ABSTRACT
The problem of transworld identity, due to Chisholm (1967), is a non-epistemic problem. It covers questions about the conditions, if any, under which an individual that relates to one possible world in a certain non-epistemic way is identical to an individual that relates to another possible world in that same kind of way. The problem of transworld identification due to Kaplan (1967) is an epistemic problem. It covers questions about how, if at all, we can reliably or successfully tell whether we have a case of transworld identity. Sometimes the problems are conflated, and conflation is abetted by unqualified talk of how we ‘individuate’ possible individuals or by unexplicated talk of ‘criteria of identity’. Even when the problems are not conflated, it often seems that the (alleged) intractability of the problem of transworld identification is supposed to have one or more of the following consequences for transworld identity: that claims of transworld identity are semantically deficient; that we do not understand claims of transworld identity; or that there are no facts of the matter about transworld identity.1 It is easy to gather from the literature the impression that transworld identity, at least, is a problem for AR but not for GR, since AR asserts the transworld identity of individuals while GR rejects this thesis in favour of the worldboundedness of individuals. I will argue, however, that this impression is highly misleading. The substantive issues generated by the problem of transworld identity do not render the position of AR obviously more problematic than that of GR. On the other hand, it is mistaken to think that the standard AR solution to the problem of transworld identification – the ‘stipulation solution’ – also provides a solution to the real epistemological problem set for AR by considerations of transworld identity.
