A enunciação em Michel Pêcheux: uma questão inquietante

RESUMO O presente estudo propõe-se a responder às seguintes questões: 1ª) Que noções de enunciação podem ser derivadas da reflexão discursiva de Michel Pêcheux nos anos iniciais de sua produção? 2ª) Qual é o estatuto que assume a enunciação no escopo da AD nesse período? Para tanto, procede a um exame de textos do filósofo francês publicados entre 1969 e 1975, a fim de investigar a problematização da enunciação nesses escritos. A análise do corpus teórico permite dele derivar diversas noções de enunciação, as quais revelam em seu conjunto um estatuto epistemológico e um estatuto teórico-metodológico da enunciação no estabelecimento das bases fundantes da AD.

All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0

(Never) Initial Words
Although some language theories that aim to go beyond the limits of language as a system of signs began to appear in the first half of the 20th century, it was in the second half of that century, especially in France, that such theories flourished. Among the questions that moved such theories in their origins, the problems of discourse and enunciation stand out, which, if not avoided by all linguists aligned with linguistic structuralism, neither were they at the center of the concerns of this epistemological paradigm, dominant in the French intellectual conjuncture of the 1960s.
More specifically, Maingueneau (2016) and Guilhaumou and Maldidier (1986) deal with the relationship between enunciation and Michel Pêcheux's Discourse Analysis (DA). These authors, although in dialogue with the research herein, for they deal with such a relationship, do not closely explore our subject of interest: the place of enunciation in the early years of DA. Thus, the present work aims to answer the following questions: 1) What notions of enunciation can be derived from Michel Pêcheux's discursive reflection in the early years of his production? 2) What is the status assumed by enunciation in the scope of DA in this period?
To this end, we proceeded to an examination of texts published by the philosopher between 1969 and 1975 in order to investigate the problem of enunciation in these writings. The selected studies correspond to two of the three periods of DA, as outlined by Pêcheux (1983). Although the author refers to his works and to those of others, we will restrict ourselves to those signed by him, alone and in co-authorship. The selected texts are grouped into three analytical categories: one related to another, and yet another, to texts that are placed in the transition from the first to the second "epoch": Chart 1 -Theoretical corpus  Da AD-1 à AD-2 AD-2 Automatic Discourse Analysis (1969/1995) Langue, "langage", discours (1971) La Semantique et la coupure saussurienne: langue, langage, discours (1971) Overview and Prospects (1975/1995) Language, Semantics and Ideology: Stating the Obvious (1975/1982) Source: the author.
All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 To answer the aforementioned questions, we have organized this study into four sections, in addition to its introduction: first, we examine how enunciation appears in DA-1; then, we evaluate the problem of enunciation in the transition from DA-1 to DA-2; next, we analyze the approach to enunciation in DA-2; finally, we present the final remarks based on the answers to the guiding questions.

Enunciation in DA-1
In a text on the theoretical references of Automatic Discourse Analysis (hereinafter, ADA-69 or DA-1), Gadet et al. (1995) 1 state that Jakobson provides Pêcheux with both theoretical and analytical elements. Indeed, the philosopher contests what he considers one of the implications of the Saussurean langue /parole opposition in the linguist's reflection: the belief in the passage from the coercion of the linguistic system to the freedom of the use of language. According to Pêcheux (1995, p.73; italics by the author), 2 "this opposition authorizes the triumphal reappearance of the speaking subject as subjectivity in action," as if the parole were a road to human freedom and the subject, a free and intentional individual.
Against this supposed unrestricted freedom attributed to the parole and the speaking subject, Pêcheux elaborates the notion of process of production of discourse, defining it as "all those formal mechanisms which produce a discourse of a given type in given 'circumstances'" (PÊCHEUX, 1995, p.77; quotation marks by the author). 3 This leads him to a conception of discourse that is distinct from the Jakobsian conception of message as a transmission of information: "[...] the term discourse [...] implies that we are not necessarily dealing with information transmitted from A to B but, more generally, with a 'meaning-effect' that occurs between points A and B" (PÊCHEUX, 1995, pp.84-85; italics and quotation marks by the author).  GA: Rodopi, 1995a, pp.63-121. 3 For reference, see footnote 2. 4 For reference, see footnote 2.
All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 As for the protagonists of the discourse, Pêcheux clarifies that points A and B designate not individual human organisms, but "determinate positions within the structure of a social formation" (PÊCHEUX, 1995a, p.85). 5 In other words, these are not empirical subjects, but social positionssuch as the positions of "boss" and "worker"positions represented in discursive processes, but in these not reflected as such, because what works in them "is a series of imaginary formations designating the positions which A and B ascribe to themselves and to one another; the image they have of their own position and of the position of the other" (PÊCHEUX, 1995a, p.85; italics by the author). 6 The referent is also an "imaginary object (a subject's viewpoint), and not a physical reality" (PÊCHEUX, 1995a, p.85;italics by author). 7 In this reflection, we see the influence of two notions dear to the theories of enunciation, the notions of intersubjectivity and reference. As in the enunciative perspectives, such notions appear in DA-1 as linked not to an existing ontological reality a priori, but to a reality projected in the discourse, based on the interlocutors' images about themselves, the other and what they speak. Therefore, DA-1, although critic of the enunciative theories for attributing onto them an idealist conception of the subject, by refusing a vision of language as a communication tool and of empirical subjects endowed with intention, who would talk about a physical reality external to language, approximates rather than distances itself from enunciativists, who are also distant from pragmatic approaches.
If Jakobson is a constant presence in ADA-69, the same cannot be said of another enunciation theorist, namely, Benveniste. Of the three references to him in the book, we are interested in two, due to the different positions they present.
In the first, Pêcheux confronts the positions assumed by Jakobson and Benveniste regarding the sentence/discourse status in relation to the lower levels of language. For Jakobson, "from phoneme to discourse, we are in the presence of linguistic signs whose dimension increases but which are still bound to the same rule of combination" (PÊCHEUX, 1995a, p.102; italics by the author). 8 Benveniste, on the other hand, recognizes "that the sentence has a very specific status, and that it is the boundary All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 separating linguistics from discourse theory" (PÊCHEUX, 1995a, p.102; italics by the author). 9 Here, Pêcheux moves away from Jakobson and leans towards Benveniste, who places the sentence as a unit of discourse, distinct from the phoneme, the morpheme and the lexeme, units of language. This distinction would be due to the fact that the signs within an utterance would respond to different rules of combination than the utterances within a discourse. However, the philosopher does not recover Benveniste's reflection on the sentence in all its complexity: "Pêcheux relies upon Benveniste when [...] he refers to the sentence as the unity of discourse and as the frontier of a domain which is irreducible to the order of grammar. But he draws no theoretical conclusions" (GADET et al., 1995, p.54). 10 In the second reference to Benveniste, Pêcheux quotes him to criticize his conception of the sentence as infinitely creative, unlimited variety. The philosopher's criticism of the linguist is based on the argument that this conception relates the discourse to the alleged freedom of speech. For Gadet et al. (1995), 11 Pêcheux's ideological doubts towards Benveniste is explained because he seems to have seen in him "a step backwards, as having effected a return of a psychological subject which had been victoriously driven from the theoretical scene by Saussure and structuralism" (GADET et al., 1995, p.54). 12 Along the same lines of reasoning, Maldidier adds that Harris' distributionalism is the greatest inspiration of ADA-69, which seems "to have as the opposite result the repression of enunciation and a 'poor' place given to Benveniste" (MALDIDIER, 1990(MALDIDIER, /2003 quotation marks by the author). 13 In summary, in DA-1, "none of the three references to Benveniste shows any real understanding of the gap opened up in structuralism by his recognition of the role of enunciation" (GADET et al., 1995, p.54). 14 Generally speaking, in this first "epoch" of the theory of discourse, the problem of enunciation arises timidly. It emerges in ADA-69 still as a nebulous notion, which meanders through other notions with more precise contours, such as those of process of production of discourse, conditions of production, discourse protagonists, referent, imaginary formations¸ langue, parole, discourse, sentence, utterance, subject.
2 Enunciation: From DA-1 to DA-2 The two texts covered in this section are not indicated by Pêcheux (1995b). 15 If we place them here, we do so for two reasons: a) both present a reflection on enunciation that cannot be ignored in a study that investigates this problem in the early years of DA; b) both are in the gap between the ADA-69, the milestone of the first "stage," and the titles Overview and prospects and Language, Semantics and Ideology, milestones of the second "stage."

Enunciation in Langue, 'Langage', Discours
According to Maldidier (1990), the article Langue, 'langage', discours was published in October 1971, in the communist newspaper L'Humanité. Although the author does not provide an in-depth discussion of this piece, we would like to address it more closely, as it seems to deal with enunciation with more emphasis than the study of Pêcheux discussed in the following subsection.
In Langue, 'langage', discours, the philosopher criticizes what he calls a metaphorical application of linguistics on other domains, through studies on drama and narrative "syntaxes" in the literary domain; studies of "lexicons" that reduce the class struggle to the idea of a conflict between dialects and jargons in the socio-historical domain; studies that go beyond the textual level, extending to the set of objects and behaviors that can be understood as structures (music, painting, cinema, fashion). Such initiatives are denounced by the author because they aim at an alleged science of sciences, as "an enterprise 'of general analysis of the human intelligible'" (PÊCHEUX, /2015 Two mistakes could result from this criticism, which Pêcheux hastens to dispel.
The first mistake is related to the fact that such criticism does not mean that these other types of "languages" cannot be objects of study, but that the investigation on such different domains requires a change of terrain and a displacement of the subjectivist problem centered on the individual into what historical materialism calls social relations.
The second mistake consists in the deduction that language as an autonomous reality disappears and that grammar is nothing more than an object of class struggle, being, according to Pêcheux ( /2015, "more convenient to conceive language (object of linguistics) as the basis on which processes are built; the linguistic basis characterizes, in this perspective, the functioning of the language in relation to itself," 17 while "the term discursive process (process of production of discourse) [is reserved] to the functioning of the linguistic bais in relation to representations [...] put at stake in social relations" (PÊCHEUX, /1995; italics by the author). 18 That articulation of processes on the linguistic basis is possible, within this basis, by the existence of "mechanisms summarized by the term enunciation, through which the position of the 'speaking subject' is taken in relation to the representations of which he is the support." Some linguists and researchers specialized in the study of texts begin to work on this point [enunciation]*, which is decisive for the future of the relations between linguistics (or theory of language) and what has been designated here by the provisional name "theory of discourse," and which rightly constitutes a sector of historical materialism probably destined for a great development (PÊCHEUX, /2015 italics and quotation marks by the author; italics with quotation marks by us). 19 Some observations are necessary regarding these final pages of the text. 17 In Portuguese: "mais conveniente conceber a língua (objeto da linguística) como a base sobre a qual processos se constroem; a base linguística caracteriza, nesta perspectiva, o funcionamento da língua em relação a si própria." 18 In Portuguese: "a expressão processo discursivo (processo de produção do discurso) [é reservada] ao funcionamento da base linguística em relação a representações [...] postas em jogo nas relações sociais." 19 In Portuguese: "articulação dos processos sobre a base linguística torna-se possível pela existência, no próprio interior desta base, de mecanismos resumidos pelo termo enunciação, pela qual se efetua a tomada de posição do "sujeito falante" em relação às representações das quais ele é o suporte. Alguns linguistas e pesquisadores especializados no estudo de textos começam a trabalhar sobre este ponto [a enunciação]*, decisivo para o futuro das relações entre a linguística (ou teoria da língua) e o que foi designado aqui pelo nome provisório de "teoria do discurso", e que constitui de direito um setor do materialismo histórico destinado, provavelmente, a um grande desenvolvimento." All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 First, the criticism of the subjectivist problematic is reinforced here. According to Pêcheux ( /2015, "the process of discourse must evidently not be confused with the act of speech of the individual speaking subject," 20 as it does not focus on one "speaking subject." Thus, enunciation should not be regarded in the individual sphere of speech (parole), but in the social sphere of discourse as a process that is determined both by the linguistic basis on which it takes place and by the class relations of a given social formation.
Second observation: enunciation is inserted here in a privileged epistemological and theoretical-methodological place. It is epistemological when assigned as a decisive point for the future of the relations between the theory of language and the theory of discourse. And it is theoretical-methodological, when indicated as "a general condition for the possibility of [discursive] processes" 21 whose study "will probably make it possible to adequately place and perhaps solve the problem of 'relations between syntax and semantics'" (PÊCHEUX, /2015; quotation marks by the author). 22 In this regard, the study of enunciative mechanisms, which enables the transition from the linguistic into the discursive, penetrates a border area, common to language and to syntax, on the one hand, and to discourse and to semantics, on the other.
Third observation: this reflection on the relation linguistic basis-discursive process and the place of enunciation in this relation seems to be linked to two research orders postulated by Pêcheux in ADA-69: a) the relationship between the variable processes of production of discourse and the invariant background of language; b) the relationship between the process of production of discourse and its conditions of production. As we can see, these two texts, two years apart, present similar definitions of process of production of discourse and enunciation. The first, from 1969, is related to formal mechanisms that produce a given discourse under given conditions of production.
The second, from 1971, is associated to internal mechanisms within the linguistic basis that make possible the realization of the discursive processes on this basis. Hence, the idea of mechanisms, which defines the process of production of discourse in 1969, seems to shift to define enunciation in 1971, while the discursive process is now defined as Therefore, enunciation, in 1971, seems to respond to the two orders of research envisaged in 1969. To the first order, it responds when taken as a condition for the possibility of articulating discursive processes on the linguistic basis. To the second, it responds when also taken as a condition for the possibility of the taking of position of the "speaking subject" regarding the social representations of which he is the support, which refers to the relationship between the discursive process and its conditions of production.

Enunciation in La Sémantique et la Coupure Saussurienne: Langue, Langage, Discours
Published two months after the text in L'Humanité, the article La Semantique et la coupure saussurienne: langue, langage, discours marks Pêcheux's "thundering entry into the field of linguistics" (MALDIDIER, 1990(MALDIDIER, /2003; quotation marks by the author). 23 In this "text of 'epistemological intervention,'" 24 published in the journal Langages and written along with the linguists Claudine Haroche and Paul Henry, Pêcheux "centrally intervened for the first time in the field of linguistics towards Saussure and against semantics [as then practiced]" (MALDIDIER, 1990(MALDIDIER, /2003. 25 It contains the essentials of the article published weeks before in the communist periodical, such as the reiteration of the necessary change of terrain or perspective required by semantics.
This change results from the adoption of a position strongly influenced by the concepts of historical materialism. The authors remind readers that a given social formation, at a given historical moment, is characterized by the mode of production that dominates it and by the relationships between classes that compose it. Such relationships correspond to political and ideological positions, which are not "created by individuals, but are organized into formations displaying relations of antagonism, alliance or 23 In Portuguese: "entrada estrondosa" de Pêcheux "no campo da linguística." 24 In Portuguese: "texto de 'intervenção epistemológica.'" 25 In Portuguese: "intervinha pela primeira vez de forma central no campo da lingüística em torno de Saussure e contra a semântica [como então praticada]." Here, Haroche, Pêcheux and Henry (1971)  For the authors, words change their meaning when they change DFs, these changes being determined not only by the nature of the words used, but mainly by the constructions in which they combine themselves, constituting discursive sequences whose terms are organized according to the conditions in which such sequences are produced. Thus, the study of these changes requires not a lexical or grammatical semantics, but a discursive semantics, regarded as "the scientific analysis of the characteristic processes of a discursive formation, [...] which takes into account the link that unites these processes to the conditions in which discourse is produced" 26 In Portuguese: "constituem indivíduos, mas que se organizam em formações que mantêm entre si uma relação de antagonismo, de aliança ou de dominação." TN. The translation used here can be found in the article Overview and Prospects (PÊCHEUX, 1975(PÊCHEUX, /1995. For full reference: PÊCHEUX, M. Overview and Prospects. In: HAK, T.; HELSLOOT, N. (ed.) Michel Pêcheux: Automatic Discourse Analysis. Translated by David Macey. Amsterdam; Atlanta, GA: Rodopi, 1995c, pp.123-187. 27 In Portuguese: "um conjunto complexo de atitudes e representações que não são nem 'individuais' e nem 'universais', mas que se relacionam mais ou menos diretamente a posições de classes em conflito umas em relação às outras." TN. The translation used here can be found in the article Overview and Prospects (PÊCHEUX, 1975(PÊCHEUX, /1995c. For full reference, see TN on footnote 26. 28 In Portuguese: "o que pode e deve ser dito (articulado sob a forma de uma arenga, de um sermão, de um panfleto, de uma exposição, de um programa, etc.) a partir de uma posição dada numa conjuntura dada." TN. The translation used here can be found in the article Overview and Prospects (PÊCHEUX, 1975(PÊCHEUX, /1995c. For full reference, see TN on footnote 26. Let us carefully read this quotation, which is undoubtedly very similar to the one present in the text in L'Humanité (cf. 2.1), but also distinct from it in some aspects.
The first aspect: if in L'Humanité enunciation is assimilated into the mechanisms that effect the articulation of the discursive processes on the linguistic basis, here, the focus seems to shift from enunciation to the utterance/enunciation relation. This shift in focus presents a change and an addition compared to the previous quote, which leads us to two other aspects that distinguish those quotes.
The second aspect: the modification refers to the social representations of which the "speaking subject" is the support. In the first quote, the "speaking subject" takes a position in relation to such representations through enunciation. In the second, he takes a position regarding these representations not through enunciation, but through the utterance/enunciation relationship. Our hypothesis is that the term utterance is, here, in synonymic relation with the term discursive sequence as a construction of words organized according to the conditions of production of this construction. In this perspective, the utterance would be the necessary materiality to the manifestation of the taking of position of the "speaking subject" in relation to social representations, a 29 In Portuguese: "a análise científica dos processos característicos de uma formação discursiva, [...] que leva em consideração o elo que liga esses processos às condições nas quais o discurso é produzido." 30 In Portuguese: "destacar a importância dos estudos linguísticos da "relação enunciado/enunciação", pela qual 'o sujeito falante' toma posição em relação às representações de que ele é o suporte, "desde que essas representações se encontrem realizadas por um 'pré-construído' linguisticamente analisável". É sem dúvida por essa questão, ligada à da sintagmatização das substituições características de uma formação discursiva, que a contribuição da teoria do discurso ao estudo das formações ideológicas (e à teoria das ideologias) pode atualmente se desenvolver mais proveitosamente." materiality articulated by enunciation as a series of mechanisms that manage such articulation.
The third aspect: the addition concerns a conditional clause absent in the previous quote and inserted in this one. Linguistic studies on the utterance/enunciation relationship, through which the "speaking subject" is placed facing social representations, are important as long as these representations are performed by a linguistically analyzable "preconstructed." This clause places the presence of a preconstructed subject to linguistic analysis as a condition for the study of the utterance/enunciation relationship and the representations linked to it. Such a preconstruction, not mentioned in the communist newspaper, is mentioned in the scientific journal, but not explained. Despite this non-explanation, the insertion of that term, along with that of the term utterance in this article aimed at linguists, seems to want to make clear to the interlocutors the importance of language in the theory of discourse, so that, "on the side of linguistics, discourse has now a firmer seat" (MALDIDIER, 1990(MALDIDIER, /2003; italics by the author). 31 From these 1971 writings, we may conclude that enunciation is no longer a matter of secondary importance in Pecheutian thought, as it is in ADA-69. On the contrary, enunciation comes here as a possible bridge between syntax and semantics, language and discourse, the individual and the social, linguistics and DA, theorization and analytical practice. Moving in this borderline among different epistemological-theoreticalmethodological instances, enunciation continues to gain space and increasingly claims the right to exist within the conceptual apparatus of DA, to which new notions, such as those of IF, DF, preconstructed and discursive sequence are being added.

Enunciation in DA-2
The gap between the two texts from 1971, analyzed in the previous section, and the two texts from 1975, analyzed in this one, covers the construction of more solid theoretical foundations for DA. This is because, although the 1971 Langages article became the manifesto of researchers engaged in discourse, it proposed a general determinism and a simple scheme of the ideology/discourse relationship: "There was still All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 a lack of links to the theory of discourse, both in terms of language and of ideology" (MALDIDIER, 1990(MALDIDIER, /2003 On the side of ideology, Pêcheux is affiliated to Louis Althusser. At the turn of the 1970s, the Althusserian reflection that impacts Pêcheux the most is present in the renowned article Ideology and Ideological State Apparatuses (Notes towards investigation). 33 In this article, Althusser argues that "all ideology hails or interpellates concrete individuals as concrete subjects" (ALTHUSSER, 2014, p.264; italics by the author), 34 being the perception of subjects as such an ideological effect, as well as the evidence of transparency of the language, which produces the impression that a word designates the same and only thing. It is through this parallel between the subjective evidence and the referential evidence that Pêcheux will attempt, in Language, Semantics and Ideology, 35 to sediment a materialist theory of discourse, a theory whose foundations are already found in Overview and Prospects. 36 On the language side, in addition to Saussuresince the beginning, Pêcheux's greatest linguistic reference -, Oswald Ducrot is also a name with which the philosopher dialogues in DA-2. Along with Paul Henry, Pêcheux reformulated the Ducrotian notion of presupposition in the field of discourse, moving it away from the logical-pragmatic assumptions underlying it and renaming it preconstructed. According to Maldidier (1990Maldidier ( /2003; italics and quotation marks by the author), "the preconstructed provides the linguistic anchorage of the taking of the interdiscourse," 37 the key concept of the theory, not yet developed, but embryonically present since ADA-69, as the "relation of the discourse to the 'already heard,' to the 'already there. '" 38 The discussion about the language/discourse relationship becomes more complex with the development of the Pecheutian reflection on enunciation in the two publications of 1975, to which we now move on.

Enunciation in Overview and Prospects
The article Overview and Prospects was published in the issue 37 of Langages, in March 1975, two months before the publication of the book Language, Semantics and Ideology. However, Maldidier (1990) warns that the article was produced much earlier than the book, so that, between the two texts, his thinking had matured. According to her, in the article, the philosopher addresses linguists and enhances the technical device, while in the book he writes for a collection organized by Althusser and focuses on theoretical aspects. Despite this observation, the author recognizes in the article an advancement in the theorizing about discourse. Surely, in it, Pêcheux proceeds to a "reformulation of the whole" of ADA-69, justifying the need for such a return based on the developments in the relationship between linguistics and DA, which made the presence of a linguist in this endeavor indispensable. The chosen linguist was Catherine Fuchs.
Pêcheux 39 begins the ADA-69 mise au point outlining the general epistemological framework that guides this update of the inaugural text: That framework is provided by the articulation of three regions of scientific knowledge: 1. historical materialism, as a theory of social formations and of their transformations, and as a theory of ideologies; 2. linguistics as a theory of both syntactic mechanisms and processes of enunciation; 3. discourse theory as a theory of the historical determination of semantic processes. It should be added that these three regions are in a sense traversed and articulated by a reference, which has to be made explicit, to a theory of subjectivity, and that theory is of psychoanalytic nature (PÊCHEUX, 1995c, p.123; italics by us). 40 As can be seen, if the epistemological status of enunciation in the disciplinary genesis of DA was only deductible from the prospective outcomes of the 1971 texts, now this status is verbally claimed. 39 TN. Although this article from number 37 of Langages was co-authored with Catherine Fuchs, we will refer to Pêcheux, as we will privilege the first part of the texttheoretical -, which Maldidier (1990) attributes to the philosopher, from what can be inferred that the second partmore technicalwas in charge of the linguist. 40 For reference, see footnote 26. In this article, Pêcheux rejects an idealist conception of ideology as a sphere of ideas and discourses, in favor of a vision that conceives it as a constitutive instance of discourse, of subject and meaning. Following Althusser, he postulates that this constitution occurs through the interpellation of the individual into a subject by ideology, an interpellation defined in terms of subjecting the subject as an ideological subject. Such process leads subjectswithout their awarenessto occupy places in the class struggle and to the continuous reproduction of relations of production, attached to practices inscribed in ideological apparatuses of the State. By affirming that "the discursive must be regarded as one of the material elements in what we have termed ideological materiality" (PÊCHEUX, 1995c, p.127), 41 Pêcheux takes up the notions of IF and DF, sustaining that the interpellation of individuals into subjects always happens through DFs, which derive from specific conditions of production and reflect the exteriority that constitutes them.
Then, Pêcheux begins to characterize what he calls forgetting 1, that is, the subject's illusion of being at the source of meaning. This subjective illusion would be linked to the production of meaning in a DF since the process of producing a discursive sequence would be ignored by the subject, who would repress the fact that any sequence, to be recognized as significant, must necessarily belong to this or to that DF, and not to himself. According to the philosopher, this process is "indissociable from the paraphrastic relationship that exists between sequences, in such a way that a paraphrastic family of sequences constitutes what might be termed a matrix of meaning" (PÊCHEUX, 1995c, p.129). 42 It is from these paraphrastic relationships that a sequence is produced. Thus, the discursive processdefined in 1969 as a set of formal mechanisms that produce a given discourse under given conditions of production and, in 1971, as the language put into operation through enunciation and in relation to social representationsstarts to be defined, in 1975, as "the paraphrastic relations internal to what we have termed the matrix of meaning inherent in a discursive formation" (PÊCHEUX, 1995c, p.131). 43 In this direction, Pêcheux sheds light on an apparent contradiction which refers to the subject's own constitution: the discursive processes do not originate in him, but necessarily take place in him. About this, the author clarifies an ambiguity underlying the All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 notion of conditions of production in ADA-69, which "designated both the effect of the positional relations in which the subject is inscribed, and its situation, in the concrete and empirical sense of that term (its material and institutional environment, the roles it adopts, consciously or otherwise, and so on)" (PÊCHEUX, 1995c, p.132; quotation marks by the authors). 44 Pêcheux dissolves this ambiguity by opposing the first definition (pertaining to the discursive) to the second (regarding the empirical) and assimilating them to the Lacanian opposition between imaginary and real. From this, he concludes about the need for "a non-subjective theory of the constitution of the subject in its concrete situation as enunciator [...] by providing an outline of the processes of enunciation" (PÊCHEUX, 1995c, p.132) 45 and for a possible articulation between this description and the subjective illusion that defines forgetting 1. In this regard, Gadet et al. (1990Gadet et al. ( /2014; quotation marks by the authors) 46 assert that "the opacity of the ADA on this capital point [the notion of subject] must be related to the 'secondary place' that concomitantly is given to the theory of enunciation such as it is developed by Benveniste." Indeed, questions related to subject and meaning oblige Pêcheux to return to linguistics, in which he rediscovers "the question of enunciation. Elided in ADA 69, evoked as a quick promise at the end of the article in the issue 24 of Langages, it is finally confronted in the 'update' of number 37" (MALDIDIER, 1990(MALDIDIER, /2003; quotes by the author). 47 This confrontation has as its starting point the need felt by Pêcheux to define the nature and the role of language in the phase of linguistic analysis provided by the "ADA device." Just as he rejects an idealist conception of ideology, the author also rejects an idealist conception of language, which identifies it with ideology, taking it as a "world view, a perception of the world, or even as the origin of the world" (PÊCHEUX, 1995c, p.134). 48 In contrast to this idealistic conception, Pêcheux proposes a materialist conception: "[...] since discursive processes are at the origin of the production of 44 For reference, see footnote 26. 45 For reference, see footnote 26. 46 TN. In the version in English of this work, organized by Tony Hak and Niel Helsloot (1995; reference on footnote 1), the article by Gadet et al. (1990Gadet et al. ( /2014 is not present as it is organized differently from the version in Brazilian Portuguese. For this reason, we present here a translation produced by us, since the passage is not found in the edition organized by Hak and Helsloot. In Portuguese, it reads: "a opacidade da AAD sobre esse ponto capital [a noção de sujeito] deve ser relacionada ao 'lugar secundário' que, concomitantemente, é dado à teoria da enunciação tal como é desenvolvida por Benveniste." 47 In Portuguese: "a questão da enunciação. Elidida na AAD 69, evocada como uma promessa rápida no fim do artigo aparecido no número 24 de Langages, ela enfim é enfrentada na 'atualização' do número 37." 48 For reference, see footnote 26. All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 meaning-effects, the language constitutes the material site in which meaning-effects are realized" (PÊCHEUX, 1995c, p.133; italics by the author). 49 This linguistic materiality, according to the author, refers to the Saussurean idea of functioning, as opposed to that of function; however, he endorses the need for a theory on linguistic systematicity that is not opposed, in the langue/parole fashion, to the non-systematic, but that is articulated in processes.
Such reflection leads to the vision of enunciation as a necessary relation between the enunciating subject and his utterance, a view refuted by Pêcheux, who sees in it two obstacles resulting from its idealist bias. It is, simultaneously, what he designates the empiricist illusion and the formalist illusion: the first would be linked to the view of the subject as the center and source of meaning; the latter, taking enunciation as a simple system of formal operations. For the author, it is necessary to remove the problem of enunciation from the circle of idealism in which enunciativists would be trapped: "Current difficulties in theories of enunciation stem from the fact that those theories usually reflect the necessary illusion that constitutes the subject" (PÊCHEUX, 1995c, p.137). 50 According to the philosopher, this occurs to the extent that such theories reproduce this subjective illusion at the theoretical level, "by following the tradition of Bally, Jakobson and Benveniste and by adopting the idea that the enunciating subject is endowed with choices, intentions and decision" (PÊCHEUX, 1995c, p.137.) 51 By reaffirming the criticism on the so-called "idealist" conception of enunciation, Pêcheux presents his own: We can state that processes of enunciation consists of a series of successive determinations which gradually constitute an utterance, and which are characterized by their ability to posit what is said and to reject what is not said. The enunciation therefore establishes boundaries between what is 'selected' and gradually specified (the elements which constitute 'the universe of discourse') and what is rejected. We thus obtain an inverted image of 'everything which it would have been possible for the subject to say (but which he did not say)' or of 'everything which conflicts with what the subject says'. The distance between the 'rejected' zone and the conscious mind varies, and questions an interlocutor may ask in order to get the subject to clarify 'what he means' may force him to reformulate its boundaries and to investigate it anew. We propose to term this effect of partial occultation 'forgetting 2', and we will regard it as the origin of the subject's impression that he really expresses his thoughts ('I know what I am  saying', 'I know what I am talking about') (PÊCHEUX, 1995c, p.138; quotes by the author; italics by us). 52 In this long as well as rich quotation, a more robust definition of enunciation appears for the first time, in the scope of the theory of discourse, of which we underline two points.
The first refers to the excerpts in italics, which deal with the utterance/enunciation relation, here retaken to be complexified. If in the 1971 article in Langages, utterance is presented as a materiality articulated by enunciation, this articulation is best explained in this 1975 text. Pêcheux proceeds to this explanation by integrating, into the discussion on the enunciative processes, the one on the syntactic mechanisms. This leads him to make changes in his conception of language, starting with the lexicon, which is no longer conceived as "a 'stock of lexical units', or as a mere list of morphemes which are unrelated to syntax; on the contrary, it must be regarded as a structured set of elements articulated with syntax" (PÊCHEUX, 1995c, p.138; quotation marks by the author). 53 Syntax, on the other hand, "no longer constitutes a neutral domain of purely formal rules; it constitutes the mode of organization (specific to a given language) of identifiable traces of the enunciation" (PÊCHEUX, 1995c, p.138). 54 Such overlap between enunciation and syntax is present in the very definition of enunciative processes as a series of successive determinations that delimit the boundaries between what is said and what is not said, so that "the linguistic analysis on which ADA relies must be essentially morpho-syntactic" (PÊCHEUX, 1995c, p.134). 55 In this regard, Leandro-Ferreira (2000) conceives syntax as a "theoretical device that makes visible the process of production of meaning and that places itself as a site for the observation of discourse" (LEANDRO-FERREIRA, 2000, pp.121-122; italics by the author). 56 According to this author, syntax gives access to the transition from the organization of language to the order of discourse, being a space of tension between coercion and freedom, that is, between the systematizable and what escapes from it. In All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 this perspective, language is not the object of the discourse analyst, but a presupposed instance in his analytical practice: "Just as the linguist has access to language through the parole, the discourse analyst has access to the order of discourse through the organization of the langue" (LEANDRO-FERREIRA, 2000, p.37). 57 If language is a presupposed instance in discourse analysis, we think it is not illicit to understand enunciation as a mediating instance of the relationship between syntax and discourse. After all, if enunciation is a series of successive determinations, with a morphosyntactic character, that little by little constitute the utterance, then enunciation operates the mediation between the linguistic organization and the discursive order by making the first the material place of realization of the meaning-effects of the latter. However, this should not lead one to believe that enunciation is reduced to syntax or that it is a space of unrestricted freedom of the subject. Enunciation is not reducible to syntax, since it is not only what enables the syntagmatization of the lexical units, but also what makes possible for the subject to take his position regarding social representations so that the enunciative articulates the linguistic and the discursive both in terms of language organization (that of syntagmatizations) and in terms of the order of discourse (that of representations). In addition, enunciation does not suppose a free and intentional subject, for, as Leandro-Ferreira (2000) clarifies, there is a constant tension between freedom and coercion, as well as the possibility of a play of/in language: "The system keeps in itself both the possibility of a play of/in language and the need for an ordering" (LEANDRO-FERREIRA, 2000, pp.108-109). 58 Such an idea that the system foresees both the possible play and the necessary ordering illustrates the interval nature of enunciation between language and discourse, besides indicating a subject who is neither the full master of language nor entirely subject to it and to the exteriority: "He establishes an active relationship within a given discursive formation; just as he is determined, he also affects and modifies it in his discursive practice" (LEANDRO-FERREIRO, 2000, p.23). 59 The second relevant point of the quote above refers to the second part in italics.
This part still involves the question of the subject, especially his relationship with 57 In Portuguese: "Assim como o lingüista tem acesso à língua pela fala, o analista de discurso tem acesso à ordem do discurso através da organização da língua." 58 In Portuguese: "O sistema guarda em si tanto a possibilidade do jogo da/na língua quanto a necessidade de um ordenamento." 59 In Portuguese: "Ele estabelece uma relação ativa no interior de uma dada formação discursiva; assim como é determinado, ele também a afeta e a modifica em sua prática discursiva." All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 language and with the Other in the enunciative process. As already mentioned, Pêcheux sees, in what he calls forgetting 2, the constitution of the utterance from the selection and the rejection of sayings by the subject, who does it more or less consciously. Indeed, the degree of consciousness is, in this first text of 1975, a distinctive feature between the two types of forgetting, which would profoundly differ from one another. In this regard, Pêcheux argues that the subject can consciously enter the zone of forgetting 2, through the means of a return onto his own speech, an anticipation of its effect, an attempt to explain or reformulate his saying, so that "zone 2, which is the zone of the processes of enunciation, is characterized by conscious-preconscious mechanisms" (PÊCHEUX, 1995c, p.138; italics by the author). 60 In contrast, forgetting 1, according to the philosopher, is located in a zone inaccessible to the subject, because it constitutes him and represses the discursive process and its specific exterior (the interdiscourse, a term not yet named). The two types of forgetting are articulated to each other by relations of contradiction, subservience or encroachment: it is an ideological forgetting and, therefore, unconscious, since ideology is constitutionally unconscious of its own existence.
Although the two forgetting are related, Pêcheux turns to the fact that the constitutive exteriority linked to the first forgetting should not be confused with "the subjective space of the enunciation, the latter being an imaginary space which allows the speaking subject its displacements within the realm of what can be reformulated, so that it continuously goes back over what it has formulated" (PÊCHEUX, 1995c, p.139; italics by the author). 61 In other words, interdiscourse (linked to forgetting 1) should not be confused with enunciation (linked to forgetting 2), which are linked by "a relation of dominance which we can describe by saying that the non-asserted precedes and dominates the asserted" (PÊCHEUX, 1995c, p.140). 62 Therefore, there is a constitutive anteriority of the interdiscourse in relation to enunciation.
Among all the texts analyzed so far, Overview and prospects is the one that further copes with the problem of enunciation. If the study published in the same magazine in 1971 gives discourse a firmer seat in linguistics, this 1975 work provides enunciation with a firmer seat in the theory of discourse. Besides inaugurating the issue of subjective illusion, "Michel Pêcheux presents there the first relationship established between the All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 enunciation and the imaginary. He opens up a perspective for the analysis of enunciative mechanisms" (MALDIDIER, 1990(MALDIDIER, /2003. 63 This analytical perspective is opened through syntax, assumed as the mode of organization of identifiable traces of the enunciation, so that "the study of the marks of enunciation is central to ADA's linguistic analysis" (PÊCHEUX, 1995c, p.138). 64

Enunciation in Language, Semantics, and Ideology
According to Maldidier (1990), the article discussed in the previous subsection, although interesting, is a transition text, while "Language, semantics, and ideology is Michel Pêcheux's great book. It presents the most complete state of the theory" (MALDIDIER, 1990(MALDIDIER, /2003. 65 In this work, enunciation is discussed in more detail in the introduction and in chapters I of the 1st part, I of the 2nd part and III of the 3rd part. In the introduction, Pêcheux identifies three main trends within the linguistics of the time (1970s): 1) the formalist-logicist trend, represented by Chomskyan generativism; 2) the historical trend, represented by the theories of linguistic variation and change; 3) the trend of speech linguistics, also called by the author linguistics of the enunciation/the performance/the message/the text/the discourse. In the latter, Pêcheux highlights the criticism regarding the linguistic primacy of communication, as well as the relevance given to style and dialogue as transgression and disruption, respectively. Although other theoretical perspectives in addition to the enunciative ones can be inserted in this third trend, we emphasize the philosopher's careful reading of some of the main foundations of Linguistics of Enunciation: a) the issue of the concept of language as a communication tool; b) the recognition of the singularity of the use of language, which suffers restrictions from the regularity of the linguistic system, but does not end there; c) the consideration of intersubjectivity as a fundamental principle, in which the self-recognition of the self occurs through the recognition of the you. Of these three foundations, the first signals a sensitive shift from the Pecheutian position in relation to the theories of enunciation: If before these were denounced by the author as promoters of a conception of language as All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 a tool, now, he seems to recognize that such conception is, on the contrary, the target of criticism by these theories.
In the first chapter of the 1st part of the book, enunciation is themed as a modern problem that addresses the ancient question of the origin of languages as a result of man's need to act over the Other and over the world. Pêcheux sees in Condillac's conception of "the gestures of the 'language of action' and the sounds of articulated language […] a resume of how the rudiments of a theory of enunciation are now beginning to emerge" (PÊCHEUX, 1982, p.29; quotation marks by the author). 66 If in the introduction the author emphasizes enunciation characters such as the non-instrumental nature of language, as well as the singularity and the intersubjectivity intrinsic to the use of language, in this reference to Condillac, he highlights the use of language itself as a defining feature of enunciation.
According to Pêcheux, the transition that took place in the 18th century philosophyin which the subject is no longer subordinate to the truth of his speech and becomes the source of that speechis radicalized by the philosophies of subjectivity of the 19th and 20th centuries. In these, subjectivity emerges as a central category and a reworking of the necessary/contingent classic Aristotelian opposition, updated by the modern objective/subjective idealist opposition. On this, Pêcheux mentions Husserl's notions of objective expression as an independent significance of the subject and the enunciative situation (the discourse of science, for example) and of subjective expression as significance which does not depend on the subject and the enunciative situation (as speeches containing personal and demonstrative pronouns and space-time deictics).
According to the author, "Husserl's philosophical consideration of subjective expressions leads him to formulate remarks strangely similar to those made by linguists dealing with the problem of the relationship between situation, enunciation and determination" (PÊCHEUX, 1982, pp.32-33). 67 This "strangely similar" proximity becomes clearer at the end of the chapter, in which Pêcheux uses the Saussurian langue/parole opposition and the Chomskyan competence/performance oppositionwithout covering them upto associate those with the ideological pair creativity/system. Such an association consists of the tension between the subjectivity that creates parole and performance, on the one All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 hand, and the systematic objectivity of the langue and competence, on the other, in which the first affects the second by overloading the intralinguistic elements (phonological, morphological, syntactic) with extra-linguistic elements (subject, meaning, situation).
This makes Pêcheux consider, as a common point to structural and generative semantics, the indifference in relation to the theory of enunciation, seen "as the theory of this residue inherent in the existence of the 'speaking subject' in situ" (PÊCHEUX, 1982, p.39; italics and quotation marks by the author). 68 The author mentions some enunciative mechanisms, such as the anaphoric designation (reference from one segment to another in the discourse), the indexical/deictic designation (reference to the extra-linguistic context) and the evaluative/emotive modalization (the subject's position with respect to the utterance's situation and content), to finally define enunciation as he understands it to be addressed in the linguistics of that time: "In other words, enunciation designates both the fact that the subject is the support for his enounced and the set of subjective effects (different psychological contents) which underlie that enounced" (PÊCHEUX, 1982, p39; italics by the author). 69 In this definition, we see Pêcheux relating the problem of enunciation to the problems of the subject (as support of the utterance) and to issues of meaning (as a set of subjective effects of this utterance), the subjectivism involved in this conception of enunciation does not please the philosopher, though, something that becomes clearer in the following chapters of the work.
For Maldidier (1990), enunciation as a residue impossible to be absorbed by the system reveals the Pecheutian critique of the dichotomous functioning of philosophical and linguistic thoughts. It is a violent criticism of idealist philosophy, "incapable of working with contradiction, shaken between the false solutions of logicism and subjectivism" (MALDIDIER, 1990(MALDIDIER, /2003. 70 This philosophical idealism, whose dichotomies are reproduced by semantics, is exemplified by a series of philosophical pairs, such as logical/rhetorical, necessary/contingent and objective/subjective: "In this affiliation, the opposition langue/parole, and more broadly speaking system/speaking All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 subject, constitutes linguistics own avatar" (MALDIDIER, 1990(MALDIDIER, /2003; italics by the author). 71 It is on the second part of the book, also on the first chapter, that a detailed reflection regarding enunciation can be found. In this reflection, the philosopher returns to the phenomena he dealt previouslyexplicatives, determinatives, determiners, deictics arguing that, if they belong to the language system and refer to the functioning of the langue in regard of itself, they also constitute "the backdrop to a 'philosophical' reflection whose development he [the linguist] has been able to follow through the questions of reference, determination and enunciation" (PÊCHEUX, 1982, p.58; quotation marks by the author). 72 Although this is the only occurrence of the term "enunciation" in the text, it is possible to note that the notion of enunciation surrounds all the discussion present on it: I say that these two elements (both linguistic phenomena and sites of philosophical questions) belong to the zone of articulation between linguistics and the historical theory of ideological and scientific processes, itself part of the science of social formations: the system of a language is indeed the same for the materialist and the idealist, for the revolutionary and the reactionary, for someone with access to a certain knowledge and for someone without that access. "But it does not follow that these various people will hold the same discourse: the language thus appears to be the common basis of differentiated discursive processes […]" (PÊCHEUX, 1985, p.58; italics by the author; italics with quotation marks by us). 73 This quote retrieves what Pêcheux had already said about the need for a linguistic basis for the constitution of the discursive process. To this end, the author makes two remarks: a) the concrete/abstract opposition cannot be superimposed on the opposition discourse/language, since discourse (materialistic notion) is not to be confused with parole (idealist notion) as the concrete use of the abstract system; b) language is indifferent to the class struggle, but social classes are not indifferent to language, which is relatively autonomousso that grammar is not the pivot of class struggle and language does not belong to the ideological superstructure of social formation -, but this relative 71 In Portuguese: "Nesta filiação, a oposição língua/fala, e mais amplamente sistema/sujeito falante, constitui o avatar próprio da linguística." 72 For reference, see footnote 35. 73 For reference, see footnote 35.
All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 autonomy does not prevent any discursive process from being inscribed in an ideological class relationship, which determines it.
In summary, if langue is condition of the discourse, it is so because it is enunciation, taken as a set of mechanisms that are simultaneously linguistic phenomena and places of philosophical reflection, which enables the realization of distinct and ideologically determined discursive processes on the same linguistic basis.
In the third part of the book, the chapter that emphasizes enunciation the most is the one entitled The Subject-Form of Discourse. In this chapter Pêcheux's starting point is the process of interpellation-identification through which ideology interpellates individuals as subjects, concealing the material character of the meaning of words and utterances. This concealment occurs, on the one hand, because of the evidence of the subject as the origin of the self and, on the other, because of the evidence of meaning as being transparent. For Pêcheux, meaning has such a material character because, in addition to appearing transparent to the subject, it is in a relationship of "constitutive dependence on what I have called the 'complex whole of the ideological formations'" (PÊCHEUX, 1985, p.111; quotation marks by the author) 74 : the interdiscourse. This dependence is clarified by the author through the formulation of two theses.
The first postulates that words/expressions/propositions change their meaning according to the positions held by those who use them so that the meaning varies according to the DF in which it is produced. According to Pêcheux, this has to do with the linguistic basis/discursive-ideological process, because if the same word/expression/proposition can have different meanings according to the DF in which it is inscribed, this means that it has no meaning of its own, but constitutes its meaning within each DF. The author adds that, just as the same word/expression/proposition changes its meaning when passing from one DF to another, formally different words/expressions/propositions can have the same meaning within the same DF, which leads him to present a fourth definition of discursive process, as "the system of relationships of substitution, paraphrases, synonymies, etc., which operates between linguistic elements -'signifiers'in a given discursive formation" (PÊCHEUX, 1985, p.112; quotation marks by author). 75 In turn, the second thesis advocates that every DF disguises, due to the transparency of its All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 meaning, the contradictory materiality of the interdiscourse that determines it. The philosopher understands interdiscourse as "this 'complex whole in dominance,'" characterized by "the fact that 'it speaks' ('ça parle') always 'before, elsewhere and independently', i.e., under the domination of the complex of ideological formations" (PÊCHEUX, 1985, p.113; quotes and italics by the author). 76 This makes meaning seem transparent and provides each subject with their "reality" as an evidence system. Enunciation appears here again at the center of a discussion about subject and meaning. According to Pêcheux, the subject-form consists of the way through which the individual is interpellated as a subject when identifying to a DF. This interpellationidentification "depends on the fact that the elements of interdiscourse (in their double form, described above as 'preconstructed' and 'sustaining process') that constitute, in the subject's discourse, the traces of what determines him, are re-inscribed in the discourse of the subject himself" (PÊCHEUX, 1985, p.114; quotation marks and italics by the author). 77 That is, although the subject is constituted by forgetting what determines them, this determination breaks out in their own speech through the two elements of the interdiscourse: the preconstructed and the sustaining process/articulating utterances (or transversal-discourse).
The preconstructed appears in the subject's discourse through the process of articulation conceived as an already-said from previous discourses, but ignored in its origin and considered to have a naturalized meaning. This process connects/chains preconstructed on the axis of what Pêcheux designates as intradiscursive, that is, "the operation of discourse with respect to itself (what I am saying now, in relation to what I have said before and what I shall say afterwards, i.e., the set of 'coreference' phenomena that secure what can be called the 'thread of the discourse' as discourse of a subject)" (PÊCHEUX, 1985, p.116; quotation marks by the author). 78 Inspired by Saussure, Pêcheux also calls this connection/chaining which structures the discursive thread as linearization or syntagmatization, a process through which he brilliantly demonstrates the relationships among language, enunciation and discourse: In another connection, let me point out that interdiscourse as transverse-discourse crosses and connects together the discursive elements constituted by interdiscourse as preconstructed, which supplies as it were the raw material in which the subject is constituted as 'speaking-subject', with the discursive formation that subjects him. In this sense it can indeed be said that "intradiscourse, as the 'thread of the discourse' of the subject, is strictly an effect of interdiscourse on itself, an 'interiority' wholly determined as such 'from the exterior'.
[…] I shall say that the subject-form […] tends to absorb-forget interdiscourse in intradiscourse, i .e., it simulates interdiscourse in intradiscourse, such that interdiscourse appears to be the pure 'already-said' of intradiscourse", in which it is articulated by 'coreference'. This being so, I think the subject-form can be characterised as realising the incorporation-concealment of the elements of interdiscourse: "the (imaginary) unity of the subject, his present-past-future identity, finds one of its foundations here." (PÊCHEUX, 1985, pp.117-118; italics and quotation marks by the author; italics with quotation marks by us). 79 About this quote, we will make two considerations. The first concerns the role of syntax in the production of discourse, which is taken up here for further study. Would the preconstructed and articulation of utterances, described in the 1975 book as the elements of interdiscourse in the subject's discourse, correspond to the elements of enunciative references mentioned in the definition of syntax formulated in the article of the same year? If so, then the preconstructed and the articulation of utterances, as elements of enunciative references, represent not only the elements of interdiscourse in the subject's discourse, but alsoor perhaps for this very reasonthe elements of the constitution of the subject and of the meaning. Once again, enunciation appears as a mediating instance between the organization of language and the order of discourse, which, however, itself undergoes a mediation, that of DF, since there is no direct relationship between the subject and the interdiscourse.
The second consideration refers to subjectivity itself, whose imaginary functioning is treated by Pêcheux in that chapter, through the notion of subject-form.
According to the author, the subject-form simulates the interdiscourse in the intradiscourse, making the interdiscourse appear as the already-said of the intradiscourse, so that this is an interiority totally determined by the exteriority. Thus, the subject-form founds the subject's imaginary unity and makes them forget what determines them in an idealist effect of autonomy that regards them as an interior without an exterior. This 79 For reference, see footnote 35. impossibility for the subject to recognize their subjection is due to the fact that such subjection is manifested in them precisely by the illusion of autonomy. According to Pêcheux, the subject-effect (the identification of the subject with the self) implies the intersubjectivity-effect (the identification of the subject with another subject), in a mutual recognition not only between the subject and the self and among subjects themselves, but also between them and the Universal Subject (what everyone knows). In this (these) recognition(s), the subject forgets the determinations that constitute them.
The first occurrence of the term enunciation, in this fundamental chapter of the theory, appears as an example of the subject-effect that makes the subject feel free, conscious, responsible and the author of their own acts: "[...] the notions of assertion and enunciation are there to designate in the domain of 'language' the subject's actions in taking up positions as a speaking subject" (PÊCHEUX, 1985, p.121; italics and quotation marks by the author). 80 Pêcheux criticizes this notion of speech acts for he believes that it translates the subject's forgetting of what determines the self and makes it seem that their taking up of position in the discourse is an original action of theirs. On the other hand, the author defends that this taking up of a position is not the cause of the subject, but "it must be understood as an effect, in the subject-form, of its determination by interdiscourse as transverse-discourse" (PÊCHEUX, 1985, p.122). 81 In other words, enunciation is not a spontaneous origin of the subject: it is, rather, the subject-form as an effect, a result of his determination by exteriority, by the interdiscourse that, when crossing itself, constitutes intradiscourse as the subject's discourse thread. However, as already mentioned, this determination of the interdiscourse is masked in the intradiscourse by the subject-form, which, by incorporating and concealing the interdiscursive elements (the preconstructed and the articulation) in the subject's discourse, produces in this subject the illusion of being in control of his enunciation and thus founds his imaginary unity (his present-past-future identity).
Based on this criticism of the notion of enunciation, Pêcheux resumesto reevaluate themthe two types of forgetting inherent to discourse. While forgetting 1 was situated in an unconscious zone, because it constitutes the subject and inaccessible to him, forgetting 2 was situated in a preconscious/conscious zone, because it is related All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 to the displacements of the subject within the reformulable. Such reflections become "inadequate, in so far as their net result is to make the preconscious-consciousness an autonomous zone with respect to the unconscious" (PÊCHEUX, 1985, p.124). 82 Emphasizing, then, the primacy of the unconscious, the author redefines forgetting 2 as the site of reformulation-paraphrase of a given DF through which the meaning is constituted, a space that covers the interdiscourse that determines that DF. According to Pêcheux, such cover-up consists of forgetting 1, which is radically occulted of this interdiscourse from the subject who is under the dominance of this DF: "Thus the space of reformulation-paraphrase that characterises a given discursive formation becomes the site of the constitution of what I have called the linguistic imaginary (verbal body)" (PÊCHEUX, 1985, p.126; italics by the author). 83 This linguistic imaginary is associated with the problem of enunciation, which Pêcheux refers to the theoretical absence of a linguistic correspondent of the Freudian imaginary and ego.
Relegated to the background in DA-1, and announced as a promising path to be followed in the transition from DA-1 to DA-2, enunciation is finally theorized in DA-2.
Both in the article and in the 1975 book, it is raised to the center of the theory of discourse, either as an integral component of its general epistemological framework, or as a theoretical-methodological problem to be faced. This discursive reflection on enunciation is accompanied, in DA-2, by a fierce criticism of the linguistic trends of that time, especially by the idealism through which they would end up, on the one hand, in logicism and, on the other, in sociologism and subjectivism, enclosing itself in dichotomies that would make enunciation either a non-theorized residue or a theoretical mirror of an egocentric and solipsist subjectivity. Such reflection is also accompanied by the introduction of new notions, such as subject-form, transverse-discourse, interdiscourse, intradiscourse, forgetting 1 and 2, linguistic imaginary.
So far, we have rehearsed a reconstitution of the enunciation trajectory in DA-1 and in DA-2. Next, we will try to present a systematization of this theoretical path.

Final Words (for Now)
All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 The purpose of this article was to follow the path of the problem of enunciation in the initial development of Discourse Analysis founded by Michel Pêcheux. This objective was formulated from two research questions, which we return now to answer: 1) what notions of enunciation can be derived from Michel Pêcheux's discursive reflection in the early years of his production? 2) What is the status assumed by enunciation in the scope of DA in this period?
As for the first question, we present below, in the form of a schematic summary, the notions of enunciation that can be derived from the analyzed texts: Chart 2. Synthesis of the notions of enunciation derived from the theoretical corpus.

DA-1
Automatic Discourse Analysis (1969)  Enunciation is removed from theorization and reduced to an instrument of linguistic analysis of the discursive surface.
From DA-1 to DA-2 Langue, 'langage', discours (1971)  Enunciation, conceived as a set of internal mechanisms based on the linguistic basis that make possible the realization of the discursive processes on it, is heralded as the decisive point for the future of the relations between the already consolidated theory of language and the emerging theory of discourse.  Enunciation is a condition for the possibility of articulating discursive processes (semantic, rhetorical and pragmatic variables) on the linguistic basis (invariant syntactic background).  Enunciation is a condition for the possibility of the taking up of position of the "speaking subject" regarding the social representations of which he is the support, which refers to the relationship between the discursive process and its conditions of production.  Enunciation, as a general condition for the possibility of discursive processes, is located in a border area, common to language and syntax, on the one hand, and discourse and semantics, on the other, so that the study of enunciative mechanismswhich allow the passage from the linguistic to the discursivewill probably make it possible to put it in an appropriate way and perhaps solve the problem of the relations between the syntactic and the semantic. La Semantique et la coupure saussurienne: langue, langage, discours (1971)  Enunciation is a set of internal mechanisms based on the linguistic basis that articulate the discursive processes on it, relating, on the one hand, to the utterance (linguistically analyzable materiality) and, on the other, to social representations (imaginary projections materially realized by the utterance).  Enunciation is a series of mechanisms that manage the organization of the terms in a discursive sequence (utterance), depending on the conditions in which this sequence is produced, which, in turn, is seen as the necessary materiality for the manifestation of the taking up of position of the "speaking subject" in relation to social representations. Overview and Prospects (1975) All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0

DA-2
 Enunciation is a series of successive determinations that little by little constitute the utterance and delimit the boundaries between the said (the selected) and the unsaid (the rejected).  Enunciation is a mediating instance of the relationship between syntax and discourse, as it operates the transition from the linguistic organization (by allowing the syntagmatization of lexical units) to the discursive order (by allowing the subject's taking up of position regarding the representations put into play in social relations), thus, making language the material site for realizing the meaning-effects.  Enunciation, linked to forgetting 2, is a zone in which the subject can consciously penetrate, characterized by a pre-conscious/conscious functioning by making possible the subject's return over his own speech, an anticipation of its effect, an attempt to make explicit or reformulate his own saying.  Enunciation is an imaginary subjective space that assures the subject his displacements within the reformulable, so that he makes incessant returns on what he formulates.
Language, Semantics and Ideology: Stating the Obvious (1975)  Enunciation is a set of mechanisms that are, simultaneously, linguistic phenomena and places of philosophical reflection and which enable the realization, on the same linguistic basis, of distinct and ideologically determined discursive processes.  Enunciation is a constitutive instance of the subject and of the meaning, since the preconstructed and the articulation of utterances, as elements of the interdiscourse and traces of the enunciative references, are re-inscribed in the subject's discourse, configuring in the discursive thread the traces of what determines the subject and, nevertheless, is unknown by him.  Enunciation is the subject's taking up of position in the discourse by being crossed by the interdiscourse, whose determination is masked by the subjectform, which, by incorporating and concealing the interdiscursive elements in the subject's discourse, produces in this subject the illusion of being in his own origin and of being the source of meaning, thus founding his imaginary unity  Enunciation is a linguistic imaginary, a verbal body constituted in the reformulation-paraphrase space characteristic of a DF and whose functioning conceals from the subject identified to such DF the constitutive exteriority of this DF (his interdiscourse), thus making him forget what determines him, as if it were an interior without exterior (forgetting 1), and also making him forget that there are other possible meanings, as if the meaning were a naturalized and transparent significance (forgetting 2). Source: the author.
It is worth clarifying that this summary is only an attemptand, as such, it is always incompleteto systematize the reading gestures made throughout this study in relation to the theoretical corpus. What we mean is that our purpose, when examining Pêcheux's five texts selected for analysis, was not to develop a concept of enunciation from DA. Our objective was to reconstruct the path taken by the enunciative problematic in the first two phases of the theory and its role in the establishment of its founding bases.
Thus, the notions listed above do not consist of enunciation concepts that we wish to incorporate into the conceptual apparatus of DA. It is, rather, about formulations on a All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 specific problem, derived from a broader and more complex reflection on language, discourse, subject and meaning.
As for the second question that moved us here, from the set of notions of enunciation derived from the Pecheutian reflection from DA-1 to DA-2, we believe that we can attribute to enunciation, within the scope of the theory in that period, an epistemological status as well as a theoretical--methodological status.
The first status involves the place of honor reserved to enunciation in the general epistemological framework that guides DA. This place is shared by enunciation and syntax, both brought together within the second of the three regions of scientific knowledge articulated in that framework, namely, linguistics (as a theory of enunciative processes and syntactic mechanisms), the two other regions being historical materialism (as the theory of formations and social transformations) and the theory of discourse (as the theory of the historical determination of semantic processes)all three crossed and intertwined by a theory of the subject with a psychoanalytic bias.
We take the term epistemology from the meaning presented by André Lalande (1999) in his Technical and Critical Vocabulary of Philosophy, in which he defines epistemology as a critical study of the principles, hypotheses and results of the sciences, which aims to determine its logical origin, its value and its objective importance. In a work on the linguistics/psychoanalysis interface, Flores (1999) reinterprets this meaning, considering epistemology as a reflection that seeks a certain degree of generality and whose "main objective is to extract a necessary link from the order of knowledge for the constitution of a paradigm" (FLORES, 1999, p.16). 84 Therefore, the epistemological status of enunciation, in the first two "epochs" of DA, is due to the emphasis given by Pêcheux to enunciation as a necessary link to the constitution of the new paradigm represented by the theory of discourse in its relationshipconflicting, but constitutivewith the theory of language.
The second status, on the other hand, encloses the role of enunciation as a condition for the possibility of the linguistic basis/discursive process relationship. What is at stake here is the interval nature of enunciation between language and discourse, between systematicity and historicity. This nature is revealed by syntax, as a way of