Inconsistency and Approval-Derivation
Something more must now be said about the important logical property of inconsistency and its relevance to the quasi-deductive derivation procedure which we have been discussing. (It has been suggested already that the judgement ‘there is nothing wrong in breaking a promise’ is something like a self-contradiction, even though we take it to be the expression of an attitude, and therefore neither true nor false.1)
In a piece of deductive reasoning it is inconsistent to assert the premisses and deny the conclusion. If you do deny the conclusion, you have in effect denied one or other of the premisses which you yourself asserted. Does anything like this apply to the approval-derivation procedure which we have been considering?
Suppose that I have expressed approval of anyone who does a service to his neighbour; and I fully admit that William did do a service to a blind man by helping him to cross the street, but I do not approve of what William did. Of course, I am not thereby denying the approval I expressed of anyone who does a service to his neighbour. An attitude of approval, or the utterance expressing it, cannot be denied, as a statement can. But there is one thing we can do with both. Both a statement, which is true or false, and an expression of approval, which is neither, can be withdrawn. To put it vulgarly, we can ‘back out of’ either of them. (We can withdraw a command too, or an invitation, though these again are neither true nor false.) If I do not approve of William's action, I am withdrawing or ‘backing out of’ the general approval of all actions of that sort which I previously expressed. To put it in another way, I am unsaying what I previously said. And surely this is just what we mean by ‘inconsistency’.
It must not, however, be supposed that it is always reprehensible (unreasonable) to unsay what one has just said. If it were, there could be no such thing as the empirical falsification of a universal assertion. ‘All crows are black, and this is a crow. But good heavens! it is a grey one.’ Here I am unsaying or withdrawing something I have just said, namely ‘all crows are black’. But I do well to unsay it, because I have now discovered that it is false.
Can anything like this happen in the sphere of approvals and disapproval? I shall now suggest that it can. If I am right, here is another important way in which the ‘logic’ of approvals and disapprovals resembles the ordinary logic of assertions and denials.
An Analogue of Empirical Falsification
‘All crows are black’ is inconsistent with ‘this crow is grey’. A reasonable person cannot assert both. Which of them is he going to reject? If he is a reasonable person (not just an incorrigible theorizer) he is in no doubt about the answer. He rejects the universal proposition, however firmly he had believed it before. The experience of a particular matter of fact has the primacy over a generalization. If there is a clash between them, it is the generalization which must go.
A general approval or disapproval obviously cannot be falsified. If the attitudinarian analysis of moral judgements is correct, ‘all actions of sort A are right’ is not an assertion, and cannot be either true or false. If so, it cannot be falsified (i.e. shown to be false) either empirically or in any other way. Nevertheless, something analogous to empirical falsification can happen to it, as I shall now try to show.
Imagine you are an Englishman living at the beginning of the sixteenth century. From your earliest years, you have been brought up to approve of the burning of heretics. An ancestor of yours, greatly respected in the family, was a member of Parliament and voted for the statute De haeretico comburendo in the reign of King Henry IV, as doubtless many other good and pious men did. You share his attitude, without any reservations. But it so happens that you have lived all your life hitherto in a remote valley among the mountains of Cumberland, where no one is sophisticated enough to hold heretical opinions. So you have never yet had occasion to apply this general approval of yours to a particular instance, though you are perfectly sincere when you express your attitude by saying ‘it is right to burn anyone who is a heretic’.
But then one day you leave your remote valley and go to London. And while you are there you actually witness the burning of a heretic. You are quite satisfied that he has had a fair trial; there is no doubt at all that he does hold highly heretical opinions. It might be expected that you would approve of this particular action which you see being done before your eyes. But not only do you fail to approve of it. You go farther. What you feel is the strongest disapproval, which could be expressed by saying ‘what is being done here is utterly abominable’.
That settles the question, at least for you. In this clash between a general moral principle and a particular moral experience here and now, it is the principle that must go. It has not been falsified (as ‘all crows are black’ is by seeing a grey one) because, on the view we are considering, it cannot be either false or true. But it has been nullified or put out of court by the particular moral experience you have had. Something rather like an empirical test has been applied to this general approval of yours, and it has not stood up to the test. It has not been refuted, since it was not an assertion. But we could quite properly say that it has been rendered unacceptable. And you will admit that in some sense you were making a mistake when you accepted it.
What sort of mistake could it be? It was this. You did not fully ‘realize’ what it was that you yourself were approving of when you said ‘it is right to burn anyone who is a heretic’. You did not know what such an action would be like if it were actually done. Something corresponding to Newman's distinction between notional and real assent2 applies here. Your approval of this principle of action was notional rather than real, although it was perfectly sincere. Nor is this at all surprising. Much of our thinking, whether about conduct or other matters, is carried on by means of words; and very frequently we use words in an uncashed manner, without realizing at all fully what an actual situation would have to be like, if these words were to be a correct description of it.3 To put it otherwise, we use a verbalized concept without realizing fully what it would be like for this concept to have an actual instance.
Where approvals and disapprovals are concerned, the tendency to think of a type of action (or of person) in this ‘uncashed’ way is sometimes called ‘lack of imagination’. And if the type of action is one in which the agent does something to another person, as it frequently is, we call this defect ‘lack of sympathetic imagination’. This is a failure to realize what it would be like to be a person to whom an action of this type was done. The capacity for sympathetic imagination varies greatly from one man to another; and no doubt this is the explanation of much of the moral disagreement which exists in the world.
This ‘notional’ approving or disapproving of a type of action, or class of actions, is something like what happens when our approval (or disapproval) of a particular action has a false presupposition; that is, when we approve or disapprove of a particular action ‘as having’ such and such a characteristic, and it does not have that characteristic, though we believe that it has.4 For example, we believe it was an example of truth-telling, because what the man said did in fact happen to be true; but he himself at the time thought it was false and intended to deceive his hearers. If we had known what he really was doing when he uttered these words, we should not have approved of what he did. We assumed that he told the truth intentionally. In a rather similar way, if you had known what kind of action heretic-burning actually is, you would not have approved of that kind of action. But this is not quite the same as saying that your approval had a mistaken presupposition. It was rather that you had an ‘uncashed’ concept of the type of action you were approving. Your failure was a lack of clarity, rather than a mistaken belief.
But it is more interesting to notice that this same ‘notional’ or ‘uncashed’ character may be found when our attitude is a neutral one, neither approving nor disapproving (the attitude which we express by saying ‘this is not a moral question at all’). This neutral attitude too may be rendered unacceptable to us in the same way, by encountering an actual instance.
Let us consider an example in which the object of our attitude is a type of person, or a class of persons. Suppose that I have heard from time to time that there are persons who forgive their enemies, but I have never met one of them, and I am told that they are not very numerous. Naturally I do not approve of them. How could one, when they are so silly? Still, being a fair-minded man, I do not disapprove of them either, since they do not do any harm to anyone except themselves. They are just eccentrics, or perhaps harmless lunatics.
But then, one day, I actually come across someone who has been very unkindly treated by another man, and forgives him for it. I get to know this person better, and find that he does this sort of thing repeatedly. Apparently he makes a habit of it, which is most extraordinary. And now I begin to change my attitude of moral neutrality. I find myself saying ‘Well, really! he is a very nice person’ (I am not compelled to use the officially-moral vocabulary) or ‘he is really rather a splendid person, after all’. And then it occurs to me that I would not mind being that sort of person myself, if only I knew how to be; and if everyone were like him, the world would be a much better place than it is.
Here again, it is a case of ‘cashing’ our previously-uncashed symbols by experience of a particular instance, and the result of this is that our previous attitude of moral neutrality is rendered unacceptable. It is inconsistent with the attitude which we now have to this particular person. And here again it is the general attitude which must go, the attitude to a class or a type. What is decisive is ‘how we feel’ about this particular person who is a member of that class or an instance of that type. Here again we have something analogous to the empirical falsification of a general assertion. But here the parallel in the sphere of assertions would be a universal assertion which is both negative and disjunctive. Our previous morally neutral attitude could be expressed by saying ‘No forgiveness of enemies is either good or bad’. A parallel to this would be ‘no jaguar is either black or white’, an assertion which is falsified when we find a black one in the forests of South America.
So much for the logical properties of approvals and disapprovals. It would seem that assertive and non-assertive theories of ethics do not differ quite so much as we might suppose, if expressions of attitude, which are neither true nor false, can nevertheless have logical properties and these logical properties are at least analogous to those which assertions have.
Intra-Personal and Inter-Personal Inconsistency
But inconsistency is of course a relational concept. Something is inconsistent with something else. What could a person be inconsistent with? He could be inconsistent with himself. That is one way of answering the question. But it needs some elucidation; ‘himself’, after all, is not exactly ‘something else’. A clearer way of putting it is this: in such a case the two ‘somethings’, between which the relation of inconsistency holds, are within a single person; both of them are included in the same person's mental history. They are a saying of his and another saying of his which unsays the first. They need not necessarily be overt or publicly-uttered sayings. They might be wholly inward or private sayings, consisting of inner speech or verbal imagery, and perhaps they are more important when they are.
We can best describe this as intra-personal inconsistency. But not all inconsistency is of this intra-personal kind. The assertions ‘all animals are carnivorous’ and ‘some animals eat only grass’ are inconsistent with each other, no matter who makes them, whether publicly or privately, and no matter whether it is or is not the same person who makes both of them. Indeed, no one need actually make either of them. In this type of case, the relation of inconsistency holds primarily between ‘assertibles’, what we usually call propositions, and only derivatively between assertings of these assertibles.
It is the same when we speak of the inconsistency between two beliefs. They need not be beliefs held by the same person. Of course, they may be; and then the person who holds both at once is an inconsistent person. But if I believe that p, and you believe that not-p, it does not follow that either of us is an inconsistent person; and the difference which there is between us might still exist even if there were no inconsistent person in the world, although your belief is certainly inconsistent with mine. Here again, the relation of inconsistency holds primarily between the propositions believed, and only derivatively between the two believings or belief-attitudes; and this relation would still be there, even if neither of those propositions were believed by anyone.
Now if we are willing to hold that inconsistency can also exist in the sphere of approvals and disapprovals, are we to say that ‘intra-personal’ inconsistency is the only sort of inconsistency which can be found there? Or could an approval or disapproval of mine be inconsistent with an approval or disapproval of yours? And if it could, would this inconsistency-relation be a derivative or consequential one, as the inconsistency between two beliefs is a consequence of the inconsistency between the propositions believed?
A Child'S Dilemma
These questions are somewhat puzzling. Let us consider a dilemma in which a child might find himself. ‘Daddy disapproves of my staying in this afternoon, and Mummy disapproves of my going out. So what am I to do?’ Is Daddy's disapproval inconsistent with Mummy's? We are inclined to say that it is. But this might be because both have a quasi-imperative character, as is shown by the child's question ‘So what am I to do?’ Of course, both are disapproval attitudes also. It is not true that their quasi-imperative character is the only character they have. But still both do have it. It is somewhat as if Daddy had said ‘Don't stay in this afternoon’ and Mummy had said ‘Don't go out’. It is true that their quasi-imperative character is not sufficient by itself to make them inconsistent. If Daddy had disapproved of Tom's staying in and Mummy had disapproved of Matilda's going out, there would have been no inconsistency at all, however much imperative force these two disapprovals might have. But in the case we are considering both the quasi-imperatives are addressed to the same person;6 and that does suffice to make them inconsistent with each other, since it is logically impossible for him to obey them both.
But though he will, no doubt, ask himself later ‘So what am I to do next time’, his first question is likely to be a different and perhaps more important one, namely ‘So what am I to think of myself?’ Is he to approve of himself for having done this action, with Daddy, or is he to disapprove of himself, with Mummy? Or his question might be ‘So what am I to feel about myself?’ If he were to consult a philosopher about this latter formulation, he would probably be told that he cannot ask such a question, since feelings are not under one's voluntary control (he could only ask ‘What do I feel?’ and he must know the answer to that already). He might also be told that in any case approval is not at all like the warm feeling you have inside when you drink a cup of hot tea on a cold morning, nor is disapproval at all like having toothache. Nevertheless, he may quite well put his question to himself in that way, if there is unfortunately no philosopher at hand for him to consult. Whichever formulation he uses, he is asking whether he is to agreewith Daddy's approval or with Mummy's disapproval, since apparently he cannot agree with both. It will probably not occur to him that a third alternative is always theoretically possible:8 he might agree with neither, and take a neutral attitude, neither approving of himself nor disapproving. But as this is pretty obviously a moral issue (it can hardly be neither right nor wrong to hit another boy on the head) the possibility of taking a neutral attitude does not need to be considered.
This time the child's difficulty is not what it was in the previous example. The trouble is not that Daddy is in effect ordering him to do something which Mummy in effect orders him not to do, so that it is impossible to obey both. Neither of them is giving him orders at all, because the action approved of by the one and disapproved of by the other has already been done, and no question of obedience arises. The question is not which of his parents he is to obey, but which of them he is to agree with.
It seems clear enough that he cannot agree with both, in some sense of the word ‘cannot’. But is this because Daddy's attitude is inconsistent with Mummy's? If they had been belief-attitudes it might have been. If Daddy believes that it will be fine by half past two, and Mummy believes that it will rain all the afternoon, these two beliefs are inconsistent with each other. But this is because the proposition believed by the one parent is inconsistent with the proposition believed by the other. The inconsistency between the two believing attitudes is derivative. Can anything similar be said about approving and disapproving attitudes? Could there be a similar derivative inconsistency between them, derived from (or defined in terms of) the inconsistency between their objects?
But how could this be so in the example we are considering, since Daddy's approval and Mummy's disapproval have the same object, an action which the child did yesterday? Still, we may notice that a similar difficulty could be raised about the inconsistency between a belief and a disbelief. If Daddy believes that it will be fine from 2.30 to 4.30, and Mummy disbelieves this same proposition, his belief and her disbelief are surely inconsistent with one another, since it is logically impossible that both should be correct. Nevertheless, though belief and disbelief are different attitudes, disbelieving the proposition ‘it will be fine from 2.30 to 4.30 ‘is logically equivalent to believing the proposition’ it will not be fine from 2.30 to 4.30’. Whatever inferences can be drawn by one who holds the disbelief that p, they are the same as those which can be drawn by one who holds the belief that not-p. Any evidence favourable (or adverse) to the disbelief that p is favourable (or adverse) to the belief that not-p, and in the same degree. If the disbelief that p is correct, the fact which makes it correct is also the fact which makes the belief that not-p) correct; and if the disbelief that p is a mistaken one, the fact which makes it so is the fact which makes the belief that not-p a mistaken one. Let us see whether these considerations about belief and disbelief will help us. Perhaps they will, if we choose a suitable terminology.
‘For’ and ‘Against’
Approval is sometimes called a pro-attitude, and disapproval an anti-attitude. Let us say, more simply, that when one approves of something one is for it, and when one disapproves of something one is against it. Daddy is for a certain action which the boy did yesterday, and Mummy is against that action. These two attitudes are certainly opposed to each other. As we have seen, we are strongly inclined to say that they are inconsistent with one another. But how are we to show that they are? Perhaps we might be able to show it, if we said that being against A is logically equivalent to being for not-A (as disbelieving that p is logically equivalent to believing that not-p). And surely this is true? If someone claimed to be against A but denied that he was for not-A, we should think that he did not understand the meanings of the words ‘for’ and ‘against’.
For instance, if someone is ‘against’ enjoying the infliction of pain on others, this is equivalent to saying that he is ‘for’ not enjoying the infliction of pain on others, or ‘for’ the absence or nonexistence of such enjoyment. But suppose that someone is ‘against’ a particular entity. Then we must say he is ‘for’ the non-existence of that entity. It may seem strange to suggest that if I am ‘against’ a particular person, I am ‘for’ the non-existence of that person. But strange and shocking though it is, surely that is what I am ‘for’, if I really am ‘against’ him in toto; in my eyes, it is just a ‘bad thing’ that he exists.9 Fortunately we are not often ‘against’ a person (or even an animal) in this total way. Ordinarily, when we are ‘against’ a person, we are against his having some characteristic (usually a dispositional one, e.g. hard-heartedness or dishonesty) which we know or believe him to have; and this is equivalent to being ‘for’ his not having it.
In the example we are considering, what Mummy is ‘against’ is a particular action. Being ‘against’ A is logically equivalent to being ‘for’ the non-existence of A. Since being done is the way in which an action exists, being ‘against’ it is being ‘against’ its being done. And since in the case we are considering it is a past action, being ‘against’ it is being ‘against’ its having been done; and this in turn is equivalent to being ‘for’ its not having been done, though in fact it was.
This phraseology may seem a little complicated. But in fact Mummy might well express her attitude by saying ‘Oh that he had not done it!’ And Daddy might express his by saying ‘How glad I am that he did do it!’ (If he were a philosopher, he would add, with Hume, that he was glad ‘after a particular manner’.)
Now the non-existence of A isinconsistent with the existence of A. It is a straightforward logical inconsistency. It is logically impossible that the same action should both have existed and not have existed at the same place and the same time, i.e. that it should both have been done and not have been done at that place at that time. In that case the inconsistency between the two attitudes of being ‘for’ the action Aand being ‘against it’ is a derivative inconsistency, rather like the inconsistency between the belief that pand the disbelief that p. It is a consequence of the inconsistency between A's having been done and A's not having been done, since being against a past action A is logically equivalent to being for its not having been done.
But there is some difficulty in explaining what we should mean by ‘it’ or ‘the action A’. If the action A has in fact been done, what could be meant by ’ its not having been done’? What could the ‘it’ be, which (we seem to be sayimg) might possibly have possessed the property of not having been done? What a strange property for an action to possess! This is a familiar puzzle about the concept ‘existence’; for being done is the way in which an action exists. But there is an equally familiar solution of it. Let us suppose that Tom hit one of his schoolfellows on the nose. Then when Mummy says ‘Oh that Tom had not done it!’ she means something like ‘Oh that the characteristic of being a hitting of a schoolfellow on the nose had not been instantiated by what Tom was doing at that time!’
Alternatively we put it this way: ‘Oh that the description “being a hitting of a schoolfellow on the nose” had not applied to (had not been true of) what Tom was doing at that time!’ Daddy, on the other hand, is glad that this same description did apply to what Tom was doing at that time, or that this same characteristic was instantiated by what he was then doing. Daddy is ‘for’ the description's applying to the action (as in fact it did) or for the characteristic's being instantiated by it (as in fact it was); whereas Mummy is ‘for’ the description's not applying, though in fact it did, or ‘for’ the characteristic's not being instantiated, though in fact it was. Whichever way we put it, it is logically impossible that the same characteristic should both have been instantiated and not instantiated by what the boy was then doing, and it is logically impossible that the same description should both have applied and not have applied to what he was then doing.
If this is correct, it is not enough to say that the two parental attitudes are opposed to each other (still less is it enough to say that they are just different). They are inconsistent with each other, in much the same way as a belief and a disbelief can be; and their inconsistency is derived from, or definable in terms of, the inconsistency between their respective objects. What Daddy is ‘for’ is logically incompatible with what Mummy is ‘for’. And that is why Tom cannot agree with both. It is not that it is just psychologically impossible for him to do so. It is psychologically possible, though it might not be very easy. He would have to get into a pretty confused state of mind first, but then he might manage it. When we say that he ‘cannot’ agree with both, we mean that he cannot do it without being (at least for the time) an inconsistent or illogical person; and his very puzzlement shows that this is what he is trying not to be.
It is worth while to add, before we take leave of this child and his troubles, that the second one (‘What am I to feel about myself? Shall I agree with Daddy or with Mummy?’) is something of a blessing in disguise. He learns an important lesson from it. He learns, or at least begins to learn, to judge for himself, and also to judge about himself, to make moral judgements about his own actions. He is learning to do his own approving and disapproving, instead of just being elated or dismayed when others approve or disapprove of him; and so he is on the way to becoming an autonomous and responsible moral being. Moreover, if he is a discerning child, he may discover that he can in a way agree with Daddy and with Mummy too. It may occur to him that Daddy approved of his action as being an act of self-defence, while Mummy disapproved of it as being a manifestation of rage. He has thereby discovered that the same action can (correctly) be described in two or more different ways, and that one might approve of it in respect of one of the characteristics it has, while disapproving of it in respect of another; and this is quite an important lesson too.
Moral Beliefs and Feelings
We have seen that there are two versions of the doctrine that moral judgements are neither true nor false, the approval analysis or attitudinarian analysis on the one hand, and the injunctive analysis on the other. Something must now be said about the relations between the two.
It has been thought that the approval anlaysis has a serious defect, from which the injunctive analysis is free. One way of putting the point is this: it may seem that on the approval analysis a moral judgement does not tell us anything, whereas on the injunctive analysis it does. On both analyses alike, it is neither true nor false. But according to the injunctive analysis, it does tell us what to do in such and such circumstances; and is not this the most important thing we could possibly be told? ‘That action was right’ would mean, roughly, ‘Go thou and do likewise’, or more fully ‘Let an action of this sort be done by anyone who is in circumstances like these’. In that case, we are being told what to do.
Before we consider this objection to the approval analysis, it is worth while to point out that although the distinction between ‘true’ and ‘false’ does not apply to expressions of approval or disapproval, there is another distinction which does. It is the distinction between sincerity and insincerity. It is perfectly possible to express approval of something (for example, by saying ‘that action was right’) when one does not in fact approve: much as one may bow to someone without actually feeling any respect for him, or smile at him when he comes into the room without feeling at all pleased to see him.
Moreover, expressions of attitude ‘convey’ one's attitude to other people, without actually telling them that one has it; and if one is insincere they ‘misconvey’ it (I am assuming that the hearer understands the language which the speaker uses). And now the distinction between true and false does come back into the picture. This conveying, or misconveying, consists in arousing a belief in the mind of the hearer. What is thus conveyed to you is true if my expression of attitude is sincere, and false if it is insincere. This belief of yours is not itself a moral belief, but a belief about a matter of fact. It is a belief, correct or not, about another person's moral attitude towards a certain action or class of actions, or towards a certain person or class of persons. What moral attitude a man has about something is a question of empirical fact, though not always an easy question to answer.
Let us now suppose that we go to a friend and ask him for advice about some moral problem. For instance, would it be right to reveal to the police something which was said to us in confidence, or to tell the judge and jury about it when we are being questioned in a court of law? Our friend replies ‘It is always right to tell the truth’. Let us suppose also that this reply is a perfectly sincere expression of his attitude, so that we now know or have learned what his attitude is, at any rate in the everyday sense of the words ‘know’ or ‘have learned’.
But was that all we wanted to know when we asked him whether it would be right to tell the truth in this particular case? We have learned that he has a feeling of approval towards invariable truth telling. Is that what we wanted to know? When we put our question to him, did we only want to find out about the state of his soul? Oh no, it would be said. We were asking him for guidance about what to do; and if this is what we are asking him for, it does not matter in the least what his feelings are, or even whether he has any. But according to the injunctive analysis he has given us what we asked for, when he says ‘it is always right to tell the truth’. For according to that analysis, his answer amounts to saying ‘Let everyone tell the truth always, no matter what the circumstances are’. So he has given us guidance about what to do.
Discussion of this Objection
What shall we say about this criticism of the attitudinarian analysis? First, it is well to point out that one may approve of something or someone (in an occurrent and not merely dispositional sense of the word ‘approve’) without expressing ones' approval publicly to other people; and the same applies to disapproval. The moral judgements which a man makes privately ‘in his own heart’ are an important part of his moral life. They are especially important when they are moral judgements about his own actions, past or present, actual or proposed, or about his own character, that is, his own conative and emotional dispositions. A moral being is one who ‘judges himself, approving or disapproving of himself. That is what having a conscience consists in. Why should we suppose that all his goods are in the shop-window and that there is nothing inside, or nothing that matters from a moral point of view? Of course, he may also go about the world telling other people what to do. But this is a less important function of a moral being.
So I would suggest that this criticism of the attitudinarian analysis arises partly from an excessive interest in publicity, a kind of Hellenic passion for the market-place, where everyone spends his day talking and being talked to. We are not always in what is called ‘the hearer-speaker situation’; and I think we should be less than human—if we were, though we should also be less than human if we were not in it pretty frequently. But even though we do confine ourselves for the moment to that situation, it is still not true that in learning ‘how X feels’ about some action, done or proposed, we have learned something entirely irrelevant to moral questions. If I admire and respect someone, it makes a great difference to me to learn ‘how he feels’ about some action I have done.
Suppose I find that he disapproves of what I did, and points out to me the characteristic in respect of which he disapproves of it. For example he says ‘it was unkind to ask for the money back at that time, though the man had promised to repay it’. Somewhat dismayed by this, I go away and think it over. I reflect on the action again, its circumstances and its probable effects on others (not only on my debtor himself, but also perhaps on his indigent family). Then I may very well find that I disapprove of the action myself, and express this disapproval to myself, privately, by saying’ Yes, it was unkind, and I ought not to have done it’; and I resolve not to do that sort of thing again. The result may be similar if I learn how he feels about something I am proposing to do, but have not yet done. He expresses disapproval of it, and then (to my surprise, perhaps) I find myself feeling about it as he does. The result is that either I refrain from doing what I proposed, or, if I do it, I feel remorse afterwards.
Moreover, we do well to pay some attention to the question ‘how does So and So feel about it?’ even when we do not particularly admire or respect him. Indeed, we may come to admire and respect him when we do learn how he feels. What is called moral insight may be found in the most unlikely quarters, in children for example; or in persons who are ignorant or ill-informed about most other matters. It may even be found sometimes in the persons called ‘publicans and sinners’.
It comes to this, then. On an attitudinarian theory, the statement ‘there are persons who have more moral insight than I have’ is equivalent to ‘there are persons whose moral attitude to some action (or some person)11 I tend to share, when I learn what their attitude is, and in respect of what characteristic of that action (or person) they have it, though I did not have that attitude previously.’ On an objectivist moral theory, which holds that moral judgements are true or false, our statement ‘there are persons who have more moral insight than I have’ would mean something different, something like this: ‘There are persons who intuitively apprehend moral facts which I cannot apprehend intuitively for myself, unless, and until, one of these persons draws my attention to them; but when he does, I can.’ For example, now that you have pointed it out to me, I ‘see for myself’ that it was wrong to take vengeance on someone who had harmed me, though at the time I did not ‘see’ anything wrong in it at all.
But we must return to the attitudinarian analysis. To complete our reply to the criticism of it which we are discussing, there is something else we must notice. When we learn how someone feels about an action or a proposed action, we are in a way being told what to do. And when we notice how we ourselves feel about some action or proposed action, we are in a way telling ourselves what to do. It does not, of course, follow that we shall actually do it. But neither does this follow when the injunction is made perfectly explicit (‘let everyone do A in circumstances C, and I am someone who is in circumstances C’).
Approval, in other words, has in itself an action-guiding character. It is relevant to the settling of practical questions, and so is disapproval. Perhaps I may quote my namesake the eighteenth century British Moralist Richard Price. He was not himself an attitudinarian moral philosopher, but he made a striking remark which is true whatever analysis of moral judgements is correct. ‘Excitement belongs to the very ideas of moral right and wrong.’12 To put it in a more modern way, it is part of the meaning or force of the words ‘right’ and ‘wrong’ that we are moved to do the action to which the word ‘right’ is applied, and we are moved to abstain from doing the action to which the word ‘wrong’ is applied.
This practical or action-guiding character of approval and disapproval is ignored if we say that they are ‘merely’ feelings, or that expressions of approval or disapproval are ‘merely’ expressions of how the speaker feels. Approval can indeed be called a feeling; or rather an occurrent approval (as opposed to a dispositional one) can be called a feeling, and so can an occurrent disapproval. If any linguistic philosopher objects to calling it so, one may reply with the ad hominem argument that this is what we do call it; and it is very natural that we should, for approving and disapproving are experiences which we ‘live through’, as fearing and wishing are. But we have not said all there is to say about them when we call them feelings, though we have said something very important. They have an action-guiding character too.
Moreover, they still have it (we might say, they most noticeably have it) when the action approved of or disapproved of is one's own, or when the person approved of or disapproved is oneself. Each of us takes account of his own approvals and disapprovals, he admits or acknowledges their relevance, in deciding what he himself shall do. What makes someone a moral being is not the practical relevance he attributes to the approvals and disapprovals of others, but the practical relevance he attributes to his own; and especially to those of them which are approvals and disapprovals of his own actions, done or proposed, and to his own emotional and conative dispositions which manifest themselves in what he himself does. This is still true when he takes account of the approvals and disapprovals expressed by others who ‘have more moral insight than he has’. For these only become practically relevant for him when he ‘makes them his own’, by reflecting on them and accepting them for himself.
A Difficulty in the Injunctive Analysis
Now according to the Injunctive Analysis, if I interpret it rightly, a moral being does have to make moral judgements about himself, and he also has to make these judgements for himself, at first hand, in an autonomous and responsible manner. There is no difficulty about the first of these requirements, but there is some difficulty about the second.
There is no difficulty about the first, because the injunctive moralists admit, and indeed insist, that what is enjoined in a moral judgement is a rule of action, such as ‘let everyone always tell the truth’. By the very nature of such rules, anyone who accepts this rule is ipso facto committed to laying the injunction on himself—if he accepts the rule at all. The rule tells anyone what do do in such and such circumstances (namely those in which one is making a statement to another person); and I am someone, and let us suppose I am now in circumstances of the sort. So if I have accepted the rule, I am committed to judging, about myself, that it is right for me to tell the truth to this other person, and also committed to ‘judging myself’, in the sense of condemning myself, if I do not.
But there is a difficulty about the second requirement, that I must judge for myself in an autonomous and responsible way. It is not enough to realize that the rule applies to myself as well as to others. Of course it does. It is of the form ‘let everyone do A in circumstances C’. And I am someone, and I am now in circumstances C. But if I go no farther than that, I have only noticed a logical truism. It is not even enough if I resolve to obey this rule whenever it applies to me, and carry out this resolution inflexibly for the rest of my days. Something more is required if it is to be for me a moral rule. I must accept the rule or consent to it. I must approve of the rule for myself.
But why should I accept it? I may know that a good many other people do accept it, and that if I fail to act in accordance with it, and only tell the truth when it suits me, they will disapprove of me. I much dislike being disapproved of. It is an unpleasant situation and makes social relations very difficult. I suspect also that other people will impose various pains and penalties on me, and I shall not like that either. So I resolve to follow this rule as a matter of expediency, and I act in accordance with it for the rest of my life. After all, honesty is the best policy in the long run, or at any rate the least disadvantageous one.
But if I accept the rule on that ground, it is not for me a moral rule at all, whatever it may be for others. If I am trying to look at the matter as a moral being, such prudential considerations are not relevant. There is only one consideration which is morally relevant when one is trying to decide whether to accept such and such a rule of conduct. It is this. Do I, myself, sincerely approve of this rule? In short, in a case like this what matters from a moral point of view, and the only thing that matters, is just what the critics of the attitudinarian analysis dismiss as irrelevant: namely ‘how I feel’ about the rule of conduct in question. When I consider this rule as carefully as I can, in a cool hour, do I myself feel approval of this rule or not? Unless I do feel approval of it, it is not for me a moral rule, even though I obey it more often than not, or even always. And the word ‘feel’ is important. It is not enough that I say the words ‘I approve of it’, even though I say them to myself privately, in inner speech. What I say to myself may still be said insincerely. Sincerity and insincerity are not merely matters of ‘public relations’. When I accept a moral rule and say to myself ‘Yes, I approve of it’, this inner speech of mine must be an expression of the way I do myself feel about the rule.
What we feel about an action, done or proposed, especially when it is an action of our own, and what we feel about a rule of action, especially in cases where it is applicable to ourselves, is by no means a trivial matter. On the contrary, there is hardly anything which is more important. I shall not try to illustrate its importance by a kind of fable.
A Fable: An Intelligent Being Without Feelings
Let us imagine an intelligent being with no feelings. We will endow him with tactual sensations and sensations of temperature. He is able to feel a table with his hand, he can feel the warmth of a fire and the chilliness of an East wind. We will also endow him with organic sensations. He can feel (i.e. experience) headaches and toothaches, smarts and itches. He can feel hungry or thirsty, if we mean by this that he can experience certain sorts of bodily sensations in his mouth or his stomach. To this extent he does have feelings.
But he cannot feel either pleasure or displeasure. Consequently, there is a sense in which he knows what pain is, and another sense in which he does not. He knows what it is to experience toothache, but he does not know what it is to be displeased by this bodily sensation, or to dislike it. He can neither like nor dislike anything.
On the conative side, he does not differ much from the rest of us. He has wishes or wants, as we do. He has intentions and makes decisions. He carries out his decisions, sometimes successfully, sometimes not. But he does not feel satisfied when he succeeds, or dissatisfied when he fails. Nor does he feel satisfied when he gets something that he wants. The want just ceases for the time being, and that is all. And if he fails to get what he wants, he does not feel dissatisfied. He just finds that the want continues. He does not know what it is to be disappointed, though we can allow him to be surprised when one of his expectations is falsified.
His most striking peculiarity, however, is that he has no emotions at all. He never feels afraid of anything, though he can run away when he is in danger. He never feels angry with anyone, though he can hit someone or expel him forcibly from the room. He cannot feel dismayed or depressed or elated. He has no pity for anyone (nor of course for himself), though he is also incapable of rejoicing at the misfortunes of others. ‘Sympathy’ is a word he cannot understand, though he may be able to ‘weep with them that weep’ according to the Apostolic injunction, if he has sufficient control of his tear-ducts, and he could quite easily sigh with those who sigh, and groan with those who groan. He feels no liking for anyone or anything, and no disliking either. He is quite incapable of affection: he does not know what it is to have a warm feeling for another person, or to feel coldly towards him either. He cannot love anyone or anything, but also he cannot hate anyone or anything. He does not know what it is to feel remorse or repentance for something he has done. He cannot feel guilty about it, as he would if he were an ordinary human being. But he can resolve not to do that sort of thing again; and when he makes such resolutions, he may carry them out more successfully than the rest of us would.
His Moral Education
What kind of a moral life would such a being have? Let us first consider his moral education. To give him every chance of growing up into a moral being, we will suppose that his parents and teachers are very good people (they do of course have feelings, like the rest of us). We will also suppose that he is an unusually intelligent and quick-witted child. He soon learns to understand commands, and it does not take him long to discover that a particular command can be implicitly general. When he is told to-day ‘wipe your dirty shoes on the doormat before you come in’ he understands that he is always to wipe them on the doormat before he comes in, whenever they are dirty. Throughout his childhood, he does what his parents and teachers tell him to do. Or rather, he always does it when it is logically possible to do it; for one command may contradict another, or it may be self-contradictory, for example when his uncle said to him ‘Always tell the truth, but don't tell it when you are forbidden to’. It is true that he cannot very well be punished if by any chance he does something he is forbidden to do, or fails to do something he is told to do, since he is incapable of suffering. But he can be shown that he will not get what he wants unless he obeys, and will get it if he does obey, and he understands this very quickly. The lesson, however, is hardly needed, since he nearly always does what he is told.
To judge from his actions, he is indeed a model child. Later he becomes a model young man too, and not just by the behaviouristic criteria which we used when we concluded that he was a model child. Now that he is older, we must consider what is going on inside him, as well as his outward actions. By degrees, he has succeeded in ‘interiorizing’ these commands which his parents and teachers gave him. He has an ‘inner voice’ of his own now, which tells him to do the sorts of things which his parents and teachers formerly told him to do, and he always or nearly always obeys it. They find that they have no need now to tell him to do anything. They feel very pleased with him, and they say so. He cannot of course feel pleased with himself, because he cannot feel pleased with anything. When they tell him they feel pleased with him, he cannot understand what they are saying. Nor can he feel grateful to them for this tribute they have paid him, because he cannot feel grateful for anything. But he has often heard this unintelligible word ‘pleased’ before, and he knows just what to do when a remark of this kind is addressed to him. He has been brought up to say ‘thank you’ in such circumstances. What else could gratitude consist in, except a disposition to utter these performatory words (for surely saying ‘thank you’ is thanking)? So he says ‘thank you’, loud and clear, and then they feel still more pleased with him.
It would not be quite fair to say that his morality is just a morality of social conformity. It is true that he only does the things he has been told to do by others in his earlier years. But he does not in any way dislike doing them. Of course we cannot say that he likes doing them either; he neither likes nor dislikes anything. Still, he does them spontaneously, in the sense that no one else makes him do them. In doing them, he is obeying orders which he gives to himself, though in content those orders are the same as those which his parents and teachers gave him previously; or very nearly the same, because (being a very intelligent young man) he has tidied up one or two inconsistencies he noticed in them, and the orders which he gives himself constitute a perfectly consistent code of conduct.
Christian theologians would say that his chief deficiency is that he is entirely incapable of love (though of course incapable of hatred also) and that he is therefore something less than a person. They might even say that someone who hates his neighbours is in a better condition, since he is at least capable of loving them and it is at least conceivable that he might be ‘converted’; whereas no such conversion is even conceivable in the case of our imaginary being. It is not possible that his present feelings towards his neighbours should be altered, because he has no present feelings towards them, or towards anything. It would be misleading to say even that he is indifferent towards them, because this would imply that he does (or at least could) like or dislike something, although he neither likes nor dislikes other human beings.
No Feelings, No Approval or Disapproval
But however important it is that our imaginary being is incapable of loving anyone, and although we may be inclined to agree that this does make him something less than a person, the defect which concerns us now is a different one. It is this. He is incapable of approval and disapproval. He cannot either approve or disapprove of anyone or anything, and therefore he does not know the meaning of the words ‘right’ and ‘wrong’ as ordinary people use them. He is quite capable of obeying his ‘inner voice’, and there is no external compulsion which makes him obey. We have supposed that he does obey it much more often than not, and it might even be that he obeys it always. But what he cannot do, or even conceive the possibility of doing, is to obey it because he approves of what it says. Nor can he disapprove of the actions he sometimes thinks of doing which would be contrary to the dictates of his ‘inner voice’, nor yet of actions done by others which his inner voice would forbid him to do, if he were in the same circumstances, though he can refrain from doing such actions, and perhaps he nearly always does refrain from doing them.
Approval, and disapproval too, has an emotional element in it. It is something that one feels. That is why it used to be said, long ago, that morality is something which concerns the heart or (in Hume's words) is ‘more properly felt than judged of. Hume's remark is no doubt an overstatement. Nevertheless, there is no morality where there are no feelings. ‘The heart’ is an essential constitutent of a moral being, though certainly he needs a head as well, since he cannot be a moral being unless he has the capacity of thinking.
The ‘inner voice’ which our imaginary creature has is no substitute for a heart, though it does resemble a heart in being inward. This is because a person must approve of what his inner voice tells him to do, if its injunctions are to be for him moral injunctions. And ordinary moral beings like ourselves, if we have such an ‘inner voice’, may sometimes disapprove of its injunctions. For in their content they are the same as the injunctions which our parents and teachers taught us to obey in our childhood and youth, and some of these might quite conceivably be disapproved of, when we consider them in our mature years. For example, someone's inner voice might tell him ‘Never treat a person with a black skin as an equal’. And then, some day, he might reflect on this injunction, and find that so far from approving of it, he disapproves of it; and then he might resolve to disobey this injunction in future, though he might well find such disobedience rather difficult, since it would be contrary to his habits and his upbringing.
It follows that if there were a rational being with no feelings or no ‘heart’, his code of conduct would not be for him a moral code at all. In its contents it might conceivably be exactly the same as what ordinary people call ‘the moral law’, and it is even conceivable that our imaginary creature might always obey every single one of the injunctions which his code of conduct contains. What he cannot do is to obey them because he approves of them. Approval, and disapproval too, are experiences which he has never had, and is incapable of having.
It is impossible that he should be a moral being. The reason why he cannot be one is that he has no feelings, no ‘heart’. He is not a moral being, however much the rest of us, who do have hearts, may approve of his outward actions.
Here ends my fable. We must now return from the world of imagination to the more thorny realm of philosophical analysis.
Lecture 7, p. 390, above.
Grammar of Assent, Ch. 4. See Series 2, Lecture 5, above.
Cf. Series 1, Lecture 8, above, on the entertaining of propositions.
Cf. Series 2, Lecture 7, pp. 397–9, above.
Lecture 7, pp. 393–4.
Imperatives can also be inconsistent with each other when addressed to the same group of persons, as when ‘Those behind cried “Forward!” and those in front cried “Back 1”’.
See Lecture 7, pp. 396–7.
It seems not to have occured to some adults either, and some of the unnecessary moral indignation by which the world is afflicted may be due to this cause.
This attitude would be the contrary of the attitude of unconditional love discussed above (pp. 392–3) and might be called ‘unconditional hatred’. We may hope that it does not often exist, but it is at any rate logically possible that it might.
Cf. Lecture 7, pp. 396–7.
Some other person? Not necessarily. A might show more moral insight than I have in a moral judgement which he makes about himself—or of course about myself.
Selby-Bigge, British Moralists Vol. II, p. 180 (Section 707).
Historia Augusta, Loeb edition, Vol. III, p. 8 (Valeriani Duo, ch. 5).
In one of Alexandre Dumas' novels about the reign of Henri III of France, two followers of the Duke of Anjou (the King's brother) are discussing the character of their lord, and one of them says of him il est sans coeur. But the Duke of Anjou, if Dumas is to be trusted, nearly always did what people urith hearts would call wrong; our imaginary being nearly always does what people with hearts would call right.
Pp. 401–3, above.