Thursday, September 1, 2011

A simple argument against divine command theories

A standard line of objection against divine command theories is centered on the counterfactual:

  1. Even if God commanded it, torturing the innocent would be wrong.
But here it is extremely plausible that the antecedent is necessarily false—that God cannot command torture of the innocent. There is still a line of argument against divine command theories that continues past this roadblock, but I think it fizzles out.

But if we replace "God commanded it" with "God didn't forbid it", we actually get a much stronger argument. Actually, let's avoid counterfactuals, since we don't understand them well enough. We can give this argument:

  1. (Premise) Necessarily, torturing the innocent is wrong.
  2. (Premise) Possibly, God does not forbid torturing the innocent.
  3. (Premise) If divine command theory is true, then it is the case that: necessarily, something is wrong if and only if it is forbidden by God.
  4. Therefore, divine command theory is not true.
The argument is valid. Premise (2) is pretty plausible. It is justified by the same kinds of intuitions as (1) was. Premise (4) is uncontroversial, though it highlights the fact that the argument is specifically being aimed at divine command theories. Pure divine will theories are unaffected by the argument.

Interestingly, I think that if the argument works, it continues to work even if one replaces "God" with "a loving God", as in Robert M. Adams divine command theory.

The big question now is with regard to (3). A quick move to defend (3) is this. Possibly, God creates a world with no agents other than himself. In such a world, God wouldn't have any reason to issue any commands. So, possibly, there is a world with no agents other than God where no such commands have been issued. (Maybe you might object that God can issue a command to himself. But why would he need to? After all, the same loving character that might lead him to issue such a command would lead him to refrain from torturing the innocent.)

Now, this particular argument might make one worry that the assent to (2) was too quick. Perhaps instead the divine command theorist should have said:

  1. Necessarily, for every created agent x, it is wrong for x to torture the innocent.
However, I don't think the quantification in (2) should be restricted to created agents.

But suppose we do grant such a restriction. I think my argument can be rescued. Add:

  1. (Premise) Possibly, there is a created agent x who is not forbidden to torture the innocent.
  2. (Premise) If divine command theory is true, necessarily: for every created agent x and action-type A, A is wrong for x if and only if A is forbidden to x.
  3. So divine command theory is false. (By 6-8)

How can I defend (8)? For an initial line of defense, imagine that God created persons whose character is such that it would be unthinkable" for them to torture the innocent. Then God might reasonably refrain from forbidden them to torture the innocent not to give them the idea.

When I tried an argument like this on our graduate students, they came up with a very nice line of defense. God might command more fundamental things, such as to love God and neighbor. Torturing the innocent is incompatible with these fundamental commands. And it might be necessary that God command these more fundamental things because being subject to such commands might be constitutive of being an agent (or at least a created agent, I guess).

We can run this line of thought in two ways. First, we might say that what is incompatible with a command counts as forbidden. Second, we might modify (8) by saying that if divine command theory is correct, then it is necessary that something is wrong for a created agent if and only if it is incompatible with some divine command. For convenience, I will consider the first line—it won't matter, I think.

But what one is commanded by God is an extrinsic characteristic of a created agent, while being an agent is an intrinsic characteristic. So it seems problematic for divine commands to partly constitute our being agents. Imagine a being just like you, with the same nature, beliefs and other intrinsic features, but whom God did not command to love God and neighbor. Such a being still believes that she should love God and neighbor to the same extent as you believe it, and has the propensity to deliberate in light of love of God and neighbor as much as you do. Wouldn't she be an agent just as much as you?

Still, one might wonder what kind of reasons God might have not to command someone to love God and neighbor. But I think answers are possible. First of all, if it were possible to have persons who love God and neighbor with any obligation to do so, there plausibly would be a value in there being such persons—and it is hard for a divine command theorist to deny the possibility of such persons. Second, it could be that by giving an agent the command to love, God might be putting in the agent's head the idea that it is possible not to love. And there could be a value to creating agents who do not have any such idea. Third, if it were possible to do so, there would be a value to creating agents who cannot sin—and creating agents who are under no commands would be a way to do that if divine command theory is right. In fact, given divine command theory, God might create a mix of agents: some who are under commands, as we are, and some who aren't.

I don't know how strong this line of thought is. And like I said, it does nothing against theories that involve solely divine will considerations and have no command (or promulgation of will) component to obligation.


Heath White said...

Here is a line of thought, not necessarily endorsed.

First, your students’ point about God commanding more general duties, and things incompatible with those commands being wrong, is absolutely crucial on any reasonable view. God does not command anyone not to torture alleged Afghani terrorists, in so many words … but if he has commanded us not to torture anyone, then obviously the more specific case falls under the more general.

Second, the worry about necessary wrongness and contingent command-issuance can be assuaged if we observe a distinction that Adams points out. He says (IIRC) that good/bad is not a matter of divine command, but right/wrong is. The latter categories always arise in social relations (natural law as a special case of law generally), while the former categories depend on intrinsic features of an agent (natural telos, or something).

Consider a population which has not yet conceived of murder (Adam and Eve shortly after fall, perhaps). Maybe God doesn’t give them any commands against murder so as not to plant any ideas. Then murder is still bad, it would still impede union with God, it’s just not officially prohibited as it were. It is morally permissible in a nugatory sense, since if murder became thinkable God would pretty quickly command against it.

To make this view go, you need to worry less about natural law and more about (super)natural teleology. I.e. “sin” is not “acts contrary to divine law” but “acts contrary to one’s (super)natural end”. I’m not sure I like it though. (We might note that Cain is given no formal prohibition on murder, but judged for it.)

Alexander R Pruss said...

a. It seems clearly false to me that right and wrong arise only social relations. You can surely do wrong on your own. If it's wrong to torture the innocent, it's wrong to torture yourself when you're innocent.

b. I think the intuition behind (2) and (1) isn't just that torture of the innocent is bad for you, unless it's that special kind of bad-for-you that on natural law views constitutes wrong. :-)

In fact, one can imagine cases where it's not on balance bad for you in an ordinary sense of "bad for you" (it still has the kind of badness that on NL views makes it wrong, of course). For instance, suppose that if you don't torture this innocent, you will be inducted into the SS, and suppose empirical data shows that being inducted into the SS is very likely to result in even greater moral degeneration than that resulting from torturing one innocent.

Marc Belcastro said...

Dr. Pruss:

In my experience, some advocates of divine command theory hold that A is permissible if it's not forbidden by God. So if God is silent with respect to A, we can conclude that A is permissible. On this account, basically, God takes an active stance on forbidden matters (by issuing commands against them) and a passive stance on permissible matters (by not issuing commands against them).

But could the divine command theorist escape your objection by holding that God takes an active stance on both forbidden matters and permissible matters? On this account, A is permissible if and only if it's permitted by God. Thus, in a world populated by no agents other than God, if God doesn't issue a command not to torture the innocent, it doesn't follow that it's permissible to torture the innocent.

Perhaps such evasive maneuvers bring the divine command theorist too close to divine will theory.

Alexander R Pruss said...

So you still would have to deny (1), but at least you wouldn't have to say that torture would be permitted. And you don't get to say that if God didn't exist, everything would be permitted. But you can say that if God didn't exit, nothing would be permitted--or forbidden. (It's like error theories of morality.)

I think the fact that you'd still have to deny (1) is a problem.

I am also not sure how a speech act of permitting works unless it's done against a background of an actual or supposed prohibition.

Marc Belcastro said...

I'm inclined to think that the divine command theorist might not have to fear the denial of (1). I wonder if it's open to her to maintain that if, per impossibile, God were to command the torture of the innocent, then torturing the innocent would be good and obligatory. Of course, she'll insist that there are no such worlds in which this obtains, but if there were, any command which God issued would be good and obligatory by virtue of His being the necessary standard of goodness. So perhaps denying (1) isn't too problematic.

Regarding your comment about how a speech act would function in such a circumstance, do you think the same would hold for a speech act prohibiting A if it were issued without a background of an actual or supposed permission?