Do atheists steal morality from the Christian worldview?

Some conservative Christian apologists argue that whenever atheists engage in moral reasoning, they are stealing from the Christian worldview. As I understand their argument, the point is supposed to be that only Christianity can metaphysically ground objective morals and values and that this would be obvious if only we were not so blinded by sin (herein: the obviousness thesis). In their unrighteousness, atheists deny Christianity while simultaneously presupposing Christianity in their moral discourse, or so it is claimed. Call this the Presuppositionalist Moral Argument (PMA). In what follows, I show that PMA is implausible because, due to the controversy over Hume’s Is/Ought dichotomy, the obviousness thesis is probably false.

Hume’s is/ought dichotomy is the view that no statement about what ought to be can be derived from a statement about what is the case. While Hume used his dichotomy to argue that moral statements should be understood non-cognitively, many contemporary metaethicists utilize the dichotomy to argue that moral facts constitute an independent domain of facts which cannot be reduced to non-moral facts. Other metaethicists argue that moral facts can be reduced to non-moral facts. At present, there does not seem to be a consensus between the two views and it is difficult to determine which side is correct. Given the debate, I think it reasonable to assign a 50/50 chance as to whether the is/ought dichotomy is true:

P(Hume’s dichotomy is true|debate) ~ P(Hume’s dichotomy is false|debate) ~ 0.5

But now the kicker: Hume’s dichotomy is incompatible with theistic metaethics (as I will show), which entails that morality can be explained by Christianity only if Hume’s dichotomy is false. There are multiple metaethical theories compatible with the conclusion that Hume’s dichotomy is false; thus, the probability of Christian metaethics is less than 0.5:

P(Christian metaethics|debate) < 0.5

Far from being obviously true, Christian metaethics is less likely to be true than the alternative.

Why should we think that theistic metaethics is incompatible with the Hume’s dichotomy? Here I quote from one of my working papers:

Although many classical theists find arguments for moral realism convincing and, from this, conclude that God exists, we argue in the following that the conjunct of moral realism and the is/ought dichotomy entail that theism is false. […]

Let’s suppose that both moral realism and the is/ought dichotomy were true. There are two candidates for a theistic metaethics: first, that morality is reducible to God’s commands or, second, that morality is reducible to God’s nature (for example, perhaps the Good is identical to God’s essence).

There are well-known reasons why it is implausible for moral truths to be reducible to God’s commands. First, this would render moral truths arbitrary due to the fact that God could not possess moral reasons for Her commands (Euthyphro’s Dilemma). Furthermore, given the is/ought dichotomy, it is difficult to see how God could possess non-moral reasons for Her commands, for this would entail that moral facts could be reduced to various non-moral facts (perhaps concerning God’s nature or desires). But it seems equally strange to suppose that God possesses no reasons at all for her commands. Second, reducing moral truth to God’s commands involves reducing moral facts to facts about commands, which violates the is/ought dichotomy.

The second option, that morality is reducible to God’s nature, seems to be the best candidate for a theistic metaethics. However, the second option entails that moral facts are reducible to non-moral facts about God (Greg Bahnsen suggests, for example, that ethical facts are reducible to facts about God’s character); thus, the second option is not a viable candidate for theistic metaethics. Moral facts might be reducible to facts about God’s essence or nature, but unless we identify facts about God’s essence or nature as moral facts, this reduction violates the is/ought dichotomy.

Perhaps the theist will argue that facts about God’s essence or nature are identical to moral facts. For example, some theists have argued that, given Divine Simplicity, God is identical to the property of Goodness, itself, in virtue of which all other moral facts can be explained. However, we find this to be mysterious, if not plainly contradictory. How is it that a specific person can be identified with an abstract property or with a universal?

The remaining option is that the ethical facts, somehow, constitute an independent domain of facts not created by God. Yet, given God’s aseity, everything else that exists is somehow caused or explained by God. […] Given moral realism and the is/ought dichotomy, the moral facts are not caused or explained by God. This is a contradiction. Thus, if moral realism and the is/ought dichotomy are true, theism is not, or so it seems. Perhaps the theist has a way of resolving this difficulty, but whatever that way might be is not obvious.

There is a third difficulty with the notion that ethical absolutes entail theism: theistic metaethics appears to render the order of explanation of moral truths incorrectly. Bahnsen has argued that God’s commands are both good and non-arbitrary because they issue forth from God’s good character, which is the ultimate measure against which all else must be compared. Put aside the difficulties with the is/ought dichotomy. Most theists conceive of God as good because She is virtuous, fair, loving, and so on; Bahnsen’s view entails that virtuousness, fairness, and lovingness are good because God possesses these traits. That’s a reversal of what seems to be the correct order of explanation.

Wes Morrison puts the point this way: “Is God good because he is loving, generous, just, faithful, kind, and so forth? Or are these attributes good-making because God has them?” The first possibility is more plausible than the second. Theists say that God is good in virtue of Her properties. If Bahnsen chooses the second, it is difficult to make sense of the notion that God is good (in virtue of what would God be good?).

Thus, Christians should not say that atheists steal from their worldview; the obviousness thesis fails.

I do not think that this is surprising. Ordinarily, when we engage in philosophical debate, we do not accuse our interlocutors of theft when we think that they made unwarranted assumptions. Instead, we say that our interlocutors have not presented a convincing case (or something similar) for their conclusions. The accusation of theft sounds, to my ears, like little more than bullying for ideological purposes.


  • May 1, 2016 - 4:34 am | Permalink

    Your comments belie a complete misunderstanding of moral/ethical grounding. The theist need not approach the is/ought problem at all. There is no dichotomy in classical (particularly Thomist metaphysics). In your comments about God being good because of a, b, and c attributes, you’re putting the cart before the horse and ignoring that it’s a car. God is not good because of a, b, and c attributes, God is Good because that is the nature of God’s character. The moment you start ascribing a, b, and c attributes to God as goodness, you’re judging God based on your own preferences defining what you think is good. In answer to your question: “How is it that a specific person can be identified with an abstract property or with a universal?” It is the nature of God to be so. Asking how is asking for a finite to judge and discern how an infinite can have a particular nature. Only the atheist has to contend with the is/ought dilemma, the atheist must show how a mere chemical reaction can have intentionality and be subject to judgement. For the atheist, all moral statements are mere statements of preference. “Saying that murdering babies is wrong.” Is nothing more than saying, “This particular action causes a chemical reaction that is unpleasant to this particular meat machine.” Meat machines that are “dancing to their DNA” can’t even have preferences. They can’t even really have identity or intentionality. Without intentionality how could an ought have any validity.?

    • Dan Linford
      May 1, 2016 - 11:38 am | Permalink

      Sam —

      Thanks for your comments! This blog post is quite old, so I wonder how you came across it. I’m now a blogger for Secular Spectrum on the Patheos network, where you’ll find several other blog posts about metaethics. Moreover, I’m well aware of the view that you’ve accused me of being ignorant of — i.e. the view that axiological value corresponds to God’s nature and deontic value corresponds to God’s commands. If you are interested in reading my in-depth responses to modified divine command theory, I encourage you to read those posts. Suffice it to say that merely declaring God’s nature (or essence, as the case may be) to be identical to axiological value is not a sufficient reply to the objection I’ve raised. Goodness is an abstract universal and therefore not a concrete particular; God is a concrete particular. As typically construed by philosophers, it’s simply incoherent to say that any abstract universal is numerically identical to any concrete particular. Thus, the burden is on the theist to demonstrate how it would even be metaphysically possible for God’s nature or essence to be identical to Goodness, itself.

      I don’t think you’ve solved this problem yourself. You stated: “Asking how is asking for a finite to judge and discern how an infinite can have a particular nature.” I take it that what you mean is that a finite being cannot determine how an abstract universal and a concrete particular are numerically identical in the Godhead because doing so exceeds our comprehension as finite beings. That’s an assertion. It may well be true, but you’ve offered nothing by way of evidence or argument for the claim other than that i.e. Thomists have believed it to be so. Moreover, the atheist could equally well say that we know we are “meat machines” (to use your terminology) and that morality consists of an objective domain of facts, but it exceeds our comprehension as to how all of this fits together. Importantly: anyone could claim that the difficulties with their worldview are merely the result of our finite natures and that, while the answers are out there, they exceed our comprehension as finite beings.

      To be sure, I don’t think we can comprehend everything. There is much that does exceed our finite understandings. But it’s not at all convincing to interlocutors to point to mystery when objections are raised. Theistic metaethical views might well be true in some way that neither of us can understand, but, if they are, why are we even bothering with this conversation? After all, if theistic metaethical views are true, but only in some way that we cannot understand, then we cannot understand those theories, and, in consequence, there is nothing to be said concerning them. This is the so-called paradox of ineffability. If x is ineffable, then we cannot say anything about x (including that x is ineffable).

      Furthermore, I find it ironic that you accuse me of misunderstanding, since your own post indicates a variety of misunderstandings. In my post, I argued that theistic metaethical views often conflate is and ought statements; that is, they often confuse descriptive statements with normative statements. In response, you’ve claimed that only the atheist faces the is/ought problem. And, on your view, only the atheist faces the is/ought problem because, as you’ve interpreted it, the is/ought problem concerns the notion that normative facts could not arise out of purely natural (or physical) facts. That is, on your view, the atheist has to explain how moral facts could arise out of “mere statements of preference”. However, you’ve misunderstood the is/ought problem.

      The is/ought problem concerns a much more general dichotomy than that between the normative and the natural. The dichotomy between is and ought statements is a distinction between normative claims (i.e. moral claims) and claims about what is the case. What Hume argued when he presented the is/ought dichotomy is that no ought statement could ever be derived from an is statement. That is, that no normative claim could ever be deductively entailed by any collection of non-moral claims. Importantly, this means that no claim about the supernatural — whether about God or otherwise — could entail moral claims. To be sure, this cuts against views which attempt to reduce morality to purely naturalistic facts — whether those are facts about well-being, the subjective attitudes of human beings, or whatever. But it also cuts against any attempt to reduce morality to purely supernatural facts, such as facts about God.

      You’ve accusing me of ignoring a variety of classical views about God’s nature. But, for whatever reason, you’ve ignored the vast number of philosophers according to whom moral facts are not “mere statements of preference”, correspond to some objective domain of facts, but do not correspond to any particular facts about God. That is, you’ve ignored the mainstream metaethical view that the moral facts are both God-independent and objective. To name only a few authors, you’ve ignored GE Moore, Rush Shafer-Landau, David Enoch, Erik Wielenberg, David Boyd, and Peter Railton. This list could go on for a very long time. This is a common ploy maintained by theists — to attempt to monolithically reduce all non-theistic views to some sort of radical subjectivism — but it’s not at all convincing to anyone that actually knows the metaethics literature.

      Lastly, your comments about intentionality are, frankly, bizarre. There are plenty of naturalistic conceptions of intentionality. But even if we did not presently have any naturalistic conceptions of intentionality, it’s hard to see how that could possibly have relevance to the post. I’ve argued that morality is independent of God. I did not argue that morality reduces to our reactive attitudes (such as our preferences) or that human beings are purely physical objects, composed of nothing more than chemicals, etc. I think the former — that morality is about anyone’s preferences (whether human or God) — is just false. Morality is not about anyone’s preferences. And while I think there is some large degree of empirical evidence for the claim that we are nothing but physical matter suitably arranged, I won’t pretend to have a favored explanation for our intentional states. I will add that everything I said is fully consistent with the notion that we actually have souls. While I do not presently believe that there are any souls, the existence of souls would not justify adopting theism.

  • Greg Reeves
    May 3, 2016 - 12:24 am | Permalink

    Great post and I am in well over my depth. However, am I curious to know how you get around the problem that you have proven roo much? That is, you seem to want to take both “Hume’s distinction is correct” and “moral realism is true”. However, unless you limit Hume’s distinction to only physical states of being (ie, you can’t go from statements of physical fact to moral fact) then don’t you run into the problem that your two statements are mutually contradictory?

    In other words, if moral realism is true, then you ha’e the “is” statement “moral facts exist”. How then can you go to “you should follow those moral facts”?

    Or “murder is objectively morally wrong” then goes to “you ought not murder”. Am I on to something here or am I missing something?

    It seems to me that if you want to reject the classical “Good is so because God is good” because of the is/ought distinction, how can you still hold to the statement, “murder is objectively morally wrong, therefore you ought not murder”? Can you reject one without rejecting the other?

  • Leave a Reply