from patient to patient, in spite of the intersubject variability in localization and orientation
of the primary auditory cortex.
In a recent study, Pantev et al.^23 observed a tonotopic organization of the human audi-
tory cortex for the Pam component (corresponding to the 30-ms intracerebral AEP com-
ponent); unlike the results of the present study, however, higher frequencies were
represented more laterally than were lower frequencies. This discrepancy may be explained
by the fact that, in the present study, only slight frequency-dependent fluctuations in
amplitude were observed for the 30-ms component, compared to those observed for the
P/N 50. Furthermore, the P1m (corresponding to the P/N 50 in this study), generated in
the primary auditory cortex, was not systematically discernible in Pantev et al.’s study and
has been shown in several MEG^44 –^46 and EEG studies to have highly variable scalp topo-
graphy. With intracerebrally recorded AEPs, on the other hand, the P/N 50 is robust and
can reach amplitudes of up to 150 V.^25 It is also probable that two different overlapping
tonotopic maps exist, each corresponding to either the 30- or the 50-ms component, and
that different investigation techniques have differential capacities to explore these maps.
Frequency-dependent fluctuations in amplitude were also observed for the 80-ms com-
ponent, which is generated in secondary auditory areas anterior and lateral to the primary
cortex.^25 In the right hemisphere, a mediolateral as well as an anteroposterior tonotopic
organization was seen, with high frequencies represented posteromedially and low fre-
quencies represented anterolaterally. High frequencies were also represented ventrally in an
area adjacent to the lateral part of the primary cortex, which was also the area in which low
frequencies were represented for the 50-ms component. While the distribution of BFs was
roughly similar for the 80-ms component in both hemispheres, this distribution was not as
well defined in the left hemisphere, where neuronal populations were often found to
respond to more than one frequency or over a broad range of frequencies. Neither
Merzenich et al.^13 nor Reale and Imig,^16 however, found a clear tonotopic organization of
the secondary cortex (AII) in their studies with nonhuman animals: several groups of neur-
ons had multiple BFs or responded over a broad range of frequencies. Evoked responses
from this area had longer latencies and were less well tuned than those from the primary
cortex.16,47In their study, Merzenich et al.^14 reported that neurons in these secondary areas
often responded over a broad range of frequencies and that the assignment of BFs was dif-
ficult, with two maxima often being observed. It is possible that a second harmonic distor-
tion produced this ‘double range’because the pairs of maxima in this study were often
harmonically related.^48
With respect to area 22, in the present study no tonotopic organization was observed in
either hemisphere. This finding is consistent with data from animal studies using single-
unit recordings^49 showing that neurons in nonprimary auditory areas in the macaque are
not selectively responsive to pure tones, but respond only to frequency-centred sound
bursts containing many frequencies.
One of the most significant findings in this study is the difference in the response select-
ivity of neurons in the right and left cerebral hemispheres. Neurons in the right auditory
cortex were more sharply tuned to frequency than neurons in the homologous region of
the left hemisphere. In a previous study, we showed that left auditory neurons were speci-
fically sensitive to the temporal features of auditory information.^50
163