
In his comment on Nyberg and Tulving's (1996) article, Shanks (this issue) argues that the dissociation evidence we presented for the multiple memory systems approach is weak. In this reply, we consider his comments, discuss some inconsistencies in his arguments, and note that he seems to have overlooked some of our points. We maintain our position that dissociation evidence constitutes one important source of support for separate memory systems, and we argue that converging evidence from dissociation studies and functional neuroimaging studies strongly indicates the existence of multiple human memory systems.