Retrieval from semantic memory an analysis of facilitory and inhibitory priming effects

The purpose of the present study was to replication Blaxton and Neely’s (1982) findings of facilitation from one semantically related prime and inhibition from four semantically related primes in a generation task, and 2) to determine if facilitation and/or inhibition are a function of the level to which the prime and the target are processed. Using Blaxton and Neely’s generation paradigm, subjects generated both primes and targets from the first two letters of words in response to a rhyme cue (nonsemantic task) and/or a category cue (semantic task). For the within-subject manipulations, subjects either generated one or four primes prior to the target, which were either semantically related or unrelated to the target. A recognition test was administered following the generation task to determine if memory for the primes and targets was a function of the level to which they were processed. The only evidence of inhibition was found on the priming trials in both the related and unrelated conditions, but not on the target trial in the related condition, as was found by Blaxton and Neely. Several explanations are offered to account for this failure to replicate. When the nonsemantic task was performed on the primes, smaller differences in reaction time between the related and unrelated targets occurred compared to when the semantic task was performed on the primes, supporting previous evidence that nonsemantic tasks on related primes reduce the facilitory effects. The manipulation of target task was inconclusive. The recognition test results indicated that memory for the primes and targets seems to be a function of the level to which they are processed. The results were discussed in terms of the spreading activation and retrieval inhibition hypotheses