From System into Action, from Homogeneous to Heterogeneous: Movements of the Founding Concepts of Dialogism, Polyphony and Interdiscourse / Do sistema à ação, do homogêneo ao heterogêneo: movimentos fundantes dos conceitos de dialogismo, polifonia e interdiscurso

Although originally based on a homogeneous perspective of language and, consequently, of the utterance and subjects – the discursive studies have developed toward the recognition of a discourse founding heterogeneity, which is able to support concepts such as dialogism, polyphony and interdiscursivity. This article intends to present this evolution through a brief theoretical retrospective, that covers the Speech Acts by Austin, as well as the constitutive heterogeneity of AuthierRevuz, the dialogism of Bakhtin and also through the polyphony of Oswald Ducrot. Derived from this turning point of the discursive studies, these concepts are considered by means of a wider reflection, which understands the institution of the heterogeneous discourse as its foundation.


Introduction
Despite the diversity of theoretical approaches developed from the foundation of a linguistic science, the positions adopted to reflect on language can be summarized in three basic paradigms: realistic, mentalist and pragmatic. United by a representational perspective, the first two conceive language as an instrument of representation of reality and thought, respectively; the meaning being maintained in both by an objective relation between the word and the entity it represents. Otherwise, the pragmatic perspective thinks language as practice shifting the sense of that binomium to build it in its own praxis.
From the three positions are, thus, built the two opposite poles that guided the debates of Modern Linguistics, which reflect a founding dissociation between logic and rethoric that has always been present in the history of Western thought: logic, observed from the Platonic and Aristotelian perspectives, focuses on the problem of language as representation -concrete or mental -by raising the question of conditions of truth in the utterance through an analysis of the proposition; rhetoric, propriety of the sophists, not recognizing the condition of language representation, turns to the study of the persuasive power of discourse, in order to aprehend language as discourse that produces effect, as a power capable of intervention in reality.

From the Theory of Enunciation to the most current approaches in Discourse
Analisys, discourse studies generally conceive language as a subject of action and, thus affiliate to a rhetorical founding discourse, identified to a pragmatic linguistic perspective. However, the successive theoretical reworkings traversed by these studies with regard to its own object -discourse -require the recognition of three phases distributed in a continuum from a homogenizing discourse, understood as a discrete and uniform entity, to the constitutive heterogeneity, which perceives multivocality as a hallmark of discourses.
Establishing a heterogeneous concept of discourse is what makes possible the introduction of the concepts of dialogism and poliphony, proposed as main theme of this edition. In this manner, perceiving that the reflexion about the theoretical grounds that support the creation of these concepts is vital to their understanding, this article proposes a chronologic presentation of some founding concepts of discourse studies and, consequently, the discussion of the theoretical basis on which they are settled.
1 The evolution of concepts: from logic to rethoric, from system into action Despite the privilege of logic having been constant in linguistic studies, it is from an approach more oriented to rethoric that J.L.Austin takes the first steps towards the establishment on enunciative-discoursive studies.
In How to Do Things with Words (1962), Austin starts from the evidence of an opposition between performatives and constatatives to study the extent to which saying means acting. Thus, opposing the position of the logic that every proposition is either true or false, the concern of this author is no longer the pursuit of truth in the analisys of the sentence, but the effectiveness of the act and the speaker´s engagement in communicative interaction, because he recognized that [...] In any circumstance of utterance, to say is always to perform three acts: 1 produce certain sounds; 2 produce certain words (which are part of a lexicon) according to certain constructions (that is, a grammar), using intonation; 3 employ elements of type 2 with a more or less determined meaning. With a given reference, sense and reference would constitute a meaning. 1 [our translation] (GUIMARAES, 2002, p.38).
The set of these three acts has been called by Austin Speech Acts and through them, he recognized the presence of the action of a speaker throughout the utterance.
Extending it to the identification of a perlocutionary and another locutionary act, the Theory of Speech Acts involved elements such as context, usage conventions and the speaker´s intention. Therefore, apart from the complexity involved in the introduction of elements which had been excluded from linguistic studies until then, Austin´s contribution to the discourse studies is owed mainly to the understanding that the use of language is always an act, portraying language as social concrete practice.
As in Austin, the concerns with the subject and the act of enunciation were also inscribed in the formulations of Émile Beneviste; in this sense the author has declared: When describing a few years ago, the forms of subjective linguistics enunciation, we indicated briefly the difference between "I swear", that is an act, and "he swears" that is just an information. The terms performative and constative had not appeared yet, but this was, after all, the substance of the definition (BENVENISTE, 1973, p.237).
The aparent equivalence between both theoretics came apart, however, in the criteria used to distinguish the acts. Unlike Austin, who presented "extralinguistic" conditions, Beneviste adopts criteria that are exclusively linguistic and formal. This formalism will characterize the Enunciation Linguistics created by him.
According to Beneviste, the transformation of language into discourse is linked to the act of enunciation, which, in turn, is established in the relation of the speaker to language. The enunciation is, thus, understood as the act by which a speaking subject appropriates the formal apparatus of language, putting it into operation. In this consideration there are two unfoldings of vital importance to the development of discourse studies: the first points to the preassumption of a subject as source of the process of convertion of the language into discourse; the second points to the recognition of this (enunciation) subject as a linguistic issue.
Considering that it "it is in and through language that man constitutes himself as a subject", Beneviste (1973, p.224) treats subjectivity as the ability of the speaker to propose himself as subject. That is when, by appropriating the language, the speaker qualifies as "I". Thus, according to the author, "it is "ego" that says ego. Here we find the foundation of "subjectivity" that is determined by the linguistic status of person" (BENEVISTE, 1973, p.224).
Given the dialogical character of language, the use of "I" is necessarily related to the existance of someone who is assigned to "you." It is in this sense that Beneviste, besides assigning the possibility of language to the introduction of each speaker as the subject, binds the occurrence of "I" to the proposition of another person, "the one who, being, as he is, completely exterior to "me", becomes my echo -to whom I say you and who says you to me" (BENVENISTE, 1973, p.225) The reciprocity involved in the dual I/you establishes a relation of polarity which, under Benveniste"s perspective, configures assimetrically since, although concepted as compliementary elements, ego is always in a position of transcendence over "you." The equality of terms is therefore restricted to the fact that they are both taken as "empty" linguistic forms that do not refer to any specific concept or individual.
They are linguistic forms which "it is in and through language that man constitutes himself as a subject" (BENVENISTE, 1973, p.226).
The personal pronouns, however, constitute only the first supporting point for addressing subjectivity. Yet, there is still the need to refer to the other classes, defined according to the instance of discourse in which they are produced.
The establishment of discourse depends on the organization of I/here/now and also the world around these coordinates. It is in this sense that the indicators of deixis, demonstratives, adverbs and adjectives, having the subject as point of reference, organize the spatial and temporal relations. Therefore, the deictic designation constitutes the first anchor point of the subject and, by allowing a first meaning to the notion of subjectivity, stands in the center of the problematic of enunciation.
Following the proposal of Benveniste, Kerbrat-Orecchioni (1980) suggests, from a theoretical double slip suffered by the term "enunciation", the consideration of two ways of conceiving the linguistics of enunciation, one "extended" and the other "restricted." To better clarify this double concept, see the words of the author herself: (a) Conceived broadly, the linguistics of enunciation aims at describing the relations woven between the enunciation and the different elements which constitute the enunciative mark. (b) Considered in a restrictive sense, the linguistics of enunciation is not interested in more than one of the parameters constituting the framework of enunciation: the speaker-writer. Such is the attitude we adopt here, at least with respect to most of our study. 2 [our translation] ( KERBRAT-ORECCHIONI, 1980, p.41-42).
Despite recognizing a linguistics of the enunciation "extended" to all the elements that constitute the act of enunciation, when positioning herself on a" restrictive" concept, the author associates to the Benveniste perspective, centralizing the problem of enunciation, and consequently, subjectivity, in the observation of the speaker.
As for this theoretical approach, there is also a strong presence of a certain formalism in studies by Kerbrat Orecchioni, defining "her" problem of enunciation as "the pursuit of linguistic procedures (shifters, modalizers, evaluative terms and so on) with which the speaker prints his mark in the utterance is inscribed in the message (implicitly or explicitly) and relates to it" 3 [our translation] (KERBRAT-ORECCHIONI, 1980, p.43).
As it turns out, the first steps towards the establishment of a theory of enunciation and hence of discourse, are given by the break with the logical concept, concerned with the question of vericonditionality of sentences. Despite the differences established between these proposals, it must be observed that in all of them there is a concern with the activity of a subject enunciator, constituted in the individual act of enunciation of appropriation of the linguistic system.

New guidelines: from homogeneous to heterogeneous
With the evolution of scientific thought, the "individualized" character that permeated the founding concepts of an enunciative theory came to be understood according to the diversity introduced in the sciences of the time. That is how, in the passage from modernism to postmodernism, the subject proposed by the rationalistic epistemology is succeded by a subject problematised by history, ideology and psychoanalysis. Following the theoretical tendencies at that time, linguistic studies broke with a concept of language and subject bond to the Cartesian-humanist perspective, to propose a concept permeated by the notion of heterogeneity.
Discovered by the West thirty years after its publication -the first texts dating from 1920 -the reflections proposed by Bakhtin 4 were fundamental in the opening of a mindset closer to the plural, the multivocal, the heterogeneous.
Understanding language as constitutive human condition, taking into account its psychic dimension -approached by consciousness and ideology -and interdiscoursivity that traverses the subject, Bakhtin starts from a notion of language that has in the social phenomenon of verbal interaction its true substance, to recognize the importance of considering, broadly, the social aspects as constituent elements of the enunciation. Accordingly, he claims the enunciation is "a pure product of social interaction, whether it is an act of speech determined by the immediate situation or by the broader context that constitutes the living conditions of a particular linguistic community" (VOLOŠINOV, 1986, p.85).
Thereby rejecting an individualized subject, the question of subjectivity comes to be observed by the Russian theorist from an intersubjective perspective. That is, what is at stake are not the relations of a transcendental subject with the language, but a subjectivity marked by ideological activity and constituted in the interaction between the self and the other. It is a social subject, inserted in the memory and history that cannot be conceived outside the relations of one to another, understood as constitutive of both the subject and the identities.
In this perspective, the notion of "other" is not configured simply as the exterior counterpoint whose existence is essential for the interaction, what is revealed is the appointment to a double consideration of "other", understood as discourse and as receiver. On this, by working in a pioneering way with the quoted speech, the author makes the following statement: What we have in the forms of reported speech is precisely an objective document of this reception. Once we have learned to decipher it, this document provides us with information, not about accidental and mercurial subjective psychological process in the "soul" of the recipient, but about steadfast social tendencies in an active reception of other speacker`s speech, tendencies that have crystallized into language forms. (VOLOŠINOV, 1986, p.117).
The fundamental question of Bakhtin's dialogism is then constructed to treat alterity not as a multiplicity of consciousness, but as a relation of decentering language itself. Being thus established as propriety of language, heterogeneity shall constitute the reflections of this author on the enunciation and the subject. It is following this type of reflection that other theorists advance: M. Pêcheux proposes the notion of interdiscoursivity, Ducrot reorganizing the concept of polyphony and Authier-Revuz establishes the concepts of constitutive heterogeneity and shown heterogeneity.
In "A Análise do Discuso: três épocas" (1997), Pecheux, summarizing the course of a discourse analysis devised by him, establishes three phases, so that in the first, there is a structuralist position, in which the subject is determined by the place from where he speaks, that is, the interior of a discoursive formation conceived as a closed structure device; in the second, recognizing the need to pinpoint the relation of discoursive formation with its exterior, there is the notion of interdiscourse. There are, however, no major changes in the concept of the subject, whereas in the third phase, the construction of discoursive objects and events is observed in intradiscourse, constituting then the issue of alteriry in discoursive identity.
It is in this third phase that Pêcheux"s proposals show consistent with Bakhtin's reflections. Regarding what has just been stated, see the following quote, in which Pêcheux and Fuchs, despite recognizing the enunciation attached to a subject, do not conceive it as an individual act anymore: If we define the enunciation as a relation always necessarily present of the utterer with his enunciation, then it clearly appears, at the very level of language, a new form of illusion according to which the subject is the source of meaning or identifies the source of sense: the subject discourse is organized by reference (direct, dissenting), or without reference to the situation of enunciation (the "I-here-now" speaker) he experiences subjectively as many sources as there are references (people, time, locations). Every language activity needs the stability of these anchor points for the subject; if this stability fails, there is a shakeup in the structure of the subject and the language activity 5 [our translation] (1997, p.174). 5 Se definimos a enunciação como a relação sempre necessariamente presente do sujeito enunciador com o seu enunciado, então aparece claramente, no próprio nível da língua, uma nova forma de ilusão segundo a qual o sujeito se encontra na fonte do sentido ou se identifica à fonte do sentido: o discurso do sujeito se organiza por referência (direta, divergente), ou ausência de referência, à situação de enunciação (o "euaqui-agora" do locutor) que ele experimenta subjetivamente como tantas origens quantos são os eixos de referenciação (eixos das pessoas, dos tempos, das localizações). Toda atividade de linguagem necessita da What is placed in the passage quoted above is the need to recognize the enunciation as a process not only individual or intentional, but from all the social and historical process in which it operates. In the words of the authors: We say that the process of enunciation consists of a series of successive determinations by which the utterance constitutes gradually and is characterized by putting the "said" and consequently reject the "unsaid". The utterance equates to setting boundaries between what is "selected" and has gradually become necessary (through the formation of the "universe of discourse"), and what is rejected. In this way, there is an empty space of "all that could have been said by the subject (but he does not)" or the field of "everything that opposes what the subject said " 6 [our translation] (PÊCHEUX and FUCHS, 1997, p.176).
The establishment of these boundaries does not only intend to impose a boundary between what is "selected" and what is not. Through them, not only the "unsaid" is excluded, but also, and especially, one points to the fact that the selection is somehow marked. Thus, the selection to which Pêcheux Fuchs refer is not restricted to the simplification of a paradigmatic relation, more than that, it is involved in the assumption of a discoursive position, because it is based on the affiliation / rejection of certain discourses. This is where interdiscoursivity is established and hence the approach with the perspective of Bakhtin.
Understanding that every discoursive formation is associated with a discursive memory and builds from formulations that repeat, refuse and transform other formulations, Pecheux and Fuchs (1997) define interdiscourse as a process of unceasing reconfiguration in which a discoursive formation is led to incorporate elements prebuilt outside. In this sense, postulating the historical determination of a non-individual meaning, the author dismisses the subject from the realm of his saying, conceiving it as a subject-effect, subjugated to the action of a radical "Other". estabilidade destes pontos de ancoragem para o sujeito; se esta estabilidade falha, há um abalo na própria estrutura do sujeito e na atividade de linguagem. 6 Diremos que os processos de enunciação consistem em uma série de determinações sucessivas pelas quais o enunciado se constitui pouco a pouco e que tem por característica colocar o "dito" e em conseqüência rejeitar o "não-dito". A enunciação equivale pois a colocar fronteiras entre o que é "selecionado" e tornado preciso aos poucos (através do que se constitui o "universo de discurso"), e o que é rejeitado. Desse modo se acha, pois, desenhado num espaço vazio o campo de "tudo o que teria sido possível ao sujeito dizer (mas que não diz)" ou o campo de "tudo a que se opõe o sujeito que disse". Given these theoretical assumptions, the enunciating subject is understood as a kind of spokesman for discourses preceding it. That is, it is produced as if to internalize, elusively, the pre-construction imposed by its discoursive formation. Thus, the proposal of Pêcheux and Fuchs (1997) recognizes the constitutive alterity of discourse, but in relation to subjectivity, gives "I" a null role, characterizing it basically by subjection imposed by its discoursive position. The establishment of discoursive heterogeneity costs, to this proposal, the invalidation of any individual responsibility of the subject.
As it turns out, the similarities between the theoretical propositions of Pecheux and Bakhtin are restricted to the recognition of the social and ideological constitution of enunciation and hence the heterogeneous character of the interdiscoursive construction.
And the distance established between these authors on the issue of subjectivity lies precisely in different interpretations of these considerations: while Pêcheux interprets them as corroborative to the cancellation of the subject, Bakhtin uses them to build a "self" among "others".
In accordance with Bakhtin's reflections on the constitution of the subject, O.
Ducrot builds his theory of polyphony based on heterogeneity. Thus, appropriating the Bakhtinian concept of polyphony, the scholar will demonstrate how the principle of dialogism works on the utterance level.
Defining the enunciation as the historical event on the appearance of the utterance, Ducrot (1977) establishes a concept of enunciation not remitted to the subject.
Thus, he puts himself in a position of building an entire theory of subject and enunciation as a representation of what the utterance brings from its enunciation.
From these propositions, the existence of a source and a target is no longer conceived as inherent to the enunciation. It therefore starts to be accounted amongst the qualifications that the meaning of the utterance attributes to the enunciation. That's because, despite the enunciation being an empirical point of view, the action of a single speaking subject, the image that the utterance builds constitutes a hierarchy of speech. Therefore, the meaning of the utterance is responsible for the assignment of one or more subjects who would be its origin, which distinguishes Ducrot into three categories: author, speaker, and utterer.
According to this perspective, the author is conceived as the empirical speaking subject that produces the utterance. Considering it as just a psycho-physical-physiological agent in the action of speaking, the author is not, then, conceived as linguistic category, and thus, is not relevant to the study of enunciation.
Otherwise, the speaker is conceived as a being that is presented by the utterance as its responsible and not necessarily is assimilated into the real author. The apparent paradox set this perspective is undone by the recognition that, differently from the author, the speaker is a discoursive being that is only established in the sense of the utterance, it is to the pronoun I -as well as the other first person marks -that it refers.
Beyond the differences between the speaker and the speaking subject, Ducrot also suggests that not all enunciations constitute themselves as products of an individual subjectivity. In this sense, he states: The presence of first person marks makes the utterance as attributable to a speaker, assimilated to the person to whom it refers, however there is the possibility to display, in a statement attributed to a speaker, an enunciation attributed to another speaker 7 [our translation] (1987, p.184).
Thus, from the notion of speaker, one should distinguish two more: the "speaker as such", indicated by (L), and "the speaker as a being in the world" whose sign is determined by (λ). While L is responsible for the enunciation, considered solely from this property; λ is a "complete" person who, among other properties, is the origin of the utterance. However, it should be noted that both L as λ are addressed by Ducrot as discourse beings whose methodological status is totally different from the empirical speaking subject. In an attempt to clarify the differences between L and λ, one should consider that: The being who is credited with sentiment in an interjection, is L, the speaker seen in their enunciative engagement. It is to λ , instead, that it is assigned in the declarative utterances, ie, when the world, among other properties, has to articulate his sorrow or joy 8 [our translation] (DUCROT, 1987, p.188). 7 A presença de marcas de primeira pessoa apresenta a enunciação como imputável a um locutor, assimilado à pessoa à qual remetem, entretanto há a possibilidade de fazer aparecer, em uma enunciação atribuída a um locutor, uma enunciação atribuída a um outro locutor. 8 O ser a quem se atribui o sentimento, em uma interjeição, é L, o locutor visto em seu engajamento enunciativo. E é a λ, ao contrário, que ele é atribuído nos enunciados declarativos, isto é, ao ser do mundo que, entre outras propriedades, tem a de enunciar sua tristeza ou alegria .
Differentiating itself from the speaking subject, but also from the (s) speaker (s), the notion of utterer is identified from the conception of "perspective center" established by Genette 9 , in which he refers to "the beholder". To Ducrot, the utterer is the person from whose point of view the events are presented. Affirming, through comparison with the narrative elements, that the utterer is to the speaker as the character is to the author, he defines the utterers as follows: I call enunciators beings who are regarded as expressing through enunciations, without being assigned precise words; if they speak is only in the sense in which the enunciation is viewed as expressing his views, his position, his attitude but not in the material sense of the term, his words 10 [our translation] (DUCROT, 1987, p.192).
While the source of a discourse is assigned to the speaker, the attitudes expressed in this discourse can be configured through utterers from whom he approaches or distances. In this sense, it is observed, in the role of the utterer proposed by Ducrot, an approximation to the notion of discoursive position established in Pêcheux. It should be noted, however, that this approximation occurs only in relation to the consideration of a discoursive position -which unites both theoretical perspectives with the proposals of Bakhtin. While for Pêcheux the subject is subjugated to a memory of unlimited discourses, the heterogeneous subject of Ducrot does intervene in his discourse with different voices, through a conscious and intentional strategy. And it is in reflection between these concepts of subject that, by changing the irreducibility of these proposals to allow consideration of one under another, Jacqueline Authier-Revuz inserts her studies.
Building her theoretical-descriptive outline from the articulation between the dialogism of the Bakhtin Circle and the propositions of Lacan in his reading of Freud, the aforementioned author presents a very specific concept of heterogeneity and constitutive presence of alterity as the core of her concept of language, subjectivity and meaning. At this point, it is important to emphasize that, although establishing a base composed of two theories with roots and consequences quite apart, Authier-Revuz promotes a consistent approach between them through a fundamental point of contact which conceives "the other " as inalienable to the constitution of identities of subjects, the ways to manifest them and establish them in and through language" 11 [our translation] (BRAIT, 2001, p.10).
Establishing itself as a counterpoint to the aforementioned theoretical propositions about the concepts of Bakhtin concerning enunciation and subjectivity, Lacan´s proposition defines the "other" in several ways, instituting a few steps in the theoretical construction of alterity. However, for the purposes of this article, it is sufficient to consider that the author: [...] placed the question of alterity, that is, the relation of man with his environment; with his desire and the object, from the perspective of a determination of the unconscious. More than "others" [Freudians] however, he sought to show that radically distinguishes the Freudian unconscious -as another scene, or as a third place that escapes consciousness -all unconscious conceptions derived from psychology. That's why I coined a specific terminology ("Other" / "other") to distinguish between the jurisdiction of third place, ie, determining the Freudian unconscious ("Other"), which is the field of pure duality ("other") in the sense of psychology. 12 [our translation] (ROUDINESCO andPLON, 1998, p.558 cited BRAIT, 2001, p.16). Unlike Bakhtin, alterity is treated by Lacan through the unconscious perspective and its double "Other" / "other" constructed from very diverse roots than the ones proposed by the Russian theorist. It is, however, precisely from this diversity that Authier-Revuz promotes compatibility between the ideological dimension and the dimension of the unconscious, not endorsing them, but recognizing them from a linguistic materiality understood as a place of exhibition and constitution of identities and subjects. 11 o "outro" como inalienável, por diferentes caminhos, da constituição das identidades, dos sujeitos e das formas de manifestá-los e constituí-los na e pela linguagem. 11 [...] situou a questão da alteridade, isto é, da relação do homem com seu meio, com seu desejo e com o objeto, na perspectiva de uma determinação do inconsciente. Mais do que "outro"s [freudianos], entretanto, procurou mostrar o que distingue radicalmente o inconsciente freudiano -como outra cena, ou como lugar terceiro que escapa à consciência -de todas as concepções de inconsciente oriundas da psicologia. Por isso é que cunhou uma terminologia específica ("Outro"/ "outro") para distinguir o que é alçada do lugar terceiro, isto é, da determinação pelo inconsciente freudiano ("Outro"), do que é do campo da pura dualidade ("outro") no sentido da psicologia.
Thus, connecting the perspectives of Bakhtin and Lacan, the author questions the uniqueness of a homogenizing conception of discourse and establishes the enunciation as inevitable place of heterogeneities or "no coincidences". In order to do so, she draws on the Lacanian category of the imaginary, understanding, as Flahaut (apud AUTHIEZ-Revuz, 1998, p.188) that: The veil [...] that we interpose between the actual functioning of speech and conscience we take from it [...] [should] not be considered only negatively, as a pure illusion without measure, obscuring the reality: the opacity is itself a reality, and what you need to recognize is that the veil (with its illusion effects), we could not live without it. It is, therefore, about taking seriously the superficial, the foam of everydayness, the zone of everything that comes to conjure the unbearable emergence of real [...] the space where they are produced and where the mediations circulate whose texture merges the symbolic and the imaginary. 13 [our translation].
Established as inherent to the constitution of the human subject, the imaginary allows the permanence of the illusion of "center" thus establishing itself as an instance of the subject responsible for ensuring the illusion of ONE to allow it to function as non-ONE. This is not a simple split or decentering of the subject, but rather a recognition of the inevitable heterogeneity in the enunciation -and consequent incompleteness -that makes the subject in its illusion of unity and dominance, negotiate with it, locating and delimiting the place of the other to circumscribe its own territory.
In other words, what is proposed in Authier-Revuz is the consideration of a heterogeneous constitution of the enunciation that, set out from the vital illusion of the imaginary, circumscribes the subject between the conscious and unconscious and introduces the other as a constitutive parameter of identity itself. As it turns out, rather than an alleged dilution or deletion, this subject "speaks and remains in the ghost under the form of the Ego" 14 [our translation] (ROUDINESCO, 1997, p.42 Cited AUTHIEZ-Revuz, 1998 Recognizing the utterance as constitutively heterogeneous is to understand it is not only according to post-Freudian psychoanalytic parameters, but also to reaffirm its social nature. That is, it is in the consideration of the complex relations between concrete interlocutors, socially and historically situated, that the imaginary intervenes to constitute subjectivity from alterity. In this sense, Authier-Revuz inscribes in his studies the proposals of Bakhtin, whose association to the Lacanian conceptions enables, to the subjection of Pecheux and the deployment of Ducrot, the constitution of a subjectivity of the non-one.
Thus, resuming and relating the points made in this section, we conclude that while Bakhtin, privileging the study of the "aesthetics of verbal creation", establishes , pioneeringly, the dialogical character of language from the relations between the "I "and the" other "and Althier-Revuz relates them to Lacan's psychoanalytic categories in order to institute a heterogeneity/alterity formed between the ideology and the unconscious, Pecheux and Ducrot resort to this heterogeneous concept more punctually: the first creates the concept of interdiscoursivity establishing relations of the "universe of discourse" and the second links the question of subjectivity to the consideration of the different voices of the discourse, instituted in the polyphonic concert of the enunciation.
Faced with such observation, it should be recognized that, despite the different perspectives adopted by the authors, all of them conceive language as an area of heterogeneity. Thus, considering the inability to dissociate language and society and, consequently, to define any externality between the subject and his discourses, alterity is established in subjectivity, heterogeneity in the enunciation and the consideration of interdiscourse as indisputable foundation for discourse studies.

Final Considerations
This article aimed, through a brief retrospective of discourse studies, to identify the foundation of concepts such as dialogism, polyphony and interdiscourse in a heterogeneous conception of language, discourse and subject.
As pointed out in the first section, despite the fact that the enunciative origin has established an important disruption with the logical perspective and initiated a pragmatic approach in Linguistics, early studies brought a concept of the individual subject as center of the linguistic production and, therefore, assimetric in relation to the other elements of the enunciation, establishing a dialogic character restricted to the recognition of "you" as an interlocutor in communication, external to the subject and discourse.
The western discovery of Bakhtin texts, in accordance with the scientific and philosophical expectations of the time, enabled a change in the homogeneizing concept of language. This also allowed the discussion about concepts such as subjectivity and enunciation to cover all the complexity these terms hold: enunciation spreads beyond the circumscription of the immediate situation, taking into consideration broader social restriction; the subject, seen as a part of this social institution, is then formed by the "other" that is represented discoursively. Only when this "other" -social, historical, ideological, psychic -is included in the configuration of the "I" subject and the discourse itself, is that it establishes the effectively dialogic character of language, brought to life in discourse through the multiple voices that constitute and shape it, explicitely or not.
Dialogism, polyphony and interdiscourse are thus concepts which, although distinct, are necessarily configured on the equivalence of a heterogeneous view of language, which deprives the exclusivity of the ego in subjectivity, conceiving it by alterity; it recognizes the presence of these other enunciative voices as constitutive of all discourse, and is thus able to identify discursive affiliations organized in order to reproduce or refute each other.