3 ms·
You might like this better. I wrote this a few years back, mostly about Parfit's personal identity philosophy: https://www.thoughtexperiments.net/people-who-div
by yoaviram 3y ago
You might like this better. I wrote this a few years back, mostly about Parfit's personal identity philosophy:
https://www.thoughtexperiments.net/people-who-divide-like-an-amoeba/ https://www.thoughtexperiments.net/people-who-divide-like-an...
- cubefox 3y agoI think the right solution is to replace the relation of personal identity across time (=) with the notion of "becoming" or "having been" (>, <). Instead of saying "Baby Peter is adult Peter" (BP = AP) we can say "Baby Peter will become adult Peter" (BP < AP) or "Adult Peter was Baby Peter" (AP > BP). Unlike =, < and > are not equivalence relations. They are transitive, but they are neither reflexive nor symmetric. This way branching persons (forks or fusions) do not lead to paradoxes. If X branches into Y and Z, it is perfectly possible that X < Y, X < Z, and Y ≠ Z. The person that is about to enter the duplicator (X) becomes, one moment later, both the person that comes out on the left (Y) and the person that comes out on the right (Z). That doesn't mean that Y and Z are the same person, they just both have been X a moment ago. This means that when we go to bed in the evening, there is a different person that comes out of the bed in the morning. But that's not so counterintuitive, since the evening person still became the morning person. This weaker relation of "becoming" (rather than the stronger relation of "being identical to") can also serve as the notion of "survival". The evening person "survived as" the morning person. Unfortunately Parfit didn't discuss this solution.