On a supposed temporal/modal parallel

Abstract

N 'On a temporal slippery slope paradox' (ANALYSIS 46.1, January 1986), D. E. Over attempts to formulate a temporal paradox of artefact identity paralleling Chisholm's well-known modal paradox. I do not believe that the parallel works but think that it fails interestingly because of an important dissimilarity between temporal relations and relations between possible worlds, consisting in the fact that the former put instants of time into a unique linear order not paralleled by any analogous linear ordering of possible worlds. In short, time constitutes a dimension in a way that possibility does not. Chisholm's paradox may be posed in the following manner. Suppose we allow (as seems plausible) that an artefact could have originated from a set of component parts slightly different from those from which it actually originated but not from a wholly different set. Then we may be invited to consider a sequence of worlds in which a succession of slight changes in the originating components is envisaged in a fashion consistent with the identity of the resulting artefacts in successive worlds in the sequence, yet such that we end up with a world in which the originating components are completely different from those of the actual world, so that, by virtue of the transitivity of identity, we are seemingly obliged after all to identify the artefact resulting in this final world with the actual artefact contrary to our initial supposition. Over's supposedly parallel temporal paradox is alleged to arise from the following principle, which he takes to be analogous to the modal supposition just discussed: