free indirect quotation of thoughts already verbalized in the character’s mind. Narratorial functions are not the way to approach the ambiguities in Coetzee’s narration, however. Any classical approach to the use of free indirect discourse or psycho-narration in this novel would find ample evidence of both sympathy and irony towards the inner life of Lurie, but it would be redundant to attribute these stances to narratorial intentions, since Lurie himself, a professor of literature and communication, seems perfectly capable of being the empathizer as well as the shrewdest ironizer of his own discourse—and more troublingly, the apparently “narratorial” discourse. A recurring stylistic feature in the narration of Lurie’s experiential flux is a search after le mot juste: “a moderate bliss, a moderated bliss” (6), “this daughter, this woman” (62), “to pass him tools—to be his handlanger, in fact” (136), “Her hips and breasts are now (he searches for the best word) ample” (59). This apparently hesitant verbalization of experience gives rise to contradictory effects, that of almost autistic repetition and that of intentional poeticization: “A ready learner, compliant, pliant” (5). The mind of Lurie appears both as an unedited flux and as a conscious, even artistically motivated editor of language. If we take another look at the beginning of Disgrace, we may notice how the authorial schematization of one’s own experience, already at play in the minds of Emma, Pamela and Madame de Clèves, is given another ironic twist. The focalizer-character’s tendency to assume narratorial activities is clear from the outset: the conventional expositionary mode that introduces the protagonist with the help of familiar categories (“a man of his age, fifty-two, divorced”) is already focalized through Lurie: internal focalization is indexed by the absence of the protagonist’s name (“…he has, to his mind…”). Moreover, as we read on, we conclude that the evocation of prototypes and even the reference to the subjective nature of

Masters of Interiority

209

this “good solution” (“to his mind”) are in fact the first evidence of the protagonist’s ability to self-ironical distancing. Again, as in the example (1) from Madame Bovary or in the anguish of the Princess of Clèves, we may notice how the figural mind exploits the convention of psycho-narration by transforming the markers of indirectness into figural self-reflection on one’s own cognitive processes (cf. “His needs turn out to be quite light, after all”; see Ron 1981: 35). Here, as well as with the earlier examples, one is reminded of Fludernik’s inspiring study on schematic language representation. In short, Fludernik claims that representation of speech and thought relies on mechanisms of schematization and typification (based on cognitive prototype modelling). Whenever narrative reproduces an utterance, mental or oral, it is not actually a matter of reproduction but of construction. As Fludernik demonstrates, this construction relies on approximation, on rendering utterances and expressive features that are likely within certain situational and narrative frames. It is through “typical” markers of expressivity that the illusion of a figural voice is projected into the narration. (Fludernik 1993: 398-408 and passim.) Consequently, in the context of fictional consciousness representation, schematization and typification are considered narratorial functions. Yet in the cases of our artful focalizers—the Princess of Clèves, Emma Bovary, or David Lurie—schematization appears (on the level of our quasi-mimetic interpretation) as intended by the characters themselves; they seem to want to evoke markers of interiority that are somehow representative of something they identify with—more or less willingly (“Technically he is old enough to be her father; but then, technically, one can be a father at twelve”). These examples seem to suggest that stylistically, the mechanisms of typification and schematization are malleable and often attributable to the alleged intentions of the focalizing characters. Schematization and typification of one’s own “narrative” and one’s own experiences demonstrate how narrator as function can be turned into narrator as style. Richard Aczel’s study on narrative voice (to my mind, one of the most progressive ones in the field) moves towards a more textually oriented approach by emphasizing the recognition of narratorial idiom (Aczel 1998: 467-468). Aczel maintains that in FID, the voice of the narrator is identifiable only through its absence, as “voice-different-from” (ibid. 478). Thus the system of voices in a narrative text is a system of differences: the narratorial idiom is identified through its deviation from the figural voice (ibid. 478, 494)—and vice versa. However, Aczel continues in the familiar narratological vein by assigning intentional discursive agency almost exclusively to the narrator and not to the figural voice.

210

Maria Mäkelä

If we follow Aczel’s argument on the centrality of stylistic differences all the way through, we may conclude that a figural voice that attempts to deconstruct these differences by incorporating narratorial conventions into its own discourse also upends the power relations of this system: the resident of the diegetic world appears, if only momentarily, as the organizer of the discursive and narrative elements of her own story. Thus in the case of Emma Bovary’s fancies (example 1), the context works reversely to Aczel’s theory of voice-as-difference: Emma’s subjective attempt to frame and narrativize her own motives is nested—chameleon-like—within the objective report on her moods and behaviour. One of the hackneyed definitions of FID has been its naturalization as narratorial mimicry. One step towards deconstructing such reductionist interpretations is offered by focalizers who create the interpretive illusion of mimicking the narratorial style: they use seemingly objective qualifiers, distance themselves from their experience, dissolve their subjectivity into generalizations; replace self-reference with authority. In such narrative situations, the absence of expressivity creates the subjective experientiality by negation. As Fludernik remarks about FID, “there is a deliberate attempt to erase stylistic difference, the difference between background and foreground” (Fludernik 1993: 331). The only expansion I would suggest to Fludernik’s argument is that there might be a deliberate attempt—not just on the narrator’s side, as Fludernik suggests, but on both sides. Or, stripped from any narrative illusions, this is the uncanny interpretive effect created by the absence of agency markers (“Emma thought”) and the contravening default understanding of narrative voices as hierarchically structured. Such reading of narrative style and of the objective-expressive binary opposition would also potentially upend the conventional reading of the Uncle Charles Principle (Kenner 1978) as “linguistic parody” of the otherwise anti-colloquial narrator (cf. Fludernik 1993: 332). In fact, Lucy Ferriss’ recent article on UCP is already well ahead in doing this when she coins the term APP, named after the short story “A&P” by Updike. In short, APP is UPP’s counterpart in first person fiction and in internally focalized narrative, referring to the contagion of the (otherwise colloquial) character-narrator’s or focalizer’s language by authorial (literary, well-formed, persuasive) idiom (Ferriss 2008: 185). The most notable difference between my own arguments and Ferriss’ concerns the interpretive effects of ambiguous discursive agency. Ferriss encapsulates her own idea by suggesting that authors lending their authorial idiom to their characters’ discursive plane “are saying, in effect: ‘However idiosyncratic my character may appear, however broken-off his or her world, we can connect and unify it if we only supply the right language’” (ibid. 190). In my interpretation, however, the uncanniness of the figural voice persists and the “ex-

Masters of Interiority

211

change of syntactic structures” (ibid. 185) never really ceases, and this instability of the figural voice as a vehicle of experientiality continues to be thematically productive. In Coetzee’s novel, these ambiguities of voice and authority are deeply entangled with the major theme of confession and remorse. Lurie refuses to express public contrition for his having an affair with a young student, but, along with the story, is made to feel remorse for an entire postapartheid nation as well as for his own sex. This ethically unstable position is reflected in the narration: Lurie is established as the only focalizer of events that seem to pass by as an uncontrollable present-tense flow, and yet it seems that it is his figural voice that controls the narrative discourse in a poignantly literary fashion, as if in artistically motivated retrospect. In fact, this ethically unstable positioning of voice comes very close to resembling the notorious first-person fabulator Humbert Humbert who, as the leading theorist of character (i.e. first-person) narration James Phelan has noted, is able to turn the discourse into a story; Nabokov is using Humbert’s act of telling as itself part of the represented action of the novel, a present-tense story running parallel to the past tense story of Humbert and Dolores […] More specifically, Nabokov uses this present-tense story to add a significant layer to the whole narrative: the ethical struggle of Humbert the Narrator. (Phelan 2005: 121)

Perhaps due to his emphasis on homodiegetic narration, Phelan’s rhetorical poetics ignores the uncanny role of third person focalizers in narrative communication – and thus all the “living to tell about it” that has no discernible speaker function attached to it.9 Yet if we consider the narration of Disgrace in the light of Phelan’s reading of Lolita, we may notice a similar tension between being in the middle of things and the re-evaluation of that experience. It is through this tension that the focalizer David Lurie, the man in the middle of things (increasingly awful things, as we learn) just happening to him, at the same time resembles an informed narrator, yet not extradiegetic and not even an intradiegetic one since there is no outer narratorial frame in the classically hierarchical sense; no signifying difference between the extradiegetic and the figural voice arises from the discourse (cf. “There is still Soraya. He ought to close that chapter. Instead, he pays a detective agency to track her down.” (9)). The persuasiveness of Lurie’s evaluations and self-irony creates a communicational fallacy and makes the boundary between homo- and heterodiegetic narration seem artificial. A literary mind such as Lurie is even likely to use the third person to refer to himself. His evident intertextual counterpart Humbert 9

For a critique of the narratological “standard theory of narrative transmission”, see Skalin 2005, and also Nielsen’s article in this volume.

212

Maria Mäkelä

Humbert, who, at the moment of the death of Lolita’s mother and Humbert’s wife, assumes the third person perspective of others who mistakenly treat him as a mourning widower, provides a plausible model. An interpretive leap from the third to the first person may find support from some postclassical studies on the so-called simultaneous narration (first-person-present-tense; Cohn 1999: 96-108, and Hansen 2008), a mode familiar from, among many others, Coetzee’s Waiting for the Barbarians (1980). Per Krogh Hansen continues Dorrit Cohn’s work on simultaneous narration by pointing out some interpretive challenges that the unnatural combination of immediacy and mediacy sets forth; this is what he notes about one of his compelling exemplary texts: “[…] even though [the] text is so obviously marked as ‘narrated’ […] one cannot consider this as a sign of a narratorial or authorial instance beyond the characternarrator, even though the latter occupies a non-naturalizable narrational situation” (Hansen 2008: 327). At a first glance, Lurie seems to occupy this same untenable position, yet one is tempted to think that precisely the use of third person is crucial here and the meanings and aesthetic effects of Disgrace would not survive translation to first person. In order to create the illusion of a discursive takeover—which, thematically, seems to generate from a mixture of denial, defence, literary sensibility and literary obsession—the narrative situation needs to cast the shadow of a possible authority (the heterodiegetic “presence”) towards which the diegetic discursive agent is reaching through narratively and stylistically “authorizing” strategies. What strikes me as an interesting difference between Hansen’s and my own examples has to do with the stylistic effects: Hansen’s firstperson-present-tense create “a disturbing apathetic sense” and indicate incapability of narrative framing (at least in psychological or ethical terms), whereas my examples from epistolary narration, canonical third-person FID and Disgrace display narrative overcapacity and disturbing eloquence in a narrative position which, by definition, should be that of an unwilling and unconscious mediator (cf. Yacobi 1981: 123-124, 2005). In fact, these anti-narrators betray such crafty design in their flux of experience that not only we experience narrativity, we get the sense of writing. As Henrik Skov Nielsen demonstrates in the present volume, one of the pitfalls of structuralist narratology is its mostly unchallenged reliance on pronominal reference even in fictional contexts. Yet I think that a questioning of the conventional functions of third- and first-person reference should not lead to a conclusion that the pronouns “I” and “(s)he” would in any literary context be interchangeable. Just as narrative takeovers and usurpations presuppose a hierarchy of voices, similarly it is the conventional use of pronouns and their reference functions that provide the ground for the uncannily narrativizing and quasi-communicating figural

Masters of Interiority

213

voices. Only through violation can they create a sense of structural loopholes and question the fixed statuses of the constructor and the constructed. Meir Sternberg has reminded us of the varying degrees of communicativeness of textual agents and their awareness of their function as narrative mediators: “[…] discoursers in narrative polarize (or shuttle) between tellers and informants: those who communicate with another about the world, as against those who lead their secret life and unwittingly mediate in the process another’s higher-level communication” (Sternberg 2005: 233). Although Sternberg rightly points out that the discourser’s status is flexible, he does not look into cases in which the leaders of secret lives reflect communicative features in their discourse. Literary narratives do not only demonstrate the fact that evocation of a voice is always in some measure an evocation of communicativeness: this loophole in structure opens up new thematic possibilities. Such techniques foreground the linguistic and literary mediacy at work in the human mind; and furthermore, display the nature of fictional minds as overly verbalized and overdetermined in their verbal design.10 Consequently, I find it fundamentally counterintuitive when Tamar Yacobi writes that the “purely informative status” of Emma Bovary and other subjects of consciousness representation “makes [them] the diametric opposite of the invariably self-conscious author, a passive participant in the text’s communicative process” (Yacobi 1987: 338). I should think that cognitive narratology would have the potential to go beyond the hierarchization of voices and explore the readerly effects of the ultimate multi-voicedness of literary communication. Yet it seems that cognitive narratologists are too keen on keeping their feet on the oral and naturally social ground of storytelling to address the sort of discursive foregrounding in fictional minds that I have been after in this article.11 When Lisa Zunshine (very persuasively) suggests that “fictional narratives […] rely on, manipulate, and titillate our tendency to keep track of who thought, wanted, and felt what and when” (2006: 5), she does not consider the manipulative effects that are brought along to consciousness representation by multi-level verbalization and narrativization processes. The downplaying of discursive agency in FID is designed against our meta-representational capacity and is not reducible to formulaic representations such 10

Cf. Jahn 1996: 247: “Interior monologues may indeed have a quality of ‘voice,’ but this is only because thought has a quality of voice, and not because thought equals voice or is a kind of voice, let alone a narrative voice.” 11 More on this skepticism towards the study of fictional minds in cognitive narratology can be found in an article where I study the phenomenon of characters constructing each other’s minds—a topic that is closely related to the current one. (Mäkelä 2006)

214

Maria Mäkelä

as “X interprets that Y tells that A assumes B to think…” and so on. As Michael Peled Ginsburg notes well before any narratological Theory of Mind applications, “[FID] deserves our attention because it makes explicit the fundamental characteristic of discourse in the novel, its double focus, its existence both as a representation of an object and as in itself an object of representation” (Ginsburg 1982: 140). Thus a formulaic representation is bound to run into its own impossibility (“X interprets/tells/assumes/represents/is X”?). Another persuasive point is made by Palmer, who criticizes classical narratology for its internal speech bias and calls for a more holistic approach to fictional minds as embodied and as social constructs (Palmer 2004: 9-12). Especially Palmer’s critique of classical narratology’s “speech category approach” (the dissection of narration into direct, indirect and free indirect discourse) is well deserved, as already mentioned earlier. However, again something is lost with the bathwater: the fact that literary minds are verbally biased, they consist of nothing but language. We hear unuttered, even unintended sentences communicating to us. A Short Conclusion: Fictions of Authority This article has been an attempt to rehabilitate the role of the figural voice as a quasi-intentional discursive agency. I have suggested that readerly intuition finds nothing alarming in fictional minds that speak to them, persuade them and try to convince them of the legitimacy of their own interpretations—these features can be found at the very heart of literary conventions. Yet there seems to be no berth for such unnatural textual agencies in narratological reasoning. The shift from classical to postclassical narratology has not significantly altered the hierarchically driven standard assumptions about narrative communication and about the fixedness of the levels of intentionality in narrative discourse. Yet the history of figural masterminding goes way back and is already thematized and even parodied by 18th-century epistolary fictions. The voice of an unmediated experience is shown to run into its own impossibility and is replaced by a subjectivity that reflects its own intermediary position both as a textual construct and as a textual constructor. By the emergence of modern novelistic practices, the status of the figural voice has been further complicated, since, in terms of naturally occurring human communication and cognition, it has become incommunicable. As a result, this experiential ambiguity is reflected in structural uncertainty—to a point where the entire notion of the structure of voices becomes debatable.

Masters of Interiority

215

Ultimately, my argument does not so much spring from an urge to criticize existing narratological categories as it is inspired by the multiple ways fictional minds seem to reflect, shadow, challenge and even carnivalize the upper level construction processes—literary creation, the narrative act, and finally, the act of literary interpretation. Yet this is also a bilateral manoeuvre: since the reader is the final constructor of voices, we need to suppose that there is indeed a readerly interest in reading figural voices against the cognitive or the communicative grain—a willingness to defamiliarize the fictional mind. As Henrik Skov Nielsen notes in the present volume, “the question is whether the reader will always try to naturalize anything—and if so, if it can always be done successfully.” Yet another contributor to the present volume, Rolf Reitan suggests several interpretive effects set forth by the use of narrative “you” as both reference and address; what to me seems to be the most appealing of these interpretations is to regard the narratorial voice in second person narratives as “some version of the creator’s voice talking (no, not talking to) his creation, or more accurately, writing his creation” (Reitan in the present volume, italics in the original). Here Reitan is going where I would also like to go: a step further from naturalizing literary communication and a step towards acknowledging the uncanny shadows that the literary terms of existence cast across imagined worlds and minds. As Nielsen reminds the narratological community in the wake of an interdisciplinary explosion, the peculiarity of fictional representation is the result of the paradoxical fact that “[a] work of fiction creates the world to which it refers by referring to it” (Nielsen 2004: 145).12 Figural voices suggest that, in a way, this goes for literary experientiality as well: a verbalized fictional mind bears in itself the process of its own twofold genesis, that is, writing and reading as literary constructions. References Aczel, Richard 1998 “Hearing Voices in Narrative Texts”, New Literary History 29.3, 467-500. Altman, Janet Gurkin 1982 Epistolarity: Approaches to a Form, Columbus: Ohio State University Press. Bakhtin, Mikhail 1981 M. M., The Dialogic Imagination: Four Essays, ed. by Michael Holquist, transl. Caryl Emerson & Michael Holquist, Austin: University of Texas Press. Bally, Charles 1912 “Le style indirect libre en français modern”, Germanisch-Romanische Monatsschrift 4, 549556 & 597-606. 12

A point originally made by Käte Hamburger; see also Cohn 1999: 12-13.

216 1914

Maria Mäkelä

“Figures de pensée et formes linguistiques”, Germanisch-Romanische Monatsschrift 6, 405422 & 456-470. Banfield, Ann 1982 Unspeakable Sentences: Narration and Representation in the Language of Fiction, Boston: Routledge & Kegan Paul. Bender, John 1987 Imagining the Penitentiary: Fiction and the Architecture of Mind in Eighteenth-Century England, Chicago: University of Chicago Press. Booth, Wayne C. 1991 The Rhetoric of Fiction, 2nd ed., London: Penguin. Bray, Joe 2003 The Epistolary Novel: Representations of Consciousness, London and New York: Routledge. Coetzee, J. M. 2000 Disgrace, London: Vintage. Cohn, Dorrit 1978 Transparent Minds: Narrative Modes for Presenting Consciousness in Fiction, Princeton: Princeton University Press. 1999 The Distinction of Fiction, Baltimore and London: Johns Hopkins University Press. Ferriss, Lucy 2008 “Uncle Charles Repairs to the A&P: Changes in Voice in the Recent American Short Story”, Narrative 16.2, 178-192. Flaubert, Gustave 1999 Madame Bovary, Paris: Librairie Générale Française. 2003 Madame Bovary, transl. Geoffrey Wall, London: Penguin. Fludernik, Monika 1993 The Fictions of Language and the Languages of Fiction: The Linguistic Representation of Speech and Consciousness, London and New York: Routledge. 1996 Towards a “Natural“ Narratology, London and New York. Ginsburg, Michael Peled 1982 “Free Indirect Discourse: A Reconsideration”, Language and Style 15.2, 133-149. Gunn, Daniel P. 2004 “Free Indirect Discourse and Narrative Authority in Emma”, Narrative 12, 35-54. Hansen, Per Krogh 2008 “First Person, Present Tense: Authorial Presence and Unreliable Narration in Simultaneous Narration”, in: Elke D’Hoker & Günther Martens (eds.), Narrative Unreliability in the Twentieth-Century First-Person Novel, Berlin and New York: de Gruyter, 317-338. Herman, David 2007 “Cognition, Emotion, and Consciousness”, in: David Herman (ed.), The Cambridge Companion to Narrative, Cambridge: Cambridge University Press, 245-259. Jahn, Manfred 1996 “Windows of Focalization. Deconstructing and Reconstructing a Narratological Concept”, Style 30.2, 241-267. Jongeneel, Els 2006 “Silencing the Voice in Narratology? A Synopsis”, in: Andreas Blödorn, Daniela Langer & Michael Scheffel (eds.), Stimme(n) in der Text: Narratologische Positionsbestimmungen, Berlin and New York: de Gruyter, 9-30. Kenner, Hugh 1978 Joyce’s Voices, Berkeley: University of California Press 1978. Laclos, Choderlos de 1961 Les Liaisons Dangereuses, transl. P. W. K. Stone, London: Penguin.

Masters of Interiority

217

Lock, Charles 2001 “Double Voicing, Sharing Words: Bakhtin’s Dialogism and the History of the Theory of Free Indirect Discourse”, in: Jørgen Bruhn & Jan Lundquist (eds.), The Novelness of Bakhtin: Perspectives and Possibilities, Copenhagen: Museum Tusculanum Press, University of Copenhagen. Mäkelä, Maria 2006 “Possible Minds: Constructing—and Reading—Another Consciousness as Fiction”, in: Pekka Tammi and Hannu Tommola (eds.), FREElanguage INDIRECTtranslation DISCOURSEnarration: Linguistic, Translatological, and Literary-Theoretical Encounters, Tampere: Tampere University Press, 231-260. McHale, Brian 1978a “Free Indirect Discourse: A Survey of Recent Accounts”, Poetics and Theory of Literature 3.2, 249-287. 1978b [Review of Pascal 1977] Poetics and Theory of Literature 3.2, 398–400. 1983 “Unspeakable Sentences, Unnatural Acts: Linguistics and Poetics Revisited”, Poetics Today 4.1, 17-45. McKeon, Michael 1997 “Prose Fiction: Great Britain”, in: Nisbet, Hugh Barr & Claude Rawson, The Cambridge History of Literary Criticism: Vol IV, The Eighteenth Century. Cambridge: Cambridge University Press, 238–263. Mezei, Kathy 1996 “Who Is Speaking Here? Free Indirect Discourse, Gender, and Authority in Emma, Howards End, and Mrs. Dalloway”, in: Kathy Mezei (ed.), Ambiguous Discourse: Feminist Narratology & British Woman Writers, Chapel Hill and London: University of North Carolina Press 1996, 66-92. Miller, D. A. 1988 The Novel and the Police, Berkeley: University of California Press. Muratore, Mary Jo 1994 Mimesis and Metatextuality in the French Neo-Classical Text: Reflexive Readings of La Fontaine, Molière, Racine Guilleragues, Madame de La Fayette, Scarron, Cyrano de Bergerac and Perrault, Geneve: Droz 1994. Nielsen, Henrik Skov 2004 “The Impersonal Voice in First-Person Narrative Fiction”, Narrative 12, 133-150. Palmer, Alan 2004 Fictional Minds, Lincoln and London: University of Nebraska Press. Pascal, Roy 1977 The Dual Voice: Free Indirect Speech and Its Functioning in the Nineteenth-Century European Novel, Manchester: Mancester University Press. Phelan, James 2005 Living to Tell about It: A Rhetoric and Ethics of Character Narration, Ithaca and London: Cornell University Press. Richardson, Samuel 2001 Pamela, or, Virtue Rewarded, Oxford: Oxford University Press. Ron, Moshe 1981 “Free Indirect Discourse, Mimetic Language Games and the Subject of Fiction”, Poetics Today 2.2, 17-39. Seltzer, Mark 1984 Henry James and the Art of Power, Ithaca: Cornell University Press 1984. Stanzel, Franz K. 1984 A Theory of Narrative, transl. Charlotte Goedsche, Cambridge: Cambridge University Press.

218

Maria Mäkelä

Sternberg, Meir 2005 “Self-consciousness as a Narrative Feature and Force: Tellers vs. Informants in Generic Design” in: James Phelan and Peter Rabinowitz (eds.), A Companion to Narrative Theory, Oxford: Blackwell, 232-252. Vološinov, V. N. 1986 Marxism and the Philosophy of Language, transl. Ladislav Matejka & I. R. Titunik, Cambridge (Mass.): Harvard University Press. Watt, Ian 1984 The Rise of the Novel. Studies in Defoe, Richardson and Fielding, Berkeley and Los Angeles: University of California Press. Yacobi, Tamar 1981 “Fictional Reliability as a Communicative Problem”, Poetics Today 2.2, 11-126. 1987 “Narrative Structure and Fictional Mediation”, Poetics Today 8.2, 335-372. 2001 “Package Deals in Fictional Narrative: the Case of the Narrator’s (Un)Reliability”, Narrative 9.2, 223-229. Zunshine, Lisa 2006 Why We Read Fiction: Theory of Mind and the Novel, Columbus: Ohio State University Press.



POUL BEHRENDT & PER KROGH HANSEN (Copenhagen and Kolding)

The Fifth Mode of Representation: Ambiguous Voices in Unreliable Third-Person Narration The last decades’ vivid discussions regarding the concept of unreliable narration have, in almost all cases, focused on first person narrators (homodiegesis). In this study, we will investigate the relevance of the concept of unreliability in relation to third-person narration (heterodiegesis). After having demonstrated that the present marginalization of heterodiegetic unreliable narration is partly a result of an unreflective submission to a certain ‘package deal’, we will focus on the relevance of free indirect discourse (FID) as a basic device in dealing with readers’ delusions, and on the under-discussed issue of authorial unreliability in third person narration. We suggest that the problem of ‘who speaks and who sees?’ requires more complex answers than usually given.1 Our aim is to show that the two different parameters by which FID and the related modes of representation of thought usually are analyzed not necessarily support each other. We shall refer to these levels as the ‘level of discourse’ (or the ‘discursive level’) and the ‘level of narration’ (or the ‘narrative level’)—the former being based on the grammatical parameters that usually determine the well-known modes of representation (that is FID, ID and DD) as definite kinds of discourse: person, inquit, tense and syntactical markers; the latter being based on parameters of voice distributions and characteristics, colliding more or less, as shall be proved, with the grammatically defined parameters of discourse. Our study encompasses two seemingly separate aspects of narrative unreliability—that of factual and that of ideological unreliability respectively. Factual unreliability, which up until now—in the few existing contributions to the field—has been regarded as unimaginable in third person

1

Our study can therefore be related to Maria Mäkelä’s article in the present issue which is concerned with the relationship between mediacy and immediacy in the third person narrative context and instances of focalization and stylistically unmarked free indirect discourse.

220

Poul Behrendt & Per Krogh Hansen

narration,2 is examined in the first part of our investigation through a reading of Isak Dinesen’s classical short story: “Sorrow-Acre”. In the second part, we will examine ideological unreliability through an investigation of FID as a representation of thought in Henry James’ muchdiscussed third-person narrated short story “The Liar”. A closer look at Wayne C. Booth’s controversial transposition of this story to a first person narration ‘in disguise’ serves to reveal the deficiencies of an analysis of unreliability carried through without any consideration of the flexible parameters of FID. Unreliability and Third-Person Narration When Wayne Booth in 1961 introduced the basic concepts upon which the study of unreliable narration has been based ever since, it was of utmost importance to him that there in principle was no difference between the unreliability of ‘teller-characters’ and ‘reflector-characters’. Booth subsumed both under the term ‘narrator’ and argued that “any sustained inside view, of whatever depth, temporarily turns the character whose mind is shown into a narrator”.3 His outset was that ‘person’ was one of the most overworked distinctions in literary studies.4 Instead, he focused on the ways in which the particular qualities of the narrator relate to specific effects,5 claiming that: If the reason for discussing point of view is to find how it relates to literary effects, then surely the moral and intellectual qualities of the narrator are more important to our judgment than whether he is referred to as ‘I’ or ‘he’.6

On these grounds, Booth investigated the reliability of ‘reflector-characters’ such as Jane Austen’s Emma, Henry James’ Strether and Lyon, and Franz Kafka’s Gregor Samsa by the same guidelines as his analyses of moral unreliability in first person narration. Following Booth’s argument, it is therefore paradoxically consistent that he in The Rhetoric of Fiction transposes the heterodiegetic narration of Henry James’ “The Liar” to 2

3 4 5 6

Cf. Cohn 2000: 311: “We must conclude, I think, that third person novels, no less than first person novels, allow for the separation of the narrator from the author, and that a discordant reading of such novels can have major interpretive implications. This critical move does not imply, however, that heterodiegetic novels allow for factual unreliability”. See also Martens 2008. Booth 1991: 164. Ibid. 150. Cf. “Afterword to the Second Edition” (1983) in which Booth commented on this judgment: “Plain wrong. It was radically underworked” (412). Ibid. 150. Ibid. 158.

The Fifth Mode of Representation

221

homodiegesis in his schematic representation of the signs of unreliability.7 This, however, compelled him to introduce a double set of narrators, an “unreliable” reflector-narrator and a “reliable [third person] narrator”; surprisingly enough, without any need for the implied author.8 Probably, Booth’s doubling of the narrative roles has contributed to the general agreement in the many studies of unreliable narration, in which the phenomenon—the technique or the effect—is exclusively reserved for narration conducted in the first person. To some theorists, the focus on first person narration has been a deliberate choice, explicated on the level of subtitles; William Riggan’s 1981-monography: Picaros, Madmen, Naïfs and Clowns. The Unreliable First-Person Narrator is one example;9 James Phelan’s Living to Tell About it. A Rhetoric and Ethics of Character Narration is another,10 which at least in principle leaves an alternative open. However, in most studies of unreliable narration, within the framework of modern narratology, it is more or less taken for granted that the concept of the unreliable narrator belongs to what Sternberg and Yacobi11 term a ‘package deal’—including the first-person narrator and excluding any third-person agency. This tendency has been supported by the fact that the developing corpus of canonized first person narrators—stretching from Goldsmith’s Vicar of Wakefield over Poe’s mad “Tell-Tale Heart” monologist and Nabokov’s Humbert Humbert to Ellis’ Patrick Bateman and Ishiguro’s butler Stevens—to a large extent dictates the direction of interest. Most critics seem unconsciously to build their arguments on the package deal, insofar as all the examples they comment on are first person narrations.12 There are also theorists who have motivated the delimitation, referring explicitly to the kind of rhetoricity on which Booth partly based his argument. Franz K. Stanzel claims that the concept of reliability shall be reserved for “characters that make verbal statements and thereby address or intend to address an audience”;13 and Monika Fludernik makes a comparable claim in her discussion of Dorrit Cohn’s outline of the differences and affinities between an untrustworthy and ironic authorial voice and its reflectors. She concludes: “The difference between unreliability and reflec-

7 8 9 10 11 12

Ibid. 348-349. Ibid. 354, 353. Riggan 1981. Phelan 2005. Yacobi 2001. See for example Wall 1994, Zerweck 2001, Olson 2003, Heyd 2006, and the long series of studies by Ansgar Nünning (1997, 1998b, 1998a, 1999, 2005). 13 Stanzel 1984: 152.

222

Poul Behrendt & Per Krogh Hansen

torization of the deadpan variety is therefore [...] a gradual one. Only firstperson narrators can be properly unreliable”.14 This kind of consideration of the issue, however, is actually rarely seen in recent studies and one might at first hand think this due to the much stricter categorizations that have characterized narrative theory after Booth, mainly promoted by Genette’s distinction between ‘voice’ and ‘mood’ (or ‘focalization’).15 True enough: If we rigorously follow Genette, it is evident that in a strict sense, a heterodiegetic narrator cannot be unreliable in the sense Booth and his aftermath has suggested, insofar as it isn’t ‘fleshed out’ as a character. This is also the main reason for Seymour Chatman’s reckoning that there is an issue to consider, but also for his unwillingness to consider it a case of an unreliable narrator insofar as the subject in question is not “a narrator at all, [but rather] a fallible filter”.16 In this claim, he was supported by Manfred Jahn in his study of “Package-Deals” (1998).17 According to Chatman, a filter character cannot “misrepresent” the story, insofar as “she is not attempting to represent it; rather she is living it. So she can hardly be responsible to the narrative in the way a narrator is”.18 What matters to Chatman in this respect, is that even though we in many cases of filtration find that “both the feeling and the language might be shared by character and narrator, that does not seem reason for arguing that the demarcation between the story world and the discourse world is blurred”.19 Thus, he claims that in “unreliable narration, the implied author constructs a narration that the implied reader must call into question. [...] In fallibility [...] the narrator asks the narratee [...] to enjoy an irony at the expense of a filter character”.20 This is not a true alternative; enjoying irony at the expense of a narrated character is no less a presupposition in unreliable first person narration than it is in third person narration with a reflector/focalizer/filter character. As we shall learn from “Sorrow-Acre”, in unreliable third person narration, the reader is no less ‘asked’ to call the story world into question, as he or she is in a first person narration. Here, as elsewhere, it is quite common that heterodiegetic narration operates with a focalization (internal, as well as external) that establishes a limited or perhaps even distorted point of view, not unlike the one we find in homodiegetic narratives. 14 15 16 17 18 19 20

Fludernik 1996: 213. Genette 1980. Chatman 1990: 150. Cf. Jahn 1998. Chatman 1990: 150. Ibid. 147. Ibid. 151. Cf. Jahn 1998: 92.

The Fifth Mode of Representation

223

Besides, through techniques like FID and indirect representation, we witness how ‘the third person’—the ‘he’ or the ‘she’ of the narrated— obtains a voice that merges with the voice of the narrator. In these cases, the effect is exactly the opposite of what Chatman says, namely that the demarcation between story world and discourse world is blurred. Therefore, we are here dealing less with a narrator who tells about the third person, than one through which the third person speaks. We can go as far as claiming that in such cases we have third person narration without any indication of an extradiegetic narrator. The extra- or heterodiegetic narrator is in the most radical cases, especially in modern examples, reduced to the role of a mediator—an indexically pointing voice without any normative stance towards the narrated. This, to a certain degree, is also the case for the examples to be found in this study. In both “Sorrow-Acre” and “The Liar”, there most evidently is an extradiegetic narrator, in James’ story explicitly judging and even excusing the reflector. However, it is just as evident, that this narrator is unstable insofar as information is withheld and distorted points of view are mediated uncommented. In Dinesen’s story, as we shall show, a close reading discloses authorial irony that makes the reader infer a rather unreliable narrating agent which—as we shall argue—is best understood as ambiguous but nonetheless authorial. Part 1 The Two Levels of Ambiguity in FID Instances of ambiguity in free indirect discourse are to be found on two different levels:  

On the level of discourse, regarding the grammatical identification of this mode of representation compared to other modes.ʹͳ On the level of narration (based on a ‘correct’ grammatical identification) regarding range and responsibility of the third person narrator’s voice as compared to the voice and range of narrated characters in FID.

In the following it shall be argued that we might approach unreliability in third person narration by considering these ambiguities. 21

Cf. McHale 1978: 264: “The ortodox position in stylistics since Bally (1912, 1914) has been to assume the primacy of grammatical features in the constitution of FID.”

224

Poul Behrendt & Per Krogh Hansen

Regarding the identification of FID on the level of discourse, the most famous example of ambiguity was put forward forty years ago by Hans Robert Jauss. In 1970, he claimed that the grounds for the trial against Madame Bovary were based partly on the prosecution’s misreading of a brand new stylistic device in Gustave Flaubert’s third person narration. That is, a reading in which Flaubert was blamed for opinions put forward by his protagonist.22 Jauss’s claim was supported eight years later in Dorrit Cohn’s seminal work, Transparent Minds.23 According to Cohn, as soon as the identification has been performed correctly, the responsibility for the content of an utterance in FID rests mainly on the character. Thus, the proof of narrated monologue, according to Cohn, is that it can be transposed to the first person, present tense, without altering anything else in the wording—that is, transposed from the authorial utterance of “narrated monologue” to the characters’ utterance of “quoted monologue”. This follows from her definition of narrated monologue as “a character’s mental discourse in the guise of the narrator’s discourse”,24 based on the claim that the narrator’s “identification” with “the character’s mentality […] is supremely enhanced” by free indirect discourse.25 Thus, only characters—never authors—are to blame for views, expressed through FID. In his 1983-monograph, Madame Bovary on Trial, Dominick LaCapra presented a different view, maintaining that in narrated monologue (which he labelled “free indirect style”) there is an unavoidable, innate ambiguity. According to LaCapra, even when FID is correctly identified, two voices are always in play simultaneously: that of the character, and that of the author-narrator. You cannot always tell which voice plays the leading part, not even when speaking of the subversive opinions of—say—Madame Bovary, because FID “renders decisive judgment about the character or story difficult to attain”.26 LaCapra’s conviction was based, not on Dorrit Cohn’s views, but on a similar work published only one year before Transparent Minds, in 1977: The Dual Voice by Roy Pascal, covering both speech and thought.27 Here Pascal stressed the continuous presence of a narrative voice in FID generally: free indirect speech is never purely and simply the evocation of a character’s thought and perception, but always bears, in its vocabulary, its intonation, its syn-

22 23 24 25 26 27

Jauss 1970. Cohn 1978: 106-07. Ibid. 14. Ibid. 112. LaCapra 1982: 59-60. Pascal 1977. Cf. McHale 1978.

The Fifth Mode of Representation

225

tactical composition and other stylistic features, in its content or its context, or in some combination of these, the mark of the narrator.28

Based on Käte Hamburger’s no narrator paradigm, the dual-voice hypothesis was ruled out by Ann Banfield in Unspeakable Sentences 1982; and later on, repeatedly, by Monika Fludernik in her great work on FID from 1993, The Fictions of Language and the Languages of Fiction.29 Even though Fludernik acknowledges the dual-voice phenomenon as a readers’ experience in many instances of FID, she ends up by supporting Ann Banfield’s view (though not her arguments). On an interpretational level, a choice is inevitable because the voice of the narrator and that of the character in FID are “entirely distinct in the frames that they evoke”.30 Nevertheless, the dual-voice hypothesis arose again in 2004 in an important contribution to Narrative. Here Daniel P. Gunn convincingly argued that the hypothesis is indispensable when dealing with Jane Austen’s Emma.31 In this novel, two voices constantly compete and collaborate with each other through the author’s use of FID, primarily as “an imitation of figural speech and thought, in which the narrator echoes or mimics the idiom of the character for the purposes of the fiction”.32 Gunn’s analyses thus contradict Dorrit Cohn’s claim, maintained as late as in 1995, that in narratives where FID predominates, the narrator is “reduced to a merely functional presence”.33 However, on both levels—the level of discourse defined by grammar, as well as on the narrative level defined by domination of voices—the general effort has been to prevent mistakes on part of the reader. Only rarely is it taken into consideration that readers’ misinterpretation might be part of an author’s scheme. A First-Time Reading of FID: Isak Dinesen’s “Sorrow-Acre” In order to explore narrative unreliability on the level of discourse, we shall use an example from the best known among Isak Dinesen’s Winter’s Tales. From the short story titled “Sorrow-Acre”, we quote a passage that plays a decisive role in shaping the reader’s view of necessities and possibilities in the story world.

28 29 30 31 32 33

Pascal 1977: 43. Banfield 1982 and Fludernik 1993. Fludernik 1993: 453. Gunn 2004. Ibid. 37. Cohn 1995: 14.

226

Poul Behrendt & Per Krogh Hansen

The tale runs from sunrise until sunset on a fine harvest day around 1775. It takes place at a Danish country manor, whose family is under threat of extinction due to the death of the Lord’s only son a year before. The Lord’s nephew, who was raised at the manor in company with his cousin, has been asked by his ambitious mother to offer himself as a hereditary substitute. Tacitly, however, he has declined. He wants to be selfreliant, depending on his own abilities and not on that of ancestry; he dreams of travelling to the new world of America. But there is also another reason: The family estate, in difference from the neighbouring properties, had gone down from Father to Son ever since a man of their name first sat there. The tradition of direct succession was the pride of the clan and a sacred dogma to his Uncle, he would surely call for a son of his own flesh and bone.34

A first-time reader of this passage would tend to understand it as an objective judgment authorized by a third-person narrator. The possible presence of the nephew as a reflector in the last sentence does not change this impression. The passage corresponds to the factual circumstance of the story world; namely, that the Lord of the manor has recently married a young ‘Lady’ from Copenhagen who had formerly been designated the bride for the lord’s late son. Nevertheless, after sunset, on the very day of his arrival to the manor, the nephew will be having a one-night stand with his newly married aunt—thereby begetting an heir to the estate. The point is that no reader of “Sorrow-Acre” has ever been able to grasp this fact when reading the story for the first time. To a first-time reader this simply never happens; first of all, because it is never explicitly told; secondly—as shown in the above quotation—because the reader is told the exact opposite. Nevertheless, as soon as the event is revealed—or communicated—it is accepted by a second-time reading of the story; for instance, it is now a widely accepted fact in the Isak Dinesen-literature of “Sorrow-Acre”.35 Thus, in the light of this untold event, several cues suddenly become conspicuous. For example, that Adam during his stay in England had been told by a gipsy-woman that a son of his was to sit in the seat of his fathers. Or the circumstance, that his young aunt in her morning-wakening scene asks a flea, caught between her fingertips, whether it is the only creature ready to risk “its life for her smoothness and sweet blood”.36 Among these cues, too, the fact that the above quotation could be interpreted just as well, or better, as FID. 34 35

Dinesen 1942, 35. Cf. e.g.: Jørgensen 2000, Engberg 2000, and Selboe 1999 & 2000, while Langbaum (1964) imagined that the nephew would marry his aunt after the lord’s death. 36 Dinesen 1942: 48.

The Fifth Mode of Representation

227

The point is, however, that FID in this case is far less conspicuous and easier to overlook than Cohn’s definitions imply. It simply makes no sense to transpose the passage into first person, present tense—which, according to Cohn, was the proof of narrated monologue; simply, because there is no implied first person in the two first clauses of Adam’s discourse (unless “their” should be transposed to “our”). Besides, regarding the last sentence, it is not a direct thought of Adam’s, but a reported conviction. The Third-Person Narrator and the Fifth Mode of Representation: Ambiguous Discourse This example is just one among several in Winter’s Tales where readers are led astray by the equivocation in FID. This is not a case of a reflector being morally unreliable, but a question of the reader’s misidentification of a mode of representation. To get a better understanding of how this functions, we have to take a closer look at what characterizes third-person narration in past tense. The distinctive features hereof can be summarized in four points, as follows:    

It is narrated by a neuter narrator (or ‘narrating function’ (Erzählfunktion) if one favours Käte Hamburger’s approach) without any gender, any name, any age or I.͵͹ It is linguistically representing the setting of the story, that is, a given situation, its time, place, history, and the behaviour of the characters involved. Giving the reader (if deemed necessary) direct access to whatever the characters may sense, think or feel, using varying modes of utterance to express their mental circumstances. Equipped with varying degrees of liberty (or restrictions) to alternate between describing the behaviour (externally) and exposing the consciousness (internally) of the characters.

Insofar as these characteristics allow for several ‘grey zones’ in-between the narrators’ and the characters’ domain, it follows that it is insufficient only to calculate with the three already mentioned modes of direct (DD), indirect (ID) and free indirect discourse (FID) in third person narration.

37

Hamburger 1993.

228

Poul Behrendt & Per Krogh Hansen

We must include an additional mode to register the intermingling of the characters and the author-narrator, schematized in the following table: Table 1: The four discursive modes of representation and the narrative modes in past tense third-person narration Level of narration: Narrative modes ĸ Narrator dominated/authorial

Character dominated/figural ĺ

ĸ External

Internal ĺ

ĸ Diegesis

Mimesis ĺ

ĸ Signs of dating Level of discourse: Representational modes

1: CID Character(in)dependen t Discourse

Person:

Third person (he/ she/they)

Inquit:

÷ Mindinquit

Tense:

Past tense

Syntax:

Main clause

Signs of deixis ĺ 2: ID Psychonarration

ĸ

ĺ

Sliding transitions (÷ inquit)

Sliding transitions (... , he thought, ...)

Third person (he/she/ they) + Mindinquit (… thought that ...) Past tense Subordinate Clause

3: FID Narrated Monologue

ĺ

Third person (he/she/ they) ÷ Mindinquit

Sliding transitions (÷ inquit ÷ verbs ÷ quotation marks)

Past tense Main clause

4: DD Quoted Monologue

First person (I/we) + Mind inquit: (… thought: ‘…’) Present tense Main clause

CID: Character-(in)dependent discourse ID: Indirect discourse FID: Free indirect discourse DD: Direct discourse

Disregarding for the time being the level of narration in the bars at the top of the table, the four modes are visualized as vertical columns, from left to right. The first one to the left is the authorially dominated, character-(in)dependent narration (CID). The inclusion of ‘in’ in brackets signals the fact that there often may be a merging of horizons between the narrator and a character. The character ‘sees’ the same as the narrator. However, the narrator ‘sees’ the character, too. By representing consciousness exclusively through behaviour and speech CID differs from the three other modes. Thus, it is a mode of representation which is independent of direct access

The Fifth Mode of Representation

229

to the minds of the characters, but still dependent on the possibility of giving some presentation of the character and his or her consciousness. The last column, to the right, is the character-dominated mode of direct discourse or ‘quoted monologue’ in the first person, present tense, which in the 20th-century novel unfolds to a genre of its own. Between these opposites, you find the two intermediaries; that of indirect discourse, where the narrator, on behalf of a character, reports or summarizes what is going on in his or her mind; and that of free indirect discourse, where the voice and individual mind of a character is supposed still to be audible through the third person narration. The column farthest out to the left contains the grammatical criteria that determine the four modes of representation as definite kinds of discourse—Person, Inquit, Tense and Syntax. Indirect discourse and direct discourse differ from the two other modes by their usage of inquits. Direct discourse differs from indirect discourse both in Person and Tense. Free indirect discourse differs from indirect discourse by Inquit and Syntax. Between the four columns, there is a no-man’s land of sliding transitions between the varying modes of representation. Most present day authors, for example, will do without marks of quotation in direct discourse of thought (as well as of speech). They may also discard the inquits in ID, deleting the initial main clause, or, conversely, insert an inquit in FID to make certain whose voice is on record. In the so-called scale models, theorists have tried to fill out the blanks between the four columns with several separate categories, though without any lasting success.38 In this context, however, the important thing is that by criteria of grammar, there is no difference between the first mode of characterindependent discourse (CID) and the third mode of free indirect discourse (FID), neither with respect to Person, Inquit, Tense, nor Syntax.39 This deceptive convergence was the grounds for the prosecution’s misreading in the Bovary-trial 150 years ago.40 It is the grounds, too, for the misunderstanding of the finishing point of “The Liar” as being authorized by 38 39

Cf. McHale 1978 and Fludernik 1993: 280-318. Cf. McHale 1978: 264: “In any case, the basic grammatical characteristics of FID—absence of reporting verb of saying/thinking, back-shift of tenses, conversion of personal and possessive pronouns etc.—do not in themselves guarantee its being unequivocally distinguished from neutral (diegetic) narration in which only the narrator’s voice is present.” 40 Concluding an analysis of ambiguities in a passage of FID, Monika Fludernik in her contribution to The Literary Encyclopedia on “Free indirect discourse” states a similar convergence: “In this case no free indirect discourse exists, we are dealing with narrative report pure and simple”. (Fludernik 2001) Cf. McHale 2009: 435: “To further complicate matters, many instances of FID entirely lack the form’s defining features so that, taken out of context, they appear indistinguishable from non-quoting narrative sentences”.

230

Poul Behrendt & Per Krogh Hansen

the narrator (more about this later). In the tales of Isak Dinesen, this convergence is used as means of narrative unreliability, as visualized in table 2 below. Table 2: The fifth mode of representation in past tense third person narratives Level of Narration: Narrative modes ĸ Narrator dominated/authorial

Character dominated/figural ĺ

ĸ External

Internal ĺ

ĸ Diegesis

Mimesis ĺ

ĸ Signs of dating Level of Discourse: Modes of Representation:

CID Character(In)dependent Discourse

Person:

Third person (he/she/they)

Inquit:

÷ Mind-inquit

Signs of deixis ĺ ID PsychoNarration

ĸ Sliding transitions (÷ inquit)

Third person (he/she/they)

FID Narrated monologue ĺ Sliding transitions (... , he thought, ...)

+Mind-inquit (thought that …)

Third person (he/she/they) ÷ Mind-inquit

Tense:

Past tense

Past tense

Past tense

Syntax:

Main clause

Subordinate Clause

Main clause

ĺ Sliding transitions (÷ inquit ÷verbs) ÷quotation marks)

DD Quoted Monologue

First person (I/we) + Mind inquit: (... thought: ‘…’) Present tense Main clause

The fifth mode of representation: AD Identity in ĺ Person, Inquit, Tempus & Syntax The mode of ambiguity and of narrative unreliability ĸ

The convergence of mode 1 and mode 3 is the fifth mode, which we in what follows will label Ambiguous Discourse (AD). This is the mode of ambiguity and unreliability on the level of discourse in the stories of Isak Dinesen (to be discerned from the dual-voice phenomenon on the narrative level of FID). Even though they may seem fraudulently alike, the relation between CID and FID on the discursive level is contradictory (either it is FID, or it is CID). But on the level of narration in FID the relation between the voice of the narrator and the voice of the character is varying according to specific narrative circumstances (see below for ‘authorial’ and ‘figural’ invasions, respectively). By causing the reader to confuse the mode of free indirect discourse (FID) with the character-(in)dependent discourse (CID) and thereby authorizing the convictions of a character to be judgments of the impersonal narrator, the fifth mode (AD) leads the reader to overlook story world facts of decisive importance for understanding the story.

The Fifth Mode of Representation

231

However, AD might just as well make the reader convinced of events to come, which in fact will not occur. Thus, it lasted fourty years before anybody suspected that the young Norwegian actress of Isak Dinesen’s anecdote of destiny, “Tempests”, takes her own life at the end of the sto41 ry. Any first-time reader would expect her to resume playing the part of Shakespeare’s Ariel, for which she was originally designated by the Danish director of a touring theatre company. She has left this company when becoming engaged to the son of a ship owner, after courageously having rescued his boat and most of his crew (as well as the theatrical troupe) from shipwreck and drowning. Later on, she realizes that it was not herself, but Shakespeare’s Ariel who performed so bravely on the stage of a crude reality at the expense of a young man’s life. In the end, she realises also that for the same reason she is unable to marry her fiancé, she is also unable to ever again play the part of Ariel on a stage of fiction—which she actually plays, consciously, as a real-life experience, when returning in despair into the arms of the delighted Prospero-director.  One of the reasons why readers never discover this story-world fact but adhere to the conviction that the heroine will continue to play the role of Ariel for the rest of the season, is that the fifth mode of representation is put into use, just a few pages before the end. While the two are embracing each other, we are told about the director and his pupil, in the equivocal manner of AD: He was not to abandon his precious possession, but she was still his and would remain with him, and he was to see his life’s great project realized.42

Just as in “Sorrow Acre”, the reader—by means of AD—is invited to rely on this not being the personal conviction of a theatre director, but a storyworld fact expressed and authorized by the third person narrator. But the thing is that, most likely, it is the other way round. Authorial Invasion in the Modes of Representation43 As part of this context the phenomenon of dual voices also applies— something which becomes obvious if the above utterance is transposed, according to the advice of Dorrit Cohn, from FID to direct discourse:

41 42 43

Selboe 1996. Dinesen 1958: 130. To some degree comparable to Roy Pascal’s concept of narratorial “usurpation” and “illegitimate intrusion” with “dire results of misreading”, cf. Pascal 1977: 105-12, and McHale 1978: 280.

232

Poul Behrendt & Per Krogh Hansen

I am not to abandon my precious possession, but she is still mine and will remain with me, and I am to see my life’s great project realized.

Nobody would speak to himself in this way. As it appears, the irony of an authorial voice in the DD-version becomes audible in the underlined words. However, to interpret the ambiguity of dual voices on the narrative level of FID, it is important to separate modes of discourse from the narrative modes (as it is, likewise, when interpreting the ambiguity of FID on the level of discourse). While the modes of discourse, as already mentioned, are grammatically defined, the modes of narration are defined by which of the two opposites dominate in the discourse of representation: the authorial voice on the one side, to the left, versus the voice of the character on the other, to the right. When looking at the modes of narration on the horizontal bars on top of the tables, we get the impression of a continuous scale from a maximum to a minimum of authorial domination (from the left, declining to the right) and of character domination (from the right, declining towards the left), just as in the case of External vs. Internal, Diegesis vs. Mimesis, and Dating vs. Deixis (a list of binaries that may be prolonged). This is correct in regard to modes of representation. However, it can be strongly modified by the relation of the representational discourse to other textual aspects. In the case of Isak Dinesen, direct discourse does not mirror any cognitive processes of the mind, nor does it mirror any individual idiom. The characters of Winter’s Tales do not think in a stream of consciousness. On the contrary, they think in brief aphorisms, with a marked authorial note (likewise, mimetic expressions or signs of deixis are not found in the stories of Dinesen nearly as often as for instance in a novel of Jane Austen; nor was there any in the quotation from “Tempests”). This contradicts the grammatically defined position of direct discourse as exclusively character-dominated. Such invasions of an authorial voice might occur in the same manner in other modes of representation; first of all, of course, in utterances of free indirect discourse, thereby supporting the fifth mode of representation (AD), that is the ambiguous fusion of the third and the first mode. In the key passage from “Sorrow-Acre”, the choice of words and the degree of Olympian survey of the hereditary conditions of the clan might just as well be those of the narrator as of the character. Even though the identification of this passage as FID turns out to be the correct identification on the level of discourse, the dual-voice analysis on the level of narration reveals that the reader’s initial misinterpretation of the passage as CID is supported by the authorial obfuscation of the character’s voice:

The Fifth Mode of Representation

233

The family estate, in difference from the neighbouring properties, had gone down from Father to Son ever since a man of their name first sat there. The tradition of direct succession was the pride of the clan and a sacred dogma to his Uncle, he would surely call for a son of his own flesh and bone.44

There are no signs of deixis, no mimetic anaphors, nothing that would traditionally mark the passage as a character-dominated FID; except for one word at the end of the passage: the word “surely”; however, only if we take “surely” to mean “probably”. If it means “certainly”, it might just as well be an authorial narrator’s choice of word. Shifting of Modes Indeed, it actually is a choice of the authorial narrator. If we concluded that the passage of free indirect discourse in the above quotation is undermined by the act of adultery at the end of the story, we would be just as wrong, as we would be right. If the reader, for example, concluded that the line of direct succession from father to son thereby is broken, he or she would be mistaken: The point is that according to the law of those days, Adam’s offspring would not be his own, but his Uncle’s. A legal fact, to which there is a hint at the beginning of the story, in which the third person narrator gives a character-(in)dependent survey (CID) of the part and importance of the female gender at a European manor down through generations: The ladies who promenaded in the lime avenues, or drove through them in heavy coaches with four horses, carried the future of the name in their laps […] Their lords might rule the country, and allow themselves many liberties, but when it came to that supreme matter of legitimacy which was the vital principle of their world [that is: the men’s world], the centre of gravity lay with them [that is: with the women].45

It all lies concealed in the concept of legitimacy, which may be legally preserved, also when biologically broken. However, even though the nephew, in his act of adultery, believes himself to be an instrument of retribution, it is quite clear, in a second reading, that the Uncle is fully aware of what is going on between his wife and the young man—and consents. What the lord calls for is not only the legitimacy in the line of direct “flesh and bone”-succession; it turns out—unfortunately only in the author’s Danish auto-translation of “Sorrow-Acre”—that he considers his nephew, literally, a part of himself, “by the blood, through the soil, and through the 44 45

Dinesen 1942: 35. Our italics = FID, underlining = CID. Dinesen 1942: 32.

234

Poul Behrendt & Per Krogh Hansen

Name itself that we both carry”, as he claims when his nephew in the late afternoon threatens to leave the manor for good.46 As a result, there are not only two but three levels involved in the experience and interpretation of narrative unreliability for “Sorrow-Acre”. There is the first level, on which the reader is lead astray by the ambiguity on the level of discourse in FID and his or her ‘willing suspension of disbelief’. There is the second level, as the reader discovers the deceit of the author-narrator, recognizing the mimicry between the third and the first mode on the level of discourse in FID. And finally there is the third level, that of the author, when the narrative deceit is discovered to be an ironic means to heightened insight into the collisions between the Old and the New World—including the reader as a part of the dilemma. In this way, the dual-voice passages, after the shift from an authorial (CID) to a character-dominated (FID) interpretation, shifts once more, including the lines that precedes the central passage quoted from “Sorrow Acre” concerning the surprising marriage of the lord with the designated bride of his late son: Adam’s Mother in her disappointment lay the blame on him. If he had returned to Denmark, she told him, his Uncle might have come to look upon him as a son, and would not have married—nay, he might have handed the bride over to him. Adam knew better. The family estate, in difference from the neighbouring properties, had gone down from Father to Son ever since a man of their name first sat there. The tradition of direct succession was the pride of the clan and a sacred dogma to his Uncle, he would surely call for a son of his own flesh and bone.47

As can easily be seen, in the first half of the passage there is FID in speech (or rather in a letter) of what Adam’s mother has written to her son, equipped with an ‘irregular’ inquit (she told him), as a sliding transition between ID and FID. This ensures that the reader does not take these words as Adam’s opinion or that of the narrator. However, the whole passage might also be interpreted as FID: Adam’s recollection of what has happened before his homecoming, doubling it all, with the words of his mother as a character’s dual-voiced FID (“nay”) embedded in an authorial narrator-dominated FID. Nevertheless, all eventual passages of FID are embedded in a laconic discourse. Thus, the beliefs of Adam and his mother regarding the old Lord, in the final shift, turns out to be quite right— both regarding the sacred dogma of direct succession, and regarding the 46

In Danish: “ved Blodet, gennem Jorden og gennem det Navn selv, som vi begge bærer”, (Dinesen/Blixen 2010: 247-48). Thus, in the author’s auto-translation several passages were added that never have been inserted by the American publisher (though Robert K. Haas promised to) into the following editions of Winter’s Tales. There is a difference of in total 41 pages between the (original) American and the (later auto-translated) Danish edition of Winter’s Tales. Cf. Behrendt 2010. 47 Dinesen 1942: 35. Our italics = FID, underlining = CID.

The Fifth Mode of Representation

235

uncle’s willingly handing the bride over to his nephew—even though not in the sense that they (or the reader) had foreseen. This means that the irony connected to narrative unreliability, in the case of “Sorrow-Acre”, is not fully discovered until the final shift on the third level, which is that of the author. The irony is never immediately audible and cannot be detected as a part of the narrator’s voice as long as it is perceived as either CID or FID. It exclusively depends on the reader’s tacit recognition. There is one main reason why readers in general do not immediately discover what is going on in “Sorrow-Acre”. On the day of his nephew’s homecoming, the Uncle—by coincidence—has arranged for the passing of a sentence in a quite extraordinary trial. The lord, as the locally supreme power, has offered to pardon the only son of a poor widow, who has been accused of arson. Yet, it is a condition, that the mother during a single day, from sunrise until sunset, can harvest a rye field (a task that would normally take a man three days). When the Lord refuses to call off this gruesome arrangement, leaving the verdict to chance and likely to cost the old woman her life, his nephew decides to leave the manor, filled with a rage, which equally—and intentionally—fills up most readers. The point is that the young man’s rage stems from the morally incompatibility of having a love affair with his young aunt, when simultaneously a poor old woman may be sacrificing her life to save her only son. Nevertheless, the nephew changes his mind once more and decides to stay at the manor as the argument with his uncle makes him realize that, by an act of adultery, he at one time may cuckold the stubborn old lord and play a role in the young lady’s life comparable to that of the old woman to the life of her son.48 The aunt, however, is never informed—neither by her husband, nor by her affectionate next-of-kin—of what is going on in the rye field or of her own part as a battleground for the power struggle between the two males. The Puzzle Picture Thus, as long as the true nature of the equivocal passages is not recognized, the fifth mode of representation (AD) has the same status as a puzzle picture in which the hidden image is still to be revealed, similar to the untold event at the end of “Sorrow-Acre”. Both demand, when recognized, a hermeneutic reorientation towards the story world as a whole. 48

Cf. Behrendt 2007: 136-139, 157-166.

236

Poul Behrendt & Per Krogh Hansen

This follows from the strong element of ‘underreporting’ and ‘underevaluating’ on the side of the author-narrator concerning the hours, which the nephew during the day spends together with his uncle’s wife. The focus of the story is almost exclusively on the events in the rye field from dawn until dusk, misguiding the reader, like the Fifth Mode, on the level of discourse as well as of narrative voices. Part 2 First-Person Narration in Disguise It should be evident by now that Chatman’s clean cut between fallible reflectors and unreliable narrators is less obvious than he suggests. Chatman’s concept relies on his communicative model, which not only identifies different narrative instances, but also suggests a hierarchy between them, where the reflector-character is subordinated the narrator, who again is subordinated the implied author. This model—with its static frames and its anthropomorphic understanding of narrator and Implied Author as acting beings—very easily turns the attention away from the dynamic interaction between these levels of narration. It might even be claimed that in some cases the hierarchy is actually reversed: The extradiegetic narrator being subordinated the reflector character. This, for instance, is the case in one of Isak Dinesen’s other Winter’s Tales, “The Sailor-Boy’s Tale”, a third person narrative in which it is not revealed until the finishing lines that the story is a first person narration in disguise that provides its title with a double meaning: “In this way the sailor-boy got back to his ship “Hebe”, which was to sail the next day at noon, and lived long enough to tell this story”.49 Even though there is a frequent use of ID in the tale, there is certainly no psycho-narration (allowing, according to Cohn, a third person narrator speaking in his/her own discourse on behalf of the character). Neither are there any examples of AD, nor of FID, in which a dual-voice phenomenon otherwise would be audible. Actually, the only instance of free indirect discourse in the story is a sentence that represents the culmination of the 16-years-old boy’s experience of life while standing in the midst of a group of dancing men, calmly expecting to be caught and receive the pun49

This is our translation of the author’s (later) Danish version of the story’s finishing lines which in the English edition sounds, less emphasized: “In this way, the sailor-boy got back to his ship, which was to sail the next morning, and lived to tell the story.” (in Danish: “Saaledes kom Skibsdrengen tilbage til sit Skib “Hebe”, der skulde lette Anker næste Dag ved Middagstid, og levede saa længe, at han kunde fortælle denne Historie”. (Dinesen/Blixen 2010: 22.

The Fifth Mode of Representation

237

ishment of death for having killed another sailor who prevented the boy from meeting a girl he only had known for 24 hours: He did not entreat his destiny, nor complain. Here he was, he had killed a man, and had kissed a girl, he did not demand any more from life, nor did life now demand more from him. He was Simon, a man like the men round him, and going to die, as all men are going to die.50

Thus, there is no irony, neither in this passage of FID, nor in the story as a whole, unlike the ten other winter’s tales. The distance between the protagonist and the narrator (and their merging in the last sentence) is similar to the distance (and merging) between the retrospective and the acting “I” in a first person narration. However, even though “The Sailor Boy’s Tale” comprises a supernatural female being, assuming sometimes the appearance of a peregrine falcon on its flight to South East Africa, sometimes of an old Lapp woman with a family in Norwegian Lofoten, it is a completely reliable narration.51 Nevertheless, it is a living proof of the unnatural fact that heterodiegetic narration might just as well be homodiegetic. The problem with Chatman’s conceptualization manifests itself in his remarks on Booth’s reading of James’s “The Liar”. Chatman, who in general appreciates Booth’s reading, blames him for confusing the issue of fallible reflector vis-à-vis unreliable narrator and says that: the protagonist, Oliver Lyon, cannot be an unreliable narrator for the simple reason that he is not the narrator at all. He is, rather, a fallible filter. Of course, the story transpires very much in Lyon’s head [... and he] rationalizes some rather deplorable behavior on his own part. Still, he does so in the privacy of his own mind, not as a representation to a narratee. Indeed, as a character he has no consciousness of the existence of a narratee [...].52

The question is, though, whether the consciousness of having a narratee can be said to be a constitutive feature of all narrators? This does not seem to be the case. First person-present tense narration (as an example where narrative unreliability nearly is the norm) is in most cases not performed as a directed communication to a narratee. It is a situated telling in an impossible situation (the moment of the action) taking the form of a mixture of registration, reflection and imagining, without any address.53 At best, we can say that the narrator narrates, not to a narratee but is overheard by a (to the narrator) non-existent listener. The same can be claimed of Oliver Lyon’s explanations of his own motives when put forward in AD. This is not to imply that Chatman is wrong when claiming that our knowledge of Lyon’s views as flawed is 50 51 52 53

Dinesen 1942: 99. Our italics = FID, underlining = CID. Cf. Behrendt 2011. Chatman 1990: 150. Cf. Jahn 1998: 93. Hansen 2008.

238

Poul Behrendt & Per Krogh Hansen

established, “not by us listening to him tell the story but by listening to a covert narrator represent his thoughts and intentions”.54 It still leaves the question of how to understand this covert narrator, ‘his’ role and function, and ‘his’ relation to the reflector-character. As we shall see, it is also questionable whether we actually should consider him ‘covert’ at all. To approach these issues, we might very well return to Booth’s reading to see how he is dealing with them. As will be known, in The Rhetoric of Fiction Booth generally takes great efforts to ascribe full control of the distance between the unreliable character-narrator and the reliable impersonal narrator to the implied author. However, the implied author seems to be of very little significance in the chapter in which Booth confronts the present issue. This may be due to the fact that he finds “The Liar”, together with “The Aspern Papers” and some other minor short stories, problematic insofar as it, according to Booth, escapes James’s control and therefore develops an “unintentional ambiguity”.55 By way of James’s notebooks, Booth looks into stories where observers—especially unreliable observers becoming acting parts of the story they observe—“are imported after the original conception of the idea has been formulated”.56 According to Booth, this gives the stories a problematic “double focus” that seems to spring from what he considers an “incomplete fusion” between two ideas—an original and one that is added at a later stage. This is the pattern Booth finds in “The Liar” (1888), supported by the differences between the first and the second edition of the story (The New York Edition, 1907-1909). According to the notebooks, James’s original interest was the wife of a pathological liar, Colonel Capadose. The idea was to tell a story of a woman who is corrupted into dishonesty as a consequence of her loyalty to her lying husband. During the development of the story things become more complex due to James’s adding an observing third person—the painter Lyon—who twelve years earlier was in love with Mrs. Capadose and therefore has a personal interest in making her confront the lies of her husband. Booth shows in his reading that the personal interest of Lyon turns him into yet another liar in the story: The motives Lyon expresses for his task are not the true ones. His real interest is directed by jealousy, and not—as several critics before Booth claimed— “inspired by the Muse of Truth”.57 We might have expected Booth to carry through this argument by referring to how the implied author arranges things and instructs the reader 54 55 56 57

Ibid. 151. Booth 1991: 316. Ibid. 340. Ibid. 350.

The Fifth Mode of Representation

239

how to get to the right understanding. However, insofar as the implied author in Booth’s conception is the normative instance that controls the development of the narrative, it is not surprising that ‘he’, in Booth’s opinion, does not play any significant role here. The story is highly ambiguous and does, in that sense, not express any clear norm. Instead, Booth (ibid. 348-49) takes control himself—that is: he ‘translates’ the story into a first person narrative and in the same stroke determines the unreliability of Lyon by comparing Lyon’s alleged view of the events with his own normative interpretation of them: After twelve years I meet again with the woman who once refused to marry me because she did not know then that one day I would be famous. She is even lovelier than ever, and I am horrified to discover that she has married an inveterate liar.

How can she endure living with such a “monstrous foible”? And how can she avoid being herself morally destroyed by contact with such a contemptible man? I admit that Capadose is not—as yet—a “malignant liar,” that he is strictly disinterested, that he has indeed a kind of code of honour in his being. I also must admit that I, too, “lie”, in a sense, when I lay on my colour as a painter. At the same time it seems to me a tragedy that a lovely creature like her should be to man of no integrity. I decide to force her to admit that she is distressed by her husband’s lies. Using methods of a subtlety that almost make me blush, I persuade them to allow me to paint Capadose’s portrait, determined to paint it in such a way as to reveal he depths of his deceptive heart.

Actually she refused the narrator because she knew that happiness would be impossible with any man as selfcentered as he. He is not so much horrified by the lying as jealous, he finds, in short, that he is still in love in love with her, and the discovery that her husband is a liar only contributes to his unconscious jealousy. He is really sure that she must regret having married a contemptible man, when she might have had someone like himself. His own lying is strictly “interested,” and he has far less integrity than Capadose.

He really decides to make her show signs of regret about having chosen wrongly. Lying to him about his motives, he persuades them to allow him to paint the husband’s portrait, determined to paint it in such a way as to expunge all the good traits of the Colonel, whom everyone else in the story likes, and allow only the dishonesty to show, thus creating a monster out of what is actually, as James himself describes him in a letter, “a charming man, in spite of his little weakness.”

240

Poul Behrendt & Per Krogh Hansen

When she sees what I have revealed, surely she will give sign that her basic integrity has not been shattered. I paint the portrait as planned, it is a masterpiece of truth. The Liar stands revealed upon my canvas in his true colours. But the Capadoses discover the portrait when they think I am away; actually I happen upon the scene and am forced to eavesdrop to protect my interests. Mrs. Capadose is shattered by the vision, seeing it truly; Capadose, somewhat more slowly seeing what I have revealed, slashes the portrait to bits. I do not try to stop him; rather I am glad that now, at last, I shall get her to admit regrets about her unfortunate marriage. But instead she support her husband in the falsehood he invents about how the portrait must have been destroyed, and she reveals unmistakably to me that she has been totally corrupted by her husband. The Liar has triumphed, and I have lost my vision of the incorruptible woman. “Her hypocrisy” is revolting.

Surely she will show some sign that she regrets her marriage, that she could imagine her happiness with her husband “more unqualified.” It is a masterpiece of the power of caricature. The Liar, stripped of all his redeeming human qualities, stands portrayed upon the canvas. Having sneaked back without any announcement, he deliberately eavesdrops. Lyon is really delighted to see her horror when she discovers the cruel “truth” of the portrait, and even more delighted to see Capadose slash the portrait to bits. Now at last she is ashamed of her husband, and Lyon has made her so; but he has made her even more horrified by his brutality. In responding to her husband she reveals unmistakably that she still loves the better man and is willing to lie for him. The vicious Liar—Lyon—has been caught in his own trap.

First of all, the problem with this way of interpreting the story as a hypocrite first person narration versus a prosecuting third person narration, is that all ‘facts’—presented as “unconscious” grounds for moral condemnation in the right column—are actually communicated quite bluntly in the story itself. Several times it is claimed by other people that Lyon is still in love with Mrs. Capadose; we are told that he keeps thinking of her twice a week, and he blames her directly for his status as a bachelor.58 Very often, however, these facts are told by way of AD, thus not only including the “reliable” third person narrator but the “unreliable” reflector himself in a sliding-zone of transition. Booth indirectly admits this in his concluding 58

Cf. James 1888: 125, 128.

The Fifth Mode of Representation

241

remarks at the bottom of the two columns: “One would hesitate to belabor what may seem obvious”.59 The only difference between his version and that of Henry James is that in James’s version there are no explicit (or clearly implicit) norms by which to condemn Lyon’s behaviour and motives. What Booth hereby indirectly shows is that the determination of a narrator’s unreliability to a large extent depends on the extra-textual framework the reader brings to the text.60 Booth, however, barely reflects this. Instead, and without motivation, he considers his own view identical to “the author’s”61, which is not necessarily the case. As Chatman notices, Booth somewhat overstates Lyon’s ‘viciousness’ and thereby actually repeats Lyon’s project, producing a caricature of Lyon strikingly similar to the caricature Lyon produces of Capadose. Secondly, Booth’s argument for his ‘columnization’ of the story’s selfevident facts, is that they have been neglected by former readers of “The Liar”, first and foremost by West & Stallman62; but while their reading was based exclusively on the 1888-version of the story, Booth’s reading is tacitly based on the New York edition, even though he pretends to be speaking of the first version only. For example, the basic binary concepts— “disinterested” versus “interested lying”, structuring Booth’s two columns and covering respectively the behaviour of Colonel Capadose and that of the reflector—are not introduced into the story by James until the 1908version.63 Here it is placed early in the story when the painter is informed about the true nature of Colonel Capadose by the old ‘lord’ whose portrait he is summoned to paint. In the first edition, the colonel is called a “thumping liar”, which is deleted in the second edition.64 Instead, the colonel’s lying is called harmless and “disinterested”, which is one of the strongest arguments against Lyon’s insisting on “the truth” of his painting.65 When Booth ridicules West & Stallman for claiming that Lyon in his intrigues is “inspired by the Muse of Truth”, he neglects that this phrase is a reference to Henry James’ thematic forerunner to “The Liar”, namely “The Story of a Masterpiece” (1868). Here, a painter makes a portrait of a 59 60

61 62 63 64 65

Booth 1991: 349. Cf. Cohn 2000: 312. See also the aforementioned studies of Ansgar Nünning for a radical take on this issue, and Hansen (2005, 2007, 2008/9) for an attempt to consider extra-textual unreliability as an autonomous (but not exclusive) narrative strategy out of four—the others being intranarrational, internarrational and intertextual unreliability. Booth 1991: 352. West and Stallman 1949. Cf. James 1908: 344. James 1908: 344, and 1888: 132. Ibid.

242

Poul Behrendt & Per Krogh Hansen

girl with whom he once was in love, thereby revealing her true nature to her present fiancé, who in despair destroys the masterpiece, but ends up marrying the girl even though he cannot love her anymore. In this story it is said about the painter, that the girl “might easily have accused him of injustice and brutality; but this fact would still stand to plead in his favour, that he cared with all strength for truth”.66 Finally, the entries regarding “The Liar” in Henry James’ Notebooks invite the unreflective reader to assume that Lyon represents the author himself in his attitude towards the married couple, corresponding to the overall absence of explicit narratorial criticism towards the artist in the first version of the story. Figural Invasion in the Modes of Representation Booth’s explanations of the fact that hardly any critic before him had spotted Lyon’s unreliability are as follows: First of all he notes that “the difference between Lyon’s voice and James’s voice, speaking behind and through the style,” does not always exist.67 Many of Lyon’s observations, comments and judgments, for instance, are sound. Secondly, he says that “to read the story properly we must combat our natural tendency to agree with the reflector”, and many have failed in this.68 Even though Booth’s explanations undoubtedly are correct, he still forgets the most important factor that might make us ignore the unreliability: namely the third person narration. As already mentioned in the introduction, Booth’s motivation for including reflector-characters in his considerations is that he does not acknowledge any significant difference between reflector and narrator characters—that is between third person or heterodiegetic narrators and first person or homodiegetic narrators. At first, this outset opens many insights into the complexities of multilayered narration, but it also—surprisingly—makes him disregard the important aspect of ‘distance’, which is examined closely in other parts of his book. On a general level it is evident that the function of third person narration is double: on the one hand it increases the distance to the narrated (a communicational axis between an extradiegetic narrator and narratee is established; it is not the acting agents who speak, but a voice placed in the intermediate space between author and diegetic universe). On the other hand, it decreases the distance to the narrated. Friedman, referring to Lub66 67 68

James 1961, I: 287. Cf. Martineau 1972: 18. Booth 1991: 352. Ibid.

The Fifth Mode of Representation

243

bock, noted this years ago when claiming that the reader “perceives the action as it filters through the consciousness of one of the characters involved, yet perceives it directly as it impinges upon that consciousness, thus avoiding that removal to a distance necessitated by retrospective firstperson narration”.69 Even though this double aspect of—a decreasing and increasing— distance in third person narration most certainly diffuses our attention towards narrative unreliability, we presumably will detect it in cases where the modes of representation of thought are strictly limited to one character. “The Liar” is restricted in this sense,70 but the picture is blurred insofar as the heterodiegetic narrator shows clear signs of both omniscience and authorial presence, while his voice at the same time blends with the reflector’s and occasionally comes close to merging with it. Uncertainty about what the narrator knows and on who speaks is noticeable from the very beginning of the story, which we will have to quote quite extensively: The train was half an hour late and the drive from the station longer than he had supposed, so that when he reached the house its inmates had dispersed to dress for dinner and he was conducted straight to his room. The curtains were drawn in this asylum, the candles were lighted, the fire was bright, and when the servant had quickly put out his clothes the comfortable little place became suggestive— seemed to promise a pleasant house, a various party, talks, acquaintances, affinities, to say nothing of very good cheer. [...] The walls were adorned with old-fashioned lithographs, principally portraits of country gentlemen with high collars and riding gloves: this suggested—and it was encouraging—that the tradition of portraiture was held in esteem. [...] It was a numerous party—five and twenty people; rather an odd occasion to have proposed to him, as he thought. He would not be surrounded by the quiet that ministers to good work; however, it had never interfered with his work to see the spectacle of human life before him in the intervals. And though he did not know it, it was never quiet at Stayes. When he was working well he found himself in that happy state—the happiest of all for an artist—in which things in general contribute to the particular idea and fall in with it, help it on and justify it, so that he feels for the hour as if nothing in the world can happen to him, even if it come in the guise of disaster or suffering, that will not be an enhancement of his subject.71

69 70

Friedman 1955: 1164. With one exception: when a strange female model intrudes into Lyon’s studio, the viewpoint for a brief moment is transposed from Lyon (who turns his eyes away from her) to Colonel Capadose (who later on accuses the same woman for having destroyed his portrait); however, it is impossible to decide whether the description of the woman is an FID of the colonel or a CID of the narrator; cf. James 1888: 216. 71 James 1888: 123. All underlinings (= (C)ID) and italics (= (F)ID) are ours; omniscience in bold; CID and FID simultaneously = Fifth Mode.

244

Poul Behrendt & Per Krogh Hansen

Even from the first lines of the story there is a sneaking ‘figural’ invasion into the seemingly character-(in)dependent survey (CID) of a nouveauriche manor and its inhabitants—because the duration of the drive from the station is measured on the expectations of the arriving painter. This tendency is increased when, at the end of the paragraph, the reflector partly takes over the narration through the ID-formulas suggestive, seemed and suggested with a sliding transition to FID. Paradoxically, in the last paragraph quite the opposite occurs, as the heterodiegetic narrator evidently expresses knowledge that Lyon does not possess: “And though he did not know it, it was never quiet at Stayes”. Furthermore, we also find remarks of nearly gnomic status, which could have been made by Lyon as well as the heterodiegetic narrator (as for instance in the general remark in the final lines of the quote about the happy state of the artist). Besides, in this text the whole extradiegetic communicational setting which characterizes most heterodiegetic narration takes on a certain ironic twist: The heterodiegetic narrator quickly establishes a communicational axis beyond the diegetic universe by referring to Lyon as “our artist”,72 “our hero”73 and “our friend”.74 Thus, he includes the reader directly in excuses for the hero’s behaviour: “Lyon’s curiosity on this point may strike the reader as fatuous, but something must be allowed to a disappointed man”.75 On top of this, in the first edition of the story the narrator on three different occasions changes identity from an impersonal to a first person narrator by referring to himself as an “I”. For instance, when he describes the way Lyon, behind a curtain at the balcony, discovers the Capadoses’ intrusion into his studio where they destroy the portrait of the colonel: “When I say he pushed it aside I should amend my phrase; he laid his hand upon it, but at that moment he was arrested by a very singular sound”.76 Half a page later this same “I” comments on Lyon’s behaviour when watching Mrs. Capadose throw herself, in a flood of tears, into the arms of her husband. This forces Lyon to take a step back again behind the curtain: I may add that it also had the force to make him avail himself for further contemplation of a crevice formed by his gathering together the two halves of the portiere. He was perfectly aware of what he was about—he was for the moment an eavesdropper, a spy; but he was also aware that a very odd business, in which

72 73 74 75 76

Ibid. 125, 126. Ibid. 125, 130. Ibid. 124, 131, 134, 135, 220. James 1888: 282 (213). Ibid. 299.

The Fifth Mode of Representation

245

his confidence had been trifled with, was going forward, and that if in a measure it didn’t concern him[,] in a measure it very definitely did.77

These instances, where the narrator manifests himself as an explicit “I”, are not included in Booth’s alleged handful of “unequivocal intrusions” in which the narrator distances himself from his reflector; that is, reveals or condemns him as “inconscient” (to make use of Henry James’ term). On the contrary, the narrator here shows a strange kind of solidarity with Lyon’s project, excusing him, once more, for having his confidence being trifled with. This is very far from Booth’s right-column summary of Lyon’s behaviour in the same situation: “Having sneaked back without any announcement, he deliberately eavesdrops”. According to the narrator, Lyon has no plan of eavesdropping; besides, he does not sneak back, but arrives at his studio without any expectations of finding the intruding Capadoses there. His eavesdropping is—in the eyes of the narrator—clearly not deliberate, neither is it a result of ‘inconscience’; but as a creation of the moment it is obviously excusable. By being occasionally present in first person, the narrator partly increases, partly decreases the distance to the reflector—as if the story were actually a first person narration in disguise; indicated, too, by the eerie way in which the narrator ‘personally’ knows what otherwise could only be known by divine omniscience—unless there were some kind of underground connection or sentimental identity between the protagonist and his narrator. Besides, by the manner in which he breaches the conventions of impersonal narration the narrator takes on the role of an author writing a story. For that reason, he might just as well be identified as—Henry James. Ambiguous Discourse: The Fifth Mode of Representation Only five years after the publication of The Rhetoric of Fiction, Dorrit Cohn published her ground-breaking article on the representation of consciousness in fiction: “Narrated Monologue: Definition of a Fictional Style”.78 Here she, too, pointed out Booth’s intent to do away with ineffectual classifications, like the “most overworked distinction” between the first and third person in fiction. She also quoted Booth for his wish, once for all to dismiss the concept of “erlebte Rede”, which in those days was a common expression for FID in the US: 77 78

Ibid. 300. Cohn 1966.

246

Poul Behrendt & Per Krogh Hansen

[…] the author who counted the number of times the word “I” appears in each of Jane Austen’s novels may be more obviously absurd than the innumerable scholars who have traced in endless detail the “Icherzählung”, or “erlebte Rede” or “monologue intérieur” from Dickens to Joyce or from James to Robbe-Grillet. But he is no more irrelevant to literary judgment.79

This statement from Essays in Criticism is the background for Booth’s doubtful claim that he cannot count more than four “unequivocal intrusions by the reliable narrator” that can be used to underline “the difference between Lyon’s picture of himself and the true picture”.80 Booth’s firm idea of what is “the true picture” has a striking resemblance with Lyon’s characterization of the caricature he is painting of colonel Capadose. This prevents him from undertaking any analysis of the many cases of ambiguous ‘fifth mode’ representation in “The Liar”, where it is uncertain whether it is the narrator (through CID) or Lyon (through FID) who speaks. This is best exemplified by the final words of the short story. After having disclosed Mrs. Capadose as yet another liar, Lyon decides to give up the project of reclaiming her: Lyon stopped at the corner, looking vaguely up and down. He would never go back – he couldn’t. She was still in love with the Colonel – he had trained her too well.81

Who is saying that the colonel, Mr. Capadose, has trained Mrs. Capadose ‘too well’? Is it the narrator? Or is it Lyon? On one level, the sentence is a repetition of a recurring formula: “Truly her husband had trained her well”; “Yes, her husband had trained her well”82—and in both cases it is rather uncertain whether it is the authorial narrator’s view of the relationship, or only Lyon’s that is expressed. In the quoted excerpt from the end of the short story, the previous sentence points in the direction of Lyon, due to the explicitly marked free indirect thought, but this is not necessarily the case. Here too, it might be the authorial narrator’s remark. Furthermore: Is it actually the colonel, to whom the pronoun ‘he’ refers, here, in the final paragraph? Is it not just as likely that it is the authorial narrator making an ironical judgment of Lyon and his pushy intrigues, forcing Mrs. Capadose to confess that she regrets having married the lying colonel and not the truthful painter? Paraphrasing Booth, we might therefore say that to recognize Lyon’s role as an unreliable narrator, we must combat our natural tendency to trust a heterodiegetic, omniscient narrator. We could even claim that the authorial narrator is not reliable in this case since he does not maintain the 79 80 81 82

Booth 1961: 60; Cohn 1966: 101. Booth 1991: 351. James 1888: 315. Ibid. 313-14.

The Fifth Mode of Representation

247

omniscient, gnomic status he at first shows signs of, but rather lets Lyon’s perspective be voiced more or less without comment. Booth finishes his reading of “The Liar” by discrediting the story, considering it “only half-developed”.83 He does this partly because he in general does not approve or value the ambiguity and unstable irony that characterizes a text like this; but mainly because he is very persistent in not acknowledging any difference between first and third person narration, and therefore does not consider the effect and function of the thirdperson voicing of the story. One cannot but wonder whether these ambiguities are as unintentional as Booth wants them to be? Is it not more likely that they are part of an authorial scheme, a reader trap, which Booth himself falls into when he judges Lyon harder than the evidence allows him to? When Booth hardly makes any reference to the normative instance that usually sanctions the unreliability, by imposing a ‘right’ set of norms and values as opposed to the unreliable narrator’s ‘wrong’ set, it may be due to the fact that this text does not have an implied author that fits Booth’s norms. Strange Voices – Concluding Remarks When Gérard Genette in Narrative Discourse presented his famous distinction between who sees and who speaks he solved problems of a longstanding confusion over point-of-view with questions regarding the function of the narrator. However, by reserving the metaphor of “voice” exclusively for the narrator, he created a new set of problems for the analysis of represented consciousness in third person narration. The three types of focalization he settled with—internal, external, and zero focalization—are much too coarse to grasp the subtleties connected with the most important of them: internal focalization. Thus, as Genette took in the concepts of DD, ID and FID, he put ID and FID together in one and the same category (discours transposé) and did not think it worthwhile to discern their fundamentally different functions in the representation of speech and thought respectively. However, as it has been demonstrated in this study, Genette’s concept of mood cannot capture the experience and interplay of voices that fascinates and deludes readers in the advanced technique of free indirect and ambiguous discourse we meet in Dinesen and James. As also stated by Monika Fludernik in 1996 (in flagrant contrast, apparently, to her formerly

83

Booth 1991: 353.

248

Poul Behrendt & Per Krogh Hansen

quoted view from 1993),84 in reflector-mode narrative the voice factor simply is constitutive for the “focalized interpretation”: Indeed, the question of ‘who speaks’ can no longer be taken literally at all because in reflector-mode narrative there is no discernible speaker function in the text; yet the narrative implicitly suggests that the discourse emanates from the focalized consciousness of the reflector character although that, of course, is the result of an illusionistic narrative technique.85

This is the best imaginable support for Booth’s unfortunately misguided claim that there is no essential difference in utterance between reflectormode and first person narrator-mode. There certainly is, but through an author’s subtle utilization of the convergence between the first mode of character-(in)dependent survey and the third mode of free indirect discourse into the fifth mode of representation (AD), strange voices emanate. They can only be perceived through the separation of the modes of discourse from the modes of narration by which the authorial voice and the voices of the characters reappear on a new level. As the grammatical features (Person, Inquit, Tense, Syntax) are parameters for the modes of representation on the level of discourse, so the designations of discourse (CID, ID, FID, DD & AD) are parameters for the dual-voice distribution on the narrative level. Likewise, the modes of narration—that is, the binary concepts of Authorial-Figural, ExternalInternal, Diegesis-Mimesis, Dating-Deixis—are parameters for breaks or intensifications of the grammatically designated distribution of voices in the spread between narrator and character, and thereby parameters for authorial or figural invasions of the modes of representation. This kind of dual-voice phenomenon transcends the traditional use of the concept and proves Booth’s idea of unreliable narration still to be highly relevant for the study of third person narration. However, unreliability in third person narration is no longer only defined as a moral deficit in a fictive ‘character-narrator’; but as a tacit, indirect relationship between an authorial narrator (or, perhaps, more precisely: an author who narrates) and a first-time reader who cannot rely on what he or she is told with regard to story-world facts. Narrative unreliability is a part of an ironic game with the reader being at one and the same time the hunter and the hunted.

84 85

Fludernik 1993: 453; cf. p. 7 above. Fludernik 1996: 344.

The Fifth Mode of Representation

249

References Banfield, Ann 1982 Unspeakable sentences, Boston: Routledge & Kegan Paul. Behrendt, Poul 2007 “Historien i historien. Læseren som hovedperson i Karen Blixens “Sorg-Agre” [”The Story in The History. The Reader as Protagonist in Isak Dinesen’s “Sorrow Acre”] in Den hemmelige note [The Secret Note], Copenhagen: Gyldendal. 2010 “Efterskrift. Dansk Genesis” [“Postscript. Danish Genesis”], in Isak Dinesen 2010. 2011 “Juryens Veto. Det skjulte paradigme under deklasseringen af “Skibsdrengens Fortælling” i de amerikanske udgaver af Winter’s Tales” [“The veto of the jury. The secret paradigm behind the degradation of “The Sailor Boy’s Tale” in the American editions of Winter’s Tales”]. Spring 30. Copenhagen: Spring. Booth, Wayne C. 1961 “Distance and Point-of-View: An Essay in Classification”, Essays in Criticism 11, 6079. 1991 [1961, 1983] The Rhetoric of Fiction, London: Penguin Books. Chatman, Seymour 1990 Coming to Terms. The Rhetoric of Narrative in Fiction and Film, Ithaca and London: Cornell University Press. Cohn, Dorrit 1966 “Narrated Monologue: Definition of a Fictional Style”, Comparative Literature 18, 97112. 1978 Transparent minds: narrative modes for presenting consciousness in fiction, Princeton, N.J.: Princeton University Press. 1995 “Optics and Power in the Novel”, New Literary History 26.1, 3-20. 2000 “Discordant Narration”, in Style 34.2, 307-16. Dinesen, Isak/Blixen, Karen 1942 Winter’s Tales, New York: Random House [Second edition 1943]. 1958 Anecdotes of Destiny, London: Michael Joseph. 2010 Karen Blixen Værker: Vinter-Eventyr. Edited by The Danish Society of Language and Literature (Nicolas Reinecke-Wilkendorf (text edition); Peter Olivarius & Henrik Blicher (commentaries), and Poul Behrendt (postscript), Copenhagen: Gyldendal. Engberg, Charlotte 2000 Billedets ekko. Om Karen Blixens fortællinger, Copenhagen: Gyldendal. Fludernik, Monika 1993 The fictions of language and the languages of fiction, London: Routledge. 1996 Towards a ‘natural’ narratology, London: Routledge. 2001 “Free indirect discourse”, in The Literary Encyclopedia, Oct. 20. www.litencyc.com. Friedman, Norman 1955 “Point of View in Fiction: The Development of a Critical Concept.” PMLA 70.5, 1160-84. Genette, Gérard 1980 Narrative Discourse. An Essay in Method, Ithaca: Cornell University Press. Gunn, Daniel P. 2004 “Free Indirect Discourse and Narrative Authority in Emma”, Narrative 12.1, 35-54. Hamburger, Käte 1993 The Logic of Literature, Bloomington: Indiana University Press. Hansen, Per Krogh 2005 “When Facts Become Fiction: On Extra-Textual Unreliable Narration”, in Fact and Fiction in Narrative: An Interdisciplinary Approach, edited by Lars-Åke Skalin, Örebro: Örebro University, 283-307. 2007 “Reconsidering the Unreliable Narrator”, Semiotica 165.1/4, 227-246.

250 2008

Poul Behrendt & Per Krogh Hansen

“First person, present tense. Authorial presence and unreliable narration in simultaneous narration”, in Narrative Unreliability in the Twentieth-Century First-Person Novel, edited by Elke D’Hoker & Günther Martens, Berlin and New York: De Gruyter, 317-338. 2008/9 “Unreliable Narration in Cinema. Facing the Cognitive Challenge Arising from Literary Studies”, Amsterdam International Electronic Journal for Cultural Narratology (ACJN). Heyd, Theresa 2006 “Understanding and handling unreliable narratives: A pragmatic model and method”, Semiotica 162.1/4, 217-243. Jahn, Manfred 1998 “Package-Deals, Exklusionen, Randzonen: Das Phänomen der Unverlässlichkeit in den Erzählsituationen”, in Unreliable Narration: Studien sur Theorie und Praxis unglaubwürdigen Erzählens in der englischsprachigen Erzähllitteratur, edited by Ansgar Nünning, Carola Surkamp and Bruno Zerweck, Trier: Wissenschaftlicher Verlag Trier. James, Henry 1868 “The Story of a Masterpiece”, in The Complete Tales of Henry James, ed. Leon Edel, 12 vols., Philadelphia and New York. 1888 “The Liar. In Two Parts”, in The Century. Illustrated Monthly Magazine, 36.1 (May 1888), Part 1, 123-35; 36.2 (June 1888), Part 2, 213-23, New York. 1889 A London life; The Patagonia; The liar; Mrs. Temperly, New York: Macmillan. 1908 “The Liar”, in The Novels and Tales of Henry James. The Aspern Papers. The Turn of The Screw. The Liar. The Two faces, New York Edition, Vol. XII, Charles Scribner’s Sons: New York, 311-388. Jauss, Hans Robert 1970 Literaturgeschichte als Provokation, Frankfurt am Main: Suhrkamp. Jørgensen, Bo Hakon 2000 Siden hen – om Karen Blixen, Odense: Odense University Press. LaCapra, Dominick 1982 Madame Bovary on trial, Ithaca, N.Y.: Cornell University Press. Langbaum, Robert 1964 The Gaiety of Vision: A Study of Isak Dinesen's Art, London: Chatto and Windus. Gunther Martens 2008 “Revising and Extending the Scope of the Rhetorical Approach to Unreliable Narration”, in Narrative Unreliability in the Twentieth-Century First-Person Novel, edited by Elke D’Hoker & Günther Martens, Berlin and New York: De Gruyter, 77-106. Martineau, Barbara 1972 “Portraits are murdered in the short fiction of Henry James,” Journal of Narrative Technique 2.1, 16-25. McHale, Brian 1978 “Free Indirect Discourse: A Survey of Recent Accounts”, Poetics and Theory of Literature (PTL) 3, 249-87. 1981 “Islands in the stream of consciousness. Dorrit Cohn’s Transparent Minds”, Poetics Today 2.2, 183-91. 2009 “Speech Representation”, in Handbook of Narratology. Contributions to Narrative Theory, edited by Fotis Jannidis, Matías Martínez, John Pier, Wolf Schmid (executive editor), Berlin and New York: Walter de Gruyter, 434-46. Nünning, Ansgar 1997 “’But why will you say that I am mad?’ On the Theory, History, and Signals of Unreliable Narration in British Fiction”, AAA. Arbeiten aus Anglistik und Amerikanistik 22, 1, 83-105.

The Fifth Mode of Representation

251

1998a “Unreliable Narration zur Einführung. Grundzüge einer kognitiv-narratologischen Theori und Analyse unglaubwürdigen Erzählens”, in Unreliable Narration: Studien sur Theorie und Praxis unglaubwürdigen Erzählens in der englischsprachigen Erzähllitteratur, edited by Ansgar Nünning, Carola Surkamp and Bruno Zerweck, Trier: Wissenschaftlicher Verlag Trier. 1998b “Unreliable Narrator”, Encyclopedia of the Novel, edited by Paul E. Schellinger, Chicago & London: Fitzroy Dearborn Publishers, 2, 1386-1388. 1999 “Unreliable, compared to what? Towards a Cognitive Theory of ‘Unreliable Narration’: Prolegomena and Hypotheses”, in Grenzüberschreitungen: Narratologien im Kontext/Transcending Boundaries: Narratology in Contextm edited by Walter Grünzweig and Andreas Solbach, Tübingen: Gunter Narr Verlag, 53-73. 2005 “Reconceptualizing Unreliable Narration”, in A Companion to Narrative Theory, edited by James Phelan and Peter J. Rabinowitz, Massachusetts and Oxford: Blackwell Pub., 89-107. Olson, Greta 2003 “Reconsidering Unreliability: Fallible and Untrustworthy Narrators”, Narrative 11.1, 93-109. Pascal, Roy 1977 The dual voice. Free indirect speech and its functioning in the nineteenth-century European novel, Manchester: Manchester University Press. Phelan, James 2005 Living to tell about it. A rhetoric and ethics of character narration, Ithaca: Cornell University Press. Riggan, William 1981 Picaros, Madmen, Naïfs, and Clowns: The Unreliable First-Person Narrator, Norman: University of Oklahoma Press. Selboe, Tone 1996 Kunst & Erfaring. En studie i Karen Blixens Forfatterskap, Odense: Odense Universitetsforlag. 2000 “Tone Selboes officielle opposition”, Nordica 17, 155-206. Stanzel, Franz K. 1984 A Theory of Narrative, Cambridge et. al.: Cambridge University Press. Wall, Kathleen 1994 “The Remains of the Day And Its Challenges to Theories of Unreliable Narration”, The Journal of Narrative Technique 24.1, 18-42. West, Ray B, Jr. and R. W. Stallman 1949 The Art of Modern Fiction, New York: Holt, Rineman and Winston. Yacobi, Tamar 2001 “Package Deals in Fictional Narrative: The Case of the Narrator’s (Un)Reliability”, Narrative 9.2, 223-229. Zerweck, Bruno 2001 “Historicing Unreliable Narration: Unreliability and Cultural Discourse in Narrative Fiction”, Style 35.1, 151-178.



BRIAN RICHARDSON (University of Maryland)

Unnatural Voices in Ulysses: Joyce’s Postmodern Modes of Narration From the outset, Joyce’s utilization of narration and narrators has attracted critical and theoretical attention; standard accounts of figural narration, “stream of consciousness,” and interior monologue have always drawn on Joycean examples. Nevertheless, I do not feel that the full extent of Joyce’s practices has yet been articulated; this is especially the case with the more unnatural or impossible kinds of narration in Ulysses. Below, I will provide an inventory of these practices, ranging from the merely original and moving to the logically impossible. Individuating Heterodiegetic Narration We may begin by noting Joyce’s “interiorization” of third-person or heterodiegetic narration, as the authorial prose describing a character often takes on the features of that person’s speech. Hugh Kenner called this the “Uncle Charles Principle” after a passage in A Portrait that states, admittedly tendentiously, “Every morning, therefore, Uncle Charles repaired to the outhouse but not before he had greased and brushed scrupulously his black hair and brushed and put on his tall hat” (Voices 16-17). The word “repaired” is decidedly hackneyed and old fashioned. Kenner’s point is that “it would be Uncle Charles’ own word should he chance to say what he was doing” (17). In these instances, third person narration ceases to be objective, neutral, or consistent; it is inflected by the sensibility of the figure it describes.1 This technique is used throughout Ulysses; thus, Reverend Conmee is described in terms that he might use to describe himself: “The superior, the very reverend John Conmee S. J. reset his smooth watch in his interior pocket as he came down the presbytery steps” (10.1-

1

For an additional, more wide ranging discussion of this general phenomenon, see Skalin.

254

Brian Richardson

2). It will be noted that this technique is both different from and similar to what Dorrit Cohn, elaborating Stanzel’s concept of figural narration, has termed “narrated monologue.” She defines it as a “transformation of figural thought-language into the narrative language of third-person fiction” (100). But the Uncle Charles principle is not a technique for rendering verbal thoughts; it is instead concerned with narrating events partly in the style of the character had that character been narrating. That is, it does not reproduce conscious thoughts; instead, it recreates a personal style of narrating. The Supernarrator Going beyond Franz Stanzel’s categories of first-person, third-person, and figural narration and Dorrit Cohn’s model of the representation of mental events as psychonarration, narrated monologue, and quoted monologue, we find that there is no established category in narrative theory for multiperson narration: that is, works with more than one narrator who relates events, none of which is ontologically primary. That is to say, we have the narrative of Esther Summerson within Dickens’ heterodiegetically narrated Bleak House. There is no problem here; the primary narrator is responsible for both Summerson’s words as well as the rest of the narrative. In the case of Faulkner’s The Sound and the Fury, however, there are four separate narrative sources (that is, the thoughts and words of Benjy, Quentin, Jason, and Dilsey); we need a new, more comprehensive category for the agent that produces all the book’s narrations. Zdzisáaw Najder has recently argued for “an unnamed supernarrator” in Conrad’s “Amy Foster,” Lord Jim, and Nostromo “who offers us the narratives of others,” and who “possesses the qualities of a person” (39). Given the multiple forms and styles of narration and narrator in Ulysses, we certainly need the figure of the supernarrator for this work. And at some points, such as the narration of “Eumaeus,” we may need to go even further. Since Joyce’s narration at times seems to transcend the framework of a single human sensibility, we might more precisely speak of “supernarration,” and thereby elude the humanist personification implicit in a concept like “narrator.” Beyond the “Arranger” The next question is how many primary narrators are there in the text, that is, beyond the various interior character narrators (the anonymous recounter of the Cyclops episode) or monologists (Molly Bloom’s stream

Unnatural Voices in Ulysses

255

of consciousness)? For some time, Joyce scholars argued for two such figures: the self-effacing primary narrator of the first half of the book and the more devious “arranger” who is responsible for the more outlandish acts of narration elsewhere in the text. The former is responsible for the line, “Mr Leopold Bloom ate with relish the inner organs of beasts and fowls” (3.1-2); the latter would be responsible for the later, playful, selfconscious line, “As said before he ate with relish the inner organs, nutty gizzards, fried cods’ roes while Ritchie Goulding, Collis, Ward ate steak and kidney, steak then kidney, bite by bite of pie he ate Bloom ate they ate” (11.519-22). David Hayman notes that this second, audacious narrator makes “a character address objects . . . letting hands talk, a tie rebel” in “Circe” (92). This radical “narrator is obliging us to accept another order of reality. He is playing on our need to naturalize and explain the strange and elusive, to close the field of experience. He is also asserting his independence, his freedom from the rules he himself has established” (92). Hayman goes on to assert that by the book’s second half, he will have assumed many faces, but still can be identified by a single impulse. This figure he calls the arranger (93). We may well ask whether two such primary narrators are enough. An existing debate centers on the myriad voice in “Cyclops.” Here are some representative samples: first, the demotic speech of the anonymous character-narrator: “I was just passing the time of day with old Troy of the D. M. P. at the corner of Arbour hill there but be damned but a bloody sweep came along and he near drove his gear into my eye” (12.1-3). Next, I’ll provide an early parody of a mellifluous, idealized style: “In Inisfail the fair there lies a land, the land of holy Michan. . . . There sleep the mighty dead as in life they slept, warriors and princes of high renown” (12.68-70). Then there is a blasphemous parody of the Apostle’s Creed: “They believe in rod, the scourger almighty, creator of hell upon earth, and in Jacky Tar, the son of a gun” (12.1354-55). There are also the Rabelaisian lists, including a bland and rather pointless roll call of clergy in attendance at a revival of traditional Irish sports (“William Delaney, S.J., L.L.D.; the rt rev. Gerald Molloy, D.D.; the rev. P. J. Kavanagh, C. S. Sp.; the rev. T. Waters, C.C.,” etc [12.927-38]); a long and rather overdone list of ladies whose names are implausibly associated with trees “Miss Virginia Creeper, Miss Gladys Beech, Miss Olive Garth, Miss Blanche Maple,” etc. (12.1268-78); and there is the more ribald enumeration of neglected Irish heroes, real and imagined, that includes Christopher Columbus, Benjamin Franklin, Patrick W, Shakespeare, and Brian Confucius (12.176-99). Karen Lawrence has identified a crucial feature of the curious narration in the Cyclops chapter that points to Joyce’s subversion of a modernist, mimetic poetics. Lawrence argues against the position articulated by

256

Brian Richardson

Marilyn French that assigns the discourse to two narrators, one a character in the story and another off-scene narrator who supplies the exaggerated materials. Lawrence points out that dual narrator position falsifies “our experience of the heterogeneity of the parodies" (102); the series of very different styles, interpolations, and amplifications cannot be reduced to a single voice or perspective; in addition, she points how even the chapter's two types of narrative stance “are not always separated from each other,” and that one form occasionally degenerates into the other (105). This clearly is the kind of ontological conflation characteristic of unnatural narration. More specifically, it resists attempts to situate the source of the discourse in the person of a single, self-consistent, human speaker, however ironic or avant-garde. Mimesis as Diegesis Two chapters in Ulysses, “Circe” and “Ithaca,” elude traditional forms of narration and model themselves on different, adjacent modes, respectively, the drama and the catechism. Not surprisingly, Joyce transforms both forms. The question of whether drama has a narrator is given particular urgency by the representations of speech, thought, event, and setting in “Circe.” This episode is written as if it were the scenario of a play, but it is a most unusual kind of drama. Manfred Jahn has argued that the stage directions and other dediscalia of a play text should be attributed to a narrator; this is clearly true of Joyce’s text, which includes distinctive narration near its beginning, “Round Rabaiotti’s halted ice gondola stunted men and women squabble. They grab wafers between which are wedged lumps of coral and copper snow. Sucking, they scatter slowly, children” (15.4-7). The narrated events become more unnatural as the drama continues and increasingly outrageous events are depicted, such as the appearance of Stephen’s dead mother, “her face worn and noseless, green with gravemould” (15.4159). Not only do the stage directions depict unusual and impossible events, the dialogue sections are likewise transformed, as inanimate objects are given speaking roles. Thus, the gas jet, a prostitute’s bracelets, and the pianola (“Best, best of all./ Baraabum!” [15.4106-7]) all utter sounds or words. The traditional relations between mimesis and diegesis and narration and description are here radically transformed or even inverted. Perhaps the most radical move in this strange text is the verbal generation of events. At one point, Bloom denigrates tobacco, Zoe retorts: “Go on. Make a stump speech of it” (15.1353). What then follows is a depiction of the figure of Bloom in workingman’s overalls, red tie, and apache cap,

Unnatural Voices in Ulysses

257

giving an oration on the evils of tobacco before an adoring populace: “Mankind is incorrigible. Sir Walter Raleigh brought from the new world that potato and that weed” (15.1356-57). The phrase “make a stump speech” precedes the event it simultaneously names and produces, as a line of dialogue sets forth a series of events. The “Ithaca” episode presents another unusual kind of narration. It employs the form of the catechism, itself an unusual genre. It is a scripted series of questions and answers performed by two or more individuals but written by neither; normally, it is a pseudo-dialogue. At the beginning of this chapter we get questions and answers like the following: Was there one point on which their views were equal and negative? The influence of gaslight or electric light on the growth of adjoining paraheliotropic trees.

This is, to be sure, a rather odd kind of catechism, more informational than doctrinal, and seems to be answering a simple question in a dehumanized, hypercorrect, scientistic rhetoric. One may well wonder who the speakers are, and what kind of narrator is producing this odd alternation of lines. The voices get stranger as the chapter proceeds, as the interlocutor and respondent become more multiform. At some points a genuine though brief dialogue starts to emerge, as the respondent provides the information desired by the interlocutor when a strictly literal answer would prove misleading (“Was this affirmation apprehended by Bloom?/ Not verbally. Substantively” (17.1017-18). These transformations continue until they conclude with the interlocutor’s call for clarification, “Womb? Weary?” (17.2319) and the respondent’s stumbling attempts to answer the text’s final three questions, With? When? and Where? The question, “When?” takes the respondent to the edge of meaning: “Going to dark bed there was a square round Sinbad the Sailor roc’s auk’s egg in the night of the bed of all the auks of the rocs of Darkinbad the Brightdayler” (17.2328-30). The last question, “Where?” leads the respondent beyond speech (and the conventions of typography) as the response is indicated merely by a large dot: Ɣ . Monika Fludernik notes that one encounters narratological difficulties here “because the questions ‘Womb? Weary?’, ‘With,’ ‘When,’ and ‘Where’ cannot be interpreted realistically or made to tally with Bloom’s posture in the bed” (“Ithaca” 94-95). Joyce’s interlocutor is an unstable and inherently protean figure (or kind of discourse) that regularly oscillates from one function to another as it invokes familiar categories like narrator and narratee in order to blur their edges or transgress them altogether. It is, I argue, another original category that deserves inclusion in a narratology that attempts to circumscribe the experiments of Ulysses and postmodern fiction in experimental narration. Again, it transcends humanistic models. Karen Lawrence notes

258

Brian Richardson

that although she uses “the term ‘narrator’ for ease of reference,” she prefers “the concept of consciousness or mind of the text, since Joyce does everything possible in ‘Ithaca’ to destroy our sense of a narrating, human voice” (183-84).2 Impossible Narration We may now go on to examine still more unusual, outrageous, and impossible acts of narration in the text. Returning to the “Cyclops” episode, we find an apparently impossible narrative situation: the character narrator describes himself leaving the group in order to urinate while the others, presumably out of earshot, discuss their mistaken belief that Bloom has just won a great deal of money by betting on the “dark” horse, Throwaway: So I just went round the back of the yard to pumpship and begob (hundred shillings to five) while I was letting off my (Throwaway twenty to) letting off my load gob says I to myself I knew he was uneasy in his (two pints off of Joe and one in Slattery’s off) in his mind to get off the mark to (hundred shillings is five quid) and when they were in the (dark horse) [. . . . ] Ireland my nation says he (hoik! phthook!) never be up to those bloody (there’s the last of it) Jerusalem (ah!) cuckoos” (12.1561-66; 1570-72).

A naturalistic recuperation of this strange passage would suggest that it might be a simultaneous re-enactment of the events that are being related, that the narrator is urinating and spitting again at the same point in his recounting as he did in its original enactment. But this passage includes the character’s own reproduction of his stream of thoughts at the time of the recounted events (“hundred shillings is five quid”); a more plausible interpretation is that Joyce has collapsed the two temporalities into a single vivid though impossible narrative stream.3 This allows the character’s past private thoughts to interpenetrate his own present tense narration. There are still more unnatural practices in this text. Hugh Kenner notes that in the Sirens chapter, Bloom reflects on the note he has written to Martha and wonders whether the postscript he had added, “I feel so sad today. So lonely,” was appropriate: “Too poetical that about the sad. Music did that. Music hath charms. Shakespeare said. Quotations every day in the year. To be or not to be. Wisdom while you wait.” At this point the narrator paraphrases and interpolates thoughts about Shakespeare which Stephen framed but did not speak two hours earlier: “In Gerard’s 2 3

See also Bernard and Shari Benstock on this subject. See Nielsen, “Fictional Voices? Strange Voices? Unnatural Voices?” in the present issue.

Unnatural Voices in Ulysses

259

rosary of Fetter Lane he walks, greyedauburn. One life is all. One body. Do. But do” (11.904). Kenner continues: “And Bloom, as if he had heard this remark or not heard it, continues, ‘Done anyhow’: meaning he has written what he has written” (Ulysses 64). Kenner goes on to explain that “some mind, it is clear, keeps track of the details of the printed cosmos”; for him, “the intrusion of this consciousness is perhaps the most radical, most disconcerting innovation of all in Ulysses” (Ulysses 64). Such unnatural interpolations are not uncommon in this work. As C. H. Peake explains, among the most extraordinary events in the hallucinatory Circe episode is the interpellation of the thought of one character into the mind of another without any possible naturalistic explanation. Thus, the formula “Nebrakada Feminum,” which Stephen read in a book a few hours earlier, is referred to by Bloom and Molly in Circe. Likewise, “in the passage where Bloom is accused of many crimes, Miles Crawford repeats the expression, ‘Paralyze Europe,’ which he spoke after Bloom had left the newspaper office” (268); Peake goes on to list numerous other examples and observes that while one or two such transgressions could be explained away or attributed to authorial carelessness, “their number makes such explanations impossible, and in most places the confusion is plainly deliberate” (269). They are not amenable to any naturalistic recuperation or even a preternaturalistic one like telepathy. Instead, Peake states that “they are part of a technique that makes no pretense of being confined to the minds, the space or the time of the characters” (269). The only mind capable of containing such thoughts is the mind of the author, he infers. We may conclude that Joyce is once again playing deviously with the conventions of representing minds in narrative fiction. The tradition states that a third-person or heterodiegetic narrator may disclose the contents of several minds, but the characters cannot perform this feat; first-person or homodiegetic narrators must in turn limit themselves to the knowledge of their own mind. In the history of narrative fiction, this rule has often been ignored (see Nielsen 2004) or, in the case of Nikolai Gogol’s “The Overcoat,” it has been travestied (see Richardson 2006: 7). What Joyce does is to combine these conventions, giving characters what seems to be an unwitting and inexplicable access to the mental data of others. The characters seem to be briefly accorded the powers of an omniscient novelist, though they never seem to realize that this is the case.4 Or Joyce is simply and flagrantly violating the rules for the representation of mental events that he has so meticulously established in the first ten chapters of Ulysses, 4

For an intriguing aspect of characters’ discursive appropriation of the presumed powers of novelists, see Mäkelä’s “Masters of Interiority” in the present issue.

260

Brian Richardson

and thereby drawing attention to the constructedness or even the artificiality of every form of narration. We find still other anomalous, unnatural bits of narration in this text. Early in the Cyclops chapter we overhear the anonymous narrator, a debt collector, recount a conversation he has had with a creditor, a plumber named Geraghty, whom he describes as “The most notorious bloody robber you’d meet in a day’s walk and the face on him all pockmarks would hold a shower of rain” (12.25-26). The narrator recounts Geraghty’s defiant talk: “Tell him, says he, I dare him, says he, and I doubledare him to send you around here again” (12.27-28). A few pages into the chapter, we find a curious semi-repetition of these lines: “I dare him, says he, and I doubledare him. Come on out here Geraghty, you notorious bloody hill and dale robber!” (12.100-01). The italicized portions constitute most of a sentence spoken by Geraghty as reported earlier in the anonymous narrator’s account; the rest is a slight variation of the phrase that this narrator uses to characterize him. But the ontological status of the second line, “Come on out here, Geraghty, you notorious hill and dale robber” is very dubious. It is very difficult to find a way to naturalize this strange sentence; instead, I argue that the two are run together as the different speakers are collapsed into each other. These paradoxes are problems as defined by (and unresolvable within) the very mimetic paradigm it seems first to evoke by playing with narrators (Brian McHale’s epistemological axis), and then to frustrate by refusing to provide the explanatory “parallax view” that would resolve the ambiguity. The narration is taking on an independent life of its own, free of the constraints of psychology, realism, or mimesis itself.5 Again, one needs to move away from the concept of a single narrator and instead revel in Joyce’s creation of narration unmoored to a single human-like entity. Authorial Self-Insertion Stephen Dedalus asserts that Shakespeare has subtly inserted himself into his work: “He has hidden his own name, a fair name, William, in his plays, a super here, a clown there, as a painter of old Italy set his face in a dark corner of his canvas” (9.921-23). And this is what Joyce, too, frequently does. There are numerous proto-postmodern conflations of author and narrator secreted within the text, from Stephen’s reported promise that he

5

For additional such examples, see Hazard Adams’ discussion of what he calls “rock” of the text that have somehow wandered into the wrong chapter.

Unnatural Voices in Ulysses

261

will produce something substantial in ten years’ time (which corresponds to Joyce’s own publication of Dubliners and Portrait in 1914, ten years after the artist as a young man made this claim) to Molly Bloom’s famous metafictional invocation articulated while she is menstruating: “O Jamesy, let me up out of this pooh” (18.1128-29). Here the presence of the author insinuates itself into his narration. Similarly, for decades critics wondered why Ulysses takes place on June 16, 1904. It is now evident the date was chosen because it was the time of Joyce’s first extended encounter with Nora Barnacle, the woman he would live with until his death. As Richard Ellmann observes, “To set Ulysses on this date was Joyce’s most eloquent if indirect tribute to Nora, a recognition of the determining effect upon his life of his attachment to her” (156). Again, Joyce frames his fictional narrative around crucial dates from his own life, as the author unnaturally penetrates his narrative fiction. Multiple Implied Authors The book’s aesthetic alters after the ninth episode, and becomes more expansive, unruly, and, by the time we reach “Eumaeus,” the sixteenth episode, even self-negating. Karen Lawrence writes that in the dead prose of this chapter, “Joyce chooses the ‘wrong’ word as scrupulously as he chooses the right one in the earlier chapters” (167). The differences are so vast that one may argue, not that we have a different narrator, but that we have a different implied author for this chapter; as Timothy Martin has observed, “the end of the novel is so different from the beginning that it might almost have been written by a different writer” (207). To some extent, this is a question of interpretation: formalists who view the novel as an organic totality will tend to infer a single implied author, however difficult it may be to convincingly fabricate a unified entity. Poststructuralists who see Ulysses as more decentered, fragmented, and heterogenous will have no trouble constructing more than one implied author. And to those who argue that there should be no problem in postulating a single narrator or a single implied author of the entire text, since, after all, a single man wrote the entire book, we might respond that the conclusion does not follow from its premises since a historical author can indeed create two different implied authors in a single text, as is demonstrated by Chaucer’s retraction, in which the same historical author denounces the successes achieved by the Canterbury Tales. One might also question the notion of a single, fixed authorial self: was the Joyce of 1914, when he started the book, the same man in 1922, when he finished it?

262

Brian Richardson

Conclusion Drawing on the work of a number of Joyce scholars, I have tried to outline the full range of Joyce’s unusual and unnatural practices of narration. These practices have all tended to move in the same general, anti-realist direction. The opposition between an objective third-person narration and a subjective first-person discourse is collapsed through the “subjectification” of the traditional means of objective narration. The multiplicity of narrative practices that seem to call for a narrator and an arranger or better, a supernarrator, is challenged by antithetical or incommensurate styles of narration that cannot be reasonably contained within a single figure, voice, or style. As Ulysses progresses, we find ever more radical, unnatural, and impossible kinds of narration. Joyce both establishes and then violates independent, autonomous sources of narration as he draws on and then moves away from a realist or humanist poetics. He abrogates virtually every binary opposition that mimetic narratives presuppose. Repeatedly, he dissolves the notion of a single, human-like narrator who is telling a story. For this text, we must replace the concept with a broader, more fluent category, as the narrator (and even a single implied author) are supplanted by a depersonalized, deindividualized idea of narration. Concerning the larger issues of narratology, we may likewise use Joyce’s example to corroborate a relatively new position that is gaining ground in narrative theory. Here, too, one can no longer presume the existence of a real human or a human-like narrator standing behind and above the discourse; in many works, there is only the discourse itself. Monika Fludernik has (1996: 278-303) identified a number of non- or transmimetic narrating situations, including those of contemporary narratives that are “no longer recuperable as the consciousness of a narrator figure because the juxtaposed material was too heterogenous either stylistically or thematically to warrant integration as part of a verisimilar stream of consciousness” (287). Marie-Laure Ryan (2001) has also embraced this position, noting that “the narrator is a theoretical fiction, and that the human-like, pseudonatural narrator is only one of its many possible avatars” (152). Such a stance is also argued for throughout my book, Unnatural Voices. We need a theoretical model of narration that can embrace the full spectrum of narrative voices, from the most compellingly human-like to the most radically decentered and posthuman. I hope this anatomy of Joycean strategies traced above confirms the importance of this position and firmly situates Joyce as a preeminent author of unnatural narratives.

Unnatural Voices in Ulysses

263

References Adams, Hazard 1990 “Critical Constitution of the Literary Text: The Example of Ulysses,” in Antithetical Essays in Literary Criticism and Liberal Education, Tallahassee: Florida State University Press, 90-110. Benstock, Bernard and Shari 1982 “The Benstock Principle”, in The Seventh of Joyce, edited by Bernard Benstock, Bloomington: Indiana University Press, 10-21. Cohn, Dorrit 1978 Fictional Minds: Narrative Modes Presenting Consciousness in Fiction, Princeton: Princeton University Press. Ellmann, Richard 1982 James Joyce, revised edition. Oxford: Oxford University Press. Fludernik, Monika 1986 “‘Ithaca’. An Essay in Non-Narrativity,” in International Perspectives on James Joyce, edited by Gottlieb Geiser, Troy NY: Whitson, 90-107. 1996 Towards a ‘Natural’ Narratology, New York: Routledge. Hayman, David 1982 Ulysses: The Mechanics of Meaning, rev ed., Madison: University of Wisconsin Press. Jahn, Manfred 2001 “Narrative Voice and Agency in Drama: Aspects of a Narratology of Drama”, New Literary History 32.3, 659-80. Joyce, James 1986 Ulysses, New York: Random. Kenner, Hugh 1978 Joyce’s Voices, Berkeley: University of California Press. 1987 Ulysses, revised edition, Baltimore: Johns Hopkins University Press. Lawrence, Karen 1981 The Odyssey of Style in Ulysses, Princeton: Princeton University Press. Martin, Timothy 1998 “Ulysses as a Whole”, A Collideorscape of Joyce, edited by Ruth Frehner and Ursula Zeller, Dublin: Lilliput, 202-14. McHale, Brian 1987 Postmodernist Fiction, New York: Methuen. Najder, Zdzisáaw 2008 “The Personal Voice in Conrad’s Fiction”, in Joseph Conrad: Voice, Sequence, History, Genre, edited by Jakob Lothe, Jeremy Hawthorn, and James Phelan, Columbus OH: Ohio State University Press, 23-40. Nielsen, Henrik Skov 2004 “The Impersonal Voice in First-Person Fiction,” Narrative 12.2, 133-50. Peake, C. H. 1977 James Joyce: The Citizen and the Artist, Stanford: University of Stanford Press. Richardson, Brian 2006 Unnatural Voices: Extreme Narration in Modern and Contemporary Fiction, Columbus: Ohio State University Press. Ryan, Marie-Laure 2001 “The Narratorial Function: Breaking Down a Theoretical Primitive,” Narrative 9.2, 146-52. Skalin, Lars-Åke [forthcoming] “Narrator Text, Character Text, and the Question of Reliability in ThirdPerson Literary Fiction”.



Index Abbott, H. Porter; 49; 53 Aczel, Richard; 8; 16; 20; 23; 27; 33; 34; 175; 176; 178; 188; 192; 209; 210; 215 Adams, Hazard; 260; 263 Agamben, Giorgio; 140 Alber, Jan; 55; 58; 67; 77; 80; 81; 127; 145; 146 Altes, Liesbeth Korthals; 3; 4; 21; 25; 27; 31; 33; 35; 127; 145 Altman, Janet Gurkin; 197; 198; 201; 215 Altschuler, Eric Lewin; 186; 188 Amis, Martin; 184 Andersen, Rikke Kragelund; 5 Aristotle; 103; 105; 106; 110; 113; 114 Austen, Jane; 220; 225; 232; 246 Auster, Paul; 94; 96; 97; 99 Bakhtin, Mikhail; 8; 9; 175; 177; 179; 181; 187; 188; 189; 206; 215; 217 Bal, Mieke; 26; 34 Bally, Charles; 206; 215 Banfield, Ann; 3; 4; 5; 14; 15; 18; 25; 27; 28; 29; 30; 31; 32; 33; 34; 136; 191; 216; 225; 249 Barnet, Sylvan; 45; 53 Baron-Cohen, Simon; 134; 137; 145 Barthes, Roland; 25; 34; 147; 148; 150; 161; 172 Beckett, Samuel; 184 Beck-Nielsen, Claus; 137; 138; 139; 142; 144 Behrendt, Poul; 9; 10; 219; 234; 235; 237; 249 Bender, John; 191; 192; 216 Benstock, Bernard and Shari; 258; 263 Benveniste, Émile; 2; 3; 18; 20; 21; 34; 35 Bergson, Henri; 46; 53 Berman, Morton; 53 Bin Laden, Osama; 140 Blanchard, Marc; 33; 34 Blixen, Karen; 234 Booth, Wayne C.; 10; 193; 216; 220; 221; 222; 237; 238; 239; 240; 241; 242; 245; 246; 247; 248; 249 Bray, Joe; 198; 199; 216 Brenkman, John; 23; 34 Brown, Kate E.; 180; 188 Bundgaard, Peer; 6; 7; 83; 85; 88; 92; 98; 99

Burto, William; 53 Butor, Michel; 8; 147; 148; 149; 153; 154; 158; 160; 161; 162; 165; 167; 171; 172; 173; 174 Carver, Raymond; 94 Cervantes; 119; 120 Chafe, Wallace; 87; 89; 92; 99 Chatman, Seymour; 5; 13; 21; 22; 23; 24; 25; 26; 29; 32; 33; 34; 185; 188; 189; 222; 223; 236; 237; 241; 249 Chaucer, Geoffrey; 261 clausbeck-nielsen.net; 7; 128; 138; 139; 142; 145 Coetzee, J. M.; 9; 207; 208; 211; 212; 216 Cohn, Dorrit; 38; 53; 59; 60; 61; 75; 80; 90; 99; 103; 109; 125; 181; 188; 191; 192; 195; 205; 212; 215; 216; 220; 221; 224; 225; 227; 231; 236; 241; 245; 246; 249; 250; 254; 263 Conrad, Joseph; 99; 254; 263 Cosmides, Leda; 7; 128; 129; 130; 131; 132; 133; 134; 135; 136; 137; 138; 143; 144; 145 Costal, Allan; 146 Culler, Jonathan; 125; 149; 172 Defoe, Daniel; 136 DelConte, Matt; 63; 64; 80; 148; 172 Derrida, Jacques; 9; 140 Dickens, Charles; 107; 109; 111; 113; 116; 117; 118; 119; 125; 254 Diderot; 121 Dinesen, Isak; 10; 220; 222; 223; 225; 226; 230; 231; 232; 233; 234; 235; 236; 237; 247; 249; 250 Doležel, Lubomír; 181; 188 Dostoevsky, Fyodor; 183; 188 Ellis, Bret Easton; 59; 60; 63; 80; 221 Ellmann, Richard; 261; 263 Engberg, Charlotte; 226; 249 Eysteinsson, Astradur; 184; 188 Farley, Ralph Milne; 154; 155; 156; 162; 173 Faulkner, William; 254 Ferriss, Lucy; 210; 216 Fitzgerald, Francis Scott; 30; 36 Flaubert, Gustave; 205; 216; 224 Fludernik, Monika; 3; 6; 8; 11; 17; 20; 27; 31; 34; 37; 53; 57; 58; 61; 62; 63; 76;

266 77; 80; 81; 103; 117; 118; 119; 121; 125; 147; 148; 149; 151; 152; 153; 154; 155; 157; 158; 159; 160; 161; 162; 163; 164; 165; 166; 168; 169; 173; 175; 176; 188; 191; 192; 194; 195; 198; 209; 210; 216; 221; 222; 225; 229; 247; 248; 249; 257; 262; 263 Frank, Joseph; 106; 125 French, Marilyn; 256 Frey, James; 136 Friedman, Norman; 242; 243; 249 Frith, Christopher D.; 134; 137; 145 Gallagher, Shaun; 129 Genette, Gérard; 2; 3; 5; 6; 13; 15; 16; 17; 18; 19; 20; 21; 22; 24; 25; 26; 29; 33; 34; 35; 36; 37; 38; 39; 40; 41; 42; 43; 44; 45; 46; 48; 52; 53; 84; 86; 87; 96; 99; 101; 117; 119; 121; 125; 127; 135; 136; 145; 152; 155; 156; 157; 158; 160; 161; 163; 170; 173; 175; 222; 247; 249 Georgakopoulou, Alexandra; 176; 178; 188 Gibson, Andrew; 3; 31; 33; 35 Ginsburg, Michael Peled; 196; 205; 214; 216 Goffman, Erving; 176; 184 Gogol, Nikolaj; 186; 188; 259 Goldsmith, Oliver; 221 Graves, Robert; 68; 69; 80 Gregersen, Frans; 97; 99 Greimas, Julien; 175 Grishakova, Marina; 2; 8; 175; 176; 188; 189 Gunn, Daniel P.; 192; 216; 225; 249 Gupta, Sunetra; 165; 172; 173 Göransson, Ulrika; 106; 125 Hamburger, Käte; 3; 4; 15; 35; 84; 87; 99; 105; 106; 110; 114; 125; 225; 227; 249 Hansen, Per Krogh; 2; 9; 10; 55; 63; 77; 80; 83; 212; 216; 219; 237; 241; 249 Hantzis, Darlene Marie; 151; 173 Harvey, David; 45; 53 Hayman, David; 255; 263 Heinze, Rüdiger; 58; 67; 80; 81 Hemingway, Ernest; 87; 88; 89; 91; 92; 97; 99 Herman, David; 166; 167; 168; 170; 173; 175; 189; 196; 216 Heusden, Barend van; 132; 146 Heyd, Theresa; 221; 250 Husserl, Edmund; 87; 99 Hutto, Daniel D.; 129; 145 Hydén, Lars-Christer; 176; 179; 189

Index Ishiguro, Kazuo; 221 Israel, Samuel Ben; 176; 189 Iversen, Stefan; 7; 55; 80; 83; 127; 129; 145; 146 Jahn, Manfred; 88; 99; 175; 176; 177; 187; 189; 213; 216; 222; 237; 250; 256; 263 Jakobson, Roman; 107; 125 James, Henry; 10; 220; 223; 229; 237; 238; 241; 242; 243; 244; 245; 246; 247; 250 Jauss, Hans Robert; 224; 250 Jongeneel, Els; 201; 216 Joyce, James; 10; 11; 253; 254; 255; 256; 257; 258; 259; 260; 261; 262; 263 Jørgensen, Bo Hakon; 226; 250 Kacandes, Irene; 8; 148; 149; 150; 152; 162; 166; 167; 171; 173 Kafka, Franz; 220 Kellogg, Robert; 81 Kenner, Hugh; 210; 216; 253; 258; 259; 263 Kjærstad, Jan; 5; 38; 43; 44; 45; 47; 48; 49; 50; 52; 53 Kraglund, Rikke; 127 Kuroda, S.-Y.; 4; 5; 13; 14; 15; 18; 25; 26; 27; 29; 35 Kushner, Howard I.; 180; 188 Labov, William; 110; 112; 125 LaCapra, Dominick; 224; 250 Laclos, Choderlos de; 199; 201; 216 Langbaum, Robert; 226; 250 Lawrence, Karen; 255; 257; 261; 263 Leslie, A. M.; 132; 133; 134; 146 Lessing; 106; 122 Leudar, Ivan; 146 Livingston, Paisley; 84; 99 Lock, Charles; 206; 217 Lubbock, Percy; 243 Lundholt, Marianne Wolff; 2 Lydon, Marie; 172; 173 Margolin, Uri; 8; 148; 159; 166; 167; 173 Martens, Gunther; 220; 250 Martin, Timothy; 261; 263 Martineau, Barbara; 242; 250 McEwan, Ian; 96; 97; 98; 99 McHale, Brian; 181; 189; 191; 196; 205; 217; 223; 224; 229; 231; 250; 260; 263 McInerney, Jay; 151; 153; 154; 161; 166; 168; 169; 173 McKeon, Michael; 106; 125; 198; 217 Mezei, Kathy; 192; 217 Miller, D. A.; 191; 192; 217 Miller, James; 180; 189 Milner, Jean-Claude; 13; 35

Index MukaĜovský, Jan; 9; 178; 189 Muratore, Mary Jo; 204; 217 Murdoch, Iris; 105; 106; 125 Mäkelä, Maria; 9; 127; 191; 213; 217; 219; 259 Nabokov, Vladimir; 135; 184; 185; 186; 189; 190; 221 Najder, Zdzisáaw; 254; 263 Nielsen, Henrik Skov; 2; 6; 15; 30; 35; 55; 58; 67; 75; 77; 80; 81; 83; 92; 127; 128; 136; 140; 145; 146; 176; 189; 211; 212; 215; 217; 258; 259; 263 Nünning, Ansgar; 39; 53; 117; 122; 125; 221; 241; 250; 251 O' Neill, Patrick; 51; 53 Olson, Greta; 221; 251 Page, Ruth; 176; 189 Palmer, Alan; 191; 195; 198; 214; 217 Pascal, Roy; 192; 196; 204; 205; 217; 224; 225; 231; 251 Patron, Sylvie; 3; 4; 5; 14; 18; 23; 27; 32; 35 Peake, C. H.; 259; 263 Perec, Georges; 147; 165; 167; 172; 173 Petitot, Jean; 85; 99 Peuter, Jennifer de; 176; 189 Phelan, James; 2; 4; 6; 38; 53; 56; 60; 64; 65; 67; 76; 77; 78; 79; 81; 101; 104; 122; 125; 126; 144; 175; 189; 211; 217; 218; 221; 251 Philippe, Gilles; 23; 35 Pilkington, Adrian; 176; 189 Pillai, Johann; 70; 74; 81 Plato; 120 Poe, Edgar Allan; 6; 56; 70; 71; 72; 73; 74; 81; 82; 91; 97; 99; 221 Pratt, Mary Louise; 126 Premack, D.; 129; 146 Prince, Gerald; 2; 8; 26; 35; 104; 126; 178; 189 Pritchard, Hollie; 70; 81 Rabinowitz, Peter J.; 38; 53; 126 Rajan, Gita; 70; 81 Reitan, Rolf; 2; 3; 8; 15; 38; 53; 55; 59; 65; 66; 67; 83; 127; 147; 170; 173; 215 Richardson, Brian; 4; 8; 10; 40; 48; 52; 53; 55; 65; 66; 80; 81; 103; 126; 127; 128; 145; 146; 148; 151; 153; 155; 158; 159; 163; 166; 167; 168; 171; 173; 175; 176; 184; 189; 259; 263 Richardson, E. Arthur; 81 Richardson, Samuel; 198; 201; 205; 217; 218 Ricoeur, Paul; 109; 126

267 Riggan, William; 221; 251 Rimmon-Kenan, Shlomith; 17; 18; 23; 26; 35; 37; 39; 53 Ron, Moshe; 205; 209; 217 Ross, John R.; 5 Rossum-Guyon, Françoise van; 150; 171; 174 Roth, Philip; 92; 93; 97; 99 Royle, Nicholas; 140 Ryan, Marie-Laure; 2; 8; 11; 26; 35; 102; 126; 177; 178; 187; 189; 262; 263 Schaeffer, Jean-Marie; 16; 35; 36 Schiffrin, Debora; 177; 184; 190 Schofield, Dennis; 148; 174 Scholes, Robert; 81 Sclater, Shelley Day; 188; 190 Searle, John; 100; 102; 104; 126 Selboe, Tone; 226; 231; 251 Seltzer, Mark; 191; 192; 217 Shakespeare, William; 89 Shen, Dan; 18; 36 Shklovsky, Victor; 46; 47; 54 Simon, Paul; 83 Skalin, Lars-Åke; 2; 7; 101; 106; 112; 114; 126; 253; 263 Smith, Ali; 179; 181; 185; 190 Smith, Barbara Herrnstein; 101; 104; 117; 126 Soelberg, Nils; 19; 36 Sokolov, Sasha; 186; 187; 190 Sperber, Dan; 132; 145; 146; 176; 190 Stallman, Robert Wooster; 241; 251 Stanzel, Franz K.; 2; 5; 13; 24; 25; 26; 27; 29; 36; 87; 90; 99; 100; 101; 126; 193; 194; 217; 221; 251; 254 Steinberg, Günter; 174 Sternberg, Meir; 56; 81; 96; 100; 212; 213; 218; 221 Story, W. W.; 70; 71; 81 Suarez, Virgil; 153; 160; 161; 168; 169; 174 Tammi, Pekka; 2; 6; 57; 58; 67; 75; 76; 81; 181; 190 Tannen, Deborah; 179; 180; 190 Thomas, Bronwen; 176; 189 Tommola, Hannu; 2 Tooby, John; 7; 128; 129; 130; 131; 132; 133; 134; 135; 136; 137; 138; 143; 144; 145 Tröhler, Margrit; 176; 190 Twain, Mark; 114; 117; 124; 126 Vega y Vega, Jorge Juan; 171; 174 Virgil; 43; 54 Vološinov, V. N.; 206; 218

268 Waletzky, Joshua; 110; 112; 125 Wall, Kathleen; 221; 251 Walsh, Richard; 3; 4; 78; 81 Walton, Kendall; 102; 126 Watt, Ian; 198; 218 Weinreich, Harald; 170; 174 Weiss, Peter; 85 West, Ray B., Jr.; 241 Wildgen, Wolfgang; 132; 146 Wilson, Deirdre; 176; 190 Witherington, Paul; 70; 81

Index Wittgenstein, Ludwig; 104 Wolf, Werner; 117; 126 Woodruff. G.; 129; 146 Woolf, Virginia; 28; 36 Yacobi, Tamar; 193; 212; 213; 218; 221; 251 Zerweck, Bruno; 221; 251 Zimmerman, Brett; 70; 82 Zunshine, Lisa; 7; 128; 129; 135; 136; 137; 144; 146; 198; 213; 218