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

physical substance.
As before, the failure of the argument from essential properties
does not entail that minds cannot be non-physical – it merely shows
that a radical distinction in properties (essential or otherwise)
between minds and canonical physical objects is not su
fficient to
establish that minds are non-physical.
2.4 ARGUMENTS AGAINST CARTESIAN DUALISM
We have now seen three arguments in favour of dualism. The argu-
ment from introspective appearance and the argument from essential
properties both seek to establish a broad mind-body dualism. Coupled
with certain common-sense intuitions concerning the e
fficacy of the
mind in bringing about changes in the body and vice versa, they
become arguments supporting Cartesian dualism. The argument from
religion seeks to establish Cartesian dualism in particular, as the inter-
action between mind and body is essential for the argument.
We should already be concerned for the theory, given that we have
actively sought the strongest arguments in its favour and have dis-
covered that none of them succeed in establishing their conclusions.
Even more troubling for the Cartesian dualist though are the follow-
ing negative arguments.
2 . 4 . 1 T H E P R O B L E M O F OT H E R M I N D S
The first objection to Cartesian dualism we will consider identifies a
problematic consequence of the view.
We have become quite adept at investigating the physical universe
and have all manner of methods and equipment at our disposal for
doing so. We are at a loss, however, when it comes to investigating the
non-physical.
If minds are immaterial, then they are clearly not investigable by
known empirical methods. Not only does this put minds beyond the
scope of science, it also means that there is no way to know whether
or not other people have minds. As far as the Cartesian dualist is able
to discern, she may well have the only mind in the universe – all other
human bodies may well just be mindless automata.
This is an epistemological concern – a concern about what we can
know – which comes with a methodological concern for the possibil-
ity of a science of mind. The objection is not insuperable, however.
The Cartesian dualist can help herself to a reply, of sorts, to each of
these concerns.

9

---

With respect to scientific methodology, she might point out that it
is not unknown to science to postulate, and investigate, unobservable
entities by examining their observable consequences. So, while she
will have to maintain that minds are simply not amenable to direct
empirical investigation, she can hold out hope that there will be
observable consequences of mentality that can be investigated,
thereby giving science indirect access to minds.
With respect to our everyday knowledge of the minds of others, the
Cartesian dualist can reason by analogy to her own mental life and its
role in mediating experience and behaviour. Presumably – she might
say – you do, in fact, think that other people have minds (it is fairly
di
fficult to get around in the world without proceeding on that
assumption). Why then do you think this? Presumably because you’ve
observed that the best explanation for the way other people behave
involves attributing mental states to them.
In other words, you know that if you have certain experiences, this
will lead to certain beliefs and desires (mental states) which in certain
situations will lead you to behave in particular ways. You’ve further
observed other people in just such situations acting in just such ways
and consequently assume that they share certain of your beliefs and
desires (like the belief that it is lunchtime and the desire for food) and
recognise that these mental states play an important explanatory role
in understanding their behaviour.
This reply to the problem of other minds appeals to an inference to
best explanation: the best way to explain the way other human bodies
move around in the world is to attribute to them the kind of mental
states I know that I have. It should be apparent, however, that while
this reply su
ffices to demonstrate the utility in assuming that other
people have minds (the assumption confers useful predictive capac-
ities), it certainly does not establish that they do. The problem of other
minds remains for the dualist.
2 . 4 . 2 O C K H A M ’ S R A Z O R
William of Ockham was a medieval philosopher and notable logician
of the early fourteenth century. You may well have heard a common
corruption of Ockham’s razor that is something along the lines of
‘the simplest explanation is often the best’. Properly construed,
Ockham’s razor is intended to serve as a methodological constraint
on theory construction.
The most accurate gloss of Ockham’s razor in the realm of meta-
physics is ‘don’t expand your ontology beyond necessity’. Another
way of putting it is to say that one shouldn’t postulate any more
10
  

---

entities than are absolutely necessary to explain the phenomena about
which we are theorising.
This can be deployed as a methodological objection to Cartesian
dualism – the contention being that the dualist does, in fact, expand
her ontology beyond explanatory necessity, that postulating non-
physical entities is not required in order to explain mentality.
This a moderately weak objection so I shall give it short thrift. At
best – if you think that the principle should constrain theory con-
struction – it entails that when presented with two explanatorily ade-
quate theories of mind, one of which postulates non-physical entities
and one of which accounts for mentality in purely physical terms, one
should prefer the latter. This will be something to bear in mind once
we have surveyed the space of available theories of mind.
2 . 4 . 3 T H E P R O B L E M O F I N T E R AC T I O N
A considerably more potent objection – one which is generally con-
sidered to be the rock on which Cartesian dualism founders – centres
on the problem of interaction.
The physical universe is held to be causally closed, which means
that every physical e
ffect has a physical cause. A physical effect
brought about by a non-physical cause would contravene the first law
of thermodynamics. While science has certainly got it wrong about
many things in the past, our theory of thermodynamics is a found-
ational theory which most of modern science rests on.
The problem here for the Cartesian dualist, if it is not already
apparent, is their contention that the non-physical mind is causally
e
fficacious in the physical world, that the non-physical mind causes
change in the physical body.
What might the Cartesian dualist say to the problem of interac-
tion? The only possible response seems to be to deny that the physical
universe is, in fact, causally closed. This, however, seems rather
implausible. Were it the case that our physical actions were caused by
non-physical minds, then energy would be added to the physical uni-
verse every time a mental action resulted in a physical action and this
addition of energy would, one might think, be measurable.
There is a theistic response available here, which is to claim that in
every case of scientific observation, an omniscient, omnipotent divin-
ity intervenes and adjusts the observer accordingly, such that we think
that energy is always conserved and that the amount of energy in the
physical universe is constant, but in fact it is constantly increasing.
Taking such a line, however, brings with it a raft of troubling epistem-
ological concerns.

11

---

There seems to be no secular way to rescue Cartesian dualism from
this objection. We can, however, advance modified forms of dualism
which retreat from the commitment to interaction.
2.5 OTHER DUALISMS
Recall from section 2.2 the four propositions [D1]–[D4] which charac-
terise Cartesian dualism. One way to recover the core ontological intu-
itions of Cartesian dualism from the damning criticism of the problem
of interaction is to give up the commitment to propositions [D3] and
[D4], leaving us in want of an account of the relation between the
physical body and the non-physical mind. This strategy leads to the
theistic dualist theories known as parallelism and occasionalism.
Another possible strategy is to give up only [D4] and maintain
a commitment to propositions [D1]–[D3]. Again, this requires a