Abstract
In each of two experiments 32 Ss were familiarized with a set of stimulus and response nonsense syllables and then learned a list of eight pairs. Each of four familiarity conditions (neither syllable, stimuli, responses, and both syllables) was represented in this list by two pairs, one learned under a repetition condition and the other under a nonrepetition condition. In Exp. 1 the pairs were presented for 2 sec. each; in Exp. 2 presentation was self-paced. The difference between learning under the repetition conditions was shown to be a function of stimulus and response familiarization. The sensitive variables were total errors in Exp. 1 and presentation time (time to learn) in Exp. 2. It was difficult to criticize these findings on the basis of the item-selection artifact usually referred to in studies of this type. The results favor an all-or-none conception of PA learning when the stimuli and responses act as single elements.

This publication has 19 references indexed in Scilit: