554 Karen Neander
or a memory of last night’s dinner.
Just as the notion of function is a peculiarity of biology relative to the more
basic natural sciences, the notion of a representation seems to be a peculiarity of
psychology (and the social sciences) relative to biology and the more basic natural
sciences.
On the Cartesian picture, the Great Divide between the intentional and non-
intentional sciences is thought to have an ontological basis, because on that picture
bodies and brains are physical systems and minds are not, and so little continuity
is to be expected between the natural sciences, which aim at explaining the for-
mer, and psychology and the social sciences, which aim at explaining the latter.
With the demise of Cartesian Dualism, the case for integrating psychology with
the natural sciences seemed to improve. According to Ontological Behaviorism,
mental terms have a behavioral analysis. And according to Reductionist (type
type) Physicalism, psychological properties are neurophysiological properties, and
bridging the divide is merely a matter of discovering these identities. However,
neither behaviorism nor reductionist physicalism remained very popular for long.
With the arrival of the view that is commonly (but in this context a little
misleadingly) known as Functionalism, the Great Divide reappeared, as if out of
a mist. Functionalism consists of two main theses. The first is that mental states
cannot be reduced to brain states because mental states are multiply realizable and
the second is that mental states can be characterized in terms of their functional
role. Pain in humans might be realized by a neurophysiological state such as (toy
example) C-fibers firing. But, says the functionalist, this does not preclude other
kinds of creatures that lack C-fibers from experiencing pain: Pain in octopi or pain
in Martians might be realized by different kinds of physical processes. What pains
have in common, on this view, is their characteristic disposition: Pain is the kind of
state that is characteristically caused by bodily damage and that characteristically
causes certain other specifiable inner states and motor outputs.
What “characteristically” means is not usually made explicit, but the kind of
function talk that is used in Functionalism is not usually understood as talk of
normative and teleological functions. Such talk is usually understood as talk of
what are sometimes known as CR functions, about which more shortly.
Functionalism, as it is standardly understood, is consistent with physicalism,
but it is also consistent with the Great Divide, especially since intentional mental
states are expected to be even more diverse in their realizations than phenomenal
states (like pain) are. Not only might a desire for peace be realized by different
states in humans and in aliens from Alpha Centauri, but it might also be realized
by different states in different humans, or by different states in the same human,
as a child develops into an adult, or perhaps even on different occasions, from
one day to the next. And just as we can study computer programming and data
systems without studying hardware engineering, we can study human psychology
without studying human neuroscience, or so the reasoning goes. In fact, it is
argued that we must preserve the autonomy of cognitive science, because it is only
if we abstract away from the diverse details of neurological realizations that we