Don't Walk Away, Renê
i.
In his First Meditation, Descartes sets about to reject, provisionally, anything about which he can entertain any doubt, so as to discover whether there is anything certain that could serve as a foundation for other knowledge. He does this by assuming for the sake of argument that everything he is experiencing is an illusion; that in reality he is not seated by a fire surrounded by familiar objects; that, furthermore, he does not even have a body with sense organs, limbs, and so on; even that there is no sky or earth to support and contain other things; and that there is, in general, not even such a thing as an extended body that takes up space and has a shape, such as could be the rudiment of the composite physical structures we encounter in actual experience, including our own bodies. Not only can we doubt particular objects and situations, since one can be as certain of their existence in dreams as in waking life, yet discover upon waking that none of it is true; but, says Descartes, we can also doubt general, mathematical truths, since I can be deceived in the midst of thinking about numbers in general as easily as I can in the midst of perceiving a particular thing.
I am very much in sympathy with Descartes on the desirability of getting to the bottom of things, if such a thing is possible, because, as he says, so many errors derive from any one false assumption that starting out with a set of prejudices about the world dooms any investigation of it from the outset. It is plausible, at least, that retreating from all the positions we have taken up about the world and ourselves to an impregnable stronghold of certain truth(s) would tell us where we stand with respect to knowledge.
I remark that Descartes' erasure of all that can be doubted is accomplished in two strokes, the first clearing away all knowledge of particular things, the second all abstract knowledge of the possibility of existing things. In the first case, it is the senses that fall under suspicion; in the second, thinking itself. But can we not then also be deceived about the argument that contains these assertions as premises?
If I can be deceived each time I add two and three, can I not also be deceived each time I think that some particular combination of premises necessitates a particular conclusion?