ABSTRACT
A basic question about the nature of memory encoding and retrieval concerns whether they are similar operations. Many cognitive theories champion the view that encoding and retrieval are comparable processes, such that retrieval is successful to the extent that encoding processes are reinstated (Bransford, Franks, Morris, & Stein, 1979; Craik, 1983; Kolers, 1973; Roediger, Weldon, & Challis, 1989; Tulving & Thomson, 1973). There is growing evidence, however, that these mnemonic operations differ in important ways. The divided attention paradigm has three primary advantages that make it ideally suited to address the similarity of encoding and retrieval. First, one can compare the vulnerability of encoding and retrieval to disruption by having subjects perform a secondary task only during encoding or only during retrieval. Second, one can examine the attentional control of encoding and retrieval by asking participants to vary their emphasis between the memory task and the secondary task across trials. If encoding and retrieval are under voluntary control, then as they are given more emphasis, memory performance should increase and secondary task performance should decrease. Third, one can examine the resource demands of encoding and retrieval, if secondary task costs are taken as an index of their resource demands (Kahneman, 1973; Kerr, 1973). The more central resources encoding or retrieval demand, the more they should disrupt performance on a secondary task.
