Time and Identity

10 Me, Again

I traveled to the conference from which this volume is drawn with John Pollock, a dear and now departed friend and colleague. I dedicate this paper to him with love and respect. He is missed. Introduction Thought about the self raises some very special problems. Some of these concern indexical reference quite generally; but there is one, having to do with identity over time, that seems to be unique to the self. I’ll be using a historical exchange between Anscombe and Descartes to raise the problem and proposing a resolution that casts light both on why self-directed thought presents a unique epistemic predicament and where Descartes’s cogito argument—still one of the most compelling and resilient arguments in philosophy—goes wrong. What Anscombe Said to Descartes Descartes begins his discussion in the Meditations with the question “what am I?” and concludes, famously, that he is a nonmaterial substance. His reasoning turns on the thesis that nothing can be true of his nature that is not made known to him in the act of thinking.1 Schematically, the argument runs thus: (1) I =def that thing whose existence cannot be doubted (i.e., that thing whose existence is made known in the act of doubting that it exists). (2) I can’t be identical to anything whose existence can be doubted. 10 Me, Again Jenann Ismael 210 J. Ismael (3) I can doubt the existence of anything for which objective criteria of identity can be provided (give me a description and I can coherently doubt that it is satisfied). (4) Hence, I can’t be identified with anything for which objective criteria of identity can be given. (Nagel 1989, 35)2 So construed, the argument is subject to an objection—often associated with Lichtenberg, but repeated by Russell, and given its most articulate expression by Anscombe—that is, on first encounter, devastating (Anscombe 1990).3 The charge is that Descartes’s argument fails because he pulls a bait and switch, overtly declaring the object of his inquiry is nothing other than that whose existence is made known in the act of trying to deny that it exists, but then tacitly appealing to a richer notion in allowing that it can be reidentified in different thoughts. In Anscombe’s words: People have sometimes queried how Descartes could conclude to his RES cogitans. But this is to forget that Descartes declares its essence to be nothing but thinking. . . . His position has, however, the intolerable difficulty of requiring an identification of the same referent in different I-thoughts. (Anscombe 1990) It is not the proclaimed guaranteed existence of the self that is problematic , for that does indeed follow if the ‘I’ in an ‘I’-thought is purely reflexive , that is, if it simply refers to itself. It is the combination of guaranteed and continuing existence that is objectionable. For, if the thinker outlives the thought—if the ‘I’ in one thought is even potentially intersubstitutable with the ‘I’ in others—its existence can’t be revealed in the mere production of an ‘I’-thought. All whose existence is so guaranteed are the individual thoughts themselves. According to the argument, Descartes is not entitled to assume the existence of an underlying continuant to supply them a common referent. On the one hand, the concept needs to be kept bare to guarantee existence. On the other hand, the bare concept doesn’t seem to be enough to support reidentification. If I cannot be something whose existence is not made known in the production of an ‘I’-thought, something whose existence can be coherently doubted while I am thinking of it, then I cannot be something that even potentially recurs. The unadorned reflexive ‘I’ of the individual thought cannot pull other ‘I’ occurrences under its referential scope. ‘I’ and Other Indexicals The difficulty can be brought out with a comparison of ‘I’ with names and other indexicals. In the case of names, ambiguity aside, criteria for identifi- Me, Again 211 cation and individuation of words substitute for criteria of identification and individuation of objects.4 The user of what Frege used to call a “logically perfect language” with a name for every object and no indexicals could be assured that the same word used on two occasions or in two expressions referred to the same object and could be intersubstituted salva veritate...