| |
Abstract:
In recent years, there has been increased interest in the
distinction between perceptual and conceptual components of
semantic memory, with several studies using semantic priming to
examine this distinction. However, these studies have produced
mixed results: one study, for example, found that perceptual and
conceptual relations contribute independently to priming effects,
whereas another failed to find any evidence for priming of
perceptually related words. Moreover, the sensitivity of priming
effects to nonassociated semantic relations has itself been
questioned. Comparisons across experiments are difficult because of
differences in stimulus list composition. These issues were
examined in four lexical decision experiments, where prime-target
word pairs were related either associatively, by shape and
function, by function, by shape, or unrelated. The context of the
stimulus list was varied across experiments. When associated items
were present, no other relation primed. When associated items were
omitted from the list, the pattern of priming effects varied
depending on which other relations were present in the list. Most
priming effects were attributed to relatedness-checking strategies:
only the most salient relation primed in each experiment. However,
evidence for automatic priming was found in one experiment:
perceptually related words primed without participants noticing
prime-target relations. Taken together, these findings suggest that
list context effects can account for inconsistencies in the
literature, and that perceptual and conceptual components of
semantic knowledge can indeed be isolated using semantic
priming.
|
