Consider claims like this:
- What it is for a person x at t0 to be identical with a person y at t1>t0 is for y at t1 to have a chain of memories leading back to x at t0 (and maybe: and there is no other competitor).
- What it is for a person x at t0 to be identical with a person y at t1>t0 is for y at t1 to have a body that has continuously developed from that of x at t0 (and maybe: and there is no other competitor).
I think that people who are attracted towards (1), (2) and claims like them don't really think of diachronic personal identity as identity, maybe because they don't think of persons as really entities (or maybe they just don't think of persons as substances).
It might be thought that a stage-theorist need not be worried about this. The stage-theorist, after all, may think that x at t0 and y at t1 are different entities, being different stages of a greater whole, and that (1) and (2) give a good account of what makes them be stages of the same whole. But I think this is only somewhat more plausible, because the parthood relation (which the stages stand in to the worm), if there is such a relation, is surely almost as ontologically basic as identity.
This does not mean that the discussion of criteria of personal identity is puerile. For one can take identity to be genuinely primitive, but still look for truths such as:
- Necessarily, a person x at t1 has a chain of memories going back to every time prior to t1 at which this person existed.
- Necessarily, there are no such chains of memories between distinct persons.