9

lnt J Phil Rel: 193-208 [1982] 0020-7047/82/0134-0193 $02.40. Martinus Nijhoff Publishers, The llague, Printed in the Netherlands.

QUINN ON DIVINE COMMANDS AND MORAL REQUIREMENTS

TIIOMAS B. TALBOTT Willamette Universit),

In a recent work of great theological as well as philosophical interest, 1 Professor Philip L. Quinn has sought to resurrect a divine colrlmand theory of moral obligation, or at least to defend such a theory as consistent and to defend it against sonle of the standard philosophical objections. But just what is a divine cornmand theory of moral obligation? According to Quinn, there are in fact "several distinct" theories, so in discussing such theories one must "specify which one is being attacked or defended" (p. 64); but common to all of them, he suggests, is the claim that "morality in some way depends, at least in part, upon the will of God" (Preface). The difference between theories, he seems to think, is a matter of how this dependence relation is understood, According to some theories, the most radical ones, God's connnanding something is the complete reason why it is morally required; but according to other (less radical) theories, it is rather a necessary part of the complete reason, the remainder of which might be that God is a wise creator, a loving father, etc. Quinn insists, however, that even the nlost radical form of the theory, even the theory that God's comlnanding something is the sole and complete reason why it is morally required, is consistent and Call in fact be defended against the traditional philosophical objections. My purpose in this paper is to examine Quinn's defense of the divine command theory and to show, if possible, why it is unsuccessful. 1 shall divide what follows into three sections. In Section I, 1 shall argue that Quinn has left obscure some essential features of the divine command theory; in particular, that he has left obscure the precise sense in which, according to divine colnmand theorists, morality is thought to depend upon the commands of God. In an effort to correct that defect, I shall try, in Section 11, to specify' an important necessary condition of any genuine divine comlnand theory; that is, of any theory according to which all moral obligations depend, either in part or in full, upon the commands of God. 1 shall then argue, in Section 111, that the divine command theory has implications that are clearly unacceptable and is therefore clearly false, that it could be true only if made utterly vacuous and trivial so that morality would no longer have the kind of importance traditionally attributed to it, and that Quinn's defense of the theory therefore fails.

194

Consider, to begin with, Professor Quinn's initial statement of the theory, what he calls "Tile Simple Theory," the kernal of which, he says, consists of three propositions: (T1 a) (TI b) (T1 c)

It is necessary that, for all p, it is required that p if and only if God COlnmands that p. It is neccssary that, for all p, it is permitted that p if and only if it is not the case that God commands that not-p. It is necessary that, for all p, it is forbidden that p if and only if God commands that not-p.

These three propositions, says Quinn, given certain further qualifications concerning the use of the term 'God,' imply that morality depends upon tile existence of God; for if God should not exist, nothing would then be forbidden and hence everything permitted. 2 But at this point a difficulty arises. As it stands (apart from any additional assumptions), (T1 a) implies only that 'it is required that p' and 'God commands that p' are, of necessity, true (or false) together; it does not imply that God's commanding p is the full reason, or even part of the reason, why p is morally required; and it certainly doesn't imply that Gods's commanding p actually brings it about that p is morally required. For suppose, to take the extreme case, that God is a necessary beiJlg and hence exists in all possible worlds, that loving kindncss, justice, and moral perfection are essential properties of God, and that God, being essentially olnniscient, necessarily knows what is morally required and necessarily commands what is morally required just because it is morally required. It then follows that (Tla) is true, in other words that 'it is rcquircd that p' and 'God commands that p' are, of necessity, true (or false) together, but not that what is morally required depends, in any unique or special way, upon the commands of God. Rather, given these assumptions, the fact that p is morally required will itself be God's reason for commanding that p. So if Quinn's "Simple Theory" is to express the idea that morality actually depends upon the commands of God, that something is morally required because God commands it and not vicc versa, some additio.aal assumptions are required as Quinn himself would no doubt acknowledge. But what further assumptions should a divine command theorist adopt at this point? One possibility, though perhaps an embarrassing one from the point of view of a divinc command theorist, is suggested by the claim, already mentioned, that (1) If God should not exist, then nothing would be morally required, nothing forbiddcn, and everything permitted. If this claim is to have any significance at all, or at least the kind of significance that a divine command theorist would attribute to it, one must presumably assume that God is n o t a necessary being and hence that

195 (A) There are possible worlds in wbicb God fails to exist. For suppose that God were a necessary being and hence could not have failed to exist; suppose, in other words, that the proposition: (2) God does not exist, is false in all possible worlds. Given this assumption, it of course follows that (2) entails every proposition whatsoever; for example, it entails not only: (3) Cruelty for its own sake is morally permissible, as the divine command theorist wants to insist, but also: (4) Cruelty for its own sake is morally forbidden, an implication the divine command theorist could hardly be expected to relish. So if the divinc command theorist wishes to endorse the clairn that (2) entails (3) but not the claim that (2) entails (4), he must adopt the assumplion that there are indeed possible worlds in which God fails to exist. Now this surely is an embarrassing result. A theory implying that the supreme being, and hence the supreme moral authority, is a mere contingent being can hardly be expected to have a very broad theological appeal. Beyond that, however, the assumption that God's existence is logically contingent, even if theologically acceptable, would give the divine command theorist only part of tile dependence relation he is looking for, and a very minor part indeed. It would, to be sure, enable him to claim that morality requircs the existence of God; for if God is not a necessary being, it does follow, given (Tla), (TI b), and (Tlc), that therc are possible worlds in which everything is permitted, namcly those in which God fails to exist. But it does nol foUow, even if God's existence is logically contingent, that morality depends, in the required sense, upon tile will or the commands of God; for it is quite possible both that ,norality requires the existence of God and that God always commands what is morally required, in part at least, because it is morally required. To see why this is so, one need only suppose, as before, that loving kindness is an essential property of God, and that an essentially loving God (logically) couldn't command certain things, such as cruelty for its own sake, and couMn'r forbid certain others, such as the genu#te love of one person for another. One need only suppose, in other words, that at lcast some of God's commands are no more a matter of "free ch.oice" than his beliefs concerning wbat is true. Given these assumptions, it still follows, even if God is not a necessary being, that a good many moral judgments are true in all possible worlds and hence necessarily true; for example, that cruelty for its own sake is nor morally required and that the gemtine love of one person for allother is not forbidden. Moreover, if we should restrict our attention to the set S of worlds in which, cruelty for its own sake is physically possible, 3 we shall find that, with respect to all of the worlds in S, either cruelty ['or its own sake is forbidden or else nothing .is forbidden. In some of the worlds in S, of course, God will fail to exist; so in these worlds, cruelty for its own sake is indeed permissible, though not morally required. With respect to these worlds,

196 perhaps we can say, speaking rather loosely, that they lack a moral order. But in the remaining worlds in S, those in which God does exist, we can presumably say not only that they have a moral order but the same moral order. They have the same moral order at least in the sense that God's will concerning cruelty for its own sake does not vary from world to world: irl each of these worlds cruelty for its own sake is morally fo,'bidden. So the sum of the matter is tllis: Even if God is not a necessary being, (Tla), (Tlb), and (Tlc) are quite colnpatible with the view that a good many hnnmn acts c'ouldH'l be morally required and a good rnany others eouldu't be forbidden; and they are quite compatible with the view that God's commands do not vary from world to world, at least not with respect to fundamental moral principles. They are quite compatible, in other words, with the view that there is but one, and only one,possible moral order. Now if there is but one possible moral order, if God's commands do not vary from world to world, at least not with respect to the most fundalnel~tal moral priuciples, little sense can be given to the claim that morality depends upon the will of God. For (';od. on this view, does not settle upon one possible set of moral principles in preference to another; there is in fact only one possible set of moral principles. Nor does it follow, .just because everything is permitted in those worlds in which God fails to exist, that God is in any way responsible for the moral principles themselves. It is quite possible, after all, that moral principles simply have no application in those worlds in which God fails to exist, just as they have no application in those worlds in which there are no sentient creatures. A theist might believe, for example, that the whole institution of morality requires a supreme (and inlpartial) judge, one who will enforce the appropriate moral principles (whatever they might be) and who guarantees the ultimate triumph of justice;and he might believe that God is the only possible supreme judge. Apart from the existence of God, hc might therefore believe, there can be no assurance that what we call "morality" always serves an individual's best interest; and apart from that assurance, there can be no real moral obligation. But while such a set of beliefs clearly implies that lnorality requires the existence of God, it by no means implies that God is responsible lbr the moral principles themselves. God may, so far as this view is concerned, simply know what moral difference his own existence makes, and he may command something because he knows it is morally required and knows that morality is in the best interest of created persons. At this point, taking a slightly different tack, a divine command theorist might want to insist that, even if God is' a necessary being, and even if loving-kindness is one of his essential properties, God is nonetheless free not to issue any commands at all -- in which case nothing would be morally required and nothing forbidden. A divine command theorist might, in other words, simply give up on the claim that (1) If God should not exist, then nothing would be morally required, nothing forbidden, and everything permitted,

197 and substitute for (1) a claim such as: (5) If God should not have issued any commands, then nothing would be morally required, nothing forbidden, and every tiring permitted. Such a strategy, at this point, would perhaps have a two-fold advantage: It would, in tire first place, enable the divhle command theorist to zero in more specifically upon the commands of God, as opposed to lhe bare existence of God and it wot.,ld, in the second place, enable him to abandon an assumption of dubious theological value, assumption (A), and to replace it with one that is perhaps more plausible; namely: (B) There are possible worlds in which God exists but issues no commands at all. But unfortunately fl~r the divine command theorist, the conjt, nction of (B) with. (TI a), (TI b), and (Tl c) no more entails that morality depends upon the will of God than does the conjunction of (A) with (Tla), (Tlb), and (T]c). Certainly the conjunction of (B) with (Tla), (TI b), and (TI c) does entail this much: that there are possible worlds in which, though God exists, nothing is morally required and nothing forbidden. But that is quite consistent, for reasons of the kind already considered, with the view that morality is independent of the will of God. If one should suppose, for instance, that the worlds in which God issues no commands are also worlds in which he creates no persons at all, it wot,ld hardly be surprising to find that there are no moral requirements in worlds such as these. Presumably there would be no point whatsoever for God to issue colnrnands in a world in which he has not so much as even created other persons. So even if there are worlds in which God exists but issues no commands, it remains altogether possible, if not highly probable, that God's reason for issuing no commands in such worlds is just that there are no created persons in them and therefore no moral requirements in them either. Of course, taking his cue from this, a divine command theorist might adopt: (C) There are possible worlds in which God exists and creates persons but issues n o comman ds. But once again, the same difficulty that has plagued us from the beginning arises in a new form. It is at least possible, surely, that, in a world W in which God creates persons but issues no commands, his reason for not issuing any commands is just that there are no moral requirements in that world. Here one might imagine, to give .just one possibility, that in W created persons have a somewhat different nature than they do in the actual world; that in W it is psychologically impossible for anyone intentionally to harm another (and perhaps also, because of their clear perception of the love of God, psychologically impossible for anyone not to worship God). Perhaps in such a world there would be no such thing as a moral requirement and no point for a moral test; so perhaps in such a world God would issue no

198 commands at all. Of course one might dcny (with some plausibility, I think) that W, it" so described, is a genuinely possible world; but even if" it were not, one point seems indisputable: Only in special kinds of circumstances, surely, only in circumstances in which issuing commands would accomplish little or no overall good, would a God who h)ves those whom he creates choose not to issue any commands at all. Indeed, if an essentially loving God is necessarily disposed to forbid cruelty for its own sake, as seems plausible to suppose, then he will also, in certain kinds of situations, necessarily issue commands against it; for he could hardly forbid such cruelty, given (Tla), (Tlb), and (Tlc), without issuing commands against it. If loving kindness is an essential property of God, therefore, there seem to be but two options: either (a) there are no possible worlds in which God creates persons and issues no commands, or, if there are such worlds, then (b) they will be just the kind of worlds with respect to which one can plausibly say, as one can with respect to W, that God issues no commands in them simply because there are no moral requirements in them. Perhaps, then, we can st, mmarize the difficulty with Quinn's Simple Theory in the following way: As it stands, his Simple Theory simply doesn't imply that what God commands is morally required because he commands it;and though God might have many reasons for commanding wt/at is morally required beyond the mere fact that is is morally required, it is nonetheless quite consistent with Quinn's Simple Theory to believe that God sometimes, or perhaps even always, commands what is morally required .just because it is morally required. If the divine command theory is to imply that morality actually depends upon the will of God, therefore, some fi~rther assumption is requked. But what assumption? If one should adopt assumption (A), the assumption that God's existence is logically contingent, it does follow, given (Tla), (Tlb), and (Tic), that moral principles have no application in those worlds in which God fails to exist; and if one should adopt assumption (C), the assumption that there are possible worlds in which God creates persons but issues no commands, it also follows that moral principles have no application in some of the worlds in which God has created other persons. But in neither case does it follow that morality depends upon the will or the commands of God; for it is still quite possible that, in the worlds in which God does exist, he commands what is morally required just because it is morally required, and in the worlds in which he issues no commands, he chooses not to do so just because there are no moral requirements in these worlds. The basic difficulty here is the familiar one that the expression "if and only if" cannot, by itself, be used to express an asymmetrical relation, and that is just the kind of relation that divine command theorists are trying to express. What is morally required, they want to say, depends upon the will of God in a sense in which the will of God does not depend upon what is morally required. And to express that sort of dependence relation, a very radical assumption is required, as we shall see.

199 II Before setting forth the required assumption, however, it is worth considering why two less radical assumptions, those made by Professor Quinn, are insufficient. Quinn's assumptions are these: First, tile term 'God' functions like a proper nalue in that it picks out the same individual in all possible worlds where it picks out any individual, regardless of how that individual may alter from world to world. And, second, at least some of the things commanded by God differ as possible worlds vary. In brief, we shall assume that the individual named 'God' and believed by theists to be actually wise, powerful, and good, if there is such an individ'ual, cl,,~es not command exactly tile same things in every world where he exists (p. 25). But unfortunately, these assumptions will in no way remove the difficulty, already noted, with Quinn's statement of the Simple "lheory. For though the first assumplion does make it clear that 'God' is being used, in (Tla), (Tlb), and (Tlc), as a proper name, it gives us nary a clue of the extent to which the individual picked out by 'God' alters from world to world. And though the second assumption does tell us that at least some of God's commands "differ as possible worlds vary~' it tells us nothing whatsoever about the kinds of differences that are relevant to a divine command theory and is therefore hopelessly unspecific. Even an opponent of the divine command theory, after all, would concede that at least some of God's commands differ as possible worlds vary; but he wouid go on to insist that God's commands differ as possible worlds vary for the same reason that they differ (if he exists) as situations vary in the actual wo,ld. If God did, for instance, command monogo,ny in the Ncw Testament after permitting polygamy in the Old, it should hardly be surprising to discover that he also commands monogomy in some possible worlds but not in others. But the reason for such differences might simply be that, while independent of the will of God, morality is nonetheless cotztextually dependent; that is, what is morally required does vary from context to context and therefore would vary from world to world. Just as the practice of eating pork might be perfectly healthy in one social context but not in another, so the practice of polygamy might be morally permissible in one social context (where it. contributes to human happiness) but not in another. Accordindy, it is quite possible that God's commands vary from world to world simply because what is morally required varies from world to world and God, being morally perfect in every world in wMch he exists, always commands what is morally required just because it is morally required. There is another possibility worth considering at this point as well, one that is easily neglected; namely that, while s o m e moral obligations depend upon the will of God, many others do not. Of course nothing in Quinn's Simple Theory requires even the concession that s o m e moral obligations depend upon the will of God; but if there is a (;od, and he is the Creator, he presumably has all the rights and privileges of the Creator. And if even a father, charged with the responsibility for

200 rearing his son, is sometimes entitled to enforce, at his own discretion, rules of his own devising, the same might also be true of a Creator, whose responsibility it is to care for all created persons. What an opponent of the divine command theory denies, however, is that all moral obligations depend upon the will of God. Just as the rights and privileges of a father must be construed as an expression of a deeper obligation, the obligation to love and care for the children whom he helped to bring into the world, so also must tile rights and privileges of the Creator be construed as an expression of a deeper responsibiliO, , the responsibility to meet the true needs of those whom lie brings into existence. Accordingly, the Creator, if he exists, no more has the right to harm created persons, or to issue commands harmful to them, than a father has to issue commands harmful to his children; and no one, whether lie be a father, a civil magistrate, or the Almighty himself, has the right to command cruelty for its own sake, or to encourage such cruelty in any way. Nor is anyone, if commanded to engage in cruelty for its own sake, ever obligated to obey such a command. Nor does the obligation to obey such a command follow from the obligation to obey one's parents, or even the obligation to obey the Creator himself. The obligation to obey God, however, unlike the obligation to obey one's parents, may nonetheless be absolute and unconditional; for if God is morally perfect in every world in which he exists, then in no world will he ever issue a command that he is not entitled to issue. So there are, then, several possibilities that Quinn has so far failed to rule out. l-yen granted (-l'la), (Tlb), and (Tlc), it is still possible that none of our moral obligations depend upon the will of God, and it is also possible that, while some of our moral oblisations depend upon the will of God, many others do not. To rule out these possibilities, one must assume not only that God's commands "differ as possible worlds vary," as Quinn does, but also that they differ in certain relevant kinds of ways. But just what makes for a relevant kind of difference at this point? The basic difficulty with the assumptions considered so far, particularly assumptions (A) and (C), gives us a clue. For the basic difficulty is that loving kindness, justice, benevolence, etc. are essential properties of God; but if these are essential properties of God, then a good many moral principles will turn out, given (Tla), (TI b), and (Tic), to be necessarily true and hence (in that sense) independent of the will of God. What this suggests, therefore, is that aa moral principles (and an moral obligations) depend upon the will of God only if God's commands are radically different in different possible worlds, so different that the very l)o#ll of his commands is different in these different worlds. What it also suggests is that, to rule out the possibility that at least some moral principles are necessarily true and therefore independent of the will of God, one must assume that God's commands are h~ ~zo ~:al' constrained by his essential properties; that his character is so radically different in different possible worlds that he is capable of comrnanding virtually any conceivable act in any conceivable kind of situation. For as Quinn himself rernarks:

201 What God can command may be a function of what his essential properties are. Thus, if he is necessarily perfectly good, there may be certain things that he commands in no possible world, llowever, we shall ignore this complication and suppose instead that God is free to command anything he chooses to comnland (p. 31). But unl'ortunately for Quinn, the complication he wants to ignore is just the complication he can't ignore and in fact doesn't ignore. By assuming that God is "free to command anything he chooses to command" including cruelty for its own sake, Quinn in effect assumes that loving-kindness, justice, benevolence, etc. are not essential properties of God; that even such awkwardly stated properties as not being malicious and not being cruel are purely accidental properties of God. Nor is there, if the above arguments are correct, some alternative assumption that might have been adopted at this point, that might have enabled him to show that all moral principles depend upon the will of God. Where 'x' is a variable ranging over every action that is logically possible and every compossiblc set of actions, the in-. dicated assumption would thus seem to be: (D) For any x, it is logically possible that God should have commanded x. And if (D) is conjoined with (Tla), (Tlb), and (TIC), then we at last do have a theory which implies that all moral principles, and all moral obligations, depend upon the commands of God. But if (D) is indeed a required addition to the divine command theory, as I have argued, it is also a fatal defect; for as we shall see in the following section, the conjunction of (D) with (Tla), (Tlb), and (Tic) clearly has unacceptable implications, implications that no reasonable person (who fully understands them) could possibly accept.

Ill Perhaps the most important implication that assumption (D) has for theology is that loving-kindness, justice, benevolence, etc. are not essential properties of God; that the individual picked out by 'God' therefore has vastly different attitudes, and a vastly different character, in different possible worlds. What assumption (D) implies, in other words, is that there are worlds in which God is malicious and cruel, no less than those in which he is loving and kind -- worlds in which, being unjust, he decidedly plays favorites (as many religious people have always believed), no less than those in which he is always even-handed, always just and fair. For there is, presumably, some connection between the kind of commands issued by God in a given world and the kind of individual he is in that world: not just any command can be seen, in a given situation, as an expression of love, or lnercy, or justice, for example. It would be nice, of course, if we could spell out clearly, at this point, the exact difference that God's love, or his justice, or his benevolence is bourld to make in an.r possible world; but the complexity of the various possible

202 worlds would seem to make this a practical impossibility. It is surely enough to observe, however, that if God's love makes any d(r at all in the commands that he issues, if it is incompatible with any conceivable command in any conceivable kind of situation, then it follows from assumption (D) that there are possible worlds in which God lacks that property; and similarly, if the property of not being malicious and cruel is incompatible with any conceivable command in any conceivable kind of situation, then there are also possible worlds in which God is malicious and cruel. Indeed, assumption (D) was introduced in the first place precisely because it seems to have implications such as these, precisely because it seems to imply that God has a radically different character in different possible worlds. Now, admittedly, if taken by itself, this assumption need not be regarded as altogether disastrous for theology. For though God is, I am inclined to think, a more fitting object of worship if loving kindness can be numbered among his essential properties - if he (logically) couldn't fail to be loving and kind - the opposing view is perhaps not entirely unreasonable. What is really important, a theist might believe, is that God is in flTct loving and kind ..... that is, loving and kind in the actual world - and the bare logical possibility that he might not have been loving or kind need not detract from the grandeur of his character in the actual world. A theist might even believe that a supreme being who could have been malicious and cruel, had he so chosen, but who in fact has chosen to be loving and kind, is, for that very reason, a more fitting object of worship than one who is (logically) incapable of any unkindness whatsoever, incapable of anything but love. And though this view is, I am inclined to think, a mistaken one, it is perhaps not entirely unreasonable, as I have said. But whatever one might think of assumption (D) when taken by itself, the conjunction of (D) with (TI a), (Tlb), and (Tlc) has implications that are simply outrageous; for the conjunction of (D) with (Tla), (Tlb), and (Tlc) implies not only that there are possible worlds in which God is malicious and cruel, but also that: (E) There are possible worlds in which God, being malicious and cruel, issues commands that are likewise malicious and cruel, and in these worlds God's commands are nonetheless morally binding. Now it is one thing to admit the logical possibility that God is malicious and cruel, or the possibility that he could have commanded cruelty for its own sake, but quite another to claim that God's commands, however malicious and cruel, are nonetheless morally binding. If in a world W in which he commands cruelty for its own sake, God were also omnipotent, or simply very powerful, the most heinous acts might indeed be excusable, and one might have grounds for doubting that moral obligation could even exist in such a world. But the implication that such acts of cruelty and torture, committed solely for the purpose of making people miserable, could have been morally required is surely outrageous - not just false

203

but obviously false .... and any theory which has such implications obviously unacceptable. Of course, in saying this, I fully recognize the difficulty of trying to refute a theory by drawing implications from it. The difficulty is that a proponent of a theory, if hard-nosed enough, might just accept all of its implications, however unattractive they may seem to his opponents, for the very reason that they are inlplications of his theory. But in the case of the divine command theory, this difficulty is not nearly as serious as one might be inclined to think; for, once the ilnplications of the theory are fully clari[ied, one is left with but two possibilities: Either (a) the theory really is false because implications such as (E) are just false, or else (b) the theory will turn out to be, as we shall see, so trivial and vacuous that no one, not even the nontheist, will find it at all unattractive or controversial. As is so often true in philosophy, the distinction between clar([~v#N a theory and r@tting it (or between clarifying a theory and establishing it) is virtually impossible to draw with any precision. Consider, in this connectkm, Quinn's own attempt to reconcile us to the implication that cruelty for its own sake (or gratuitous cruelty, to use his own expression) could have been morally required, lie begins by attacking the reliability of intuition. He points out, in the [irst place, that moral intuitions "fail to produce agreement about controversial issues, as recent actual cases involving abortion, euthanasia and similar problems show quite clearly;" suggests, in the second place, ttlat intuition is especially unreliable when one is forced "to go beyond actual moral problems into the world of the merely possible" (p. 58); and concludes, finally, that any objection to the divine command theory that presumes upon the reliability of intuition is nothing less than ntoral dogmatism. But there is, he suggests, "no reason why a divine command theorist should subscribe to a view which licenses moral dogmatism on the part of its critics" (p. 60), and, besides, possible worlds in which God commands gratuitous cruelty might in fact be quite different from the actual world: Possible worlds with the greatest over-all similarity to tile actual world in which God commands what we call 'gratuitous cruelty' might yet be very unlike tile actual world, so dissimilar that intuition is an unreliable guide to what is required and what forbidden there. It might be, for example, that in such worlds what we call 'gratuitous cruelty' provides cathartic release for its perpetrators without causing pain to its victims, for over-all similarity is not necessarily tied to sameness of causal laws. And the respects in wt:ich such worlds differ from the actual world might have a great deal to do with how tMngs acquire moral status .... If the critic insists that his intuitions about the matter must settle the question, then surely our divine command theorist can with justification reply that this tactic begs the question against intuitions shaped by and congruent with divine command theories (p. 60). But if such a line of defense seems convincing, I want to suggest, it is only because it tends to conceal from us the full implication of a divine command theory; and the way to counter such a defense, therefore, is simply to make tile issues at stake clearer. To that end, perhaps several comments are in order:

204 (1) The first point to be made, at this point, is that certain implications of tile divine command theory, implications that a critic will inevitably find unacceptable, must be distinguished carefully fiom others that are easily confused with them. An unacceptable implication, as we have seen, is that gratuitous cruelty could have been morally required; but this ilnplication lnUSt not be confused with the one that Quinn acknowledges, one that is admittedly quite innocuous: that there are possible worlds in which what we call 'gratuitous cruelty' is morally required. For of course what u,e call 'gratuitous cruelty' may not be actual cruelty at all, particularly not if it "provides cathartic release for its perpetrators without causing pain to its victims," as in Quinn's example. Consider the act of bashing someone over the head with a baseball bat. Presumably we would call this 'an act of cruelty,' or 'an act of gratuitous cruelty' if done simply for one's own pleasure, by reason of the effect that such acts typically have, or at least are typically believed to have, in the actual world. But .if a situation should arise - and here it matters not whetller it be in the actual world or in some merely possible world - if a situation should arise in which bashing so,neone over the head causes him no pain (and inflicts upon him no harm), thcn what we could ordinarily call 'an act of cruelty' is clearly not, in this case, an act of cruelty at all, only something mistakenly described as cruelty. Nor is it rcmarkable, surely, or in any way controversial, that actions mistakenly described as cruelty are sometimes morally required. What the divine command theory implies, however, is that actual cruelO,, not .just anything that happens to be called 'cruelty,' could have been morally required; that God could have commanded cruelty in the full knowledge that it does cause pain to its victims and does nothing for the perpetrator, save but to make him a more vicious individual; in the full knowledge that, if obeyed, such a comrnand would increase markedly the overall balance of human suffering over human happiness in the world. ]'hat is the implication so offensive to critics of the divine command theory, and it .is just that implication which Quinn has unfortunately failed to address. (2) Consider now the matter of the reliability of intuition. To judge this matter fairly, one must, at the very least, distinguish between two kinds of cases: (a) those in which the person making a moral judgment is fully informed concerning the relevant facts of a case, and (b) those in which he is not so informed. If a person is not fully informed concerning the facts of a case 9 if, for instance, he is ignorant of some of the important consequences of an act - then one need not postulate a bizarre possible world, with wildly different causal laws, in order to show that intuition is going to be an unreliable guide. But then presumably no one would expect of intuition that it be a reliable guide in cases such as these. What one expects, if one is inclined to trust intuition at all, is that intuition is a reliable guide in those cases, or at least in some of those cases, where one is fully informed concerning the relevant facts; that is, in those cases where one is fully informed concerning the consequences of an act, the context in which it is (or could be) performed, the motives behind it, etc. Indeed, Quinn's own example is based upon the

205 assumption that intuition is' reliable in cases such as these. It is based upon the assumption that almost anyone would agree to the following: that if "what we call 'gratuitous cruelty' provides cathartic release for its perpetrators without causing pain to its victims," then it is perhaps not morally reprehensible after all. In this example, therefore, it is Quinn who presumes upon the reliability of intuition who presumes that we can all discern, and agree upon, the moral relevance of causing pain to others and producing good in the world (e.g., the cathartic release). But if Quinn's own example presupposes the reliability of intuition, it hardly gives us grounds for mistrusting intuition, and certainly not in those cases where one is fully informed about the details of another possiblc world. (3) Contrary to what Quinn would have us believe, moral judgments about possible cases are often less dubious, and far easier to agree upon, than those about actual cases. For one thing, an assessment of actual cases inevitably presupposes an account of the facts, which account cannot, as it might in merely possible cases, simply be assumed to be true. Take, for instance, tile issue of capital punishment. 111 assessing tile practice of capital punishment, one cannot simply assume that this practice will, or will not, act as a deterrent against murder, or that it will, or will not, lnake a society less humane; for these issues arc themselves an essential part of the dispute to be resolved. Moreover, whereas possible cases can often be simplified, in various ways, for the purposes of illustrating a specific point, actual cases are apt to involve complexities, such as knotty conceptual problems, that cannot be eliminated by the mere expedient of more precise stipulations. Here it is perhaps the abortion issue that comes most readily to mind. Even the most cursory review of the literature on abortion will uncover a bewildering array of problems, some of which involve border-line case questions that may indeed be unanswerable: When does the organism that begins developing at conception become a human person? 9 and what, exactly, are the criteria of personhood anyway? At what point can an individual be said to have genuine interests and therefore rights - and what are the rights of an individual (whether he be an unborn child or a famous violin player) whose life happens to depend upon the physical body of another? That such questions should generate controversies is hardly surprising, but these controversies are no more grounds for wholesale moral skepticism than they are for skepticism concerning the existence of other persons. And in particular they certainly provide no grounds for skepticism concerning non-controversial matters, such as the immorality of inflicting pain upon others simply for one's own pleasure. Accordingly, the important distinction, if one's concern is with the reliability of our most basic moral intuitions, is not the distinction between actual cases and possible cases, for many possible cases may be less controversial than some actual cases. The important distinction is the one between controversial cases, those involving unresolved factual disputes and unresolved (perhaps even unresolvable) conceptual problems, and non-controversial cases, those concerning wMch of our ordinary moral intuitions are clear and decisive.

206 (4) But while the criticism of the divine command theory, like that of so many other theories, no doubt requires, at some point, an appeal to fundamental moral intuitions, it by no means follows that a critic of the theory can only resort to dogmatic assertion at this point. For one thing, there is the matter of clarification, as wc have said. Since proponents of the divine command theory inevitably conceal from themselves (and others) some of its more horrendous implications -- implications such as (E) above .... part of the critic's strategy must simply be to clarify the neglected implications in the confidence that no one who fully understands them is likely to accept them. Moreover, if a proponent of the theory should perhaps be willing to accept all of its implications, which seems unlikely, the critic will then simply point out that the theory has become utterly trivial and vacuous, utterly uninteresting; and this can perhaps be done in the following way. One implication of the theory, the critic will point out, is that God's commands not only differ as possible worlds vary, but, as we have seen, the whole point of his commands, and therefore the whole pohlt of the rules embodied in his commands, likewise will differ as possible worlds vary. In some words, those in which God is loving and just, the point of his commands will presumably be to promote the welfare of persons -- or more generally the welfare of all sentient creatures in part, perhaps, by promoting relationships between persons that arc just and fair. This does not mean, of course, that the right act in such worlds is necessarily one that produces the greatest good for the greatest number, nor even that the consequences of an act are more important than motives in determining the rightness of an act. But it does mean that God's purpose in issuing commands, in such worlds, is a loving one, and therefore that the whole point of the rules embodied in his commands is to promote welfare of some kind (and some degree). In other words, however, those in which God is malicious and unjust, his purpose will of course be quite different. In these worlds, the whole point of God's commands will presumably be to maximize discord, conflict, and misery in the world. Moreover, not only will God's purpose in issuing commands be different in different possible worlds, but the very notion of a ration',d motive for obeying God's commands will be different in different possible worlds. In those worlds in which God is loving and just, a rational motive for obeying his commands would presumably be love of God, rooted in an awareness of his perfections, or perhaps even love of fellow man (or love of other persons). But what could possibly count as a rational motive for obeying God in those worlds in which he is malicious and unjust? One possibility might seem to be fear of the consequences of disobedience. But as it turns out, not even fear will provide a rational motive for obedience; for if God is truly malicious and unjust, hc might ultimately punish even those who obey him, no less than those who disobey him. Accordingly, in those worlds in which God is malicious and unjust, a rational motive for obeying him is perhaps simply inconceivable. Be that as it may, if the whole pohzt of the rules embodied in God's commands is radically different in different possible worlds, then the divine command theory surely is utterly vacuous and trivial. It is vacuous and trivial because, given

207

such differences in different possible worlds, no sense whatsoever can be given to the claim that (6) Necessarily, if God commands that p, then p is morally required, beyond what might be given to the claim that (7) Necessarily, if God commands that p, then God commands that p. At least Quinn has given us no indication whatsoever of what it would mean to assert (6) beyond what it means to assert (7). (5) Another way for a critic to bring out the utter vacuousness of the Divine Command Theory would be for him to make the divine command theorist a present of the term 'morality' and then proceed to argue in the following way. First, he would introduce an alternative term, for example that of 'sorality,' and explain that the point of sorality is just what the point of morality is in lhose worlds in which God is loving and just. He would insist, however, that the point of sorality, unlike that of morality, remains the same in all possible worlds containing persons; so in some possible worlds what is soratly required (in a given situation) will of course be closer to what God would command (in that situation) if he were loving and just than to what he in fact does command. Secondly, the critic would acknowledge that, if God happens to exist, and happens to be loving, kind, and sorally perJk~ct, then what is sorally required and what is morally required lnay indeed be the same in the actual world. But he would go on to point out that, given Quinn's Simple Theory, the whole point (and hence the very nature) of morality, unlike that of sorality, is different in different possible worlds; so conflicts between what is sorally required and what is morally required are at least possible. In a world in which God commands gratuitous cruelty, for instance, gratuitous cruelty, though morally required, will not be sorally required (and may even be sorally forbidden). Finally, the critic will point out that, given our explanation of sorality, and given Quinn's Simple Theory, sorality is obviously superior to morality, at least in this sense: that if a conflict should arise between what is sorally required and what is morally required, what is sorally required is obviously more humane and more just. Moreover, even in the case where the same act A is both sorally required and morally required, a person whose desire it is to promote human welfare and justice will do A because it is sorally required, not because it also happens to be morally required. Hence sorality is, in this sense, superior to morality ; indeed, given Quinn's Simple Theory, morality no longer seems to have the kind of importance traditionally attributed to it. I conclude, therefore, that Quinn's defense of the divine command theory is unsuccessful, and my argument can perhaps be summarized as follows: Quinn's Simple Theory, (Tla), (TI b), and (Tlc), implies that a//moral obligations depend upon the will of God only if an additional assumption is adopted, something like assumption (D); but the conjunction of (TI a), (Tlb), and (Tlc) with assumption

208 (D) has implications that are utterly outrageous, implications such as (E). Moreover, if a divine c o m m a n d theorist should perhaps be willing, at this point, to 'bite the bullet,' so to speak, if he should be willing to accept the full implication of his own theory including such implications as (E), he would succeed only in making th.e theory utterly w~cuous and trivial, so that morality would no hJnger have the kind o f importance traditionally attributed to it. Despite Quinn's ingenious efforts, therefore, the divine c o m m a n d theory rernains just as inadequate as most philosophers have always believed it to be.

NOTES 1. Philip L. Quinn, Divine Commands and Moral Requirements, Oxford, 1979. All page references arc lrom this work. 2. In this essay, I shall use the term 'morally forbidden' as if it were synonymous with "morally wrong' or simply 'immoral'. That an action is morally forbidden, in other words, should not bc tlaought to imply that some pers'on or some authori O, has in fac~ forbiddcs it. 3. As I am using the term 'physically possible', an action A is physically possible in a world W only if persons exist in W and it is physically possible in W for a person P to engage iJ~ cruelty towards others only if persons other than P also exist in W.