This is a chapter from Nathan Nobis’s Ph.D. dissertation Truth in Ethics and Epistemology: A Defense of Normative Realism.


CHAPTER 4:  Hare’s Universal Rational Prescriptivism

4.1. Introduction.

For over half a century, R.M. Hare developed a non-cognitivist position on the nature of ethics called “universal prescriptivism.”[1] Although Hare’s position was developed in many books and articles over his long career, my focus will be on his most recent work, his final defenses of universal prescriptivism.[2]

The position and the arguments in its favor are complex due, in part, to the fact that universal prescriptivism is both a meta-ethic, i.e., a theory of the semantics, metaphysics, epistemology, logic and psychology of moral judgments, and a normative ethic, i.e., a theory about what morally ought to be done.[3] Thus, it’s a view about what we are saying when we say that some action ought to be done, and a view about what we ought to do. Unlike many ethicists, Hare thought that answers to these two kinds of questions are highly interdependent. While I think Hare was right in thinking that these kinds of questions are so related,[4] I will argue that Hare’s answers to both kinds of questions are mistaken. Thus, I argue that the arguments in favor of universal prescriptivism are unsound.

My first focus will be Hare’s meta-ethic. He argues that universal prescriptivism provides a better account of the nature of moral judgments, compared to naturalist, intuitionist and emotivist meta-ethical theories. These theories attempt to explain what it is we are doing (or trying to do) when we make moral judgments: are we trying to state truths? If so, what kind of truths? Or are we (merely) expressing our emotions? Or are we doing something else?

At the time, philosophers who answered these questions saw them as intimately related to the question of the meanings of moral terms: different views on the meaning of moral terms yielded different understandings of what we are doing (or trying to do) when we use them. Hare argues that, although all these theories provide some true insights into the nature of the meaning of moral terms, universal prescriptivism is the superior theory. His argument for prescriptivism can be seen as, in part, an argument from the elimination of the other theories of the meanings of moral terms.

While, compared to ethics, there has been less concern about the meanings of moral terms and what we are doing (or trying to do) when we make epistemic judgments, comparable questions can be asked. I argue that Hare’s objections to naturalist, intuitionist and emotivist meta-ethical theories suggest analogous objections to naturalist and intuitionist and emotivist meta-epistemological theories. I argue that since some of these implications for the meanings of epistemic terms are false, Hare’s arguments against them are not sound. However, since parallel arguments to these were given against realistic meta-ethical theories, this shows that some of these arguments are not sound also. Thus, I defend some versions of moral realism from Hare’s objections. As a moral and epistemic realist, I have no interest in defending emotivism of either kind, so my discussion of emotivisms will be limited to what is required to undermine Hare’s arguments.

            After I defend moral realism by way of objecting to Hare-like parallel arguments against epistemic realism, I consider the possibility that this defense fails and Hare-like arguments against epistemological naturalist, intuitionist and emotivism are sound. Suppose Hare showed that the meanings of epistemic terms are not that what an epistemological naturalist, intuitionist or even emotivist says they are. If that is the case, then, if epistemic terms are meaningful, we might be left with something like an epistemological universal prescriptivism. I briefly attempt to develop what such a view would look like, and then argue that is unreasonable to accept: we have better reason to think the meanings of epistemic terms are not what epistemic prescriptivism implies they are. But if it this is so, then the arguments in its favor (which are parallel to those given for ethical prescriptivism) are weak and so at least some of Hare’s objections to ethical naturalism, intuitionism and/or emotivism are weak. Again, moral realism is defended by way of defending a kind of epistemological or intellectual realism.

Next I turn to Hare’s normative ethics. The main concern of what we might call Hare’s normative ethics was when a moral judgment, or a prescription, is “rationally acceptable.”[5] Hare’s focus was moral “oughts.” Here I observe that Hare’s position on what we morally ought to do, and what we should believe about what we morally ought to do, and the considerations he gives in its favor, seems to essentially depend on claims about non-moral oughts, in particular epistemic oughts and other intellectually evaluative language. This observation is similar to one of Sturgeon’s observation about Hare’s position on reasons for action:

There is something anomalous about a position like Hare’s . . noncognitivist about moral judgments but cognitivist and naturalist about reasons for action.  For talk of reasons or rationality displays the same features—difficult controversy, plus a typical “endorsing” role for the terminology—that are supposed to support noncognitivism about morality; and it is hard to suppose that judgments about reasons for action (or for feeling) are so different in this respect from judgments about reasons for belief.[6]

I aim to exploit this anomaly to undermine Hare’s position. The meaning of these non-moral oughts and evaluations needs to be understood: I argue that if they are understood in a manner comparable to how Hare understands moral oughts, then his overall position on the nature of morality is ultimately groundless: there is no reason why anyone should accept it. However, if Hare understands these oughts in a more robust, realistic manner, then there is no reason why moral oughts cannot, and should not, be understood in the same manner. So either Hare’s position undermines its own epistemic justification, or it has room to understand moral values as a moral realist does, in which universal prescriptivism is false. Yet again, moral realism is defended.

4.2. Ethical and Epistemological Theory

Hare took himself to be developing an “ethical theory.” What did Hare mean by that term? He explains: 

The expression ‘ethical theory’ covers attempts to say what we are asking when we ask moral questions. What do we mean by the words or the sentences that we use in moral discourse: what is the nature of the moral concepts or of morality?  If successful, these attempts will have implications for another, epistemological, question which also belongs to ethical theory: how should we set about answering our moral questions rationally?  Or can there be no rational way – is it just a matter of how we feel or what the current mores dictate?  On the other hand, if there can be rational discussion of moral questions, does it demand that there be a truth about them, or a set of facts, that can be discovered?[7]

To provide partial and quick answers to these questions, Hare thought that the meanings of moral words, sentences, and concepts is such that there can be, and is, a “rational” way to answer moral questions: their answers are not just a matter of how we feel or what current customs dictate. However, although there is a rational way to answer moral questions, he denied there are moral truths or facts to be discovered. So, to consider a possible example, if the question is whether any abortions are morally impermissible, Hare’s answer might have been that at least some abortions are morally impermissible (and that this is answer is, or can be, “rational”), but that it’s not, strictly speaking, true that some abortions are morally impermissible or a fact that some abortions are morally impermissible. This provides some negative understanding of his position (i.e., in terms of what it denies); a more positive understanding of how he understood judgment will be provided below. 

            If there are ethical theories, there are then, in the same sense, epistemological theories. Following Hare, we would think of the goals of epistemological theory, in this sense, in something like the following manner:

The expression ‘epistemological theory’ covers attempts to say what we are asking when we ask epistemic questions. What do we mean by the words or the sentences that we use in epistemic discourse: what is the nature of the epistemic concepts?  If successful, these attempts will have implications for another, epistemological, question which also belongs to epistemological theory: how should we set about answering our epistemic questions “rationally”?  Or can there be no “rational” way – is it just a matter of what we happen to believe, or what most people believe?  On the other hand, if there can be rational discussion of epistemic questions, does it demand that there be a truth about them, or a set of facts, that can be discovered?

A possible set of answers to these questions would include claims that the meanings of epistemic words, sentences, and concepts are such that there can be, and is, a “rational” way to answer epistemic questions: their answers are not just a matter of how we feel or what we currently believe. And, although there is a “rational” way to answer epistemic questions (say, e.g., about whether some belief is rational, or justified, or known, or some inference ought to be accepted, and so on), there are no epistemic truths or facts to be discovered. That is a possible view, and understanding and evaluating it (as well as its moral analogue) would require a fuller understanding of what is meant by “rational.”

While many of these epistemic concerns in the above passage are familiar to epistemological inquiry, I should mention that some of the concerns about ethical theory do not obviously transfer to epistemological theory, in the analogous sense. First, while concerns about the nature of “morality” are common, it’s not obvious what the analogous concerns are: they are certainly not the nature of “epistemology,” since that is a philosophical discipline. Perhaps they are concerns about the nature of knowledge or justification, or what we ought to believe, or the nature of correct reasoning, or the nature of evidence. Perhaps the analogous concern would be the nature of all these concepts and all the claims we make using these concepts. Either of these options would make “epistemic theory,” in this sense, quite similar to more common conceptions of epistemological theorizing, or those conceptions that many epistemologists would offer.

A second difference here, related to the epistemic upshot of a successful understanding of epistemic concepts, is that epistemic epistemology is a field of inquiry about itself whereas moral epistemology is an epistemic inquiry about another field or topic, viz. morality. Despite this difference, however, we can still ask epistemic questions about epistemic concepts and epistemic evaluations, e.g., whether we know that we know, and justifiably believe that we justifiably believe. Insofar as, on common views, epistemic judgments are of propositions with epistemic properties or predicates as constituents, and propositional knowledge is possible (and actual) there seems no deep reason why such epistemological knowledge or justified belief is not possible (and actual, if there is any knowledge or justified belief in the first place). 

Thus, although there are differences between ethical and epistemological theory in Hare’s sense, comparable questions can be asked. Since this is so, comparable answers might be given, although their plausibility might differ. And some comparable answers about the meanings of moral and epistemic terms have been given. Let us consider some of these answers and Hare’s objections to them. Hare only asked questions about ethical theory, but questions very similar to his can be asked about epistemological theory.

4.2.1. Against Ethical and Epistemological Naturalisms

In this section I consider Hare’s objections to ethical naturalism and intuitionism, the two kinds of what he calls “descriptivist” ethical theories. On these theories, the meanings of ethical terms are the same as the meanings of descriptive terms, terms for which their meanings are determined, as Hare says, entirely by the “syntax and truth conditions” of the descriptive expression.[8] The notion of a descriptive expression will be clarified below.

I argue that his arguments against this kind of ethical view suggest analogous arguments against epistemological naturalisms and intuitionisms. Thus, if Hare’s arguments show that ethical descriptivisms are false, then they might show that epistemological descriptivisms are false also. And if epistemological descriptivisms are false, then epistemic terms, if they have meaning, have a meaning beyond their truth conditions. If this is so, that points us in the direction of an epistemic non-descriptivism, either an epistemic emotivism or an epistemic prescriptivism. Whether this implication should be resisted will be a focus of discussion later in the chapter. 

First, let us consider naturalistic descriptivisms; later we will consider intuitionistic or non-naturalistic descriptivisms. In both ethics and epistemology, philosophers have argued that the meanings of ethical and epistemic terms are the same as, and exhausted by, the meanings of some naturalistic, in principle empirically-verifiable expression. These views are naturalisms. So, to say that an action is morally required is to say that it has some natural property (e.g., maximizing overall pleasure of all affected by the action), and to say some belief is justified or epistemically is to also to say that it has some natural property (e.g., being caused by a belief forming process that typically produces true belief) and there is no other aspect of their meaning: the terms’ meanings are “wholly determined by [the] truth-conditions” for the naturalistic expression.[9] These theories are a kind of what Hare calls descriptivist theories: on them, to use a moral term is to describe something, i.e., to say that is has some specified property, in this case a natural one.

Some ethicists, especially those from the early twentieth century, said that claims that relate a moral (or epistemic) property and a naturalistic one are analytic and so reasonably believed a priori: e.g., some might have said that reflection on the meaning of “morally right” reveals a naturalistic analysis. Other philosophers, particularly contemporary ones, who accept naturalistic analyses of ethical and/or epistemological terms, have claimed instead that these are synthetic claims that are reasonably believed not merely by reflecting on their meanings; they might argue that their (alleged) truth is seen by way of some argument to the best explanation, or inductive generalization, or some manner beyond merely reflecting on the meaning of the terms. Hare argues against both these kinds of naturalism in ethics; he doesn’t seem to find the latter kinds of naturalism to be much of an improvement over the former.  He writes:

I have put the distinction between naturalism and intuitionism in terms of the different kinds of truth-conditions they impose on moral statements. My distinction is therefore broad enough to cover both the old-style ‘refutation of naturalism’ due to Moore . . , and the new-style naturalism whose chief habitat is Cornell.  . . [T]he way I have put the distinction will make it apply to both the old and the new naturalisms.[10]

Thus, Hare does not seem to care whether a naturalist claims that her theory is a theory about the meaning of moral terms or whether it is a theory that posits an equivalence between moral and naturalistic terms (or properties), and yet is not a statement of meaning.  For the sake of argument, since Hare believes his arguments are not affected by this distinction, I will assume he is right.

One of Hare’s arguments[11] against naturalist descriptivisms is that if they are true, then many cases of moral disagreement are, in fact, not moral disagreements; since this is a false implication, this shows the theories to be mistaken.[12] To illustrate this kind of argument, suppose person A and person B accept different naturalistic analyses or definitions of “morally required”: each accepts a different meaning of the moral terms and what its naturalistic truth-conditions are. Person A believes that “morally required” means “produces the most pleasure of all alternatives” and person B believes that “morally required” means “will contribute to the stability of society.” Suppose A and B are discussing the morality of a token act of abortion. If A says this act was morally obligatory and B says that this act was not morally obligatory, then it would appear that they disagree about whether this act was morally obligatory.

However, as Hare understands naturalism, it falsely implies that they do not disagree. This is because, if naturalism is true, according to Hare, then A said that this act produced the most pleasure of all alternatives and B said that this act contributed to the stability of society. If that is what was said, however, then they are not disagreeing: they are simply talking past each other about the morality of abortion, and they even might both be saying something that’s true. Or their conversation might lead them to a disagreement about the meaning of the term “morally obligatory.”  However, since they are disagreeing about the morality of this abortion, and this abortion is not both morally obligatory and not morally obligatory, this shows that naturalism is a false view about the meanings of moral terms. Since they are disagreeing, they are each saying more than, out of one’s mouth, that the abortion produces the most pleasure and, from the other’s, that this abortion will not contribute to social stability. 

Hare illustrates this idea in the following using an example of two cultures who accept different naturalistic analyses of “wrong”:

. . if we follow the naturalists, we shall have to say that the sense of the word in the two cultures are entirely different.  This will have the consequence that they are not contradicting one another; for fighting might be wrong in the sense of the word used in one culture, but not wrong in the sense of the word use by the other.  The people in each culture will be right in their own sense of the word “wrong.”[13] 

However, if each party to the dispute is right, then they are not disagreeing. But they are disagreeing so Hare concludes that naturalisms are false: the meaning of a moral expression is, at least, not solely that of the meaning of a naturalistic express (since if it were, then these wouldn’t be cases of disagreement), and so to evaluate something morally is not merely to say that it is has some natural property. Hare might allow that attributing a natural property is involved in making a moral evaluation, but that it is not the entire truth, as naturalism says it is. 

One might respond to this kind of reasoning by claiming that, in the dispute between A and B, at most only one has the correct definition of “morally obligatory” since it does not mean both “produces the most pleasure of all alternatives” and “will contribute to the stability of society”: only one of those definitions could be right, at most.[14] On Hare’s understanding of naturalism and “meaning,” however, this seems to be false or irrelevant since they both fully mean what they say, respectively, even though their views on the meanings of moral terms are inconsistent.

Assume, however, that there are correct understandings of the meanings of moral terms, that B’s definition is correct and that this abortion will not contribute to the stability of society. If so, then this abortion is, apparently, not morally obligatory and so B has said something that’s true. And, A, the utilitarian, apparently has said something that is false because she accepts a false naturalistic equivalence between “morally obligatory” and “maximizes happiness”: she he the wrong view on what the moral term means. She just doesn’t understand what ‘morally obligatory’ means and that’s why her moral evaluation is mistaken. Hare’s reaction to this case, which is founded on the assumption that there are objective meanings to words, seems to be the following:

What has happened in this case is that a substantial moral principle, that one ought to do what will contribute to the stability of society, has got promoted into an analytic truth, true in virtue of the meaning of ‘ought.’  But it is not an analytic truth. If it were, then the [pro-abortionist] would win the argument because the [anti-abortionist] would be contradicting themselves, by saying something which the very meaning of ‘ought’ establishes as false.[15]

Hare’s thought seems to be that the meaning of a term is determined by how people use it, but that no substantive moral conclusions follow from this. Everyone should be able to agree that, when some people use the term “right,” they mean “maximizes utility,” and when others use the term “right,” they mean “contributes to social stability.” That’s just a descriptive truth about linguistic behavior and people’s conceptual backgrounds. Hare seems to be thinking, however, that none of these facts entail that rightness is any of these natural properties, and so none of the (apparently) disputing parties in the cases above can say to the others, “You are mistaken in your moral evaluation because you do not accept my definition of the moral terms.”

So he apparently understands naturalism as a view about what people mean by various terms, allowing that different people can accept different meaning and there’s no question for which claim about meaning is correct.  Something like that is needed to avoid a response in the cases above to the effect of, “Person B simply does not understand the meaning of ‘morally obligatory’ so what he says is false. But person A does understand the meaning and is saying something that is true, so at least they are not both saying something that’s true.” If meaning is understood on an individualistic basis, then the “talking past each other” objection holds and shows that there’s more to the meaning of moral terms than the meanings of natural terms. If this kind of objection does not show this, then objections below will do so in a clearer manner.

            Before we consider these further objections, we might observe that analogous epistemological naturalisms have the same results regarding disagreement. Consider one naturalist, person C, who understands “justified” as “produced by a reliable belief-forming process.” Consider another hypothetical naturalist, person D, who understands “justified” as “highly probable relative to the believer’s observation.” Let us grant that both these latter expressions are naturalistic ones; they could, in principle, be empirically verified. Suppose that they are discussing whether some token believer is justified in accepting some specified proposition. C says that he is, and D says that he is not. On an epistemological naturalism analogous to the moral naturalisms that Hare discusses (and thinks he refutes), C has said that a reliable process produced this belief and D has said that this belief is not probable, relative to this believer’s observations. These claims might both be true: let’s suppose that they are. If so, however, then they are not disagreeing: they are merely talking past each other. But since they are disagreeing about whether this belief is justified, this shows that epistemological naturalism, in this sense, is false. The meanings of epistemic expressions are not solely the meanings of naturalistic expressions, and to evaluate something epistemically is not merely to say that it is has some natural property. Perhaps attributing a natural property is involved in making a moral evaluation, but that it is not the entire truth, as this kind of naturalism says it is.[16] 

            A number of other considerations suggest that moral and epistemic terms’ meanings are not equivalent to merely naturalistic expressions. First, coming to believe that some action or belief has some natural property does not seem to “close” the question of whether it ought to be done, or ought to be believed. Frank Jackson expresses this concern:

. .  the open question argument has equal putative force in both contexts.  Just as no amount of information couched in purely descriptive terms seems to close the question of what is morally right to do, so no amount of information couched in purely descriptive terms seems to close the question of what it is rational to believe or do. It always seems to make sense to ask, even after all the descriptive information is in, what one ought to do [or believe], independently of whether the ought is given a moral or [epistemically] rational slant.[17]

This suggests that the meanings of, on the one hand, naturalistic expressions and, on the other, moral and epistemic expressions are not the same. If they were, then the question of whether some belief with some specified naturalistic feature(s) are justified, or have some other epistemic status, would be “closed,” in virtue of the very meanings of the terms.  But these questions are not “closed”, i.e., it can make sense for competent speakers to ask them about any naturalistic property that a moral or epistemic property might be identified with, so these kinds of naturalisms are false.

            A final concern that some have used to argue against naturalisms involves the motivational factors that they find strongly connected to evaluative judgment.[18] To many it seems that if one judges that one is morally required to do some action, then one must have some motivation towards doing that action. This suggestion has some plausibility, since, intuitively, it would be odd for someone to judge that he really ought to do something, but feel no “pull” at all towards doing that action. Similarly, we might think that it would be odd for someone to judge that some belief (which she currently does not accept) is rational or justified for her to have, but feel no pull or desire towards believing that proposition.

These phenomena suggest, however, that naturalisms are mistaken. This is because, according to many critics of naturalism, describing something in naturalist terms does not even seem to have this related motivational impact: to say that something has some natural property is, apart from background desires, motivationally inert. This seems true with moral naturalisms, and it also seems true with epistemological naturalisms. To think that a belief is produced by a reliable process, or is in conformity with the design plan for some module of the mind, or is highly probable, or has any other naturalistic feature does not have the motivational impact that thinking it has some epistemic property does. To the former, naturalistic observations, an indifferent, “So what?” response is unremarkable; for the latter, epistemic observations, an indifferent, “So what?” response would be remarkable. Motivational indifference to natural properties is understandable, but motivational indifference to moral and epistemic properties is puzzling. This suggests that the two kinds of terms don’t mean the same thing and so that naturalisms are false.

            Thus, Hare provides some good reasons to think that naturalistic ethical theories are mistaken, in terms of capturing what “we mean by the words or the sentences that we use” in moral discourse and the nature of these concepts. These good reasons, however, seem to apply to epistemological theories also: the meanings of epistemic terms are not identical to the meanings of naturalistic terms. Thus, if any descriptivism is correct, it would have to be a non-naturalistic, intuitionist descriptivism.

Before I move on to these theories and Hare’s arguments against them, I want to note that Hare gives an argument against naturalistic descriptivisms that is peculiar and unconvincing. (He gives a similar argument against intuitionist descriptivisms also, and I will present this below, after I present his strong arguments against intuitionism.) He claims that naturalism “collapses” into moral relativism, that moral relativism is false, and so naturalism is false also.[19] He seems to think that if any kind of naturalism is true, then “what will determine the truth or falsity or moral judgments will be the particular truth conditions accepted in a given society as determining the moral words”[20] or “the linguistic usage of native speakers of the language.”[21] This, of course, would make moral truth relative to, i.e., metaphysically determined by, community acceptances or linguistic practices. And this might imply that what’s right in one society would be wrong in another, merely because of different linguistic usage and what moral concepts are accepted. Hare thinks this result is false and that even descriptivists would agree.

Hare seems to assume, however, that on any naturalism the majority’s view on the truth conditions of a term determines the truth conditions of the term, or somehow makes the analysis true by definition and thereby unquestionable. While some naturalisms might accept that, it’s not clear that all naturalisms do or would have to.  Furthermore, since it seems that some naturalists could easily think that the majority of society fails to understand what a moral term means or has a false understanding of what it is for something to be right (obviously, many utilitarians think that), Hare’s criticism is misplaced: it might refute some naturalisms, but not all kinds of naturalisms that reject this assumption. Since this argument against moral naturalism is pointless, and there are so many other arguments to try to show that moral (and epistemic) terms do not mean natural terms, I will not consider analogous versions of it against possible epistemological naturalisms.

4.2.2. Against Ethical and Epistemological Intuitionisms

Hare argues that ethical non-naturalism, or intuitionism, is mistaken: it provides a false view of the meaning of moral terms. Hare describes intuitionism as “the view that [the truth conditions of a moral statement] are the possession [by actions, people, etc.] of specifically moral, sui generis properties which cannot be defined without introducing some moral terms into the definiens.”[22] So, one possible intuitionism might say that the truth conditions for “rightness” are determined by the truth conditions for “goodness,” and that the truth conditions for something being good (or the meaning of “good”) can only be stated in other moral terms, e.g., perhaps being valuable, being admirable for its own sake, being such that it ought to exist, and so on. These latter moral terms, intuitionists claim, either cannot be defined, or they can be defined only in other moral terms. 

There are analogously non-naturalistic, or intuitionist, epistemological views. On these views, the truth-conditions or meanings of epistemological terms likewise cannot be defined without introducing some epistemic terms into the definiens.  So, on a possible non-naturalistic view, the meaning of “justified,” or its truth conditions, is the truth conditions of “there is sufficient evidence for,” which might be further explicated in terms of “ought to be believed,” which, in term, might be either claimed to be not definable or definable only in further epistemic terms. These explications might lead to a web of interrelated epistemic concepts, none of which can be defined, or can reasonably be defined, in purely naturalistic terms. 

Hare’s objection to intuitionism in ethics is that he thinks it implies an objectionable sort of moral relativism.  He thinks intuitionism provides a relativistic-resulting understanding of how we tell that something is good that can lead to logically incompatible, yet equally truthful implications, which is impossible. He explains:

Intuitionism is the view that the truth conditions of moral judgments, which give them their meaning, consist in conformity with the data on which we have to base our moral reasoning, and with which its conclusions have to square; and these data are the common moral convictions that all morally educated people have. Since these convictions will vary from one society to another, the effect of intuitionism is, again, to anchor our moral reasoning to something relative to particular societies. True, there are convictions which are common to most societies; but there are others which are not, and no way is given by intuitionists of telling which are the authoritative data.[23]

He thinks that to support their moral judgments, “intuitionists are appealing to nothing objective, but only to their own and other people’s thoughts, and these will vary from one person and society to another,”[24] thus resulting in an unacceptable relativism. He basically thinks that intuitionists’ view about the meanings, or truth conditions, of moral terms gives rise to a false moral epistemology that shows that the view of meaning is mistaken. If the meaning of “morally good” cannot be identified with the meaning of any naturalistic expression, then to identify that some state or condition (e.g., being happy) is morally good is, on some intuitionisms, just to carefully consider that state and just “see” that it is good. Moral terms’ meanings are such that well educated people can just see when something is right or good: that’s how they know or reasonable believe that something is good. But different people and societies sometimes just “see” things differently and intuitionism provides no way to adjudicate these competing visions other than by appealing to further intuitions. To add to the problems arising from moral conflict, Hare claims that intuitionists claim moral infallibility: he thinks they claim that, “the mere occurrence of the experience [i.e., a moral intuition] guarantees the truth of the moral statement.”[25] However, it is doubtful that an intuitionist has to think that, that they should think that (since they surely realize that intuitions can conflict, which casts strong doubt on infallibility) and that many intuitionists even did (or do) accept this kind of infallibility.

To return to the initial questions of “ethical theory,” asked above, Hare thus thinks that intuitionism implies that there is no way to answer our moral questions “rationally” since (on Hare’s interpretation of ethical intuitionism) what’s right and good is “just a matter of how we feel or what the current mores dictate,” and so nothing rational. Since Hare thinks moral questions can be answered rationally, and thinks this irrationalism follows from the intuitionistic theory of the meaning and truth conditions of moral terms (i.e., he seems to think that accepting epistemological skepticism about moral judgments is inconsistent with accepting the intuitionistic theory of meaning or truth conditions), he rejects intuitionism.

Ethical intuitionists would likely resist this argument, in part by arguing that Hare oversimplifies and misrepresents their positions. For now, however, for the sake of argument I will concede that it is a good argument against intuitionism in ethics, or at least against the kind of position that Hare calls intuitionism.  However, if it is a good argument, then the comparable argument can easily be made against the analogous epistemological intuitionism. How can we tell that a belief is justified, or that there is sufficient evidence for some claim, or that some belief is rational, or that some inference is warranted, or that some theory ought to be accepted, or some explanation is the one we should believe, or that some belief is, in fact, known?  A kind of epistemological intuitionist would argue that, given the meanings of these epistemic terms, we just look, reflect and see whether something – typically, some belief and believer-in-context – has any of these properties: intuition tells us so. Intuition, or what might also be called rational insight, seems to be the best way to describe the phenomena of coming to see that some belief has some epistemic quality. 

A Hare-like response to epistemological intuitionism is to observe that since intuitions about what’s reasonable, justified and known sometimes vary from one society to another, and among individuals, the effect of epistemological intuitionism is, again, to anchor our epistemic judgments to something relative to particular societies and individuals. While there are epistemic convictions that are common to most people and societies, there are others that are not, and no way is given by intuitionists of telling which the authoritative data are.[26] Some might respond to this by observing that different people might evaluate beliefs differently because they have different evidence and that all these evaluations might be correct because of the differing evidence.

To avoid the full effect of this response, for one, we can observe that some epistemological intuitions can vary even among people with very similar backgrounds and so, we might think, have similar evidence. Second, we can focus our attention to cases where we might think the evidence is about the same for all, e.g., philosophical examples where what it sought is a verdict on whether some belief would be justified or known, or a judgment on what the person should believe, and so on: here intuitions about what epistemic evaluation is correct would vary also. A defender of epistemological intuitions, a critic of the Hare-like response above, would have to argue that there are no cases of competing epistemological intuitions. It is unlikely that this argument would succeed; the arguments might imply some kind of false epistemological infallibilism, and they might also falsely imply that evidence is essentially “private” so we cannot see and evaluate others’ evidence. So it seems that Hare-like doubts about epistemic intuitions are as strong as his doubts about ethical intuitions.  

Although this gets complicated since it turns epistemic theory upon itself, one might think that epistemic intuitionism implies that (to return to the initial questions of “epistemological theory,” asked above) there is no way to answer our epistemic questions “rationally” since (on this epistemological intuitionism that is analogous to Hare’s understanding of ethical intuitionism) something’s epistemic status is just a matter of how we feel (or how things seem to us) or what the current epistemic standards or popular opinion dictate. If this kind of view should be rejected, and it follows from the kind of intuitionistic theory of the meaning and truth conditions of epistemic terms that we are discussing here, then this view about the meaning and truth conditions of epistemic terms should rejected. Since I am assuming, for the sake of argument, that the analogous argument against ethical intuitionism showed ethical intuitionism to be mistaken, I will assume that this parallel argument shows epistemological intuitionism to be mistaken. Epistemological intuitionism yields a false epistemology, which shows it to be a false view about the meaning and truth conditions of epistemic terms.

Some of the arguments against naturalism in ethics and epistemology suggest similar arguments against intuitionisms. First, two intuitionists who appear to be engaged in moral debate might actually be talking past each other if each accepts a different version of intuitionism and so is attributing (or denying) a different sui generis moral or epistemic property. Since we might suppose that these intuitionists would really be disagreeing, this would show, at least, that there’s more to making a moral evaluation than attributing a non-natural property. Second, if moral and epistemic judgments have an intrinsically motivational quality about them, then there is more to the meaning of a moral term than whatever meaning sui generis, non-natural terms have, if these terms stand only for non-natural properties. On standard views, properties (or apprehension of them) do not have any motivational influence on their own, so if moral and epistemic judgment necessarily involves motivational factors, then there is always something going on more than attributing properties.

Thus, the considerations Hare gives against naturalistic and intuitionistic ethical descriptivisms suggest comparable arguments against naturalistic and intuitionistic epistemological descriptivisms. Since we have tentatively found the arguments against epistemological descriptivisms to be as strong as the arguments against ethical descriptivisms (which we have presumed to be strong, in part to see how Hare reasons to his universal prescriptivism), we are left only with the option of some kind of epistemological non-descriptivism, if we wish to acknowledge that epistemic terms are meaningful in some manner. Some might take this as a reductio against Hare’s arguments: if arguments parallel to those he gives against ethical descriptivism take us to epistemological non-descriptivism, then something has gone wrong. While this response might be attractive, I hope to bolster it with positive concerns against epistemological non-descriptivism, which I develop below, as well as the considerations against epistemological emotivism – another kind of non-descriptivism – that I developed in the previous chapter.

In rejecting ethical descriptivisms, both naturalisms and intuitionisms, Hare concluded that “there is a further element in [moral terms’] meaning, the prescriptive or evaluative. . which is not so determined [by truth conditions], but expresses prescriptions or evaluations or attitudes which we assent to without being constrained by truth-conditions.”[27] My concern now is whether a position on the meaning of epistemic terms that incorporates such a feature is plausible. I will try to develop such a position and argue that it is not plausible. If epistemological non-descriptivism is not plausible, this suggests that there is something wrong with the arguments above against epistemic descriptivisms. But if there is something wrong with Hare’s arguments against epistemic descriptivisms, this suggests that there is something wrong with Hare’s arguments against ethical descriptivisms, since the arguments are parallel.

I will argue that Hare’s arguments against both ethical and epistemological intuitionism are weak. The acceptability of a kind of ethical intuitionism is further bolstered by the fact that Hare’s position presupposes a kind of epistemological intuitionism, and its defense depends on it. Since epistemological intuitionism (a version of epistemic realism) is an acceptable view (independently, and even according to Hare, apparently), ethical intuitionism (a version of moral realism) would seem to be acceptable also. The arguments against it were not as strong as they might have seemed and so Hare does not provide good reason to reject moral realism.

4.2.3. Against Ethical and Epistemological Emotivisms.

Before we consider Hare’s preferred version of ethical non-descriptivism, we should briefly consider his arguments against emotivism: our treatment can be brief since I have no interest in defending emotivism of any kind. Furthermore, I developed and criticized versions of epistemic emotivism in my previous chapter.

According to Hare, emotivists believe that moral judgments are “the expressions of irrational or at least non-rational attitudes of approval or disapproval” and that “the only questions one can reason about are factual ones.”[28] He claims emotivism implies then there can be no rational argument about fundamental moral questions: people have their emotional responses and no responses are better or worse than any others, none are more or less rational than others, and there is no way to argue that someone’s responses are mistaken (because they cannot be mistaken). Emotivists like Ayer and Stevenson agreed, but since Hare thinks that there can be rational arguments about ethics, in virtue of the meanings of moral terms, Hare thinks emotivism is false: he calls it an “irrationalist” doctrine.[29] This assessment seems correct. If to judge something to be wrong or right is just to express one’s positive emotions about it, as the “Boo! Hooray!” caricature of emotivism has it, then it is not at all clear how one might reason about fundamental moral matters. 

            In my previous chapter, I argued that the considerations Ayer and Stevenson give in favor of their emotivisms suggest epistemic emotivism also. I argued that we have better reason to reject such a view than accept it, once we reflect critically on what seems true about epistemic evaluations. I argued that this too is an irrationalist doctrine, since it implies that judgments about what’s justified, known and reasonable are only expressions of emotion. That this is an irrationalist theory becomes is especially clear when we apply it to our understanding of the nature of argumentation. If epistemological emotivism is true, then to judge that some claims provide a good reason for another claim, or that some set of premises justifies a position (which is not merely to say that the position is a logical consequence of the premises), is just to express one’s emotions in favor of having some belief, and to disagree with these assessments is just to express contrary emotions. Jonathan Kvanvig offers a related observation:

[A]rguments and explanations presuppose the truth of epistemic norms, and if the norms themselves are given nonalethic status, then the explanations and arguments are simply defective in virtue of the fact that their presuppositions are not true. Good arguments presuppose logical and epistemic norms.[30]

In sum, if epistemological emotivism is true, then we are in fundamental error about the nature of reasoning, since there are no truths about how we ought to reasons: there are merely different emotional preferences that people have about accepting and rejecting various complex claims involving logical relations, none of which are better or worse than any other. This is a consequence that is hard to accept. If Hare accepts it, it will surely make a difference for how we should understand his insistence that we should be consistent in our judgments, as we shall see.

Fortunately, however, there is little antecedent reason to accept epistemological emotivism, since there is little to recommend the premises that might be given in its defense (and so there is little to recommend the premises given in favor of ethical emotivism). Fortunately, also, however, if epistemological emotivism were true, then there would also be little or no reason to accept it. If the doctrine were true, then to say that “epistemological emotivism is the most reasonable view,” or one “we ought to accept,” would be for that someone to merely express his or her positive emotions about our believing epistemological emotivism, or to say that to try to cause us to accept the view. But it wouldn’t be true that, in response to such expressions, we should change our views, so there would be no problem in ignoring such expressions.  Thus, epistemological emotivism generates many puzzling implications, none of which have any force beyond people’s emotional expressions. In my previous chapter, I presented a number of reasons to think that epistemological emotivism is false, all of which are applicable here also.

Since these irrationalist implications follow from premises given in defense of ethical emotivism, this provides further reason to reject the view, and Hare might have agreed. Hare had more arguments against emotivism, but since I have no interest in defending emotivism, I will move on to his preferred version of non-descriptivism.

4.3. “Rational” Universal Prescriptivism?

Thus far, I have argued that Hare’s arguments against ethical descriptivisms suggest analogous arguments against epistemological descriptivisms, and that if the former are sound, then the latter are sound.  Someone might respond that since arguments against epistemological descriptivism are not sound, the arguments against ethical descriptivism are not sound either. For example, someone might think that a kind of epistemological intuitionism is true: there are some propositions which, when examined carefully, one can just “see” that they are justified and that they ought to be believed, just like (some might argue) that there are some states of affairs that that one can just “see” are morally good.

Although Hare thought otherwise, an advocate of this kind of epistemological and ethical intuitionism could be a fallibilist; she can, and should, admit that she might be mistaken in her intuitions: she can be open to the possibility that she has missed something and isn’t seeing things as they are. If this intuitionist perspective is a viable one about epistemic evaluations, then it would seem to be a viable one about ethical evaluations. If that is so, then Hare has not refuted ethical intuitionism.

I will attempt to develop this response in light of critical discussion of Hare’s universal prescriptivism. I will argue that this position and the considerations Hare gives in its favor support this response. To do this, I must explain the main feature of universal prescriptivism.[31] The view is not as easy to present because it is not always ideally clear whether Hare intends to state, on the one hand, how moral terms are used and how they can be used and, on the other hand, how they should be used. I will argue that if Hare has claims about how they should be used, these claims ultimately undermine his ethical non-descriptivism.

However, if the position has no claims about how we should use moral language, then his position is a merely descriptive “logic” of the relations between moral terms and so has no implications for which moral evaluations we ought to accept, or how we ought to reason in light of our understanding of these logical relations between moral terms. If that’s true, then universal prescriptivism does not appear to be an ethical theory in the sense Hare claims it is. Furthermore, Hare does not seem to take himself as a mere “moral logician”: he does not seem to take himself to be merely observing various logical relations among moral evaluations, but never saying what judgments people ought to accept. Either way, universal prescriptivism is so doubtful that it should not be seen as a viable alternative to some version of moral realism.

One of Hare’s main concerns is showing that moral thinking can be a “rational” enterprise. Exactly what he means by “rational” is never entirely clear (and so I must work with an intuitive, undefined notion of the term, although I think it can often be linked to a distinctly epistemic sense of the notion), but he thinks moral thinking can be rational, i.e., we can reason about the acceptability of various moral evaluations and principles because of the meanings of moral terms.

Hare thought that the meanings of moral terms are such that they are “governed by logic.”[32] This seems to be a descriptive claim; it is a claim that there can be logical relations between moral judgments: moral judgments might logically imply other moral judgments in conjunction with other moral judgments and non-moral premises. Hare thinks, as we have seen, that moral judgments are not descriptions; he thinks they are closely related to commands, or are a kind of command. And commands can have logical implications. Consider this set of sentences:

(1)   Take all the boxes into the house.

(2)   This is a box.

(3)   Take this box into the house.

Sentence (3) is, in some sense, a logical consequence of (1) and (2), even though (1) is a command and so is neither true nor false. The claim that this box should not go into the house would be, in some quite plausible sense of the term, inconsistent with (1): they can’t be jointly performed or satisfied.

            As mentioned above, Hare thinks that moral judgments, which he calls “prescriptions,” are like commands, i.e., that they are commands with special features that not all commands have (these features will be explained below). Consider the moral judgment “We ought not to hurt babies for fun.” Hare would understand this judgment as something like the following:[33]

(1)   No hurting babies for fun, please.

Consider this empirical, descriptive claim:

            (2) Maggie is a baby.

And from (1) and (2) it follows, logically, that

            (3) No hurting Maggie for fun, please.

Sentence (3) is, in some sense, a logical consequence of (1) and (2), even though (1) a command is neither true nor false. The claim that Baby Maggie ought to be hurt for fun is, in some sense, inconsistent with (1).

For those who would claim that, since a “premise” in these “arguments” is neither true nor false, the “argument” cannot be “valid” (or that it isn’t even an argument, strictly speaking), Hare would likely suggest that this person needs to expand their notion of argument and validity beyond the common merely truth-functional notions. The common definition needs to be amended to reflect the fact that there are valid logical relations among prescriptions.

            So Hare’s basic claim is that there are logical relations among moral terms, and that set of sentences containing a moral prescription can be logically inconsistent or inconsistent. He writes:

. . if (as is certainly the case) there can be logical inconsistency between contradictory prescriptions, someone who wants the totality of the imperatives, or in general prescriptions, that he (or she) accepts to be self-consistent will have to observe the rules which govern consistency.[34] 

An important thing to note about this claim, however, is that it is only the observation that for a set of prescriptions to be consistent, these prescriptions have to conform to the logical rules that determine consistency. That’s just to say that someone who wants consistency will need to have consistency to get consistency. These are uncontroversial claims about the logical nature of prescriptions.  However, they are not claims about prescribers, or the people who make prescriptions. Should they want consistency in their imperatives, or ought their imperatives be consistent, or should they believe that their imperatives should be consistent? These questions need to be explained, but both affirmative and negative answers are not helpful for Hare. I develop this response in a number of ways below, in response to a variety of Hare’s claims about consistency.

Negative answers to the question of whether our imperatives should be consistent suggest that, ultimately, all Hare has done is observed the natural fact there are logical relations among imperatives, but has not claimed that recognition of these relations should influence our behavior or which prescriptions we should accept. If that is all Hare has done, however, his theory has no implications for which moral evaluations we should accept. If that is so, then Hare (and others) have been very mistaken about the reach and power of his theory. His theory is only this: “here’s what’s consistent with what, but I have nothing to say about what you should do.” This wasn’t Hare’s theory.

If observation of these natural, logical features should influence our behavior, however, that might vindicate naturalism in ethics and epistemology because that would suggest that natural features can be identified with an evaluative characteristic. If this is true, this undercuts Hare’s main arguments against ethical naturalism and, by extension, epistemological naturalism.

And if people should want consistency in their imperatives, i.e., their imperatives should be consistent, Hare has a dilemma on his hands: if this “should” is a moral “should,” then, on Hare’s view, it’s an imperative. We can ask whether we, morally, should accept it. This question might take us down a regress of further moral “shoulds,” i.e., moral imperatives, which are based on only further imperatives. This would be a puzzling consequence, and might also lead to the kind of relativism that Hare (and others) likely reject. The reason for this relativism is that imperatives are not an objective feature of the world: rather, they originate in people and people make different imperatives. If moral judgments were based in imperatives “all the way down” and different imperatives can be accepted all the way down, the incompatible prescriptions might be equally justified. This is hard to understand, and Hare’s view is designed to avoid this result. So it is doubtful that moral imperatives depend on further moral imperatives “all the way down,” so to speak.

If, however, people should want consistency in their imperatives, i.e., their imperatives should be consistent, and this is not a moral should, then we can ask what kind of should it is. Hare seems to describe it as a “should” of “rationality,” albeit of an unspecified kind. We can then ask if these kinds of “shoulds” are imperatives/commands, or if they are ever literally true. If they are merely imperatives or commands, then “rationality” is founded solely on commands. Again, this is a counterintuitive and puzzling consequence, which might lead to a kind of relativism about rationality, which Hare seems to reject.

If, however, if it is literally true, as a matter of “rationality,” that we should have consistent beliefs and imperatives, then there are going to be facts or properties that make propositions that express these truths true. But if there are such facts and properties, then, some kind of “rational” or epistemological realism is true and the arguments against versions of it, developed earlier, are not sound. But if some kind of epistemological or rational naturalism or intuitionism is true, it is hard to see why an ethical naturalism or intuitionism could not also be true. These morally realistic views claim that there are moral facts or properties. Hare states, like Stevenson and Hume, however, that “as soon as we start asking what it is for a moral quality or fact to exist in the world, we get lost” (p. 451). But if we “get lost” looking for moral qualities and facts, we get equally lost looking for the epistemic, intellectual, or “rational” facts that Hare’s position seems to appeal to in generating practical implications. But if these kinds of facts or qualities can be found, then it would seem that moral qualities can be found also.

Hare says something that might be seen as a response to the kind of objection I am raising to his position. He says:

[T]here do not have to be moral facts in the world in order for us to develop a theory of moral reasoning any more than there have to be logical facts to substantiate logical reasoning. The necessities which constrain our reasoning are formal necessities.[35]

Logical claims are made true by formal necessities. However, claims (if they are claims) to the effect of our beliefs ought to conform to these formal necessities, or we are unjustified in accepting claims that do not conform to them are not formal necessities. If claims like these pertaining to belief are true, they are not true in virtue of form or syntax. If they are true, however, then there are aspects of the stance-independent world that make them true. And if there are such aspects of the world to make claims like these true, then it would seem that there could also be aspects of the world that can make it the case that some moral judgments are literally true. If that is so, then Hare’s arguments against descriptivism are unsound.

            Related to the claim that moral prescriptions can be in logical relations is Hare’s claim that moral judgments are “universalizable.” Hare explains this notion in the following manners:

A statement is universalizable if and only if it commits anyone who assents to it, on pain of inconsistency, to accepting that there is a universal principle which holds, and which applies equally to any situation exactly similar in its universal properties. . . [Universalizability] has the consequence that one cannot without self-contradiction say that someone ought to do something in a situation, but that a precisely similar person ought not to do it.[36]

Again, we see in this statement a descriptive claim about the logical relations that hold between prescriptions and the “universal principles” that hold between them. However, Hare seems to assume that someone’s assenting to inconsistent prescriptions would be that is bad and ought to be avoided: perhaps it is for this reason that he describes inconsistency as a “pain,” not a merely neutral state. However, if inconsistency is bad and ought to be avoided, and this is literally true or a fact, then that might suggest that all the arguments against there being moral truths or facts were not unsound after all.

            Let us consider another statement of universalizability:

One cannot with logical consistency, where a and b are two individuals, say that a ought, in a certain situation specified in universal terms without reference to individuals, to act in a certain way, also specified in universal terms, but that b ought not to act in a similarly specified situation. This is because in any ‘ought’ statement there is implicit a principle which says that the statement applies to all precisely similar situations. This means that if I say, ‘That is what ought to be done; but there could be a situation exactly like this one in its non-moral properties, but in which the corresponding person, who was exactly like the person who out to do it in this situation, ought not to do it’ I contradict myself. This would become even clearer if I specified my reasons for saying why it ought to be done: ‘It ought to be done because it was a promise, and there were no conflicting duties.[37] 

Should we not contradict ourselves? If not, then there’s no problem doing so: Hare has made the novel observation that moral commands can contradict and be consistent, but it is not true that consistency should be sought, or that consistence is better than inconsistency.  If we should avoid contradicting ourselves, and we should strive for consistency, and these claims are literally true, i.e., they do not consistent in someone’s urging “All commands being consistent, please!” then it would seem that there can be room for moral judgments being straightforwardly true also. 

            Let us consider one final statement of Hare’s, to better understand the view:

These purely logical features of moral statements have the consequence that if we accept a set of universal moral principles, and accept that the facts of the situation are as they are, then we cannot avoid assenting, at the bottom end of the structure to a set of singular prescriptions for actions. To refuse to assent to these would be to involve ourselves in self-contradiction.  . . [W]e are constrained to assent to the prescriptive conclusions if we assent to the premises. If we do not want to assent to the conclusions, we have to alter the premises; and if we do not want to do that, we have to put up with the conclusion.[38]

Again, we can ask whether Hare is merely making a descriptive point about the logical relations among moral concepts, or if he is saying that it is better for one’s prescriptions to be in one logical condition (e.g., consistency) over another (e.g., inconsistency, especially recognized as such). If he is merely doing the former, then his position has no substantive implications for how we ought to reason about moral matters.  But if he’s doing the latter, and making a “rational” evaluation, and this is not itself a neither-true-nor-false prescription, then, again, we might wonder why moral evaluations couldn’t be either true or false also, and so either ethical naturalism or intuitionism is true.

            What I have tried to argue is that, if Hare’s position is true, then there are truths or facts about rationality. This notion of rationality has been left undefined (since Hare does not define it), but it seems to be related to epistemic or intellectual evaluations. And I have suggested if epistemic or intellectual evaluations can be literally true on Hare’s position, then that would make room for moral evaluations being literally true. That would, of course, imply that Hare’s position is false.  I have also argued that if Hare’s position does not allow for epistemic or intellectual truths or facts, then it is an irrationalist position: if that is so, then it should not be accepted either.

            I have not yet considered, however, what an epistemic prescriptivism would look like. Alan Millar and Jose Luis Bermudez suggest a characterization of such a view.  They explain:

On this account if, for example, you condemn someone’s reasoning as irrational you are saying, in effect, ‘Let me not reason that way if in his shoes.’ Two people could agree on the naturalistic properties possessed by a thinker but disagree on whether that thinker is rational. What would they be disagreeing about is how to reason in that thinker’s shoes. But the disagreement may not be one which can be resolved by appeal to norms which are treated as objectively authoritative by everyone.[39]

This is a possible view that is somewhat similar to Hare’s prescriptivism since it yields epistemic imperatives. It is hard to see how it could be conclusively refuted, especially in what some would describe as a non-question begging manner, as the theory might call into question some of the most basic assumptions about reasoning. The final claim of the passage is unclear: perhaps everyone would not treat some epistemic or intellectual norms as objectively authoritative: perhaps there are no norms that everyone would treat as objectively authoritative. But those are empirical claims about people’s acceptances. 

The important question is whether there are epistemic or intellectual norms that are objectively authoritative for everyone. The common view is that there are; and if there are not, then some kind of intellectual or epistemic relativism or nihilism would be true. Elsewhere in this work I discuss these kinds of views and reject them, in part because they are epistemically self-undermining: if they are true, then it’s not (objectively) true that they are reasonable to hold. I also presented reasons to regard them as false: there are many epistemic phenomena for which they do not easily jibe with and the realist picture more readily accommodates.

However, since there is little positive reason to accept the kind of Hare-like epistemic view that Millar and Bermudez suggest (but do not defend), then it is reasonable for most people, probably, to continue believing that there are epistemic norms that are objectively authoritative for everyone. However, if the semantics, metaphysics and epistemology of these claims can be adequately secured, then comparable semantics, metaphysics and epistemology for moral claims can likely be adequately secured also.

4.4. Normative Ethics and Normative Epistemology.

In my introduction I said I would discuss Hare’s normative ethic, i.e., his theory on which moral judgments we ought to accept and what we morally ought to do. This is useful and necessary since there are normative epistemologies, or theories on which beliefs ought to be accepted.  Hare’s theory about what we ought to do stemmed from his view about universalizability. Basically and very, very roughly, he thought that like cases should be treated as like cases and so, given the “logic” of moral terms, we should make the moral evaluations we would be willing to make, were we in the shoes of anyone affected by our actions.  In effect, it’s a kind of “golden rule” ethics. Hare states it simply:

What maxims we can adopt, or what moral judgments we can accept, will then depend on what we are prepared to prescribe for all like situations (whether, for example, were we the unfaithful husband or his deserted wife.[40]

This is not entirely clear since Hare does not seem to want to be making only a descriptive claim about what maxims we “can” adopt, but rather which maxims we should adopt. The idea is that if one can accept a principle and its results from all affected perspectives, then it is “rationally” acceptable, and if, and only if, it is rationally acceptable should one accept it. Hare got into trouble with the “can” in this claim because it seems possible that someone could accept principles that would cause him or herself severe harm: e.g., perhaps a committed Nazi would be willing to accept principles that would require him to be executed even if he turned out to be a Jew. Hare’s theory might not have adequate resources to say why this is a bad principle; if so, some would regard this as a refutation of the view. This might lead us to reconsider morally realistic positions, and re-assess Hare’s arguments against them. 

My objections to this kind of “golden rule” ethics is not that it is false, but that, first, if it is true and, second, we intellectual ought to believe it, then it is hard to see why there couldn’t be true moral evaluations also. If we ought to judge like cases as like cases and this is true – it’s not merely a prescription or an expression of emotion – then it is hard to see why there couldn’t be true moral evaluations also. If Hare’s theory is epistemically justified for somebody or rationally ought to be believed by somebody, then there are facts of this kind also. But, as I argued earlier, if we recognize evaluative facts of this kind, then it’s hard to see why there could not be facts of a distinctively moral kind as well.

If, on the other hand, it’s not true, or neither true nor false ought that one to be consistent, then Hare’s position might be ultimately groundless. If no evaluative “oughts” or “shoulds” of any kind are true, the implications of this would be astounding. I’ve argued, however, that we have good reasons to think that some evaluative judgments are true. I think that these often simple reasons are far stronger than the complex ones Hare gives in favor of his universal prescriptivism. Thus, while the evaluative world might be prescriptions “all the way down” so to speak, we have better reason to deny this than accept this. And if is correct, then while we might find great epistemic value in golden-rule considerations and trying to see things from others’ points of view, the truth might well be these maneuvers can put us in a better position to see the independent moral facts.

4.5. Conclusion.

I have argued for all of the following: first, that Hare’s universal prescriptivism and the arguments in its favor presuppose epistemic or intellectual evaluations: Hare’s understanding of the nature of moral evaluations depends on an understanding of epistemic or intellectual evaluations. I have argued that Hare’s arguments against ethical descriptivism suggest analogous arguments against epistemic descriptivism. I have argued that understanding epistemic or intellectual evaluations in a non-descriptivist manner leads to rational unacceptable consequences; my discussion from the previous chapter on Ayer supports this also. Since epistemic evaluations are meaningful and rational, they therefore should be understood along descriptivist lines. But since this is so, then moral evaluations can be understood along descriptivist lines also. Thus, the arguments against ethical descriptivism are not as strong as Hare thinks: the various considerations he has offered against these theories can, and should be, resisted.



[1] Although Hare resisted being called a non- cognitivist and an anti-realist, and said that there were perfectly good senses in which moral judgments were true and known (see Hare [“Universal Prescriptivism” 451]), I will still refer to his position as non- cognitivist and anti-realist because it is clearly inconsistent with moral realism.

[2] For Hare’s early work, see Hare (The Language of Morals), Hare (Freedom and Reason) and Hare (Moral Thinking). His more recent work, which helpfully summarizes the main arguments of the early work, includes Hare (“Universal Prescriptivism”) and Hare (Sorting Out Ethics).

[3] My uses of the terms “judgment” and “evaluation” are intended to be neutral between cognitivist and non-cognitivist understandings of moral evaluations. 

[4] For arguments that one’s meta-ethic has consequences for one’s normative ethic, and vice-versa, see Sturgeon (“What Difference Does it Make”).

[5] See Hare (“Foundationalism and Coherentism in Ethics” 191).

[6] See Sturgeon (“Critical Study of Gibbard” 403).

[7] See Hare (“Universal Prescriptivism” 451).

[8] See Hare (Sorting Out Ethics 42).

[9] See Hare (“Universal Prescriptivism” 452).

[10] See Hare (Sorting Out Ethics 65).

[11] In his over fifty years of writings on these issues, Hare gives more arguments than I discuss here. My focus, therefore, is on his most recent arguments since we might think that he thought these were the strongest and most important.

[12] See Hare (Sorting Out Ethics 52-55, 58-60).

[13] See Hare (Sorting out Ethics 69).  Here is another of Hare’s examples of this point: “If one lot of people say that abortion is wrong and another lot says it is not wrong, they are not differing merely in their linguistic usage. They are expressing different moral opinions.  This shows very clearly what is wrong with naturalism.  What is wrong is that it pretends that what are in fact substantial moral principles are nothing more than linguistic rules. Naturalism confuses learning morals with learning a language.  But the two are very different. If I have grown up thinking that abortion is wrong, I have acquired more than a mere linguistic skill.  I have acquired a moral principle, an attitude to abortion.” See Hare (Sorting out Ethics 69).

[14] My discussion below is intended to address the worry that the above arguments presuppose (falsely, the objector claims) that if naturalism is true, then, perhaps, “right” means “maximizes pleasure” and “right” means “is approved by all.” The objector claims that “right” cannot mean both these things.  However, as Hare seems to understand naturalism, it allows that, for person A, “right” can mean one thing and for another person, “right” can mean something incompatibly different. Hare’s objection is to naturalisms of this kind: the objection is that they do not allow for genuine disagreements between people who understand the meanings of moral terms differently.

[15] See Hare (“Universal Prescriptivism” 453).

[16] Hare allows for moral terms to have a naturalistic meaning and that these can vary – these are the non-moral considerations that a moral judgment depends on, or those to which we will appeal to in making or justifying a moral judgment – but that the “prescriptive” meaning (explained below) is universal for moral judgments. It’s this prescriptive meaning that enables people from different societies to disagree, even though they relate moral terms to different natural terms.

[17] See Frank Jackson (“Non-Cognitivism, Normativity, Belief” 100).

[18] I develop this concern in Chapter 5 in a discussion of Mackie’s views.

[19] See Hare (Sorting Out Ethics 65).

[20] See Hare (“Universal Prescriptivism” 453).

[21] See Hare (Sorting Out Ethics 66).

[22] See Hare (Sorting Out Ethics 82).

[23] See Hare (“Universal Prescriptivism” 454).

[24] See Hare (“Universal Prescriptivism” 454).

[25] See Hare (Sorting Out Ethics 94).

[26] For arguments from these considerations to the conclusion that epistemic intuitions are unreliable, see Weinberg, Nichols and Stich (“Normativity and Epistemic Intuitions”) and Nichols, Stich and Weinberg (“Meta-Skepticism: Meditations on Ethno-Epistemology”).

[27] See Hare (“Universal Prescriptivism” 452).

[28] See Hare (“Universal Prescriptivism” 455).

[29] See Hare (Sorting out Ethics 117).  Here Hare also mentions Gibbard’s views and states that, in his view, they are not “irrationalist” like Ayer and Stevenson’s are.

[30] See Kvanvig (The Value of Knowledge 176).

[31] Another feature of prescriptivism is its motivational internalism.  See Hare (“Universal Prescriptivism 457-458): “A speech act is prescriptive if to subscribe to it is to committed, on pain of being accused of insincerity, to doing the action specified in the speech act, or, if it requires someone else to do it, to willing that he do it.” Although this would establish, on standard assumptions about motivation, that moral evaluations are not (purely) beliefs, since my arguments below do not concern motivational issues, I will not address them any more.

[32] See Hare (Sorting Out Ethics 42).

[33] This locution is one concrete way to present what Hare thinks someone means when they make a moral judgment. This meaning can be presented with alternative formulations of the imperative, however.

[34] See Hare (“Universal Prescriptivism” 455)

[35] See Hare (Sorting Out Ethics 7).

[36] See Hare (“Foundationalism and Coherentism in Ethics” 192-193).

[37] See Hare (“Universal Prescriptivism” 456).

[38] See Hare (“Foundationalism and Coherentism in Ethics” 193).

[39] See Millar and Bermudez (Reason and Nature 9).

[40] See Hare (“Universal Prescriptivism” 460).