Contenu du sommaire : Unité(s) du texte, sous la direction de Dominique Legallois

This article presents the results of a study on the emergence of anaphora in children's speech. Concentrating on a dialogical perspective we show that the anaphoric value of children's first pronouns emerges earlier than is usually claimed in the literature. Previous research on language development comes to the conclusion that cohesion is precocious, but that intradiscursive relations are mastered quite late. Our quantitative results (drawn from the analysis of one hundred and five sequences of adult-child conversations) show that 90% of children's pronouns are used for second mentions of discourse objects. Pronouns appear in contexts of high discursive continuity and shared attention. These linguistic units are therefore specialized very early. Our conclusion is that children do not start with a deictic use of pronouns and use them anaphorically later on. Moreover, first anaphoric devices are based in dialogue. The unity of the text could therefore spring from the unity of the dialogical exchange.

Adding a modifier in the anaphoric NP can impede the good formation of the associative anaphora. If, in the famous example of the village and the church (Un village… L'église…), we substitute l'église romane for l'église, the anaphoric sequence becomes deviant. According to G. Kleiber, this kind of sequence is not well-formed because there is no stereotypical relation between the two referents. We can find however well-formed associative anaphors with a non stereotypical adjectival modifier in the anaphoric NP. In this case, two factors, which are linked to the value of existential uniqueness that the definite article presupposes, are decisive: 1° the restrictive or descriptive feature of the adjective; 2° if the adjective has a restrictive feature, the stereotypical amount of “parts”.

The various types of semantic and pragma-semantic relations obtaining between either two or three successive or separated text sentences within a discourse are not without effect on the interpretation and the discourse functioning of anaphors occurring in the second and third members of such groups of sentences. The article attempts to pinpoint the exact influence of these relations on the character of the anaphoric relation(s) which may also connect them. The interpretation (or “resolution”) of the anaphoric relation(s) at issue will flow from this integrative effort; and in turn, the resolution of the anaphors in the non-initial sentences in the group will make it possible to satisfy one of the conditions for application of the coherence relation selected itself.

This chapter explores lexical repetition and its text-organizing function in French written discourse (non-narrative texts). On the basis of Patterns of lexis in text (1991) by M. Hoey, I show how an analysis of lexical relations in texts can reveal the relatedness of adjacent or non-adjacent sentences, which produce elaborate patternings, and reticulate organisations. In this way, lexial units, sentences, nets, and hypotexts constitute the units of text.

In Cohesion in English (1976), Halliday & Hasan gave a small part to lexical collocation, though emphasizing it as a programmatic priority. Referring to the recent debates within and on the edge of textual linguistics (Adam, Rastier) and to our own praxis in the matter of computer-assisted and statistic approaches, we intend to show collocation (cooccurrence) as a central aspect of textuality. In this purpose it must be seized as a whole, which means statistically, and then invested back into the apparatus of textual analysis as a main way of expert browsing. Some instances and outlooks are given upon Balzac's Le Père Goriot.

The purpose of this paper is to examine the descriptive and epistemological relevance of the concept of textual homonymy. More particularly, the paper studies the central postulate of this concept: the invariance of ?textual signifier? (vs signified). To illustrate these problems, we chose to examine some aspects of the poetic texts.

Textual units, particularly semantic units such as themes and actors, do not correspond to the image of discrete units which the logico-grammatical tradition bequeathed to us. Textual units are manifested not in words but in passages related to genetic and hermeneutic practices: They are stabilized moments of these practices.
In conformity with the theory of semantic perception, units are forms drawn against backgrounds: they undergo diffusive movements (transformation of forms into backgrounds) and summoning movements (transformation of backgrounds into forms). This study intends to contribute to a general theory of these transformations.

Most of the cognitive psychology models of text processing aim to getting to the bottom of the cognitive invariants that refer to processes or to the representations over which they operate. These objects that provide observable elements and that are analyzed mainly by linguistics constitute an essential means for the study of linguistic activities implied in the processing of texts. Nevertheless, the research studies carried out in the field of text comprehension have not always proved to be effective. Many comprehension problems are not related to the students cognitive malfunctioning, but to ethnocentric models. That is why it is necessary to take into account the contexts, and to work on the basis of comparative data in the creation of models of cognitive processing of texts and their units.