all propositions about mental states can be reduced without loss of meaning to propositions that exclusively use the language of physics to talk about bodily states/movements.
Hard behaviourism was created by logical positivists like Carnap and Hempel. They applied the verification principle to the philosophy of mind. They argued that because mental states are private and so unverifiable, mental terms (linguistic attempts to describe mental states) can only be meaningful if they apply to something public and verifiable like behaviour. So hard behaviourists analytically reduce mental terms to behavioural terms.
propositions about mental states are propositions about behavioural dispositions (ie propositions that use ordinary language).
Ryle’s soft behaviourism claims that mental terms can be analysed into behavioural dispositional terms. So it is not a reductive theory nor does it try to tie mental terms to particular individual behaviours, but to a ‘disposition’ to behave in a certain way.
How someone behaves might not be what their disposition to behave was. They might have had a disposition to behave in a certain way but other factors caused them to behave differently. For example if we know someone has a disposition to eat messily but didn’t in a certain situation. We might figure out why by analysing the situation – perhaps there was someone there they were trying to impress by being mannered. We could observe whether they act more polite than they usually do in other situations also involving that person.
In other words, someone’s behavioural disposition can only be figured out by observing a lot of their behaviour in many contexts.
The distinctness of mental states from behaviour (including Hilary Putnam’s ‘Super-Spartans’ and perfect actors)
Hard Behaviourism attempts to reduce mental terms like ‘pain’ to behavioural terms like ‘wincing’. However, what if someone decided not to display the pain behaviour of wincing? Then there would be pain with no pain behaviour – a mental term without a behavioural term to reduce it to. However the hard behaviourist claims all mental terms can be reduced to behavioural terms.
Ryle’s soft behaviourism doesn’t have this problem because Rlye could still claim the person feeling pain still has a disposition to pain-behaviour, but for some reason it hasn’t resulted in behaviour. So even if there is no pain-behaviour, there is still a behavioural disposition into which Ryle can analyse the mental term of ‘pain’ into. So soft behaviourism still works.
Putnam argued that if a race of ‘super spartans’ existed with a culture that involved never displaying pain, then they wouldn’t even have the disposition to pain behaviour. However, they would still have pain. Therefore there would be a mental term which cannot be analysed in terms of behavioural dispositions. Therefore soft behaviourism is false.
Ryle responds that in such a culture the term ‘pain’ could never be learned because there would be no observable behaviour for people to learn from. Such a culture could therefore not exist.
Putnam responds that it’s logically possible for the super-Spartans to be born ‘fully enculturated’.
Issues defining mental states satisfactorily due to (a) circularity and (b) the multiple realisability of mental states in behaviour
Hard behaviourism claims that mental terms are analytically reducible to behavioural terms. So the mental term ‘scared’ can be analytically reduced to the behaviour ‘running away’.
However while being scared might cause person X to run away, it might cause person Y to attack. So the mental term ‘being scared’ can be multipally realised in different behaviours. But if ‘scared’ reduces to ‘running away’, it can’t also reduce to ‘attacks’, therefore hard behaviourism is false.
Similarly, one behaviour could multipally realise different mental states. X might run towards something to give it a high five, Y might run towards something to cuddle it.
Soft behaviourism argues that mental terms are analysable in terms of behavioural dispositions. Ryle argued that dispositions are not tied to individual behaviours but require observation of multiple behaviours in multiple contexts to determine. So, he would argue you could ask the person running towards something why they are, or study their facial expressions, or observe what they do when they have reached it and take that into account in determining what their disposition is. So, because Ryle doesn’t reduce mental terms into particular behaviours, he doesn’t have the issue of multiple realisibility
Ryle’s expansion of behaviourism by bridging mental terms to dispositions creates a problem of its own, however. It might appear obvious to bridge the mental belief that it’s raining with the disposition to carry an umbrella, for example. However, the bridge between that belief and that disposition assumes various other mental states, such as the desire not to get wet. Trying to bridge that mental state to a disposition will however simply make more assumptions of the existence of other mental states which will themselves require bridging to a disposition, and so on forever. Ryle attempts to make dispositions a more expansive loose basis for behaviourism than mere instances behaviour, but in the process creates an infinite chain of explanation and so can’t have a final analysis. Because of this, the mental states he is attempting to analyse will always be analysed into something which also involves mental states, making it a circular analysis. It’s like explaining X with an explanation that involves X. That is not a valid explanation or analysis.
The mind is finite so presumably this chain of bridging will end at some point.
The asymmetry between self-knowledge and knowledge of other people’s mental states
Both types of behaviourism claim that we gain knowledge about a mind by observing behaviour. Since everyone’s behaviour is public, including my own, my knowledge of my own mind should therefore be symmetrical, meaning equivalent, to my knowledge of other people’s mind. However I know more about my own mind than I do others. So my knowledge of other mind and knowledge of my own mind is asymmetrical therefore behaviourism is false.
Ryle responds by agreeing that there is asymmetry but disagreeing that that proves behaviourism false. He argues the asymmetry exists because we observe more of our own behaviour than that of others.
Arguably Ryle misses the point of the objection which is that we gain our self-knowledge through introspection, whereas I cannot use introspection to gain knowledge of the minds of others. So Ryle might be able to explain the asymmetry regarding the total amount of knowledge, but he cannot explain the asymmetry of the ways of gaining knowledge that are available for my own mind compared to the minds of others.