that a single intact pathway to the right auditory cortex is not a sufficient substrate for the
normal perception of modulation. The presence of normal processing of IRN by I.R. pro-
vides an important control to allow the exclusion of other possible causes of the deficit. The
IRN task demanded similar attention and working memory for simple acoustic features
using a similar two-alternative forced-choice paradigm. The presence of such dissociated
deficits is critical to these studies,^28 and studies showing a general deficit in performance in
a range of tasks do not allow such interpretation in terms of distinct perceptual processes.
In terms of the psychophysical examination of higher-order temporal processing in sub-
jects with lesions, these studies suggest that the deficits are produced by temporal lobe
lesions that involve superior temporal lobe areas beyond the primary auditory cortex.
Additionally, subject I.R. has a large right frontal lesion implicating this area in the percep-
tion of sound sequences and low-rate modulation. More studies of such patients using
detailed modern structural imaging and systematic psychoacoustic testing will character-
ize the necessary structures for higher-order temporal processing. Such studies will never
be supplanted by functional imaging, from which sufficientprocessing mechanisms cannot
be immediately inferred.
Functional imaging
Fine temporal structure
The missing fundamental stimulus has been used in MEG experiments29,30where a
mapping of perceived pitch has been demonstrated. For reasons discussed above, this map
cannot be interpreted unambiguously in terms of the temporal structure of the stimulus,
as template models could also explain the perceived pitch. But the mapping of perceived
complex pitch to auditory cortex, derived by whatever means, is convergent with the lesion
data discussed above. In the study of Langner, distinct representations of the frequency and
temporal domains were suggested, with different orientations in the superior temporal
planum. This would allow extensive processing in auditory cortex of spectrotemporal
properties such as pitch as well as spectral (timbre) and temporal (attack) instrument
characteristics. These are all features of individual notes.
IRN has also been used in functional imaging experiments, with the aim of identifying
brain areas where the brain activity varies as a function of the fine temporal regularity of
the stimulus.^2 In an initial study with PET in collaboration with Patterson, such a relat-
ionship was demonstrated in both auditory cortices. On the right, a region of activation
was demonstrated in the medial part of Heschl’s gyrus, containing the primary auditory
cortex. On the left, the activation was situated more laterally and may correspond to sec-
ondary auditory cortex; these asymmetries are of interest in view of the lesion data sug-
gesting a particular involvement of the right primary auditory cortex in the processing of
temporal regularity. The activity shown in the PET study was argued to correspond to neural
activity occurring after temporal regularity is converted into a more stable code in the
ascending pathway, a process called temporal integration. The study is critically dependent
on this assumption, but not on the exact mechanism by which such temporal regularity is
stabilized, the simplest possibility being based on autocorrelation as mentioned above.
173