| |
Abstract:
ERP waveforms to repeated words, faces and objects are more
positive than first presentations, but there is no such difference
for orthographically illegal non-words, scrambled faces, and
nonsense patterns. We examined ERPs to initial and repeated
presentations of line drawings of structurally possible and
impossible objects in a target detection task. In Experiment 1, the
non-target objects were possible and impossible geometric figures
and the targets were real-world objects or combinations of parts of
real-world objects. In Experiment 2, the non-target objects were
real-world objects and the targets were geometric figures. At
frontal sites the repeated possible and impossible non-target
items, in both experiments, evoked a more positive ERP waveform
(250-350ms) than did first presentations. In contrast, at parietal
sites the repeated non-target items, in both experiments, evoked a
more negative ERP waveform (300-600ms) than did first
presentations. The topographic distribution of this parietal
repetition effect was different from the target P300 effect. The
brief frontal positivity to repeated items may reflect a
facilitation of conceptual or lexical access during object
categorization. The polarity of the parietal repetition effect was
the reverse of what is usually found with stimulus repetition,
although it is consistent with some earlier work using visual
stimuli. It may reflect a facilitation of a perceptual
representation or activation of an episodic memory trace.
|
