Summation of activation: evidence from multiple primes that converge and diverge within semantic memory

Six experiments addressed the combinatorial influence of multiple related primes in naming, lexical decision and relatedness judgment performance. Primes either converged on a single semantic representation (e.g., LION-STRIPES-TIGER) or diverged onto distinct semantic representations (e.g., KIDNEY-PIANO-ORGAN). The facilitatory influence of 2 related primes was well predicted by the sum of the influences from the single-related-prime conditions (a) for both convergent and divergent primes, (b) in lexical-decision and naming, (c) across varying prime-target stimulus onset asynchronies, and (d) under target-degradation conditions that increased the priming effects. The relatedness-judgment task yielded an additive pattern of priming for convergent prime conditions; however, an underadditive pattern of priming was found for divergent prime conditions. Discussion focuses on the role of attentional systems that modulate the type of information used to perform a given task.