More isolationism ...

Inasmuch as we've concretised abstractions (mind/body, quality/quantity ...) and literalised metaphors (inner/outer, introspection/extrospection ...), dualism is built into our language. And inasmuch as it is built into our language, a dualistic necessitarianism is generated. 'This is how it has to be,' we say (Philosophical Investigations, §112). Thus an early-modern picture holds us captive. And we can't get outside it, for it lies in our language, and language seems only to repeat it to us (ibid., §115). It's an unwholesome picture, too, which has much discomposed Lady Philosophy. She can hardly sleep nowadays, worrying that the men she sees are really 'hats and coats concealing automatons' (Descartes, Meditation II), that bodily visibilia do ever but 'suggest and infer' mental invisibilia (Berkeley, Alciphron), that even if she believes elsewise, she cannot know elsewise. And, alackaday, this is how it has to be—yes? No! If we are already given over to that disputable seventeenth-century dualism, or to a metaphysic logically-isomorphic therewith (e.g. today's scientistic you're-your-brainism), then the foregoing is foregone. But we needn't give ourselves over. (And don't let's.)

' "I can't know what is going on in him" is, above all, a picture. It is the convincing expression of a conviction. It does not give the reasons for the conviction. They are not obvious.' (Philosophy of Psychology, XI, §326.)

Wittgenstein

Outer goings-on are not, as it were, 'symptoms' of inner goings-on, correlated therewith in a contingently 'loose and separate' way (to borrow from Hume); rather they're logico-grammatical criteria thereof and therefor (Zettel, §466, PI, §580, Blue and Brown Books, p.25, et alibi). 'The fluctuation in grammar between criteria and symptoms makes it look as if there were nothing at all but symptoms' (PI, §354; cf. BB, p. 113, Z, §438)—i.e., nothing at all but empirico-inductive correlations, e.g. that between barometric pressure and rainfall (PI, §354)—yet this isn't so. Bodily perceptibilia aren't empirico-inductive ('symptomatic') evidences for mental imperceptibilia; they're conceptuo-criterial evidences. We do not reason by induction from the bodily to the mental, from the outer to the inner, travelling through a mind-body interspace. 'We do not see facial contortions and make inferences from them (like a doctor framing a diagnosis) to joy, grief, boredom' (Z, §225). If a man thrown into a blazing fire screams in agony, onlookers don't say, 'Probably he's in pain—very probably even. But, then, maybe he's pretending.' (No, 'one can't pretend like that' [ibid., §570].) And 'if I see someone writhing in pain with evident cause, I do not think, "All the same, his feelings are hidden from me" ' (PP, XI, §324), or, 'His mind is in pain, but what has that to do with his body?' (PI, §391).

Consider infancy. An infant is an altogether self-revealing creature. His inner's his outer, his outer's his inner. If he is mewling and squirming fretfully then we know that something's wrong, although we mayn't know just what. (We know that something's wrong; we believe that it's colic.) And if he is smiling peacefully, or giggling gleefully, we do not say, 'Notwithstanding, maybe he's severely perturbed within.' For we learn self-control and self-restraint, concealment and suppression, dissimulation and dissemblance. (And these actions are now virtuous, now vicious.) It doesn't make sense to say that we can do them ab initio. 'A child has much to learn before it can pretend' (PP, XI, §363). And 'lying is a language-game that needs to be learned like any other one' (PI, §249).

'A child has hurt himself and he cries; then adults talk to him and teach him exclamations and, later, sentences. They teach the child new pain-behaviour.
   ' "So you are saying that the word '"pain"' really means crying?" On the contrary: the verbal expression of pain replaces crying, it doesn't describe it.' (ibid., §224.)

The mind-body connection is not contingent, but that's not to say that it's necessary. It's neither/nor. For there's no literal connecting to speak of. We're not talking about an efficient causality; we're talking about a formal be-causality. Criterial evidences are not inductive—so '[we] do not say that the evidence makes what goes on within us only probable' (MS., 169)—but that's not to say that they're deductive. It's neither/nor again. And although third-person experiential propositions ('so-and-so believes X, feels Y, desires Z ...') are defeasible, certainty is nevertheless possible. Defeasibility is not defeat. (Omnia praesumuntur legitime facta donec probetur in contrarium.)

'I can be as certain of someone else's feelings as of any fact. But this does not make the sentences "He is very depressed," "25+25=50," and "I am 60 years old" into similar instruments. A natural explanation is that the certainty is of a different kind. This seems to point to a psychological difference. But the difference is a logical one.' (PP, XI, §330.)

Moreover, defeasibility is circumstantial. It has context, setting, background and foreground. Suppose a man is bleary-eyed and sniffling. He may be grief-stricken, he may be dicing onions, etc. To answer 'Why is he thus and so?' we need to look at the 'outer' and not the 'inner'; we need to look around him.

'Pain-behaviour and the behaviour of sorrow—these can only be described along with their external occasions. (If a child's mother leaves it alone it may cry because it is sad; if it falls down, from pain.) Behaviour and kind of occasion belong together. (Z, §492.)

'An "inner process" stands in need of outward criteria.' (PI, §580.)