In the spirit of my series of posts on the basis of morality (which I plan to continue), I thought I would include the following thoughts about the divine command theory, which were prompted by recent conversations at Randal Rauser’s blog. That discussion began because I claimed that the notion that God is the source of morality is noxious. Now, perhaps that was a poor choice of words given that it is prone to cause offense. In any event, I think it is true that this idea is a false one and is not conducive to clear thinking about morality. Let me explain:
I think that the divine command theory (DCT) facilitates poor moral reasoning. The reason is that it asks us to focus on morally irrelevant features when we reason about matters of moral concern. This, at least in some instances, can blind us to significant moral truths.
The divine command theory says that our moral obligations are grounded in the commands of God. The most widely defended modern version of that view stems from the work of Robert Adams. Adams’ version of the DCT says that obligations are grounded in the commands of a God who is perfectly loving. I don’t think that this modified divine command theory (as it has come to be called) is particularly relevant to what I have to say here, but my criticism will be directed at this view (or something near enough to it).
The problem with the divine command theory that I want to highlight is that it claims that the moral features of actions are specifiable solely in terms of God’s commands. What makes an action morally obligatory is the fact that God commands that we do it. What makes an action morally wrong is the fact that God commands that we not do it. The problem is that these are morally irrelevant features. In other words, the fact that God commands an action has nothing to do with what makes the action wrong, nor indeed can it. What makes an action wrong is the fact that it harms conscious beings. Now, I grant that this is a very general statement, but I don’t think that I need to be any more specific (except by way of illustrating what I mean with some examples) to establish my point.
I made two claims in the previous paragraph: (1) God’s commands do not have anything to do with what makes an action morally wrong. (2) God’s commands cannot have anything to do with what makes an action morally wrong. The second is obviously stronger than the first and while I think that the second can be established via a straightforward argument, I will not attempt to do so here. Rather, I will focus my efforts on establishing (1). To do this I will use Randal Rauser’s sample version of the DCT that he provided in the discussion I mentioned:
Sample divine command theory of ethics: God is maximally good and loving and has commanded his creatures to love one another. This divine command constitutes the moral obligation that all creatures have always to love one another.
Let’s call this view “Rauser’s Divine Command Theory of Ethics” (RDCT). This sounds fine and Randal’s point is that it is difficult to see how anyone could find anything objectionable about this view. I think it is objectionable because it misidentifies the morally relevant features of actions. So, what, on RDCT, makes it obligatory to love one another? Is the fact that creatures are deserving of love? Is it the fact that loving one another facilitates peace and harmony? Is it the fact that being loved is a necessary condition for a creature to achieve its highest good? All of the features mentioned in these questions are reasonable candidates for what makes it the case that we should love one another. But on RDCT, the answer to all of these questions is, No. What makes it the case that we are obligated to love one another is the fact that God commands it. Full stop.
This point cannot be overemphasized. In the discussion following his original post, Randal pointed out, correctly, that it is perfectly consistent with RDCT that God commands that we love one another because we are conscious agents with dignity. While that is true, the important thing is that, on RDCT, this is merely an account of the reasons why God commands love; it is not an account of what makes love right. To see this we need only notice that the fact that we conscious agents with dignity cannot, on any version of the DCT, by itself make it the case that we are obligated to love one another. All versions of the DCT claim that moral obligations are grounded in God’s commands such that, in the absence of a divine command with respect to a particular action, that action is neither obligatory nor wrong. All versions of the DCT are thus committed to the following dependence principle:
Dependence Principle (DP): Moral obligations depend for their existence on God’s commands. If God offers no commands with respect to a particular action, then that action has no deontic properties.
So, what does make it the case that we ought to love one another on RDCT? The fact that God commands it. Nothing else could make it the case that it is obligatory to love one another, on the DCT. As I said, this is a perfectly general point that applies to all versions of the DCT; the whole point of the DCT is that obligations are grounded in God’s commands. His commands make it the case that we have moral obligations. This is the problem that I want to focus on.
The problem is that, contrary to (DP), God’s commands in fact have nothing to do with what makes an action morally wrong. In my most recent blog post, I attempted a detailed account of the basis of moral obligation. The simple answer is that actions that cause harm are wrong because they cause harm to conscious beings, actions that help are right because they help conscious beings. Think about how odd (DP) is. It implies that in the absence of a divine command concerning whether we should not do it, rape is not wrong. That is ludicrous. Rape is wrong because it causes physical and mental harm and it interferes with the agency of a conscious person. End of story. The suggestion that God’s commands have anything to do with is, quite frankly, bizarre. I can’t see why anyone would think that. I can’t see why anyone would think that, in the absence of a command not to torture and kill small children, that torturing and killing small children is not wrong.
There are really two problems here: First, it seems pretty clear that acts like torture, rape, and murder are wrong because of the harm that they cause to conscious beings. Second, I can’t see how God’s commands could change anything. If it is not already wrong to torture children just in virtue of the fact that torture causes serious short-term and long-term physical and emotional suffering, then I don’t see how God could make it wrong just by commanding that we not do it. This is a pretty magical ability that God must have in order to take an action, which has no moral features in the absence of his commands, and make it wrong. How does that work?
So, the divine command theory just gets morality wrong. By claiming that what makes an action right or wrong are the commands of a perfectly loving God, DCT misidentifies the relevant moral features. And it doesn’t just sort of get things wrong, the error is radical. DCT claims that what makes actions wrong is not the harm caused by them; this is an extreme error. Now, this is obviously a problem in that it is a false theory, but it is also a problem because it leads to faulty moral reasoning. Let’s return to RDCT. It tells us that we are obligated to love one another. Now, I want to know whether I should give money to a panhandler on the street. What does RDCT tell us about that? Well, it claims that I need to love the panhandler. But what does that entail? Perhaps loving him means that I should give him any disposable income that I have until he is in a position to provide for himself. Perhaps if I love him, I just need to buy him a meal and send him on his way. Perhaps I should ask if he is a drug addict and then, if so, pay for his drug-treatment. Or perhaps if I really love him I should help him score his next hit. What should I do?
I am not claiming that the answer to this question is easy. I am only claiming that RDCT (or any version of the DCT) facilitates faulty thinking about it. How do I know what love requires? Well, since God, on RDCT is perfectly loving, I need to look to him. Well, then presumably, I need to find out what God commands me to do in this instance. Suppose that God commands me to ignore the panhandler. Is this impossible? I don’t know, but I don’t see anything self-contradictory about the following: God believes that each of us needs to be responsible for ourselves. Loving others entails not burdening them with our problems. Thus God commands that we not give money to panhandlers.
What is the problem here? What God commands is not relevant to what I should do. And to the extent that I worry about what God commands, I am not worrying about what is morally relevant. Further, how do I know what a perfectly loving omniscient being will command? For all I know, there are goods that are beyond my ken. Perhaps, from God’s perspective, loving a drug addicted panhandler entails torturing him until he gives up his drug habit. (For more on the fact that we cannot know that a loving God won’t command torture, read this.) But none of this is relevant. The relevant questions concern what I can do to help and avoid harming this person. That is what matters, and worrying about God’s commands only facilitates confusion. The DCT blinds us to the significant moral truth that, in difficult matters of moral concern, we should be focusing on what we can do to avoid harm and facilitate well-being.
Another example: Should I eat animals that have been raised and slaughtered in a modern factory farm? If the DCT is true, then I need to find out what God commands about this. There are a few problems here: First, how am I going to discern what God commands about eating animals? Second, this is a waste of time; I should be thinking about the effects of my actions on conscious beings. Third, it is all too easy to reach the wrong conclusion if all I am concerned with is what God commands. Again, knowing what God’s commands are is difficult. But it is all too easy to come up with rationalizations about what God commands. I can pull passages from the Bible in which God (allegedly) gives humans dominion over animals and use that as part of a justification for concluding that God would not command that we not eat animals (even those raised in factory farms). But this is faulty reasoning (and, at least according to me, the wrong answer); what matters is harm, not God’s commands.
Now, obviously any case of moral reasoning can be infected by self-serving motives; that is not unique to cases of reasoning about God’s commands. But I think that a theory that misidentifies the morally relevant features in such a radical way as the divine command theory does is especially prone to such problems. The problem is that God’s commands literally have nothing to do with the deontic features of actions. Now, if we had some perfect access to God’s commands, we might then use his commands as a guide to what is right since he, being omniscient, would know what is required of us. But, of course we have very imperfect access (if indeed we have access at all) to God’s commands, even on the assumption that God exists. If thus seems to be (a) a waste of time to worry about God’s commands and (b) in our interests to avoid focusing on God’s commands since such concern can lead us astray.
not not a philosopher 
8 comments
Comments feed for this article
May 12, 2014 at 8:17 pm
Matthew Flannagan
Jason, I guess my initial question is this, where does a divine command theorist have to deny that properties like harm *make* actions wrong.?
What standard DCT of Adam’s Evans, Craig and so on maintain is that the property of morally required by God is * identical* with the property of being required by God. Now, strictly speaking we don’t say of two properties that are identical that one *makes* the other come into existence. That would suggest that something can be the source of itself, which seems absurd.
So a DCT does not need to deny that harming others makes things wrong, or that things are wrong in virtue of having these properties. All they need to say to get that conclusion is that a loving and just God prohibits certain actions in virtue of the fact it has these properties, that’s not implausible, presumably a loving and just being would do that.
What a DCT needs to deny is that the property of “helping” or “harming” are *identical* with the properties of being morally required or prohibited. That seems to me plausible, moral requirements are after all “requirements” and it while its true we have requirements to not harm the suggestion that the property of not harming itself *is* (in the sense of identity) a requirement, just seems to me quite bizarre, natural facts about harming simply are not the sort of thing that can be a requirements, requirements are intrinsically prescriptive they tell us what to do, and natural properties such as “being poisonous” don’t by themselves tell us to do anything.
May 13, 2014 at 12:56 pm
jbthibodeau
Matthew,
You make an important point in this connection in your paper responding to Sinnot-Armstrong’s arguments against DCT; namely that makes’ is ambiguous. My point is that, on the DCT, obligations are constituted by God’s commands and thus, in the absence of God’s commands, there are no obligations. I find this to be implausible. In the absence of a divine command not to rape, rape still involves the intentional interference with the agency of a conscious being and still involves deliberate physical, emotional, and psychological harm. All of these things seem sufficient to make rape wrong (to constitute the wrongness of rape).
Now, you say that natural properties are not the sort of thing that can be requirements since requirements are intrinsically prescriptive. (Aside: David Brink said something connected to this in a seminar I took in grad school, which I’ve always found useful: “There are basically two responses to philosophical arguments: The ‘Of-Yeah?’ response and the ‘So-what?’ response.” I am going with the “Oh-yeah?” response here: Natural properties can’t be prescriptive? Oh-yeah? Who says?). Why can’t natural properties be prescriptive? Let’s take a closer look at this claim.
You believe that supernatural properties can be prescriptive. On your view, the property of being commanded by God constitutes an action’s obligatoriness. But this property, being commanded by God is just a property, as it the property of being harmful. The divine property is different in two ways: First, it is a supernatural property rather a natural one. Second, it is a property that involves the activity of a conscious agent. Neither of these properties seem very good candidates for explaining the prescriptiveness of moral obligations. Or, rather, neither of them seem to explain it any better than the natural property of being harmful.
Let’s look at these two factors in turn; we’ll take the second first. The fact that an action is commanded by a conscious agent is not sufficient to make it morally wrong; at least not in general. Indeed, it is easy to see, at least as far as human agents are concerned, the fact that an action is commanded by a conscious agent is quite irrelevant to what makes it wrong. Suppose Barack Obama commanded that New Zealanders stop talking with accents. Obviously his command would be ridiculous and could not make the action required. Even if, out of love, Obama commanded that New Zealanders give more of their income to international charities, the command would not make it obligatory to do so. The example is silly but the point is perfectly valid: in general, the commands of a human being do not constitute moral obligations. So, this gives us prima facie reason to suspect that the commands of any conscious agent cannot constitute moral obligations. Commands just don’t seem capable of the task that the DCT claims for them. To put the point differently: The property of being commanded by Obama is not prescriptive (it is merely descriptive), so, similarly, the property of being commanded by God is merely descriptive. Thus the onus is on the DCT defender to explain how God’s commands are different.
On to the other factor: the supernatural character of divine commands. I see no reason to think that supernatural properties are better candidates than natural properties for accounting for moral oblgiations. If the problem is the prescriptive character of moral obligations, then I want to know why supernatural properties are capable of having a prescriptive character while natural properties are not. If natural properties cannot tell us to do anything, then why can supernatural properties do so? It is not enough to say that supernatural features are not limited in the way that natural features are. We need to know what it is about supernatural features that enable them to do what natural features cannot.
So, I want to make two points: First, if it is true that natural properties cannot be prescriptive, then I see no reason to think that the same is not true of supernatural properties. If prescriptiveness is difficult to account for, we don’t solve the problem just by hypothesizing supernatural properties, do we? Second, I don’t agree that natural properties cannot be prescriptive.
The example you chose to illustrate your point is not a good one. The property of “being poisonous” is not a property of actions, but of objects, or stuff (for lack of a better word). Being harmful is a property of actions. And I see no reason why that property cannot be prescriptive. In the first person case, I recognize that the fact that an activity is harmful to myself is a reason not to engage in it. (Yes, this is a practical reason rather than a moral one, but even practical reasons have a prescriptive character.) The property of being poisonous is not prescriptive but that is because it is not a property of actions. So, the example gives us no reason to think that natural properties of actions cannot be prescriptive.
May 14, 2014 at 4:30 am
Matthew Flannagan
Jason,
The word “makes” is ambiguous in these contexts. Normally when we saying one thing A makes another thing, B, to be the case. We think A and B are distinct things. A is prior to B, and B comes A about because of B. So for example if I state “the pressure of against the glass made it shatter” No one would take this statement to be saying that pressure against the glass is identical with shattering the glass, this would confuse cause and effect. What we are saying is that two distinct things happen the pressure on the glass and the glass smashes and the second thing happens in virtue of the former.
In other contexts however we use the word more loosely in a constitutive sense. So for example if I say “what makes it the case that John drinks a glass of water is that he drinks a glass of H20” this does not suggest that water and H20 are distinct things the former existing prior to the latter and causing the latter to exist. We are rather pointing to the sense that they are the same ( an identical) thing.
Now take your statement the statement “rape is wrong *because* it involves the intentional interference with the agency of a conscious being and still involves deliberate physical, emotional, and psychological harm.” Your suggesting this is an obviously true claim and a divine command theorist must deny it.
I disagree, If we take this as an uncontroversial statement ( and I can think of counter examples which suggest its not) the question arises as to in what sense is it uncontroversial. I think it’s clear that its not uncontroversial in the constitutive sense. When a person says this they are not saying wrongness is identical with the physical properties in question. They are rather saying that the property of wrongness supervenes on these physical properties in some way. They are suggesting that, there is a necessary connection between these properties and wrongness so that they are coextensive and latter exists in virtue of the former. And in this sense of the word a divine command theorist can agree entirely. He can say in all possible worlds where harms like this exist a loving and just God would prohibit those actions.
Of course the divine command theorist cannot hold that in any possible world the property of wrongness is identical with these properties. But I doubt that, in that statement ““rape is wrong *because* it involves the intentional interference with the agency of a conscious being and still involves deliberate physical, emotional, and psychological harm.” Is obviously true, in *that* sense of the word because.
Ok two things here.
First, note this counter factual test is compatible with both reading “because” in a supervenience sense or a constitutive sense. If the natural and deontic properties in question are distinct properties which are necessarily co extensive and the latter exist in virtue of the former. It will still be the case that it’s impossible for these natural properties to exist without the deontic ones existing. So, one can accommodate the insight your proposing here in a way entirely compatible with DCT.
But second, looking at the counter factual itself here I am not sure the counter factual you mention is false. Suppose there is a possible world where God fails to prohibit rape this will either be due to the fact (a) God does not exist in that world or (b) God exists and does not issue the command in that world.
Now if the former (a), is true and God does not exist in the world in question, then no physical properties will exist in that world either . Theists believe that God is the creator and sustainer of the universe, and so all physical properties depend on him for their existence. Hence a world in which God does not exist and issue commands is a world where the natural properties of causing harm don’t exist either. I don’t think its problematic to claim that in a world where no conscious agents exist no actions are wrong. This is because in such a world there are no actions.
On the other hand if (b) is the case and God does exist but fails to issue the command then we have a hypothetical situation where a being who is fully informed ( omniscient) perfectly rational, loving, just and impartial could permit a rape. But in *that* context it’s not obvious that rape would be wrong. What seems counter intuitive is not that rape would be permissible in this context, but that there is a it’s possible world that a loving and just omniscient person could ever permit such an action.
Finally let’s look at the issue of prescriptivity, First, I agree that supernatural properties are not in and of themselves better candidates for being prescriptive than natural ones. Its clear to me that some natural properties can be requirements. For example there are legal requirements and requirements of reason and these are constituted by natural facts . My point is that causal properties actions like “causing harm” are not identical with requirements of any sort, we can have requirements to not do actions which have these properties but they are not themselves a requirement.
Where I disagree is your second claim that activity of “conscious agents” is not a good candidate for prescriptivity here is why: persons are paradigmatic cases of beings that can make requirements upon others. People can require things of other people, persons can tell others what to do, make demands on peoples conduct conduct, prescribe people’s behavior and so on.
But the kind of natural properties you’re referring to cannot do these sorts of things. The example I used came from Richard Garner’s article “On the genuine queerness of moral properties and facts” published in the Australasian Journal of Philosophy,68:2
I don’t think your two examples refute this. Take first the Obama case you note correctly that if “Barack Obama commanded that New Zealanders stop talking with accents” that would not be sufficient to make the action *morally* required. I agree, but that does not mean that Obama could not under certain circumstances require new Zealander’s to do this or demand they did. Suppose the US annexed New Zealand and Obama made a federal order prohibiting New Zealand accents, then New Zealanders would be required by law to refrain from speaking accents. We would not be morally required to do this, but that does not mean that Obama did not make a demand upon New Zealanders or prescribe their behavior or require them to do something. It’s just that his simply that his requirements lack the requisite authority, we don’t have a compelling reason to follow them. So, this example does not dispute my point that the conscious agents can make demands or require things of people, where as unconscious causal properties like “harming” can’t prescribe or require things. “Harm” is not that sort of thing.
Your second example actually underscores my point. You write
In this example, the harmfulness of the action does not prescribe or demand you not do it. What , what prescribes or requires you to refrain from is your practical reason. In other word’s, it’s what you a conscious person, yourself prescribe in light of rational reflection on certain facts and goals, or what you ( a conscious person) would prescribe if you were informed or reflected correctly. Consequently, the property of being required is not identical with the property of being harmful, it’s a distinct property constituted by facts about what rational agents prescribe. When you say the action is required *because* its harmful, the “because” is not being used in the constitutive sense, it refers to a distinct property which is the basis for the requirement but not the requirement itself.
However, also this is exactly the same sense in which a DC theorist would can claim that things are required because they are harmful. The fact something is harmful means that some body else a rational person ( God) would place requirements upon people behavior to refrain from it.
May 16, 2014 at 12:51 pm
jbthibodeau
There are two different issues upon which we disagree: First, that God’s commands constitute moral obligations. Second, that natural properties such as the propensity to cause harm are the right kind of property to generate moral obligations. Each of these issues has several different aspects that are important to identify and keep as separate as we can. I am going to start by talking about each of these issues as they connect to the issue of prescriptivity. In a second comment to follow, I will more fully address the issue of whether the DCT is compatible with our moral intuitions.
The prescriptive character of obligations: I’ve made two claims that I want to defend in this connection: (1) The commands of conscious beings are not the right kind of thing to constitute moral obligations, (2) The capacity to cause harm is the right kind of thing to constitute moral obligations.
(1) You correctly point out that persons are paradigmatic kinds of things that can place requirements on others. With this I completely agree and indeed it is a large part of the reason that I think that harmfulness makes such a good candidate for accounting for prescriptivity (more on this below). However, my claim was not that persons cannot impose requirements upon us, but rather than the commands of persons (or conscious agents, as I initially put it) are poor candidates for accounting for the precriptivity of moral obligations. The example that I used was supposed to show that, in general, commands do not generate obligations. Now, you are quite obviously right that, in some circumstances, commands do generate obligations (but that doesn’t change the fact that, in general, they don’t). Two things about this, however: as you note, typically the requirements generated by human commands are not moral requirements but rather legal requirements. Obama’s fictional command to New Zealanders to stop speaking in non-American accents would generate a legal requirement at best. But, as a matter of fact, Obama couldn’t do this. Even if somehow New Zealand became a part of the United States, Obama’s command would be ineffectual because the President is not a legislator; he cannot create laws. This is not just a minor point; the larger and relevant point is that, at least post-Enlightenment, the world has come to believe that the authority to create legal requirements can only reside in a duly elected legislator or legislative body. This shows us that, typically, human commands are not the kind of thing that generate even legal requirements. So, and this is the second point, atypical and specific factors must be in place in order for the commands of any person or persons to constitute legal obligations. This is true in the case of governmental regulation and in the armed forces, the two places where commands (or command-like utterances) typically do generate requirements.
Now, it is also important to note that the special factors that must be in place for commands to generate requirements are institutional in nature. The order of a colonel places his subordinates under the obligation to obey the order because the colonel occupies an institutional role. And the force of authority that his command enjoy stems not from his own person but from the institution to which he and his subordinates belong. Similar things can be said of the authority of legislative bodies: There are institutions that grant legislators their authority; the authority resides not in the person(s) of the individual legislator(s) but in the institutional office that the individual(s) occupy; and the scope of the authority is always limited by the scope of the institution.
So, in general, human commands do not typically generate any kind of requirement; when they do, the requirements are legal requirements at best; and they require the societal infrastructure of pre-existing institutions. God’s commands are thus not good candidates for generating requirements of any kind since (a) most commands of persons don’t generate requirements, (b) at best it would seem that they could generate legal requirements, but (c) there are no institutions that grant God the authority to issue commands that generate requirements.
(2) I think that the capacity to cause harm is the right kind of property to generate moral requirements. Here I want to make one point: the analysis that I am recommending, according to which (roughly) the fact that an action causes harm to conscious agents is enough to make it wrong, offers an understanding of morality as concerning precisely situations in which persons (conscious agents) impose requirements on others. That is to say, if what makes an action wrong is the fact that it is harmful to conscious agents, then moral obligations really are imposed on us by other persons. Moral authority is, on my view, ultimately grounded in persons.
But notice that the authority of persons is not institutional in any way, and thus it does not extend to commands (And this accords with the fact that special circumstances are required in order for commands to generate requirements). Agency and the capacity to be harmed are natural and essential features of persons. Persons impose obligations on others by there shear existence, full stop. Notice also that this level of description enables us to respond to Mackie’s queerness objection. Mackie is correct that most natural properties do not and cannot impose imperatives on us. But not all natural objects are created equal and not all natural properties are created equal. Rocks and trees, colors, sizes, and shapes, cannot impose requirements; but that is because they are not the kind of thing that matters morally. When we reach the level of persons who are conscious beings with agency, things change dramatically. Conscious states have morally relevant features, agency is a morally relevant capacity. So, it should be no surprise that conscious agents have the capacity to generate moral requirements. At the appropriate level of description, the prescriptivity issue really is no problem at all.
May 17, 2014 at 12:32 pm
jbthibodeau
The counter-intuitive consequences of DCT: In the original post, I argued that the DCT is wrong, in part because it asserts the Dependency Principle (DP) [Moral obligations depend for their existence on God’s commands. If God offers no commands with respect to a particular action, then that action has no deontic properties.] My argument concerning DP has two aspects: First, I argued that DP is implausible because it denies that what makes an action wrong is the harm it causes. Second, I argued that it is difficult to see how, if an action such as rape was not already wrong (in virtue of the harm caused), it is difficult to see how God’s commands could change anything. My most recent comment goes part of the way to establishing the second of these two points.
Part of your criticism of my analysis, as I understand it, has involved a criticism of DP. In particular, you have pointed out that a DCT defender can accept that, in one sense of ‘makes’, the harm of rape makes rape wrong. You’ve pointed out that the DCT is only committed to the claim that the harm of rape constitutes the wrongness of rape. And you’ve provided some reasons for thinking that the harm of rape cannot constitute the wrongness of rape (to which I responded in my most recent comment). You’ve also addressed the counterfactual part of DP. You’ve given reasons for thinking that the counterfactual does not have any counterintuitive consequences; in particular you’ve asserted that a DCT defender can assert that there are no possible worlds in which rape occurs but is not wrong. So, in this comment, I want to address the counterfactual and the counterintuitive consequences that I believe follow from it.
First, I think that it is worth pointing out that any defender of any moral theory can maintain that the expressions ‘obligatory’ and ‘commanded by God’ are necessarily coextensive (and the same with ‘wrong’ and ‘forbidden by God’). For example, a utilitarian who is a theist can certainly maintain that God commands all and only obligatory actions and forbids all and only wrong actions, that God’s nature (in particular, his moral perfection) is the same in all possible worlds, and that God exists in all possible worlds. Such claims have the requisite consequence. Let’s call the conjunction of utilitarianism with these claims about God ‘Theistic Utilitarianism’ (TU).
So, both TU and DCT each has the consequence that ‘wrongness’ and ‘commanded by God’ are necessarily co-extensive. So what differentiates them as theories? Obviously it is that DCT asserts DP while UT denies it (and instead asserts some other principle of dependency). It seems to me, therefore, that a defender of DCT, in order to defend the aspects of his theory that makes it distinct, must accept DP and its consequences.
Now, one of the apparent consequences of DP is that there are possible worlds in which rape is not wrong. This is something that defenders of DCT have denied is a consequence of DCT. One way to deny this is to maintain that since God exists in every possible world there are no worlds in which rape occurs and is not wrong, according to DCT. I am not sure that this is precisely your view since, in your comments you acknowledge the possibility of a world in which God does not exist. But I want to address it and argue against it because I think that, in any event, what I say will have a great deal of relevance to your view.
First, the most natural way to test a dependency claim (of the nature of DP) is to try to construct possible worlds in which the properties/objects, which are allegedly dependent, are present but the object/properties upon which they are allegedly dependent are not. If we succeed in constructing such a possible world, then the dependency claim is false. To construct such a world requires that we be allowed to suppose that, in the imagined world, the object/properties upon which the dependent features depend do not exist. In the particular case of DP, this entails that we imagine a world in which God and/or his commands do not exist. Since, as I say, this is the standard/best way to test dependency claims (at least of a metaphysical nature), there must be a strong presumption in favor of allowing, as a possible world, a world in which God and/or his commands do not exist.
Now, as I indicated, you do allow for this since you claim, “Theists believe that God is the creator and sustainer of the universe, and so all physical properties depend on him for their existence. Hence a world in which God does not exist and issue commands is a world where the natural properties of causing harm don’t exist either.” So, I think you agree that we can imagine worlds in which God does not exist and use our reasoning about such worlds in our evaluation of the DCT. However, notice that this is not the only possible world that is relevant to our inquiry concerning the implications of DP. Let’s consider a world in which God exists but does not issue any commands. In this world, God disapproves of rape but does not issue any commands about it. Why would God not issue any command? Well, perhaps he believes that the cognitive distance between himself and his creation is so vast that any attempt at communication would be ineffectual and/or that there are no people with whom he can adequately convey his commands. Surely such a world is possible. But, on DCT, in this world, rape is not wrong because God does not command that we not rape.
Next, consider your claim that God is the sustainer of the universe. I take it that you believe that this claim entails that there are no possible worlds in which actions with natural properties such as harmfullness exist and in which such actions are not wrong. I don’t think that this inference is defensible. Consider the following claim, which I’ll abbreviate as (S): in every possible world in which there exists a physical universe, there also exists a sustainer of that universe. S does not imply that God exists in every possible world nor that there are no possible worlds in which a physical universe exists and God does not exist. So, I think we can consistently imagine a world in which God does not exist and in which actions that cause harm exist. We can do so even acknowledging that S is true. This world, call is WΦ, has a sustainer (some supernatural being other than God), and contains a physical universe and contains creatures that are capable of engaging in actions that cause harm. According to DCT, rape is not wrong in this world.
WΦ is a genuine possible world. It is consistent with the claim that God is the sustainer of the (actual) physical universe and with the claim that every physical universe requires a sustainer. However, in WΦ, rape is not wrong. This is a consequence of DP. Since DCT is committed to DP, DCT has the counterintutive consequence that there are possible worlds in which rape is not wrong.
May 21, 2014 at 6:17 am
Matthew Flannagan
Jason
Let’s look first at the issue of prescriptively. Here you state the commands of people do not generate obligations or requirements independent of certain institutional facts. I’d make two responses here.
(1) As I read Mackie I think prescriptivity involves two things: First, it involves a “call for action or refraining from an action” it in your word “imposes an imperative” on us. Second, is that it provides an overriding reason for action so that the action is something we have to do
As I see it the word, “requirement” can be used in both senses. Mackie tends to conflate these but they seem to me to be separate. My son for example often demands for me to give him my computer. This involves a demand or call for action and issues an imperative to me. However, it does not give me a reason to give him my computer or impose a rational requirement for me to give him the computer.
Now when I stated that moral requirements are requirements I meant to focus specifically on this first sense. Hence the quote from Garner which stressed that the. The genuine queerness of moral properties resides in their alleged “intrinsic imperativeness” there “action-guiding or action-directing” nature.
Now it seems to me that its clearly true that in that sense of the word, commands of conscious agents are paradigmatically prescriptive *in that sense* and are so quite apart from institution infrastructure. Consquently,the fact that Obama can’t, absent certain institutional facts, generate a binding legal requirement is beside the point, because even absent such facts he can call for new Zealander’s to do something and issue demands, they may not be legally binding on us but that doesn’t mean he can’t make them.
My point is the existence of harm or suffering, by itself intrinsically, does not issue an imperative. A however command does. The fact that harm exists can tell us what to do if there is an imperative already in existence which prohibits some action class which the harm falls into. But it does not itself prescribe anything
(2) That said, I am unconvinced by your suggestion that the imperatives imposed by agents cannot, absent institutional facts generate requirements in the second sense either. I cited one example in my previous post, requirements of practical reason are as I see it constituted by imperatives laid down by a person. Standard accounts of practical reason envisage an requirement of practical reason as a imperative that you would issue your actual self if you were fully informed and rational. What a fully informed rational counterpart of yourself would tell you do do. This suggests to me that in fact conscious agents *are* paradigmatic examples of what imposes practical requirements upon people. In theory a person can impose requirements on people’s behavior provided they are informed rational and have goals that fit the right way with the interests of the person.
May 25, 2014 at 7:27 pm
Staircaseghost
“What we are saying is that two distinct things happen the pressure on the glass and the glass smashes and the second thing happens in virtue of the former.”
… where “in virtue of” is cashed out in terms of the behavior of our model given counterfactual data.
When people try to discuss modal concepts, there is a constant danger of confusion when they are defined circularly e.g. defining cause using other modal concepts in the definiens like “because of” or “in virtue of” or “makes”.
“I think it’s clear that [Jason’s theoretical identification] is not uncontroversial in the constitutive sense. When a person says this they are not saying wrongness is identical with the physical properties in question. They are rather saying that the property of wrongness supervenes on these physical properties in some way.”
Like you, I don’t think Jason’s proposed intertheoretic identification is accurate. And I agree that when someone says something like “moral properties are constituted by X-properties” that they are saying at least that the former supervene on the latter. But the formal relation of supervenience is inclusive of the identity relation — every set of A-properties identical to B-properties supervenes on B-properties, and vice versa — so I disagree that you can automatically conclude that people making this kind of metaethical supervenience claim aren’t making an identity claim.
“They are suggesting that, there is a necessary connection between these properties and wrongness so that they are coextensive and latter exists in virtue of the former.”
This is where the modal confusion starts to kick in. Once you apply the formal notion of coextension, “in virtue of” becomes formally meaningless. To say the glass shattered in virtue of the pressure just is to say that the model’s outputs given that amount of pressure always include shattered glass. Thus, the “in virtue of” claim adds no new information to a pre-existing coextension claim.
“And in this sense of the word a divine command theorist can agree entirely. He can say in all possible worlds where harms like this exist a loving and just God would prohibit those actions.”
There is no “this sense”; it is equivocal. By “possible worlds” you either mean a) logically possible, in the sense of coextensivity by way of analytic identification or b) nomologically possible, the “in virtue of” sense. Choose B, and you lose the strong, constitutive, necessary coextension. Choose option A, and you lose any meaningfully informative sense of “in virtue of” or “makes”, as well as committing yourself to the highly improbable claim that any talk of gods commanding otherwise would be not only false, but literally contradictory.
Ah, but didn’t you specifically say you were talking about a “loving and just” god? Definiendum, meet definiens. Of course only a morally good lawgiver gives morally good laws, just as only gazorpazorp lawgivers give gazorpazorp laws!
But DCT is supposed to say these commands are moral, full-stop, which is why Jason’s evil-god objection has its force. Restricting LPWs with good laws to LPWs with good lawgivers makes for a meaningless and uninformative tautology. No one who did not already know the rules for correctly deploying the concept of moral wrongness would be able to correctly deploy the concept of “worlds where a loving and just God would prohibit X” to separate them from LPWs with evil gods. You are simply instructing him to condemn gods who give evil commands, but presumably he knew one ought to do that already.
This is why it’s clear to me as an Expressivist that theists and naturalists alike should give up trying to make intertheoretic identifications between moral properties and any purely descriptive properties. The act of ascribing moral properties always involves an ineliminably noncognitive component of approval or disapproval, and any* purely descriptive predicate (natural or supernatural) is ex hypothesi going to be logically and empirically compatible with taking any emotive attitude towards it one fancies.
[*] Barring trivial cases of self-reference like “it is true that I approve of X”.
May 26, 2014 at 6:29 am
Matthew Flannagan
Staircase ghost
1 You write ” When people try to discuss modal concepts, there is a constant danger of confusion when they are defined circularly e.g. defining cause using other modal concepts in the definiens like “because of” or “in virtue of” or “makes”.” I agree this would be circular if I was trying to give a reductive definition of modal terms, but I wasn’t. I was simply describing two possible ways to interpret the word “makes” using terms known to the reader, no attempt at a formal definition was attempted.
You go on to suggest the term “in virtue of” is “formally meaningless” as it “adds no new information to a pre-existing coextension claim.” Here I simply disagree, the phrase “in virtue of” adds to co-extension the idea that there is an asymmetric dependence relationship between the properties in question and that is not included in the pre-existing co-extension claim. Co-extension is compatible with symmetrical relationships and relationships where there is no dependence relationship between the properties in question.
2. Regarding Supervenience you write that I cannot “automatically conclude” that can people “making this kind of metaethical supervenience claim aren’t making an identity claim.” Because “the formal relation of supervenience is inclusive of the identity relation” But I didn’t automatically conclude those making this claim aren’t making an identity claim. What I said was if the claim is to the uncontroversial and obvious claim Jason suggests it is, it can’t include the identity claim. Moreover, I did not “automatically” conclude this. I provided an argument that the identity sense was implausible.
(For what its worth, its not given that , in this context “the formal relation of supervenience is inclusive of the identity relation” one important argument against divine command theories, proposed by Mark Murphy is based on the idea that supervenience moral properties supervenience on non-moral properties, and hence the properties in question are distinct. )
3. You suggest that by possible worlds I must mean either (a) logically possible, in the sense of coextensivity by way of analytic identification or (b) nomologically possible. I am actually not convinced that analyticity and nomological possibilities are the only two options here. But putting that to one side, I am suggesting the relationship is analytic in sense (a) You suggest doing this means I “you lose any meaningfully informative sense of “in virtue of” or “makes”, and commit myself to the “the highly improbable claim that any talk of gods commanding otherwise would be not only false, but literally contradictory.” This is however false. First, given God is essentially loving and just compassionate it *is* contradictory to talk of God commanding certain types of harmful actions, possession of such character traits is logically incompatible with commanding certain types of actions.
Second, a loving and just person would not just fail to command certain things he would be motivated to do so because they had these properties its having these properties would be the motivation and practical reasons he had for the command. Hence, there is a meaningful sense in which such a being would issue certain commands in virtue of its natural properties.
4. This brings me to your last criticisms. You state I a state this is a case “Definiendum, meet definiens. Of course only a morally good lawgiver gives morally good laws, just as only gazorpazorp lawgivers give gazorpazorp laws!” But that’s just false. If I had claimed that moral goodness was defined in terms of what a morally good person commanded you would have a point. But no divine command theorist does this, they identify specifically “deontological properties” of being obligated, being required, being wrong. With the commands of a person who has ( among other things) specific character traits such as being loving, just, impartial and so on. There is nothing circular in this. Its only circular if you mistakenly assume that the concept of being loving, or compassionate is the same as the concept of being morally required. Similarly your claim that “But DCT is supposed to say these commands are moral, full-stop” Is simply false false almost no DCT attempts to say Gods commands are *moral* full stop they purport to give an account of a specific type of moral property, specifically deontological properties. This is true of both historical and contemporary versions of the theory.
Finally your last objection that “No one who did not already know the rules for correctly deploying the concept of moral wrongness would be able to correctly deploy the concept of “worlds where a loving and just God would prohibit X” as far as I can see carries little force. That objection notes that people must already *know* a lot about morality and understand the concept of right and wrong to be able to identify whether a command comes from a loving and just God. But that only shows that our concepts of right and wrong are epistemically prior to our knowledge of whats loving and just. A DCT does not deny this it asserts rather that moral obligations are ontologically dependent on God.
This simply confuses epistemic with ontological priority. One thing A can be ontologically dependant on B, while our knowledge of B is dependent on A. I have to for example already know a lot about water and how it works to test the hypothesis that water is H20, that doesn’t disprove that there is an ontological dependence of water and H20.
Finally, I don’t agree that “the act of ascribing moral properties always involves an ineliminably noncognitive component of approval or disapproval” It seems to me you can get amoralists and paradigmatically evil people who express approval or indifference towards acts which are morally wrong.