"They that deny a God destroy a man's nobility, for certainly Man is of kin to the beasts by his body; and if he be not of kin to God by his spirit, he is a base and ignoble creature." - Francis Bacon, Of Atheism

PLATO'S DILEMMA
Consider the following doctrines:

a) God is good.

b) God wills us to do what is good.

c) God is the basis of ethics.

The first and second of these are held true in Christian, Judaic, and Islamic thought. They are not, however, essential to the idea of a creator of the universe. There could be a transcendent creator of the universe who has no moral attributes, or who is indifferent to our moral state. But, in this chapter, we shall consider those varieties of theism which argue for a tight connection between God and morality. This takes us to doctrine c). What it asserts is that a proper understanding of the nature of morality must make reference to God, an idea that, in one form or other, has played a crucial role in theistic thought. Expressed like this, however, the idea is somewhat vague, and open to more than one interpretation. Here is one, not very plausible, interpretation: morally correct behaviour depends on belief in God. If this implies, as it seems, that atheists are liable to behave badly, then we can reject it quite quickly. Atheists may have exactly the same views about what counts as good and bad, and may behave just as well, or as badly, as theists. A more interesting interpretation of c) goes as follows: in deciding what is right, the theist must make use of some religious image. For example, the Christians may take Jesus as the model on which to base their idea of right conduct. In this sense, faith informs moral conduct.

However, in this chapter, I want to pursue a quite different, and more metaphysical, interpretation of c), namely the idea that the existence of God explains the existence of moral values. The fact that certain acts are good or bad depends in some way on God and his properties. Some theists have appealed to this explanatory role in arguing for the existence of God. The argument, knows as the 'moral argument' for God, goes roughly like this:

The moral argument

1 There are moral values

2 The existence of those values depends on the existence and nature of God.

Therefore: God exists.

Premise 2), note, is one way of rephrasing doctrine c).

It will be helpful at this point to make a distinction between ethics (or 'first-order' ethics), and meta-ethics (or 'second-order' ethics). Ethics is concerned with such questions as what the best kind of life would be, or what I ought to do and which rules to adopt by which I could decide what I ought to do. Meta-ethics, in contrast, is concerned with the status of ethical judgements. 'Is theft wrong?' is an ethical question. 'Are there objective moral values?' is a meta-ethical question. Some theists believe that there is an important connection between God and ethics, in that in deciding what to do I must make appeal to God. What in contrast we are now concerned with is the idea that there is an important connection between God and meta-ethics, in that a proper understanding of what is involved in ethical judgements must involve appeal to God.

The aim of this chapter is not so much to undermine the moral argument for God's existence, though this will be one result of the discussion, but rather to cast doubt on the idea that the existence of moral values depends on God. The problem with doctrine c), given this interpretation of it, is that it makes difficulties for our understanding of a) and b) - a problem first noted by Plato.

In his dialogue Euthyphro, which is concerned with the nature of piety, Plato presents us with the following question: 'How are we to understand the idea that God wills us to do what is good?' There are two answers we could give to this question:

A God wills us to do what is good because certain acts are good, and he wishes such actions to be performed.

B An act is good only because God wills it.

It seems that, whichever way we answer the question, we get an unhappy result - or at least an unhappy result for the theist. Suppose we choose answer A): God wills what is good because, independently of his will, it is good. It seems to follow from this that moral values exist independently of God. That is, even if God had not existed, there would still have been moral values, so the basis of ethics has nothing to do with theism. This goes against the theist assertion that ethics is informed by the fact that God exists, in that we cannot divorce questions concerning what is right from questions concerning God's design for us. If, on the other hand, we opt for answer B), and say that something is good by virtue of the fact that God wills it, then the assertion that God wills us to perform good acts just reduces to the unenlightening assertion that God wills us to do what he wills us to do. Of the three doctrines which we began this section with, then, b) appears to conflict with c), under its metaphysical interpretation. There is also an apparent conflict between a) and c), for, if ascribing goodness to something just means that God wills it, then the assertion that God is good becomes the curious and morally empty assertion that God wills that he be as he is. This can hardly be represented as one of the foundation stones of the religious life. This problem is neatly captured by Bertrand Russell in his essay 'Why I am not a Christian':

"If you are going to say, as theologians do, that God is good, you must then say that right and wrong have some meaning which is independent of God's fiat, because God's fiats are good and not bad independently of the mere fact that he made them. If you are going to say that, you will then have to say that it is not only through God that right and wrong came into being, but that they are in their essence logically anterior to God." (Russell 1957, p. 19)

This amounts to more than just a criticism of the moral argument for God (which was Russell's target at that point in his essay), for we can present these reflections in a more structured way to provide an argument against theism on the ground that it contains an inconsistency. It cannot be true both that the existence of moral values depends on God and that the statement 'God is good' makes a morally significant assertion. Consequently, theism is false, because it is incoherent. Making the moves absolutely explicit, we can construct an argument for atheism which exploits a version of premise 2) of the moral argument for God. We can call it the 'meta-ethical argument for atheism':

The meta-ethical argument for atheism

1 If theism is true then 'God is good' is morally significant.

2 If theism is true then God plays an explanatory role in ethics.

3 If 'God is good' is morally significant, then moral goodness must be independent of God.

4 If God plays an explanatory role in ethics, moral goodness cannot be independent of God.

Therefore: Theism is false.

How exactly is the conclusion reached? Taking it more slowly, 1) and 3) together imply 5):

5 If theism is true then moral goodness must be independent of God,

However, 2) and 4) together imply 6):

6 If theism is true then moral goodness cannot be independent of God.

Putting 5) and 6) together, we obtain:

7 If theism is true then moral goodness both is, and is not, independent of God.

Theism, in other words, is self-contradictory and hence false.

We can construct an exactly parallel argument, substituting 'God wills us to do what is good' for 'God is good'. By doing this, we capture the challenge to theism posed by Plato's dilemma.

Now, in order to reach the conclusion of the meta-ethical argument, we have had to ascribe to the theist a particular philosophical doctrine, namely that the existence of moral value is explained in some way by the existence and properties of God. Not all theists will necessarily assent to this meta-ethical assertion. But for the more philosophically inclined theists, this is an important aspect of God. God, for them, explains the existence of many things, and, since God is also a moral being, it is entirely natural to suppose that moral value somehow resides in God. So theists have a reason to defend premise 2) of the meta-ethical argument. What they must do, then, is to resist its apparent consequences. To do this, they must attack the argument at some point. The most controversial premise of the argument, I suggest, is 3). To see whether or not it is true, we need to examine the concept of goodness.

DESCRIPTIVE VERSUS PRESCRIPTIVE MORALITY
Consider once again Plato's question: 'Does God will us to do what is good because, independently of him, it is good, or is it that what is good is so only because he wills it?' The assumption here is that we must choose one or the other, not both, and at first sight it does indeed seem that we cannot say 'yes' in response to both of these questions. However, the word 'good' can be defined in more than one way, and the possibility arises that, in one sense of 'good', acts are good independently of God's willing us to do them, and that, in another sense of 'good', acts are good because he wills them. John Mackie has suggested that, by making a distinction between different senses of 'good', the theist can escape Plato's dilemma (though not other snares, according to Mackie, who is no apologist for theism). If this is so, then the theist may also be able to escape the meta-ethical argument. So let us look at the distinction in question.

We can distinguish between the descriptive and the prescriptive elements in morality. It is the case that x is a good thing to do for y, in a purely descriptive sense, if x benefits y in some way: it enables y to live a happy life, say, or it contributes to the stability of the society in which y lives. 'Good' in this descriptive sense attaches to a wide variety of actions. It is a good thing for y to eat, to dress according to the weather, to communicate, to refrain from ending the lives of a large number of people, etc. Morally good actions constitute a subset of this class. But, in addition to this descriptive sense of 'good', there is a sense which carries an implication of requirement: there are certain things that one ought to do. To recognise that x is good in the prescriptive sense is to recognise an obligation to do it. Of course, we can appropriately talk of 'ought' even if we are using 'good' in a merely descriptive sense: one ought to eat, to remain upright when walking, and so on. But the 'ought' here is only hypothetical, or conditional upon some purpose. One ought to eat if one wants to stay alive, one ought to act on one's desires if one wants them to be satisfied. In contrast, the prescriptive sense of 'good' carries with it an unconditional 'ought': one ought to refrain from killing, full stop, not merely if one wants to avoid censure. These two senses of 'good' are not necessarily in opposition to each other, though it would be possible to maintain that only the descriptive sense of 'ought' is legitimate, and that therefore the only kind of obligation there is is a conditional one. On this view, one ought to behave morally only if one wants to bring about a certain end, such as the greatest happiness of the greatest number, or the leading of a fulfilling life. So, in making the distinction, and holding that there are actions which are good in a prescriptive sense, we are not just defining words, we are assenting to a certain conception of morality. What this conception of morality is will emerge as we proceed, but, for the time being, we will consider the suggestion that the distinction undermines Plato's dilemma.

Something can be descriptively good, i.e. in our best interests, whether or not God wills it, and so in this sense is good independently of God's will. Of course, if theism is right, even descriptive goodness is not entirely independent of God, since he created things in such a way that doing certain things would be good for us, but that God wills us to do what, as a matter of fact, is good for us does not reduce to the triviality that he wills us to do what he wills us to do. He wills us to do what is descriptively good (i.e. good for us) because he is benevolent: he wants us to do what will enable us to live in harmony with others compatibly with pursuing the goal of self-fulfilment. So what is descriptively good is independent of God's will, in the sense that we do not have to refer to God in defining it, although it may be that God is the causal explanation of why, as a matter of fact, certain things are good for us. However, we can say quite consistently with this that the fact that something is prescriptively good is not independent of, but constituted by, God's will. The additional element of requirement that attaches to morally good actions is provided by God's requiring us to do these things. So he wills us to do what is descriptively good, such as to eat and refrain from killing each other, but, in addition, places on us an unconditional requirement to perform certain of these descriptively good actions. He does not require us to eat, but he does require us to refrain from killing. The fact that there is a distinctive group of descriptively good actions which are morally valuable does depend directly on God's will. The good, x, is prescriptively good because God requires it, not merely because he wills it. The theist can then present his as God's crucial explanatory role in ethics.

Consider now the assertion that 'God is good'. The meta-ethical argument presents us, in effect, with the following dilemma: either 'God is good' is morally insignificant, or moral goodness is independent of God. Can we employ the descriptive/prescriptive distinction to avoid both horns of this dilemma? Mackie's suggestion is that, when we attribute goodness to God, this is to be understood in purely descriptive terms, that God is disposed to do things which are good for us. It should not be taken as meaning that God does what is required to do, for, since only God can place this requirement on any agent, it would mean that God requires himself to do what he does, and this has no moral content to it whatsoever. So both Plato's dilemma and the meta-ethical argument can be seen as conflating two distinct senses of 'good'.

A way out of Plato's dilemma and the meta-ethical argument has been provided for the theist. But is it the right way out for the theist to take? I suggest not. The descriptive sense of 'good' is surely too weak to capture the theist's conception. If the only coherent meaning we can ascribe to 'God is good' is that he is disposed to do good things for us, then he is good only in a conditional sense. If we value certain things, then God is good. If we do not, then he is not. No theist is likely to settle for such a weak rendering of 'God is good'.

So far we have considered a) and c) of the theist's doctrines as if they were quite independent propositions. But if, instead, we think of them as being closely related, another way of disarming the meta-ethical argument suggests itself. When we say 'God is good', part of what we mean, no doubt, is that God is at least analogous to a morally good member of the human race. That is, God has some properties that we would consider good in a human. For example, just as good parents (in the sense of morally good, not just effective, parents) would take care of their children, so God takes care of his creation. But this cannot exhaust the goodness of God, for there are few analogies we can draw between ourselves and a divine being. What else is implied in the statement that God is good may be that, unlike humans, God is a source of moral value. That is, God's goodness in part consists of the fact that he is the basis of ethics. Since it is not trivial that God plays such a role, it cannot be trivial that God is good; in fact it is highly morally significant, because it points to the source of moral obligation. What effect do these considerations have on the meta-ethical argument? They appear to undermine premise 3):

3 If 'God is good' is morally significant, then moral goodness must be independent of God.

We can now see that this need not be true at all. The theist can, surely, hold that 'God is good' is morally significant because it identifies the source of moral obligation, which implies that moral goodness is not independent of God. So 3) should be replaced with:

3* If 'God is good' is morally significant then moral goodness is not independent of God.

But, of course, this completely blocks the meta-ethical argument.

MORAL REALISM AND MORAL SUBJECTIVISM
The atheist has another line of attack, however. The attempt to show that theism is internally inconsistent when it comes to facts about morality may have failed, but the atheist can instead attempt to show that the most plausible theory concerning the basis of ethics leaves no room for God to play any significant role in the explanation of moral value.

Let us begin by considering a problem to which ethics should have an answer. It is agreed by everybody that moral properties go hand-in-hand with certain non-moral, or natural, properties. Suppose that I freely and deliberately deprive someone of their livelihood - for example, by burning down the local tea shop. I do this simply because I want to, not because there is any conflict between the owners' interests and mine - for example, because I disapprove of tea-drinking, or because I wish to avenge some harm they have done me. That this act of mine is wrong there can be little doubt. Suppose now that everything that one could say about this act, apart from its being morally wrong, could also truly be said about another act, done by somebody else. Then, if the first act is wrong, so is the second. The general principle on which this rests is that moral properties (such as being wrong) supervene on certain natural properties (such as causing gratuitous injury): i.e. that if two acts share all the relevant natural properties, they also share the same moral properties. This is not to say that moral properties are nothing but natural properties, merely that there is a systematic correlation between them. Now here is the problem: what makes it the case that certain moral properties supervene on certain natural properties? Why should there be this invariable connection?

This problem is particularly acute for the moral realist. Moral realism holds that it is an objective, mind-independent fact that an act, defined according to its natural properties, has the moral value that it does. Consider suicide. To define it as the intentional taking of one's own life is to define it in a morally neutral way, according to its natural properties. Now, suppose one thinks that such an act is wrong. The problem is then to explain why any act with these natural properties is wrong. Theism has an answer to this question: it is because of God that acts with certain natural properties also have a certain moral property. God condemns acts with natural properties x, y and z, and that is what makes such acts wrong. But now another problem arises: how do we become aware that certain acts have the moral properties they do? Do we have some special faculty of moral intuition which makes 'visible', as it were, these moral properties? Does God reveal to us, every time we witness an act, his approval or condemnation of it? In contrast to the strangeness of these ideas is a natural account of moral properties which the atheist can offer, and which answers both the problem of moral knowledge and the problem of the relationship between moral and natural properties in one step. If we accept this natural account, then there is little room for God in ethics.

The account the atheist can give goes as follows. Moral properties are a reflection of our own feelings of approval or revulsion. These feelings may to some extent be innate, but no doubt others are socially conditioned. Leaving that issue on one side for the moment, the reason why moral properties supervene on natural properties is that acts with certain natural properties (e.g. the deliberate taking of human life) tend to cause in us feelings of revulsion, pity, etc., and thus lead us to condemn the act. Saying 'this is wrong' is simply an expression of that feeling, so moral properties are not properties which acts have in addition to their natural properties, and whose connection with natural properties is therefore mysterious. Once we accept this, we can see that there is no problem about moral knowledge. We 'know' that acts with certain natural properties are wrong simply because they cause certain feelings within us. This view is called moral subjectivism.

There are a number of variations on this central idea. The atheist could draw an analogy between moral properties and colours. The colour of an object supervenes on certain facts about its surface structure, in particular on facts about the way in which the (in themselves colourless) atoms are arranged. We might initially be puzzled by this: why should two things identical in terms of the micro-structure of their surfaces have identical colours, when viewed under the same conditions? The accepted answer, of course, is that the arrangement of atoms determines which wavelengths of light hitting the surface are absorbed and which are reflected. This, in turn, causes us to have certain sensations when the reflected light hits our retinas, and in response we attribute a certain colour to the object. This does not mean that colours are just 'in the head': they are genuine properties of the object, but properties of a certain kind, namely dispositions of the object to affect us in certain ways. Exploiting this analogy, the atheist could say that the moral properties of acts were genuine properties of the acts themselves: dispositions that those acts had to affect us in certain ways, to evoke admiration or anger. Alternatively, the atheist could say that moral values were just in the head, and so mind-dependent, and that any moral judgement, such as 'this is wrong', was equivalent in meaning to 'I disapprove of that'. Such a theory of moral judgement is known as 'emotivism'. We need not get into the issue of whether emotivism is the correct account of moral values. The atheist does not need to commit himself to a particular theory about what people really mean when they give utterance to moral judgements. The important point is that he can give an account of what the connection is between the natural properties of an act and the moral property we ascribe to that act, an account which does not make our moral knowledge entirely puzzling.

What role does God have to play in all this? None whatsoever, it seems: moral knowledge and supervenience can apparently be explained without reference to God. However, the account is not actually incompatible with God's playing some explanatory role in ethics, though the role is less significant perhaps than the one originally envisaged. We can make room for God in the atheist's account by assigning to God the job of so constructing us that we respond emotionally in the way we do to certain natural properties of acts. God, as we might put it, is responsible for our moral psychology. This is a somewhat less central role than the one we canvassed at the end of the previous section, that the wrongness of an act was constituted by God's disapproval, but it does show that theism can be made consistent with moral subjectivism. However, alternative and more plausible accounts are available of how we come to respond as we do to the natural properties of acts. Our moral psychology may be the result of biological or social evolution: societies in which certain emotional responses are reinforced may be more likely to survive than other societies, and that is why these responses have become the norm. Again, God is not pushed out altogether, for the theist can point out that the mechanisms of biological and social evolution themselves call for explanation, and God is needed to fix the laws as they are. All this will eventually have repercussions in the development of moral beliefs, but now God is being placed at some distance from the facts which constitute moral values. To retain the connection between God's goodness and the facts which constitute moral goodness, the theist must insist that God does have a moral design for the world, and that this design is reflected in the laws that he makes. He does not directly implant moral notions into people's minds. We must be careful, however, not to reintroduce the notion that God desires the best for us because God is good, for then we are faced with Plato's dilemma all over again.

It seems that we can square theism with different accounts of the relation between moral and natural properties, of how we can have moral knowledge, and of how we come to have the moral psychology that we do, even though we have had to present a) as a consequence of c), rather than a doctrine with significance in its own right. But there are two further features of our conception of morality which are somewhat harder to square with theism, and to those we now turn.

PLURALISM AND AUTONOMY
Many of us live in societies which are pluralist in their political, social, religious and moral outlook: we live side by side with people who have radically different views, and in a genuinely pluralist society the coexistence of these different views is a peaceful one. Indeed, we may not only tolerate pluralism, we may actually welcome it and regard debate between rival ideologies as beneficial to all sides. Even where there is no common ground upon which rival parties can discuss the relative merits of their views, as is the case typically with different religions, it is still possible to see the presence of incommensurable views as an enrichment of one's culture. If we take this view, then we are likely to view the breakdown of understanding and tolerance between different ideologies as deeply unfortunate. Now, if it is possible for a pluralist society to be a stable one, as many people believe, then it is hard to see such a society as simply an intermediate stage in the development of a society with a single ideological outlook. Yet, if God has a moral design for the world, and so construes the world that his ideal will eventually be realised, then a pluralist society can at best be seen as a stage on the way to this goal, and not a desirable end in itself. Further, if God is responsible for our moral psychology, or for the conditions which determine our moral psychology, then many of us, it seems, have developed the wrong moral responses to the natural properties of acts: we condemn as wrong some things which are either right or of no moral significance, and acquiesce in some things which are wrong. Is the mechanism by which we develop our moral views then imperfect? Surely not, or else the universe would simply be a bungled experiment. Can religion help us to form the correct view? Not if a number of equally compelling, but incompatible, religions are on offer.

When we reflect that the coexistence of different moral values within a society may be preferable to an ideologically monolithic one, moral relativism becomes a plausible position. According to relativism, there are no absolute moral truths, which are the same from one context to another, but, rather, there is a given moral judgement that will only be true relative to a particular context, where 'context' means a group of people sharing an outlook and culture. That is to say, moral values are in part constituted by moral attitudes, which will vary from context to context. If relativism is correct, then the theist faces an uncomfortable choice: either God has properties which are good with respect to one society but not with respect to another, or God is morally neutral. Now pluralism does not entail relativism, but they are natural partners.

I want, finally, to consider whether there is room for God to play another explanatory role in ethics, namely the role of providing a rationale for genuinely moral actions.

Why should I refrain from harming others? 'Because God commands it.' What kind of rationale for action is this? If I do something because I believe that God commands it, then I must believe that I should obey God. But here there is a danger of regress: I should obey God because I should obey an infinitely wise, benevolent being, and I should obey an infinitely wise, benevolent being because I should…and so on. To avoid the regress, the theist must justify obedience to God, not by appealing to some other set of obligations, but by showing how God could have a special moral authority. The authority must be special because, in general, if I do something only because someone else requires it, and not because of the features of the act itself, then I am not behaving morally. I may appear to be behaving morally, for example, if I dash into a burning house to save someone lying unconscious on the first floor. But if I turn out to have done this simply because I was told to, and I would have been just as happy to recite 'Twas brillig, and the slithy toves did gyre and gimble in the wabe' if I had been told to do that instead, then my action is not moral. So why should acting in response to God's commands make my action a moral one?

One answer is that acting in response to God's commandments is acting morally because God defines what is good: x is good if and only if God commands it. So to judge that x is good is, ipso facto, to judge that God commands x. This is why obedience to God is a special case: it imparts a moral aspect to my actions that obedience to no other authority could impart. So I do act because of the intrinsic worth of the act. However, there is still reason to think that performing an act because one believes that God commands it gets in the way, as it were, of acting morally.

Compare the two desires: the desire to subordinate oneself utterly to the wishes of some authority, so that everything one does is eventually an unconscious reflection of those wishes; and the desire that one's behaviour should reflect one's own ideals, to act because one thinks it is right, independently of the will of any other individual. Which is the better ideal, as far as our moral development is concerned? The atheist insists that the second desire is the better one. For the atheist the moral ideal is autonomy, or self-government. The truly moral agent is one who wishes to be his own master, not the instrument of some other power, and not to trust the deliverances of some supposed authority, but to work out for themselves the rightness of certain kinds of behaviour. But, if we value autonomy, then we distance God from morality: what God wants will not feature essentially in out deliberations. If it does, we will still want to ask whether God's wishes reflect what we believe is right. The danger of a morality which subordinates the agent's wishes and beliefs to those of an authority is that it can be based on fear of the consequences of transgression. But a morality based on fear is no morality at all.

To value moral autonomy is not necessarily to subscribe to some controversial meta-ethical theory. It is not, for example, to embrace moral scepticism and deny, or doubt, the existence of moral values. It is more likely to go together with an honest attempt to work out moral values. The autonomous agent is no more likely to be amoral than an unquestioning believer. Nor does autonomy necessarily lead one to moral relativism, the view that moral values vary from context to context. The autonomous agent may well believe in the existence of objective moral values. Autonomy would then consist in working out what those values are.

We have not shown that autonomy is inconsistent with theism. After all, God may want us to be free agents, working out our own reasons for doing things and doing them because we want to and because we see them as having intrinsic worth, just as parents want their children ultimately to be self-governing. But if this is God's design for the world, then the consequence is that morality need not make essential reference to God. So this is one more reason to suppose theism to be explanatorily redundant.

SUMMARY
The idea that the existence of moral values depends in some way on God creates difficulties both for the doctrine that God is good and for the doctrine that he wills us to do what is good. The problem is this: we can, apparently, only make sense of these doctrines if we think of goodness as being defined independently of God. But if it is so defined, then God does not, after all, explain the existence of moral values. Thus, theism can be presented as containing an inconsistency. This we called the 'meta-ethical argument' for atheism. It is, however, possible to avoid the contradiction if we argue that God is good precisely because his existence explains the existence of moral value. The goodness of God is thus radically unlike the goodness of ordinary moral agents.

This raises a further problem, however. If God is the basis of moral values, then such values must be objective, and we are, therefore, faced with the following questions: 1) How do we come to be aware of these moral values, if they exist entirely independently of us? 2) Why do moral facts supervene on natural facts? 3) How can the existence of objective moral values be reconciled with the existence of different conceptions of what is right? These difficulties are not faced by the atheist, who can provide the following account of moral knowledge: acts with certain natural properties tend to cause in us feelings of revulsion which, in turn, lead us to describe those acts as wrong. So the 'wrongness' of an act is simply the disposition of the act to cause a feeling of revulsion in us. But our reactions are not entirely biologically programmed: they are, in addition, influenced by our culture, hence the variety of moral systems.

There is, however, more than one way in which God could play an explanatory role in ethics. If we think of his role in these terms, that an act is good only by virtue of the fact that God wills it, then we trivialise the assertion that God wills us to do what is good. However, his role might be a rather more indirect one. The existence of God is quite compatible with the atheist's account of our moral psychology, for God may have caused us to react in the ways the we do to acts with certain natural properties. However, it is much harder to reconcile theism both with the existence of morally pluralistic cultures, and with the view that such cultures are, in a sense, preferable to morally homogeneous cultures.

The final problem we discussed was that of autonomy: if the best kind of moral agent is autonomous, i.e. self-governing, then we should not appeal to the will of God as a motivation for performing the right action. Autonomous morality is, by definition, independent of God, so, in one sense, God is not required as a basis for ethics.