hope of understanding them on the hypothesis that the functional or-
ganisation of the mind is a hierarchy of modules.
3.3 Brain-scanning evidenceOur detailed understanding of how modules develop and function will
certainly be enhanced in the future by greater use of brain-scanning
techniques. But as yet this is not the most important source of evidence for
modularity. Partly this is because there are limitations on the tasks which
patients can perform while their brains are being scanned (their heads may
have to be motionless, for example; and obviously they cannot be playing
football!). But a more important problem for all scanning techniques is
that the resulting pictures of neural activation are always produced by a
subtractionof background neural activity. (A raw picture of brain activity
occurring at any one time would just be amess, with activation of many
diVerent areas around the cortex; for there is always too much processing
going on at once.) First a scan is taken while the subject is performing the
target activity (listening to a piece of text, say), and then a further scan of
the same subject is taken in which everything else is, so far as possible, kept
the same. The latter is subtracted from the former to obtain a picture of
those brain areas which are particularly involved in the target activity.
It is obvious that brain-scanning can only begin to be useful, as an
experimental technique, once we already have a set of reasonable beliefs
about the functional organisation of the mind and the modular
organisation of the brain. For otherwise, we cannot know what would be
appropriate to choose as the subtraction task. (For example, should the
subtraction task for text-comprehension be one in which subjects listen
to music, or rather one in which they receive no auditory input? The
answer will obviously depend upon whether speech-comprehension and
musical appreciation are handled by distinct systems; and upon whether
any ‘inner speech’ in which subjects might be engaging in the absence
of auditory input would implicate the speech-comprehension system.)
But as our knowledge advances, brain scans seem likely to prove an
invaluable tool in mapping out the contributions made by diVerent brain
areas to cognitive functioning.
One further caveat to be entered about what brain-scanning can reveal
concerning modularity, is that one must be wary of assuming that a
module will always be located in a speciWc region of the brain. This is
because the notion of a module is itself essentially functional. What a
moduledoesis more vital thanwhereit gets done. Indeed, to philosophers
of mind brought up on orthodox functionalist accounts of the mind, such
as ourselves, it is somewhat surprising that cognitive functions should turn
Developmental rigidity and modularity 61