Synopsis

1. Consciousness

1.1 Approaches to understanding the nature and implications of 'consciousness' prior to the domination of the computer model of the brain/mind:

1.11 Cartesian dualism.

1.12 Logical behaviourism.

1.2 Approaches to the question Can purely physical systems - eg computers - experience 'consciousness'?

1.21 Jackson argues that there are qualia (and so defends epiphenomenalism).

1.22 Jackson's target: the identity theory (of consciousness) articulated by Place.

1.23 Machine functionalism offers an understanding of 'consciousness'.

It thinks of the brain on the analogy of a computer running a program. At any one time, a certain step in the program will be being calculated. It is this step - a 'calculation' in effect - that is identified with conscious experience. (Though there is no suggestion that all the steps of a program as it runs are experienced.

1.24 Teleological functionalism offers another understanding of 'consciousness'.

Staying with the machine running a program' analogy, teleological functionalism holds that a phenomenal experience is to identified with the biological function being performed by a particular subroutine of the program running on the brain.

Concepts include

• software
• program
• 'levels' of description
• different hardware platforms running the same program
• type/token distinction

2. Instrumentalism

We ordinarily explain our own and others' behaviour in terms of beliefs and desires. (This type of explanation is dubbed 'folk psychology). But what is a belief? What does a person who takes the computer/brain analogy seriously think a belief might be?

Dennett defends the idea that when we explain someone's behaviour in terms of belief and desires we are not invoking internal states. We are invoking theoretical constructs which have been elaborated because of their explanatory power.

According to his terminology, when we try and explain the behaviour of a 'system' - eg a person - by attributing to them beliefs and desires, we are adopting towards them the 'intentional stance'. As far as some 'systems' are concerned, the best way to predict their behaviour is to adopt the intentional stance towards them - to attribute to them beliefs and desires. Humans are often systems of this kind. Sophisticated computer systems may be also.

Dennett thinks that great swathes of human behaviour can successfully be predicted by taking up towards human beings the intentional stance. He says this must be because there is a pattern to human behaviour which we pick up on and exploit with our explanations invoking beliefs and desires.

According to this view, the test of whether a system should have beliefs and desires attributed to it is whether this helps predict its behaviour, not whether there are internal states (states of the brain, or states of consciousness) of any particular kind in it.

It leaves room for attributing beliefs and desires to computers, if they developed behaviour of the requisitely sophisticated kind - so bypassing any question of whether they could be 'conscious'.

3. Language

Might machines to be got to use language as human beings do?

3.1 Understanding is more than manipulating uninterpreted symbols.

John Searle presents what is called the Chinese Room argument designed to show that there is more to understanding a language than manipulating uninterpreted symbols. (It is often claimed that computers achieve their effects by manipulating uninterpreted symbols.)

What objections might be made to Searle's argument? How good are they?

To respond to Searle, you have to think of the internal states of physical systems (eg brains) acquiring meaning. How can they be thought of as doing so? Jerry Fodor offers one possible answer in "A Theory of Belief". The Churchlands offer another in 'Stalking The Wild Epistemic Engine'.

3.2 A distinction between syntax and semantics is often made.

(Early) Chomsky's computational understanding of 'grammar'.

3.3 The representational theory of thinking

According to the computer/brain analogy, what is thinking? The repesentational theory of thought is one answer: what is going on when a person is thinking is that states of bits of the brain represent propositions and these are manipulated in various ways.

There is another theory, which is the one sketched by the Churchlands in 'Stalking the Wild Epistemic Engine'. They argue that there a good deal of 'thinking' which cannot plausibly be thought of as involving representations of propositions. They think it involves representations but not representations of propositions.

3.4 The 'language of thought'

It is argued that there are reasons for rejecting the Churchlands' view and for insisting that thinking must be quasi-linguistic. Fodor's 'Why there has to be a language of thought' is the central paper here.

 

4. Intentionality

Any theory of thinking has to account for the fact that thoughts seem to 'point' - possess intentionality.

What is intentionality? How can we envisage physical systems such as computers (or bits of them) acquiring it?

Besides thoughts, language items (eg sentences) seem to have intentionality.

What is the connection between the fact that language items have intentionality and the fact that thoughts have intentionality? - The thesis that it is thinking that confers intentionality on language items (eg by uttering them with something in mind).

But then where does the intentionality of thought come from?

The idea that physical systems like brains (or bits of them) acquire intentionality through their sense experience, and/or through evolution.

En passant:

• Brentano's thesis that consciousness=intentionality.

5. Connectionism

Maybe the brain is not a serial (von Neumann) computer, but a computer of a different type - a connectionist net, or network of connectionist nets.

What is a connectionist net?

What objections are there to thinking of the brain as a serial computer?

The representational theory of thinking assumes a von Neumann type computer model of the brain, so to argue that the brain is not a computer of this type attacks this theory of thinking.

En passant:

• Does the connectionist net have light to throw on generality?
• How goal-seeking is programmed into a computer

6. Eliminativism

This is the thesis that explaining our own and others' behaviour in terms of beliefs and desires - folk psychology - is completely mistaken and that the advance of neuroscience will clarify what concepts we should employ instead (to give us better predictions). Folk psychology it is claimed throws little or no light on many features of human behaviour. It will be replaced by neuroscience just as Aristotelian physics gave way to Newtonian (which has of course itself given way since).

Is folk psychology a theory?

7. The status of Folk Psychology

It is argued that beliefs (which form a central part of Folk psychology) are made what they are partly by the context of the believer. Putnam argued the same point with his water / twater scenario.

If established, this point would threaten to undermine both

• folk psychology, in the sense that it implies that you can't tell what your beliefs are just by introspecting, and
• the representational theory of thinking, which holds that beliefs are or are grounded in brain states (or states of bits of the brain).

Stich points to a widespread assumption about psychology: that it must explain human behaviour solely in terms of properties which are grounded in physical states of the brain (dubbed 'the principle of psychological autonomy').

VP

Updated 24:03:03