| |
Abstract:
In ERP studies of masked priming where subjects are not aware
of the prime word, words preceded by a string of X's elicit more
negative brainwaves within 250 ms when compared to words preceded
by words (Grossi & Neville, 1999, 2000). In this study, we
manipulated the nature of the prime in order to investigate whether
these effects were due to physical characteristics of the primes or
more abstract linguistic information. In a masked lexical decision
task, words and nonwords were preceded by unrelated words,
pronounceable nonwords, strings of letters, false fonts or strings
of X's. Primes were presented for 66 ms and masked so that subjects
were unable to identify them. ERPs showed that words preceded by
strings of X's were more negative in the 90-150 ms window at
occipito-temporal sites as compared to words preceded by all other
primes, suggesting that this effect might be due to differences in
brightness or other physical dimensions of the stimuli. In the
120-250 ms window, different ERP effects were obtained when targets
were preceded by words and pronunceable nonwords as compared to
letter strings, false fonts or X's; the different scalp
distribution and timing of these effects suggest that the retrieval
of letter and wordform information involve non-identical neural
systems.
|
