God's Call: Moral Realism, God's Commands, and Human Autonomy. By John E. Hare. Grand Rapids, Mich.: Wm. B. Eerdmans Publishing Co., 2001. Pp. x + 122. $14.00 (cloth). ISBN 0-8028-3903-7.

In God's Call John E. Hare is seeking "an account of God's authority in human morality" (vii). The three chapters -- one on each of the topics mentioned in the subtitle -- range widely over a variety of philosophical topics, both systematic and historical. I shall begin with a careful look at his first chapter, which is the longest and most complex, and then comment more briefly on the second and third chapters.

Hare understands moral realism as "the view that moral properties such as moral goodness are real" and moral expressivism as "the view that moral judgments are . . . orectic" -- that is, that their role is to express some act or disposition that belongs under the general heading of orexis, which "cover[s] the whole family of emotion, desire, and will" (viii). Hare's approach in the first chapter is to set forth the twentieth-century debate between realists and expressivists as a gradual convergence between the two positions, beginning with a statement of each view in its most unqualified form and proceeding through a series of "concessions" on each side. Each concession brings the two sides closer together, until in the end we are left with an intermediate position that Hare calls "prescriptive realism," which is meant to preserve the insights of both expressivism and realism.

Now the first thing to be noted is that as Hare has defined realism and expressivism, they are not contradictories. They do not even belong to the same domain of questions. Moral realism, as he defines it, is a position in moral ontology; to deny it is to be, not an expressivist, but an anti-realist. It is less clear how we are to understand expressivism. It may be a claim about the meaning or use of moral judgments. In that case, it is a position in moral semantics; to deny it is to be, not a realist, but a descriptivist. On the other hand, it may be a claim about the causal function of moral judgments in guiding action. In that case, it is a position in moral psychology; to deny it is to be, not a realist, but a fool, since only a fool would deny that the making of moral judgments plays a role in guiding action. And there is a third possibility: that it is a claim about the way in which orexis permits or impedes access to the objects of moral judgment. In that case, it is a position in moral epistemology; to deny it is to be, not a realist, but a rationalist.

So the whole framework of the discussion is vitiated from the outset by Hare's failure to keep distinct kinds of question separate. And, inevitably, the story of the particular "concessions" that each side is said to have made to the other involves further confusion between questions of different types. One such confusion can be seen in Hare's discussion of "the first realist concession": Iris Murdoch's concession that human beings are by nature selfish. Hare writes, "It is a concession to subjectivity in that she recognizes and indeed stresses that accurate moral perception needs obedience, a selfless attention, a pure heart; but our moral thinking is in fact usually characterized by the opposite of these -- by the root inclination to favor ourselves unjustly" (11). But it is no more a concession to subjectivity about morals to say that our moral perception might be obscured by perverse desire than it is a concession to subjectivity about astronomy to say that our perception of the moon might be obscured by clouds. The ontological question is one thing; the epistemological question is quite another.

Similar confusions bedevil Hare's discussion of the other concessions made in the twentieth-century debate. Now since prescriptive realism is presented as, roughly, the position that one is left with after all the concessions have been made, Hare's failure to preserve the distinction between different sorts of questions means that it is difficult to see clearly what prescriptive realism is supposed to be -- to see, that is, which questions it is intended to answer, and what that answer is. The best hope of getting clear on the matter comes on pp. 46-48, where Hare lays out prescriptive realism using an example he had already set forth in his introduction. Suppose Peter judges that his marriage is worth saving and that God wants him to try to save it. According to Hare, Peter's judgment involves at least the following aspects:

(1) It involves claiming "a Kantian kind of objectivity. . . . He is judging that people like him should respond to this kind of situation in this kind of way" (47).

(2) It also involves claiming "objectivity in a different sense, claiming that he is responding to a pull by the relationship that is really there outside his present imperfect attempts at evaluation" (47).

(3) In making the judgment Peter also "endorses the feeling of pull," thereby "endorsing not just his feeling on this particular occasion, but the whole set of norms that prescribe this kind of response" (47).

Take (1) first. It is clearly not part of the content of Peter's judgment that other people ought to behave in a certain way; his judgment concerns himself, not others. Even if R. M. Hare were right that the requirement of universalizability falls right out of the logic of moral language, it would still not follow that the making of a moral judgment about one person's obligation in any way involves the making of a moral judgment about anyone else's obligation -- even though the truth of any judgment of the first sort would always entail the truth of some judgment of the second sort.

Perhaps the best way to interpret (2) is as saying that the marriage really is worthwhile whether Peter feels it to be worthwhile or not. Now if Mackie is right, the commonsense understanding of evaluative language is that such language purports to represent features of the world that are independent of our representations of them. If Peter shares this commonsense understanding, his judgment that his marriage is worth saving is to be analyzed as a claim about a feature his marriage has independently of his judgment that it is worth saving. Interpreted in this way, (2) does seem plausible, if unremarkable.

(3), by contrast, is not plausible at all. I am not sure I understand what it means to endorse a pull, but it must involve some intentional state directed at the pull; and Peter's judgment about his marriage is not such a state. But I am more concerned about the claim that his judgment involves endorsing "the whole set of norms that prescribe this kind of response" (47). To judge "I ought to try to save my marriage" is not to endorse any system of norms. One may well not be thinking about more general norms at all. There need not even be some determinate set of norms such that one would not (occurrently) make the particular judgment unless one acknowledged (dispositionally) precisely that set of norms. The first-order judgment about the value of preserving one's marriage is one thing; a second-order judgment about the validity of the general norms that require the first-order judgment is something else altogether.

The pity here is that almost nothing of what Hare says in his first chapter is necessary for his view about the relationship between morality and the divine will, which he lays out in the second chapter. The view is strongly influenced by the moral philosophy of John Duns Scotus. According to Scotus as Hare understands him, the ultimate end of human beings is a loving union with God. But God is not constrained to choose any particular route by which we are to reach that end. The route God in fact chose is the moral law, which is only one of a great number of possible routes to our end. Thus, "God's call to us does prescribe a reality independent of our prescribing, namely the divinely chosen route to our final end. By saying that the reality is independent of our prescribing it, I mean that the route is there whether we try to follow it or not. . . . But saying that the route has independent reality is not saying that it has necessity" (84). Rather, what makes that route the right one is simply the fact that God chose it.

In the third chapter Hare takes up a challenge to this Scotistic view. The challenge comes from Kant -- or rather, Hare says, it comes from a standard but mistaken reading of Kant. Kant's argument is that morality cannot be derived from "a divine and supremely perfect will" (102). For we cannot intuit God's perfection, so we must derive our notion of God's perfection either from our moral concepts or from our non-moral concepts. If we derive it from our moral concepts, our account of morality as stemming from God's perfect will turns out to be crudely circular; but if we derive it from our non-moral concepts, our account of morality will be "drawn from such characteristics as lust for glory and domination and bound up with frightful ideas of power and vengefulness" (102) and thus will not be an account of morality at all.

This argument, Hare says, "is always taken as an argument from autonomy against the idea that we can think of God as the source of our moral obligation" (89). Hare addresses the argument in two ways. First, he offers a interpretation of autonomy that takes into account "those passages found throughout Kant's writings where he describes God as the head of the kingdom of which we are mere members, and where he says we should recognize our duties as God's commands to us" (94). On Hare's interpretation, our role is that of "recapitulating in our own wills the declaration in God's will of our duties. This is how we are lawgivers; we declare a correspondence of our wills with the law (which we do not create). For me to will the law autonomously is to declare it my law" (96).

Second, Hare argues that Kant's argument can be understood properly only if we see that Kant was not arguing against divine command theory generally, but only against the particular version of it defended by Crusius. Hare seems to regard the argument as decisive against Crusius (105-8), though unsatisfactory if taken as an argument against divine command theory generally (114-5). But it seems to me that if Kant really had Crusius in his sights, he was guilty of some of the most spectacular ignoratio elenchi ever perpetrated in the history of philosophy. First, Crusius does not think we intuit the divine perfection. We know it either inferentially through natural theology or testimonially through Scriptural revelation. And yet it does not follow that his account of morality is either crudely circular or else contrary to morality. Crusius holds that what is good for us depends, not on God, but on our nature. What is morally good or obligatory is that we pursue perfection, not for its own sake, or even out of love for God, but "out of obedience to [God] as our creator, sustainer, and lawfully commanding superior and lawgiver" (Guide to Rational Living §173). God's perfection is in part a matter of his commanding (and necessarily commanding) what is -- independently of his will -- good for his creatures. It is not a matter of God's doing what is morally obligatory, since only dependent beings are subject to obligation. Thus Crusius's account of morality is not circular. But neither is it blatantly contrary to morality. Indeed, Crusius warns that "We should guard against the mistaken belief that divine punishments and rewards are necessary so that the law may be obligatory, in that fear of the former and hope of the latter would drive man to obedience. . . . For through this . . . all true obedience would be destroyed" (§194). So whatever merit there may be in Kant's argument, it has no force at all as a reply to Crusius.

The great virtue of this book is that Hare consistently tries "to do philosophy through its history" (x). This is not the work of a contemporary analytic philosopher who occasionally decorates his arguments with a historical reference; nor is it the work of a historian of ideas who occasionally analyzes an argument. On the contrary, Hare works out his most engaging philosophical insights through his historical investigations, and he develops his most provocative readings of the history of philosophy by attending carefully to philosophical analysis. Although I have raised some objections to the specifics, there is no denying that Hare's approach produces distinctive and interesting results that repay close examination.