Authors:

Henson, R. N. A., Norris, D. G., Page, M. P. A. & Baddeley, A. D.

Reference:

Year of publication:

1996

CBU number:

3339

Abstract:

In three experiments, young adults saw either six or seven consonants presented sequentially, which they attempted to recall in the same order, immediately after the last consonant had disappeared. The findings replicated those of similar experiments by Baddeley (1968): Though phonologically confusable consonants were harder to recall in the correct position, their presence in lists where they alternated with phonologically nonconfusable consonants had little effect on the probability of correctly recalling the nonconfusable consonants. This applied whether consonants were read silently and responses written (Experiments 1 and 2) or whether they were vocalised and responses spoken (Experiment 3). These facts are hard to reconcile with models that assume serial recall proceeds by chaining along a sequence of associations between items. Further analysis of errors revealed that most transpositions were over short distances (the locality constraint), that phonologically similar items tended to swap with each other (the similarity constraint), and that repeated responses were rare and usually far apart (the repetition constraint). Though these observations have been reported elsewhere, taken together under the conditions of the present experiments, they prove troublesome for most current models of serial recall, even those that do not employ chaining. A new model is described that is consistent with these constraints and that simulates the detailed pattern of errors observed.