It is common to reject divine command theories of metaethics on the basis of the Euthyphro problem, the essence of which is that such views render morality unacceptably arbitrary. In what follows, I will argue that when it comes to the Euthyphro problem divine command theories are not uniquely prey to it: all forms of objectivist moral realism are exposed. I will then go on to argue that there is no need for anyone to worry, as the Euthyphro problem is not really a problem at all. Obviously, if I am correct this does not mean that a divine command theory of metaethics is true, just that one prominent reason for rejecting it is unsound. However, I will go on briefly to sketch some reasons why divine command theory deserves to be taken more seriously in contemporary metaethics.
The structure of the paper is as follows. In section 1, I define my terms and in section 2 I outline the problem at the heart of the Euthyphro. In section 3, I explain why the non-naturalist moral realist faces the Euthyphro in the guise of problems with moral supervenience. In section 4, I explain why naturalist moral realism faces the problem in the guise of problems with the non-trivial local moral supervenience thesis. In section 5, I explain why the Euthyphro is not a problem. Finally, in section 6, I briefly outline some of the virtues of a divine command analysis of morality before concluding.
1
Divine command theories of metaethics come in different shapes and sizes. Here I use the term to mean any view that identifies moral properties with the property of being commanded or favoured by a god of some sort.Footnote 2 By ‘a god’ I mean some kind of a supernatural agency possessed of a considerable amount of control or influence over us. Note that on my usage it is not the essence of divine command theories that the god in question be the god of any religion. The god certainly need not be God. True, most contemporary divine command theorists have God in mind and sometimes – often in fact-divine command theory is defined as the view that moral goodness is what God commands (favours, loves, etc.).Footnote 3 But this saddles the view with religious baggage that it need not have and works to make the theory less, not more plausible.Footnote 4 It is also worth remembering that the original Euthyphro objection targeted divine command theories understood in my manner. The Euthyphro dialogue does predate Christianity after all. I should mention that for convenience I will also treat divine command theory as a form of metaethical realism rather than an error theory.Footnote 5
By objectivist moral realism (moral realism hereafter) I mean the view that moral claims are truth apt, that at least some moral claims are true, and that their truth-makers are attitude-independent properties of either a natural or non-natural kind.Footnote 6 So, non-subjective naturalist views according to which moral properties are, or reduce to natural properties qualify, as do non-naturalist views according to which moral properties are sui generis.Footnote 7
2
The Euthyphro problem was first raised by Plato in his dialogue of the same name. It is not my object here to engage in a textual analysis of the actual dialogue. The core problem (whether it is to be found in Plato's dialogue or not) is that if morality is whatever a god favours us doing, then morality is rendered unacceptably arbitrary. As Cudworth put it ‘divers modern theologers do not only seriously, but zealously content… that the arbitrary will and pleasure of God… by its commands and prohibitions, is the first and only rule and measure [of moral value]’.Footnote 8
Why is this kind of arbitrariness a problem? There are, I think, two fundamental concerns. If morality is the commands of a god it becomes metaphysically possible for even the most horrendous deeds to be morally right. For instance, there is a possible world in which Hitler's deeds were right and praiseworthy despite having been performed with the same intentions and having all the same consequences as they did here, for there is a possible world in which the god approves of such deeds. This is supposedly absurd. Call this the horrendous deeds concern.
The other concern is that if two numerically distinct acts that have all the same natural features can differ morally, then the morality of each has become some kind of brute addition. What we are favoured doing one day we may not be favoured doing the next (the god's tastes may change) and thus the wrongness of any deed cannot be said to be due to its possession of various natural features. After all, a numerically distinct act possessing exactly the same features could have a different morality. The act is wrong because the god happens to disfavour it, rather than because, say, it causes someone to suffer or manifests a particular character trait. The morality of a deed has come ungrounded. Call this the grounding concern.
Both concerns are confused and misguided. However, before addressing them directly I will seek the sympathies of a wider audience by showing that these self-same concerns can be raised about all forms of moral realism.
3
Non-naturalist moral realists hold that morality cannot be naturalistically explained. Moral properties are sui generis. However, if moral properties are not reducible, then there seems no principled way of ruling out the possibility of a naturalistically identical world to this one in which moral properties are differently arranged. So, there could be a naturalistically identical world in which Hitler's deeds have the property of being right and praiseworthy (the non-naturalist faces the horrendous deeds and grounding concerns). This is a well-recognised problem for the non-naturalist. Tellingly, however, it is not referred to as the Euthyphro. Rather, it is expressed in terms of the non-naturalist having problems explaining the moral supervenience thesis.
To clarify: the global version of the moral supervenience thesis says that if two worlds are identical in all of their natural properties, then they must be identical in any moral properties they have as well.Footnote 9 This is supposedly a conceptual truth about morality. Non-naturalism does not entail the truth of this thesis and so seems incapable of explaining why it is necessarily true. And yet, as is often pointed out, denying it would mean moral properties would be brute additions and that it would be metaphysically possible for horrendous deeds to be morally right. Here is Michael Ridge expressing both the horrendous deeds concern and the grounding concern:
(S) [the global moral supervenience thesis] is extremely plausible, to the point that someone who denied it would thereby betray incompetence with normative concepts. To deny (S) would be to allow, for all that has been said so far, that it could have been the case that the world was exactly like the actual world in all of its non-normative and descriptive features, yet Hitler's actions were not wrong. Since all the non-normative and descriptive facts are the same in this possible world it will still be true that Hitler killed the same people, had the same intentions, etc. Such bare normative differences seem inconceivable.Footnote 10
So, on the face of it non-naturalism seems to render morality arbitrary every bit as much as divine command theory. Interestingly, however, this ‘problem’ for the non-naturalist is seen as just that: a bit of a problem. It is not seen as a decisive refutation of the view.Footnote 11
We seem to have a double standard in operation. Non-naturalist views are recognised to have a significant problem – no question of that – but we do not find them summarily dismissed on this basis. Divine command theories on the other hand share the same problem, but are considered decisively refuted by it. Perhaps this is because of a general queasiness about positing gods, a queasiness that is not felt to the same degree about positing a sui generis moral landscape. But then we are just dealing with a prejudice aren't we? For in terms of, say, ontological extravagance there seems little difference between the two views. Of course, non-naturalists sometimes claim that their non-naturalism does not involve positing some extra kind of thing or dimension to the universe. Saying something does not make it so, however. Perhaps the non-naturalist is not committed to substance dualism. But they must be positing something extra or they are in no way distinct from the naturalist. As Finlay puts it:
Non-naturalism has to involve metaphysical commitments if it is to distinguish itself from semantic and ontological antirealism. It is puzzling why non-naturalists would think this antimetaphysical stance is consistent, or why, given their realism, they would wish to avoid metaphysical claims.Footnote 12
Though as Finlay then goes on to suggest ‘it may be that they are conflating mild, merely metaphysical commitments (to a domain of properties) with radical, supernaturalist commitments to other kinds of self-subsistent substance’.Footnote 13 Would positing a god be any less ‘mild’ than positing a domain of irreducible moral properties? Consider: a god need be no more metaphysically queer or extravagant than, say, a minotaur or Bigfoot. As Bedke puts it:
Bigfoot, if he were to exist, would be metaphysically naturalizable, and his discovery would not call for an expansion of the kinds of entities, properties or relations that we have come to recognize in the world. His kind would be more of the same, for presumably he would be a flesh and blood creature with a biological history, at base composed by the entities recognized by our best physics.Footnote 14
Perhaps some will deny that a naturalizable creature can be a god, but that is both false (the ancient Greek gods lived at the top of mount Olympus and if they existed would have been odd, but naturalizable creatures – we still refer to them as gods) and irrelevant. Just call it ‘strange powerful creature command theory’ then. It is ‘strange, powerful creature command theory’ to which the Euthyphro objection was first offered, after all.
Perhaps non-naturalism is not considered decisively refuted by its exposure to the Euthyphro because it is thought to have the resources to be able to deal with it, even if no one has quite figured out how yet. But ahead of careful debate it is hard to see why anyone should assume non-naturalism has any greater resources available to it than a divine command theory. And when one considers how the non-naturalist deals with their version of the Euthyphro their attempts so far seem either to fail or seem equally available to a divine command theorist. For instance, Zangwill attempts to deal with non-naturalism's Euthyphro by denying that the global moral supervenience thesis needs to be explained.Footnote 15 If that is a viable way of dealing with the problem it is equally open to the divine command theorist to refuse to explain things. What is good for the goose is good for the gander.
Russ Shafer-Landau attempts a non-naturalist explanation of the moral supervenience thesis. He says this:
According to the sort of ethical non-naturalism that I favour, a moral fact supervenes on a particular concatenation of descriptive facts just because these facts realize the moral property in question. Moral facts necessarily covary with descriptive ones because moral properties are always realized exclusively by descriptive ones. Just as facts about a pencil's qualities are fixed by facts about its material constitution… moral facts are fixed and constituted by their descriptive constituents.Footnote 16
But as Ridge points out:
Crucial to the success of this strategy is the idea that constitution is distinct from identity. Otherwise, the view on offer would collapse into a form of reductionism, with normative properties being identical with certain constellations of descriptive properties. In itself this is not a problem, since constitution by is indeed distinct from identity with…. A clay statue is constituted by the lump of clay, but the lump of clay is not constituted by the statue; the clay is instead constituted by certain molecules, which themselves are not constituted by the lump of clay, but by various other subatomic particles and so on.Footnote 17
It is vitally important to note that in this analogy the statue is analogous to the wrongness, and the clay is the natural property or properties on which the moral property is supervening (after all, we don't want to say that the property of ‘being a statue’ is a non-natural property, for that would make Shafer-Landau's view indistinguishable from naturalism, so ‘being a statue’ is analogous to the non-natural property of wrongness instead). Note that if I take a hammer to the statue it will cease to be a statue, but it will remain a lump of clay. In other words, there can be two possible worlds that are alike in terms of their clay content, but unalike in terms of statue content (there cannot be two naturalistically identical worlds in which the lump of clay composes a statue in one but not in the other, but that's irrelevant). By analogy then, there can be two possible worlds alike in terms of their natural properties, but that differ in their moral ones. Shafer-Landau's attempt to show how non-naturalism can entail the moral supervenience thesis seems to fail, then.Footnote 18
4
Metaethical naturalist realism is the view that moral properties reduce to, or just are, natural properties. It is common to see the moral supervenience thesis as providing a powerful source of support for such views.Footnote 19 For unlike non-naturalist views, naturalism entails the truth of the global moral supervenience thesis. If moral properties just are, or reduce to, natural properties, then it is trivially true that two worlds that are identical in their natural properties must be identical in their moral properties as well. If Hitler's deeds are wrong in the actual world, there is no naturalistically identical possible world in which they are right. In other words, naturalist views are widely thought not to be exposed to the Euthyphro.
Matters are not so straightforward. It is true that naturalist views entail the global moral supervenience thesis. However, what often goes unnoticed is that they do not entail the non-trivial local moral supervenience thesis and yet exactly the same reasons for thinking the global moral supervenience thesis is a conceptual truth (namely the horrendous deeds concern and the grounding concern) speak in favour of the non-trivial local moral supervenience thesis being one as well. In other words, if you cannot vindicate the non-trivial local moral supervenience thesis, then you face the Euthyphro. And naturalism cannot explain the non-trivial local moral supervenience thesis and so therefore faces the Euthyphro.
Let us first distinguish the non-trivial local moral supervenience thesis from the trivial local moral supervenience thesis. According to the trivial local moral supervenience thesis it is impossible for acts within a possible world to differ in terms of their moral properties without also differing in some natural respect. This thesis is trivially true because the subvening base includes an act's spatio-temporal properties. Spatio-temporal properties are natural properties. However, there can never be two acts that share all the same natural properties within a possible world. The thesis is therefore trivially true. For it amounts to nothing more than the thesis that one and the same token action has the same moral properties that it has. No one has any difficulty explaining why one and the same token act has the moral properties it has. So, although naturalism entails the trivial local moral supervenience thesis, so does non-naturalism and divine command theory. It is a trivial thesis, after all.
The non-trivial local moral supervenience thesis excludes spatio-temporal properties from the subvening base. According to this thesis, it is impossible for two acts that differ spatio-temporally and in no other natural respect, to differ morally. In other words, if act a1 is morally wrong today/here, then a numerically distinct but otherwise naturalistically identical act, a2, cannot be right tomorrow/over there. Now note that unless one affirms the non-trivial local thesis one will have to allow that it could be the case that in the future or in some different location there could be an otherwise identical Hitler performing otherwise identical acts as those performed by the 1930s and 40s Hitler and these acts could be right and praiseworthy. Alternatively, just imagine there is a mirror solar system to this one on the other side of the universe and that all the same acts happen there as happen here. Deny the local non-trivial moral supervenience thesis and you will have to allow that mirror-Hitler's acts over there could be morally right and praiseworthy consistent with our Hitler's acts being wrong and blameworthy. If mirror-Hitler's deeds are morally right despite being performed with the same intentions and having the same consequences then (supposedly) we cannot say that the Hitler of the 1930s and 40s deeds were wrong in virtue of their intentions and consequences etc. In other words, deny the non-trivial local moral supervenience thesis and you face squarely the horrendous deeds concern and the grounding concern: you face the Euthyphro.
Naturalist reductionist views do not entail the non-trivial local moral supervenience thesis and so cannot explain it. All metaethical naturalist views entail that it is impossible for two acts to differ morally without differing naturally. Yet if two acts differ in terms of their spatio-temporal properties they do differ in some natural respect because spatio-temporal properties are among an act's natural properties. The sad fact is the naturalist does not have a principled explanation of why spatio-temporal properties should be excluded.Footnote 20 It is no good just stipulating that spatio-temporal properties don't count. Anyone can stipulate a problem away. A divine command theorist can just stipulate that the god's will does not vary, and a non-naturalist can just stipulate that moral properties are necessarily where they are. It is no good appealing to the supposed absurdity of allowing that an act it would be wrong to perform here, might be right to over there other things being equal. For that is to do no more than to appeal to the intuitive appeal of the non-trivial local moral supervenience thesis. If the reductionist is allowed to make such appeals then the same privilege must be extended to the non-naturalist and the divine command theorist. Similarly, it is no good appealing to some normative moral theory that does not allow spatio-temporal properties to make a difference, such as utilitarianism. Such theories simply presuppose that spatio-temporal properties cannot make a difference rather than explaining why they do not. Nothing stops the divine command theorist from just helping themselves to the supposed truth of utilitarianism (or whatever) and concluding that the god of morality always and everywhere commands us to maximise happiness; nothing stops the non-naturalist from doing the same.
The simple fact is that when it comes to the non-trivial local moral supervenience thesis the naturalist is no better placed to explain it than the metaethical non-naturalist or the divine command theorist. If you cannot explain the non-trivial local moral supervenience thesis you face the Euthyphro. Naturalism therefore faces the Euthyphro.
5
A problem shared is still a problem. However, it does mean more people have an interest in a solution. Happily the solution to the Euthyphro is easy. Take horrendous deeds. Aside from the concern about grounding (which we will deal with separately) why is it absurd to allow that it is metaphysically possible for there to be a naturalistically identical world to the actual world but in which horrendous deeds are morally right? Presumably it is absurd because we have the very clear and distinct intuition that such acts are wrong, an intuition so clear and distinct it is simply inconceivable that it might be mistaken. Not all of our moral intuitions are like this, but that is precisely why horrendous deeds are singled out: we have incredibly strong intuitions about what, say, Hitler did, and it just beggars belief to suppose that intuitions of this strength might not have something answering to them in reality. Intuitions of a certain strength cannot exist absent the object of awareness, in other words.
If that is true, then we have our solution to the horrendous deeds problem. Our moral intuitions are mental states and mental states are natural properties. Thus, in all naturalistically identical possible worlds to this one we – or rather, out counterparts there – have the same moral intuitions, and thus will also have the strong moral intuition that horrendous deeds, such as Hitler's, are wrong there. By hypothesis it is impossible for intuitions of this clarity and distinctiveness to fail to have something answering to them in reality, and thus it follows that in all those possible worlds there will be something answering to those intuitions, be it a god's disfavouring, or a non-natural moral property. In all naturalistically identical possible worlds to this one Hitler's deeds are clearly and distinctly intuited to be wrong, and so are wrong. The god disfavours them there (or they have the non-natural moral property of wrongness), for if she did not such intuitions would not (could not) exist.
Of course, there will be those who reject the idea that certain moral intuitions cannot exist absent the object of awareness. They will reject what I have said above and will maintain that it is metaphysically possible for clear and distinct intuitions to fail to have anything answering to them in reality. But by hypothesis those who hold that powerful moral intuitions can fail to have anything answering to them in reality can hardly at the same time insist that it is inconceivable that horrendous deeds might be morally right. For on what basis could they do so? And so someone like that cannot run the horrendous deeds objection. To put it another way, if you think the horrendous deeds objection is a good one, then you have your solution to it: it is a good objection only because it is inconceivable that our intuitions about such deeds could fail to have something answering to them. But if that is so, then there are no naturalistically identical possible worlds in which such deeds are anything other than wrong. On the other hand, if you think that it is possible for intuitions of the clarity and distinctiveness of our intuitions about horrendous deeds to fail to have anything answering to them, then aside from concerns about grounding you cannot consistently reject a metaethical theory for allowing there to be a possible world in which such intuitions fail to have something answering to them. For what possible objection can you have to that? The horrendous deeds concern is therefore either no problem at all, or it contains the seeds of its own solution.
What about the grounding concern? I maintain that the grounding concern is confused. It is enough that some of a token act's natural features have given rise to its moral properties on that particular occasion for it to be true to say that the act is wrong in virtue of those features. There is no need for them to have done so necessarily. Take me. Sometimes I like a curry because of its spicy heat. Sometimes I dislike a curry because of its spicy heat. I'm changeable like that. I like curry one day, and dislike it the next. Does it follow from this that my like or dislike of curry has nothing to do with any of its natural properties? No, when I dislike curry I dislike it because of its spicy heat. That's a natural property of the curry I am eating. It is why I dislike it. I dislike it because of its spicy heat. When I like curry, I like it because of its spicy heat. That's the same natural property of the curry. Nevertheless, when I like the curry I like it because of its spicy heat. My like or dislike of curry is grounded in its natural features. Yet I like it one day and not the next. It would be absurd for someone to insist that if on a particular occasion I dislike curry because of its spicy heat that I must now dislike curries hereafter, and dislike them because of their spicy heat hereafter. It would be absurd to insist that unless that is the case I cannot truly say on any given occasion that ‘I like this curry because of its spicy heat’ or ‘I dislike this curry because of its spicy heat’.
If moral norms are the commands of a god then it seems clear that this god dislikes it when we hurt others for fun. Such acts are wrong, then, because they involve hurting others for fun. That natural feature of such acts that is inspiring the god of morality to disfavour us performing them, and her disfavourings just are the wrongness of such deeds. To the question ‘why do I dislike curry today’ comes the answer ‘because of its spicy heat’. To the question ‘why does morality condemn hurting others for fun’ comes the answer ‘because such acts cause others to suffer’. The fact that tomorrow morality might wish us to hurt others for fun, and may even wish us to because of the suffering such deeds cause, does not alter the fact that today such deeds are wrong, and wrong because of the suffering they cause. All of the same applies, with small adjustment, if instead of a divine command view of ethics one endorses a non-naturalist or naturalist metaethics. For example, if the natural properties of an act cause or in some way produce the moral property of wrongness one day, but the property of rightness the next, it remains true that on that day when they produced wrongness the act was wrong because of its natural properties, and on that day when they produced rightness the act was right because of its natural properties.
The often made criticism of divine command theories that we would have to say that an act is wrong just ‘because a god disapproves of it’ is false, or at least grossly misleading. The same objection could be made against the non-naturalist. To the question ‘why is this act wrong’ could come the answer ‘because it has the non-natural property of wrongness’. But that is like answering the question ‘why are my shoes full of water?’ with ‘because your shoes are full of H2O’. However, normally when we ask ‘why are my shoes full of water?’ we are not inquiring what the water that is in the shoes is made of, but how the water has come to be there. Similarly, when we ask ‘why is this act wrong?’ we are not asking what the wrongness is made of, but rather what it is about the act that lead to it having this property. The answer to that question is not ‘because it has the non-natural property of wrongness’ or ‘because a god disapproves of it’ but will be something like ‘because it causes someone to suffer’ or ‘because it breaks a promise’. That on another occasion an otherwise identical act might not have the property of wrongness does not undermine the veracity of the earlier explanation.
Although what I have said above is, I think, sufficient to show that the grounding objection is misguided I want to say something else. It is often suggested that moral properties can never be just brute additions. Yet that just does not ring true. Yes, I grant that often it seems an act is wrong because of its possession of certain features. But sometimes an act just seems wrong and, well, that's that (and of course, on some occasions an act seems wrong and it is unclear what natural features of the act are responsible for having got it this way). To the question ‘why is Xing wrong’ comes the answer ‘because it just is’. Could not that answer sometimes be true? In other words, if we pay attention to the actual moral appearances it seems – seems to me anyway – that sometimes the morality of a deed really is just a brute addition. Not always, often more of a story can be told – patterns can be detected. Morality has discernible tastes, so to speak. But morality has brute preferences too.
I do worry that there are a set of mantras about morality that are being taken by most to be beyond question yet operate to blind us to the moral appearances. For instance, consider the mantra that if two numerically distinct acts are otherwise naturalistically identical they must be morally identical as well. This commits us to a morality that is fixed across time and space. And thus any moral landscape that there may be has a uniform and unchanging nature. Yet as our undergraduate students like so often to point out to us, different things have appeared right and wrong to different people at different times and places. Taken at face value this is prima facie evidence of a changing moral landscape, not a fixed one. Consider: if painters from one age produce paintings of a certain hill with trees on it, and painters from another paint pictures showing the hill without trees on it we have prima facie evidence that the hill had trees on it at one time and no trees on it at another. We could not draw this conclusion if we held it to be a conceptual truth about hills that whatever they have growing on them now, they must always have had growing on them. We would have to come up with complicated diffusing explanations of the differing representations of the hill. But why hold such a bizarre thesis to be true? When it comes to the one about morality the support comes in the form of the grounding and horrendous deeds concerns. But as I hope I have shown these concerns are confused. If I am correct contemporary metaethics is blind to some moral appearances. Look at the moral landscape. Morality does appear to be variable. And even if it does not – even if the diffusing explanations of varying moral intuitions over time are the better explanations of them – a metaethical theory should not be faulted for having the resources for being able to accommodate varying intuitions were any to occur. Divine command theories of metaethics do not entail that morality changes over time, they merely allow that it can do so. This is a virtue, not a vice. All other forms of moral realism allow it as well. Again, this is a virtue rather than a vice. In short, then, all moral realists face the Euthyphro, but it is not really a problem. Far from it, it is the product of silly dogmas and sloppy thinking.
6
Nothing I have argued above suggests divine command theories of metaethics are true. Rather, if my arguments go through I have merely shown that the most common reason for rejecting such theories is unsound. Yet my object in this paper is to get divine command theories of ethics taken more seriously in contemporary metaethics. To that end let me briefly outline why I think such views are so plausible. And I want to stress, I have no religious commitments and never have had. My desire to promote divine command theory derives entirely from a sincere (perhaps misguided) belief that it does a good job accounting for morality's core features.
Morality appears to be normative and objective. That is to say: it directs us, and those directives do not appear to be directives we are issuing to ourselves and others. A divine command theory of metaethics seems well placed to be able to accommodate these normative and objectivist pretentions. For one might credibly suppose that only an agent, someone minded, can be in the prescribing, commanding and favouring business. Directives, favourings, wishes – these things are all essentially mental. The chair I am sat in cannot bid me sit in it. It does not have a mind. And the same holds true for the natural features of any act or state of affairs. The fact an act hurt someone cannot itself be the disfavouring of me performing it. It may give rise to a disfavouring of the act being performed, but it cannot itself be the disfavouring. The non-agential natural world cannot bid us do things. No good trying to pass the normative buck to normative reasons, for exactly the same concerns apply to those. A normative reason is a favouring of doing (or believing) something, and so the same point can be made: normative reasons do things only a person could do: they favour us doing things. There is a normative dimension to the universe. But the universe as revealed to us by the natural sciences is not in the favouring and directing business. So much the worse for naturalism then. The world as presented to us by the natural sciences cannot accommodate the normative dimension.
Non-naturalist metaethical views seem to fair no better. For the idea of an extra-mental non-natural favouring just wafting about all by itself seems unintelligible. Labelling such things sui generis normative properties does not help. Korsgaard says ‘that is the problem with realism: it refuses to answer the normative question’.Footnote 21 She is too generous. We are not getting silence from the non-naturalist, we are getting a landscape made of extra mental favourings. The problem is that landscapes are dead, lifeless things that can no more care about what we do or what happens than a pebble can. That seems as true of a non-natural landscape as of a natural one. In short, then, morality seems to do things only an agent can do: it favours and instructs. Thus, to make sense of the apparent normativity of moral norms we must at least suppose them to be the directives and favourings of some agency or other.
When we wonder what we are morally commanded to do, we are not wondering what we command ourselves to do. We seem, rather, to be trying to make ourselves sensitive to something outside. Moral norms have objective pretensions. In addition, when I wonder what morality bids me, you, and someone who lived a thousand years ago do, it is the same morality – the same source of biddings – I am talking about, and that I take everyone else to be talking about (the same being true of normative reasons in general). As such, the evidence is against ourselves being the common unifying source of moral commands (and/or normative reasons). If we are to respect appearances the source of morality's favourings and directives must be a single agent who has existed across all time. And our moral sense, or faculty of intuition is a sense – albeit not entirely accurate – of what it is that this creature favours us doing and being. On this analysis, she is the voice of our conscience, a voice most of us find it impossible to turn off or ignore entirely. If all that is correct, and if she exists – so, (if our moral impressions have something answering to them in reality and do not constitute some kind of hallucination), then she has considerable influence over us all, or at least over all of us who have fairly reliable consciences. It seems no abuse of the term to describe a creature like this as a kind of god. Thus, we arrive at a divine command theory of metaethics. Not through religious conviction, but a disinterested concern to account for morality's core features.
Of course, I am not for one moment suggesting that there are no problems facing a divine command metaethics. The above was a mere sketch designed merely to showcase some of the virtues of the view. Concerns remain. For instance, the view saddles morality with an extravagant ontology (though as I argued earlier, arguably no more extravagant than that proposed by non-naturalism). However, to think important concepts won't come with extravagant ontologies seems more like wishful thinking than anything else. When I read a version of this paper at a conference an audience member reported that the best objection to what I had just argued was that no gods exist. Yet that seems presumptuous. If the best analysis of normative reasons in general turns out to be a divine command one (according to which Reason – the unifying source of all normative reasons – is an agency of some kind) then surely we would have very strong grounds for concluding that such a god exists? For normative reasons certainly appear to exist – (all intellectual inquiry presupposes them) and there cannot be any reason to doubt such appearances (not wholesale anyway) for any such doubt would have to be a normative reason (or a consideration that gives rise to a normative reason if one prefers). A normative reason surely cannot be evidence that normative reasons do not exist anymore than a live rabbit can be evidence that live rabbits do not exist. And thus if a divine command analysis of normative reasons turns out to be best, there seems to be nothing for it but to conclude that the god in question exists. No good appealing to an apparently godless scientific worldview, for all of science presupposes the existence of normative reasons.
7
Divine command theories of metaethics are routinely dismissed because they face the Euthyphro problem. However, all forms of metaethical objectivist realism face the Euthyphro. A problem shared is still a problem. Yet on reflection the Euthyphro problem is not actually a problem at all. I conclude not that divine command theory is true – (though I have tried, albeit very briefly, to sketch some reasons why one might think that it is), but that it at least deserves to be taken more seriously in contemporary metaethics.
