How Can a Commentary Say Something New? Foucault, Derrida, and the Supplement

 1. What Is a Commentary?

How can a commentary say something new?

The question seems paradoxical. A commentary comes after something has already been said, and its relation to that earlier discourse is what makes it a commentary. Yet in The Order of Discourse, Foucault describes commentary as a procedure that makes possible the production of new statements. The commentator can say something that has not been said before, but does so by returning to a text that has already been established.

This gives commentary a peculiar structure. It depends simultaneously on repetition and novelty. The earlier discourse provides the point of reference that allows the new statement to be recognized as a commentary, while the commentary itself extends, interprets, and potentially alters what can be said about that discourse. Its novelty therefore does not come from leaving the earlier text behind. It comes from returning to it.

Foucault's own opening gesture in The Order of Discourse already anticipates this problem. He imagines that he might enter a discourse that was already underway rather than having to begin speaking himself. The desire is to find himself "carried along" by a discourse that precedes him. Speech, in this image, does not begin with the individual speaker. The speaker enters a field of statements that is already there.

Commentary gives this situation a more specific form. To comment is to enter an existing discourse and produce another statement within its horizon. The commentator does not simply repeat what has been said, because commentary would then have no productive function. Yet the new statement cannot simply detach itself from what came before, because its status as commentary depends precisely on that relation.

The result is a peculiar form of repetition: the commentary returns to the earlier text in order to make something else possible.

This is why Foucault can describe commentary as a means of saying "what has been said" while nevertheless producing something new. The earlier text becomes a point of return rather than a point at which discourse comes to an end. What has already been said becomes the material from which further statements can be produced.

The important question, then, is not whether commentary repeats or adds. It does both. The interesting question is how these two operations are connected.

2. What the Supplement Makes Visible

Derrida's notion of the supplement offers a useful way of seeing this connection.

A supplement appears to be something added to an already existing thing. It comes afterward and therefore seems secondary. If something requires a supplement, however, its apparent completeness begins to look less secure. The addition can reveal a dependency that was not visible when the original was considered on its own.

Derrida's analysis of the supplement in Of Grammatology, particularly in his reading of Rousseau, concerns the relation between what is taken to be original and what is treated as an addition. Writing, for example, is assigned a secondary position in relation to speech: speech appears primary, while writing is introduced as a supplement to something that already exists. Derrida's analysis complicates this hierarchy by asking what the need for a supplement tells us about the supposed completeness of the original.

The value of the concept here is not that commentary and supplementation are identical. It is that the supplement gives us a vocabulary for thinking about the peculiar status of what comes afterward. Something can arrive later and yet affect how what came before is understood. What appears secondary can reveal something about the structure of what appeared primary.

This helps illuminate the paradox of commentary.

A commentary is chronologically secondary. The text comes first, the commentary afterward. But the commentary can make something visible in the earlier text that was not previously articulated. It can establish a new relation among passages, introduce a distinction, or bring a previously unnoticed implication into view. Once that has happened, subsequent readers may encounter the earlier text differently.

The later statement has therefore become part of the history through which the earlier statement is understood.

This does not mean that the commentary changes the earlier text as a material object. Descartes's Meditations, for example, remains the same text whether or not Foucault or Derrida reads it. What changes is the field of interpretation in which that text subsequently appears. The commentary adds something to the history of the text, and that addition can change what readers are able to see in what preceded it.

The supplement makes this possibility easier to grasp because it draws attention to a structure in which the secondary element does not simply follow the primary one. Its arrival can disclose something about the relation that was already there.

Commentary can therefore be understood as a form of productive afterlife. What comes later remains dependent on what came before, but what comes later can also transform the possibilities of reading what came before.

3. Commentary in Practice: Foucault, Derrida, and Descartes

The exchange between Foucault and Derrida provides a particularly clear example of this structure.

In History of Madness, Foucault interprets Descartes's First Meditation as part of the history through which madness is excluded from the domain of reason. Derrida later returns to this interpretation in "Cogito and the History of Madness." The important point here is not to reconstruct the entire disagreement over Descartes. What matters for the present argument is the form of Derrida's intervention.

Derrida does not approach Descartes as though Foucault had never read him. He approaches Descartes through Foucault's reading. The earlier interpretation therefore becomes the point of departure for another reading.

The sequence is not simply: Descartes → Foucault → Derrida.

It is:

Descartes → Foucault's reading of Descartes → Derrida's reading of Foucault's reading.

That difference matters. Derrida's intervention returns to a text that Foucault has already interpreted, but the return does not merely reproduce Foucault's interpretation. It raises questions about how Foucault has identified the relevant moment in Descartes's argument and about what follows from that identification.

The later reading consequently sends us back to the earlier text.

This is commentary in a particularly revealing form. A text has already been interpreted, yet the interpretation itself becomes something that can be interpreted. The new reading does not simply add another layer to a fixed object. It changes the route by which the earlier object is encountered.

The same structure can be seen in the subsequent history of the exchange. Foucault's interpretation becomes an object of Derrida's commentary; Derrida's intervention can then become an object of further commentary, including Foucault's later responses. What initially appeared to be a linear sequence becomes a chain of returns:

text → commentary → commentary on commentary → further commentary.

At each stage, something new can be said because something has already been said. The earlier statement provides the material and authority for the later one, while the later one alters the interpretive situation in which the earlier one is read.

The example therefore gives a concrete form to the problem Foucault identifies theoretically. Commentary does not simply prolong discourse after its completion. It creates a way for an existing discourse to generate further statements without ceasing to be recognizable as the discourse to which those statements belong.

The supplement helps make the same structure visible from another direction. What comes later can remain dependent on what precedes it while nevertheless changing the significance of that precedence.

4. The Commentary That Never Ends

Once commentary is understood in this way, its relation to the origin becomes clearer.

Chronology still matters: one text can precede another, and an author can write before a commentator. But chronological priority does not determine the final significance of what comes first. An earlier text can remain the point of departure for interpretations that it could not itself have anticipated.

This is why the opening image of The Order of Discourse is so suggestive. Foucault imagines entering a discourse that is already speaking. The speaker does not occupy an absolute beginning. He takes up words that have already circulated and places them in a new configuration.

Commentary formalizes this situation. It is a way of entering an existing discourse without simply repeating it. The commentator inherits a text, but inheritance is not passive preservation. To repeat a discourse in the form of commentary is to place it in another context, establish new relations within it, and make further statements possible.

Derrida's supplement helps us understand why this process can have consequences for what came first. The later element may appear secondary precisely because the earlier element is assumed to be complete without it. Yet the addition can reveal that the earlier text had possibilities that had not been exhausted by its previous readings.

The commentary therefore occupies an unusual position. It is secondary in sequence but potentially transformative in interpretation. It comes after the text, yet it can alter the conditions under which the text is subsequently encountered.

And because the commentary itself can become the object of another commentary, there is no necessary final point at which the process must stop. The commentator's statement becomes someone else's text; what was once a commentary becomes an object of interpretation; what appeared secondary becomes part of what a later reader must now inherit.

The question with which we began can therefore be answered in a more precise way. A commentary can say something new because its relation to what has already been said is productive rather than merely repetitive. It returns to an earlier discourse, but in returning it opens that discourse to another formulation.

The supplement gives us a way to see why this matters. What comes afterward may remain dependent on what came before while nevertheless revealing something that could not be seen in quite the same way before the addition was made.

The history of discourse is consequently not simply a sequence in which one text follows another. It is also a history of returns. What has been said can be said again, but the return may change what can be heard in it.

The origin remains first in time. It need not, however, remain final in meaning.

References

Derrida, J. (1978). Cogito and the history of madness. In A. Bass (Trans.), Writing and difference (pp. 31–63). University of Chicago Press. (Original work published 1967)

Derrida, J. (1997). Of grammatology (Corrected ed., G. C. Spivak, Trans.). Johns Hopkins University Press. (Original work published 1967)

Foucault, M. (1972). The archaeology of knowledge and the discourse on language (A. M. Sheridan Smith, Trans.). Pantheon Books. (Original work published 1969)

Foucault, M. (1978). History of madness (J. Khalfa, Ed.; J. Murphy & J. Khalfa, Trans.). Routledge. (Original work published 1961)

Foucault, M. (1981). The order of discourse. In R. Young (Ed.), Untying the text: A post-structuralist reader (pp. 48–78). Routledge & Kegan Paul. (Original work published 1971)