ABSTRACT
There is general agreement that recognition decisions are based on multiple components or sources of information. There is, however, no agreement on these components nor on the kind of data that would support one component over another. One reason for this problem is that some of the proposed components are more or less directly tied to operationally defined tasks while others are defined primarily with respect to explicit or implicit assumptions about the structure of the memory trace and the processes that operate on that structure. When these distinctions about the level at which the construct is defined are ignored it leads to confusion. As an alternative to continuing the debate at these multiple levels it is shown how many of the ideas about the component processes in recognition can be incorporated into a general theoretical framework based on distributed storage or PDP models of memory. Such a framework provides a basis for assessing the contributions of the proposed components to the recognition decision.
