In Edward Regis, Jr., ed., Gewirth's Ethical Rationalism (Chicago, III.: University of Chicago Press, 1984).

Do Agents Have to be Moralists?

R. M. Hare

This paper will be almost entirely about just one section of Professor Gewirth's book [Reason and Morality, 1978], that called "Generic Rights and Right-Claims" (2.7, pp. 63-103). In this section Gewirth defends a crucial step in his argument, and one which certainly needs defense, since on the face of it it is a daring one. But before I start criticizing his book, I must pay tribute to one very great virtue it has: he does not, like the authors of many recent books on moral philosophy, make constant appeals to our moral intuitions, but seeks to found his argument solely on conceptual truths. This means that, if sound, the argument should be able to be shown to be so, and similarly if it is unsound. His book is therefore much more worthy of the attention of those whose aim in doing moral philosophy is to find a secure method of moral reasoning -- who are not content to rely on moral convictions, however firm or widely shared, which may only be the contingent effects of a particular writer's upbringing, and admit of no test but the appeal to more convictions of the same sort.

We must first put this section in the context of Gewirth's entire argument. He conveniently summarizes this on p. 171:

By way of brief summary: an agent is a person who initiates or controls his behavior through his unforced, informed choice with a view to achieving various purposes; since he wants to fulfill his purposes he regards his freedom and well-being, the necessary conditions of his pursuit of purposes, as necessary goods; hence he holds that he has rights to freedom and well-being; to avoid self-contradiction he must hold that he has these generic rights insofar as he is a prospective purposive agent; hence he must admit that all prospective purposive agents have the generic rights; hence he must acknowledge that he ought at least to refrain from interfering with his recipients' freedom and well-being, so that he ought to act in accord with their generic rights as well as his own.

The structure of the argument, then, is this: the concept of an agent is shown to involve an agent's having purposes; if he has purposes, he must treat as necessaryry goods the necessary conditions (freedom and well-being) for the pursuit of these purposes; hence he must claim rights to freedom and well-being; hence he must accord the same rights to all agents who have the same reasons (being prospective purposive agents) for claiming them as he has; hence he must acknowledge that he ought himself to respect these rights in others.

I shall be concentrating on the two steps in the middle: 'hence he must claim rights to freedom and well-being; hence he must accord the same rights to all agents. . .'. I shall argue that the first of these steps is only valid in a somewhat weak sense of "rights", if such a sense exists, in which a right-claim is not necessarily universalizable (i.e. in which we may claim a right without being thereby committed to extend it to others who have the same reasons for claiming it as we have). If there is no such sense, then the step is invalid. But even if there is such a sense, it is not that which is required in order to make the next step valid, viz. 'hence he must accord the same rights to all agents who have the same reasons for claiming them as he has'. In short, Gewirth is guilty of equivocation.

The first of these two steps gets its clearest explanation on page 80, which I shall take as epitomizing all Gewirth's longer explanations:

Suppose some agent were to deny or refuse to accept the judgment
(1) 'I have rights to freedom and well-being.'
Because of the equivalence between the generic rights and strict 'oughts,' this denial of (1) would entail the agent's denial of
(2) 'All other persons ought at least to refrain from interfering with my freedom and well-being.'
By denying (2), the agent would have to accept
(3) 'It is not the case that all other persons ought at least to refrain from interfering with my freedom and well-being.'
But how can any agent accept (3) and also accept
(4) 'My freedom and well-being are necessary goods'?
That he must accept (4) we saw above; for by virtue of regarding his purposes as good the agent must also a fortiori value his freedom and well-being as required for achinieving any of his purposes. Hence, insofar as he is a purposive agent, that is, an agent who wants to achieve the purposes for which he acts, he must want his freedom and well-being to be kept inviolate, so that they are not interfered with by other persons. He must want this, moreover, not as a mere favor from other persons but as setting a requirement for their noninterference that they are obligated to obey, such that from his own standpoint as a purposive agent, severe censure and even coercion are warranted if they violate the requirement. Hence, the agent must accept (2). Consequently, since (2) is logically equivalent to (1), the agent contradicts himself if he denies (1). He must therefore accept, on pain of contradiction, that he has the generic rights.

We may start by asking why Gewirth is so confident that an agent cannot accept (3) and also accept (4). By 'necessary goods' he means, it seems, 'goods which are necessary to the achievement of the agent's purposes'. Let us admit for the sake of argument that the agert must want his purpose to be achieved (at any rate want it ceteris paribus), and that if one wants the purpose to be achieved, one must want whatever is a necessary condition for its achievement. But does it follow that one must therefore think that other persons ought to supply these necessary conditions, by refraining from interfering? Can I not want something, without thinking that I ought to have it?

Gewirth makes a great deal of play with the difference between moral and prudential 'oughts, and says several times that the 'ought' in question here is a prudential one. It may be questioned whether it can be, as he phrases (3). For when I say that other persons prudentially ought to do something, I must mean that they ought to do it in their own interest; I cannot mean that they ought to do it in my interest. So the step from the perhaps true premise that the agent wants something to the conclusion that he prudentially ought to seek it (and whatever is a necessary condition for it) may be all right; but this does not entitle us to conclude that he must think that other people prudentially ought to give it him; for it would not be true that they ought, and there would be no reason why he should think so, if to give it him would not be (and were known by him not to be) in their interests.

It seems to follow from this that Gewirth cannot consistently take 'ought' in (3) prudentially. And indeed he often, as in this passage, takes it in what must be a moral sense. He says that the agent musr want noninterference "not as a mere favor from other persons but as setting a requirement for their noninterference that they are obligated to obey, such that from his own standpoint as a purposive agent, severe censure and even coercion are warranted if they violate the requirement". There are several expressions here which consort extremely ill with prudential 'ought'. Other people do not automatically have obligations to supply what the agent prudentially ought to seek; there is no requirement on them to supply it (the prudential requirements are all on him, not on them); and he is not warranted in censuring them, only himself, if he fails to get what he needs through his own (prudential) fault.

Not only is it almost unavoidable to take all these expressions in a moral sense here, but Gewirth needs to, if he is going to make good the next step in his argument. For what he is going to do is to show (as vvould be quite legitimate if it had been established by the present step that a purposive agent is committed to claiming moral rights) that he is also committed to extending these rights to other people. Moral rights, like moral 'ought's, are universalizable, in the sense that if one claims them for oneself, one is committed to extending them to other similar people in similar situations. There is indeed a sense in which prudential judgments are universalizable (I cannot say "I ought, in my own interest, but you ought not, in your own interest, although we and our situations are precisely similar"; for if we and our situations were precisely similar, your interest and the means to satisfying it would also be the same). But this sense does not help Gewirth's argument. For if all he had shown was that an agent must claim that there is a prudential requirement on him to seek the necessary conditions for achieving his purposes, the universalization of this claim would only yield the claim that there is a prudential requirement on other similar agents in similar situations to seek the necessary conditions for achieving their purposes. Since they and their situations are similar, the necessary conditions will also be similar. But what Gewirth needs to substantiate is the universalized claim that there is a requirement on anyone, including the original agent, to seek the necessary conditions for achieving the purposes of anyone else who is similar and similarly placed. The agent would be committed to this claim if it were true in the moral sense of 'ought' that an agent is committed to claiming that all other persons ought to supply the necessary conditions for achieving his purposes. But this has not been shown.

In general, it would appear, from his frequent use of such expressions as 'his prudential due' (72), 'must hold, from within his own prudential standpoint in purposive action, that he is entitled' (73), and the like, that he does not realize how big a gap there is between prudential and moral requirements. The latter are what he needs for the next step in his argument; "But the most he has established by the preceding step is the former. As he says himself,

In order to be a right-claim the demand must purport to be valid, justified, or legitimate; this is why it asserts that something is due the demander, that he is entitled to it, and not merely that he wants, values or insists on having it. . . . But since his ground is prudential, not moral, why should these other persons accept or credit his demands? (p. 72)

The expression 'from within his own prudential standpoint in purposive action' occurs frequently in different variants. Gewirth seems to think that, provided that the agent remains within this standpoint, he will have to claim that certain things are 'prudentially due' to him. This may be so; but it is not enough for the next step in the argument. In order to make this, the agent has to be got to claim that certain things are morally due to him, not just "prudentially due"; and moral claims, although they may be made by individuals, are not made from within their individual standpoints; that is what universalization is all about. They are made from a standpoint which, it is claimed, can be shared by anybody. Even if it is true that the agent, from his standpoint, has to claim certain things, he is not committed to claiming that anybody else, from his different standpoint, has to accept these claims. He could only claim this if the original claim was made from a standpoint which could be shared by anybody; but many individual purposes are such that, though those who have them must pursue them, they are at odds with the purposes of others; and those others may therefore, from their standpoints, reject his claims, just as he might reject similar claims of theirs.

Gewirth sometimes seeks to avoid these difficulties by laying stress upon the prescriptivity of the claims to rights which an agent must make. But this misses the point, which is about universalizability, not prescriptivity. We may readily admit that the agent has to prescribe that his purposes be fulfilled. Otherwise they would not be his purposes. He must therefore want them to be fulfilled. But the question is, Must he prescribe and want the similar purposes of others to be fulfilled in similar circumstances? If not, his prescription is not a universal one, and therefore not moral. And if it is not, then the next step becomes illegitimate. For the next step relies on an insistence that, if what are claimed are rights (i.e. that others ought or ought not to do certain things), then it must be allowed that the same would apply if the roles were reversed. But this will only be so if the prescription to which the agent is committed is a universal one. And this has not been shown. It has been shown that he must prescribe that his purposes be achieved, and that the necessary conditions for achieving them be supplied. But it has not been shown that he must prescribe this universally -- that is, prescribe that this be so whoever is in the roles in question.

It would thus be perfectly consistent for someone to admit that he was a purposive agent, and therefore bound to assent to singular prescriptions, and even to claim rights in the weak sense, if there is one, which is nonuniversalizable, but to refrain from prescribing universally, i.e. claiming rights in the strong, universalizable sense which compels him to accord them to others too. This would make him somewhat similar to the "amoralist" described in my books Moral Thinking (chap.10) Freedom and Reason (pp. 100f.).1 It would be nice if Gewirth's argument could show this position to be inconsistent; but its failure to do so is more troublesome for him than for me. Because of his pretensions to 'objectivism', which I shall be discussing in a moment, he really does need, having rightly abjured appeals to moral intuitions, to produce logical or conceptual arguments against the amoralist. I, by contrast, though I can offer only prudential arguments to persuade people not to become amoralists (arguments which require as premises substantial beliefs about the world as it contingently is), can take comfort from this apparent weakness in my position, because it at least preserves my bona fides as a nondescriptivist, while permitting me to do all that can be done, and all that is in practice needed, to counter amoralism. So Gewirth, a rationalist descriptivist, falls short of aims which he cannot avoid having, whereas I, a rationalist nondescriptivist (an animal that has no place in Gewirth's bestiary) can survive such a reverse.

It is therefore time to examine, by way of a postscript, why it is that Gewirth gets himself into this trouble. The clue is to be found in the early pages of the book, where, after quite rightly proposing to argue for a rational morality on conceptual grounds without appeal to moral intuitions, he slips too easily into a much more disputable claim that this involves finding a 'correspondence-correlate', i.e. determining "what facts, objects, or relations may serve as the basis of the truth-value of moral judgments" (5). He thinks that he has to adopt such an out-and-out correspondence theory of moral truth because his prejudices have not allowed him even to contemplate the possibility of a rationalist non-cognitivism or nondescriptivism. That this is the cast of his thought comes out in such phrases as 'noncognitivists and other antirationalists' (xi). Because he is unable to take seriously the idea that rationality in choice and therefore in prescription might consist in something different from rational fact-finding, he thinks that he has, in order to establish the rationality of moral thought, to discover in the world facts or objects to be the correlates of true moral judgments.

Even his own theory, though it suffers from the flaw in argument that I have been trying to expose, does not need to be stated in this questionable way as a correspondence theory. He sets out to establish on conceptual grounds alone a theory which makes some moral judgments rational. He also, as we have noticed, makes crucial use of the idea that by virtue of being agents we are committed to certain prescriptions (viz. that the necessary conditions be realized for achieving our purposes). All he needed to do was to show that these prescriptions, to which as agents we are committed, commit us in turn to certain moral prescriptions from which no agent who is consistent in his thinking can dissent. The notion of moral truth has no necessary part to play in this argument. So it was open to him to state his theory without reference to it, and entirely in terms of prescriptions.

It is, indeed, quite possible for prescriptivists like myself to find a place for the notion of the truth of certain moral judgments;2 but we do not need it in our argument. Gewirth thinks that he needs it; and this nay be why he has taken his short cut to the refutation of amoralism. Someone who believes in moral facts, but who rejects intuitionism (the theory that we directly discern them), will need to show by conceptual means that everybody who reasons consistently is committed to acknowledging these facts. It is not open to him to take the less ambitious course, which I take myself, of showing by conceptual means (including some conceptual means which are not available to descriptivists) that there are two options: the moralist option, which, once adopted, constrains us into a set of moral principles such as we are familiar with; and the amoralist option, which, though logically invulnerable, is not going to be adopted by any human being who thinks seriously and reflectively about his own powers and his own interests in the light of the available facts, which do not need to include any moral facts. This less ambitious program is not open to a descriptivist; he has to find a way of proving moral conclusions to amoralists, and therefore may be tempted too easily into thinking he has found a proof when he has not.


Notes

1 R. M. Hare, Moral Thinking: Its Levels, Method and Point (Oxford: Clarendon Press, 1981); Freedom and Reason (Oxford: Clarendon Press, 1963).

2 See my Lindley Lecture, "Some Confusions about Subjectivity," in Freedom and Morality, ed. J. Bricke (University of Kansas, 1976). In this lecture, however, I unwittingly traduced Professor Gewirth; see my correction in Philosophy 52 (1977), last footnote to my "Geach on Murder and Sodomy."