Minds and Computers : An Introduction to the Philosophy of Artificial Intelligence

particular mental states along the following lines. To be in a particu-
lar mental state is to be in a state which is apt to be caused by certain
stimuli and apt to cause certain behaviour.
Providing a definitional taxonomy of mental states is then simply a
matter of identifying the stimuli which are apt to cause each particular
mental state and the behaviour each is apt to cause. To be in pain, for
instance, is to be in a state which is apt to be caused by, inter alia,
burning my hand and is apt to cause, inter alia, the removal of my hand
from the source of heat and further hand-discriminating behaviour.
The causal theory conserves, from its philosophical predecessor,
the intuition that there is a crucial connection between mentality and
behaviour. It does not, however, make the problematic identification
between mental states and dispositions to behave. To be in a mental
state is not, for the causal theorist, just to be disposed towards certain
behaviour. Rather, to be in a mental state is to be in a state which is
36
  

---

apt to stand in certain causal relations mediating stimulus and behav-
iour. The causal theorist is not an eliminativist about mental states.
The obvious next point of inquiry is to determine precisely which
states are apt to have these causal properties. In other words, now that
we have characterised the role of mental states by giving a conceptual
analysis of mental state terms, it is time to locate the occupants of
these roles by making substantive identifications.
The causal theory, as it stands, is ontologically neutral – it does not
commit us to any particular ontology. One could, for instance, be a
causal theorist but still maintain that the things which occupy the
roles of mental states are immaterial. This would, of course, require
an unusual account of causality but we have already seen that this is
a problem for the dualist.
Although the causal theory is, strictly speaking, ontologically
neutral, the talk of causality does pave a fairly obvious path to a
material identification of mental states. This is precisely what is pro-
vided by Australian materialism.
5.2 THE IDENTITY THEORY
Australian materialism rose to prominence in the late 1950s with the
publication of two very influential papers: Place’s Is Consciousness a
Brain Process? (1956) and Smart’s Sensations and Brain Processes
(1959).
Australian materialism makes strict identifications between types
of mental states and types of neural states. In other words, to be in a
certain type of mental state just is to be in a particular type of neural
state.
This is an analysis of mental states that aims to provide an interthe-
oretic reduction. Types of mental states, according to the Australian
materialists, smoothly reduce to types of neural states. They do not
hold to eliminativism with respect to mental states but, rather, seek to
make scientific identifications of the correct referents of our mental
state terms.
So the Australian materialist does not believe, as did the behav-
iourist, that we are simply mistaken in using mental state terms as if
they refer to substantive entities. Where the behaviourist held mental
state terms to be akin to terms like ‘witches’ and ‘phlogiston’ – to be
shown to be mere ‘folk’ terms by the progress of scientific discovery –
the Australian materialist holds that mental state terms are akin to
terms like ‘lightning’, to be identified with physical phenomena in
accordance with our scientific theories.
 
37

---

At this point, the various names by which the theory is known
should make a lot more sense. It is a reductive materialist theory that
makes type–type identifications between mental states and certain
physical states – namely, neural states, or states of the central nervous
system – hence the appellations ‘reductive materialism’, ‘identity
theory’, ‘type physicalism’ and ‘central state materialism’.
One clear advantage of the theory is that it provides a solution to
the problem of other minds. We can tell whether other people actually
have mental states simply by investigating their brains. Having a type
of neural state, on this analysis, just is having a type of mental state,
so other minds are readily identifiable and empirically amenable.
Another clear advantage of the theory is that it confers a scientific
methodology for investigating mentality. If we want to know about
the mind, we should do neuroscience. In particular, we should seek to
determine which types of neural states obtain as which types of
mental states.
As a materialist – or physicalist – theory, Australian materialism also
satisfies Ockham’s razor. The Australian materialist admits only phys-
ical substance into her ontology. At least, qua Australian materialist,
this is the case – she may well have other reasons to expand her ontol-
ogy, but these won’t be reasons which pertain to her theory of mind.
A final selling point lies in the theoretical fit with the causal theory
of mind. To the extent that one holds that the causal theory is a
correct analysis of mental states, one finds an advantage in the provi-
sion, by Australian materialism, of candidates that are apt to have
precisely the causal powers held to be characteristically defining of
mental states.
If we, then, marry the conceptual analysis of the causal theory with
the substantive identification of Australian materialism, we get the
following account of mental states. To be in a type of mental state is
to be in a type of neural state which is apt to be caused by certain
stimuli and apt to cause certain behaviour.
Despite the numerous advantages of Australian materialism, there
are, as always, a number of philosophical objections we can mount
against the theory.
5.3 ARGUMENTS AGAINST AUSTRALIAN
MATERIALISM
Let’s begin with some fairly weak objections to Australian material-
ism. We might argue that we have the capacity to introspect our mental
states and that when we do so, we learn about our mental states. We
38
  

---

don’t, however, learn anything about our neurophysiology through
introspection, so mental states can’t be identical to neural states.
There is a clear reply to this objection. It straightforwardly begs the