Contenu du sommaire : L'ancrage énonciatif des récits de fiction, sous la direction de Gilles Philippe

Communication, fiction and the expression of subjectivity. The expression of subjectivity, though it is by no means confined to fiction, has mostly been studied in the fictional discourse. One of its expressions is the so-called represented speech and thought (RST), which has been widely discussed by Banfield. Her major thesis is that, in a given RST utterance, the subjectivity of no more than one subject of consciousness can be expressed. This has been contested, notably by Ducrot. In this paper, I compare RST with two other ways of reporting speech, namely direct discourse and indirect discourse. I focus on singular personal pronouns and their use in reporting subjectivity and I conclude that the pronouns used to designate the subject of consciousness in a RST utterance - whether they are or not third person pronouns - are very similar in both their function and their interpretation to the first person in speech. However, given that first person pronouns in the same sentence can be used to designate distinct individuals, Banfield's claim that there can be no more than one subject of consciousness in a given RST sentence does not seem to hold. What is more, contrary to her claims, there can be RST utterances where the subject of consciousness is represented by the first or second person pronouns, which means that RST is a means of communication. Finally, the expression of subjectivity, far from being restricted to fiction, is merely the manifestation of a general cognitive tendency in humanity: the intentional stance.

Enunciative inconsistencies in fictional narratives. Because of the specificity of their functioning, fictional narratives often contain sentences with enunciative inconsistencies, that is with marks of, as it seems, more than one enunciative node. When considered in isolation, such sentences may often be regarded as barely acceptable; but in context, the reader will usually overlook the inconsistency. It doesn't mean that s/he makes do with marks of two enunciative nodes in one utterance: as a matter of fact, either the reader cancels the enunciative inconsistency by ignoring one of the marks, or the weakest enunciative mark is given a new non-enunciative status in the sentence. All this proves to be very important in the current debate on the enunciative functioning of fictional narratives.

C'est as a point of view marker in fictional narratives C'est has a double representational value in narratives: firstly because it signals the introduction a new object, secondly because it implies a perceiving or thinking "subject of consciousness" that validates the first representational value. This is why c'est is enough to introduce by itself a light point of view effect in the narrative and may become a full point of view marker, when coming along with a perfective to imperfective tense shift. All this is obviously of highest importance in the reading of narratives, since such linguistic devices strengthen the readers's referential illusion and identification with the character or narrator, hence legitimize possible interpretative inferences. effect in the narrative and may become a full point of view marker, when coming along with a perfective to imperfective tense shift. All this is obviously of highest importance in the reading of narratives, since such linguistic devices strengthen the readers's referential illusion and identification with the character or narrator, hence legitimize possible interpretative inferences.

Borderline beings and narrative "angelism". This article, based on some concepts of discourse analysis and of the linguistics of enunciation, deals with fuzzy relationships between narrator and character in novelistic narration. The author studies especially two notions that he has already introduced in previous works: "generic character" and "witness- narrator", in order to give them a more precise definition. A distinction is made between "unidentifiable character" and "any member of a group", as enunciative positions in the narrative. The next step is the introduction of a new notion, "narrative angelism", which makes it possible to analyse some short passages beyond the scope of classical narration. Two examples (from Zola and Mauriac) are commented. It is argued that such phenomena, which are very few, are narrowly bound with the resources of a given language (for example, French writers can use the subject pronoun "on") and with the aesthetic universe of the works in which they appear.

Variations of the enunciative utterer-centered anchor points and fictionalization of an Albert Camus autobiographical narrative In Camus's works, one can find three versions of the same autobiographical anecdote. The rewriting of an identical scene as fiction in The Stranger, clearly as fact in Reflections on the guillotine, and as mixed discourse in Camus's posthumous autofictional novel The First Man may be seen as a textbook case. This article examines the enunciative anchoring in each of these forms of discourse in order to study whether narrative fiction offers specific enunciative features that are linguistically identifiable.

Fictional enunciation and referencial constructions. Basing our analysis on a text of fiction, Le Bavard by Louis-René Des Forêts, we show that, in fiction, signs of subjectivity like the deictic pronouns or adverbs have a specific enunciative functioning since the referential adjustment system that is compelled in the dialogue, cannot work anymore, the construction of ref erencial values being fictitious. Relying on the works of E. Benveniste and A. Culioli, we suggest a linguistic analysis of the utterances of these personal and temporal forms in that text to reveal that the formal apparatus of enunciation may make use, in fiction, of the fundamental instability of the referential system.