1I am about to put together various things in this paper: lexicographical definitions of the word language (and its French equivalent langage), Michel Foucault’s archaeology, some thoughts from Husserl and Saussure, and finally some contemporary uses of the notion of mediation. If I were to adopt a traditional approach in the history of ideas, it would take a great many pages to bring these things together in a homogeneous and coherent whole, for example as a theory or a system of philosophical thought. But my semiotic approach, as it can be broadly defined as a structuralist approach, is not looking for homogeneity and coherence. It does not produce comparisons between these things, but only seeks to discover and show in what way they are comparable. Or, as developed in a recently published book of mine (Badir 2022), the semiotic approach tends to establish relationships of applicability between things considered in their diversity. In this way, these things fully preserve their singularity and their power of alterity. Incidentally, but fortunately, this way of seeing can be quite sparing in phrases.

1 | Expression, communication: two conceptions of language

2Structuralism was, from the beginning, linked to a reflection about the relationship between language and thought. Cassirer, who probably introduced the very term structuralism into the philosophical debate,1 was very attentive to the progress of the linguistics of his time. And, although he placed the foundation of structural ideas in Kantian schematism, he was aiming for its most innovative and promising extension in the philosophy of grammar, which was developed by Kant’s disciple Wilhelm von Humboldt.

3Dictionaries have captured the presence of a relationship between language and thought by making the former the expression of the latter. In doing so, they have, in a way, enshrined this relationship in the common meaning of the word language. To take up here a distinction proposed by Georges-Elia Sarfati (2021) between canon (stated), vulgate (diffused), doxa (explicitly normative) and ideology (normative in an underlying way), dictionaries have transformed the philosophical vulgate into a commonly shared doxa.

4However, at the turn of the twentieth century, a remarkable change occurred in the definition of the word language. As regularly as language had previously served for the expression of thought, it is by then defined by its function of communication between human beings. This change can be observed in both French and English dictionaries, suggesting that the doxa in question goes beyond idiomatic usage and national cultures. Dictionaries, however, are obviously only witnesses of a paradigm shift that has taken place in philosophical conception.

5Without any concern for completeness, here is a sample of definitions in both languages. Lets’ start with a review of French dictionaries of the nineteenth century:

Dictionnaire de l’Académique française, 6e édition (1835) : « Emploi que l’homme fait des sons et des articulations de la voix, pour exprimer ses pensées et ses sentiments ».
Bescherelle (1856) : « Pouvoir donné à l’intelligence de se manifester par des signes. »
Littré (1873) : « Proprement, emploi de la langue pour l'expression des pensées et des sentiments. »
Hatzfeld & Darmesteter (1895) : « Expression de la pensée par la parole ».
Petit Larousse (1906) : « Emploi de la parole pour exprimer les idées ».

6An exception to this doxa comes from Laveaux (1820), which seems to be far ahead of its time, where the two functions are intertwined with each other:

« Manière dont les hommes se communiquent leurs pensées, leurs sentiments ».

7English dictionary of the nineteenth century:

Webster (1886) : « Human speech; the expression of ideas by the voice; sounds, expressive of thought, articulated by the organs of the throat and mouth ».

8At the turn of the twentieth century, Century (1895) advances both functions:

Century (1895) : « The whole body of uttered signs employed and understood by a given community as expression of its thoughts; the aggregate of words, and of methods of their combination into sentences, used in a community for communication and record and for carrying on the processes of thought ».

9This is also what the French dictionaries published in the twentieth and twenty-first centuries do:

Petit Larousse : « Faculté que les hommes ont de communiquer entre eux et d’exprimer leur pensée au moyen de signes vocaux (la langue), qui peuvent éventuellement être transcrits ».
Robert : « Fonction d’expression de la pensée et de communication entre les hommes, mise en œuvre au moyen d’un système de signes vocaux (parole) et éventuellement de signes graphiques (écriture) qui constitue une langue ».
Trésor de la langue française informatisé : « Faculté que les hommes possèdent d'exprimer leur pensée et de communiquer entre eux au moyen d'un système de signes conventionnels vocaux et/ou graphiques constituant une langue ».
Wikipédia francophone : « Le langage est la capacité d’exprimer une pensée et de communiquer au moyen d’un système de signes ».

10The contemporary English dictionaries are even more radical: they retain exclusively the communication aspect in their definition:

Oxford : « The method of human communication, either spoken or written, consisting of the use of words in a structured and conventional way ».
Collins : « A language is a system of communication which consists of a set of sounds and written symbols which are used by the people of a particular country or region for talking or writing. »

11Thus, at the time when structural linguistics was being developed, the very idea of language had changed radically since Humboldt's time. Is Cassirer right to maintain that “the program of structuralism developed by Viggo Brøndal [whom he quotes in his article] is, in truth, very close to Humboldt's ideas” (Cassirer 1945, p. 117)? It is true that Humboldt’s comparatism differs from Kant’s universalism and, in this respect, can probably prefigure a revised canon in the conception of language. Nevertheless, it is only at the turn of the twentieth century that the canon is transformed into a scientific vulgate, then into a sociocultural doxa.

12In this essay, I would like to measure this transformation in the conception of language. To do so, I will start by giving it a framework, that of archaeology conceived, as I will explain, as a structural model of historicity. I will then compare two theorizing projects that were almost contemporary, namely Husserl’s and Saussure’s, in order to determine the stakes of the transformation. Finally, I will pose the hypothesis of a new transformation of the conception of language based on such archaeological model.

2 | The historicity of an archaeology

13To situate the preponderance of a concept, as it has been able to establish a scientific paradigm in a social history, requires the consideration of a long, wide and, more particularly, deep stretch of time. The social changes that can be correlated to such a concept cannot be subjected to schemes of strict causality, but are rather measured by their duration and the anticipation of their development. There would then be a way to consider an archaeology, in the sense that Michel Foucault (1966) gave to this term, relative to language in its relations to thought and to social life.

14Let us recall what were the stakes of the archaeological approach in order to provide a base to the reasoning that we seek to carry out here. Hubert Dreyfus and Paul Rabinow (1984, p. 37) confided that Foucault had thought of giving the title “An Archaeology of structuralism” to the book now known as Les mots et les choses. In the end, the formula “An Archaeology of Human Sciences” (“Une archéologie des sciences humaines”) was retained, but only as a subtitle for the book, while his next work was entitled An Archaeology of Knowledge. Whatever the nuances of orientation presented by these formulations, archaeology is indeed the theoretical concept presiding over the project. It consists in proposing a history of knowledge, or rather a history of a relationship to knowledge (this relationship established, to put it simply, between words and things), in accordance with three ages: a central age, another one that precedes it and a third one that follows it. The central age, called “classical,” provides a principle of intelligibility. In regard to knowledge, this principle corresponds to representation, by which a thing is known by its name. Before this classical age, knowledge had been developed on the basis of a relationship of similarities established between things, words not being themselves fundamentally different from things; but the mutation that classical science has undergone is such that this knowledge soon became obsolete. As for structuralism, —whose action on knowledge in general, and on science in particular, Foucault ended up recognizing, stems from the emergence of the human sciences, —it confirms, by the crisis of representation that it generalizes, the passage to another age, called “modern”. From then on, it is easy to understand why knowledge, human sciences and structuralism, even if they are not held for equivalent concepts, are consigned in one and the same archaeological project: knowledge designates the general object put under study, human sciences stands for the decentering that is operated in relation to an object posed as central, i.e. science, while structuralism suggests the contemporary revealer of this decentering.

15So, it is a certain history of knowledge that the archaeological project instantiates, a history in which for each age a conception of knowledge is established, as well as its social and cultural conditions: something like the forms of its historicity. Because they have knowledge as their object, Foucault describes such historical forms as epistemes, and the theory that animates them is that of the sign —hence the fact that structuralism is not only the horizon of the archaeological project but also its starting point, which reveals its own historical situation.

16One last remark about the archaeological project, for it is important for the path that I am undertaking: the functions of the sign, that is to say the type of relation that is established between words and things, are ahistorical in themselves. For instance, representation is the very definition of the sign according to Saint Augustine: aliquid stat pro aliquo; thus, this definition was known well before the classical age. Besides, the observation of similarities between things, with the principle of analogy that serves as their driving force, is not an epistemic act that would have lost its vitality in the nineteenth century, far from it. Lastly, when Saussure defines the sign as the relation of an acoustic image, or signifier, to a concept, or signified, he argues against a “nomenclaturist” conception of the relation between a word and a thing, thereby bringing its representational function under criticism, but not as if languages had ever functioned on any other principle than the one he brings to light. In short, if these semiotic functions serve the archaeological project, it is not because of the possibility of their historical emergence, but because they serve as principles of intelligibility of knowledge in a logic of sedimentation where science is posed as a central element. Such a history could only know three ages, because centrality only supposes a before and an after. The model is therefore quite structural in its concept, although the demonstration contained in Les mots et les choses consists in making the shift of preponderance of epistemes concomitant to cultural changes.

3 | Theory of sign and theory of language

17On the basis of the model of the archaeology of knowledge, which makes epistemes succeed one another on the basis of various conceptions of the sign, we shall look at the change that occurred at the turn of the nineteenth and twentieth centuries between two conceptions of language as a possibility of a succession of doxas such as is likely to construct an archaeology of discourse in social interaction. The parallelism to be made between the two approaches will appear more clearly in the table below, showing also the areas to be questioned for the archeological model to be convincing: the first column presents the form of decentering that communication has imposed on discourse, and the phenomenon which was revealing of this decentering (see 4.2); the last column shows what conception of language (as testified by the revealing phenomenon) succeeds communication (see 4.1).

Object aimed at by the description	Conceptual status of the meaning aimed at in the object	Basis of relationship of meaning	Central and peripheral conceptions of this relationship
An archaeology of knowledge (or: of the human sciences, of structuralism)	in accordance with epistemes	based on sign as	–1) resemblance 0) representation +1) signification
An archaeology of discourse in social interaction (or: of [decentering object?], of [revealing phenomenon?])	in accordance with doxas	based on language as	–1) expression of thought 0) communication +1) ?

18Before turning to these issues, which will be discussed in the next section, it is worth highlighting the differences between the two archaeological projects. There are two main ones.

19The first difference consists in the conceptual status of meaning from which the objects of archaeology are aimed —conceptual status of meaning or, if one prefers, framework of intelligibility according to which archaeology intends to describe the object it gives itself. The episteme is a silent background. It is necessary to bring forth the positive forms in a very particular way so that the negative space of meaning which underlies them may reveal its role as “historical a priori” or as “condition of reality” (Foucault 1969, p. 167). The historical analysis contained in Les mots et les choses opens up the archives of three positive sciences —general grammar, the analysis of wealth, and natural history— in order to show how representation serves as the common epistemic background from which they are thought, developed, and eventually fractured. On the other hand, in order to account for a doxa, the archaeological investigation can be confined to the consultation of dictionaries, because a doxa is known by all —except for the precise definition of what it is, namely a shared and normative space of meaning. It would, of course, be interesting to determine how communication has become a vulgate in the philosophical (more globally, speculative) discourse on language, but this cannot have any direct impact on an archaeology of discourse in social interaction. Not only has communication always been a function related to discourse, but also what ended up imposing it as a norm in social interaction does not only result from a change of ideas on the theme of language; all the technical and social evolution accomplished in the nineteenth century would have to be called upon for its advent. In other words, a doxic archaeology does not have to make the document or the archive the center of its methodological problems. Rather, what is important is the interpretation of the technical transformations, more particularly the media transformations, in the configuration of a society and a culture.

20The second difference could take up a lot of our time, because there is nothing simple about it. It consists in recognizing a difference between the conception of sign and the conception of language. I will try to make this difference a little clearer by commenting on the epistemic projects of two founders of sciences, namely Husserl's phenomenology and Saussure's semiology. Against all expectations, a conception of sign is attributable to Husserl, while Saussure develops a conception of language. This is, in any case, what I intend to argue.

21The coupling of these two differences in the archaeological project allows us to say that language inscribes a certain relation in the doxa of discourse, whereas the sign inscribes a certain relation in the episteme of knowledge. An archaeology consists in holding this relation as central, without which it would not reach its descriptive objective, while also showing other relations, not only possible but effective, historically effective, at the edge of this preeminent relation.

Husserl

22The perspective of the phenomenological analysis undertaken in the Logical Investigations (1901) is that of a criticism of knowledge. It starts with some “linguistic discussions” (1970, p. 165) deemed necessary and from those tries to produce “essential distinctions” (id., p. 183). Expression, a word retained by Husserl in preference to sign, although in certain points of view synonymous, testifies eloquently to the orientation taken by the analysis. The expression can be deployed “in isolated mental life” (ibid.), that is to say that it will allow to put aside all that, in language, does not belong to this activity where the consciousness is sufficient to itself. In the First Investigation, Husserl thus tries, with a remarkable meticulousness, to release “meaning simpliciter” (id., p. 200) by distinguishing it from other aspects of the expression, namely the reference (or “fulfilling sense”), the communication and the expression as such (in its “meaning-intention function”).

23The activity of discernment leads to a stratification which is necessary to the logical construction of a theory of knowledge. Husserl recognizes that language (or discourse) has a communicative function, but the expressions of this language have a role to play “in uncommunicated, interior mental life” (p. 190). One thus observes that between the two functions of language noted by the dictionaries, Husserl establishes a hierarchy: there is no communication possible without expression, whereas expression continues to play a role independently of communication, i.e. in “isolated mental life”.

24Moreover, as Husserl’s definition of expression becomes more precise, it leads to a reduction on the part played by the human being in expressing their thought through language, and also the part played by thought itself. The reduction undertaken by the phenomenological analysis is as follows: from language to sign, then from sign to the intentional meaning such as it is conceived as “ideal essence” extracted from “psychological and grammatical connections” (p. 224). Husserl has thus elaborated a theory of sign, which is devoted to serve as a foundation to science, and notably challenges the medieval conception of the two-sided sign (p. 199).

Saussure

25Semiology apparently presents a theory of sign, since it too can be opposed to the medieval conception of sign as aliquid stat pro aliquo. Indeed, Saussure’s sign is two-sided, but the functional link between the two sides has been broken. At least, between the signifier and the signified, there is no function that can be used in the framework of a theory of knowledge. The arbitrariness of the sign points in some measure to this: the definition of the sign is free of any theorization that would make it an instrument at one’s disposal.

26In fact, semiology offers a theory for language, not for sign. It imparts a principle of intelligibility to the linguistic method by classifying the facts of speech among other semiological facts. As is now well known, the semiological character of language, better than its psychological and sociological characterizations, accounts for what defines the facts of speech as objects of study.

27Moreover, the semiological classification does not serve a theory of knowledge, still less does it found a metaphysics, —even if others, following Saussure, were and still are keen to follow the theory of language in its epistemological and ontological extensions— but it does open up a field of study: semiology will study “the life of signs within society” (Saussure 1959, p. 16). Thus, not only does the theory of language support the study of linguistics by placing the facts of speech in a category where its defining features are made more apparent, and consequently where its description can be more coherent, it also extends the linguistic method, which, starting from general linguistics, proposes itself as a descriptive model of other facts.

28Finally, neither expression of thought nor communication need to be taken into account in the definition of language for the linguistic description, which supposes in any case to suspend any effect of hierarchy between them. The definition of sign, however, imposes the necessity of two speaking subjects, invested with psychological and social connections, as is highlighted by the expression “within society” in the descriptive project attributed to semiology. As for the use of language, it “requires the presence of at least two persons” and the circuit of speech cannot be described as complete without this “minimum number” (id., p. 11).

29To sum up : Husserl takes a single speaking subject as a starting point, and this subject is further reduced to a purely ideal and logical instantiation, while for Saussure the presence of two speaking subjects is the minimum required. Husserl starts from the old doxa of language in order to elaborate a theory of sign by reducing the role of language, whereas Saussure, in his theoretical thinking, enshrines a conception of language in which the communication of two speaking subjects becomes preeminent.

4 | Mediation, as a third conception of language

30Archaeological modelization, as we have seen, leaves open the possibility of a third conception of language. Our commentaries on Husserl's and Saussure's theoretical assertions have made it possible to retain a criterion, namely the number of speaking subjects that they involve, through which the first two conceptions can be opposed. If the expression of thought depends only on one speaking subject, communication, on the other hand, implies two. To consider a third conception, it would be enough to increase this number. In this last section, I would like to argue that mediation, as it tends to spread in ordinary discourse today (at least in French-speaking societies), can serve as an appropriate term for this ‘more-than-two-speaking-subjects’ conception of language.

31Such a hypothesis allows us to complete the last column in our table in the following way:

[An archaeology of discourse in social interaction in accordance with doxas based on language as:] -1) expression of thought: one speaking subject 0) communication : two speaking subjects +1) mediation : more than two speaking subjects

32In the very recent past, there has been an arising of the use of mediation: suddenly, it has appeared everywhere, without it being clear how or why. In the past, the word was used within rather precise frameworks of intelligibility, for example in geometry or in music, the best known framework being that of international law, where mediation was only offered or attempted (what it achieved was called otherwise, agreement or settlement of a dispute). For it to leave these specialized uses, and for it to no longer refer simply to the easily identifiable action of a mediator, mediation had to be elevated to the rank of a concept. Actually, in the philosophical conception and the theorization of human sciences, mediation has been able to abstract itself from its ordinary contexts of application to claim a more general scope. In a socialized discourse, a discourse that does not yet belong to the ordinary language, but that gives voice to a “zeitgeist” propagating in the stratified space of a given society, the extension of mediation is by now acquired, taken up and enshrined by institutions of all sizes and of all kinds, up to the most legitimate ones. It names realities that are not only dependent on conceptions and discourses, and thus contributes to the concrete organization of social life.

33The process by which a signifier acquires a conceptual signified in order to infiltrate the sphere of social discourse and ultimately serve as a referent exposes the very movement of the doxa. What is called “mediation” today covers various social realities in different sectors of activity, notably job search (intercultural mediation, social mediation), corporate work (psycho-social mediation), professional expertise (operational mediation), group therapy (family mediation) or cultural animation (cultural mediation), each of these sectors allowing the identification of a role, generally professionalizing, directly attached to its institution: the “mediator”. Unless one is involved in one of these fields oneself, one will probably have difficulty in seeing precisely what type of job or function is thus configured by mediation. But this is precisely the indication that mediation has become a doxic object: its generalization and propagation make its use legitimate, without it always seeming necessary to question the relevance of its particular applications.

4.1. In social interaction

34We would like, firstly, to show that mediation, as it is being instituted, instantiates a relation to more than two speaking subjects.

35We will start from a (relatively) simple example, that of school mediation. Appearing at the end of the 2000s, in France as well as in Belgium and Quebec, school mediation is presented by its actors as a means of preventing violence in secondary schools. Its purpose is directly linked to educational objectives, since violence, particularly in the form of harassment, is the main cause of student dropout and absenteeism. Such a reality is obviously not new and the school institution has always taken it into account in its mission. Mediation is simply the recent form in which this reality appears by increasing previously unintegrated situations: “discipline” problems today no longer target only the relationship between students and teachers, but also students among themselves as well as students with people outside the school, in particular with family members.

36School mediation is a place for speech and dialogue: speaking out is encouraged and facilitated, dialogue is developed in a configuration of activities in which authority (at least moral authority) nevertheless has a predominant role to play. Compared to other roles linked to discipline (principal, prefect of studies), the role that is designated by the word mediator does not predetermine, or at least much less so, the institutional status of the actor responsible for making this place operational and active. And yet the dialogue instituted in a school mediation does not presuppose equality between the interlocutors. On the contrary, the relationship of enunciation is asymmetrical, as much by the age of the speakers (child/adult), the statutory function (with or without professional function) as the role in the interaction (request and answer or, in terms of service, assisted/assistant). Nor does it presuppose complete freedom between them: the mediator, through the professional function they fulfill and the employer who delegates it to them, remain an institutional representative, constrained in their speech by this very function2. In fact, in a school mediation, subjects other than the interlocutors necessarily express themselves, namely the authorities of the school and, more globally, the “school system”, with its rules, with the laws that institute its mission, with the social norms (citizenship, in particular) that promote it. These subjects are individualized or social, directly identifiable within the institutions involved or assignable because of the responsibilities conferred on them, in all cases they are authentic speaking subjects, because they have the means to make their voice heard. If a mediation were to bring together only a student and a mediator, a third subject would remain presupposed in its very mechanism. This third subject has, moreover, the latitude to speak in turn when the mediator returns, as their function encourages them to do, to their institutional partners or to the parents of a student: whether they report on their work or mentions a problem that they cannot solve alone, or pleads the student’s cause, the mediator is indeed a sort of translator between two subjects who, without their intermediary, seem to have difficulty understanding or even speaking to each other. The mediator is on both sides a relay: in front of the student, they relay an institutional word without, however, giving way entirely to it, because it is their responsibility to maintain a dialogue with the student; in front of the school authorities or the parents, they relay the student’s word by making it audible and intelligible thanks to the status conferred on them. There are thus at least three speaking subjects involved in each speech apparatus provided for in school mediation.

37A second example, less expected than the first one, will allow us to test the stability of the relationship maintained by mediation. Presented by Bénédicte de Villers and Véronique Servais (2016), it relates to therapeutic mediation which is brought by an animal, generally a dog, in the context of hospital care. This type of mediation is sometimes used when the patient refuses care and closes to any communication. The animal is supposed to bring the patient out of their apathy and to make them accept the administration of the planned care. The animal thus acts as a mediator between the patient and the hospital institution. How can this mediation be considered as a speech apparatus? First of all, it leads the patient to express themselves and to be understood, by gestures, affectionate words, of the animal; then, very often the patient projects a series of emotions on this supposedly mute and innocent being, this infans that the patient makes speak through their own voice; finally, the animal obviously did not come alone, a master, that the institution solicited and who is qualified to this service, accompanies them. There is thus a “dispositive dynamic” which prepares for the dialogue in which the patient engages thanks to and with the animal. Within this dynamic, the role of mediator is unceasingly transposed from one actor to the other: from the animal to the patient (mediation of voice) and, even before that, from the institution to the animal and its master (mediation of the care which must accompany the cure). It follows that therapeutic mediation proceeds to a conciliation, provoked not so much by language as by a return to language. The apparatus is thus a speech apparatus, because it is only through it that speech occurs. And, here again, at least three speaking subjects are involved, even if one of them —the animal— may remain silent throughout the process.

38A series of shorter examples will show the impact of mediation in digital media. While mass media (press, radio, television) have been described as communication media, digital media can indeed be considered as mediation media. We will evoke the well-known examples of Wikipedia and social media (considered as a whole) before reporting on the less usual case of political propaganda broadcast on Facebook.

39Wikipedia presents itself as an encyclopedic project. At first glance, nothing distinguishes it from printed encyclopedias, except that it is “free”, which is facilitated by the digital media, without being specific to it (there are paid online encyclopedias, and printed encyclopedias are freely available in libraries). The particularity of Wikipedia lies in the fact that its articles can be written by anyone who wishes to do so; this is its “freedom”: everyone is free to participate, as an anonymous author, in a collaborative editing of the encyclopedia. However, the authors are led and constrained in their writing. If there is no editorial management, strictly speaking, it is nevertheless true that the writers are supposed to subscribe to an editorial protocol, that they are organized in a network and that they exercise a certain control over each other. If, for example, writers of the same article disagree about the information to be given, it is expected that a mediation among senior editors will be set up to resolve the conflict. It is therefore a speech apparatus that no longer corresponds to the author-editor-reader chain, because with Wikipedia any reader can also assume the role of author and editor. The meaning of such an apparatus can only be understood if a community takes part in it. Probably this community was on the horizon of any encyclopaedic project, but Wikipedia, by using the possibilities of digital media, integrates it directly into its realization.

40Social media such as Facebook, Reddit, LinkedIn or Pinterest, among others, are obviously speech apparatuses, designed for the sharing of information and diverse expressions. The written verbal language is next to the forms of communication by pictograms (in particular emoticons) that these media have made available to users and that have soon reached, as a result, a maximum degree of convention. From speech apparatuses, these media have thus passed to the concrete disposition of signs aligned as on a keyboard. But these pictograms are for the most part signs of communitarization: the evaluation, the sharing, the public commentary have no more direct content than that of a performative act of adhesion to a community. The specificity of the mediation carried out by the social media is thus to institute a relationship of language between subjects who exchange less between them than they speak among them.

41Massimo Leone (2017), investigating jihad, has analyzed the impact of such an apparatus when it passes through the hands of a propagandist. The art of persuasion aimed at rallying candidates for jihad does not so much use rhetorical means as it does the very properties of the media: the contents are almost instantly copied and made familiar by the users. Thus, the propagandist technique is no longer that of a revealed secret; on the contrary, it takes on the appearance of a transparent mediation— where each user willingly assumes the role of mediator— and of a consensual and comforting content. It is only in a second moment that the propagandist regains control of the mediation, by cutting the chain of it by the transfer towards a private channel of communication.

42From these examples, we can see that mediation does not simply instantiate a language relationship between three speaking subjects, that is the speaker, the mediator and the listener. One would consider more surely that it mobilizes a community to which the speaking subjects adhere by means of the speech apparatus that it sets up. Mediation is consequently this language approach where the subject speaks before a community, present or represented, as soon as he or she is entitled by this one to express themselves.

43If such a definition confirms the existing link between mediation and language, it leaves however in the shadow what mediation does to language. Actually, it radically changes its conception: by putting at the heart of the process of language a form of dialogue that cannot be summarized as an interlocution between an interchangeable speaker and a listener, it redistributes in depth the roles by which it was customary to instantiate it.

4.2. The critical scope of mediation

44In order to make the archaeological model fully applicable to discourse, we will now ask ourselves what the decentering of discourse through mediation actually implies, and what might have been the revealer of this decentering. The way in which we have described the processes of mediation leads us to plausible answers for these two questions. Decentering consists in the fact that discourse has become inseparable from the scene on which it occurs; this is what we have designated as the “speech apparatus”. Mediation, as a conception of language, emphasizes the praxeological conditions through which discourse manifests itself. Finally, social networks, along with other media apparatuses specific to the Internet (in its “Web 2.0” version as well as in its “Web3D” version), have certainly served as a revealer, not of the existence of these apparatuses, but of their impact on the forms and conditions of speech.

45The first column of the synthesis table can therefore be completed.

Archaeological function of the object	Object of the description
• General object • Decentering object • Revealer	An archeology • of knowledge • of human sciences • of structuralism
• General object • Decentering object • Revealer	An archeology • of discourse in social interaction • of speech apparatuses • of digital apparatuses for social interaction

46One may thus expect that those who have highlighted mediating processes in social interaction have also sought to deconstruct the communication model. Christine Servais (2016) has shown how French philosophers, at the turn of the twenty-first century, have produced a critique of the communicational doxa through thoughts elaborated around the theme of mediation, questioning, among other things, the role of technique, the relation between the speech apparatus and the discursive process, the essential part of the institutional third party, as well as the confrontation in discourse with the otherness of others. Against an irenic conception of language underlying the concepts of expression and communication, these philosophers offer a polemologic conception of language which makes room for the dispute and the misunderstanding.

47This critical discourse does not turn mediation into a standardized philosophical concept. Jean-François Lyotard conceives an immediate mediation, Bernard Stiegler a mediation distinct from any mediatization, Jean-Luc Nancy a mediation without mediator, and Jacques Derrida a mediation more original than the immediate. Servais concludes to the fundamental ambiguity of the theme of mediation in the philosophical discourse. This is precisely a reason for a doxic recovery where the notion is both trivialized and equivocal: any manifestation of speech is likely to pass for mediation. If therefore philosophy uses the notion of mediation with a critical aim, this notion has gained at the same time a doxic scope, that the digital apparatuses of social interaction, in their way, have enshrined.

48In this respect, the parallel to be made with the archaeology of knowledge seems more blurred, since it is not from a single place that, starting from mediation, an epistemic critique, on the one hand, and a conception of language, on the other hand, have developed, whereas structuralism is indeed supposed to have been this critical activity which led the heterogeneity of signification to be interpreted as the generalized malfunction of the episteme of the representation3.

5 | Brief conclusion

49What is at stake in this discussion? Once again, the parallel proposed with the archaeology of knowledge can help us ascertain this. At the beginning of Les Mots et les choses, Foucault develops a commentary on Velasquez’s painting, now known as Las Meninas, in which the power of the king is imposed in the very conception of knowledge; representation is the episteme in which the world is seen as it is: organized around royal power. Similarly, communication is the doxic conception from which the world assembles and thrives: by liberating the world from aristocratic hierarchy, communication promotes the equality of exchanges, following a democratic ideal. As for mediation, it supports a socio-political model that I have proposed, in an earlier article (Badir 2018), to call “ecological.” The double imposition of the recognition of the non-reducible character of otherness, on the one hand, and of the integration of action and interaction in a collective, on the other hand, attenuates the democratic egalitarian and libertarian aim and valorizes the diplomacy between inherently different beings, which corresponds to the very function of mediation.

Notes

• 1 In his article “Structuralism in Modern Linguistics” published in Word in 1945. For a presentation, see Rastier (2018).
• 2 Thus, even if the mediator is a fellow student, as is the practice in Quebec, equality is not assumed, since this would presuppose an equitable distribution of freedom of speech. Moreover, the Quebec school institution has foreseen that two mediators be present in any mediation situation, in order to clearly maintain an asymmetry in the interlocution.
• 3 But isn’t it also because structuralism made signification a new epistemic paradigm that the archaeologist did not make it the last word of his investigation and abjured, shortly after the release of Les Mots et les choses, his adherence to structuralism?