H. P. Grice, "Stevenson's 'Meaning'"

Stevenson’s major contribution to philosophy was his development of emotivism, a theory of ethical language according to which moral judgments do not state any sort of fact, but rather express the moral emotions of the speaker and attempt to influence others. Stevenson’s emotive theory of ethical language Stevenson always stressed that his work did not include any substantive moral judgments, but rather comprised “analytic ethics,” or what is now commonly called “metaethics,” the branch of moral theory that is about ethics and ethical language. What do we mean when we say that something is good or bad, or right or wrong? On the face of it, we are describing, attributing to the thing some property, goodness or badness, or rightness or wrongness. What could these properties be? How do we find out about them? Much of philosophical moral theory explores various answers to these questions. Stevenson thought that questions about the nature of moral properties were misplaced. Our moral judgments do not, at least primarily, describe at all. Uttering moral sentences has a different function: to express emotions, and to influence or invite others to share them. All of his main contributions appeared in Ethics and Language, 1944, and a collection of papers, Facts and Values, 1963. Distinguish between expressing a certain state of mind and saying that one is in it. If I say, “Ann Arbor is in Michigan,” I express my belief that Ann Arbor is in Michigan, but I do not say that I believe such a thing. For what makes what I said true? Not that I really do believe that Ann Arbor is in Michigan; only the fact that Ann Arbor really is in Michigan. Stevenson’s theory of ethical language, in a nutshell, was that when I say, “Inequality is bad,” I have expressed a certain negative moral attitude toward inequality, though I have not said that I have it. It should be clear why Stevenson stressed that his theory was “analytic” or metaethical, and did not contain any substantive moral judgments. For by claiming that moral judgments serve to express emotions, he had not expressed his own moral emotions at all. Besides expressing the speaker’s attitude, Stevenson said, moral statements also “create an influence,” they invite the audience to share in the emotion expressed. Thus, “x is good” is akin to “Let us approve of x.” Moral exhortation, after all, is commonly Blackwell Companions to Philosophy: A Companion to Analytic Philosophy Edited by A. P. Martinich, David Sosa Copyright © Blackwell Publishers Ltd 2001 used to try to persuade the audience to share the speaker’s suggestions, and moral judgment is often a call to action. Furthermore, in context, ethical statements can come to have some secondary descriptive content; in Victorian England, for example, calling a woman “virtuous” implied that she was chaste. So a Victorian moralist could manage to describe a woman, and not merely to evaluate her (express his emotional attitude toward her and invite others to share it), by calling her “virtuous.” Some advantages of emotivism Stevenson’s theory was enormously influential in the middle of the twentieth century. Taking its cue from Ayer’s short remarks on ethics in Language, Truth and Logic, Stevenson’s theory added sophistication and subtlety (see AYER). In “The Emotive Meaning of Ethical Terms,” Stevenson sets out some criteria for a successful analysis of moral terms, explaining that what he calls traditional “interest theories” of ethical terms fail one or more criteria. These interest theories include the views of Hobbes, whom Stevenson understood to have defined “good” to mean “desired by me,” and Hume, whom he interpreted as defining it to mean “desired by my community.” In the first place, we must be able sensibly to disagree about whether something is “good.” This condition rules out Hobbes’s definition. For consider the following argument: “This is good.” “That isn’t so; it’s not good.” As translated by Hobbes, this becomes: “I desire this.” “That isn’t so, for I don’t . . .” In the second place, “goodness” must have, so to speak, a magnetism. A person who recognizes X to be “good” must ipso facto acquire a stronger tendency to act in its favour than he otherwise would have had. This rules out the Humean type of definition. For according to Hume, to recognize that something is “good” is simply to recognize that the majority approve of it. Clearly, a man may see that the majority approve of X without having, himself, a stronger tendency to favour it . . . In the third place, the “goodness” of anything must not be verifiable solely by use of the scientific method. “Ethics must not be psychology.” This restriction rules out all of the traditional interest theories, without exception. Emotivism appears to be well prepared to satisfy these criteria. First, people can genuinely disagree when one states that X is good and the other states that X is not good; they are disagreeing in attitude, as Stevenson puts it, not in factual belief, but this is genuine disagreement just as plainly as we may disagree when I suggest that we go to the movies tonight and you suggest that we go have a few drinks instead. Second, and perhaps most significantly, if the judgment that X is good is an expression of favorable attitude toward X, then it is clear why anyone making such a judgment will have a tendency to act in favor of X. Finally, while adding to our knowledge by scientific investigation may sometimes resolve certain ethical issues, there can be deeper disagreements that are left untouched by scientific methods. Our emotional attitudes may differ in the face of converging empirical knowledge. A further attraction of emotivism is that it dissolves knotty-looking metaphysical problems of metaethics. Consider the question of whether moral properties are natural properties or some other special sort. G. E. Moore famously argued that moral properJAMES DREIER 176 ties could not be natural properties, and later John Mackie argued for skepticism about the existence of moral properties on the grounds that they could not be natural ones, and the metaphysics and epistemology of non-natural properties is too spooky (or “queer,” as Mackie said) for sober philosophy. Some metaethicists have tried to show how moral properties could be part of the natural world after all, but it is difficult to explain just how our linguistic habits and practices could determine just which natural property moral wrongness could be, given the wide diversity and disagreement in moral values among different people and cultures at different times. Emotivism resolves the issue by denying that moral predicates, like “wrong” and “good,” serve to pick out properties at all. They serve as markers of mood or emotion instead. So the metaphysics of alleged moral properties is avoided if we adopt Stevenson’s view. Some difficulties for emotivism Emotivism is not without its difficulties, and the main ones were leveled at Stevenson soon after he began to publish his views. One criticism was offered by Brand Blanshard in a paper called “The New Subjectivism in Ethics.” Blanshard complained that emotivism has an obviously false implication. When I see a rabbit with its foot caught in a trap, I might say (or think) “That’s a bad thing.” I would then, plausibly, be expressing my negative emotion toward the pain of the rabbit. But suppose I then contemplate the situation in which I myself become very jaded and cease to care about the suffering of sentient animals. Do I (now, actually) say, “Well, in that case, the suffering of the rabbit would not be a bad thing at all”? No, of course not. But emotivism implies that this is how I should think. So emotivism is false. This criticism is instructive, though it is not correct. It illustrates two important points about Stevenson’s theory. First, the fact that we would ordinarily say one thing or another is very important, according to Stevenson’s approach. He would never have replied, “We might not say such a thing in that case, we might steadfastly deny it, but we would be mistaken.” His theory was supposed to account for our ordinary judgments, and not to reform those judgments. So it is important whether Blanshard’s example really does show that emotivism sometimes contradicts our ordinary ethical judgments. However, the criticism is unsound, because Stevenson’s approach does not, in fact, imply that we do or should judge that the suffering of the rabbit would not be at all bad if we were jaded and uncaring. To think that it does imply such a thing is to mistake emotivism for a poor relation, subjectivism. The subjectivist thinks that “bad” means (something like) “apt to cause a negative emotion in me.” So to call something bad, according to subjectivism, is to say that it causes a negative emotion in oneself. But Stevenson took great pains to distinguish his own view from subjectivism, and he gave very similar examples to show why subjectivism is incorrect. According to emotivism, remember, calling something bad is not saying that it does or doesn’t do anything – that would be to describe the thing. Ethical language does not (primarily) describe a thing or an emotion or the speaker, it expresses the emotion of the speaker. When I contemplate the situation in which I heartlessly feel no sorrow over the rabbit’s suffering, I (right now, actually) feel rather bad about that, and if I were to express my emotion I would say, “That would be a bad thing.” C. L. STEVENSON 177 Probably the most influential criticism of Stevenson, the criticism that later emotivists (and fellow non-descriptivists, see below) have been most concerned to address, was a problem noticed by Peter Geach and John Searle. It is sometimes called the “embedding problem.” To put it succinctly, the problem is that even if emotivism really does tell us what somebody does when she asserts a simple moral sentence like “It is wrong to kick cats,” it does not seem to tell us what such a sentence means. For there is more to the meaning of a sentence than the facts about what is accomplished or expressed by an assertion of it, since we can use sentences without asserting them, in unasserted or embedded contexts. There are many kinds of unasserted contexts. Here are a few examples; notice that in no case would someone sincerely uttering the entire sentence be asserting that it is wrong to kick cats. If it is wrong to kick cats, then it is wrong to kick Tibbles. Either it is wrong to kick cats, or there is nothing wrong with kicking people. I wonder whether it is wrong to kick cats. Do you mean to say that it is wrong to kick cats? Many other kinds of examples could be given, but the idea is clear enough. Critics of emotivism point out that what Stevenson said about the emotive meaning of ethical terms does not seem to explain how a sentence like “It is wrong to kick cats” embeds into these complex contexts. What, that is to say, does the sentence contribute to the complex whole? One thing is clear enough: someone uttering any of the four example sentences above could not be said to be expressing a negative emotion toward kicking cats. So something more must be said. Stevenson himself never seems to have taken this problem to heart, so he never said much of anything by way of reply. But some later non-descriptivists have said more (see below). The embedding problem may appear to be a kind of technicality, and perhaps it is, though many philosophers have taken it very seriously. The final criticism I will mention seems to cut deeper into the spirit of emotivism. Stevenson said, and emotivism gains much of its plausibility from this idea, that a person who sincerely asserts or believes a moral judgment must necessarily feel some sort of emotional tug, so that whoever judges something good must be emotively in favor it, and whoever judges something bad must be against it or inclined to avoid, or would try to eliminate it. But we may wonder whether this claim is true. Isn’t it possible to judge sincerely that something is good, but feel no sympathy or other “pro-attitude” toward it whatsoever? There is no uncontroversial answer. Some find it obvious that such a thing is possible, while others are at least at first inclined to wonder what the questioner could possibly have in mind. But if we tell a background story it starts to seem very plausible that Stevenson may have overstated the connection between moral judgment and emotion. Surely it is imaginable that someone could be a self-avowed and sincere amoralist. Such a person would have no interest at all in moral values or rules, and might even be perfectly forthright in admitting so. Yet amoralists could surely learn to recognize which things are good and bad, even if the normal concern with such things might seem quaint or misguided to them. So they could with perfect sincerity and understanding manage to judge that giving to charity is morally good, or that breaking promises is bad, and they could make those judgments without any emotion or JAMES DREIER 178 motivation or tendency to promote the “good” things or discourage the “bad” ones. All of this seems possible. Doesn’t it show that emotivism is a mistaken theory? Perhaps not. In the paragraph above, the words “good” and “bad” are in quotation marks. It is plausible that amoralists use these and other moral words in what R. M. Hare called the “inverted commas sense,” really mentioning them rather than using them. Amoralists cannot say (sincerely, at least) that charity is good, so instead they say that charity is what most folk call “good.” We (moralists or amoralists) can certainly mention emotive words without expressing their emotive meanings. Some related theories Hare’s theory is not emotivist, though it is a close ally. According to Hare, the main function served by moral judgments is prescription (see HARE). So Hare agrees with Stevenson that we do not fundamentally describe things when we call them good or bad, and he even agrees that moral judgments could be called “expressions of emotion,” since prescriptions are expressions, in a sense. But Hare cautions against taking Stevenson’s idea too literally. In The Language of Morals, he writes: We speak of expressing statements, opinions, beliefs, mathematical relations, and so on; and if it is in one of these senses that the word is used, the theory, though it tells us little, is harmless enough. But unfortunately it is also used in ways which are unlike these; and Ayer’s use (in speaking of moral judgements) of the word “evince” as its rough synonym was dangerous. Artists and composers and poets are said to express their own and our feelings; oaths are said to express anger; and dancing on the table may express joy. Thus to say that imperatives [or moral judgments] express wishes may lead the unwary to suppose that what happens when we use one, is this: we have welling up inside us a kind of longing, to which, when the pressure gets too great for us to bear, we give vent by saying an imperative [or moral] sentence. Such an interpretation, when applied to such sentences as “Supply and fit to door mortise dead latch and plastic knob furniture”, is unplausible. In the 1980s and 1990s Simon Blackburn and Allan Gibbard developed versions of emotivism (or in Gibbard’s more general terminology, “expressivism”) grounded in the same root ideas as Stevenson’s theory. These theories are more sophisticated in various ways (in particular they make good headway into the embedding problem mentioned above), and they have to some extent supplanted Stevenson’s emotivism, though as inheritors, not as refuters. Bibliography Works by Stevenson 1944: Ethics and Language, New Haven, CT: Yale University Press. 1963: Facts and Values: Studies in Ethical Analysis, New Haven, CT: Yale University Press. (A collection of papers. Essay Two is especially useful as an introduction. Essay Eleven contains the most mature version of the theory.) Works by other authors Ayer, A. J. (1936) Language, Truth and Logic, New York: Oxford University Press. (Chapter 6 contains the germ of emotivism, which Stevenson developed into a full and sophisticated theory.) C. L. STEVENSON 179 Blackburn, S. (1984) Spreading the Word, Oxford: Oxford University Press. (Chapter 6 especially develops a variation on emotivism designed to address prominent objections.) Blanshard, B. (1949) “The New Subjectivism in Ethics,” Philosophy and Phenomenological Research 9, pp. 504–11. Geach, P. T. (1960) “Ascriptivism,” Philosophical Review 69, pp. 221–5. (An influential, but somewhat technical, objection to emotivism.) Gibbard, A. (1990) Wise Choices, Apt Feelings, Cambridge, MA: Harvard University Press. Goldman, A. and Kim, J. (eds.) (1978) Values and Morals: Essays in Honor of William Frankena, Charles Stevenson, and Richard Brandt, Dordrecht and Boston: Reidel. (A collection of critical essays; those on Stevenson’s work are quite accessible. Contains a comprehensive bibliography of Stevenson’s writing.) Hare, R. M. (1952) The Language of Morals, Oxford: Oxford University Press. (A work contemporary with Stevenson’s, with important similarities and contrasts.) Searle, J. (1969) Speech Acts, Cambridge: Cambridge University Press. JAMES DREIER 180 181 15