or melody processing.46,49 The pattern of coupled pitch and ataxia measures, and
decoupled pitch and loudness measures, suggests that the role of the cerebellum is to monitor
and optimize the acquisition of information in sensory modalities (audition). Patients and
controls did not differ in digit span memory. This indicates that the differences between
patients and controls in pitch discrimination were very likely not due to any working mem-
ory deficits that may possibly be associated with cerebellar damage.
An important independent corroboration of the implications of these data is provided
by a recent study examining cerebellar degeneration patients’performance on a variety of
visual tasks.^43 The authors report that the patients were impaired selectively for the detec-
tion of visual motion, speed, and direction, as dissociated from eye movement control and
from any other attentional, memory, or cognitive task components. It should be noted that
there are also neurological studies suggesting that the cerebellum may be instrumental in
other nonmotor behaviours. These behaviours include judging the timing of events, solv-
ing perceptual and spatial reasoning problems, and generating words according to a
semantic rule.47,78–^83 In terms of the sensory acquisition hypothesis we are testing, however,
impairments in these specific tasks may be more related to the dependence of these tasks
on well-controlled information processing (e.g. sensory acquisition) than to different spe-
cific roles for the cerebellum in each of those behaviors.
General issues in interpreting brain activations
Given the rapid emergence of promising new neuroimaging methods for the study of
music performance, perception, and cognition, it may be useful in the present volume to
discuss current and general methodological issues for interpreting the role of activated
brain areas in the processing of music. Mapping elementary operations onto particular
brain areas and mapping those operations/areas onto system-level models of specific musi-
cal and other cognitive tasks requires a convergence across paradigms (behavioural tasks)
and independent experiments. No single task or paradigm used in neuroimaging or neu-
rological investigations can isolate an elementary operation. Likewise, no single group of
subjects (i.e. one study’s sample) can fully capture the location and distribution of a brain
area in a population. Thus, mapping localized operations entails sifting through multiple
studies.
In functional neuroimaging, for example, the analysis of multiple studies, or metanaly-
sis, can be rigorously and quantitatively performed^84 because the data are customarily
reported in stereotaxic coordinates.23,30A growing number of rigorous metanalyses with
important implications for mapping operations and systems, as well as for psychological
models, are being conducted. For example, such analyses have been performed in motor
processes and supplementary motor areas,^85 in oculomotor control and frontal eye fields,^86
in speech motor-sensory processes and the mouth area of primary motor-sensory
cortex,87,88in visual information processing,^89 –^91 in visual attention,^91 and in memory.^92
Improved methods for such meta-analyses are currently being developed and vali-
dated.84,87,88,93Using such quantitative stereotactic-based metanalyses will yield important
insights into the relationship between neural mechanisms supporting music and those
supporting other cognitive processes.
262