How far, if at all, can analogies with machines help us to assess theories of the relationship between mind and body?

How far, if at all, can analogies with machines help us to assess theories of the relationship between mind and body?

Physicalism is the view that everything that exists is physical, or depends on something that is physical. For a physicalist, the mind is not a substance or a thing. It is more accurate to talk of mental properties, mental events, mental states and processes. We can then say that these properties (etc) are possessed not by a mind, but by a person or brain, which are physical objects. In short, mental states are brain (physical) states, says the type identity theorist. Therefore, thinking a thought or feeling an itch is exactly the same thing as certain neurones firing.

This explicit relationship between mental and brain states, however, seems too confined. Hilary Putnam provides an objection to type identity theory saying that mental properties cannot be identical to physical properties because the same mental property can be realised by different physical properties. Type identity theory is not multiply realisable. For instance, the type identity theorist may say that being in pain is C fibres firing. According to Putman, this cannot be the case since the brain states that relate to pain are different in different species, but pain is still the same mental property. Therefore, ‘being in pain’ cannot be exactly the same thing as being in any one of these different physical states (C fibres firing) because a squid can have the mental state ‘being in pain’ even though it does not have C fibres. Thus, type identity theorists are guilty of chauvinism.

Token identity theory agrees that mental states are physical events, but rejects the claim that mental properties are identical to physical properties, thus, avoiding criticism from Putnam’s argument from multiple realisability. Instead, token identity theory maintains that while mental properties are not identical to physical properties, each individual occurrence (‘token’) of a mental property is identical with the occurrence of a physical property. So, the mental property ‘being in pain’ is not identical with any physical property i.e. C fibres firing, however, each time the mental property ‘being in pain’ occurs in a particular creature, it occurs with and because of the occurrence of some physical property. In a different creature or on a different occasion ‘being in pain’ may occur with and because of another physical property e.g. D fibres firing. Nevertheless, this mental property only ever occurs with and because of the occurrence of some physical property. By avoiding the problem of multiple realisability, token identity theory can be said to be a better theory of the mind than type identity theory. However, this leaves us with the problem that no brain state need correspond to any sort of mental state or behaviour whatsoever. This cannot even be tested for, since it may be argued that we can never have the same thought twice and, therefore, never the same brain state. Even though this criticism does not disprove the theory, it means that there is nothing to suggest beyond doubt that brain states are mental states. Though it is highly liberal, it seems that it is better to side with token identity theory than it is to side with type identity theory.
Functionalists did just that. Functionalism is usually understood to be a token identity theory, in that functionalism agrees that brain states are responsible for mental states, but disagrees that they are identical with them. For functionalists, mental states are states that exist between input, such as a stimulus experience, and output, such as behaviour. To characterise a mental state, we need to describe its typical inputs and outputs. Therefore, we can say that mental states are functional states.
The property of ‘having function x’ is a property that can occur in many different physical things. For example, ‘being a mousetrap’ is a functional property. There are several different mousetraps, built in different ways, using different methods and materials. Likewise, ‘being in pain’ is also a functional property, in that there are a lot of different physical ways, different brain states that could be ‘being in pain’. This might not just vary from species to species; it could vary from one individual to another, or even in one individual, from one time to another.
To help understand this idea, consider the usual functionalist example of a computer. Imagine that you ask a computer to add the numbers 3 and 7. On one level - at a low level - what is happening in the computer is dependent on the hardware; on another level - a high level - the computer's software is calculating the answer. Since, historically, computers have had different hardware that works in different ways; we cannot describe the process of calculation as the activity of hardware. Instead, the functionalist argues, the process of calculation is simply realised by the hardware. Therefore, the software is a function of the hardware. So, in terms of the brain, mental states are dependent upon brain states or neurological activities in the same way that the functions of a computer program are dependent upon computer hardware. With this computer model, brain states are analogous to the hardware and mental states to the software. Functionalists say that brain states help to realise mental states and that the mind is, in fact, a total functional system. By this, they mean that a mind is made up of the total possible functional (mental) states that it can have and that these interact with the inputs and each other in order to produce outputs or behaviour.
The functionalists’ use of a computer model which describes the mind as a multiply realisable total functional system lends itself to arguments for Artificial Intelligence (AI). AI is a research project in computer science that aims to create computers that display behaviour that is intelligent. Some philosophers and scientists argue that a test for whether a computer is intelligent is the ‘Turing Test’. A person, a computer and another person (the interrogator) are each in a different room. The person puts the same questions, in turn, to the person and computer, not knowing which is which. If, after five minutes, the interrogator cannot tell from his conversations with the person and computer which is which, then the machine has passed the Turing Test.
Whether the Turing Test is a good test for intelligence is very controversial. Some say that all the test really does is expose how little we know about what makes people people. The fact that it's easy to deceive people into thinking a computer is human doesn't actually teach you anything about the difference between computers and humans; we still feel as if there is something more to humans. Some say that the something that computers lack is consciousness, they cannot be genuinely intelligent.
John Searle argues that Intentionality is not reducible to functions. Many mental states are ‘about’ something, objects or events in the world. For example, I might have a belief about Paris, a desire for chocolate, be angry at the government, or intend to go to the pub. In all of these cases, my state of mind is directed towards and object, the thing I am thinking about. This idea of directedness is known as Intentionality. To illustrate his argument, Searle describes a room with two holes in the wall; through one, pieces of paper are passed in (inputs), through the other, pieces of paper are passes out (outputs). There is someone in the room who has to respond to the inputs by sending the outputs. The inputs are questions in Chinese; the person doesn’t understand Chinese, but has a rule book which correlates every question with an answer. The person finds the output that is that answer, and sends the piece of paper out. The room as a whole- a system- ‘behaves’ as if it understands Chinese. But it doesn’t; the person doesn’t, the rulebook doesn’t, the room doesn’t. Even if the person memorised the rulebook, he wouldn’t understand Chinese, says Searle. The person would not know what the questions mean. Searle argues that this is what real Intentionality requires. Yet the room performs the same functions as someone who does understand Chinese. So performing functions isn’t enough for understanding meaning, for real Intentionality.
Some functionalists have rejection Searle’s conclusion: the room does understand Chinese. To illustrate their point, functionalists develop Searle’s thought experiment: the person inside the room is replaced by a computer that does the same thing. Suppose we then put the computer inside a robot, which interacts with the world. We add a programme for correlating visual input (through a camera in its eye) to output. So, the robot is now capable of naming things in Chinese, as well as answering questions in Chinese. Functionalists ask if it is still obvious to conclude that the robot doesn’t understand Chinese.
Searle would still say that it doesn’t. Artificial intelligence isn’t intelligence at all; at best it is a simulation of intelligence. The Chinese room performs the function that a human would perform. Likewise, other computers demonstrate a range of mind capacities such as monitoring and controlling bodily processes, purpose, logical thought, memory, the ability to learn and so forth. However, although, the simulation of Intentionality is convincing, Searle maintains that there if a distinct difference between authentic Intentionality and the simulated kind. This difference is consciousness. Without consciousness, a series of functional interactions remain meaningless to the robot, or computer, even though they look meaningful to us. But intentional states are meaningful from the inside: they are meaningful to the creature that has them.
Though analogies with machines cannot help us to assess theories of the relationship between mind and body since machines appear to lack consciousness and consequently intentionality, they do highlight the similarities in which humans and computers produce certain outputs. For example, most people perform long division by following rules, where the only intentional aspect if knowing that the rules will provide the answer to a question about how many times one answer ‘goes into’ another.