By recording event-related EEG potentials from patients with frontal,
parietal, or temporo-parietal lesions, we wanted to determine the cort
ical area that is relevant to the P3(b) component, replicating the app
roach used by Knight and associates [9,11,12,37] who found reduced P3(
b)s in patients with temporo-parietal lesions. They used auditory and
somatosensory stimuli. We wanted to replicate their findings in audito
ry and visual oddball tasks and analysed potentials evoked by targets
and by standard stimuli. Temporo-parietal patients' P3(b)s were reduce
d with auditory targets and lacked a distinct Pt maximum with visual t
argets. Further, auditory N1 was reduced both with targets and standar
ds, P3 to visual standards and P2 to auditory standards were reduced.
Parietal patients' P3(b)s differed only slightly from the control grou
p, being somewhat reduced over the lesioned hemisphere with visual sti
muli. Their P3 to visual standards was, however, reduced to the same e
xtent as was the temporo-parietal patients'. Frontal patients did not
differ from the control group both in their P3(b) and in their P3 to s
tandards but had a number of conspicuous features in modality-specific
components. In sum, our results on P3(b) (as well as on N1) replicate
Knight's findings, confirming his conclusion that integrity of the te
mporo-parietal junction is critical for P3(b). Implications for hypoth
eses on P3 are discussed.