A Divine Copp Out

I promise to avoid bad puns in future titles.

David Copp advances a “society-centered” theory (SCT) of the justification of moral standards in Morality, Normativity, and Society.  It’s a standard-based theory.  “Standard” is used in this context to name a rule or imperative, for example, “Do not torture innocent children to death.”  Imperatives have no truth conditions.  But this does not prevent a moral claim from having truth conditions.  A claim endorsing the standard, like “It is wrong to torture innocent children to death” expresses a proposition about the standard.  It asserts that a standard that calls upon people to avoid torturing innocent children to death is appropriately justified.  This proposition is true only if there is a standard that calls upon people to avoid this behavior, and this standard is appropriately justified.

There are, of course, many different accounts of the conditions under which a moral standard is appropriately justified.  The correct account will be the one that gets things right about what actually justifies the standard; it will determine the truth conditions of moral propositions.  Copp’s moral theory is society-centered.  This means that the theory justifies moral standards in terms of the needs of a society.  By the lights of SCT, a moral standard is justified upon the condition that the society, given its needs, would be rationally required to select the standard as part of its moral system.  By the lights of SCT, a moral proposition endorsing the standard, asserting in effect that the standard is justified, is true if and only if it would be rationally required for the society to select the standard given its needs.  The proposition really is true if SCT provides the correct account of the conditions under which a moral standard is appropriately justified.

Some have complained that SCT is vulnerable to the objection that it implies counter-intuitive normative propositions.  Whether it is will be contingent upon what standards a society would be rationally required to choose were it in very different circumstances, possessing very different needs.  For example, the objection may go, by the lights of SCT, torturing innocent children to death would not be wrong if our society were in circumstances in which its needs would be better served were we to have a moral system that permitted torturing innocent children to death.  This, however, conflicts with our established, pre-theoretical moral beliefs.  Therefore, SCT does not provide the correct account of the conditions under which a moral standard is appropriately justified.  Furthermore, any theory that implies that anything goes in morality is wildly counter-intuitive.

The most troublesome and popular of objections to the Divine Command Theory of morality (DCT) is just this same “anything goes” objection (AG).  DCT is thought to imply the following conditional:

(AG)                If God were to command that we torture an innocent child to death, then torturing an innocent child to death would be morally right.

Since, as most people think, AG is false because it would be morally wrong for us to torture an innocent child to death under any set of circumstances, moral standards cannot be appropriately justified in terms of divine commands.  DCT does not provide the correct account of what justifies a moral standard and so does not provide the relevant truth conditions for moral propositions.  Any theory that implies that anything goes in morality is wildly counter-intuitive.

It seems true that AG is an implication of DCT.  Therefore, proponents of DCT require some way of affirming consistently that DCT provides the correct account of what justifies moral standards, that AG is an implication of DCT, and that this result is not morally counter-intuitive.  Is there any way to do this?  David Copp has suggested one in the context of his own “society-centered” theory (SCT) of the justification of moral standards.  The way he defends SCT against criticisms that it has morally objectionable normative implications can be adapted for a defense of DCT.

Copp responds to the criticism that his theory implies counter-intuitive normative propositions by noticing a scope ambiguity in statements having the following form, where X refers to something that conflicts with our established, pre-theoretical moral beliefs:

(ΑΓ)                 If circumstances were otherwise, then Xing would be morally right.

According to one interpretation of this, the scope of what Copp refers to as the “moral operator” is narrow:

(ΑΓ1)               If circumstances were otherwise, then morality would be such that Xing is

morally right.

The moral operator could also have a wide scope:

(ΑΓ2)               Morality is such that, if circumstances were otherwise, then Xing would be morally right.

Given the structure of the theory, ΑΓ1 is an obvious implication of SCT.  If society were radically different such that its needs would be better served by torturing innocent children to death in some cases, then a standard that called for this would be justified.  However, whether ΑΓ2 is true depends upon the actual needs of society.  According to Copp, the moral proposition expressed by ΑΓ2 is false by the lights of SCT.  SCT has it that torturing innocent children to death is morally wrong, and would be morally wrong in any society.  Therefore, there is no conflict between SCT and our established, pre-theoretical moral beliefs.

Disambiguating ΑΓ and being explicit about the position of the moral operator serves to distinguish a second-order account the conditions under which a moral standard is appropriately justified, a metaethical theory proposition, from a first-order expression of a moral proposition.  ΑΓ1 is an example of the former.  As such, it is not in competition with, and cannot conflict with, a first-order moral proposition.  The proposition “Morality is not such that, if circumstances were different, then torturing innocent children to death would be morally right” is in competition with, and does conflict with, ΑΓ2.  But SCT does not imply ΑΓ2.  According to Copp, SCT implies the moral proposition that torturing innocent children to death is wrong, period.

Therefore, normative objections to SCT, objections to the theory made on moral grounds, fail to notice the difference between a first-order expression of a moral proposition and a second-order metaethical account of what justifies moral standards.  Copp’s strategy suggests an important lesson:

LESSON:        Established, pre-theoretical moral beliefs can conflict with a normative proposition, but they cannot conflict with a metaethical proposition. 

The normative objection to DCT suggested by AG makes this same mistake.  Proponents of DCT can make use of the lesson in order to show that the “anything goes” objection to DCT fails.

Copp’s strategy suggests that AG is ambiguous between the following:

(AG1)             If God were to command that we torture an innocent child to death, then morality would be such that torturing an innocent child to death is morally right

(AG2)              Morality is such that, if God were to command that we torture an innocent child to death, then torturing an innocent child to death would be morally right.

AG1 provides the narrow reading where the moral operator governs only the consequent.  AG2 provides the wide reading of AG where the moral operator governs the entire conditional. 

If DCT provides the correct account of the conditions under which a moral standard is appropriately justified, then AG1 must be true.  Since the theory justifies moral standards in terms of God’s commands, if God commanded the torture of an innocent child to death, then morality would be such that the torture of that child to death is morally right.  But this is simply a theoretical point.  AG1 is a second-order, metaethical proposition, not a normative proposition.  As such, it is not in competition with, and cannot conflict with, any of our established, pre-theoretical moral beliefs.  In this case, the relevant pre-theoretical moral belief is “Morality is not such that, if (circumstances were different and) God were to command that we torture an innocent child to death, then torturing an innocent child to death would be morally right.”  DCT does not imply that this belief is false.  This belief is a moral proposition, like the one discussed above.  Like that one, it asserts that a moral standard prohibiting torturing innocent children to death is justified.  This proposition is true by the lights of DCT if God has issued a command to avoid the behavior.  If DCT provides the correct account of what justifies moral standards, then the proposition really is true.

If this is the case, that is, if our pre-theoretical moral belief is true by the lights of DCT because there is a standard forbidding that anyone torture an innocent child to death, and this standard is justified in terms of God’s commands, then the obvious implication is that by the lights of DCT, AG2 has to be false.  AG2 and our pre-theoretical moral belief are first-order moral propositions.  They conflict with each other and so they cannot both be true.  Since our pre-theoretical moral belief is true by the lights of DCT, AG2 has to be false by the lights of DCT. 

The purported problem for DCT is that it is supposed to be incompatible with many of our established, pre-theoretical moral beliefs.  This incompatibility is merely apparent.  “Do not torture innocent children to death” is a moral standard.  Proponents of DCT claim, roughly, that moral standards are justified in terms of God’s commands.  This claim is a metaethical proposition offering an account of what has to be true about a moral standard in order for the moral standard to be justified.  As we discovered in the earlier lesson, it is not possible that a metaethical proposition could conflict with any moral belief, not even firmly established, pre-theoretical intuitions.  A claim expressing a proposition about the standard, like “It is wrong to torture innocent children to death” may conflict with our moral beliefs, but in this case it simply does not. 

Someone could respond that the pre-theoretical intuition is actually not offered as a first-order, substantive moral judgment.  Rather, it is a second-order, metaethical theory claim.  The idea is that what people really object to about DCT is that if God’s commands had been different, then morality would have been such that a standard permitting the torture of an innocent child to death is justified.  Understood that way, the intuition really is in conflict with AG1 and so the theory.  However, this move would be somewhat disingenuous.  Originally, the intuition was offered by way of an argument that DCT has counter-intuitive moral implications; it is most typically presented as an objection made on moral grounds.  Defenders of DCT should take such arguments seriously.  But it is no surprise that, and defenders of DCT need not take seriously the fact that many will have theoretical intuitions incompatible with DCT.  That is just to say they have a different theory and that is no argument against DCT.  It is merely the expression of disagreement.  Every standard-based moral theory offers a different account of what makes moral standards appropriately justified.  Even if the proposed standards are very similar under alternative accounts, the facts that justify the standards will be different if the accounts will be different.  Therefore, it would be very strange indeed if people did not find accounts that justify standards in ways different than their own to be counter-intuitive. 

Therefore, the only way to make the anything goes objection to DCT stick is to deny that it is possible to make a distinction between a metaethical position and a substantive normative position.  However, this is implausible.  Such a distinction is commonplace in ethics and made quite easily.

31 Replies to “A Divine Copp Out

  1. Kyle, this is a really interesting post, and I have one question about the Copp argument. Essential to his argument, it appears, is inserting a moral operator into a claimed consequence of his theory. What justifies this insertion? Is the claimed consequence supposed to be ambiguous?

  2. Jon,
    Thanks for your question. Do you mean the insertion of the moral operator into (ΑΓ)? That proposition is ambiguous as to whether it’s a first-order moral proposition or a metaethical theory proposition. Inserting the moral operater makes it clear which. If the person lodging the anything-goes objection intends the former, then the moral operator should be inserted in the antecedent. That’s (ΑΓ2). If he intends the latter, then the moral operator should be inserted in the consequent. That’s (ΑΓ1).
    (ΑΓ2) could be used to form a moral objection to DCT or SCT, if any proponents of those theories actually accepted that proposition. But they don’t. (ΑΓ1) might form the basis of some sort of objection to DCT or SCT, but not the basis of a moral objection.
    Does this help or have I misunderstood your question?

  3. Kyle,
    Simon Blackburn uses an argument similar to yours to defend his quasi-realism (in Ruling Passions, e.g.) Let me try to explain why, ultimately, I am unconvinced.
    Suppose I advanced the following claim about justification in mathematics: a mathematical claim (like 2+2=4) is justified by the fact that Stephen Hawking believes it. (Call this the Hawking-Centered Theory, or HCT.) An obvious objection is that 2+2 would equal 4 even if, in fact, Hawking believed that 2+2 equaled 5. Thus HCT seems to conflict with our established, pre-theoretical mathematical beliefs. However, it seems I could defend HCT with a version of the argument you have suggested. I can admit that, if circumstances were otherwise, mathematics would be such that 2+2=5; but this is a meta-mathematical claim, and does not conflict with anyone’s pre-theoretical mathematical beliefs about anything. (If it does conflict with anyone’s beliefs, it must conflict with their theoretical beliefs; but this is “merely the expression of disagreement” about the justification of mathematics, and I “need not take seriously the fact that many will have theoretical intuitions incompatible with” HCT.) I deny, however, that mathematics is such that, if circumstances were otherwise, 2+2 would equal 5.
    This defense seems clearly unconvincing: HCT is a bad theory, and it is a bad theory precisely because it implies that 2+2’s being 4 is in some strong sense contingent on Hawking’s believing that it is. The problem, I think, is this: the belief that ‘2+2=4 even if Hawking believes otherwise’ is as much a straightforward mathematical belief as ‘2+2=4.’ Perhaps this is easier to see if we switch back to the moral case. ‘Killing a person is wrong even if he has rationally consented to it’ is as much a normative moral claim as is ‘killing a person is wrong.’ The specification of circumstances does not make a moral claim into a nonmoral meta-ethical claim. But if this is so, then (T2) ‘Torturing a baby to death is wrong even if God commands it’ is as much a moral claim as (T1) ‘torturing a baby to death is wrong.’
    I assume that Kyle will want to agree that (T2) is a moral claim but deny that DCT implies it, by holding it to be the same sort of claim as his (AG2) — and indeed, to conflict with this claim — but not his (AG1). But I do not see how this can be sustained. If DCT is correct, then morality is such that, if God were to command that we torture an innocent child to death, then torturing an innocent child would be morally right; to deny this is just to deny DCT. (Put it this way: if DCT is true, then it is true that if God were to command X, X would be morally right. That’s practically a definition of DCT. So how does adding the phrase ‘morality would be such that’ alter the truth value of the claim?)
    Return to the mathematical case for a moment. One cannot hold that mathematical truths depend on the beliefs of Hawking – so that, if Hawking were to change his mind about 2+2’s being 4, 2+2 would no longer be 4 – and simultaneously deny that mathematics is such that if Hawking were to come to believe that 2+2 did not equal 4, 2+2 would no longer equal 4. Similarly the DC Theorist must, I take it, admit that if, tomorrow, God were to command the torturing of an innocent child, the torturing of an innocent child would be the right thing to do. So while I will admit that (AG1) and (AG2) are different, I do not see how the DC Theorist can hold one of them while denying the other.

  4. Troy,
    I don’t know if you saw my previous comment before you posted yours, but I think it’s relevant. There might be really good reasons to reject DCT; they just aren’t the reasons advanced by anything-goes objectors.
    So I agree that HCT is a bad theory. And I agree that it’s a bad theory “precisely because it implies that 2+2’s being 4 is in some strong sense contingent on Hawking’s believing that it is.” DCT might be a bad theory, too. Maybe it’s a bad theory precisely because it implies that torturing innocent children to death being wrong is in some strong sense contingent on God’s prohibiting it. But there are different ways of prosecuting this thesis. One way is to follow anything-goes objectors. Theirs is a moral objection to the theory. The defense I’ve invoked is supposed to block that sort of moral objection. The moral proposition that anything-goes objectors concerned with isn’t advanced by DC theorists.
    There might, however, be other ways to prosecute the thesis that DCT is a bad theory because it holds that the truth or falsity of moral propositions are in some strong sense contingent on God’s commands. Some people think, e.g., that DCT is vulnerable to the open question argument or that if DCT is true then there’s no real disagreement between Hare’s missionary and the cannibals he confronts. I haven’t said anything about non-moral objections to DCT.

  5. Kyle,
    But I think that my objection is a moral objection to DCT, because DCT must imply claims like ‘Torturing infants would be morally right if God commanded it,’ and this is clearly a moral claim and, by the lights of our pre-theoretical intuitions, a false one. I think I wasn’t really clear about this in my earlier comment. And I ended that comment by suggesting that I understood the difference between (AG1) and (AG2) – which, looking back on it, I don’t really think I do.
    What I really want to take issue with, then, is your claim that it is possible, and indeed easy, to “make a distinction between a metaethical position and a substantive normative position” that helps the DCT avoid the anything goes objection. Of course, we can distinguish between metaethics and normative ethics in terms of subject matter. But I don’t think we can always say, for every ethical claim, whether it is metaethical or normative; at least not if we assume that a claim classified as metaethical will therefore have no moral implications. For instance, the metaethical claim that all claims of the form ‘X is morally wrong’ presuppose nonexistent objective values and are therefore false seems quite clearly to imply that ‘Torturing babies for fun is morally wrong’ is false, and thus that torturing babies for fun is not morally wrong. And similarly, the metaethical claims involved in DCT seem clearly to imply that X would be morally right if God commanded it, for any X, and this is quite clearly a normative moral claim.
    It is because of this that I think anything goes objector gets it exactly right. But perhaps you could explain the difference between (AG1) and (AG2), and why DCT is committed to one but not the other?
    Troy

  6. Troy,
    Sorry I wasn’t responsive to your objection. Your analogy with HCT made me think you were making a different point.
    In my original post I did admit that DCT must imply claims like AG. But you say that AG is clearly a normative moral claim. This isn’t clear to me. AG isn’t obviously a claim about some normative matter concerning what to do or refrain from doing. Instead, it could be a claim about what has to be true about a particular moral standard in order for that moral standard to be justified. It’s only adequately justifed in terms of God’s commands. If AG is understood this way, AG is true whenever DCT is true.
    There is a way to understand AG according to which it may be a claim about some normative matter, but it’s odd I think. It says that it’s normatively true that torturing infants would be right if God commanded it. This claim, I think, is so odd that I’m not actually sure that it should be understood as a genuinely normative claim concerning what to do or refrain from doing. But even if it is, it might be false, even if DCT is true. In order for this very odd claim to be true, God would have to issue a standard to that effect.

  7. I’m also not sure I see the difference between the two claims. I think I understand the claim, that if God commanded torturing babies it would be morally right. It seems analogous to someone (perhaps a hedonistic egoist) claiming that if torturing babies were fun it would be morally right. I see no ambiguity as yet in either version. And I’m not sure what adding ‘morality is such that’ into an otherwise seemingly OK counterfactual in two different places does to its meaning.
    Wouldn’t it be a good objection to the latter view that fun or not, torturing babies is wrong?
    Is the idea that the DCT is really the claim that what is wrong is what god actually commands? Then I could see how we might distinguish two claims where there seem to be only one. We might mean (1) if the actual world were such that God had commanded torturing babies it would be right, or we might mean (2) that in any world where God had commanded torturing babies it would be wrong. The actualized version of the DCT would seem to endorse (1) but not (2). But are holders of the DCT really holding an actualized version of it. (I think I can see that someone who accepts Copp’s theory might go this route, but I’d be surprised if DCT theorists would.)
    It might help me if you would say whether AG2 entails AG1.
    FWIW I’m a bit skeptical of the need to introduce moral operators to disambiguate sentences that seem to involve predicates applied to act types. I have the feeling I may be embarrassing myself by missing something here, but . . .

  8. Kyle,
    You say that there is a way of understanding AG that makes it a normative claim, but that on this reading it is a very odd claim. On this reading, “It [AG] says that it’s normatively true that torturing infants would be right if God commanded it.” I’ll agree with you that this is an odd claim, because while I know what ‘true’ means, I really don’t know what ‘normatively true’ means.
    You give a hint of what you take it to mean in your previous paragraph: “But you say that AG is clearly a normative moral claim. This isn’t clear to me. AG isn’t obviously a claim about some normative matter concerning what to do or refrain from doing.” So on your view, if a claim is normatively true, then (i) it is a claim about a normative matter concerning what to do or refrain from doing, and (ii) it is true. Now, I think it’s a bit misleading to use the phrase “normatively true” to denote these two quite separate properties because it suggests that this is some sort of special kind of truth, which it isn’t. (It’s as if we said the sentence “Andy is angry” is ‘emotionally true’ because it concerns emotions, and is true. The sentence isn’t emotionally true, it’s just true.)
    But anyway, there is a more serious problem, which is that if DCT is right, then AG seems to possess these two properties. For (ii) if DCT is right, then ‘torturing a child is right if God commands it’ is true. And (i) whether or not DCT is right, ‘torturing a child is right if God commands it’ is indeed “a claim about a normative matter concerning what to do or refrain from doing.” We can’t disqualify it as a normative claim simply because it specifies certain conditions for, as I pointed out above, most if not all normative claims specify certain conditions. Thus, ‘killing is morally permissible if it is in self-defense’ doesn’t cease being a normative claim just because it specifies that the killing is justifiable under certain conditions. So why, then, should the fact that we specify that one may and indeed should torture children under certain conditions — i.e. if God commands it, but not (presumably) otherwise — disqualify ‘torturing children is morally right if God commands it’ as a normative claim?
    Perhaps someone will say that this claim is irrelevant, since God hasn’t actually commanded the torturing of children. But this won’t work. Suppose we objected to Extreme Patriotism, the theory that action A is morally justified if and only if one’s country commands one to A, on the basis that it would imply the truth of TC: torturing children is morally right if one’s country orders it. The Patriot could hardly respond by saying that, since his country has never yet commanded the torturing of children, TC is not a normative moral claim, but merely a theoretical or hypothetical one. The point of the objection is that torturing children is wrong whether or not one’s country has ordered it – and that is a moral claim, and one that is clearly incompatible with Extreme Patriotism. The upshot: Claims about what would be morally justifiable under certain hypothetical circumstances are still normative moral claims. Thus, claims about which moral claims are justified under certain circumstances either are or directly imply normative moral claims – which means meta-ethics and normative ethics are too closely connected for Copp/Swan/Blackburn-type arguments to succeed.
    Besides, even if one’s country has never yet commanded the torturing of children, it might do so tomorrow. And so might God – unless, of course, we think there are objective moral standards He is bound by. But to think that, obviously enough, is to give up on DCT altogether.

  9. I realize that “torturing innocent children to death” is just being used as an example to explain other things, but have you noticed that our medical system tortures innocent children who often die anyway?
    Doctors don’t take on patient care responsibilities and need to have a better picture of how much suffering accompanies each invasive procedure, as midwives did; when women did the doctoring they did the nursing, too.
    Society’s needs have clearly made the immoral moral in this instance. It is not unusual for fundraisers and prayer groups to form around children who are close to death. Children are tortured so that society can avoid or postpone the grief of the death of one so young.
    Kids who experience lots of heel sticks grow 25% more pain receptors there, experiencing increased sensitivity to pain for a lifetime. Preemies also suffer lifelong medical problems resulting from excessive medical “care”
    Doctors often are expressing their intolerance of sick people more than a desire to heal people. If you want to read about the torture of innocent loonies, see Mad in America by Robert Whitaker. Another case of what’s good for society is torture for the individual.
    Getting paid less than enough to live on is a physically painful experience. Being paid less than enough to take care of your kids is even worse. But how else will society force young people into the military, creating future security guards who protect only the rich?
    There are many cases where what benefits society is completely unethical, if moral by a narrow definition.

  10. Mark said, “It might help me if you would say whether AG2 entails AG1.”
    AG2 doesn’t entail AG1. AG1 is equivalent to the assertion that it’s a metaethical fact that (AG). It provides the putative truth conditions for normative claims. Since, according to DCT, the truth conditions for normative claims are set by God’s (perceptive) will, if God willed some infant’s torture, then it would be right to do it. AG2 is equivalent to the assertion that it’s a normative fact that (AG). I said this was an odd claim, but that anyway it’s not entailed by DCT. Remember from the original post, and my previous comment responding to Troy, that I think of a normative claim as a claim about a standard calling for some kind of behavior. A normative proposition claims that the standard calling for that behavior is justified. AG2 is such an odd claim because the standard which it is about (“torture some infant if God wills it”) is a standard calling for behavior conditional on the very standard contained in it (“torture some infant”) being such that God has willed it. It would only be true if this very odd, complex standard was such that God has willed it, since, according to DCT, that’s the only way it could be justified. So it’s not just that God hasn’t willed infant torture. It’s also that God hasn’t willed the very odd, complex standard in AG2. And given its strange complexity, he’s very unlikely to will standards like that. He issues ones like “Thou shalt not steal” and so on.
    But Troy says:
    “We can’t disqualify it as a normative claim simply because it specifies certain conditions for, as I pointed out above, most if not all normative claims specify certain conditions. Thus, ‘killing is morally permissible if it is in self-defense’ doesn’t cease being a normative claim just because it specifies that the killing is justifiable under certain conditions. So why, then, should the fact that we specify that one may and indeed should torture children under certain conditions — i.e. if God commands it, but not (presumably) otherwise — disqualify ‘torturing children is morally right if God commands it’ as a normative claim?”
    The strange sort of complexity isn’t involved in any claim that’s conditional; it isn’t involved in the claim about killing in self-defense. The difference is that the “if it’s in self-defense” condition isn’t a metaethical account providing the putative truth conditions for normative claims. The “if God wills it” condition is.
    I think you just deny there to be a metaethical DCT that concerns a different domain of discourse than normative propositions.

  11. Jerio, nice post.
    Kyle and Troy, I’m having a hard time getting my head around AG1 and AG2 (as phrased–there must be a better way), but now that you’ve both done all this work debating the two sides I’m inclined to agree with Kyle about the legitimacy of the normative/metaethical distinction regarding DCT. The hard part for me in trying to find any plausibility in DCT is along the lines of Troy’s statement: “unless, of course, we think there are objective moral standards He is bound by. But to think that, obviously enough, is to give up on DCT altogether..” DCT detaches “wrong” from “bad” in this sense: torture is (I hope we can agree) obviously and horrifically bad for the child who is tortured. To say the torture is wrong only because God says so, while being bad for the child is irrelevant to its wrongness, is sufficient to put DCT far down on the list of plausible moral theories as far as I am concerned.
    FYI, there is another PEA Soup thread on DCT where I posted a couple days ago with comments about an article by Philip Quinn, also relating in part to the “anything goes” challenge to DCT. Quinn says that God, being good (I’m paraphrasing), is indeed bound to will what is good, but that DCT is not thereby refuted. Well, maybe this leads to circularity or vacuity, but even if not, it still means that “right” and “good” are only coincidentally linked by the fact that God is good and therefore wills the good. If God had turned out to be bad, then it would (by DCT) be right to do bad. There are far more illuminating ways to characterize the good/right link than “God says so.”

  12. I’m having a lot of trouble with the (AG1) vs. (AG2) distinction, too. Here’s my best try at understanding the difference; Kyle (or someone), please tell me if this is right.
    Compare the legal situation in the United States. The advocate of (L1) is a philosopher of law; the advocate of (L2) is a lawyer, an expert in US law.

    (L1) If the framers of our Constitution had preferred a constitutional monarchy, then US law would have been such that the king in parliament is supreme.
    (L2) US law is such that if the framers of our Constitution had preferred a constitutional monarcy, then the king in parliament would be supreme.

    These are plainly different. (L2) is false (my imaginary ‘expert’ isn’t much of one!), but (L1) is probably true. The difference seems to be this. (L1) hypothesizes a different set of legal-system-determining facts, while (L2) sticks with the actual set of legal-system-determining facts and hypothesizes a different set of facts for the actual legal system to apply to.

    If this is an analogous distinction to the distinction between (AG1) and (AG2), then I understand it. (I doubt that morality is like law in the right way to make the analogy go through, but I at least understand the view that morality is like that.)

  13. I, too, am struggling to see the relevant difference between (AG1) and (AG2). I don’t really understand what you, or Copp, might mean by this so-called moral operator “morality is such that”. Take this sentence, for example:

    Morality is such that snow is white.

    I’m inclined to say that this is true; since snow is indeed white, everything has the property of being such that snow is white — including morality (if such a thing there be). However, if we understand the phrase “morality is such that” in this way, then (AG1) and (AG2) turn out to be equivalent. So you must have some other meaning in mind. Presumably, you would say that snow’s being white is necessary but not sufficient for morality’s being such that snow is white. So what else is required? Perhaps you’ll say that it’s required that the proposition “snow is white” be, in some sense, a “moral” proposition. But then I’d like to hear more about what this sense is.
    I should say that this is also a question for Jamie: why do you say that your (L2) is false?

  14. I have another worry. You say that (AG2) is a disambiguation of (AG). But I don’t see how that could possibility be the case. Surely, if (AG) is ambiguous, then (AG2) must be so too.
    Here are the two statements:

    (AG) If God were to command that we torture an innocent child to death, then torturing an innocent child to death would be morally right.
    (AG2) Morality is such that, if God were to command that we torture an innocent child to death, then torturing an innocent child to death would be morally right.

    Now, in order to understand (AG2), we need to understand the the clause following “morality is such that”. But that clause just is (AG). Hence, if there are multiple ways of understanding (AG), there must be multiple ways of understanding (AG2) as well. Since (AG) is a part of (AG2), any ambiguity in the former must also infect the latter.

  15. Bob: You might be interested in Mark Murphy’s Stanford Encyclopedia entry on theological voluntarism — http://plato.stanford.edu/entries/voluntarism-theological/#2 (I don’t know how to do links in the comments section). He lists a number of considerations in favor of DCT — historical, theological, and theoretical.
    Jamie: Yes, I think that’s analogous. In my most recent response to comments above I tried to explain a little better why the distinction would make a difference to the DC Theorist. But you doubt that morality is like law in such a way that would allow it to work for them. What’s the law like in virtue of which allows the distinction to work there?
    Campbell: “Snow is white” asserts neither a normative nor metaethical proposition. The sentence “Morality is such that snow is white” is false because it asserts something about morality that isn’t true. It’s not a normative fact that snow is white. And if Jamie will allow me, L2 is false because it asserts something that’s false about US law. You can check the books if you like.

  16. Campbell,
    I think that (L2) is false because the law does not include any conditional laws to the effect that whatever the founders (would) prefer is thereby law. It just seems intuitively false, though. My explanation of why is a hypothesis.
    Kyle,
    The difference is that we make the law, and we could have made it differently. The counterfactual about what the law would have been in different circumstances, then, need not be a proposition of law itself. On the other hand, we do not make morality, in the relevant sense. The counterfactual about what would have been wrong or right in other circumstances, then, seems to be itself a proposition of morality.
    Sketchy, huh?

  17. Hi Jamie. I’ve noticed this “two-token commenting” going on before, but I can’t really figure out what the cause is. Using the preview function does not seem to send the message–at least it hasn’t for me. I’ll check into this a bit more.
    Dan

  18. Hi all,
    Interesting discussion!
    Perhaps it’s useful to think about the nature of the “anything goes” objection a bit. As Kyle described the objection, part of the alleged difficulty for DCT is that it implies
    (AG) if God were to command that we torture an innocent child to death, then it would be right for us to do so.
    Suppose DCT does imply this; what’s the problem, exactly?
    Kyle suggests that the conditional is false “because it would be morally wrong for us to torture an innocent child to death under any set of circumstances.” I think this is on the right track; the following claim seems true to me:
    (A) Necessarily, it is morally wrong to torture an innocent child to death.
    My intuition is that not only is torturing innocent children to death in fact morally wrong, it couldn’t possibly fail to be morally wrong.
    The problem, then, is that (AG), *together with the assumption that it is possible that God commands us to torture an innocent child to death*, entails that it is possible that it is morally right for us to torture an innocent child to death – and this conflicts with (A).
    (One of the standard responses to the “anything goes” objection in the literature involves arguing that necessarily, God does not command us to torture an innocent child to death).
    The upshot is this: If this is the right way to understand the “anything goes” objection, then disambiguating (AG) into (AG1) and (AG2) doesn’t help, since, as far as I can see, both (AG1) and (AG2) (together with the assumption that it is possible that God commands us to torture an innocent child to death) conflict with (A).
    Erik

  19. Jamie: Thanks, that makes sense. But yes, I think it’s sketchy in the sense that according to DCT that’s almost exactly what morality is like. Of course with DCT it’s not us who makes morality; rather, God establishes the truth conditions for normative claims. But since the DC Theorist claims that God “makes” morality in that sense, he can just as well claim that the counterfactual about what would have been wrong or right in different circumstances of God’s willing isn’t itself a normative claim.
    Of course you can say that morality just isn’t like that. You must, then, have a different metaethical theory in mind. For obvious reasons, this doesn’t surprise me. But, in any case, the relevant normative objection to DCT won’t work.
    Erik: You say, “Kyle suggests that the conditional is false ‘because it would be morally wrong for us to torture an innocent child to death under any set of circumstances.'” This quote from my original post is positioned there as a normative proposition, and so for the reasons presented there and in various responses to comments, not inconsistent with DCT.

  20. Kyle,
    Thanks for your reply. That’s helpful. It seems that what you mean by “morality is such that P” is something like “it is a moral fact that P”.
    But what about my second worry? You haven’t said anything about this. Do you accept that, if (AG) is ambiguous, then (AG2) is so, too?

  21. Kyle,
    Well, let me try again. In your initial post, you said that DCT implies
    (AG1) If God were to command that we torture an innocent child to death, then morality would be such that torturing an innocent child to death is morally right.
    So, let’s assume that (AG1) is true. Now suppose that God commands that we torture an innocent child to death. I take it that it would follow that
    (i) Morality is such that torturing an innocent child to death is morally right
    from which it presumably follows that
    (ii) It is morally right to torture an innocent child to death.
    So, I’d put the objection to DCT this way:
    1. If DCT is true, then (AG1) is true.
    2. If (AG1) is true, then it is possible that it is
    morally right to torture an innocent child to death.
    3. But it’s not possible that it is morally right to
    torture an innocent child to death.
    4. Therefore, DCT is false.
    Premise (2) depends on the assumption that it is possible that God commands us to torture an innocent child to death.
    I’d consider premise (3) to be a pre-theoretic, moral belief, and I’m willing to concede that (AG1) is a
    metaethical proposition. Nevertheless, given the assumption identified above, (3) and (AG1) are incompatible.
    I guess I still don’t see anything in the preceding posts that makes trouble for this argument – though of course it’s entirely (epistemically) possible that I’ve missed some subtlety that defeats the argument. If so, perhaps someone could explain which of (1)-(3) above is false and why.
    Erik

  22. (Warning: long post ahead. You might want to grab a cup of coffee first.)
    I think Jamie’s legal analogy is useful, and lets us finally see the distinction between meta-ethical and normative claims that the AG1/AG2 distinction was supposed to capture – but also helps us see why this approach won’t help the DC Theorist. Rather than the constitutional monarchy example, let’s choose one that makes the analogy with morality a bit more perspicuous:
    (L2*) “U.S. law is such that if the framers had disapproved of women drinking coffee, women’s drinking coffee would be illegal” is obviously false: the laws of the United States say many things, but they nowhere make claims about what would be illegal under conditions that do not obtain and (because the framers are dead and have no chance of changing their minds) have no chance of obtaining. But this is compatible with holding true the claim that (L1*) “If the framers had disapproved of women drinking coffee, U.S. law would have been such that women’s drinking coffee would be illegal.”
    So ‘U.S. law is such that x’ means ‘according to U.S. law, x’ or ‘x is a claim that is stated in U.S. law.’ Similarly, ‘morality is such that x’ must mean ‘according to morality, x’ or ‘x is a claim that is explicitly stated in morality.’
    This indicates that we are taking ‘morality’ as the name of some specific code. So suppose that we take it that way: ‘morality’ is the name of the code that is given, say, in the New Testament (assuming there is a clear and consistent one). Now nowhere in the New Testament does it say that torturing babies is right if God says it is (I hope). So, even though might be true that if God had wanted us to torture babies, he would have let the disciples know (through Jesus) and so this would have been stated in the New Testament, it does not follow, and is not true, that the New Testament states any such counterfactual fact. Indeed the New Testament need not even contain any general counterfactual claim, to the effect that whatever God had commanded would have been morally right.
    All well and good so far, but there is a big problem: the moral claim offered by the moral objection to DCT is not supposed to conflict with any claim of ‘morality.’ That is, no one (so far as I know) who advances this objection to DCT is claiming that the actual moral code that the DC Theorist is justified by God’s willing it contains false moral claims. So to respond on this basis is to attack a straw man. What the moral objection to DCT claims is that the account of moral justification provided could just as easily have justified moral codes that would have been flawed –so that, however the justification of morality works, it can’t possibly work this way.
    The problem is all the more pressing because, unlike the framers of the Constitution, God is still around (on most accounts that have Him existing at all) and so perfectly capable of commanding, at any given future point, the torturing of children. So, confronted with the question ‘Would torturing children be right if God commanded it?’ the DC Theorist must answer ‘yes’ (unlike the person who follows the moral code the DC Theorist is proposing a justification for, who might be able to evade the question by saying, ‘My moral code doesn’t cover that contingency.’ But the challenge, again, is aimed at the DC Theorist, not the moral follower. That’s why distinguishing between various placements of ‘morality is such that’ won’t help).
    Return to the legal case for just a moment. The distinction between (L1*) and (L2*) (or between L1 and L2) is understandable, but is presented against a background in which no question of justification is raised. Suppose, however, that it was presented by someone defending FCT – the Framers’ Command Theory – according to which our current set of laws is justified precisely because (and, presumably, only to the extent that) it is in line with the Framers’ wishes. It would be a clear objection to this theory that it would allow for the justification of unacceptable laws: for instance, it implies not just (L1*) but (L3*):
    (L3*): Had the framers of the Constitution disapproved of drinking coffee, a law prohibiting drinking coffee would have been justified.
    Now, Kyle will presumably object that the corresponding moral claim
    (AG*) Had God commanded the torture of children, the torture of children would have been justified.
    is not true, at least when interpreted as a normative moral claim, because a normative moral claims is true if and only if God wills it, and God hasn’t willed (AG*). But again, this is no more persuasive than the FC Theorist’s arguing that because (L3*) is not a true legal claim (i.e. not implied by the laws themselves), it cannot be challenged by our intuitions about the law. (L3*) is, indeed, not a legal claim in that sense, but it is a claim about what laws would be justifiable; thus it can indeed be challenged by our legal intuitions.
    LESSON: You can’t simultaneously hold that moral claims are true only when God wills them, and provide a meta-ethical account of when “moral standards are appropriately justified” which justifies some moral claims that God has not willed (like ‘torturing babies would be right if God willed it’), unless you are drawing a very strong distinction between a true moral claim and a justified one. But that distinction (unlike the distinction between ‘true belief’ and ‘justified belief’ in epistemology) seems very implausible. Moreover, even if we allowed it the moral criticism of DCT could still be made, on the basis, not that it’s implausible (and monstrous) to think that ‘torturing children is right’ could ever be true, but that it’s implausible (and monstrous) to think ‘torturing children is right’ could ever be justified. (Is this now a meta-ethical rather than a normative claim, because it refers to justification? I don’t think so: if the claim that torturing children cannot be justified under any conditions is not a moral claim, I don’t know what is. So yes, I am skeptical about the deep divide between normative and meta-ethical claims that Kyle is relying on – and I hope my reasons for being skeptical are by now clear.)

  23. Campbell,
    Sorry – I had just missed your second worry. I think it (as well as some of the earlier comments here) convinces me that I didn’t in the original post present the disambiguated claims in the clearest way. Nonetheless, I think there is an important difference between AG understood as a metaethical claim and AG understood as a normative claim. If this is allowed, then I think the strategy for defending DCT works: the DC Theorist affirms that it’s a metaethical fact that (AG); he denies that it’s a normative fact that (AG).
    Troy,
    You now seem to accept the idea that there is a difference between the two claims, but deny that it can be used to much effect in a defense against the *real* anything-goes objection. But I do think your argument implies that the anything-goes objection isn’t really a normative objection. You say:
    “What the moral objection to DCT claims is that the account of moral justification provided could just as easily have justified moral codes that would have been flawed –so that, however the justification of morality works, it can’t possibly work this way.”
    You have intuitions at the theoretical level according to which morality can’t work this way. As I wrote to Jamie, and in the original post, I’m not at all surprised that people have theoretical intuitions about the way moral standards are justified or about what it is that establishes the truth conditions for moral claims that are out of line with DCT. You and others advocate some different theory.
    Now maybe you have those theoretical intuitions about how morality can’t work because of certain of your normative beliefs. Regardless of where these theoretical intuitions come from, I don’t see how this makes the objection to DCT any less question-begging. I also think this is the right way for the DC Theorist to respond to Erik’s argument.

  24. Kyle,
    Your strategy is to dismiss objections to DCT by saying they are based on “intuitions at the theoretical level” and therefore “question-begging.” I take it, then, that a similar and equally effective defense could be mounted of what I will call the DCT+PA theory. DCT+PA identical to DCT except that it allows the Pope to overrule God on certain matters – including, interestingly, the torturing of children. Thus, DCT+PA claims that if the Pope authorized the torturing of children, morality would be such that torturing children would be morally right. This seems like an obviously flawed theory but, having granted your premises, I can’t think of any way to argue against it. If I point out that torturing children is not right under any circumstances, and so would not be right even if the Pope authorized it, one could say that that is just a moral claim, and so cannot conflict with any claim of DCT+PA. In order to conflict with DCT+PA, a claim must be a metaethical claim, i.e. a claim about justification – but on your account, to raise such a claim just means that I have “some different theory” about justification, and is therefore to beg the question. (As you wrote earlier, “defenders of DCT need not take seriously the fact that many will have theoretical intuitions incompatible with DCT. That is just to say they have a different theory and that is no argument against DCT. It is merely the expression of disagreement.”)
    What’s going on is that DCT, though it is a metaethical position, does imply some normative moral claims, and these do conflict with our pretheoretical moral intuitions in a very deep way. (The same is true, of course, of DCT+PA.) Let me set out the argument in more detail. (Quotations from Kyle are italicized.)
    “Standard” is used in this context to name a rule or imperative, for example, “Do not torture innocent children to death.” Imperatives have no truth conditions. But this does not prevent a moral claim from having truth conditions. A claim endorsing the standard, like “It is wrong to torture innocent children to death” expresses a proposition about the standard. It asserts that a standard that calls upon people to avoid torturing innocent children to death is appropriately justified.
    So:
    [D] ‘X is wrong’ iff ‘don’t X!’ is appropriately justified.
    [D’] ‘Torturing children is wrong’ iff ‘Don’t torture children!’ is appropriately justified.
    Presumably, moreover, the following also obtain:
    [D1] ‘X is right’ iff ‘X!’ is appropriately justified.
    [D1’] ‘Torturing children is right’ iff ‘Torture children!’ is appropriately justified
    Now let’s go to what I take to be a noncontroversial implication of DCT:
    If God were to command the torturing of children, then ‘torture children!’ would be appropriately justified. (call this DCT-I)
    But: if ‘torture children!’ is appropriately justified, then torturing children is right (by [D1’]).
    So: if God were to command the torturing of children, then it would be true that ‘torturing children is right.’ (DCT-I + D1’)
    So: ‘torturing children is right’ would be true under some circumstances.
    But this conflicts with our pretheoretical moral judgment that torturing children would not be right under any circumstances. Thus, DCT has unacceptable moral implications and should be rejected.

  25. Troy,
    This isn’t right:
    “Your strategy is to dismiss objections to DCT by saying they are based on “intuitions at the theoretical level” and therefore “question-begging.””
    There are many objections to DCT that don’t make this mistake. I only dismiss the anything-goes objection to DCT this way. Actually, the full thesis is that the anything-goes objection prosecuted at the normative level can be dismissed because it’s based on a false assumption. It’s false that DC Theorists are committed to it being a normative fact that (AG). The anything-goes objection prosecuted at the theoretical level can be dismissed because proponents of the theory needn’t be worried that others have theoretical intuitions according to which they deny that it’s a metaethical fact that (AG). This is simply an expression of disagreement at the metaethical level. It obviously won’t connect up with someone who endorses DCT, especially one who endorses the theory for various historical, theological, or theoretical reasons (reasons, by the way, that don’t seem to be available to the defender of DCT+PA).
    I think the most interesting issue here concerns the putative normative implications of a metaethical view. I take normative and metaethical beliefs to be distinct, such that they can’t conflict or undermine each other (the LESSON from the original post). Some disagree with this. Ronald Dworkin (“Objectivity and Truth: You’d Better Believe It”) has argued against emotivism based on that theory’s putative normative implications. I consider this to be a bad argument against emotivism (but, of course, not all arguments against emotivism make this mistake).

  26. Kyle,
    Are you sure you want to endorse this general claim? You wrote:
    “I think the most interesting issue here concerns the putative normative implications of a metaethical view. I take normative and metaethical beliefs to be distinct, such that they can’t conflict or undermine each other (the LESSON from the original post).”
    Even if the strategy of isolating certain metaethical theories from normative criticism worked (for instance by saying that what is right is what is actually disapproved of by God or me or society, so that counterfactual disapproval did not change the truth of a moral claim) it can’t be generally true that all metaethical theories cannot be refuted by normative criticisms.
    Take for example a certain kind of reductive metaethical view. Goodness is condusiveness to human flourishing. Rightness is the property (of actions) maximally promoting goodness. The ‘is’ here is the ‘is’ of identity. This seems like a paradigm metaethical view.
    But if I have the normative intuition that some act-type (say culling unhealthy humans from the population) is always wrong, and I think it is still wrong when it promotes overall flourishing, then I would seem to be justified in rejecting that theory for what I think is one of its normative implications. That the one view is normative and the other metaethical doesn’t seem to me to insulate either from the other.
    best,
    Mark

  27. Mark,
    In the situation you describe, I think I’d want to endorse the idea that you have a normative belief that leads you to have, or implies that you have, theoretical intuitions which are incompatible with utilitarianism. I think this keeps the two domains of discourse as distinct as I need.
    I wouldn’t deny that you could thereby be justified in rejecting utilitarianism, since such intuitions can provide a kind of internal justification. But I do think it would be fine for utilitarians to dismiss objections based on them as mere expressions of disagreement within the metaethical domain.

  28. Kyle,
    I agree that I have intuitions that are incompatible with utilitarianism. But there are reductive *metaethical* theories that entail utilitarianism, and I was trying to describe one of those. My claim was that since these metaethical theories entailed utilitarianism, and the intuitions conflicted with utilitarianism, there was a conflict between the metaethical theory and the intutiions. How that conflict should be resolved is not really relavent to the point which is just about whether as a general matter we can isolate the metaethical from the normative in such a way that views one domain can’t be relevant to the justification of views in the other.
    best,
    Mark

  29. Thanks, Mark and sorry; I was calling that reductive metaethical theory you described utilitarianism. I shouldn’t have done that.
    I should have said you have a normative belief (“culling unhealthy humans from the population is always wrong”) that leads you to have, or more likely implies that you have, theoretical intuitions which are incompatible with the metaethical theory you described. So this isn’t actually a normative objection to that theory. It’s a theoretical objection according to which you deny that theory has it right about the way morality really *is*. If I’ve described this correctly, then the relevant normative and metaethical claims are distinct.
    Again, you might be perfectly justified in rejecting the theory you described based on your theoretical intuitions. But that doesn’t mean it constitutes a strong objection to the metaethical theory that its proponents should take very seriously. You just disagree with them at the metaethical level.

Comments are closed.