sight-read the score while listening to its errorful performance. On passive listening trials,
subjects listened to a chorale performed without errors; they were not reading a score. On
each scanned trial of the harmony, rhythm, melody, and passive listening tasks, subjects
heard a different Bach chorale. On rest baseline control trials, subjects fixated on a fixation
point (neither reading a score nor listening to music). The image data during the primary
three tasks were contrasted with those from listening to the errorless performance of a Bach
chorale without reading a score, as well as to the rest baseline control. Analyses were con-
ducted as described earlier for our preceding PET study.
To highlight some of our principal findings, the melody, harmony, and rhythm condi-
tions each showed a distinct pattern of distributed, statistically significant brain activity
compared to passive listening control (p0.001). Often, each condition activated different
subareas of a particular major brain area. In broad anatomical terms, score reading and
comprehension of melodic, rhythmic, and harmonic features of a piece were supported by
processes in both cerebral hemispheres. More specifically, melodic comprehension acti-
vated each hemisphere equally, whereas harmonic and rhythmic comprehension activated
more of the left hemisphere than the right one. In addition, the rhythm condition activated
comparatively few brain areas outside the cerebellum.
All three tasks strongly activated bilateral cerebellum, primarily the lateral hemispheres,
with little overlap in activated area across the three tasks (Figure 17.4). Given the lack of
explicit (and apparently implicit) movement, this cerebellar activity is likely related to sen-
sory or cognitive processing, not to motoric processing (see below). Furthermore, activa-
tion of cerebellum in the rhythm condition was twice that in harmony and melody
conditions. An involvement of the cerebellum in some way in processing rhythm is consis-
tent with recent neurological and neuroimaging data.42,47–49Cerebellar activations during
the processing of melody and harmony are relatively novel findings but are consistent with
data above from our first study and with other recent studies showing lateral cerebellar
activation for processing of tonal information (see later).
One area commonly activated by the harmony, melody, and rhythm tasks was in right
fusiform gyrus. Because score reading was a central process common to all three conditions,252 Figure 17.4 Significant blood flow activations in the cerebellum as musicians sight-read a J. S. Bach chorale score
and listen for performance errors in harmony, melody, or rhythm.^46 These are group-averaged PET images for
each task (contrasted with passive listening control) overlaid on anatomical MRIs. PET data are z-scores displayed
on a colour scale ranging from 1.96 (yellow;p0.05) to 4.0 (red;p0.0001). (See Plate 8 in colour section.)