The reader who defends Descartes against Foucault, and then asks what the certainty of the instant would need in order to found anything
You asked two things. First, what Derrida’s critique of Descartes was, in brief points. Then, because the first answer did not show it, where the critique starts and how it is built, “like a→b; b→c.” The chat’s answer to the second question is the core of this page. Derrida’s starting point is not the validity of the cogito but the requirement of foundation. He grants Descartes the instantaneous certainty and asks what it would take for that certainty to found anything. The chain then runs: pure self-presence, repeatability, iterability, the “I am” as already a sign, the divided instant, and self-presence as an effect of différance.
Two registers, which the chat keeps apart. In “Cogito and the History of Madness” (1963) Derrida’s target is Foucault, and he defends Descartes: the First Meditation does not expel madness, the demon surpasses it, and the cogito holds even if I am mad. The real critique comes later, in Speech and Phenomena and Of Grammatology (1967), where the cogito is the paradigm of the metaphysics of presence.
What this review adds. The chat gave no page references, and Derrida’s own books are not in your folder. So this page anchors every step of the chain in a passage of Descartes that is held and checked. It also quotes Derrida’s key sentence, “I am originally means I am mortal,” from the one book in the folder that reproduces it with a page number. And it follows the chat’s closing observation that Derrida’s leverage falls exactly on the seam of your two-tier reading (Chat 5). Everything that rests on Derrida’s unheld texts is marked.
“What is the starting point? Explain how he builds the argument, like a→b; b→c.”
The question asked is structural: not what Derrida concludes, but the order of his steps and the premise from which they begin.
The question refused is whether the cogito is a valid inference. The chat is explicit that Derrida “never says the inference fails. It’s quasi-transcendental: the condition of possibility (repeatability) is simultaneously the condition of impossibility.” That is why this critique sits so awkwardly beside the nine objection-types of Chat 1. It does not deny any premise. It asks what the premises must presuppose in order to do their job.
The conversation is the spine. The held works below supply the text of Descartes on which each step of the chain turns, and the secondary accounts that fix the Foucault–Derrida exchange.
| Work | What it contributes here | Enters at |
|---|---|---|
| Chat, “Derrida’s critique of Descartes” (DEscartes project, 27 July 2026): two answers | The three registers; the chain A→F; the tie to Hintikka and to the two-tier reading. | throughout |
| Descartes, Meditations I–III and V, and The Search for Truth, in CSM vol. 2 | The madness passage (2:13); “whenever it is put forward” (2:17); “for as long as I am thinking” (2:18); the divisible lifespan (2:33); the proof that slips away when attention turns (2:48); the dreaming substitute (2:407). | §§01, 05, 08 |
| André Gombay, Descartes (Blackwell Great Minds, 2007), p. 26 and nn. 15–17, p. 138 | A compact, neutral account of Foucault’s remark, Derrida’s “minimalist” reply, and Foucault’s rejoinder, with bibliography. | §02 |
| Harry Frankfurt, Demons, Dreamers, and Madmen (1970; 2007), ch. 1 | Reads the First Meditation as a dialogue between a naive empiricist and his critic. Derrida’s reading of the madness passage takes the same shape. | §01 |
| David E. Johnson, Kant’s Dog: On Borges, Philosophy, and the Time of Translation (SUNY, 2012), p. 202 | Quotes Speech and Phenomena (1973: 54) for “I am originally means I am mortal.” | §05 |
| Jaakko Hintikka, “Cogito, Ergo Sum: Inference or Performance?” Phil. Review 71 (1962), 3–32 | The independent “performatory” reading the chat says Derrida lands near. | §07 |
| Jean Wahl, The Idea of the Instant in Descartes’s Philosophy (1920; ed. and trans. Schrift, Edinburgh, 2026), with essays by Frédéric Worms | The instant reading forty years before either: the cogito “not … an argument” but “the affirmation of an instantaneous certainty.” | §07 |
Read this once; everything after it is elaboration.
| Register | Text | What is claimed | Relation to Descartes |
|---|---|---|---|
| 1 · Madnessagainst Foucault | “Cogito and the History of Madness” (lecture 1963; L’Écriture et la différence, 1967) | The First Meditation does not exclude madness. The demon is a “total madness” beyond the empirical kind, and the cogito holds even if I am mad. | Defends him. Exclusion comes only later, when the cogito must be said, remembered and guaranteed. |
| 2 · Presencethe real critique | Speech and Phenomena; Of Grammatology (1967) | The cogito is the paradigm of the metaphysics of presence. Self-presence presupposes repetition, iterability and the trace. | Deconstructs him. The ego is not an origin but an effect. |
| 3 · The animal | The Animal That Therefore I Am (2006) | Cartesian subjectivity is built on a sacrificial exclusion of the animal: the bête-machine, reaction versus response. | Genealogical. An exclusion at the origin, after all. |
Derrida grants the certainty and attacks the foundation. In the chat’s words: “Derrida grants Descartes the instantaneous certainty (which is why he lands near Hintikka). His question is different: what must be the case for that certainty to function as a foundation?”
That makes Derrida’s critique the mirror image of Huet’s in Chat 1. Huet says the cogito takes time, so memory creeps in and the certainty fails. Derrida says the certainty holds in the instant, but an instant founds nothing, and whatever makes it last is already a sign. Both locate the trouble at the join between the moment and its retention. Huet treats the join as a defect of the proof; Derrida treats it as the structure of any self at all.
The chat’s first answer: the 1963 essay, where the target is Foucault and Descartes is defended.
Descartes does not exclude madness; he radicalises it. The exclusion Foucault saw comes later, when the cogito has to be kept.
Before the critique there is a defence. Derrida’s first engagement with Descartes is an argument against his teacher Foucault about two sentences of the First Meditation, and in it Descartes comes out more radical, not more repressive.
The defence matters for what follows. The later critique keeps the defence’s central result, that the cogito holds in the instant whatever my condition, and relocates the problem to what comes after the instant.
The sentences at issue come just before the dreaming argument. The meditator considers whether he might be like the insane, and appears to dismiss the thought at once:
“Unless perhaps I were to liken myself to madmen, whose brains are so damaged by the persistent vapours of melancholia that they firmly maintain they are kings when they are paupers … or that their heads are made of earthenware, or that they are pumpkins, or made of glass. But such people are insane, and I would be thought equally mad if I took anything from them as a model for myself. A brilliant piece of reasoning! As if I were not a man who sleeps at night…”
AT 7:18–19 · CSM 2:13 checked
“A brilliant piece of reasoning!” is ironic. The voice that dismisses madness is itself mocked in the next breath, and dreaming is brought in to do the same work more thoroughly. Frankfurt’s translation has “How eminently reasonable!” (Frankfurt p. 14 checked). His whole reading treats the First Meditation as a dialogue between a naive empiricist and his critic, the dialogic structure Derrida’s reading of the passage also depends on.
Descartes’ unfinished dialogue, The Search for Truth, stages the same substitution and gives the reason for it:
“…ask if you have never seen one of those melancholic individuals who think themselves to be vases … To be sure, a good man would be indignant if you told him that his beliefs cannot have any more rational basis than theirs … But you cannot take it amiss if I ask whether you are not, like all men, liable to fall asleep…”
AT 10:511 · CSM 2:407–8 checked
The motive is rhetorical, not metaphysical: madness would offend, and dreaming is universal. Gombay cites this passage for exactly that point (next section).
The chat compresses the quarrel into one line: “Foucault claimed the First Meditation expels madness (the glass-bodied man) in a founding gesture of modern reason. Derrida denies this.” Gombay’s account, which is held, fills it in:
In 1961, in his Histoire de la folie à l’âge classique, Michel Foucault adduced the curt dismissal by Descartes of the madness-hypothesis as further of evidence of a phenomenon which he argued was one of the striking events of the early seventeenth century – “le Grand Renfermement,” “the Great Confinement”… Gombay, Descartes p. 26 checked
According to Derrida, Descartes’ abandonment of the hypothesis that he might be insane is a move of no philosophic significance: the hypothesis is given up simply because the supposition that he might be dreaming is a more convenient substitute; more convenient, in that dreams, unlike madness, are a condition with which everyone is familiar… Gombay p. 26 checked
Foucault replied in “My body, this paper, this fire,” added to the second edition of Histoire de la folie (1972, pp. 583–603; English in Essential Works vol. 2, pp. 393–417), as Gombay records n. 17, p. 138 checked. His argument was that, for a seventeenth-century reader, even supposing oneself mad would disqualify the philosopher altogether. Gombay’s verdict on the quarrel is sober: “Given the sparse and sybilline text, there is faint hope for resolution” p. 26.
Gombay presents Derrida’s reply as deflationary: madness is dropped merely because dreaming is more convenient. The chat presents it as the opposite: the demon “surpasses madness,” hyperbolic doubt is “a ‘total madness’,” and Descartes “internalises madness rather than excluding it.” Both halves are in Derrida’s essay. The dismissal of the madman is, for him, a naive objector’s move that the meditator overrides, and the hyperbole of the demon then goes further than madness ever could. Gombay reports the first half; the chat stresses the second. Neither is a misreading, but they are not the same reading.
The chat’s summary of the 1963 essay has five points. The third is the one the later critique inherits:
The third and fourth points reappear, transformed, as steps B and D of the later chain. The instant is invulnerable. Everything that makes it usable (saying it, keeping it, guaranteeing it) is where the trouble enters. not held The formula “whether I am mad or not” is the chat’s paraphrase of Derrida; the essay itself could not be checked.
The chat’s second answer, set out as the chain you asked for. Each step is paired with the passage of Descartes it turns on.
Six steps, from “the cogito must be self-certifying” to “self-presence is an effect.” No premise of the cogito is denied. What is denied is that the cogito can be both instantaneous and foundational.
“I didn’t get Derrida’s critique. What is the starting point?” The chat’s answer begins by removing a misunderstanding: Derrida is not arguing that the cogito is invalid.
Every objection in Chat 1’s Type 5 asks whether the cogito survives the doubt. Derrida’s question comes one step later. Grant that it survives: can a certainty of that kind do the job Descartes gives it, the job of the Archimedean point from which the sciences are rebuilt?
Descartes himself states both the invulnerability and its limit in the same breath. The certainty is tied to the act, and it lasts exactly as long as the act does:
“…this proposition, I am, I exist, is necessarily true whenever it is put forward by me or conceived in my mind.”
AT 7:25 · CSM 2:17 checked — Latin quoties a me profertur, vel mente concipitur; the chat’s “as often as it is conceived” renders quoties more literally than CSM’s “whenever”
“I am, I exist — that is certain. But for how long? For as long as I am thinking. For it could be that were I totally to cease from thinking, I should totally cease to exist.”
AT 7:27 · CSM 2:18 checked
“For how long?” is Descartes’ question before it is Derrida’s. The first certainty comes with an expiry clause.
The chat’s six steps are given verbatim in the left column. The right column adds the passage each step leans on: Descartes where the premise is his, Derrida where the book that quotes him is held.
“This isn’t an external refutation — he never says the inference fails. … The system’s own founding requirement demands the very thing it must exclude.” Repeatability is what makes the cogito a foundation, and repeatability is what lets absence, delay and the sign into it.
“In hearing-oneself-speak, the signifier seems to evaporate at the instant of signifying — the illusion of unmediated auto-affection. Speech gets privileged, writing (the visible trace) repressed. The cogito is the philosophical crown of that acoustic illusion.” not held
The chat’s two asides: Derrida and Hintikka, and Derrida and your two-tier reading.
Three readers independently make the cogito instantaneous: Wahl (1920), Hintikka (1962) and Derrida (1963–67). Derrida alone asks what the instant costs, and his answer lands on the seam between the tiers.
“Where this bites your two-tier reading: Derrida’s leverage is precisely the seam.” The chat’s last paragraph connects this conversation to the rest of the project.
The chat ends its first answer with a footnote: “Derrida’s ‘valid at the instant it is thought or uttered’ reading lands remarkably close to Hintikka’s performative interpretation, published a year earlier and entirely independently.” The dates bear this out: Hintikka’s article is in the January 1962 Philosophical Review, and Derrida’s lecture was given in March 1963.
…the inconsistency (absurdity) of an existentially inconsistent statement can in a sense be said to be of performatory (performative) character. It depends on an act or “performance,” namely on a certain person’s (here, a’s) uttering the sentence… Hintikka, “Cogito, Ergo Sum: Inference or Performance?” p. 12 checked
There is a third, older member of the family, and your folder holds it. Jean Wahl’s 1920 thesis, newly translated, reads the cogito as instantaneous forty years before either:
‘The Cogito ergo sum is not, therefore, an argument’ but ‘the affirmation of an instantaneous certainty’ or even ‘a judgement, an argument [but] grasped in an instant’. Worms, “From One Instant to Another,” quoting Wahl, in Wahl, The Idea of the Instant in Descartes’s Philosophy p. 12 checked
It is, of course, the instant of the ‘cogito’: ‘I think, I am’, which Descartes himself says is only true ‘as long as’ I am thinking. Worms, “The Vital Instant,” ibid. p. 2 checked
For Wahl, the instant is the motif of Descartes’ whole “metaphysical experience” (Worms, p. 13). For Hintikka, it is what lets the certainty escape the charge of inference (Chat 4). For Derrida, it is what the certainty cannot outlast without becoming a sign. Worms’s own essay opens by tying the reality of the instant to “the experience of death” (p. 1), the very link Derrida draws with “I am mortal.”
The chat’s closing paragraph, quoted in full because it is the bridge to Chat 5:
Derrida’s leverage is precisely the seam. Tier one — the punctual intuition — he concedes. Tier two — the retained, general, propositional knowledge — is where Descartes must import the divine veracity to secure memory. Derrida’s claim is that the trace is already operating at tier one, which is why tier two ever became necessary. The guarantor is a symptom of a structural gap, not a repair of a contingent weakness. The chat’s second answer, final paragraph
The chat gives the late work a single line: The Animal That Therefore I Am attacks the bête-machine and the distinction between reaction and response, arguing that Cartesian subjectivity is constituted by “a sacrificial exclusion of the animal (‘carno-phallogocentrism’).” not held The page records it for completeness. Its bearing on the cogito is genealogical rather than logical: it answers the 1963 essay’s claim that the exclusion was not at the origin by finding one there after all, in another place.
What was checked, what could not be, and where the thread goes next.
The chain is sound as a reconstruction. Its Cartesian premises are all on the page in CSM. Its Derridean steps rest on books not held, except the key sentence of step D.
Derrida’s critique cannot be answered the way the objections of Chat 1 are answered, by defending a premise. It concedes the premises. What it contests is whether an instant can be a beginning.
1 · “Minimalist” versus “total madness.” Gombay and the chat characterise the 1963 reply differently (§02). A reader citing either should cite the essay itself, Writing and Difference, trans. Bass, pp. 31–63 (Routledge 2001), as Gombay gives it; Griffith’s bibliography, also held, gives pp. 36–63 of the Chicago 1978 edition.
2 · The chain is about the instant, not about the demon. Nothing in steps A–F depends on the demon hypothesis. It would apply to any self-certifying instant. That is its strength as philosophy and its weakness as Descartes scholarship: it cannot say why the Meditations needed the demon at all.
Derrida does not deny that the cogito is certain. He denies that a certainty confined to the instant can be a foundation without already being the kind of repeatable sign the doubt was meant to exclude.
The next three chats take up the instant from the inside: Chat 3 asks exactly how the demon is kept out of it; Chat 4 examines Hintikka’s performance; Chat 5 examines the seam this conversation ends on.
Click a node to collapse or expand its branch. Drag to pan; hold Ctrl or ⌘ and scroll to zoom.
The interactive argument map is available only in the web version; it reproduces, as a collapsible tree, the structure of the sections above.