¹ We are thinking in particular, in the French sector, of efforts undertaken by the reviews Littérature and Poétique.

² By the generic expression "analysis of discourse", I mean here the research and work issued from two horizons which are different but nevertheless convergent by their target and their results. In the first place the studies done by linguists on the corpus created by the various categories of discourse (political, legal, religious, etc.), studies which aim at defining the specificity of discourses by the analysis and the correlation of their different structural levels (syntactical, lexical, semantic, pragma tic). These studies are all inspired by the theory and the method defined in the 1950's by the American linguist Zellig S. Harris. In the French sector there can be found diverse illustrations in many issues of the review Langages (especially in issues No. 13, 23, 37, 41 and 52) and in several works such as those by M. Pêcheux (Les Vérités de la Palice, Paris, Maspero, 1975) and P. Henry (Le Mauvais Outil, Paris, Klincksieck, 1977). But with the expression "analysis of discourse" I also mean the masterful studies by M. Foucault of the discourses in the so-called "human" sciences (Les Mots et les choses, l'Archéologie du savoir, L'Ordre du discours), decisive studies in that to them is owed the conception, for the first time clearly defined and rigorously analyzed, of discourses as social practices (on this notion see note 3).

³ Like many others, I am borrowing this notion from M. Foucault who defines it as follows: "What is called ‘discursive practice' can now be defined. It cannot be confused with the expressive operation by which an individual formulates an idea, a desire, an image; nor with the rational activity which can be used in a system of inference; nor with the ‘competence' of a speaking subject when he forms grammat ical phrases; it is an ensemble of anonymous historical rules, always determined in time and space, which have defined for a given epoch, and for a given social, economic, geographic or linguistic realm, the conditions for the exercise of the enunciative function." (L'Archéologie du savoir, Paris, Gallimard, 1969, p. 153-154).

⁴ I hope I will be pardoned this expression which smacks somewhat of jargon, preferred here to the more traditional one of work, loaded with ideological implica tions, certain ones of which (and I will explain why later) seem to me to be a little suspicious. For the meaning given to this expression see below § IV.

⁵ A notion which I am borrowing once more from M. Foucault who defined it thus: the "[discursive] formation systems… are not constraints which had their origin in the thinking of men nor in the play of their representations; nor are they determinations which, formed at the level of institutions or social relations or of the economy, would be transcribed by force at the surface of discourses. These systems —and we have already stressed this—reside in the discourse itself […] By formation system should be understood a complex network of relations which function as a norm. […] To define a formation system in its particular individuality is thus to characterize a discourse or a group of statements by the regularity of a practice" (L'Archéologie du savoir, pp. 97-98).

⁶ We are here referring principally to the works of D. Slakta, particularly to L'Ordre du texte (Etudes de linguistique appliquée, No. 19, 1975), and to Sémiologie et Grammaire de texte, 1980, Parix X (to appear)—which have, it seems to us, given distinction to these two orders for which E. Benveniste was the first to establish with clarity (see Problèmes de linguistique générale II, chapter III) most rigorous theoreti cal status.

⁷ By structure of a discourse, we mean principally the following elements: specific positions for subjects, certain formal structures (narratives, arguments, etc.) proper to the statements, a semantic structure, an ensemble of references, a system of norms, a specific secondary semiotic functioning and intertextual functioning.

⁸ Cf. especially H.R. Jauss, Pour une esthétique de la réception, Paris, Clallimard, 1978 (French translation).

⁹ Cf. K. Stierle, "Identité du discours et transgression lyrique", Poétique 32, 1977, p. 426.

¹⁰ Cf. E.W. Bruss, "L'autobiographie considérée comme acte littéraire, Poétique 17, 1974.

¹¹ "Any literary notion referring to the inner unity of a work […] is null and void, given the theoretical presuppositions which we have recalled earlier". (M. Pêcheux and C. Fuchs, "Mises au point et perspectives à propos de l'analyse automatique du discours", Langages 37, March 1975, p. 28).

¹² Cf. art. cit.

¹³ Cf. T. Todorov, "La notion de littérature", Les Genres du Discours, Paris, Seuil, 1978.

¹⁴ We are thinking evidently of the Méditations, of the Harmonies poétiques et religieuses, of La Chute d'un Ange by Lamartine, of the Destinées of Vigny, and of many collections of Hugo, in particular Les Contemplations, Dieu, La Fin de Satan, etc.

¹⁵ Cf. R. Barthes, "Le discours de l'Histoire", Information sur le sciences so ciales, VI, 4, 1967.

¹⁶ Here we can once again recall the importance of religious inspiration in French poetry of the 16th century, from Marot's Psaumes (1541) to Sponde's Sonnets sur la Mort (1588).

¹⁷ Relations which are illustrated, each in its own way, by two works which are otherwise as diverse as Discours by Ronsard and Tragiques by Aubigné.

¹⁸ Ronsard, Hymne de l'Automne.

¹⁹ Ibid.

²⁰ L. Althusser, "Idéologie et appareils idéologiques d'État", La Pensée, No. 151, 1970.