Among the many topics covered in Sven Bernecker’s impressive study of memory is the relation between memory and personal identity. Bernecker uses his grammatical taxonomy of memory and causal account to defend the claim that memory does not logically presuppose personal identity and hence that circularity objections to memory-based accounts of personal identity are misplaced. In my comment I investigate these claims, suggesting that the relation between personal identity and memory is more complicated than Bernecker’s analysis suggests. In particular, I (...) argue that while he shows that some memories do not presuppose personal identity he fails to show that those that are appealed to in memory-based accounts of personal identity do not, and that the features of his view that allow him to define memory without reference to personal identity also obscure important features of memory that must be part of a complete account. (shrink)