next up previous contents
Next: Philosophy Lecture 2 Up: Philosophy Tutorial Previous: Mind-body dualism   Contents


Only matter exists. The mental is not distinct from the physical. But the question remains of how the phenomena of the mind are to be explained: sensations, perceptions, memories, beliefs, desires, thoughts, intentions, emotions, imaginings.

(Earliest recorded materialist - Thales of Miletus 6th C B.C. - all things are composed of water. Further developed by the Atomists - Leucippus of Miletus and Democritus of Abdera, in 5th and early 4th C B.C. respectively)

1. Eliminative materialism There are no such things; the terms are nonsensical, outmoded or scientifically fruitless - they should be eliminated.

2. Reductive materialism The above phenomena do exist, but they are only complicated forms of matter in motion.

a. Behaviourism: These ``mind'' terms refer to behaviour. Mental states should be analysed in terms of environmental stimuli and behavioural responses. There are two kinds of behaviourism:

(i) Methodological behaviourism
(ii) Logical or de-facto behaviourism

(i) Methodological behaviourism: Behaviourism is the only valid method; the only fruitful way to study psychology. This was the position implicitly adopted by most psychologists. Mental states may sometimes not be reduceable to behaviours but they are merely ``epiphenomenal'' - a byproduct of physical events - playing no causal role themselves and so irrelevant to behaviour.

(ii) Logical or de-facto behaviourism: It is the semantic version of the radical behaviourism of J.B. Watson which (according to Fodor) says that behaviour does not have mental causes. Behaviour is just the observable response of the organism to stimuli. The role of psychology is to catalogue the laws that determine the causal relation between stimuli and responses. But eg. in S carries an umbrella because it is cloudy, what about intervening mental events like observation and expectation? As psychology has matured, the framework of mental states needed to account for behaviour has grown more, rather than less, elaborate.

Logical behaviourism (paralleling logical positivism) developed in the early '60s out of dissatisfaction with both dualism and radical behaviourism. It holds that attributing a mental state (say thirst) to an organism is the same as saying the organism is disposed to behave in a certain way (eg. to seek or to drink water). One needs an open-ended, perhaps an infinite, set of behavioural hypotheticals to spell out the behavioural disposition expressed by the mental term.

Alexander Bain in the 19th Century (see sec. [*]) held that beliefs were behavioural disposition. More recently dispositional states were proposed by Gilbert Ryle (1949) The Concept of Mind.

In the physical sciences too dispositional terms like ``fragile'' are used but the manifestation of the disposition (the breaking of a fragile glass) always involves event-event causation. In logical behaviourism there is no equivalent of such causation (since mental causes - desires, beliefs ... are disallowed) so the notion of causation remains less robust than in the physical sciences. Mental states remain mere heuristic devices: the only facts allowed by behaviourism are to do with S-R relations.

b. Central-state identity theory (CSIT): Emerged around the same time as logical behaviourism. Mental states and activities are identical with bodily states (usually, the central nervous system (CNS)). Being in a certain mental state is identical to being in a certain physiological state. There are two kinds of central-statism:

(i) Methodological central-statism: the only useful way to study psychological phenomena is to study the CNS.

(ii) Logical central-statism: terms referring to mental states are (or would eventually be shown to be) synonymous with neurological terms. Thus one day eliminative materialism would be achieved.

By default in philosophy b. is taken to refer to b.(ii)

Fodor's account in Scientific American, Jan. 1981:

Identity theory does not hold as synonymous the statements John has as headache and John is in such-and-such brain state. Rather it says that these statements are rendered true (or false) by the same neurophysiological phenomena. (i.e. that something more than neuron activity is implied by the first? or both? of these statements?)

Identity theory could be held either as a doctrine about mental particulars (John's current pain or Bill's fear of animals) - this is token physicalism - or as a doctrine about mental properties (having pain or being afraid of animals) - type physicalism.

* Physicalism: Usually intended as a metaphysical thesis, it says that everything is physical or it supervenes on the physical. Used when materialism is thought of as too restrictive: it includes forces, energy etc. in addition to matter. Its main application has been in the philosophy of mind. (Seems a highly problematic concept, is avoided by many. I too, departing from Fodor, avoid it henceforth.)

Token identity says that particular or token psychological events are to be identified with token physical events. Token identity asserts only that all mental particulars that happen to exist are neurophysiological. It does not rule out the possibility of machines and disembodied spirits having mental properties.

Type identity holds that each kind or type of psychological state is to be identified by a type of physical state (for the identity theorist) or behavioural event (for the behaviourist). Type CS identity asserts that mental states can be reduced to, or replaced by, brain states, implying that all mental events that could possibly happen in any system have a neurophysiological counterpart. Since neither machines nor disembodied spirits have neurons, they could never have mental properties. Type identity requires mental properties and neurophysiological processes to be identical.

Problems for type identity: (For next 2 examples see Stillings et. al. pp. 326-327). John and Bill both remember eating vanilla icecream. By CSIT (type identity), they should share the same brain state. But suppose (extreme example) John had lost the left half of his brain and Bill the right half of his, in an early accident, and due to brain plasticity they had recovered with little loss of function. How could the memory they form later of vanilla icecream be identically represented in their brain? (Here CSIT fails but logical behaviourism prevails - Bill and John would share the same behavioural dispositions.)

Type identity in the context of behaviourism shows a contradiction in the following case: Mary and Sue share a desire for a pet unicorn, so they should be expected to show similar behavioural dispositions. But suppose Mary is painfully unassertive and lacks resources, while Sue is a brash city girl ... Their behaviour would be quite different. How could they be said to share the same desire? (Here CSIT prevails - subject to neurological verification? - it allows Mary and Sue to share a desire.)

Token identity is a weaker but more acceptable doctrine. It leaves open the possibility that there could exist information processing systems with the same psychological properties as human beings but not the same physical organisation. Accepting this doctrine calls for a relational account of mental properties that abstracts them from the physical structure of of their bearers. But such a relational account is not allowed by CS identity theory.

Logical behaviourism does offer a relational account of mental properties (in terms of dispositions). It also allows other info-processing systems (though in type identity): if a headache is characterised by a certain pattern of stimulus and response, then another system exhibiting the same S-R pattern would also be said to have a headache.

Functionalism agrees with behaviourism that psychological states are (at least partly) defined in terms of the input and outputs and that they are independent of the particular physical realisations. It also agrees with C-S identity theory that psychological states are correlated with physiological states. It differs from both logical behaviourism (common form) and CSIT in subscribing to token rather than Type identity.

Functionalism allows a mental state to be defined in terms of other mental states.

Theory Mental state is eqt. to Type/token identity Other ip Ss  
LB Behavioural disposition B T yes  
CSIT Physiological state P T no  
F B (relational) + P (causal) t yes  
  + other mental states M      
  (relational and causal)      
  Leaves the details to further research      

A Functional explanation: A headache is identified with a mental state that, for example, causes a disposition for taking aspirin in people who believe aspirin relieves headache; causes a desire (in non-masochists) to rid oneself of pain; often causes English-speakers to say ``I have a headache, etc.''; is brought about by overwork, eye-strain and tension ... the list can be added to indefinitely.

next up previous contents
Next: Philosophy Lecture 2 Up: Philosophy Tutorial Previous: Mind-body dualism   Contents