Hemispheric differences in effects of meaning similarity and meaning dominance on semantic priming: A divided visual field study

Based predominantly on semantic priming studies with divided visual field (DVF) presentation, current models of hemispheric differences in word semantic processing converge on a proposal that left hemisphere (LH) processes focus word meanings to their core by inhibiting less related meanings, whereas right hemisphere (RH) processes keep less related meanings active. The inhibition process supported by LH processing is assumed to apply to two distinct semantic processes: (a) narrowing of a single word meaning (inhibition of less related features and words), and (b) elimination of incompatible/conflicting meanings of an ambiguous word.; Semantic priming studies investigating hemispheric differences in these two processes have relied on associated prime-target pairs, which might have been problematic for two reasons. First, association might reflect lexical co-occurrence of word forms rather than effects of semantic relatedness; therefore, these studies might have confounded lexical and semantic priming effects. Second, in studies of ambiguous words dominant items were strongly associated whereas subordinate items were weakly associated; therefore, these studies confounded dominance and degree of relatedness.; To address these confounds, this study conducted two semantic priming experiments with central prime presentation, DVF presentations for targets, and a 750 ms SOA. Experiment 1 investigated the effect of degree of semantic similarity on priming, using non-associated, prime-target pairs that were controlled for lexical co-occurrence. Experiment 2 investigated effects of meaning dominance on priming with non-associated prime-target pairs. Results are consistent with high-similarity priming for left visual field (lvf) and possibly for right visual field (rvf) targets, and with high-dominance priming for rvf and lvf targets, suggesting that LH (and RH) processes mediate effects of semantic similarity and dominance.; However, priming effects in both experiments were very small. Thus, priming effects might have reflected that prime-target relatedness was less than expected, indicating that LH processing does not inhibit less related meanings, which is consistent with other studies using central primes. Additionally/alternatively, larger priming effects in other studies might derive mainly from association rather than semantic similarity. Finally, the small priming effects could be due to some aspect of the experimental procedure that might have made these experiments less sensitive to semantic priming.