The aim of God's Own Ethics Footnote ¹ is to provide an account of God's ethics and to consider the relevance of that account for how we ought to think about the force of the argument from evil. (By an ‘ethics’ I mean an agent's dispositions to treat certain considerations as favouring or disfavouring actions of different types, and as favouring or disfavouring action of those types in particular ways.)

In raising the question of what God's ethics is, it is helpful to appeal to some conception of God. Among the attractive conceptions of God, there are three that are central to the book's argument: God conceived as an absolutely perfect, or ‘Anselmian’, being, God conceived as that being who is supremely worthy of worship, and God conceived as that being who is wholly worthy of allegiance. I am asking, then, what we should think about God's ethics, in each of these ways of conceiving of God, and what we should think about how the attribution of such ethics to God, conceived in each of these ways, affects how we ought to think about the argument from evil against the existence of God, conceived in each of these ways.

The book has two parts. In part I, I ask: What is the ethics of an Anselmian being? And does this ethics provide the needed motivational premise to get a powerful version of an argument from evil up and running? In part II, I ask: Is an Anselmian being, with the ethics that I ascribe to that being in part I, supremely worthy of worship? And is an Anselmian being, with the ethics that I ascribe to that being in part I, wholly worthy of allegiance? If not, then is it possible for the Anselmian being to take on a contingent ethics that would make that being worthy of such worship and allegiance? And does this contingent ethics provide the needed motivational premise to get a powerful version of an argument from evil up and going?

The main conclusions of part I, which consists of the first six chapters of the book, are as follows. First, the ethics of an Anselmian being is not our ethics, an ethics of familiar welfare-oriented moral goodness. The Anselmian being's ethics is much more discretionary than ours is with respect to promoting the welfare and perfection of creatures: while the increasing and preventing of setbacks to the welfare and perfection of creatures always gives justifying reasons for the Anselmian being to act, they do not give requiring reasons for that being to act.Footnote ² While the Anselmian being can rationally promote the well-being/perfection of creatures, that being needs no reason not to. Second, while the Anselmian being's ethics is much more discretionary than our ethics with respect to promoting the welfare/perfection of creatures, the Anselmian being's ethics is less discretionary in another way: the Anselmian being, I claim, necessarily never intends evils to creatures (where an evil is the absence of a due aspect of welfare or perfection in a creature). And, third, this conception of the Anselmian being's ethics undermines massively the prospects of the argument from evil against the existence of an Anselmian being.

The first chapter focuses on how we should understand the notion of an Anselmian being. To be an Anselmian being, on this account, (a) that being must exhibit each pureFootnote ³ perfection to its limit, with no trade-offs among perfections required, and (b) that being's perfections should be conceived absolutely, rather than defined in relation to logically prior limits of metaphysical possibility. Such a conception of what it is to be an Anselmian being has a methodological upshot, which is that we should conceive the perfections more expansively, and not prioritize judgements of metaphysical possibility over judgements of value.

I claim that the ethics of an Anselmian being is not an ethics that is oriented towards creaturely well-being in the familiar way. On a familiar welfare-oriented ethics, one is disposed to treat setbacks to the well-being of rational (and perhaps all sentient) creatures as something to be prevented, and so to prevent such setbacks unless there exist considerations that make it appropriate not to prevent them. Why might we ascribe this familiar ethics to the Anselmian being? Perhaps because the Anselmian being must be loving in a certain way, or perhaps because that being must be morally perfect. Chapters 2 and 3 are devoted to giving reasons to reject these bases for ascribing the familiar welfare-oriented ethics to the Anselmian being.

In chapter 2, I argue that the Anselmian being's necessary love for creatures does not extend beyond any love that such a being must have in virtue of its moral goodness. Any love for creatures that the Anselmian being exhibits is either entailed by the value of those creatures or not so entailed. Any love that is entailed by the value of the creatures themselves counts as love that is required by moral perfection, and so does not go beyond moral perfection. Any love for creatures that is not so entailed, and so could go beyond it, has to be reason-giving for the Anselmian being, or else the Anselmian being will not be free; and such love will have to be itself a divine perfection, or else we should not ascribe it to the Anselmian being. So if there is love that the Anselmian being must exhibit towards creatures that goes beyond what is entailed by moral perfection, such divine love must be reason-giving and it must be a divine perfection. I reject both of these: (a) love as orectic state is not itself reason-giving, and (b) it is implausible that such love could be a divine perfection, as it lacks an intrinsic maximum, and no perfection of an Anselmian being can lack an intrinsic maximum.

In chapter 3 I argue that the Anselmian being is not morally perfect, at least if the relevant norms of moral perfection include norms that direct the Anselmian being to exhibit a familiar welfare-oriented ethics. For the Anselmian being to be bound by such norms, the Anselmian being must have decisive reasons to comply with them. (This is true both because the decisive reasons–rightness tie is a truth of metaethics, and because it would be contrary to the freedom of the Anselmian being if that being necessarily acts in some way that it does not have decisive reasons to act.) But none of the familiar accounts – Hobbesian, Humean, (non-wild) Kantian, or Aristotelian – of how the welfare of one being gives another agent reasons to promote it apply to the Anselmian being; if any such view is true, then we should deny that the Anselmian being has such reasons. The only accounts that stand a chance are those that appeal to the intrinsic value of creatures. But we have reason to think that the notion of creaturely intrinsic value is at odds with the conception of God as an absolutely perfect being. God's being absolutely perfect includes God's being the source of the value of all beings distinct from itself; things are good in virtue of standing in some relation to the Anselmian being. But any goodness that is exhibited in virtue of standing in a relation to some other being is not intrinsic goodness. So, given the existence of an Anselmian being, creatures must lack intrinsic value.

The argument of chapters 2 and 3 is primarily negative, giving us reasons to deny that the Anselmian being's ethics is our ethics of familiar welfare-oriented moral goodness. In chapters 4 and 5, I present a positive account of the ethics of the Anselmian being.

We know that the Anselmian being is perfect as an agent: the Anselmian being acts perfectly on that being's reasons for action, and the Anselmian being does not act pointlessly. What do we know, though, about the Anselmian being's reasons for action? We know that the Anselmian being creates all that exists that is distinct from that being, and that knowledge gives us a basis to think that the existence and perfection/well-being of creatures gives at least justifying reasons for to promote it, where a justifying reason is something that makes rational the performance of that action. But that does not give us a basis to believe that the existence and perfection/well-being of creatures gives the Anselmian being requiring reasons to promote it, where a requiring reason makes non-performance irrational in the absence of reasons to the contrary. And we have reasons from the perfection of the Anselmian being to expand our conception of that being's discretion, and thus to deny that such reasons are requiring. Chapter 4 offers a model of the goodness of creatures that makes this view intelligible, a classical model in which creaturely goodness is non-intrinsic, a participation in divine goodness.

While the Anselmian being's ethics is much more discretionary than our ethics with respect to promoting the welfare/perfection of creatures, the Anselmian being's ethics is less discretionary in another way: the Anselmian being necessarily never intends evils to creatures. Promotion is only one response to value; respect is another. Though the Anselmian being's reasons to promote creaturely goodness and to prevent setbacks to it are merely justifying, there is a respect response – refraining from intending evils – that the Anselmian being has decisive requiring reasons to perform. (Again, an evil is the loss of a due perfection or aspect of well-being.) The argument for the thesis that the Anselmian being never intends evils is that intention involves making the intended object a success condition on action, and it is incompatible with the holiness of the Anselmian being for the success of its agency to be constituted by evil. The Anselmian being of course allows various evils, and in a later chapter I argue that the Anselmian being can make use of such evils without intending them. But the excellence of the Anselmian being's agency is never constituted by evil, and so the Anselmian being never intends evils.

As I argue in chapter 6, this conception of the Anselmian being's ethics undermines massively the prospects of the argument from evil against the existence of an Anselmian being. The argument from evil of course requires a motivational premise, that God would be motivated to make use of God's omnipotence and omniscience in certain ways, were God to exist. This premise is in standard formulations supplied by an appeal to God's moral goodness, where that moral goodness is understood in the familiar welfare-oriented way. But if the argument thus far is correct, then we cannot appeal to such a motivational premise. The Anselmian being's reasons for looking after our well-being are justifying only. Because the Anselmian being would need no reason not to prevent setbacks to our well-being/perfection or not to provide us with greater well-being/perfection, the argument from evil is a failure against the existence of an Anselmian being. Note that this response to the argument from evil differs from that of the sceptical theists, who assume that God is bound by that familiar morality but add that we do not know what further reasons God has that modifies the application of those norms. My view is based on theses of ethics, not on theses of moral epistemology.

While the argument from evil against the existence of the Anselmian being fails, one might think that the mode of response to that argument that I employ is self-undermining: in showing that the argument from evil is a failure as pressed against the existence of the Anselmian being, I have made use of a conception of the Anselmian being's ethics that undercuts that being's worthiness of our worship and of our allegiance. As being God might plausibly be understood in terms of being supremely worthy of worship or being fully worthy of allegiance, one might object that the Anselmian being has been shown not to qualify as God. Part II deals with this set of concerns.

The main conclusions of part II are as follows. The Anselmian being's not sharing our ethics in no way undermines the necessary worthiness of that being to be worshipped. By contrast, the Anselmian being's not sharing our ethics does undermine the necessary allegiance-worthiness of that being with respect to us. Allegiance-worthiness is a matter of fit between agents’ ends, and this fit is not necessary between us and the Anselmian being. It is possible, though, for the Anselmian being to take on a contingent ethics that makes that being allegiance-worthy; there are various such ethics, and some of them do not mirror our familiar welfare-oriented moral goodness. And some contingent ethics that the Anselmian being could take on are such that the Anselmian being would, while being fully worthy of our allegiance, nevertheless still lack requiring reasons to prevent this-worldly setbacks to our perfection/well-being. Thus the argument from evil against God conceived as that being who is fully worthy of allegiance fails as well.

In chapter 7, I argue that the Anselmian being's not sharing our ethics in no way undermines the worthiness of that being to be worshipped. Worship is expressive action: it expresses the massive superiority in value of the worshipped over the worshipper. Those actions that are appropriately worshipful express the absolute greatness of the object of worship, and thus glorify the worshipped and (self-)abase the worshipper. Given this understanding of the appropriateness of worship, nothing said in part I at all calls into question the notion that the Anselmian being is supremely worthy of worship, for that being's absolute greatness was in no way called into question in part I. (It is important to distinguish questions regarding the worthiness of a being to be worshipped from questions about whether one could rightly undertake an act of worship on some occasion or in some set of circumstances. It is compatible with the view defended in this chapter that although the Anselmian being is fully and unsurpassably worthy of worship, there could be circumstances in which one could not rightly worship that being.)

But the Anselmian being's not sharing our ethics does undermine the necessary allegiance-worthiness of that being with respect to us. The main forms of allegiance are alliance and obedience. And our best accounts of the conditions of appropriate alliance and appropriate obedience involve the parties sharing ends, either ends that the parties do have or ends that they should have. But the Anselmian being's ends may properly vary radically from our own. Thus it could be no more than a contingent matter that the Anselmian being is the proper object of our allegiance.

The basis of allegiance to the Anselmian being is, I argue in chapter 8, the good of ‘religion’: it is a good analogous to the good of friendship, but which involves subordinating one's own will to the will of a divine being. Religion is, however, not a good that is necessarily available to us: it is a contingent matter whether we can rightly align ourselves with the divine being's will. One way that we could rightly do so would be if the Anselmian being contingently took on an ethics of familiar welfare-oriented moral goodness. But this is not the only way. So long as the Anselmian being does not will that we act contrary to the moral law that necessarily binds us and has a will with respect to our action such that we act better on our reasons by following the will of the divine being rather than by acting on our own discretion or on some other proximate standard for conduct, then that great good is available to us. And the Anselmian being can take on such an ethics that exhibits those features, either by binding itself via principles of rational intending or by subjecting itself to the norms of a practice, most saliently by making promises and giving commands.

Must any Anselmian being to whom we owe full allegiance have the sort of ethics, albeit contingently, that supplies the motivational premise for a powerful version of the argument from evil? In chapter 9, I say No. Some of the possibilities for a contingent ethics that the Anselmian being could take on that would make that being fully worthy of allegiance – e.g. an ethics of familiar welfare-oriented moral goodness, contingently adopted by the Anselmian being – would obviously re-establish the argument from evil with whatever force one takes it to have. But not all such contingent ethics re-raise the problem in this way. One salient possibility is that in which the Anselmian being promises each created person that his or her overall good will be secured if he or she subordinates his or her will to the will of the Anselmian being (and, perhaps, also promises each created person that every one of one's fellow created persons will receive this offer as well). Such a view meets the conditions for full allegiance-worthiness without, I say, re-raising the problem of evil. This conception – of a perfect being who has made such covenants with and promises to human beings – is, I claim, a plausible understanding of the Christian God, who did not have to make possible our overall good but nevertheless has indeed committed to make that good available to us.