Inhibitory processes in sequential retrieval: evidence from variable-lag repetition priming

Performance of simple mental arithmetic provides a sensitive measure of subtle inhibitory and excitatory processes operating in sequential retrieval. Specifically, the probability that an error (e.g., 6 x 7 = 36) matches an earlier answer is below chance for the preceding trial (negative error priming), greater-than-chance for intermediate lags (positive error priming), and returns to chance after about 10 intervening trials (approximately 60 sec) (Campbell & Clark, 1989). Negative error priming suggests that inhibitory processes initially suppress just-produced items and temporarily mask excitatory effects of recent retrievals. To test further the inhibition hypothesis of negative error priming, we examined repetition priming effects at variable lags. The results demonstrate similar non-monotonic functions for error priming and repetition priming as a function of lag, consistent with the hypothesized decay of inhibitory memory processes.