Michel Foucault’s essay “What is an Author?” has been one of his most widely read pieces since its appearance in Josué Harari’s Textual Strategies (Cornell, 1979), and it was as such that I recently encountered it in my department’s introductory literary theory course. Through what in another vocabulary we might call its “deconstruction” of the concept of “author,” this essay signals a sea-change in the way in which we read and interpret texts. Roughly speaking, instead of understanding the text as an emanation of the author’s unitary intention, we are to situate it in the context of the “discursive practices” of its time, which allow certain things to be said, certain “subject-positions” to be maintained, among which is the historically determined “author-function.” Foucault, unlike Derrida, does not seek to tease out of the self-contradictions of these practices the paradoxical cultural mechanism that lies behind them; he is not a fundamental anthropologist. His intention is less to propose a new way of reading than to articulate an epochal change that is already taking place. But lest this reference to intention might appear to challenge his thesis–or to reveal the naivete of my own habits of reading–I will not pursue this line of argument. Surely a theoretical metatext need not follow the same rules, or be read by the same method, as the texts to whose reading it refers. The question as to whether Foucault is indeed the “author” of this essay may be dismissed as perverse, since his text expounds a clear, logical idea that may be read like any discourse on method.

Or does it?

True to his historical method, Foucault notes the dichotomization of the unitary notion of authorship as authority in early modern Europe. In the place of the old indiscriminate guarantee given by the author as “father” of his text, different criteria are henceforth appropriate to different families of texts. On the one hand, scientific discourse stands or falls by its internal logic; on the other, literary discourse bears the imprint not merely of authority but of the author’s unique creative intuition. We should note that Foucault’s exclusively discursive orientation leads him to omit from consideration the most fundamental category of textual authority: that of the witness, whose authority comes not from subjective but from objective experience. Memoirs of people involved in historically important or simply unusual events always find an audience. These authors have authority because they were where we were not; however banal the form of their texts, their content prevents them from being reduced to the mere intertextual play of discourses. Foucault’s notion of “discourse” tends to make us forget that discourses not merely take up positions but also convey information.

The Greeks too had known texts sustained by logic alone; the proofs in Euclid‘s Elements did not depend on their “father”‘s authority. But in modern scientific discourse, not merely the constructions of mathematics and logic but the real world became subject to the rigor of the scientific method. Consequently, texts that did not follow this method had to be guaranteed not by their author’s mere authority but by his personal vision. Instead of being transmitted “authoritatively” from God to the reader by the intermediary of the author, the text became the reflection of a personal revelation that he alone was able to put into words.

It is this personalized view of authorship that flourished in the Romantic era and that Foucault sees today as coming to an end. The author remains, to be sure, an indispensible legal persona who owns the economic rights to his text and stands responsible for it before the law. What is new is that this external relationship of possession is no longer deemed to correspond unproblematically to the authorial subject position within the text. Whatever the author’s legal status as its creator, and however the “implied author” may trumpet his authorial identity (today, such trumpeting would probably be perceived as ironic), the text is understood not as a unitary emanation of a unique intention, but as an “intertextual” matrix of pre-existent discourses woven together, to use the familiar text = tissue metaphor, according to discursive practices over whose overt and implicit rules no individual has more than marginal control.

What is at stake here is but another moment of the eternal dialectic between the world of signs and the world of things, les mots et les choses. Foucault’s argument would no doubt have been clearer had he posed the problem in originary terms. Instead, his text is inextricably both a revelation about sign-systems in general and a revelation about our historical self-consciousness of these systems. Its factual, indeed, authoritative tone–no doubt the secret of its popularity–implies an irreversible passage from illusion to reality: the romantic notion of the author is a myth and our current anti-authorialism, a demystifying step toward the truth. Yet at the same time, the periodization of “épistémès” in the absence of any transhistorical truth-criterion implies that we are simply witnessing the passage from one arbitrary mode to another. The “death of the author” is one more revelation that the sign never reveals the thing, the map is never the territory; but this very disillusion is itself no revelation of ultimate truth, merely another illusion like the others. This paradoxical mode of thought is far more clearly articulated by Derrida, who always remains in the margins of the same phallogo-metaphysical center, albeit at the cost of Foucault’s attention to historical specificity. Only GA, it seems, is able to synthesize the universal and the historical by beginning at the minimal origin of both.


Should Foucault’s article then be read, in the mode of de Man‘s “blindness and insight,” as exemplifying the text’s blindness to its own laws of composition, as failing to situate its own revelations within the conceptual framework it establishes? But as de Man was wont to say of “texts,” as opposed to mere chains of arguments, Foucault’s essay contains within itself the sign of its own author’s ironic understanding of the paradox implicit in its composition.

In the final section of his essay, Foucault introduces a new category of author, neither the impersonal author of a scientific discourse, who holds no authority outside of the logic and empirical accuracy of the text itself, nor the “personal” author of a literary discourse, whose only authority is his unique subjective experience of reality. This new category is that of the “founders of discursivity”–the French fondateurs de discursivité is no more euphonious. Foucault considers only two thinkers worthy of this designation: Marx and Freud. Not only have these writers created new “paradigms” but, as is not the case for the creators of Thomas Kuhn’s “scientific revolutions,” their followers are forced to continually “return to the origin” to reevaluate their own founding works. A new development of psychoanalysis is not merely a new development of ideas originated by Freud–it implies and requires a new reading of Freud’s works themselves. And (to a lesser extent today than in 1979, when massive state apparatuses were still devoted to their study) this is also true of the works of Marx.

Constant rereading and reinterpreting were required by the old “authorities” as well, both literary (Homer, Shakespeare…) and philosophical (Plato, Hegel…). What is new is the scientific pretension of these new “discursivities,” which purport, to paraphrase Marx’s own famous words, not merely to speculate about the world, but to (permit us to) change it. The modern “discursivities” are, as Karl Popper and his disciples have pointed out, invulnerable to “falsification” but, unlike religions, to which they are often disparagingly compared, they provide reasonably coherent explanations of human experience without recourse to transcendental entities. In a word, they are anthropologies.

Foucault presents this new category in a matter-of-fact manner, as though merely acquainting us with one more historical wrinkle in the concept of authorship, and then goes on to reach thisNathalie Sarraute-like conclusion:

We would no longer hear the questions that have been rehashed for so long: “Who really spoke? […] With what authenticity or originality?” […] Instead, there would be other questions, like these: “What are the modes of existence of this discourse? Where has it been used, how can it circulate, and who can appropriate it for himself? […] And behind all these questions, we would hear hardly anything but the stirring of an indifference: “What difference does it make who is speaking?” (p. 160)

But within the category of “founders of discursivity,” it certainly does make a difference who is speaking. Marx and Freud are only two exceptions, but they are human, not divine; if they can “found discursivities,” then so can others. If a given thinker’s work is not a “discursivity,” is this inherent in the work itself, or an artifact of its reception? The thinker in question can only keep developing his ideas in hope that others will find it worth their while to prolong them. As he continues to write, he conducts himself as though he too were a “founder of discursivity.”


It is well and good that we henceforth be expected to read discourses without recourse to the naive notion of authorship as existing equivalently and unproblematically on both sides of the barrier between reality and representation. The textual paradigm in which the author is “the same” within and outside the text is no longer acceptable. But in the case of “founders of discursivity,” what exactly does it mean that we must return to their works as the ultimate ground of each new theoretical development? What is it in “their” discourse that is not replaceable by any combination of other discourses? These are not poets whose esthetic intuition brings together unforgettable combinations of signifiers.

It is clearly their impact not just on academic thought but on thought in general that caused Foucault to single out Marx and Freud. Their texts are traces by means of which we witness, not a historical event, but an intellectual revelation. Heidegger, Hegel, even Nietzsche, whatever their influence on professional and semi-professional intellectuals, have not deeply touched the general population and have not, therefore, transformed beyond recognition the way we think about ourselves. Darwin, however controversial he remains, founded not a “discursivity” but a branch of the natural sciences.

The intellectual revelation, like the religious revelation, is important not in its content alone. In either case, the vision revealed by the founder has an irreversible effect on our self-conception–our anthropology. It is a unique experience of thought, not because it reflects a subjective intuition that cannot be duplicated, but because it has so affected the categories by which we think that we can no longer put ourselves in the place of the founder as he wrote his text. His words are endowed with special prestige because they not merely reflect but themselves constitute a historical event; as we read them, we experience the founding link between the world of ideas and the world of reality. As in the originary scene, the sign is the “trace” of the event, but this trace is itself the event.

The “discursivities” are not bound by the regional limits of the human sciences; their generativity is that of the scene of representation itself. Both Marx’s contrast between infra- and superstructure and Freud’s more subtle and convoluted opposition between the conscious and the unconscious purport to reveal to us the hidden, originary nature of our relationship to the objects of our (material or sexual) desire. The founder’s intuition is that of a fundamental anthropology, a conception of the human. If every writer is a potential “founder of discursivity,” then no writer’s work can be reduced to the mere interplay of discourses held together by an “author-function.”


This does not imply that we should forget all we have learned about the separation between words and things and go back to reading texts as emanations of a transcendental authorial intention. But the fact remains that, at the origin of the discourses or “discursivities” that we really care about, we seek, despite all caveats, something very much like this intention–a moment of equivalence between experience and signification, between word and imaginary thing, that we cannot reproduce without returning to the event in which it originated. We remain curious of authors’ lives because we are grateful for any clue that can relate worldly experience to the construction of a new cultural paradigm. The ostensive human presence that we expel from discourse as écriture returns as the experience of foundation. Conversely, the more we see a text as a mere tissue of preexistent discourses, the less it means to us, and the more we are led to deconstruct it to the point at which we encounter the founding intuitions of the discourses from which its different strands are derived. At this point, we become once again witnesses to the trace of a human experience that we respect enough to seek the reality behind it.

But we do nothing else even in our everyday communications. Our most banal words attempt to convey an experience of emerging signification. Believers and unbelievers alike bear witness to the faith inherited from our originary ancestors that all revelatory experiences are mutually communicable, that all renew the pacifying effect of the originary sign. However incommensurable the “author-function” may be from any worldly interaction, both author and reader know each other to be, not “functions,” but human beings. What we are primarily linked to and separated from by the deferring mediation exercised by all signifying practices is not “language” or even God, but each other.