The Moral Argument for God

In his dialogue with Euthyphro, Socrates asks, “Is the pious loved by the gods because it is pious? Or is it pious because it is loved by the gods?” To put the dilemma in more modern terms: does God command an action because it is good, or is an action good because God commands it? If the former is true, then the good exists independently of God. If the latter is true, then a question arises concerning the arbitrariness of moral obligations. There seems to be nothing especially moral about obligations that are determined by sheer will, for God could command anything – genocide, torture, rape, and so on – and it would be good simply by the fact of his commanding it.

Recently, a number of philosophers and theologians have attempted to undermine the dilemma by claiming that God’s will, far from being arbitrary, issues forth from his perfectly good nature. One such philosopher is William Lane Craig, who articulates a revised version of Divine Command Theory as part of an argument for the existence of God. Without God, the claim goes, there is no foundation for morality. Ethics is illusory. The existence of objective moral values and obligations constitutes evidence for God.

While some atheist philosophers have upheld these claims, perhaps the most notable being J.L. Mackie,1 this article will focus instead on whether there is any reason to believe that moral realism requires the existence of God. I contend that there is no such reason, and I will endeavor to argue for it by first considering the theistic meta-ethical theory advocated by Craig, offering objections, and then discussing an alternative, secular meta-ethical theory.

I. Divine Command Theory

In his many debates and writings, William Lane Craig presents the moral argument for God as follows:

  1. If God does not exist, then objective moral values do not exist.
  2. Objective moral values do exist.
  3. Therefore, God exists.

Premise 1 suggests two things: first, that atheism cannot provide a foundation for objective moral values, and second, that theism can provide such a foundation. The bulk of Craig’s essay, “The Indispensability of Theological Meta-ethical Foundations for Morality,” supports the first contention by drawing on the claims of atheistic nihilists and arguing that human existence is valueless on the naturalistic worldview. In support of the second contention, much less is provided, but Craig does sketch a brief account of how objective moral values are allegedly grounded in God:

On the theistic view, objective moral values are rooted in God. God’s own holy and perfectly good nature supplies the absolute standard against which all actions and decisions are measured. God’s moral nature is what Plato called the “Good.” He is the locus and source of moral value. He is by nature loving, generous, just, faithful, kind, and so forth.2

Often two analogies are employed to try and illustrate this proposed relationship between God and morality. One analogy has it that the claim ‘God is good’ is like the claim ‘water is H2O.’ The word ‘water’ had meaning before the discovery of its chemical composition, and it was obviously not understood as H2O. Nevertheless, it has all along been true, as chemists eventually learned, that water is dihydrogen monoxide. Another analogy looks to the meter bar in Paris, by which the standard length of a meter is measured. Both these examples are meant to give some idea of how God serves as the standard of moral goodness. Thus, an important facet of Craig’s view is that goodness equates to a resemblance to the nature of God, i.e. generosity is good because God is generous.

In answer to the arbitrariness objection, the Divine Command Theorist will say that because God’s nature is the ultimate standard of goodness, and love is part of the divine nature, it could not be the case that God would command an unloving act like rape. God can no more command that an immoral act be good than he can create a rock too heavy for himself to lift. To do so would be to contradict his own nature.

One might be tempted to ask in response how we know God’s nature or obtain moral knowledge. However, this revised form of Divine Command Theory focuses on ontology rather than on epistemology. Craig is not concerned with moral knowledge, he is instead concerned with providing a sound metaphysical foundation or ground for morality. If there is no God, he wants to argue, there is fundamentally nothing to make us good, nothing constituting moral values or moral obligations.

Defending the 2nd premise, Craig appeals to our moral intuitions. “Some things,” he says, “are really wrong,” and there is “no more reason to deny the objective reality of moral values than the objective reality of the physical world.”3 Many philosophers have conceded that solipsism and external world skepticism have not been convincingly refuted. Nevertheless, we do seem to presuppose that the physical world is a reality, as we also seem to intuitively grasp that other people have minds in some sense like our own. We should trust our moral intuitions, Craig suggests. The secular moralist must either acknowledge that God is the ground of these objective moral values or she must abandon morality altogether.

II. Problems with Divine Command Theory

What can it mean to say that generosity is only good because God is generous? What it does not appear to mean is that generosity is ever good for its own sake. If that is the case, generosity is good independent of God. It is not so much, then, that generosity is good as it is that God is good, and because God so happens to be generous, generosity is therefore good. This puts the Divine Command Theorist in the awkward position of having to affirm the goodness of God before identifying any of the virtues or traits of God as being good. But then what exactly is it about God possessing a certain trait that makes that trait good?

In his paper, “God and the Ontological Foundation of Morality,” philosopher Wes Morriston draws this problem out further, asking

Why is being like God the standard of moral goodness? For this idea to have content or plausibility, it must be spelled out in terms of the characteristics that are included in God’s moral nature. Perhaps the following formulation would get the job done.

To the degree that anyone resembles God with respect to love, generosity, justice, faithfulness, kindness, and so forth, that person is morally good.

The trouble is that this makes it look as if love and generosity and justice and the rest are doing all the work in the proffered account of moral goodness, leaving God no significant role to play.4

Divine Command Theory inverts the way that many theists seem to conceive of God’s good nature. God, to their minds, is good because he is just, loving, faithful, etc. In fact, Craig elsewhere suggests that the reason why God’s nature constitutes the good is because he is, by definition, “worthy of worship,” and only a morally perfect being can be so worthy.5 This can be rather difficult to digest. We are good only to the extent that we resemble the character traits of God, and we are encouraged to understand the perfect goodness of God in light of the character traits he possesses, but God is not actually good because of these traits.

Morriston notes above that what we ought to look to are the traits or characteristics of God’s moral nature. This is important because we would not want to think that all of God’s traits are good by virtue of his possessing them, for some of those traits are non-moral. Omnipresence is not a moral trait like generosity is a moral trait, so it would not make much sense to suppose that omnipresence is good because God has it. But without further explanation as to how God possessing certain traits makes them good, there does not seem to be a useful way of distinguishing between God’s moral and non-moral traits. What makes God’s nature any kind of moral standard is the central question Divine Command Theory must address.

If God’s various moral traits are not what make him good, how can we be certain that he would not command anyone to commit murder, genocide, rape, and so forth? One might well assume that God is all-good and still believe that there are justifying reasons for why God commands horrific acts. Walter Sinnott-Armstrong observes that, “we cannot know that [God] would not command rape unless we assume that rape is immoral for some independent reason. But if there is such an independent reason against rape, then that reason is what makes rape morally wrong, and the command itself is superfluous.”6 Though it has been argued that God’s morally perfect nature prevents him from commanding rape, something of the opposite actually appears to be more at home with Divine Command Theory: a morally perfect being will have morally sufficient reasons for its commands. Of course, this successfully rebuts the arbitrariness objection, but at the fairly significant cost of granting God justifying reasons for any command, and in this respect the assertion that “God is good” seems rather empty.

The water and meter bar analogies highlight the questionable nature of the supposition that God grounds, or serves as the foundation of, morality. The measurement of a meter is an arbitrary designation, a convention reached by agreement among societies and cultures. The point may be that the ‘meterness’ of an object just is the degree to which it measures up to the meter bar, like goodness just is the degree to which something measures up to God, but an end of explanation is reached in both these cases. It cannot be pressed any further for why the meter bar is the standard of meterness, or why God is the standard of goodness. As seen above, Craig does suggest that God’s good nature be understood in terms of the fact that he is “worthy of worship.” Worthiness, however, is an evaluative concept, and supporting the claim that God is the Good, the source of objective moral value, with a further value statement is not particularly helpful.

The same sort of problem arises with the grounding of moral duties and obligations. Craig writes:

Someone might demand, “Why are we obligated to do something just because it is commanded by God?” The answer to that question comes, I think, by reflecting on the nature of moral duty. Duty arises in response to an imperative from a competent authority. For example, if some random person were to tell me to pull my car over, I would have absolutely no legal obligation to do so. But if a policeman were to issue such a command, I’d have a legal obligation to obey. The difference in the two cases lies in the persons who issued the commands: one is qualified to do so, while the other is not.

Now, similarly, in the case of moral obligations, these arise as a result of imperatives issued by a competent authority. And in virtue of being the Good, God is uniquely qualified to issues [sic] such commands as expressions of His nature.7

The existence of God gives us objective moral duties because, according to Craig, God is the right kind of authority. What makes God a “competent authority,” or what makes him worthy of worship? God just is qualified for the job. The water analogy comes to an end of explanation, too, and it would be inaccurate to say that H2O is the foundation of water when the concept “water” had meaning long before the discovery of its chemical composition. Water just is H2O. These are all identity claims, and none of them actually ground anything in an external foundation – they are brute facts rather than foundations for ethics, water, or meterness. But if this is right, why should we think that the brute ethical facts at the heart of Divine Command Theory are any more plausible than other brute ethical facts?

III. Moral Realism Without God

Moral philosopher Erik Wielenberg has defended a non-theistic, non-natural alternative to Divine Command Theory that sees moral properties as supervening on non-moral properties.8 There are ethical facts, on his view, and these facts are not analytically reducible to natural facts (that is, they are non-natural), nor are they reducible to supernatural facts involving God (that is, they are non-theistic). Some ethical facts are metaphysically necessary brute facts, requiring no foundation, much the same as the brute facts mentioned in the preceding section. These brute ethical facts inform other, non-brute ethical facts, and may be seen as the axioms of morality.

One strength of this alternative is its simplicity. Goodness supervenes on generosity, love, justice, and other virtues, such that they are intrinsically valuable. Love is good for its own sake, rather than because it is instantiated to the perfect degree in the perfect being. Generosity and justice are comprehensible to us, whereas the notion that the good is a person, and all his good traits are only good because they belong to him, borders on the unintelligible. When the good is identified with a perfect intelligent agent, the amount and severity of certain evils and suffering in the world becomes anomalous as well – a problem not faced by non-theistic, non-natural moral realism.

The primary objection to this alternative appears to be that its values are groundless. Craig writes,

What does it mean to say, for example, that the moral value Justice simply exists? I don’t know what that means. I understand what it is for a person to be just; but I draw a complete blank when it is said that, in the absence of any people, Justice itself exists. Moral values seem to exist as properties of persons, not as abstractions – or at any rate, I don’t know what it is for a moral value to exist as an abstraction. Atheistic moral realists seem to lack any adequate foundation in reality for moral values, but just leave them floating in an unintelligible way.9

The view articulated by Wielenberg is not atheistic in that it demands any sort of commitment to atheism. Similar meta-ethical conclusions have been endorsed by Christian philosophers, including Wes Morriston, who defends it in his aforementioned paper. Non-natural, non-theistic moral realism is more appropriately called non-theistic because it rests on no assumptions about a deity or about supernatural facts.

Does this alternative theory result in “floating” and “unintelligible” values? If the objection is that it provides no external foundation of morality, this must be granted. However, we have already seen that the theistic ethicist is in no better standing, as the identity claim that God is good is itself not grounded in anything outside of God. Along the same lines, Wielenberg points out that the Divine Command Theory relied upon by Craig posits at least one value, the good, that exists not as a property of a person, but as an actual person. Nonetheless, justice is instantiated on Wielenberg’s view through the obtaining of certain states of affairs, to which the existence of people may be as irrelevant as the existence of dinosaurs is to the obtaining of certain states of affairs about dinosaurs.

The demand for further explanation is a demand that can be made in regress of any moral theory, including Divine Command Theory, and so an endpoint must inevitably be reached. As Morriston says, “No matter what story you tell about the ontological ground of moral value, you must at some point come to your own full stop.”10 This is even conceded by Craig, commenting on his exchange with Sinnott-Armstrong, “the difference between the theist and Sinnott-Armstrong is not that one has an explanatory ultimate and the other does not. It is rather that the theist has a different explanatory ultimate.”11 As Wielenberg notes, though, Craig persists in demanding an explanation for the supervening of moral properties on non-moral properties, even while his own meta-ethical theory leaves certain supervenience relations unexplained.

In the preceding section, I drew attention to two questions: how is God good without reference to his character traits, and how does God’s goodness make some of his character traits good? As I argued there, no explanation for these questions seems to be forthcoming. It just is that God is good, and it just is that his goodness makes his character traits like love and justice good. But how, then, is this view any more plausible than the meta-ethics advocated by Wielenberg and Morriston? On the contrary, it looks as if there are some advantages to non-theistic, non-natural moral realism that are lacking on Divine Command Theory, as already indicated. A Platonist meta-ethics like Wielenberg’s is not without its own issues and unanswered questions, one being the problem of how we get moral knowledge from causally inert properties. But Craig’s own view identifies God with a moral property, and for all the above reasons, it seems less persuasive than the alternative discussed here.

IV. Conclusion

Dmitri Karamazov famously mentions in Dostoevsky’s novel The Brothers Karamazov that, ‘Without God, everything is permitted.’ This thought is far from uncommon in our world today, where most Americans are still quite reluctant to consider electing an atheist to the presidency.12 Religion and morality continue to be regarded as inseparable in the eyes of many, who think atheism to be incapable of accommodating any kind of robust morality. I have tried to show that such sentiments are misguided, for they often rest on expectations that can be challenged, or on deficiencies which are inescapable in even a strongly formulated divine command ethics. It is also worth noting that the 2nd premise of the moral argument is not uncontroversial, though it is absent from critical evaluation in this article. Intuitionism has been argued against not just by Mackie, but by Alasdair MacIntyre, Walter Kaufmann, Alexander Miller, and many other philosophers. Moral ontology in general is a complex subject with many live debates and ample room for disagreement. Craig’s meta-ethics, while indebted to Christian philosophers like William Alston and Robert Adams, is not shared by other prominent Christian philosophers, including Richard Swinburne, Linda Zagzebski, and Wes Morriston, the last of whom we have heard a bit from here.

In an essay on “Religious Ethics Versus Humanistic Ethics,” atheist philosopher Kai Nielsen comments on this intersection of religion and morality.

“God” by definition is “a being worthy of worship,” “wholly good,” “a being upon whom we are completely dependent,” These phrases partially define the God of Judaism and Christianity. This being so, it makes no sense at all to speak of judging God or deciding that God is good or worthy of worship. But the crucial point here is this: before we can make any judgments at all that any conceivable being, force, Ground of Being, transcendental reality, Person or what not could be worthy of worship, could be properly called “good” and even “the Perfect Good,” we must have a logically prior understanding of goodness. That we could call anything or any foundation of anything “God,” presupposes that we have a moral understanding, an ability to discern what would be worthy of worship, perfectly good. Morality does not presuppose religion; religion presupposes morality.13

The moral argument for God risks not only morality, but also theology in its attempt to ground value and obligation in God. Goodness is evacuated of meaning, moral responsibility is reduced to a matter of obedience to an external authority, and the “Most High” is elevated to a standard that has no moral content or significance in its ascription of worthiness, holiness, and perfection. Whether one is a theist, an atheist, an agnostic, or anything else, then, what may be most important is appreciating and pursuing a morality beyond the bounds of ideology.


Also see the article Why Be a Good Person? for more on morality, theism, and atheism.


1. J.L. Mackie, The Miracle of Theism (Oxford: Clarendon, 1982), p. 114-118.
2. William Lane Craig, “The Indispensability of Theological Meta-ethical Foundations for Morality.” Foundations 5 (1997): 9-12.
3. William Lane Craig, in William Lane Craig and Walter Sinnott-Armstrong, God? (Oxford: New York, 2004), p. 18.
4. Wes Morriston, “God and the Ontological Foundation of Morality.” Religious Studies 48 (2012): 21.
5. Craig and Sinnott-Armstrong, p. 69.
6. Ibid, p. 106.
7. William Lane Craig, Does Theistic Ethics Derive an ‘Ought’ from an ‘Is’? (June 13, 2010).
8. Erik J. Wielenberg, “In Defense of Non-Natural, Non-Theistic Moral Realism.” Faith and Philosophy 26, no. 1 (2009): 23-41.
9. Craig, in Craig and Sinnott-Armstrong, p. 19.
10. Morriston, p. 29.
11. Erik J. Wielenberg, “An Inconsistency in Craig’s Defence of the Moral Argument.” European Journal for Philosophy of Religion 4/4 (Winter 2012): 68-69.
12. Michael Lipka, Americans are somewhat more open to the idea of an atheist president, Pew Research Center (May 29, 2014).
13. Kai Nielsen, Atheism & Philosophy (Prometheus: New York, 2005), p. 217.