Inhibition from related primes in semantic memory retrieval: A reappraisal of Brown's (1979) paradigm.

The present study was an attempt to replicate and extend Brown's (1979) finding of inhibition in retrieval from semantic memory produced by presentation of semantically related primes. Subjects' speed and accuracy in answering generalknowledge questions (e.g., Who was the first man to walk on the moon?) were measured when the question was preceded by one of four different prime types: neutral (ready), unrelated (alligator), semantically related (John Glenn), or correct (Neil Armstrong). One group of subjects was given all four of the different prime types across the 96 questions, while a second group received only the first three types (i.e., correct primes did not occur). For the first group of subjects, latency to answer the questions^was greater following related primes than the other types, replicating Brown's finding of inhibition from semantically related primes. However, for subjects who did not receive any correct primes, latencies were not greater following related primes than following neutral or unrelated primes. Thus the inhibitory priming effect reported by Brown is not due to automatic spreading inhibition (as he suggested). Rather, the inhibition seems due to strategic factors involved in evaluating the prime to determine if it is the correct answer to the •