Memory and personhood

As in the last post, I am going to argue against the conjunction of the following two views:

1. CogSkill: Personhood requires developed and complex cognitive skills, of a sort that fetuses lack, and that are had by no mammals other than humans or at most are had by humans, and some other primates and/or cetaceans.
2. NonStage: When a person comes into existence, that person is a new entity that comes into existence.

In the previous post, the arguments against the view that the non-personal infant ceases to exist at the advent of personhood were decisive. So I will dismiss this option, and assume that the view is that after the advent of personhood, there are two beings, one of which existed prior to the advent of personhood and the other of which is the person.

Now, surely there are some animals, say dogs, that lack the skills posited by CogSkills have memories of having had particular experiences in the past. It is very plausible, then, that human infants prior to gaining these skills will also have some such memories. Suppose, then, that Sally is an infant, prior to the advent of the relevant complex cognitive skills, who has a memory of some experience. Surely she does not lose all of her experiential memories when she comes to gain these skills. On the contrary, surely some memories play an on-going role in developing the cognitive skills, and remain at least for a short time after the skills are present and a person has come to be present.
After personhood has been achieved, on the view we are thinking about, there will then be two subjects of some experiential memory of a pre-personal experience E. One subject will be a person and one a human animal. But only one of these two subjects will be having a veridical memory. For the person is mistaken in remembering having had the experience E. For the person never had E, since the person didn't exist when E occurred. The human animal, on the other hand, did have E, and her memory of E is veridical. The idea that here we have two subjects of memory, one veridical and one not, seems quite absurd. And this is a reason to reject the conjunction of CogSkills with NonStage.