Target interaction effects: implications of the parallel distributed processing assumptions

Abstract

[Abstract]: Three experiments looked for target interaction effects and asked whether these would only be found when list membership information was relevant to the retrieval task. Subjects studied relatively short lists of words some of which contained two similar targets (thief and theft) or two dissimilar targets (thief and steal) associated with the same cue (ROBBERY). Subjects were anticipating a running memory span task that was sometimes interrupted by the provision of the extralist cue. Target interaction effects were found in cued recall. In contrast, there was no evidence for target interactions with free association. These experiments plus two others also showed that cued recall and free association could be further differentiated by the probability of producing a non-presented item, the benefits of a second presentation at study, the interfering effects on running memory span, and by the right hand tail but not the leading edge of the latency distribution. The results are discussed in reference to theories which address the issue of how semantic information derived from the extralist cue is combined with list membership information.