Automatic Semantic Priming: Effect of Category Exemplar Level and Word Association Level

In three experiments we evaluated semantic priming when the prime was masked below naming threshold and the target was named. Experiment 1 found that when word primes were masked and word targets were named, prior knowledge of the related pairs did not alter semantic priming. However, semantic priming within categories occurred only when the prime stimulus was the first category exemplar. Experiment 2 found that when masked pictures were used as primes, semantic priming for word targets was highly sensitive to the category exemplar level of the prime but not to the category exemplar level of the target. Word association norms collected in Experiment 3 did not support the hypothesis that the effect of category exemplar level was mediated by the strength of word association. Experiment 4 found significant semantic priming for masked picture primes and within-category word targets, regardless of the level of word association between the prime and target. Experiment 4 also demonstrated semantic priming for high word association targets that were not members of the same semantic category. For all experiments, the subjects with the longest average reaction times also showed the largest semantic priming effect for naming word targets. We hypothesized that viewing one of the highest ranking category exemplars activates the memory representation of the category, perhaps because such prototypic exemplars are contained within the category concept itself.