Abstract

In Morals by Agreement, David Gauthier tells what I take to be the most convincing story yet about what it is we might be doing when we are thinking, talking, and acting morally, a story that can be extended, I submit, to the more general case of acting defensibly. I begin this enquiry, then, with a largely supportive rehearsal of Gauthier's story, including its domain of discourse, its motivation, its methodology, its conception of rationality and, finally, its pivotal move. ;What shall be of especial interest, however, is less the substantive details of Gauthier's account than its explanatory schema, the central feature of which being that there are certain 'games'--games paradigmatic of the human interpersonal condition--the winning strategy for which in-volves the performance of an operation on some feature of one's own psychology. Of these the most interesting will prove to be what might be called 'second-order' self-modifications, that is, operations performed on one's own psychology that are designed to ensure the development of--and/or once developed, the protection of--some of these aforementioned first-order modifications. ;But of these, in turn, I will be focussing on second-order self-modifications designed solely to ensure the unreadoptability of some feature of the original psychology. I call these features--that is, those it behooves us to render unrecoverable--'self-effacing'. And my claim is that a comprehensive analysis of the rationality of self-effacement can make good on a promise of incalculable explanatory power. ;In the course of this analysis, however, it will emerge that there is a paradox within the rationality of self-effacement, the resolution of which may require a modest revision--but a revision nonetheless--to the conception of rationality with which we started out. This will thrust us squarely into the longstanding debate--only now armed with an agenda of our own--over whether prudence is to be characterized strictly in terms of current preferences which are simply higher-ordered, or rather in terms of first-order preferences which can also be anticipated and/or remembered. ;I conclude with a series of comments of what ramifications, if any, our findings might have: for the completion of the meta-ethical program with which we started out, for the debate over the conditions on moral considerability, for naturalized epistemology--and, therein, for the residual problem of skepticism--and, finally, for the possibility of yet another convincing story about what it is he might be doing when we are thinking, talking and doing philosophy itself.