7. Análisis de datos
7.8 Grado decimo
5.1.1.4.1
M elo d y C ondition
W hen the learners used their newly acquired skill o f sight-reading, they showed
activation in right superior parietal cortex. A group o f non-leamers, who were also
scanned twice, using an equivalent inter-scan interval, did not show any difference in this
brain region, thus ruling out the possibility that the changes were due to non-specific
factors associated with being scanned twice.
In Chapter 3, empirical evidence illustrates the importance o f visuo-spatial
transformations in music reading. The dorsal visual processing stream, within which
superior parietal cortex resides, is known to be im portant for the coding o f the spatial, as
opposed to featural aspects o f visual stimuli (the "what"/"w here" distinction") (Damasio &
Benton, 1979; Gross, 1973; MaunseU & Newsome, 1989; Mishkin & Ungerleider, 1982;
Pohl, 1973). A distinction has also been made between the visual perception o f objects
versus the control o f action upon those objects (the "w hat"/"how " dichotomy) (Goodale
& Milner, 1992). W hether the distinction made is one o f "what versus where or "what
versus how", sight-reading for keyboard performance falls squarely within the functional
remit o f the dorsal stream. First, the information relevant for performance is contained in
the position o f the note on the stave ("where"); second, musical performance relies on the
use o f this positional information to guide selection o f the appropriate keypress ("how").
An activation similar to one found in the present study, in right superior parietal
cortex, was also seen in Sergent's P E T study o f musical sight-reading (Sergent et al., 1992).
activated right superior parietal cortex, but this region was not activated in the reading
alone or listening alone condition. The replication o f this finding in the present experiment
in which £MRI was used, as opposed to PET, and novice pianists were used, as opposed to
professionals, suggests that the common activation across the two studies is likely to reflect
the specifically visuospatial translation element o f music reading, and, furthermore, that
such activation may be independent o f skill-level.
5.1.1.4.2
R h ythm C ondition
The absence o f a Training Effect in the Rhythm condition may be due to two
independent factors. First, the pre-training and post-training tasks may have been too
similar for differences to be found. While reading music for rhythm requires visual
decoding based on the features o f the musical notation, (a stage which was not required in
the pre-training session), the notation is converted into a representation o f the num ber o f
beats specified before mapping this notion o f duration onto the m otor response. It is
therefore possible that early differences in the decoding stage, in advance o f this duration
to response m apping are too subtle to be noticed. Second, there was only a small
difference between the experimental and control trials; reading o f rhythmic notation was
also required for control trials, post-training, though to a lesser extent than for the
experimental trials. Further, control trials were sometimes bars o f quavers, or sometimes
crotchets or mimins and hence were not entirely predictable. This is likely to have
increased the demands made upon rhythm reading. However, this was a necessary
constraint o f the experimental design since response rates had to be matched across
experimental and control trials over the scanning session as a whole.
Since the implicit music reading task did not involve sight-reading, differences in
activation, related to the presence o f task-irrelevant musical notation, after training, must
reflect an effect o f learning at an automatic level. Such differences were obtained in the
left supramarginal gyrus, postcentral sulcus, medial parietal cortex, right cerebellum and
right frontal pole. O f these, the activation in left supramarginal gyrus was the m ost
striking, surviving at the level o f p < 0.05 (corrected). Unlike the explicit music reading
task, in which the Training Effect was characterized by a relative increase in the Trial
Effect across the two scanning sessions, the Training Effect in left supramarginal gyrus in
the implicit music reading task was characterized by a Trial Effect that was present only
after training. The Training Effect was independent o f any behavioural change in the
performance o f the feature detection task.
An explanation o f the activation change in supramarginal gyrus rests on the
following argument. During the course o f training, participants learned to make specific
keypresses in response to particular musical notes. The visual appearance o f musical notes,
post-training, may be automatically and unconsciously interpreted as an instruction to act.
One o f the earliest psychological studies o f music reading (Bean, 1939) was designed to
establish whether the musical notation evoked mental imagery o f one sort m ore than
another. Finding little evidence to support the idea that a musical note on the score can
elicit auditory imagery. Bean concluded.
“a note on the page means to most performers the act ofpressing a key, rather than sound”
For the purposes o f performing the feature detection task (up/dow n discrimination
o f a visual target), preparation o f the learned musical response would be inappropriate and
While the preparation and execution o f the task-relevant response was com m on across
both the pre- and the post- training sessions, the impHcit preparation o f a music-specific
m otor response would have only occurred post-training.
The left supramarginal gyrus is thought to be im portant in processes related to
"m otor intention". WhQe a network o f areas including the posterior parietal cortex o f the
right hemisphere have been demonstrated to subserve visual attentional processes
(Corbetta, 1993; Corbetta, 1998), imaging and transcranial magnetic stimulation (TMS)
studies have highlighted the importance o f m ore anterior parietal cortex o f the left
hemisphere for m otor orienting (Rushworth, Paus & Sipila, 2001; Rushworth, Ellison &
Walsh, 2001; Rushworth, Krams & Passingham, 2001), a hypothesis that is supported by
observations that patients with damage to left inferior parietal cortex have difficulty in
m otor sequencing (Harrington & Haaland, 1992; Kimura, 1993; Rushworth, Nixon,
Renowden, Wade & Passingham, 1997) and in the representation and awareness o f
movements (Sirigu, Daprati, Pradat-Diehl, Franck & Jeannerod, 1999).
T he involuntary effect o f musical literacy, as reflected in activation changes in the
implicit music reading task, is supported by the behavioural results o f these same
participants, described in experiment 2 o f Chapter 2 (section 1.1.2). Post-training, the
response time taken to perform an explicit task (mapping from num bers to fingers) was
increased when the numbers were superimposed on incongruent musical notes. Thus,
although no behavioural differences were seen, post-training, between the musical and
non-musical conditions o f the implicit music reading task, the training-related changes in
performance o f the Stroop task lend support to the notion that activation changes seen in
left supramarginal gyrus in the implicit music reading task may reflect response preparation.
by the fact that, in this task, both control and experimental trials required explicit
preparation and execution o f a m otor response.
Additional evidence that the presentation o f musical stimuli can elicit implicit
activation o f m otor-related cortical areas comes from studies within the auditory domain.
Expert pianists have been shown to have an involuntary increase in m otor cortex activity
when listening to piano pieces and performing a decoy task requiring detection o f a wrong
note in a piece o f familiar music (Haueisen et al., 2001). The decoy task was designed to
emphasise the perceptual rather than the production components o f musical processing,
thus ensuring that any motor-related activations were genuinely involuntary. Activation o f
primary m otor cortex during an explicitly perceptual task supports the idea that music
listening as well as music reading can implicitly affect the musical production system.
Activation o f supramarginal gyrus was also seen in one o f the subtractions o f
Sergent’s P E T study o f sight-reading (Sergent et al., 1992). Activations related to score
reading were subtracted from activations related to score reading whilst listening to a
performance o f the same music. This was interpreted to suggest that supramarginal gyrus
is involved in a visual-auditory mapping process. However, activation o f supramarginal
gyrus in the present study is not consistent with such a view since there was no auditory
component. An alternative explanation for the involvement o f supramarginal gyrus in
Sergent's study may be that score reading alone was not sufficient to activate m otor
preparatory systems but the conjunction o f musical information from visual and auditory
modalities may have been functionally suprathreshold for eliciting music production-
5 .