ABSTRACT
Classic dissociations between priming tasks and episodic memory tasks (Tulving, Hayman, & MacDonald, 1991; Tulving, Schacter, & Stark, 1982; or between implicit and explicit memory tasks, see Schacter, 1987, for a review) have a natural interpretation in the model. The implicit tasks can generally be done by the semantic lexicon (and perhaps the familiarity monitor), whereas the explicit memory tasks, including recall, source and contextual assessments, recognition requiring binding, and recollection are handled by the convolution/correlation module. There are some paradigms that have traditionally been attributed to either explicit or implicit memory that do not require binding-but that, under some circumstances might benefit from it. The model allows us to analyse these multiple determined cases. Performance on those paradigms might be enacted by either module-although with potentially different fine-grained results. Some classification learning experiments could use either (see the above section, and Metcalfe Eich, 1982, for illustrations). Similarly, the false memory paradigm results could be obtained whether binding was used or not (see Dodhia & Metcalfe, 1999, for CHARM simulations of this paradigm).
