Wittgenstein and Strawson on Other Minds
WHEN, in the Philosophical Investigations, Wittgenstein writes of ‘criteria’ and when, in Individuals, P. F. Strawson writes of ‘logically adequate criteria’, they do not mean the same thing. But in both cases a relation is involved which is neither a matter of evidence contingently being evidence for something else nor a matter of entailment holding between propositions. It is important to the understanding of Wittgenstein’s work that the issue is not obviously one of a relation between propositions. I shall later argue that certain points follow from Wittgenstein’s views which are about logical relations between propositions, but that the core of his view cannot be put this way. Given this caution, there is a measure of agreement between the two philosophers which may be put as follows: that some human being is behaving in a certain way or has certain things happening to him entitles us to assert that he is in some mental state. And, to repeat, the entitlement does not arise out of empirically ascertained regularities nor is it a matter of entailment. (I am here using ‘mental state’ loosely to include pains, intentions, beliefs, and even dispositions; the looseness should not matter for present purposes.)
Another area of agreement between Wittgenstein and Strawson is the thesis, in Strawson’s terms, that the concept of a person is logically primitive, that a person (or human being) is neither an animated body nor an embodied anima. Both philosophers reject the idea that it is not the very same thing to which I refer when I say ‘Jones is fat’ and ‘Jones is thinking of home’. With human beings it is not as it seems plausible to say it is with, e.g., the name ‘Iceland’ in the pair of sentences ‘Iceland is an island’ and ‘Iceland is bankrupt’. It can seem quite appropriate in this case to hold that the apparent unitary referent of ‘Iceland’ is only apparent. Bodies of land cannot be bankrupt and states cannot be bodies of land, have geographical properties. Thus, it seems plausible to say, a sentence such as ‘Iceland is a bankrupt island’ is a zeugma, as is ‘The cape is stormy and loose about her shoulders’. (I expect the right answer here is that ‘Iceland’ refers to a country, and that countries, rather like Strawsonian persons, just do take predicates from two different categories. This is why I say it ‘seems plausible’.)
Further, and to make a brief historical digression which will remind us of the origins of our problem about other minds, the duality of reference which we might find plausible with ‘Iceland’ is, Descartes would insist, also the case with human proper names. Moreover, just as the state of Iceland might continue to exist if relocated in a portion of Canada, the Cartesian real person is only contingently related to its body. (It is harder to find a parallel for existence without any body at all; perhaps a government in exile after a revolution might do.) I mention Descartes because, while we all admire him for appreciating that the relation between person and body is not like that of a captain to his vessel, we do not perhaps appreciate sufficiently just how he arrived at the view that the relation was ‘mysterious’. I shall argue shortly that representing a human body as a moving and changing thing in space-time, as Descartes would have done, is a device which focuses us sharply on the problem of other minds.
To return to Strawson. For him the fact that certain behaviour takes place is logically adequate for saying that certain thoughts, feelings and intentions are present. But, as is notorious, there is some difficulty in seeing just what this thesis comes to. In fact, it is not clear whether criteria should be called instances of behaviour. It is not clear whether the criteria for a P-predicate being true consist in a number of M-predicates being true or whether the criteria are something quite other than either P-properties or M-properties. Nor are we justified in equating Strawson’s symbol ‘P’ with ‘psychic’ or ‘mental’; though we are justified in equating his symbol ‘M’ with ‘physical’ or ‘bodily’. The sort of case which comes to mind most readily with Strawson’s view is one where a man is writhing and groaning and where that settles for us the question whether he is pain. But the question arises whether ‘writhing’ and ‘groaning’ are M or P predicates. Strawson would classify them as P predicates. But then they apparently need criteria; and one cannot see what these criteria are to be.
The temptation is to suppose that we must eventually get to M-predicates being true as the criteria for P-predicates being true. I wish to explore this temptation. But a word of apology to Strawson is in order first. The temptation is, I believe, itself a symptom of the reluctance to accept P-predicates as just as well-grounded and basic as M-predicates. Being part of a certain tradition in philosophy, we are subject to this temptation. It is very likely a deep-rooted prejudice connected with idolizing the physical sciences. Since I believe that Strawson has contributed greatly to withering this prejudice, it would be unjust of me to claim that the picture I am about to present is an accurate representation of his views. But Strawson does not do enough to show us what is wrong with the picture; and Wittgenstein’s strength, relative to this problem, lies in the fact that he can free us from the grip of the picture. I believe also that it is a widespread interpretation of Strawson to read him as saying that the truth of M-predicates is the ground for the truth of P-predicates.
Suppose then that we take seriously the suggestion implicit in the temptation, the temptation to hold that M-properties constitute the criteria for P-properties. To do this, let us imagine a mode of description for a phase in the life of a human being which consists solely of chemical, spatio-temporal, topological and physiological predicates. We need topology to deal with the alterations in configuration of what we ordinarily call a human face, alterations we ordinarily call grimaces, frowns, smiles, sneers, etc. For convenience, let us call this mode of description the Cartesian mode, the allusion being to Cartesian coordinate geometry. Immediately we can see that we are precluded from using any terms such as ‘writhing’, ‘groaning’ or any of the terms mentioned just above. Moreover, since our predicates in the Cartesian mode are all Strawsonian M-predicates, we cannot specify actions, but only bodily movements. Hence, we cannot justify a claim that the object being described is ‘raising an arm’; we can only say that a certain appendage is changing its spatial position relative to the rest of a certain body.
The Cartesian mode is available to us if, for certain purposes, we choose to adopt it. As I have already suggested, it was something along these lines that Descartes had in mind when he said the relation between body and mind was ‘mysterious’. And we can see why he thought so. For once we have done things this way, it will be impossible to build a bridge to our ordinary ways of talking about human beings. There are several things which it might be thought I am suggesting here which I am not suggesting, or which, at least, do not follow from what I am saying; and they are worth a brief mention. First of all, I am not claiming, a priori, that it would be impossible to establish correlations between actions, writhings, groans, smiles, etc., and the elements described in the Cartesian mode. I think there are bound to be enormous difficulties about such a project. But, even if further conceptual inquiry shows that the project is incoherent and not merely unlikely to occur, what I have said does not establish this, nor does my point make determinism impossible. Nothing I have said rules out the possibility of finding the sorts of correlations which would enable us to exhibit the counterpart in the Cartesian mode of any act or mental state as occurring when and only when (or only when) some antecedent set of states—in the Cartesian mode—occur. So long as we accept that some bodily movement or change occurs in connexion with actions and mental states, we cannot rule out, without a lot more argument, the possibility that these movements and states are determined by antecedent conditions. If the envisaged programme is possible, it seems reasonable to hold that any determinist should be happy. Hence, the identity theory of sensations and mental states with brain or bodily states, while perhaps of some interest in its own right, is not a requirement for determinism.1
Still, we can see that nothing in the Cartesian mode can give us grounds for our ordinary ways of talking about other human beings and ourselves. At most, we might get a theory off the ground about the causes of our responsive behaviour. Such a theory would have to be tied in with a theory about our perception of those bodily movements and changes. Now, where the prejudice I mentioned earlier asserts itself most perniciously is in the temptation to believe that, strictly speaking, all we observe in reality is what can be described in the Cartesian mode. This raises problems about the concept of observation, and I do not want to go into any detail about that. But if, as I think would be agreed, we would implausibly say that we were observing a gambit in chess if we knew nothing of the rules, we would then with equal implausibility say that we observed the Cartesian counterpart of a frown if the description of that frown involves topology and we knew nothing of that science. Similarly, in observing a man lifting his arm above his head, we are not observing an increase in the angle between an upper bicep and the left or right side of an upper torso. Further, concerning what we can observe and perceive, there is no barrier to saying that we see pain in a man’s face or hear anger in his voice. We also observe the gracefulness of his walk or the firmness of his stance, attributes which are revealing of personality and character.2 If it is replied to this point that we do not literally see his pain or hear the anger, an explanation is owing of what ‘literal’ is contrasted with here. It is not figurative to talk this way as it would be to talk of floorboards groaning or of causing suffering to a flower by cutting its bloom. Such modes of speech are figurative precisely in contrast to the way we talk about human beings.
As I said above, this picture of our situation, one in which we find no way to bridge the gap between the Cartesian mode or M-predicates and the way we talk of human beings, is not a representation of Strawson’s view or anything which follows from his view. I presented it to illustrate a difficulty in that view, namely that it is not easy to stop oneself from interpreting it along these lines. The picture of our situation represents the correct line of development if it is held that M-properties are the criteria for P-properties. But if they are not, and cannot be, it would appear that whatever these logically adequate criteria for P-predicates are, they cannot be stated with either M-predicates or with P-predicates. And this is a puzzling result. There is a passage in Strawson which hints at the right direction we must take to resolve the puzzle.
If one is playing a game of cards, the distinctive markings of a certain card constitute a logically adequate criterion for calling it say, the Queen of Hearts; but, in calling it this, in the context of the game, one is ascribing to it properties over and above the possession of these markings. The predicate gets its meaning from the whole structure of the game. So with the language in which we ascribe P-predicates. To say that the criteria on the strength of which we ascribe P-predicates to others are of a logically adequate kind for this ascription, is not to say that all there is to the ascriptive meaning of these predicates is these criteria. To say this is to forget that they are P-predicates, to forget the rest of the language-structure to which they belong.3
The suggestion here is that it is the entire system of P-predicates of which we gain a mastery in understanding what persons are. And there is perhaps a further suggestion in Strawson that we just do learn to apply this system without the system having to have a foundation in something else. I should want to argue, and I think that this is Wittgenstein’s view, that we grow into a mastery of this system. It is even misleading to say we are taught it, though we may be said to learn it. The system of P-predicates contrasts sharply with the Cartesian mode in that the latter does involve teaching and learning in the most banal sense of those terms.
I suggested that we just do use the system of human predicates without having to find a foundation for it. I think this is the import of many remarks of Wittgenstein’s, of which I will cite only a few. First, consider what he writes on page 223 of the Philosophical Investigations:
If I see someone writhing in pain with evident cause I do not think: all the same his feelings are hidden from me.
And on the following page:
‘But, if you are certain, isn’t it that you are shutting your eyes in face of doubt?’—They are shut.
Now what is noteworthy here, relative to Strawson, is that, even if we do call writhing and groaning and the evident cause the criteria or the grounds for a man’s having feelings, these grounds include both Strawsonian P-properties and M-properties. While Wittgenstein certainly has something like a theory about criteria—he contrasts criteria with symptoms for example—he is not obviously using the term ‘criteria’ as a central tool in his investigations of the problem of other minds. What I mean is that we cannot say that, for Wittgenstein, criteria are always something different from necessary and sufficient conditions, always something special which is related to the issue of other minds. Criteria may very well be necessary and sufficient conditions with respect to some matters. If they are not with respect to other minds, then it does not adequately illuminate Wittgenstein’s views about other minds just to say that he resolves the problem with the idea of a criterion. The concept of a criterion is a formal concept. Knowing what it comes to in any area is knowing what counts as this or that in that area.
We should read what Wittgenstein says about criteria keeping in mind always what he says about grammar. In the Blue Book,4 he writes: ‘It is part of the grammar of the word “chair” that this [here we must, I think, imagine the speaker actually performing the action of sitting down in a chair] is what we call “to sit in a chair”.’ The comparable point with respect to an issue about other minds would involve a remark such as ‘It is part of the grammar of the word “headache” that this [and we imagine someone condoling with someone and offering him an aspirin] is what we call “to sympathize with and help someone with a headache”.’ So long as criteria are construed as something we observe which justify us in saying something, we shall miss Wittgenstein’s point. Suppose someone (in a philosophy class perhaps) argues that a table might be in pain. It will be part of the answer to such an argument to ask what could count as an expression of pain coming from a table; but it is also part of the answer to note that we have no idea what will count as sympathizing with a table or feeling relief from anxiety when the table is all right again. It would also be to the point to ask what would count as enjoying seeing the table suffer, what will count as sadism with respect to tables.
The tendency to persist in interpreting criteria as observable features which justify making statements is a persistence of the Humean tendency to see ourselves in the world as onlookers. As much as anything, Wittgenstein is trying to break us of this philosophical habit.
The passages I have already quoted indicate that Wittgenstein is not concerned to justify the way we do carry on. It is more a matter of reminding us of pervasive features of our lives. What he writes at paragraph 445 applies throughout the Investigations: ‘What we are supplying are really remarks on the natural history of human beings; we are not contributing curiosities however, but observations which no one has doubted, but which have escaped remark only because they are always before our eyes.’5 We are told some things about how it is not. It is not a matter of having opinions or beliefs that the bodies around us are human. Wittgenstein puts this point dramatically: ‘My attitude towards him is as towards a soul; I am not of the opinion that he has a soul.’ We do not have a situation where we can allay the doubts of the sceptic by producing further evidence or grounds. Recall the passage quoted earlier about the man writhing with the cause of his pain evident. Wittgenstein says our eyes are shut to doubt. To have opinions is, by definition we might say, to be open to doubt.
Wittgenstein’s remarks about criteria have been interpreted in the following way: When a man is writhing with cause of pain evident, we are entitled to say that he is in pain unless specific reasons for doubting are offered. If all the putative doubter points out is that the proposition that a man is writhing and groaning does not entail that he is in pain, this point carries no weight and there is nothing odd about the fact that it does not. To think the lack of entailment does carry weight is to assume that the grounds in this region of our lives must conform to models taken from elsewhere. I do not think that Wittgenstein would disagree with what is said here. Consider page 224 of the PI: ‘Am I less certain that this man is in pain than that twice two is four? Does this shew the former to be mathematical certainty? Mathematical certainty is not a psychological concept. The kind of certainty is the kind of language-game.’ But this, as I said earlier, seems not so much to be the heart of Wittgenstein’s view as a kind of consequence of it. The trouble is that the interpretation in terms of entitlement to assert unless specific reasons are offered against the assertion, does not get at the force of the sentence, ‘They are shut’, cited earlier. This is clearly not a resolution to be decent about the feelings of others. To be utterly beastly about others’ feelings is consistent with our eyes being closed to doubt. Indeed, how could the sadist take his pleasure if he doubted?
Let us consider a case where a story about entitlement to assert unless specific reasons for doubting are offered is just the right story. The contrast between such a case and the case of the suffering of another human being should then be evident. Consider the legimate inference from ‘He promised to F’ to ‘He will F’. When you ask me if Jones will be at the party and I say confidently that he will, I can support my claim by replying to your query ‘How do you know?’ just by saying ‘Jones promised to come’. And here, in the absence of a specific reason for doubting, my claim is justified. It is no rebuttal to point out the obvious truth that promising does not entail performance. But suppose we try to fit into this story the idea that our eyes are closed to doubt. If I knew Jones very well and had acquired great respect for him as a man of his word, it might very well be that my eyes would be closed to doubt. I might even say ‘My eyes are closed to doubt here; nothing you say will make me change my conviction that he will be here’. But this avowal of mine is now an avowal of faith in Jones. And this is surely not so in the kind of case Wittgenstein asks us to imagine. It is not due to some general faith in human nature that, in such cases, doubt simply does not get any purchase. If human nature enters the picture anywhere, it is that it shows itself in the fact that we do not doubt in situations like this.
If criteria have to be interpreted as a matter of logical relations between propositions, the interpretation suggested does seem to be the only one available. Sometimes this interpretation is expressed by talking of what is normally so, and this is contrasted with what is usually so. This has the merit of indicating that the relation is not a matter of empirically ascertained regularities. This is why I think that the thesis can be said to follow from what Wittgenstein says. But he is going further than to tell us what justifies us in saying something in the absence of specific reasons for doubting. It is more as though he is telling us of points in our lives with each other where we are helpless and cannot even think meaningfully about justification and grounds. It is tempting to say that Wittgenstein is indicating places where, so to speak, the gap between reasons and causes virtually closes up. Yet I am not sure this is the right general characterization of his view. Let me develop my reasons for saying this, however. I think the reasons are correct even if it is not a matter of reasons and causes after all.
What I have in mind is that there is a kind of scale running from, at one end, ‘take to be’ and ‘believe to be’ through ‘respond to as’, ‘regard as’, ‘treat as’, to, at the other end, ‘describe as’.
One end of the scale is ‘describe as’. I have already mentioned describing a stretch in the life of a man in purely mechanical, physiological and topological terms. This is a possible activity just as it would be possible to describe a chess game purely in terms of the physical properties and spatial relations of the pieces. The range of choice open to us for ‘describing as’ is enormous, as many choices are available as there are modes of description in our language; and the stuff of new ones is there too. I have suggested already that one reason the problem of other minds disturbs us is that we come to think that we have to get from one available mode of description to another. And we think that we have to find in the one mode the stuff out of which to build the bridge to the other. But if we consider when we might actually have to justify or give reasons for talking of someone in a certain way, the whole issue is surely the other way about. If we were engaging in describing a human being strictly in the Cartesian mode, the question ‘Why are you talking that way?’ would get an easy hold. Of course, the question might also have a ready answer. I may just have found to my surprise and delight that his movements can be described in the Cartesian mode, that a certain vocabulary works, that I have mastered it. But the same question directed at someone talking as we ordinarily do about a human being could get a grip only if the person questioned were supposed to be doing something else. Imagine a man trying to use rigorously the Cartesian mode. He slips up and begins to talk of what the human being there is doing and feeling. So he is asked ‘Why are you talking that way?’ The answer might be ‘Oh, I slipped up; got interested in him’. We should note with respect to the problem (so-called) of bridge building, that we can no more get from one way of talking to the other than the other way around, if getting from consists in finding grounds in one mode of description for the other.
So we have available to us, on the level of ‘describing as’, a good deal of choice and we can imagine different interests and purposes to which these choices are appropriate. But, that we can describe in different ways, adopt different modes of description, does not show that we are regarding what we describe in a certain way or treating it in a certain way or seeing it in a certain way or responding to it in a certain way. Least of all, does it mean that we are taking it to be what we describe it as. We do not cease to believe a man is before us when we adopt the Cartesian mode; or, at least, that we are using the Cartesian mode does not itself show that we cease to treat him as a man or respond to him as one. This might happen to someone though.
Any mode of description is, at least, a collection of predicates and of any such collection the general point Strawson made about P-predicates will apply. That is, that all there is to the meaning will not be the criteria for the application of these predicates. Their role in the system, in the wider setting will be relevant as well. A similar point, I believe, is made by Wittgenstein when he discusses children in our society playing with toy trains; they will be children who also know about real trains. Some children in a primitive tribe might be taught merely how to run the toy trains. Wittgenstein says of such a possibility that the primitive children’s activity will have a different sense. If we were, for example, to give definitions of words like ‘pain’ and ‘groan’ which would enable us to use them in the Cartesian mode, thus relocating these words in a strange setting, a similar change in sense would take place.
To each mode of description, there is a range of attitudes or responses to which the language, the vocabulary of the mode of description is internally related. We might say the vocabulary expresses the interrelationship between us and the things we are talking about, though it does not describe that relationship (describing it is what I am trying to do here). A mode of description, a vocabulary, presents a creative mind with an opportunity, and an attitude of, say, wonder, might be evoked by any vocabulary if it were deployed by the right man. Some attitude terms seem general enough that we may be loath to rule out their applicability relative to just about any mode of description. But this does not seem to be so with every case. An attitude of reverence or religious awe, for example, could hardly be expressed with any mode of description.
While we may adopt and follow the rules of various modes of description, it does not follow that the understanding I have of these rules enables me to feel just any way about what I am describing or to respond in just any way. When I attribute understanding of a vocabulary to a man, there is an implication that he is susceptible to certain kinds of responses. And while there may be choices as to how to go about describing a situation, there is not the same kind of choice as to how I shall feel or as to what feelings I can make intelligible to others. (We might say that a way of talking can grab you by the heart, virtually pull out of you certain responses.) In many cases, if a way of talking does not affect a man in certain ways, we shall be entitled to doubt whether he understands the words he is using. Hamlet cannot say ‘How weary, stale, flat and unprofitable seem to me all the uses of this world’ and then inform us that he is glad to be alive.
At one end of the scale are ‘take to be’ and ‘believe to be’. Now we do not have a choice about that. And this is because if what is there is a human being, then it is one and you cannot, on pain of madness, mistake it for something else. Or if we do, as in bad light in a forest where we think it is a tree, we have ordinary error. But if I cannot grasp how to describe something as this or that, it does not follow that I have made an error about what it is. What I meant by ‘on pain of madness’ comes out if we consider briefly ‘treating as’ and ‘responding to as’. A man might very well treat another man as merely a piece of stuff (say a slave trader loading a ship efficiently). But he does not necessarily take the slave to be a piece of stuff; or, if he does, this is not shown merely by the fact that he treats him that way. We should have to know more about our slave-trader first. We might want to know, for example, whether he is racked with unconscious guilt. To say the trader takes the man in the hold to be just a piece of stuff is actually exculpatory, allowing a clear plea of insanity. That the trader treats the slave in a certain way, an inhuman way even, does not show that he responds to him in a way got at by ‘as a piece of stuff’. He may, sometimes, slip into pangs of compassion as he distributes water to his captives. We may not be able to say, in any general way about the trader—and especially not about ordinary deckhands—just how it is with him or them towards the cargo in the hold. It is vastly too easy to say they do not believe them to be human beings.
If a man takes an x to be an F in one situation, but to be a G in another (where F and G are incompatible and no change has occurred) he is mistaken in a straightforward way in one case or in both. If a man treats another as a hunk of stuff in one situation but as a human being in another, this is not contradiction or contrariety, but a different sort of human muddle. Even that is too strong; for often there will be nothing sinister in a situation where it is plausible to say one man is concerned about others only as matter (consider a pilot making calculations to assure a safe takeoff; still, this is motivated by concerns which are human).
It might be thought that when I am ‘describing as’, I must also be right in my description and thus be taking the things to be what I describe them as. But this is a mistake. When I ‘describe as’, all that needs to be so is that what I describe is as I describe it. Suppose we adopt the Cartesian mode in connexion with a cricket player going for a six. We shall leave out all reference to intention and action and make no mention of aims internal to playing cricket. If I describe his swing as a motion of a certain kind including flesh, bone and wood, the Cartesian mode predicates must be true of what is happening for me to get it right. But what I am describing is still his swing, an action of his. I do not forget this in adopting the Cartesian mode. With human beings, there is a primacy to our ordinary ways of identifying and describing them. We take what we see or believe what we see to be actions even if we can describe them as mere bodily movements. We do not slip into the Cartesian mode from our ordinary responses as we can easily slip in the other direction.
It is worth noting here that one factor contributing to the prejudice I mentioned earlier in this essay is that human beings can be much more easily brought under the Cartesian mode than inanimate objects in rerum natura can be brought under our human ways of talking. In these cases, we shall readily concede that we speak figuratively (as with groaning floorboards). This universality of applicability is a powerful inducement to believing that the Cartesian mode alone gets at how things ultimately are. It is easy to see how this universality leads to this belief; but it is a different matter to establish that it justifies it, and I cannot see why it should be thought to.
What I have said about the primacy of our human vocabulary of action and feeling suggests that this vocabulary is not one we can choose to use or not to use, but that it is constitutive of our human nature. Of course we can, in many circumstances, choose not actually to say anything in this language. But Wittgenstein’s view, as I understand it, is that we have no choice in taking things this way, in the way got at by the vocabulary. The responses we have and the related attitudes can only be understood by considering the language in which they express themselves, along with the ways of acting which, with the vocabulary, constitute a form of life.
1 I must confess that I write here with my tongue veering towards, if not completely in, my cheek. I do think the envisioned project is incoherent. Suppose a determinist said that, on the basis of the bodily states of a group of people at a given time, he could predict their bodily movements in a year’s time. He does this without any reference to psychological or mental concepts, leaving it to us to supply the action descriptions. Apart from the stock objections about informing the subjects about the issue, thus allowing them to change their minds, it should be noted how hard it is to stop from having to take into account a huge portion of the universe. For example, a falling meteorite in Australia might cause the death of a relative of one of the subjects, and lead to him or her having to be elsewhere than predicted. The same point obviously applies to predicting the movements of inanimate objects on earth. We just get the old Laplacean picture, requiring a God’s-eye view, a mere assertion that the thing could be done. Still, the weaker point that, after the fact, what happened can be exhibited deterministically, seems to me all a determinist really has to insist on.
2 Miss Cora Diamond has pointed out to me that while nothing does seem to be said by saying that we do not literally see pain in a person’s face, it is plausible to talk here of a different sense of ‘seeing’. And this does distinguish such cases from the case of seeing pieces on a chessboard and seeing a threatened gambit. In the latter case, there is no inclination to talk of different senses of ‘seeing’. Similarly, if a man raises his hand at a meeting to attract the speaker’s attention, we may say either that we see him raise his hand or that we see him trying to attract the speaker’s attention. Again, it is not a different sense of ‘see’ here. J. L. Austin cautioned philosophers against multiplying senses beyond necessity, but, in the case of seeing pain, it certainly does seem a natural thing to say that the sense is different. Of course, a different sense is not thereby a non-literal use.
3 Individuals, P. F. Strawson; London, 1959; page 110.
4 The Blue and Brown Books, L. Wittgenstein; Blackwell’s; Oxford, 1959; page 24.
5 Cf. the remarks of C. S. Peirce about metaphysics scattered thoughout his writings, that it is an empirical study but about such pervasive facts that they are hard to notice. For example: ‘It is, on the contrary, extremely difficult to bring our attention to elements of experience which are continually present’. Collected Papers, Vol. I, page 55.