For Descartes, is madness fundamentally different to dreaming?
“Descartes replies by quoting the case of dreams which produce eccentricities as great as those of madness, but to which we are all, each one of us, exposed,” Foucault correctly summarizes Derrida's argument.
Foucault responds first absurdly, suggesting that somehow this objection proves Foucault's own point. It does not, and Foucault makes no argument to suggest that it does.
For Foucault, Descartes claims madness is not the same as dreaming because not all of us can claim we may be mad, we are not all mad (perhaps shorn up with the dis-analogy that mad people interact with actual other people, unlike dreamers).
Was Descartes rhetorically claiming not to be mad?
Descartes.. may doubt his senses, he may doubt his body, but he never takes up madness as his own subjective position
Claims Foucault. Derrida replied
Descartes [says]... since I am here, writing, and you understand me, I am not mad, nor are you, and we are all sane... he feigns to rest in this natural comfort.. to unsettle himself from it and to discomfort his interlocutor.
The blog claims that Derrida conclusively disproves that
Descartes considered mad hallucinations about having a pumpkin for a head as something separate, something fundamentally different, from the dreamer's hallucinations that he is not asleep in his bed but sitting his drawing room... this absurd position about madness
But that conclusion only works, I think, if Descartes was not really - only rhetorically - rejecting madness and saying he wasn't mad (unless he did successfully account for all knowledge, not just his own, which would be a very "totalitarian" claim).
“Philosophy presents itself as the universal criticism of all knowledge (the first postulate), without any real analysis of the content or the forms of this knowledge,” Foucault says [in claiming that it is totalitarian]... [but in fact] Descartes' meditation... is nothing but an analysis of the form knowledge takes.
For what it's worth, I don't think dreaming is analogous to madness here, that while mad people may - as much as anyone - know with certainty they exist, if I am mad then someone else is sane, unlike dreaming, and this changes the nature of doubt, whether or not a mad person could have written Descartes' meditations.