Controlled versus automatic processes in retrieval -induced forgetting.

摘要

Retrieval-induced forgetting refers to memory impairment produced by the retrieval of related information. It has been assumed that retrieval-induced forgetting is based on "active" inhibition produced by retrieval competition. Further, it has been assumed that such inhibition is cue-independent---that is, what is suppressed is the item representation rather than the association between the cue and target. The present experiment examined the extent to which this phenomenon requires executive control. In particular, the present experiment investigated whether retrieval-induced forgetting is the result of controlled or automatic processes. Controlled processes are assumed to be resource-dependent whereas automatic processes are assumed to be resource-free. If retrieval-induced forgetting is produced by controlled processes, performing a resource-demanding secondary task during the retrieval practice phase would reduce or eliminate forgetting. In contrast, if retrieval-induced forgetting is the product of automatic processes, performing a secondary task would have little effect on forgetting. These predictions were tested using Anderson's retrieval-practice paradigm with or without a resource-demanding secondary task. The results showed that performing a secondary task reduced the amount of retrieval-induced forgetting. Relative to the standard full attention condition, both within- and cross-category retrieval-induced forgetting were reduced approximately by half in the divided attention condition in which participants performed a secondary task during the retrieval-practice phase. The present experiment therefore provided the first direct evidence that retrieval-induced forgetting represents a form of controlled inhibition.