Friday 4 June 2021

Norman Malcolm’s ‘Wittgenstein: The relation of language to instinctive behaviour’

As well as advising a UCLan philosophy student on his dissertation, my reason for returning to read the literature around On Certainty (OC) is one of those periodic realisations that a change of view in one area forces a change elsewhere. I realised that my quick summary, the sort of thing I used to say, at least on my previous interpretation of it, won’t do as an endorsing account of On Certainty.

The sort of thing I tend to say is that OC is radical because LW separates the game of giving and asking for reasons, of knowledge and doubt, from certainty. A framework of both animal and inherited certainty is the condition of possibility of knowledge claims and also of claims to doubt since doubts, too, take work and a context to make them reasonable. (On this, I like to remind myself that claiming that one didn’t know one’s suspiciously suddenly dead partner’s bank balance will not get waved through as it might in an epistemology class. One can be assumed to have picked up knowledge - rubbing off on one like an infectious disease - unless there is good reason why not.)

But implicit in my own past understanding of the sort of thing I tend to say was the idea that not only was knowledge distinct from the sort of certainties that plays a role in OC – let’s follow the literature and call them ‘hinge propositions’ – but that knowledge itself wasn’t certain. I thought this, if I am honest now, because I wasn’t much interested in epistemology and so had never really thought about any necessary conditions one could offer for knowledge (even if one eschews illuminating sufficient conditions). I suspect I subscribed implicitly to some sort of JTB analysis of knowledge and simply accepted that the justification condition didn’t need to be very strong. Reliabilism was sadly rife in the department in which I did my PhD (Cambridge HPS) and so perhaps this persuaded me in the same direction.

Since then, however, I’ve become much more attached to the McDowellian and Travisian view that attempts to combine fallibilism with the idea that if one knows something, one cannot be wrong. Not just because of the facticity of knowledge but because of the luck-eliminating exclusion of counter-factual possibilities. Given that their view is an action at a distance of Oxford Realism, my sick-leave project is to rethink On Certainty through the conceptual spectacles of Oxford Realism.

I am, however, still at the personal archaeological stage and hence looking at the sources I recall liking. At the top of the list is Norman Malcolm’s ‘Wittgenstein: The relation of language to instinctive behaviour’ (Malcolm 1982).

Malcolm’s main theme is that language is grafted onto primitive, animal behaviour. Malcom takes this to contrast a cognitivist or representationalist view, citing Chomsky and Fodor. This takes up the first 5 sections of the 12 section paper. In §6 he describes Moore’s comments about being certain that he has clothes on and then that he knows this. He quotes Moore thus:

How absurd it would be to suggest that I did not know it, but only believed it, and that perhaps it was not the case! You might as well suggest that I do not know that I am now standing up and talking - that perhaps after all I’m not, and that it’s not quite certain that I am! (Moore 1959: 146-147). (Malcolm 1982: 12)

In §7, Malcolm sets out his account of Wittgenstein’s criticism of Moore. It includes:

If Moore had declared to his audience that he did not have clothes on, was not standing up, did not have two hands, etc., the people there would have regarded him not as mistaken, but as mentally disturbed (OC 155). Why is this? Apparently because we expect an adult speaker of the language, who is in possession of his faculties, to be able in normal circumstances, to say straight off, without looking for evidence, whether he has clothes on, or whether he is sitting or standing.
This helps to bring out the fact that Moore is wrong in saying that he knows these things. For when a person contends that he is not merely convinced of something but knows it to be so, we expect him to be able to produce evidence in support of this distinction. But Moore was in no position to do this. As Wittgenstein says, if what a person believes ‘is of such a kind that the grounds which he can give are no surer than his assertion, then he cannot say that he knows what he believes’ (OC243). (ibid: 13)

The first part of this quotation reiterates Moore’s unexceptional certainty about himself and his state of dress. But the second part moves to suggest that Moore does not know this, opening up a gap between knowledge and certainty which forms a key part of what I tended to say about OC. That’s how I took this, 30 years ago. But now it seems to me that something else has also been introduced: that Moore is wrong to say he knows this. Malcolm says: When a person contends that he knows something to be so, we expect him to be able to produce evidence in support of this. But Moore was in no position to do so.

A little later Malcolm says:

The case is exactly the same here as with Moore’s assertion, ‘I know that that’s a tree’, which he reiterated in philosophical discussion. Wittgenstein said of this: ‘I know that that’s a tree’. Why does it strike me as if I did not understand the sentence?, though it is after all an extremely simple sentence of the most ordinary kind?... As soon as I think of an everyday use of the sentence instead of a philosophical one, its meaning becomes clear and ordinary (OC 347). Outside of ordinary contexts neither ‘I am certain’ nor ‘It is certain’ (that I have two hands) has a clear meaning. (ibid: 14)

Again the focus is on saying ‘I know’. Note also the immediate connection to certainty. ‘I am certain’ is treated here as akin to ‘I know’. But in this case Malcom continues with:

Nor can a person who is going about his affairs in ordinary life be said to assume, or take for granted, or presuppose, that he has two hands. (ibid: 14)

But this seems to me an entirely separate matter and not obviously true. Why can we not say that such a person ‘takes for granted’ his having hands? Of course, we will need a context in which this will be a sensible thing to say. But perhaps just this piece of philosophical anthropology is that context. A person may check some thing but not others. It may not come as news to anyone on Earth that an ordinary person, albeit a Cambridge philosopher, does not generally check that they have hands. But that fact itself may be interesting in a piece of philosophical anthropology, as a reminder of something always before us. Malcolm seems – precisely – to be interested in this, in this very paper.

Later Malcolm adds:

Wittgenstein says: ‘I would like to reserve the expression “I know” for the cases in which it is used in normal linguistic exchange’ (OC 260). Of course the same should hold for ‘I believe’, ‘I am certain’, ‘I agree’, ‘I assume’, and also ‘I do not doubt’. (ibid: 18)

Note again that this concerns first person present tense avowals, not underlying states of knowledge. Note also that there is an immediate link to ‘I am certain’. There is no distinction here between ‘I know’ and ‘I am certain’ with respect to reserving their first person present tense uses to the uses in ordinary circumstances (a critical remark against Moore but not yet telling). 

The underlying theme of Malcom’s paper is the absence of doubt that underlies the use of words and which Malcolm suggests characterises hinge propositions. He argues that it is difficult to articulate because, in some sense, it cannot be put into words.

Wittgenstein conceives of the absence of doubt that exists at so many points in the daily course of our lives, as not something ‘hasty or superficial’ but as ‘something that lies beyond being justified or unjustified; as it were, as something animal’ (OC 358,359). It is too fundamental to be either ‘unjustified’ or ‘justified’. It underlies any mastery of words in which a procedure of justification could be framed. This fundamental thing is so fundamental that it is difficult, or perhaps impossible, to describe it in words. One would like to characterize it in mental terms - to call it knowledge, or belief, or conviction, or certainty, or acceptance, or confidence, or assumption. But none of these expressions fit. All of them have their appropriate application within various language-games. Whereas Wittgenstein is trying to call attention to something that underlies all language-games. But can’t one give a characterization at least in negative terms of this fundamental thing? Wittgenstein attempts this in many passages. A formulation he frequently resorts to is ‘the absence of doubt’. (ibid: 17-18)

There are indications in On Certainty that Wittgenstein is dissatisfied with every attempt to characterize this fact that is so fundamental to language, thought and action. Suppose that a customer tells a greengrocer that he wants ten apples, and the shopkeeper proceeds to count them out. Wittgenstein says:
If the shopkeeper wanted to investigate each of his apples without any reason, in order to play safe, why doesn’t he have to investigate the investigation? And can one speak here of belief (I mean belief as in religious belief, not conjecture)? Here all psychological terms merely lead us away from the main thing (OC 459)
In order to have ‘absolute certainty’ must not the shopkeeper try to determine not only that these things are apples, but also that what he is doing is trying to find out whether they are apples, and in addition that he is really counting them? And if the shopkeeper doesn’t do this, is this because he ‘believes’, or ‘knows’, or is ‘certain’, or is ‘convinced’, or ‘assumes’, or ‘has no doubt’, that these are apples and that he is counting them? No. All psychological terms, says Wittgenstein, lead us away from ‘the main thing’ (die Hauptsache). (ibid: 19)

Hence a connection back to the saying vs. showing distinction that seems to be found in the Tractatus (‘seems’ because of the ongoing debate as to whether it is voiced in propria persona).

Am I not getting closer and closer to saying that in the end logic cannot be described? You must look at the practice of language, then you will see it (OC501).
Logic cannot be described! I take this to mean that it is not appropriate for Wittgenstein to say either that he ‘knows’, or ‘believes’, or is ‘certain’, or is ‘convinced’, or ‘assumes’, or ‘does not doubt’, that his name is L. W., or that this is called a ‘hand’, or that the law of induction is true. None of these terms are correct. What does it mean to say: ‘You must look at the practice of language, then you will see it’? What do you see? Well, you see the unhesitating behaviour with which a person signs his name at the end of a letter or gives his name to a bank clerk; or uses the word ‘hand’ in statements; or makes inductive inferences; or does calculations; and so on.
What you see is this unhesitating way of acting. This is the ‘logic’ of language that cannot be described with psychological words. It is too ‘primitive’, too ‘instinctive’, for that. It is behaviour that is like the squirrel’s gathering nuts or the cat’s watching a mouse hole. This is why Wittgenstein says it is something animal (OC359). (ibid: 19-20)

Again, however, Malcolm thinks there is a difficulty in putting something into words. In this case, his own attempt to describe the absence of doubt that, he thinks, underpins Wittgenstein’s account.

Just now I have spoken of this human behaviour as ‘unhesitating’, and previously as ‘confident’. But isn’t this a strained use of these words? When I sign my name to a cheque in the normal way, is it correct to say that I do it ‘confidently’ or ‘without hesitation’? It seems not. If, with advancing senility, I sometimes forgot my name, or sometimes write it unconfidently, then an observer could report of me, ‘This time he wrote his name confidently, without any hesitation’. Or if I had been given a drug that is supposed to produce amnesia, an observer might report, with surprise, ‘Why, he stated his name quite confidently!’ But the way in which a normal adult normally comes out with his own name and with the names of familiar objects, cannot be called either ‘confident’ or ‘unconfident’, ‘hesitating’ or ‘unhesitating’. (ibid: 20)

Malcolm is surely right to say that in these latter specific contexts concerning senility or being drugged, it would make sense to describe an action as ‘confident’ or ‘unhesitating’ and that, in other ordinary circumstances, the very same fluency of action would not call for these descriptions. The attempt to compliment one’s partner as they signed their signature in ordinary circumstances using one of these words would likely backfire.

But it is not clear that there is no context here to use these words in Malcolm’s attempt to describe normal doubt-free action.

I think that this point generalises to Malcolm’s comments about Wittgenstein’s criticism of Moore saying that he knew he was clothed. Without some appropriate context, such an avowal is otiose. It is unclear what Moore would be saying. Equally, without some context, it is unclear what is being said in the case of a philosopher saying, of a tree before them, that ‘that is a tree’ or ‘I know that that is a tree’. But it does not follow that the same philosopher does not know that it is a tree. His knowing that it is a tree might be cited in evidence against him if his defence depended on there having been no tall objects in the vicinity. The opposing barrister might say: ‘You could hardly have failed to spot the tree in front of you in good light and clear weather?!? So you could hardly have failed to know that there was indeed a tall tree available for your nefarious purposes?!? Your defence that you did not know this is a tissue of lies!’

Nor does it follow that Moore’s certainty is not the certainty of knowledge. In normal circumstances and when in full view of a partner it would be bizarre to say ‘I am wearing clothes’ or ‘I know that I am wearing clothes’. But it does not follow that one does not know that one is wearing clothes. It just does not require stating.

Malcom suggests that ‘when a person contends that he is not merely convinced of something but knows it to be so, we expect him to be able to produce evidence in support of this distinction’. Moore himself acknowledged that there was nothing he could do to establish or to prove his having two hands. He shares Wittgenstein’s view that nothing, in normal circumstances, could be used to establish this because nothing would have better credentials. He differs from Wittgenstein and Malcolm in not thinking that this disbars this as knowledge. So how plausible is Malcolm’s claim (on Wittgenstein’s behalf)? 

Note, first, that it concerns the use of the avowal ‘I know’ rather then whether knowledge is in play and so it offers no support to the point at hand (whether Moore knows, not whether his saying ‘I know’ has a clear sense). But, second, even as a comment on avowals it seems wrong. While one may be convinced of something but not know it (because of the facticity of knowledge) to say that one is convinced of something but that one does not know it sounds very strange. Eschewing knowledge implies a lack of conviction. Avowing conviction implies that one takes oneself to know (whether or not correctly).

In sum, although I didn’t think this 30 years ago, it now seems to me that Malcolm makes awfully heavy weather of all this. The difficulties that resolute readers of Wittgenstein highlight concerning whether one has actually said anything by simply asserting the word string ‘I know…’ without suitable circumstances and an intelligible purpose do not carry over third person ascriptions of knowledge or third person ascriptions of an absence of doubt. Malcolm’s very success in outlining a philosophical context enables his own third person descriptions to carry a content even as he disowns it. And nothing shows the falsity of the idea that Moore knows these things even if his own avowal fails.

(Moore 1; Wittgenstein 1; Malcolm 0)


Malcolm, N. (1982) ‘Wittgenstein: The relation of language to instinctive behaviour’ Philosophical Investigations, 5: 3-22