From The Philosophy of C. D. Broad, ed. P. Schilpp, Tudor Publishing Company, (New York, 1959).

BROAD'S ANALYSIS OF ETHICAL TERMS

William K. Frankena

As befits a Knightbridge Professor of Moral Philosophy, C. D. Broad has written a great deal on ethical topics over a long period of time. He has not, however, published any systematic work of any length on ethics. Most of his writings on ethics are either brief or incidental to a treatment of other topics, and many of them are discussions of the views of others rather than statements of his own. His only long work on ethics is largely historical, interpretative, and critical, and offers very little in the way of a constructive system of moral philosophy.

But, though Broad has not written one of this century's important original books on ethics, he has exerted a considerable influence on its moral theory. Possibly no one has done more recently for the emendation of the human understanding in matters meta-ethical than he has. Few if any worthwhile contributions to ethical theory have appeared in English since 1930 which have not been significantly affected by Broad's writings. His younger contemporaries have depended heavily on his expositions and critical discussions of other peoples' views, and have found his classifications and analyses of possible theories, his definitions, distinctions, terminology, and machinery useful or stimulating. To many of them he has seemed to throw a clear light over "the main problems of ethics" which they have regarded as preferable to any answers to those problems which he might have given under the cover of greater darkness. I am myself one of these younger contemporaries whom Broad has helped in his own way, and if I now go on to ask a few questions or to complain a little about what he has written, I hope I may be forgiven.

In this paper I shall ignore Broad's views on normative and psychological questions, and concern myself only with his opinions on meta-ethical ones, that is, with his analysis of ethical concepts, characteristics, terms, judgments, or sentences. This is appropriate, for Broad's own interest in ethics, when not historical, has been a theoretical and not a practical one. My purpose, however, is not to evaluate Broad's meta-ethical views, but only to try to determine what they are. It may come as a surprise to many that there should be any problem about this, since Broad is always so incomparably clear. This is because people have associated him with the position advanced in FTET, and have not followed carefully his subsequent writings. When one does study all of his works in sequence, however, one finds that there is a real difficulty in determining what Broad's analysis of ethical terms is, or even whether there is any such thing at all. Before 1934 his position is clear enough, although it is only tentatively espoused, but after that his writings are singularly non-committal. One senses that his position has changed, but one is hard put to it to say just how. At any rate, I have found the problem of deciphering Broad's views after 1934 most intriguing, and propose here to write out my attempts at solving it, in the hope that they may be of interest to others and move Broad to tell us what his views really are.

I shall begin with an account of Broad's meta-ethical position prior to 1934, and then go on to give a chronological review of his later writings.

I

It is, as we shall see, hardly fair today to tax Broad with simply assuming that ethical judgments are cognitive in the sense of being true or false. It is just, however, to say that before 1934 he always did assume that they are. In an article on ethics in 1914 he says, "Any proposition, whether about goodness or anything else, is either true or false."1 This statement is not made as part of a discussion of the nature of moral judgments, but, even so, it shows how implicitly he then believed that such judgments are cognitive. When he returns to ethics in 1928, and in the years immediately following, he again takes it for granted that ethical terms stand for characteristics, and is only concerned with the nature of these characteristics, our cognition of them, and their relations to one another and to non-ethical characteristics. In FTET, for instance, he defines ethical characteristics as "whatever characteristics are denoted by the words 'good,' 'bad,' 'right,' 'wrong,' 'ought,' and 'duty,' and by any other words which are plainly mere synonyms for some word in this list," and then he adds that "the first and most fundamental problem of pure ethics is whether these characteristics are unique and peculiar," without ever asking whether ethical terms do denote characteristics at all.2 That he did not ask this question then is, of course, not surprising, for it had not yet come to the fore as it has since -- in fact, Broad was himself one of the first to call attention to it in 1934.

In accordance with this assumption Broad usually distinguishes, during the years before 1934, only two kinds of meta-ethical theories: naturalistic and non-naturalistic ones. The former he generally defines as theories which hold that ethical characteristics can be analyzed without remainder into non-ethical ones, the latter as theories which deny this.3 As a result he has to call naturalistic certain views which Moore had regarded as metaphysical and not as naturalistic, namely, views which define ethical characteristics by reference to metaphysical or theological ones. Thus he regards Spinoza and Paley as naturalists -- Paley as a theological naturalist! Sometimes, however, he suggests a rather different definition. He defines a "natural" characteristic as "any characteristic which either (a) we become aware of by sensing sensa which manifest it or by introspecting experiences which manifest it; or (b) is definable wholly in terms of such characteristics and the notions of cause and substance," and a "non-natural" characteristic as one of which neither (a) nor (b) is true.4 On this account a naturalistic theory would not have to say that ethical characteristics can be analyzed into non-ethical ones; it could claim that goodness, etc., are unanalyzable natural properties. Non-naturalistic theories, on the other hand, would not need to insist that ethical properties cannot be analyzed into non-ethical ones; since there seems to be no reason for thinking that ethical properties would be the only non-natural ones, a non-naturalist might argue that ethical properties can be analyzed into non-natural non-ethical ones, which is what Moore's metaphysical moralists held.

In one place in Five Types Broad speaks of three "analyses" of "x is good:" the phenomenalist analysis, the causal analysis, and the a priori concept analysis.5 It is clear that he would regard the first as a form of naturalism, and the last as a form of non-naturalism, but what of the second? "The causal analysis would be that goodness is the property which causes a thing to be generally approved by men." Such an "analysis" does not exhibit goodness to us, and, strictly speaking, it does not analyze it either. It only describes it in Russell's sense; it defines goodness by giving a definite description. For all it says, then, goodness itself, if ever known by acquaintance, may turn out to be natural or it may turn out to be non-natural. Broad does not evaluate this theory here or elsewhere in his writings, and he goes on in FTET to subscribe to the a priori concept analysis. I mention it because it shows that even before 1934 he took seriously at least one type of theory which, like the emotive theory, cannot be labelled naturalistic or non-naturalistic, although it is still a theory which regards ethical judgments as true or false, and ethical terms as in some sense standing for characteristics, though not naming them.

In the 1914 article Broad takes no stand on the issue between naturalism and non-naturalism, being concerned with the views of Moore and Russell about the relation of the rightness of an action to the value of its consequences. In the 1928 article he is concerned with the "analysis of some ethical concepts," but the only strictly ethical concept he takes up is rightness. With regard to it he says,

I very much doubt whether "rightness" can be defined. I am almost certain that it cannot be defined in non-ethical terms. And I see no reason to think that it can be defined in terms of other ethical concepts, such as "good." At any rate, I do not know, and cannot think of, any satisfactory definition. . . . Rightness is a species of fittingness or appropriateness . . . but [this] is not a definition of it. For, so far as I can see, rightness is a quite unique kind of appropriateness, just as red is a quite unique kind of color.8
Here, then, Broad is almost certainly a non-naturalist, at least about rightness. In Mind and Its Place in Nature (1929), again, while considering the logical status of ethical arguments with factual conclusions, says,
I assume . . . that there are certain purely ethical characteristics, i.e. charcteristics which cannot be identified with or defined in terms of non-ethical or "natural" characteristics. I should consider that the characteristics of being "itrinsically good" or "right" or "a duty" are examples of purely ethical characteristics.7

The non-naturalistic view thus adumbrated is more fully presented, though still rather briefly and with characteristic tentativeness in FTET (1930), partly in passing and partly in a concluding confessio fidei.8 (1) "I . . . think it very likely, though not absolutely certain, that Ethical Naturalism is false, and that ethical characteristics are sui generis." (2) Our concepts of ethical characteristics are a priori and not empirical. Sometimes this is asserted categorically; sometimes hypothetically, as being almost certainly true if ethical concepts are sui generis. Always it is added that if ethical concepts are a priori, then they are the work of reason, although it is admitted that reason cannot form these concepts unless experience provides it with suitable occasions in the way of feelings of approval and disapproval. (3) There are synthetic necessary propositions connecting ethical with certain non-ethical characteristics, and they can be seen to be necessary by reason through intuitive induction. This again is sometimes asserted categorically, and sometimes hypothetically, as being true if ethical naturalism is false. (2) and (3) together make Broad a rationalist and an intuitionist, of what he calls the milder sort, both about ethical concepts and about universal ethical judgments. (4) "I am almost certain that 'right' and 'ought' cannot be defined in terms of 'good.' But I am not sure that 'x is good' could not be defined as meaning that x is such that it would be a fitting object of desire to any mind which had an adequate idea of its non-ethical characteristics."

The grounds on which Broad maintained this non-naturalistic theory are only scantily indicated. There is no mention of the naturalistic fallacy. The open question argument used by Sidgwick and Moore is carefully weighed and found wanting.9 No examination is made of possible naturalistic theories other than that of Hume, except by way of a restatement and discussion of Sidgwick's critique of four such theories.10 In discussing Hume, however, Broad gives an interesting argument which can be generalized against all naturalistic theories. Perhaps because he assumes that all ethical theories are cognitive, he interprets Hume, I think incorrectly, as maintaining a certain kind of naturalistic theory which he calls "phenomenalistic." According to Hume, he says, "goodness is the characteristic of being generally approved by men."11 Commenting on this position, Broad remarks that Hume's arguments against the rationalists neither refute his opponents nor prove his own case.

. . . it remains possible that he is right and they are wrong. I cannot profess to decide the question here; but I will end by pointing out one consequence of Hume's view. This is that every dispute on questions of right and wrong is capable of being settled completely by the simple method of collecting statistics . . . by experiment, observation, collection of statistics, and empirical generalization. This seems to me simply incredible. I should accept the view that there is a point in any ethical dispute . . . beyond which further argument becomes futile. . . . But . . . the logical consequences of Hume's theory is not [this. It] is that all such disputes could be settled, and that the way to settle them is to collect statistics of how people in fact do feel. And to me this kind of answer seems utterly irrelevant to this kind of question. If I am right in this, Hume's theory must be false.12
We may take this as objecting to all psychological theories of the sort which Broad later calls "public" or "transsubjective" on two scores: (a) that not all ethical disputes can be settled by argument, (b) that none of them can be settled by "experiment, observation, collection of statistics, and empirical generalization." But the same two objections would seem to rule out all other forms of naturalism, as well as what Broad calls the causal analysis, leaving only non-naturalism as a possibility for him, since he has not yet envisaged the emotive theory or any other kind of non-cognitive theory.

Some points in the sort of view presented in FTET are worked out by Broad in EMcP, Volume I (1933).18 He expounds two epistemological theories to which a non-naturalist may resort in order to explain the formation of a priori concepts: the theory of innate ideas and that of non-perceptual intuition. He also gives an elaborate and interesting treatment of other topics relating to intuitionism, namely, the nature of analysis, the kinds of definition, and the meaning of simplicity or indefinability. All of these discussions are valuable to anyone trying to understand the intuitionist position. Broad prefaces them, however, with a remark in which one senses a growing uncertainty about non-naturalism in ethics. He says,

. . . although it is quite clear that no naturalistic analysis of ethical characteristics with which I am acquainted is satisfactory, it is certainly not clear to me that none could be satisfactory. Still, we may say that the view that our concepts of ethical characteristics are a priori is quite plausible enough to be worth consideration.14

II

Before 1934, then, Broad was a cognitivist in his conception of the nature and function of ethical judgments, and, with some uncertainty, an intuitionist or non-naturalist. After that, however, he no longer takes it for granted that ethical terms stand for characteristics or properties. He still talks, a good deal of the time, as if they do, but he always mentions at crucial places the possibility that they do not. Apparently he was led to take this possibility seriously through reading an unpublished work by A. E. Duncan-Jones in which an emotive theory of ethical terms was expounded, but it was, of course, kept before his mind's eye by Ayer, Russell, Stevenson, and the others who propounded non-cognitive theories after 1933.15 However, Broad not only is aware throughout this period of the possibility of what he calls "the Interjectional Analysis" of ethical utterances; he might be aware of this and yet be a convinced intuitionist, as Ross and Ewing are. He may also be giving up the non-naturalism of his earlier works. At any rate, he never again commits himself, even tentatively, to any form of non-naturalism, though he nowhere explicitly rejects it in all its forms; and he shows occasional signs of being inclined to some other kind of position, though he nowhere accepts any such position in so many words.

Thus we come face to face with the problem of this paper: what position about the analysis of ethical expressions, if any, is Broad inclined to subscribe to after 1934, and on what grounds? To answer this question it seems best to make a chronological review of his ethical writings from 1934 on, in an effort to follow act by act the play of Broad's mind.

The first essay which bears on our problem is entitled, "Is 'Goodness' a Name of a Simple Non-Natural Quality?"16 Here Broad states Moore's theory of value and reviews it "in the light of our present knowledge and beliefs." His article is very clear and helps a great deal in seeing just what Moore's position implies, and what questions are involved in evaluating it. But when we ask what it reveals about Broad's own state of mind, it turns out to be very difficult to interpret. It can be regarded as being wholly noncommittal on Broad's part, for he does not say that he denies or even doubts any of the six propositions he ascribes to Moore, but it can also be interpreted as questioning or implicitly rejecting every one of these propositions.

Three parts of the essay are important for our purposes. Discussing the question whether "good" stands for a characteristic at all, Broad introduces Duncan-Jones' emotive theory, which is very similar to that proposed a little later by Ayer. Following Duncan-Jones he points out that it has a ready explanation for the fact that all attempts to define ethical words in purely non-ethical terms seem to leave out something essential, namely, that this is because "the interjectional, rhetorical, or imperative force which the original sentence derived from the ethical word in it, has Vanished."17 Then, on his own, he argues "that this theory may be further supported by reflecting on how we learn ethical words as children," concluding that Duncan-Jones' view is "quite plausible enough to deserve very serious consideration" and must be refuted before one can take up the question whether "good" stands for a simple non-natural quality.18 Moore, he goes on to claim, does not refute it. But Broad does not refute it either; he only supposes "for the sake of the argument" that it is not true, and, even so, in the rest of the discussion he mentions the possibility that "good" does not stand for a property at every point at which it is relevant.

Continuing with the question whether the property denoted by "good" is analyzable, Broad does not say either that it is or that it is not. He insists that it cannot be proved to be unanalyzable, and shows at some length that the attempt to prove that no analysis can be satisfactory "involves some very fundamental and difficult logical points."19 He seems to feel, though he does not assert, that these difficulties cannot be overcome, but he does not draw the conclusion that goodness is not a simple quality. He agrees that naturalistic analyses seem either to leave something out or to be too complex, but adds, "I am not much impressed with the importance of a widespread feeling that a proposed analysis is unduly complex." As for the feeling that such analyses are inadequate -- he reminds us again that Duncan-Jones' theory will account for it. And, of course, he adds that, even if "good" does not name a natural property, it may still be analyzable "partly in ethical and partly in non-ethical terms."

The last question is whether or not "good" stands for a non-natural characteristic. Broad first finds Moore's account of the distinction between natural and non-natural properties unsatisfactory, and offers the alternative epistemological definition of the distinction which was quoted above. Next he contends that, if goodness is simple, then it is "almost certainly" non-natural. I find this contention doubtful, but so far all is clear. Now, however, Broad says, " . . . I do not think that it has been proved or could be proved that 'good' is the name of a simple quality. Indeed, I am now going on to argue that there are considerable epistemological difficulties in holding that 'good' is the name of a simple quality."20 He then argues that, since it is almost certain that goodness is non-natural if it is simple, then one who regards it as simple almost certainly must hold (1) that we can have intuitive ideas of non-natural properties, and (2) that there are a priori notions and our notion of goodness is one of them; adding that, if one believes also that goodness has necessary connections with certain natural properties, then he must hold (3) that we can make synthetic a priori judgments.

I find this puzzling. In the last sentence quoted, Broad implies that one who holds goodness to be simple is involved in serious epistemological difficulties, namely (1), (2), and (3), and this seems to be intended as throwing doubt on Moore's position. But, on the other hand, Broad does not himself take (1), (2), or (3) to represent serious epistemological difficulties. About (1) he writes, "although [the principle that we cannot have intuitive ideas of non-natural properties] does seem to me highly plausible, I am not prepared to accept it as self-evident. I am therefore not prepared to conclude that no characteristic of which I can have an intuitive idea could be non-natural."21 In connection with (3) he says similarly, "I do not find this principle [that there can be no synthetic a priori judgments] in the least self-evident myself. . . ."22 As for (2), he does not say so here, but in other passages, some written earlier and some later, he makes it quite clear that he also does not find the view that there are no a priori concepts self-evident, and even shows signs of believing that there are such concepts.23

The point is that the "epistemological difficulties" which Broad finds in Moore's position will lead one to reject this position only if one is inclined to be an empiricist, and Broad does not seem much inclined to be one. He does seem to find empiricism "plausible," but this can hardly mean that he is ready to accept it. But then, although his article does bring out the epistemological consequences of the view that goodness is simple, and so may confirm the opponents of this view in their denial of it, it does not tell us what Broad thinks. So far as I can see, he says nothing which is incompatible even with Moore's form of non-naturalism. But, even if he means to reject this, he may still be subscribing to a non-naturalism which defines "good" in terms of "fitting," although, of course, almost everything that he says about Moore's theory will also apply to any other non-naturalistic position. But what he says is also compatible with an emotive theory, and even naturalism does not seem to be definitely excluded by it.

III

Broad's inaugural lecture, Determinism, Indeterminism, and Libertarianism, also published in 1934, may give us some light on his state of mind. Broad argues here that our ordinary moral judgments involve the notion of categorical obligation or obligability, and that categorical obligability almost certainly entails more than indeterminism and very probably entails libertarianism. But libertarianism is "self-evidently impossible." "It is therefore highly probable," he concludes, "that the notion of categorical obligability is a delusive notion, which neither has nor can have any application."24 We need not evaluate the argument, for it is only this conclusion that concerns us. While it is somewhat hypothetical, it seems to commit Broad to saying that the non-naturalistic kind of ethics to which he had subscribed is very probably mistaken, centering as it does on the ordinary notion of moral obligation. In fact, stopping where it does, it suggests a kind of ethical scepticism, according to which morality involves the notion of a relational characteristic of obligability which cannot be reduced to non-ethical terms but which is delusive in the sense of not applying to anything.25 This seems to be what one of Broad's colleagues felt when he said after the lecture, "If that is what you really believe about your subject, I should think that, if you had any duties, the first of them would be to resign the Chair." Broad's reply is amusing, but it hardly helps to end our quest for his analysis of ethical terms: "I can see what he meant, but I have seldom allowed conscientiousness to degenerate into fanaticism and I have continued to draw my salary ever since."26 I still wish I could see what he meant.27

With the interpretation of the 1934 article on Moore in mind, one reads with interest Broad's contribution to a symposium on the question "Are there Synthetic A Priori Truths?" (1936).28 For if he here finds it unlikely that there are, then one has some reason for concluding that he there meant to be throwing doubt on intuitionism in general. And, indeed, Broad does say here that "the theory that all ostensibly a priori propositions are analytic . . . has enough plausibility to be worth serious consideration."29 His main point, however, is that the principle that all truths are either analytic or empirical must itself be either a priori or empirical. If it is a priori, then it is a synthetic a priori truth, for it can hardly be regarded as analytic. To be consistent, therefore, one must hold that it is an empirical generalization about statements. But then it is only probably true and to determine its truth

. . . what is wanted is a detailed examination of (and not vague 'gas' about) logic, arithmetic, algebra, geometry, and ethics, in order to see whether representative propositions in each of these ostensibly a priori subjects can all be shown to be synthetic empirical propositions about certain kinds of linguistic usage.30

Since the paper stops at this point, we are left wondering. Does Broad believe that all ostensibly a priori propositions can be shown to be analytic in the sense of being empirical propositions about certain kinds of linguistic usage or does he not? Most readers probably assume that he does not. But he does not say so, and he does say that this theory is plausible enough to warrant serious consideration! In short, his discussion here does nothing to unveil the mystery we are probing.

In 1938, in his EMcP II, Part II, Broad makes two statements relevant to our problem:

I think it must be admitted that, when this confusion [mistaking a necessary and reciprocal synthetic connection between goodness, on the one hand, and some set of non-ethical characteristics, on the other, for an analysis of the former in terms of the latter] is avoided, no proposed definition of "intrinsically good" or "intrinsically bad" in wholly non-ethical terms seems to be at all plausible.31
This seems to rule out the possibility that Broad is moving toward some form of naturalism. He then says, about Osborne's view that "x is intrinsically good" means "x is an object which it is fitting for a human being to desire as an end,"32
It seems to me that, if there is a pair of characteristics of which "intrinsically good" and "intrinsically evil" are names, it is quite plausible to suppose that they can be analysed in this way. Any reader who is interested to know what I have to say on the whole subject may consult my paper, "Is 'Goodness' a Name of a Simple Non-Natural Quality?"33
This tells us that the 1934 paper contains Broad's views on the analysis of "good" but helps little to tell us what those views are. It does say, however, that if "good" stands for a property, and the italicized "if" indicates that Broad has a real doubt about this, then it stands for the non-natural property of being fitting to desire. From the two statements taken together then, it seems reasonable to infer that Broad would choose between the emotive theory and Osborne's form of non-naturalism, rejecting naturalism and Moore's kind of non-naturalism.

But which position does he prefer, that of Duncan-Jones or that of Osborne? In "Ought We to Fight for our Country in the Next War?" written "some years before the outbreak of the second World War," Broad gives an excellent and typically intuitionist or non-naturalist account of ethical arguments and reasoning, but perhaps we should not read anything into this fact. He then writes about the question stated in his title,

. . . I have no idea what is the right answer to this question, and, if I had, I should not be able to prove it to people who accepted different ethical principles and premisses from those which I accept. I am not sure indeed that it is the kind of question to which there is an answer, even laid up in Heaven, as Plato might say.34
The first part of this passage could have been written by an intuitionist; it is, in fact, reminiscent of Broad's comment on Hume. The second part, however, could not, for an intuitionist would have to think that there was an answer at least in heaven. Indeed, as Broad implies in his remark on Hume, a naturalist would have to think there was an answer too, though probably not in heaven. This passage, therefore, seems to indicate that Broad is disposed to choose the side of Duncan-Jones. This interpretation, however, appears to be cancelled out by the paper on "Conscience and Conscientious Action" (1940).35 It is true that Broad is careful to speak, not of "moral cognitions," but of "ostensible moral cognitions," thus avoiding a cognitivist commitment, and that he goes on to work out definitions of "having a conscience" and "being conscientious" which can be accepted by an emotivist. In less guarded moments, nevertheless, he talks like an intuitionist, and, what is more to the point, he refers to emotivists as "ethical sceptics" in such a way as to imply that he is not one of them.

IV

Broad's next piece is a review, in 1940, of Ross's Foundations of Ethics.36 He says that he is not impressed by Sir David's argument to show that "good," in at least one of its senses, is the name of a non-natural quality, and that he believes the following account of the facts to be equally plausible. Admiration is an emotion which it is appropriate to feel in the presence of any of a large number of natural characteristics like courage and intelligence. " 'To be good' . . . simply means 'to have one or other of the natural characteristics which make an object fit to be admired.' " Each of these natural characteristics counts as a fitting-making characteristic, and there is no need to have both a non-natural quality of goodness, grounded on these natural characteristics, and a non-natural relation of fittingness grounded on this non-natural quality, as Ross has. Even if we discount the fact that Broad merely claims that this account and that of Ross are equally plausible, which may be going too far, we learn here only that Broad rejects the Moore-Ross view that "good" in one of its senses stands for a simple non-natural quality. The sentence quoted suggests a form of non-naturalism similar to that of Osborne and Ewing, but it does not necessarily commit Broad to such a position. Everything depends on what he would say about the term "fit," and this he does not indicate. In 1942 Broad published a revised version of part of his 1934 paper on Moore, introducing some important changes but still leaving us in the dark about the drift of his own thinking.37 Then he argued that maybe "good" does not stand for a characteristic at all, but if it does stand for a simple one, then it almost certainly stands for one which is non-natural, and there are "epistemological difficulties." Here he repeats that "it is by no means certain that ['good'] is the name of a characteristic at all," and he argues as before that, if it names a simple characteristic, then it almost certainly names one which is non-natural. But he adds that if "good" names a complex characteristic there is also good reason to think that it names a non-natural one. For, he says,

I know of no proposed definition of goodness in purely natural terms which is in the least plausible. But . . . it would not be unplausible to suggest that "x is intrinsically good" means that x is something which it would be right or fitting to desire as an end.38
In other words, if "good" stands for a characteristic at all, that characteristic is probably non-natural.39 Broad does not, however, say that then Moore is involved in epistemological difficulties, as he did earlier. He first comments that "anyone who saw reason to doubt the existence of non-natural characteristics . . . might fairly use this conclusion as an argument to show that 'good' is not a name of a characteristic at all."40 Does he say this because he doubts the existence of non-natural properties? In the next sentence he writes, "It will be worth while to develop this line of argument a little further." Then he points out as before that Moore's position involves admitting a priori concepts and synthetic a priori truths, and again says that the empiricist denial of these epistemological doctrines is highly plausible but not self-evident. Is he denying Moore's view then? If so, on what ground? And which view is he inclined to espouse instead, an emotive theory or Osborne's kind of non-naturalism? If the latter, why does not what he says about Moore's theory hold against his own? All that seems clear is that he is still rejecting naturalism.

Next in the sequence of Broad's ethical writings appear two pieces on evolutionary ethics -- a discussion in 1942 of C. H. Waddington's views on the relation between science and ethics, and a review in 1944 of Julian Huxley's Evolutionary Ethics. Nothing of interest here occurs in the latter.41 In the former, however, Broad ends by stating as follows a view which he claims is what Waddington really had in mind.42

  1. There is a certain group of interconnected emotions which may be called "ethical." Examples of these are moral approval and disapproval, feeling of guilt, feeling of obligation, and so on. An ethical belief is a belief which is toned with one or more of these emotions. Such emotions act as motives for or against doing actions towards which they are felt, and so we have specifically moral motivation.
  2. The study of young children shows that in the main ethical emotions become attached to actions which hinder or promote the adjustment of the child's social relations with his family in general and his parents in particular. He acquires a moral motive against doing the former and for doing the latter.
  3. A certain kind of ethical emotion becomes attached to a certain kind of action through the child doing such actions impulsively or instinctively and then finding that the reactions of his parents are satisfactory or unsatisfactory to him.
  4. From this we infer that the "function" of ethical emotions is to enable individuals to live in social relations with each other; just as the "function" of the lungs is to aerate the blood, and that of the heart to distribute it throughout the body.
  5. In particular cases a type of action which is detrimental to social harmony may have become associated with an approving ethical emotion, or one which would conduce to social harmony may have become associated with a disapproving ethical emotion. In such cases we say that ethical judgments about such actions are "false." This just means that these particular ethical judgments fail to perform that "function" which is characteristic of ethical judgments as a whole in human life. To call an ethical judgment "false" would be like calling a certain state or process in the heart or lungs "unhealthy" or "abnormal."
  6. A study of the genesis of ethical emotions and beliefs in the infant and of the part which they play in making family-life possible suggests to us the function of such beliefs and emotions in the life of the race. But in order to determine the latter more precisely it is necessary to consider the main trend of change in social relations throughout human history. We then recognize that the "function" of ethical beliefs and emotions is to keep human social relations changing in this direction and to prevent them from deviating from it or reverting within it. To call a particular ethical belief "false," then, means that it fails to perform this, which is the characteristic function of ethical belief as such.
This passage is interesting because the view here expounded is similar to the view which Broad himself seems to accept in the essay to be taken up next. Here, however, he only throws it out "as a suggestion for critics of Dr. Waddington to consider." He does not criticize it himself, but it looks as if he would, for he says,
I do not propose to criticize it myself here and now. But I would conclude by asking [Waddington's critics] to look with a very attentive eye at the notion of "function," which plays so large a part in my statement of the theory. I wonder whether this has not teleological and perhaps even ethical overtones which carry us beyond the methods and presuppositions of ordinary natural science.

The last sentence appears to imply that Broad is of the opinion that "ethical overtones" lie "beyond the methods and presuppositions of ordinary natural science." Indeed, earlier in the paper he asserts categorically that an ethical belief cannot be supported or refuted by the methods of natural science, that is, by sense-perception and introspection aided by experiment and generalized by problematic induction. In other words, Broad is still against naturalism.

V

We come thus to the most important of Broad's writings on ethics after 1934, namely, "Some Reflections on Moral Sense Theories" (1944-45),43 which I find exciting because in it Broad seems to be propounding and defending a kind of naturalism after all. He considers here only "deontic sentences" like "That act is right," and begins by distinguishing three "analyses" of such sentences: (1) interjectional or emotive theories, (2) subjective theories (autobiographical, transsubjective, etc.), (3) objective theories (some naturalistic, others non-naturalistic). Of all these theories, he says, only three are worth serious consideration: the interjectional theory, the moral sense form of the transsubjective dispositional theory, and objective theories of a non-naturalistic kind.

About the interjectional theory he says very little, namely, that it "can be swallowed only after one has undergone a long and elaborate process of 'conditioning' which was not available in the eighteenth century."44 This only explains why Richard Price would not have taken the theory seriously; it does not necessarily imply that Broad himself has not undergone the conditioning required to swallow it.

The only objective or non-naturalistic theory which Broad takes up here is the naively realistic form of the moral sense theory. By this he means the view which thinks of deontic cognition as a naive realist thinks of the perception of yellow. This view Broad definitely rejects, after discussing it at some length, and he seems at the same time to be rejecting Moore's theory, for he remarks that Moore in Principia Ethica was probably tacitly assuming something like a naively realistic interpretation of both ethical and perceptual judgments.45 But does he mean to be rejecting all other forms of non-naturalism too? He does not say, for he is concerned only with moral sense theories, and does not seem to regard the Osborne-Ewing form of non-naturalism as naively realistic. He does remark once more that,

. . . if ethical terms . . . are simple and non-naturalistic or are complex and contain a non-naturalistic constituent, then the concepts of them must be wholly or partly a priori . . . and such judgments as "any act of promise-keeping tends as such to be right" must be synthetic and a priori. Now it is a well-known and plausible epistemological theory that there are no a priori concepts and no synthetic a priori judgments.
But then he adds,
If I am right, anyone who feels no doubt about this epistemological theory can safely reject the analysis of moral judgments which makes them contain non-naturalistic constituents. On the other hand, anyone who feels bound to accept that analysis of moral judgments will have to reject this epistemological theory.48
He does not tell us whether he accepts this epistemological theory or that analysis of moral judgments.

Later on in the essay, however, he finally makes clear what he thinks of epistemological objections to non-naturalism. About the argument that there are no a priori concepts he writes,

For my part I attach very little weight to this argument. I can see nothing self-evident in what I will call for short "Hume's Epistemological Principle," and I am not aware that any conclusive empirical evidence has been adduced for it. It seems to me to be simply a useful goad to disturb our dogmatic slumbers, and a useful guide to follow until it begins to tempt us to ignore some facts and to distort others. I am inclined to think that the concepts of Cause and of Substance are a priori or contain a priori elements . . .47
As for the contention that there are no synthetic a priori propositions -- he observes that "Such an argument would have different effects on different persons," indicates that he is neither unmoved nor overwhelmed by it, and concludes that "it is rather futile to rely on a general argument of this kind."48 We must infer, then, that he does not really mean to argue against Moore on epistemological grounds, as he seems to in the 1934 paper.

By the transsubjective dispositional form of the moral sense theory, Broad means the view that "x is right" means "Any normal person would feel a peculiarly moral pro-emotion towards x, if he were to contemplate it in a normal state." It is then a naturalistic theory similar to that which Broad ascribes to Hume. But Broad does not therefore attack it, as one might expect. Instead, he shows how it may be defended against attack. He states the three main difficulties of the theory, including some that Sidgwick and he had in mind in criticizing such theories as Hume's, and then proceeds to show, not that the theory cannot answer them, but that it can. The three questions are,

(i) Can it deal with the fact that judgments like "That act is right" seem always to be grounded upon the supposed presence in the act of some non-ethical right-inclining characteristic, such as being the fulfilment of a promise? (ii) If so, can it deal with the further fact that the connexion between a right-inclining characteristic and the rightness which it tends to convey seems to be necessary and synthetic? And (iii) can it deal with the fact that it seems not only intelligible but also true to say that moral pro-emotion is felt towards an act in respect of the characteristic of rightness and moral anti-emotion in respect of the characteristic of wrongness?40

We need not run through Broad's answers to these questions in detail. The important point is that he concludes that the naturalistic theory under discussion can deal plausibly with these facts and then stops. He does not actually say that he accepts the theory, but he seems inclined to. Two passages are of especial interest. In the first Broad is explaining part of the answer to the second question, and what he writes sounds very like the view which he ascribes to Waddington in one of our earlier quotations.

The defensive part of the argument might take the following line. Civilised men throughout human history have been assiduously conditioned in infancy and youth by parents, nurses, schoolmasters, etc., to feel moral pro-emotions towards acts of certain kinds and to feel moral anti-emotions towards acts of certain other kinds. Moreover, if we consider what kinds of acts are the objects of moral pro-emotions and what kinds are the objects of moral anti-emotions we notice the following facts about them. The former are acts whose performance by most people on most occasions when they are relevant is essential to the stability and efficient working of any society. The latter are acts which, if done on many occasions and by many people, would be utterly destructive to any society. On the other hand, the former are acts which an individual is often strongly tempted to omit, and the latter are acts which he is often strongly tempted to commit. This is either because we have strong natural impulses moving us to omit the former and to commit the latter, or because the attractive consequences of the former and the repellent consequences of the latter are often remote, collateral, and secondary. It follows that any group of men in which, from no matter what cause, a strong pro-emotion had become associated with acts of the first kind and a strong anti-emotion with acts of the second kind would be likely to win in the struggle for existence with other groups in which no such emotions existed or in which they were differently directed. Therefore it is likely that most of the members of all societies which now exist would be descendants of persons in whom strong moral pro-emotions had become attached to acts of the first kind and strong anti-emotions to acts of the second kind. And most existing societies will be historically and culturally continuous with societies in which such emotions had become attached to such acts. These causes, it might be argued, conspire to produce so strong an association between such emotions and such acts in most members of every existing society that the connexion between the emotion and the act seems to each individual to be necessary.50
About this answer Broad says,
No doubt this line of argument will produce different effects on different persons. For my part I am inclined to attach a good deal of weight to it.51

In the other passage, which comes at the very end of the essay, Broad is concluding his answer to the third question.

What is happening when a person is said to be feeling a first-hand moral emotion towards an act in respect of his belief that it is right or that it is wrong? I can give only a very tentative answer to this question, based on my own imperfect introspection of a kind of situation with which I am not very familiar. It seems to me that in such cases I do not first recognise a quality or relation of rightness or wrongness in the act, and then begin to feel a moral pro-emotion or anti-emotion towards it in respect of this knowledge or belief. What I seem to do is to consider the act and its probable consequences under various familiar headings. "Would it do more harm than good? Would it be deceitful? Should I be showing ingratitude to a benefactor if I were to do it? Should I be shifting onto another person's shoulders a burden or a responsibility which I do not care to bear myself?" In respect of each of these aspects of the act and its consequences I have a tendency to feel towards the act a certain kind of moral emotion of a certain degree of intensity. These emotional dispositions were largely built up in me by my parents, schoolmasters, friends and colleagues; and I know that in the main they correspond with those of other persons of my own nation and class. It seems to me that I call the act "right" or "wrong" in accordance with my final moral-emotional reaction to it, after viewing it under all these various aspects, and after trying to allow for any permanent or temporary emotional peculiarities in myself which may make my emotional reaction eccentric or unbalanced. By the time that this has happened the features which I had distinguished and had viewed and reacted to separately have fallen into the background and are again fused. They are the real mediating characteristics of my moral pro-emotion or anti-emotion; but I now use the omnibus words "right" or "wrong" to cover them all, and say that I feel that emotion towards the act in respect of my belief that it is right or that it is wrong.52

What makes this passage especially interesting is the fact that, although he does so tentatively, Broad seems to be giving as his own an account of what happens in a first-hand deontic judgment which cannot be accepted by an intuitionist, but only by one who holds what he calls an "Emotional Reaction" theory. Thus he seems at last to have parted company with the non-naturalists altogether.

In doing so, if this is what he is doing, Broad appears to think of himself as adopting an emotional reaction theory of a naturalistic kind, namely, a transsubjective dispositional form of the moral sense theory, as he so winsomely calls it. It should be remarked, however, that what he says is also compatible with an emotional reaction theory of a non-cognitivist kind. In fact, the passages quoted seem to me to fit in with some kind of interjectional analysis better than with any naturalistic one.

It is possible, then, that Broad has not clearly distinguished and chosen between the above two kinds of emotional reaction theory, even if he has given up non-naturalism. I may, however, be misinterpreting him in suggesting that he has done the latter, for he does not openly declare himself and is characteristically non-committal throughout most of the essay.

VI

In any case, the end of our quest for Broad's analysis of ethical terms is not yet, though we have not far to go. There is the well-known paper, "Some of the Main Problems of Ethics" (1946).53 In it Broad writes as if he were subscribing to "the Objective Analysis" of ethical judgments, as he himself points out, but he reminds us that ethical terms may not stand for characteristics at all, and he stops on occasion to explain what account emotional reaction theories will give of the matter in hand, for example, of the notion of right-making characteristics. He does, in one place, point to a certain difficulty for "any form of the Emotional Attitude analysis," but it is a difficulty to which he himself has supplied an answer in the paper on moral sense theories.54

Of more interest for our purposes is a paper on "Hägerström's Account of Sense of Duty and Certain Allied Experiences" (1951).55 Hägerström is one of the pioneer proponents in this century of an interjectional or emotive analysis of ethical utterances, and Broad has recently translated a volume of his essays into English. In the present paper Broad gives a somewhat sympathetic exposition of Hägerström's theory, calling it a form of "ethical subjectivism" or "ethical positivism," and claiming that "none of its Anglo-Saxon adherents has made so thorough and ingenious an attempt ... to show how the various aspects of the admitted facts can be fitted into the theory." Near the end, presenting Hägerström's views about rightness and the sense of duty, Broad says, "Hägerström first considers in detail and rejects, on what seem to me to be adequate grounds, various attempts to identify [i.e. define] 'rightness.' "56 Now, I take Hagerstrom to be rejecting all naturalistic definitions of 'right' in favor of a non-cognitive theory. Is Broad here doing likewise, rejecting even the kind of moral sense theory to which he seemed earlier to subscribe? If so, is he favoring an interjectional analysis or a non-naturalistic one? Or is not that kind of moral sense theory included among the "various attempts to identify 'rightness' " which he is denying? Or is he not distinguishing between it and the emotive theory? It seems to me quite possible that he is not, since they are both emotional reaction theories. This possibility is borne out by the fact that the positive account of moral emotions and judgments which Broad ascribes to Hägerström is strikingly similar to that which he himself suggests at the end of the moral sense paper. In fact, both accounts seem to be compatible with either an emotive or an inter-subjective moral sense theory. Even if we assume that he has given up non-naturalism then, we cannot be sure which of these two other kinds of theory Broad is inclined to hold.

Our last exhibit, a review of S. E. Toulmin's Examination of the Place of Reason in Ethics, is no less puzzling.57 Toward the end Broad restates Toulmin's estimate of the strong and weak points in the subjectivist, imperativist, and objectivist accounts of moral indicatives, and of the places of each of them in a correct and adequate account of moral phenomena. Since he says, "I find myself in general agreement with much in Toulmin's position as thus summarized," I shall quote his summary.

  1. Moral phenomena in general, and the experiences which we express by moral indicatives in particular, are unique and peculiar. The only satisfactory way to investigate them is to do so directly. If we try to force them into moulds derived from reflecting on non-moral phenomena and the verbal expressions for these, we shall inevitably distort them.
  2. The experiences which are expressed by moral indicatives resemble in certain respects judgments asserting an emotional reaction of the speaker towards an object, and in certain respects the experiences which are naturally expressed by uttering interjections or sentences in the imperative. But in each case there are unlikenesses which are as important as the likenesses.
  3. Each of the three theories has arisen through concentrating on the resemblance to one of these non-moral parallels and ignoring the unlikenesses to it and the resemblances to the other non-moral parallels.
  4. The two theories which admit that moral indicatives express judgments, viz., the objectivist and the subjectivist theories, agree in making a certain tacit assumption. They both assume that two judgments about the same thing can logically conflict only if they refer to one and the same property, which one person assigns to the object and the other denies of it. Seeing that moral judgments can logically conflict, the objectivist concludes that the words "good," "right," etc., must be names for properties of a peculiar kind. Seeing that there are no such properties, the subjectivist concludes that moral judgments cannot logically conflict and therefore can only assert or deny that the speaker is reacting emotionally in a certain way to the object.
  5. Mr. Toulmin rejects this common assumption. He holds that moral judgments can logically conflict, but that they do not assert or deny a property of an object. In order that they may logically conflict "all that is needed is a good reason for choosing one thing rather than the other. Given that, the incompatibility of 'This is good' and 'This is not good' is preserved. And that, in practice, is all that we ever demand" (p. 43).
  6. Mr. Toulmin thinks that the imperativist makes the same tacit assumption as the objectivist and the subjectivist. But the imperativist's reaction is to deny both their alternatives and so to conclude that moral indicatives do not express judgments of any kind.
  7. This line of thought is made plausible by concentrating upon singular sentences describing a concrete perceptible fact, like 'The cat is on the mat,' and taking them as the type of all sentences which can possibly express judgments. Here and here only it is sensible to talk of a correspondence between the elements and the structure of the sentence, on the one hand, and those of a certain fact to which the sentence refers, on the other, and to say that truth or falsity consists in the concordance or discordance between sentence and fact. Since moral indicatives plainly do not answer to this pattern, it is assumed that they cannot be true or false, i.e., that they cannot express judgments. But, then, it must be noted that the vast majority of sentences which admittedly express judgments obviously do not fit into this pattern.58

One cannot help wondering how much of this Broad accepts. Does he, for example, agree "that moral judgments ... do not assert or deny a property of an object?" If so, then he is here rejecting both naturalism and non-naturalism, in favor of a view which he would have to call "interjectional," though Toulmin would object to this label.59 But, perhaps, this is not part of what Broad agrees with Toulmin about. In the next paragraph he goes on to say that he does not clearly understand Toulmin's positive account of what is expressed by moral indicatives.

One aspect of this is stated in Chapter VI under the heading "Gerundive Concepts." These concepts fall under the general formula "worthy to be treated in a certain way." . . . Mr. Toulmin states definitely that gerundive concepts cannot be identified with or defined wholly in terms of de facto subjective attitudes. . . . There is nothing particularly new or startling in this aspect of the theory. It has been very fully developed, e.g., by Sir W. D. Ross and by Dr. Ewing. "Worthiness to be treated in a certain way" is in fact our old friend "fittingness" and, as such, I have no quarrel with it.
The last part of this passage seems to imply that Broad has not given up non-naturalism after all, at least not the Osborne-Ewing form of it. But, if this is so, it is difficult to see how he can be agreeing with very much in Toulmin's position.

Yet, in a sense, it is dramatically fitting that the last episode in our story should be this review of Toulmin. For this review may itself reflect something of the state of mind in which Broad has been since 1934 -- unable to decide finally between the three received types of ethical theory, seeing truth and error in each of them, and not discerning any tenable fourth alternative. This would explain why he finds so much to agree with in Toulmin's book, and at the same time why Toulmin's view does not seem clear or satisfactory to him. It would also explain why his ethical writings after 1934 have the character we have found them to have.

VII

We must now see what our story adds up to. Is there an "analysis of ethical terms" to which Broad is subscribing, at least tentatively, after 1934 or at the present time? It may be that there is not -- that he is a sceptic in meta-ethics, if not in normative ethics. There is no such analysis which he explicitly accepts even in the cautious way in which he accepted non-naturalism in 1930. On the other hand, there is none which he clearly rejects with any kind of finality, so that he may be a non-naturalist, or an emotivist, or a naturalist. It may also be, of course, that Broad has been and still is undecided, sometimes favoring one analysis, sometimes another. We have seen that for every one of the main types of meta-ethical theory, there are some places where Broad seems to lean toward it as at least plausible and others where he seems to oppose it. But, though it is possible that Broad is a non-naturalist still, or that he is uncertain, or that he is a sceptic, it is in my opinion reasonable on the basis of the above evidence to adopt the following hypothesis: that

(1) Broad has been doubtful of non-naturalism since the early thirties and has now essentially given it up, though on what grounds is not clear,

(2) he is still reluctant to accept any form of naturalism according to which ethical disputes can be settled by "experiment, observation, collection of statistics, and empirical generalization,"

(3) he has been moving, uncertainly but definitely, toward an emotional reaction theory of some sort, but

(4) he is undecided as between an interjectional form of the emotional reaction theory, like those of Duncan-Jones, Ayer, and Hägerström, but involving a unique moral emotion, and a transsubjective dispositional form of the emotional reaction theory, also involving a unique moral emotion. That is, he is undecided between the theory that a moral judgment simply expresses a unique moral emotion of the speaker, and the theory that it asserts that any normal person will feel this peculiarly moral emotion toward the object in question if he contemplates it in a normal state. In the paper on moral sense theories Broad appears to prefer the latter, but the rest of the time he prefers the emotive theory to any form of naturalism. It may be, however, that he is not always concerned to keep the two theories apart in his mind, for as we have seen, the passage at the end of the moral sense paper where he comes closest to accepting an emotional reaction analysis is compatible with both of them.

In fact, there may not be much difference in effect between the two theories just described. On either of them one can adopt Broad's favorite notion of right-making characteristics, and one can account for the apparent necessity of the connection between them and "rightness" along lines which Broad himself indicates. Again, on both of them, ethical disputes are "settled" in the same way, namely, by an empirical investigation into the presence or absence of these right-making properties. On neither view would the dispute directly involve any further enquiry of a statistical sort into what people approve. This is clear in the case of the emotive theory, but it may come as a surprise that it can be true of any emotional reaction theory of a naturalistic kind. Even Broad seems usually to think that for any theory of the latter sort an ethical enquiry is really a statistical one. Yet, when he explains how the transsubjective dispositional form of the moral sense theory can account for the apparent necessity of judgments like "Promises ought to be kept," he also shows that it does not necessarily make such judgments statistical. According to this theory such a judgment says, "Any act of promise-keeping would tend to call forth a moral pro-emotion in any normal human being who might contemplate it when in a normal state." Now the latter statement, Broad points out, may be contingent or it may be analytic, depending on the definition of "normal," and he himself seems to prefer to regard it as analytic, like "Any sample of pure water boils at 100°C. under normal atmospheric pressure," even though it is "founded on a whole mass of interconnected empirical generalizations."60 If it is contingent, then it is statistical and Broad's objection to Hume applies, but if it is analytic this is not true.

It may be argued that the dispositional view has the advantage of allowing ethical judgments to be true or false and ethical disputes to be capable of being settled as scientific questions are settled. This will be true if the accepted definition of "normal person" includes "approving of the kind of action in question when in a normal state." But then, if one is really doubtful of the validity of the ethical judgment under discussion, he will simply shift his disagreement to the definition of "normal," and, if he does, how will the new dispute be resolved? Broad says that the definition of "normal," like that of "pure water," is founded on a mass of empirical generalizations, among which will be generalizations about what is approved. Then a statistical study is involved at least at second remove, but, in any case, it cannot be claimed that the definition follows logically from the conclusions of such a study.

If this is so, however, an emotivist may contend that he can do as well as a naturalist, if he is given the same empirical generalizations about what we approve. For then he can claim that "agreement in attitude" will follow upon "agreement in belief," that ethical disputes can therefore be "settled," and that, if ethical judgments cannot be true or false, they can be justified or unjustified, reasonable or unreasonable. On both views, the statistical generalization that people tend to approve certain sorts of action or qualities of character will be of central importance, but on neither will it be a working part of any ethical dispute -- it will not constitute a direct reason for any ethical statement but will only be presupposed by the whole process of giving reasons for such statements.

To conclude -- I may be entirely mistaken in my hypothesis about Broad's later views. If so, I should be only too happy to be told the truth about them, and so, I believe, would many of his other readers. If not, it would be equally gratifying to know which form of the emotional reaction theory he prefers, if he has any preference, and to have a fuller statement of it. In either case, it would be good to hear the grounds on which he takes the position he takes and rejects those he rejects. For example, if he holds the transsubjective dispositional form of the moral sense theory, on what ground does he do so? Is that analysis of rightness a rendition of what we ordinarily mean or is it like the definition of pure water as what, among other things, boils at 100°C. in an atmospheric pressure of 76 centimeters of mercury? If the latter, how would he defend his analysis against others?


1"The Doctrine of Consequences in Ethics," Int. Journal of Ethics, 24, 1913-14, p. 298.

2 P. 257.

3 Cf. FTET, 257.

4 "Is 'Goodness' a Name of a Simple Non-Natural Quality?," Proc. Arist. Soc, XXXIV, 1933-34, 264.

5 Pp. 109-110.

6 "Analysis of Some Ethical Concepts," Journal of Phil. Studies, 3, 1928, p. 295.

7 P. 487.

8 Cf. pp. 178-179, 268-270, 281-283.

9 Pp. 173-174.

10 Pp. 166-170.

11 P. 109.

12 Pp. 114-115.

13 Cf. pp. 38-53,111-127.

14 P. 47.

15 The emotive theory may also have been called to Broad's attention by R. B. Braithwaite, "Verbal Ambiguity and Philosophical Analysis," Proc. Arist. Soc., 28, 1927-28, and by W. H. F. Barnes, "A Suggestion about Value," Analysis, 1, 1933.

16 Proc. Arist. Soc., XXXIV, 1934, 249-268.

17 P. 252.

18 P. 253.

19 Pp. 254-259.

20 P. 267.

21 P. 267.

22 P. 268.

23 Cf. EMcP, I, 47, and a quotation given below from "Some Reflections on Moral Sense Theories."

24 EHP, p. 217.

25 Of course, Broad may be thinking that fittingness is different from moral obligation and not delusive.

26 Op. cit., pp. x-xi.

27 In his class lectures at about this time Broad recurs to the causal analysis or descriptive theory of the meaning of ethical judgments, according to which "x is good" means "There is one and only one characteristic or set of characteristics whose presence in any object which I contemplate is necessary to make me contemplate it with approval, and x has that characteristic or set of characteristics." He says he regards this as "the most satisfactory account of what I mean when I make a first-hand judgment of the form 'This is good' or 'This is bad'. . . ." This seems to mean that he has given up intuitionism, without accepting naturalism or the emotive theory. The descriptive theory, he points out, leaves open the question whether there is a characteristic answering to the description or not, and whether, if there is, it is natural or non-natural. It is true that he applies it only to "good" and "bad," and not to "right" and "ought," but presumably the same reasons that lead him to apply it to the former would also hold of the latter. Then these unpublished lectures may be taken to confirm the hypothesis that Broad moved away from non-naturalism after 1934, though perhaps not in the direction of naturalism.

28 Proc. Arist. Soc. Supp., XV, 1936.

29 Op. cit., p. 110.

30 Ibid., p. 116.

31 P. 659.

32 See H. Osborne, Foundations of the Philosophy of Value, (1933), 94.

33 Loc. clt.

34 EHP, p. 240.

35 Ibid., pp. 244-262.

36 Mind, 49,1940. See p. 237.

37"Certain Features in Moore's Ethical Doctrines," The Philosophy of G. E. Moore, ed. by Paul A. Schilpp. See pp. 57-67.

38 Pp. 64-65.

39 For a brief discussion of his argument see my "Some Arguments for Non-Naturalism about Intrinsic Value," Philosophical Studies, I, 1950, 56-57.

40 P. 65

41 This review in Mind is reprinted in Readings in Philosophical Analysis, ed. by H. Feigl and W. S. Sellars.

42 "The Relations between Science and Ethics," Proc. Arist. Soc., XLII, 1942, 100G-100H.

43 Proc. Arist. Soc., 45. This article was occasioned by a reading of Richard Price and is reprinted in Readings in Ethical Theory, ed. by W. S. Sellars and J. Hospers. My page references are to this volume.

44 P. 365.

45 P. 370.

46 Pp. 363-364.

47 P. 378.

48 P. 379.

49 P. 376.

50 Pp. 379-380.

51 Loc. cit.

52 Pp. 387-388.

53 Reprinted from Philosophy, 21, in Readings in Philosophical Analysis, ed. by H. Feigl and W. S. Sellars, pp. 547-563.

54 See op. cit., p. 561.

55 Philosophy, 26, 99-113.

56 P. 110. One of these "adequate grounds" seems to be the open question argument which he found inadequate in FTET.

57 Mind, 61, 1952.

58 P. 100.

59 Broad should also object to it if he accepts (2).

60 Sellars and Hospers, Readings in Ethical Theory, p. 382.