The Fictions of Language and the Languages of Fiction

Monika Fludernik

• Publisher: Routledge
• Year: 2003
• Print ISBN: 9780415092265
• eBook ISBN: 9781134872862
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

Monika Fludernik presents a detailed analysis of free indirect discourse as it relates to narrative theory, and the crucial problematic of how speech and thought are represented in fiction.
Building on the insights of Ann Banfield's Unspeakable Sentences, Fludernik radically extends Banfield's model to accommodate evidence from conversational narrative, non-fictional prose and literary works from Chaucer to the present.
Fludernik's model subsumes earlier insights into the forms and functions of quotation and aligns them with discourse strategies observable in the oral language. Drawing on a vast range of literature, she provides an invaluable resource for researchers in the field and introduces English readers to extensive work on the subject in German as well as comparing the free indirect discourse features of German, French and English.
This study effectively repositions the whole area between literature and linguistics, opening up a new set of questions in narrative theory.

1
Contexts: the reporting of speech and thought acts

This is a chapter of contexts philosophical, linguistic, literary and narratological. Speech and thought representation has been an issue of central concern for a variety of very different disciplines, and an issue that was discussed under a number of titles specific to the concerns of those disciplines. The survey presented here is designed to locate phenomena of quoting and reporting in their multiple interdisciplinary perspectives and to demonstrate why they became such a problem within so many separate paradigms. The following presentation additionally aims at providing a frame of reference for the chapters that follow. On the one hand, the philosophical discussion constitutes a kind of cas limite of the problems of reported discourse—it, for instance, helps to explain some of the preoccupations of studies on indirect discourse by linguists such as Barbara Hall Partee (1973b, 1984) and other contributors of Linguistics and Philosophy (Dowty 1982,1986; F.Heny 1981a, 1982; B.Richards 1981,1982; P.Peterson 1982; Wreen 1989). Like deixis, the topics of speech representation and intensional logics have acquired a central, potentially disruptive relevance as truth-semantically marginal issues that are threatening to undermine the general applicability of truth semantics. For the present book, this philosophical perspective constitutes a horizon of my own inquiry, which centres on more linguistic and literary concerns. The philosophical perspective is so important because of its influence on several more narrowly relevant philosophical and linguistic studies which grew out of questions originally conceived within philosophy. Thus Castañeda’s seminal work on consciousness and the self (Castañeda 1966, 1967a, 1967b, 1968) needs to be situated within a debate between philosophers of truth semantics.¹ Likewise, speech act theory, a philosophical discipline in its origins, developed directly from the insufficiencies of the truth-value analysis of what Austin then called ‘performative’ utterances. Even more specifically, the linguistic issue of deixis poses some crucial philosophical problems, and it is interesting to observe how, for instance, Jakobson departs from Peirce’s remarks on Russell’s ego-centric particulars (Jakobson 1956/1971). The philosophical debate about deixis and naming between Frege, Russell, Peirce, Strawson, Kripke, Quine, Putnam and others is regularly invoked by standard linguistic studies of deixis, as already by Bühler or Benveniste.

That quotation and report should thus be issues which challenge the concept of truth values and therefore of analycity transfers the discussion about reporting others’ discourse on a meta-disciplinary plane, where inductive (linguistic) and deductive (truth-semantical) methodologies redefine each other’s proceedings. Ultimately, linguistics then starts to restructure the philosophical discourse about language, and linguistic discourse in turn becomes aware of its methodological inadequacies. After all, the claim to describe language empirically entails the ability to step outside language. If one departs from the presupposition that our thought is structured by the very system of language that one seeks to analyse, then the observer’s intrinsic relatedness to the object of observation is nowhere more apparent than in the linguistic attempt at an objective description of language. If linguistic empirical observation becomes feasible at all, it does so on the basis of users’ concepts of (functional) difference and structural meaningfulness. Chomsky’s reliance on informants’ intuitions thus constitutes an ultimate baring of the devices of linguistic research—linguistic units have always been discovered empirically because users intuited their existence, and they did so because they felt that those units ‘made a difference’.

Pragmatic research and discourse analysis have recently demonstrated the limits of prescriptive systematicity in linguistics, but even their apparently so unassailable empiricity (supported by computerized analyses of oral discourse) does not escape the snare of the observer’s entanglement with the object of analysis. No linguistic analysis, not even the most scientific, can eschew the requirements of relevance (cp., for instance, Ducrot 1989, chs 5–6); description may be infinite, but the parameters of a truly valid characterization of oral discourse eventually reduce themselves to the very same dependency on what human ears and minds pick out as relevant. As in other disciplines (Einsteinian physics, post-Heisenbergian quantum mechanics, or medicine), empiricity here encounters both its limits and its ultimate self-determinacy. Models of language, like models of the universe, cannot but remain provisional theories that may help to explain, or help to direct, but can no more lay claim to an ulterior truth than can the theoretical systems of mathematics or medicine or astronomy. Language, like everything else, needs to be studied in its linguistic and human context. Recent developments in cognitive linguistics have in fact taken cognizance of this boundedness of language within cognitive parameters, proposing a ‘natural’ perspective from which language and cognition are structured in existentially relevant terms. Although such a model (see, for instance, Dressler 1989,1990; G.Lakoff 1990; Langacker 1990) does not escape the ‘binding’ of subjectivity, it does account for human situatedness in language on the theoretical plane— certainly a major advance over non-pragmatic linguistics, which generates an illusion of empirical objectivity.

This is a book in which method and methodology are everywhere centrally at issue, and nowhere does this perhaps appear more clearly than in the approaches characterized in this chapter. Reported speech and the representation of consciousness are crucially central marginal issues which constitute a challenge for the respective disciplinary paradigms. Before considering the central side issue of free indirect discourse—a test case within the paradigm of speech and thought representation—we will therefore place the reporting of others’ discourse within the frames that have hitherto been used to contain it and which it has reframed in turn.

1.1
DIEGESIS AND MIMESIS—LITERARY APPLICATIONS IN THE GUISE OF GRAMMAR

I am starting my survey of contexts with the Platonic distinction between diegesis and mimesis (Republic III, 392D-394E), which closely correlates with the pre-linguistic analysis of direct and indirect discourse, oratio recta and oratio obliqua. If I say pre-linguistic, I mean to suggest that no formal (linguistic) analysis of the direct and the indirect is provided in The Republic; this distinction is simply presupposed as intuitively present to Plato’s audience. In fact Plato does not offer any grammatical definition of the direct and indirect; his terminology is entirely in terms of narration versus speeches, with ‘impersonation’ contrasted disfavourably with the narrator’s ‘speaking in his own voice’. Plato’s formulations have given rise to numerous conflicting interpretations, and the results of these can be observed today in narrative theory, in philosophy and even in basic linguistic analysis. If Plato’s text did not in fact suggest any linguistic distinctions, the ‘natural’, intuitive linguistic interpretation of the classic passage helped to confer upon his statements the status of fundamental observation. As a consequence, later scholars have frequently been blinded as to the specificity and decidedly literary quality of Plato’s remarks.

The narratological analogy between direct and indirect speech representation on the one hand and mimetic unmediated representation versus mediated summary by the narrator on the other derives directly from Plato’s example text and has helped to confuse the issue of speech and thought rendering with the question of narrative representation and mediation in general. There are then—at least factually—two quite different areas that have become irretrievably bound up with each other: grammar, i.e. the different linguistic ways of representing speech; and the status of direct and indirect discourse in narrative texts, and their relation to the narrative instances involved (narrator, characters).

The confusion between these two levels or areas of application is entirely due to Plato’s original example, the plea by Chryses to Agamemnon to release his daughter (Iliad I, 15– 21). In the Iliad this plea is rendered in direct speech, whereas Plato transforms it into indirect discourse (393E ff.). The example is noteworthy on several counts. Although the attack against mimesis in The Republic is directed against the impersonation on stage of lowerclass characters, Plato’s illustration of mimesis is taken from the Iliad, a narrative text in which the notion of mimesis can ostensibly be argued to have shifted in signification. Secondly, in the example provided, the seer addresses Agamemnon, and the narratorial description of the utterance is therefore in terms of the illocutionary speech act of imploration: ‘he prayed that’. Rather than a neutral ‘He said that’ the narrator’s indirect representation of Chryses’ plea utilizes a descriptive and semantically loaded paraphrase doubly removed from the original utterance.

The Iliad introduces this plea with an inquit phrase ‘he made prayer’ (lísseto) followed by direct speech. Plato’s (Socrates’) indirect discourse rendering, on the other hand, has êúxeto (‘he wished’) followed by an AcI (accusative with infinitive) construction that includes both Chryses’ good wishes for the Greeks’ speedy defeat of Troy, their safe return home, and his wish that they would exchange his daughter for the ransom he has brought. Note that the Greek employs one verb (‘to wish’) both for the polite formality and the actual plea. Plato’s rewriting of the passage then continues with Agamemnon’s threat to Chryses (also given in direct speech in Iliad I, 26–32). This has two AcI constructions, one for Agamemnon’s command to Chryses to leave immediately and the other for his refusal to release Chryses’ daughter. True to the direct speech in the Iliad, the indirect representation mentions the sceptre and fillet of Apollo, who protects Chryses —that is to say, it mentions the metonymical symbols of his authority—and it also renders Agamemnon’s chilling threat that Chryses’ daughter will grow old as Agamemnon’s concubine. However, in the indirect, the order of these two discourse elements is reversed. Moreover, except for the words geráô (‘to grow old’) and the sceptre and fillet of the god (skêptron, stémma), the vocabulary of the indirect rendering is changed as much as the syntax (from imperative to AcI and optative). Plato’s example therefore fully conforms to the results of speech act analyses of indirect discourse (e.g. Ryan 1981b), according to which the indirect need only represent the illocutionary force of an utterance but not necessarily its propositional content, much less its lexical or syntactic material. In indirect discourse the reporter must not distort the import of the original utterance, but s/he can easily substitute his or her own language and linguistic structure for the original, if there was one. Beyond the slippage from dramatic to narrative mimesis, Plato’s example therefore also explicitly touches on the issue of narrative mediation, of the narrator’s position as source and guarantor for the story-universe.

Book Three of the same dialogue [Republic] does not reduce all literature to such a degrading function of representation (mimesis). Here Socrates distinguishes the poet’s monologue, the characters’ dialogue, and the alteration between authorial and figural speech as three styles of presentation(diegesis): the commendably pure, the dangerously mimetic, and their rather acceptable combination. The apparent contradiction points, I believe, to the dual nature of poetic discourse. Literature as an art is representation (mimesis) even though it can assume the style of straightforward, non-imitative presentation; literature as verbal discourse is presentation (diegesis) even though it can assume the style of mimetic representation. Attempting to grasp ideas as timeless, universal substances, the thinker in Socrates and Plato had to prefer the abstractive power of verbal presentation to the impersonating representation of particular events. Yet Socrates liked to convey his philosophical vision through imaginative parables, and Plato embedded those parables into the mimetic medium of dialogue. Their poetic method of philosophizing implies that verbal art, by representing action, also presents the verbal approximation of vision.

(Hernadi 1971b:18)

In the standard formulation of diegesis versus mimesis, mimesis is aligned with the imitation not only of speech but also of action. Both Genette and Chatman, for instance, regard an objective description of action as equivalent to a ‘mimesis of events’ (Genette 1980:164; Chatman 1978:33). Diegesis, on the model of indirect speech, is then aligned with narrative mediation (and, implicitly, distortion) of the primary events (including speech acts), and it highlights the narrator’s subjective style and his or her (ideological, evaluative) point of view. Diegesis therefore potentially draws attention to itself qua narration. Although Genette and Chatman then go on to discuss speech and thought representation in greater detail, the dichotomy of diegesis vs mimesis tends to conflate not merely the grammatical issue of indirect vs direct discourse with the epic vs dramatic generic distinction a la Plato; on account of the privileged position which mimesis occupies, the diegesis/ mimesis opposition additionally suggests an equation of indirect speech with mediation tout court, and (since everything that is not mimetic is by definition adulterated) it ends up viewing narrative report of dialogue as intrinsically unreliable—a necessary evil.

There are two major confusions at work in this development. When Plato opposes narration (diegesis) to mimesis (imitation), he does so within a generic contrast of drama (all imitation) versus the dithyramb (all diegesis) and the epic (mixed), and the diegesis is here not simply ‘narrative’, but very explicitly the author’s direct voice. For Plato, of course, does not distinguish between a narrator and the author, which is the only way that the narrator of the epic would become identified with the speaker of elegiac verse. In narrative theory Plato’s distinctions have been forgotten, with the result that diegesis has come to signify narration, per se, that is to say the narrator’s rendering of the story in everything except clearly defined discourse not attributable to this enunciator, i.e. the characters’ directly quoted speech and thought acts. This schema makes it very difficult to conceptualize action (or, for that matter, description). Genette, in his discussion, concentrates on the medium of imitation as his primary concern, and he therefore excludes action from the realm of pure mimesis since only characters’ discourse can be mimetically represented (by a quotation in direct speech):

In contrast to dramatic representation, no narrative can ‘show’ or ‘imitate’ the story it tells. All it can do is tell it in a manner which is detailed, precise, ‘alive,’ and in that way give more or less the illusion of mimesis—which is the only narrative mimesis, for this single and sufficient reason: that narration, oral or written, is a fact of language, and language signifies without imitating.

Unless, of course, the object signified (narrated) be itself language. […] The truth is that mimesis in words can only be mimesis of words. Other than that, all we have and can have is degrees of diegesis. So we must distinguish here between narrative of events and ‘narrative of words.’

(Genette 1980:164 [1972:185])

Chatman (1978:33), on the other hand, insists on the concept of a mimetic action report. He therefore counts ‘objective’ or ‘un-narrated’ renderings of events as belonging to the mimetic (non-diegetic) realm.² The consequence of this move is likewise to emphasize narration’s quality of mediation, hence the term ‘non-narrated’ for direct speech quotations. Both Genette and Chatman underline that original direct discourse becomes ‘adulterated’ in narrated ...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Preface
• Acknowledgements
• Abbreviations and symbols
• Typographical conventions
• Introduction
• 1 Contexts: the reporting of speech and thought acts
• 2 Establishing the object of analysis: an introduction to the free indirect
• 3 Indirect and free indirect discourse: aspects of anaphoricity and shifting
• 4 Indirect and free indirect discourse: deictic features and expressivity
• 5 A scale of forms: speech and thought representation in context
• 6 Dual voice and stylistic deviation
• 7 Language and consciousness: Ann Banfield’s Unspeakable Sentences and beyond
• 8 Schematic language representation
• 9 Consequences and conclusions
• References