New Paper Replies to (Some of) Morriston's Important Critiques of Divine Command Theory

Brian Davis, Richard and W. Paul Franks. "Counterpossibles and the ‘Terrible’ Divine Command Deity", Religious Studies (forthcoming). The pre-publication version can be found here (No-citation or circulation rules apply).

Here's the abstract:
In a series of articles in this journal, Wes Morriston has launched what can only be considered a full-scale assault on the divine command theory (DCT) of morality. According to Morriston, proponents of this theory are committed to an alarming counterpossible: that if God did command an annual human sacrifice, it would be morally obligatory. Since only a ‘terrible’ deity would do such a ‘terrible’ thing, we should reject DCT. Indeed, if there were such a deity, the world would be a terrible place – certainly far worse than it is. We argue that Morriston's non-standard method for assessing counterpossibles of this sort is flawed. Not only is the savvy DCT-ist at liberty to reject it, but Morriston's method badly misfires in the face of theistic activism – a metaphysical platform available to DCT-ists, according to which if God didn't exist, neither would anything else.

Their key points are very similar to those made elsewhere by Baggett & Walls. I think the underlying point is worth addressing, though, viz., that the criticism would have more force if (i) it was tied to a specific account of the semantics of counterpossibles, (ii) that account is plausible, and (iii) the account entails the (non-vacuous) truth of the damning counterpossibles leveled against DCT. (An example of such a "damning counterpossible" would be, "if (per impossibile) God commanded rape, then rape would be obligatory.")

A few thoughts about the current state of the discussion on the topic:

1. I don't think it would take too much trouble to point to a relevant account of the semantics of counterpossibles that would meet desiderata (i)-(iii) above. See, for example, Barak Krakauer's recent account. Like most other accounts, his extends the Lewis/Stalnaker similarity or "closeness" relation so as to include impossible worlds. However, unlike many other accounts, his distinguishes between "nearer" and "farther" impossible worlds in a way that's sufficiently principled and fine-grained to determine the truth-values of the relevant counterpossibles here. I'd bet dollars to donuts that such an account would vindicate Morriston's intuitions (which I share) here.

2. I think the proponents of DCT make it easy on themselves with the examples of counterpossibles they consider against DCT. So, for example, instead of examples of commands that go strongly against the nature of a morally perfect being, how about something like, "If God commanded nothing one way or the other about rape, then rape would've been a morally neutral act". It seems to me this sort of counterpossible (if indeed it is such, and not an ordinary counferfactual or subjunctive conditional) is much harder to handle via the current sort of reply offered by Davis & Franks, Baggett & Walls et al.

3. For what it's worth, my own view is that the deepest point about the arbitrariness horn of the Euthyphro dilemma (which is at the heart of the current discussion at issue here) is that DCT entails that nothing is intrinsically right or wrong. So, for example, if DCT is true, then slavery and rape aren't intrinsically wrong -- wrong in themselves. Rather, their wrongness essentially depends upon the commands of a God. But the problem is that many have the intuition that rape, slavery, etc. are intrinsically wrong. As such, these intuitions are pieces of data that legitimately count against DCT for those who have them. And I don't think any amount of progress with respect to the formal semantics of counterpossibles is going to do anything to touch this point.

4. Those responding to Morriston's critique sometimes make things easy on themselves by failing to meet some of their dialectical obligations. To see this, consider the following two types of dialectical context:
(a) Proponent to critic: I've shown that *you are unreasonable* to think that DCT is false or unjustified. 

(b) Proponent to critic: I've shown you that *I'm not unreasonable* to think that DCT is true or justified.
Both dialectical contexts arise in the current discussion here. However, proponents of DCT have  sometimes responded in a way that's only relevant to dialectical contexts of the weaker type (b). Perhaps this comes out most clearly when proponents appeal to theistic activism to try to block the relevant counterpossibles from coming out true.

5. A perhaps small(ish) point: I find the polemical tone -- in both replies mentioned above -- to be extremely distasteful. It smells too much like apologetics, and not philosophy -- a point touched on recently by Gregory DawesI suspect I'm not alone on this. 


Angra Mainyu said...

Good points, EA.
I agree with 2-5 (I would need more time to consider 1).

Just two cents:

Leaving Morriston's “simple argument aside”, and on the issue of the argument using a subjunctive, there is the question of whether, in order to use counterpossibles or alleged counterpossibles in his argument, Morriston must show that there are non-trivial counterpossibles, and also provide a method for assigning truth - as Davis and Frank claim on page 4.

I'm not sure why that would be so, but on a related note, a number of defenders of DCT either claim or imply that there are non-trivial counterpossibles, and even assess some of them intuitively, without providing any method.

For example, here Craig seems to endorse, at least as the most plausible view, that there are counterpossibles that are non-trivially true or non-trivially false. Moreover, he even offers one example of each, without identifying the method for assigning truth-values – he just intuitively assigns them.

Given that, I would say DCTists who spouse such views should refrain from raising the objection against Morriston's argument raised by Davis and Frank, or explain why Morriston must meet that burden but they do not.

There is another difficulty for DCTists who are also Molinists and libertarians – not an unusual combination:

Given any actual person A such that A had a moral obligation to choose X under conditions C (which include the past of the actual world up to the point of the choice), and such that A actually freely chose X, one has:

P2(A,C,X): If A were to freely chose ¬X under conditions C, then A's actions would be morally praiseworthy. [alternative: “if A had freely chosen...etc]

I would say P2(A,C,X) is false, for any such A, C and X.

But it seems to me a libertarian who rejects [the libertarian version of] the PAP seems to have no basis for asserting that P2(A,C,X) is false in all actual cases, unless she (implicitly or explicitly) accepts that one may properly assess that P2(A,C, X) is false regardless of whether it's a counterpossible, or unless she regards that even if PAP is false as a necessity claim, as it happens the PAP holds in all past actual cases (but then, why would that be so?).

On the other hand, if a libertarian accepts libertarian PAP, then under the assumption that the PAP and libertarian freedom are true, then this paper you linked to earlier shows, under very plausible conditions, that Molinism is false.

Or let's suppose the DCTist/libertarian/Molinist is also a Christian, and she holds that if Pilate were in the biblical conditions, he would betray Jesus.
Then, one may consider the following:

P3: If, under the exact biblical conditions, Pilate freely chose not to betray Jesus, then God would send Pilate to Hell for eternity as a punishment for his failure to betray Jesus.

If the Christian also holds that P3 is false, then she's committed to non-trivial counterpossibles. Moreover, if she offers no method for assigning truth-values but still assigns truth-values and says P3 is false, it seems to me she ought not to claim that Morriston – or anyone else making a similar argument – must provide such a method, or explain why Morriston has that burden but she does not.

Angra Mainyu said...

Some further problems:

On page 8, they assess Morriston's claim that if a completely truthful and omniscient being said that 2 + 2 = 5, then 2 + 2 would equal 5, and try to infer O1 from a principle (p→q; so p > q), where “>” is counterfactual implication.
Then, they go on to say that this is a disaster for Morriston's argument against DCT, because as it happens, by Morriston's way of assigning values to the counterpossibles, this would make the counterpossible “If God were to command an SS, then an SS would be morally obligatory” would be true.
However, it seems to me that Morriston, in his 2012 paper, already says that.
In fact, Morriston seems to be saying that while Craig's reply to the “horrible command” objection fails, a different reply plausibly succeeds, and Craig's DCT escapes that kind of refutation.
Then, Morriston goes on to argue against Craig's version of DCT on different grounds.
So, it seems to me that Davis and Frank misunderstood Morriston's second paper.

However, this leads to other issues, like:

1. Is Morriston correct in his assessment of what's driving his own intuitions in assessing counterpossibles?
2. Assuming that he is correct, is that the correct way of assessing counterpossibles?
3. Assuming that the answers to both 1. and 2. are correct, are Morriston's – and/or Davis and Frank's, but they argue on similar basis – arguments in support of the truth of DCC under those conditions, persuasive?

Leaving 1. and 2. aside, briefly on 3., Morriston says that if a morally perfect being were to command X, then he would have morally sufficient reasons for commanding X, and then, the person receiving the command ought to obey. But it's not clear to me that necessarily, the person receiving the command ought to obey.
Even if the commander has morally sufficient reasons to issue a command, it does not follow that the person receiving the command ought to obey – not even if the person receiving the command knows that the commander has sufficient reasons to obey.

For example, let's say that Joseph is a brutal tyrant, and commands Bob – a general plotting to overthrow him in order to save countless people from horrible undeserved torture and death – to have innocent Alice killed.
Bob – a general - properly reckons, based on past behavior, that if he fails to command those holding Alice that she be killed, Joseph will have him killed, and then torture Alice to death slowly and horribly. That would be much worse for Alice than a quick death. Apart from that, it will be much worse for countless people, since without Bob, the coup will fail, and Joseph's reign of terror will likely last for much longer.
So, Bob commands that Alice be killed, for those reasons – mostly to save her from something much worse. However, the person receiving the command – say, Tom - is part of another plot to save some of Joseph's victims, and he has the means to smuggle Alice out of the country and make it look as if she was killed – he's done this before safely and repeatedly, and he's also promise his allies to do that again.
It seems to me, then, that Bob plausibly has sufficient moral reason to command that Alice be killed, but Tom – who receives the command -, does not have a moral obligation to comply.
This would not be any different if Bob were morally perfect, and if Tom knew it.
Granted, it might be argued that if Bob were omnimax (i. e., also omnipotent and omniscient), then he wouldn't give a command that is not morally obligatory to follow. But – say - a skeptical theist would have trouble saying that, and in any case, something else needs to be said.
Of course, this is leaving aside 1. and 2.; one may well challenge Morriston's take on counterpossibles.

Angra Mainyu said...

I've been thinking a bit more about Morriston's assessment (in his paper “God and the ontological foundation of morality”) that, based on the procedure he offers for assessing the truth of counterpossibles, Craig's theory and similar ones escape refutation, and I'm now more convinced that he's not right – and the same goes for Richard and Frank's similar argument, for the same reasons.

In particular, Morriston's procedure seems to be that a counterfactual is true when the content of the antecedent entails the consequent not just because the antecedent is impossible, but because of how it's connected logically to the consequent (page 7 of his paper). Further, he says that if a perfect being were to command that we eat our children, he would have a morally sufficient reason, and we [morally] ought to do it.

But it does not seem to be connected in that way in the case of the command to eat our children – he would have a morally sufficient reason, but that does not entail we wouldn't have a morally sufficient reason to disobey, or even a morally compelling reason to disobey.

Perhaps, some variants highlight the problem for DCT even further:

Let's consider the statement “If a perfect being commanded Alice to eat her children purely for fun, Alice would have a moral obligation to eat her children purely for fun”.

As in the previous case, the antecedent does not appear to be logically connected to the consequent in the specified way, so the procedure does not seem to render the counterpossible true, and intuitively, it seems clearly false. But moreover, it seems to me that the only reason someone might thing the conditional is true is that the antecedent is impossible, given that:

a. The perfect being's command provides no good reason whatsoever for Alice to eat her children purely for fun, since if she were to eat them partly because of the command, (or taking into consideration that the perfect being might have reasons, etc.), she would not be eating them purely for fun.

b. The fact that a perfect being commanded that action would not entail that fun is a morally compelling reason for Alice to eat her children, and even though it's intuitively implausible, all other things equal, that a perfect being would issue a command to a person who has no obligation to obey, not all other things are equal, and it's even more implausible that Alice would have a moral obligation to eat her children purely for fun.

Those considerations apply if we substitute “God” for “a perfect being”, it seems.

On the other hand, applying Morriston's proposed procedure, a metaethical theory that holds that our moral obligations are constituted by God's commands, is committed to the truth of the conditional about eating children for fun, because under such DCT, the required connection between antecedent and consequent is present.

There is another sort of commands that seem to work against DCT in this context, like the following examples:

C2: If God commanded Bob to always behave immorally, Bob would have a moral obligation to always behave immorally.
C3: If God commanded Alice to kill God just for the sake of destroying goodness, Alice would have a moral obligation to kill God just for the sake of destroying goodness.

Also, I think your example (i. e., "If God commanded nothing one way or the other about rape, then rape would've been a morally neutral act"), or a variant (e. g., “rape for fun”) works too.