5 / 06
Image of birds flying. Image of birds flying.

#228 Formulating the Moral Argument

August 29, 2011

Dr. Craig,

I have what I hope will be a helpful suggestion rather than a question. It concerns your formulation of the moral argument for God's existence. (I should also note that I can't take credit for the original insight since a fellow graduate student in philosophy brought it to my attention).

Your formulation of the moral argument is as follows:


1. If God does not exist, objective moral values and duties do not exist.

2. Objective moral values and duties do exist.

3. Therefore, God exists.

The problem with this argument is that if the conclusion is correct, then [1] is a trivial truth. Most theists (including yourself) claim that God is a necessary being. In other words, the proposition "God exists" is a necessary truth. If this is correct, then the proposition "God does not exist" is contradictory and hence a necessarily false truth (akin to saying "The husband is not married"). But everything follows from a contradiction. To clarify, if the antecedent of a conditional is necessarily false, then the conditional as a whole will always be true, regardless of what the consequent is. So the conditional "If God does not exist, then the married bachelor danced with the square circle" is true.

Perhaps a simple way to fix this problem is the flip the conditional and create a modus ponens:


1*. If objective moral values and duties exist, God exists.

2*. Objective moral values and duties do exist.

3*. Therefore, God exists.

One might argue, however, that this formulation has its own problems. We ideally want the first premise to express the proper sort of relationship between morality and God, and one could argue that [1*] is not strong enough to do that. To demonstrate this, take the conditional "If I ate eggs and toast this morning for breakfast, then God exists." This conditional is true, since every time that the antecedent is true, the consequent is true. And since I did, in fact, eat eggs and toast for breakfast, then I have shown that God exists! This is unsatisfactory because me eating eggs and toast doesn't really have anything to do with God's existence.

I think that this argument is flawed. I would argue that me eating eggs and toast does have something to do with God's existence, since it is impossible that I exist or do anything without God. Perhaps the point that the above object is trying to make is that no one would accept that what I eat has anything do with each other unless they already believed in God. This may be true for eggs and toast, but it is not the case with morality! Your defense of [1*] is intended the show why there IS a connection between God and morality. Thus, the above reformulation seems satisfactory.

Still there is another reformulation that we could try:


1`. Objective moral values and duties exist if and only if God exists.

2`. Objective moral values and duties do exist.

3`. Therefore, God exists.

This argument seems to avoid some of the concern that your original argument faced; on the other hand, I am not sure if you would endorse [1`]. One might argue that it is not true that moral duties must exist if God exists. For example, one may claim that moral duties arise from the commands of God, and it is possible that God did not create anything and thus did not issue any commands. Perhaps this last formulation is better:


1~. Objective moral values exist if and only if God exists.

2~. Objective moral values do exist.

3~. Therefore, God exists.

Formulations 2 or 4 seem to be able to your best options. Granted, I am writing this on the spur of the moment and may have missed some obvious weakness of these formulations. I think what is really going on in the moral argument is that you are arguing that a belief in objective moral values and duties commits one to a belief in God. The trick is to get this accurately expressed in a few pithy premises.

I understand that you may have other reasons for keeping the argument in its current format. The big reason being that no one would consider [1] a trivial truth unless they already accepted the conclusion, and the argument is primarily aimed at those who do not already accept the conclusion. It seems, however, that the argument should not be ENTIRELY aimed at those who do not believe in God. Ideally it should be able to be used by theists to confirm their pre-existing belief in God as well. Regardless, I thought that bringing this to your attention might be of assistance,and I also hope that the discussion may be beneficial to others interested in the moral argument.

I appreciate and admire your work.

All the best,


Flag of United States. United States

Photo of Dr. Craig.

Dr. craig’s response


Thanks, Blake, for your thoughts on this! For readers who lack a background in logic, let me clarify the problem. The nub of the problem is that logically the truth of an “If . . . , then . . .” statement is a function of its antecedent and consequent clauses. That is to say, the truth of the whole statement is determined by the truth of its parts. Let’s symbolize an “If . . . , then . . .” statement as pq. The truth value of pq is a function of the truth values of p and q. Letting “T” stand for “true” and “F” for “false,” we can determine the truth value of pq as follows:

p q p → q

As you can see, pq comes out false only when the antecedent clause is true and the consequent clause is false. (I know that seems weird, but that’s the way “→” is understood logically.)

So what does this imply for the first premiss of the moral argument?

1. If God does not exist, objective moral values and duties do not exist.

Well, since God does exist, the antecedent clause is false. Therefore, no matter what the consequent clause is, (1) comes out true! (Look at our truth table above.) So it is also true that

1′. If God does not exist, objective moral values and duties do exist.

A falsehood implies anything! So for the theist, (1) and (1′) are said to be trivially true or vacuously true.

For that reason alone, Blake, your attempts to reformulate the argument by flipping or modifying the antecedent and consequent clauses really don’t help much because you’re still using the same truth functional connective “→.” So take your (1*), for example. Since, as our truth table reveals, the whole conditional statement comes out true if the consequent is true, then it doesn’t matter whether the antecedent clause is true or false. So we could have just well have asserted

1**. If objective moral values and duties do not exist, God exists.

As your toast and eggs example illustrates, it doesn’t matter what the antecedent is; so long as the consequent is true, the whole statement comes out true.

Ultimately, I don’t think any of this matters, either logically or dialectically. Logically, the moral argument as I have stated it remains sound, having true premisses and being logically valid. Nor is there any danger of someone’s employing a trivially true premiss like (1′) to formulate a sound argument for the opposite conclusion, since from (1′) and (2) nothing follows logically. Dialectically, the theist would be in trouble only if the sole reason he believes (1) is true is because he believes the conclusion of the argument to be true and so is begging the question. But no proponent of the moral argument claims that the reason we should accept (1) is because it is trivially true. Rather he appeals to the reasons atheists themselves give for thinking (1) to be true. As you observe, the atheist like Nietzsche, Russell, or Sartre does not regard (1) as merely trivially true. So dialectically, the theist is in a comfortable position in presenting this argument to non-theists.

In any case, I should say that I think (1) is actually a disguised counterfactual. These are “If . . . , then . . .” statements in the subjunctive, rather than the indicative, mood. So the non-trivially true premiss is actually

1′′. If God did not exist, then objective moral values and duties would not exist.

In other words, if there were no God, that is to say, if atheism were true, then there would be no objective values and duties. Unlike conditional statements in the indicative mood, the truth value of a counterfactual is not a function of its constituent clauses. The counterfactual connective is symbolized “counterfactual” in order to differentiate it from “→.” It is not a truth functional connective. So (1′′) is not merely trivially true in virtue of the falsity of its antecedent clause. The conclusion then follows from (1′′) and (2), as before.

But here a new problem arises. You’ll notice that thus far I haven’t addressed your modal concerns, concerns about the necessary falsehood of the antecedent clause. Here those concerns do become relevant. For if we believe, as I do, that God is a metaphysically necessary being, then the antecedent clause of (1′′) is not merely false, but necessarily false, that is to say, impossible. But on the customary semantics for counterfactual conditionals, counterfactuals with impossible antecedents are trivially or vacuously true. That is to say, on the customary semantics it is also trivially true that

1′′′. If God did not exist, then objective moral values and duties would exist.

We seem to be right back where we started!

Fortunately, a good many philosophers today quite rightly recognize that some counterfactuals with impossible antecedents are non-trivially true or false. For example, the theist will say that it is clearly non-trivially true that

4. If God did not exist, the universe would not exist.

and it is non-trivially false that

5. If God did not exist, the universe would still exist.

Such counterfactuals are sometimes called counterpossibles. In the case of the moral argument, I think that (1′′) is a true counterpossible.

This has occasioned some discussion of late. In their recent book Good God (OUP, 2011) defending both Divine Command Morality and the moral argument for God’s existence (recommendation for all you interested in the moral argument), Jerry Walls and David Baggett discern the counterfactual nature of my first premiss and express reservations about so formulating the moral argument because it concedes too much to the non-theist, namely, that there still would be a world of personal agents acting in ways that we do if there were no God. Instead, they want frame the argument as an inference to the best explanation: the best explanation of objective moral values and duties is that God exists.

While I’ve no problem with the moral argument formulated as an inference to the best explanation, I don’t share their reservations about my formulation and, indeed, I think that when one gets down to showing that the best explanation of objective moral values and duties is God one will almost inevitably slip into arguing that atheistic explanations are not as good because given atheism, objective moral values and duties would plausibly not exist. For example, Walls and Baggett later in their book lapse into precisely this sort of argumentation, claiming,

Nietzsche is an example of someone who grasped far reaching implications that follow from rejecting God’s existence. There’s something refreshingly bold and honest about the way he saw among such implications a radical revision of traditional values and the loss of the necessary foundations for morality classically construed. . . . Whereas Nietzsche understood the transcendent foundations of classical morality and the implications of atheism, contemporary atheists who think God can be painlessly eliminated from our ontology and yet remain sentimental about moral discourse fail to take Nietzsche seriously enough (pp. 202-3).

But these Nietztschean implications are exactly what (1′′′) asserts! My formulation of the argument has the dialectical advantage of meeting the atheist where he is and asking whether morality would be objective if God did not exist. We don’t want the moral argument, after all, to collapse into the cosmological argument, dependent on asserting something like (4). We want it to stand on its own as an independent argument. Therefore, there’s nothing wrong with allowing the atheist the impossible supposition that we live in an atheistic world and then asking whether objective moral values and duties would exist.

So if this is the correct formulation of the first premiss of the moral argument, why don’t I present it that way? Well, I’ve toyed with the idea. But I figure that using subjunctive conditionals would just confuse popular audiences. As you observe, the problems we’ve been discussing are unique to the theist because he regards God’s existence as metaphysically necessary. But that is dialectically irrelevant to presenting the argument to non-believers, who do not, of course, share that assumption. No non-theist will harbor such reservations about the argument because he doesn’t believe that God exists. So for simplicity’s sake and because nothing hangs on it, I’ve continued to present the argument using conditionals in the indicative mood.

- William Lane Craig