I’ve been carrying on a conversation with Matthew Flannagan about the arbitrariness objection to the divine command theory. You can find my first post on the issue here and Matt’s response here. In this post I am going to continue my defense, against Flannagan’s objections, of the arbitrariness argument (AA):

1. Either God’s commands are arbitrary or they are grounded in reasons.
2. Arbitrary commands cannot ground moral obligations.
3.  If God’s commands are grounded in reasons, then it is those reasons, rather than God’s commands, that ground moral obligations.
4. Either way, God’s commands are superfluous; they do not ground moral obligations.

A very brief summary of the conversation to this point. Flannagan, in a paper published in Philo [1], argued that AA depended on a false Principle PI: If A is constituted by B and someone has reasons, r, for bringing about B, then A is constituted by r. Flannagan has very good reasons for thinking that PI is false, and I agree with him that it is. However, I don’t think that AA [2] relies on PI. Thus, in my previous post I argued that, because of the rational nature of moral obligations, AA would be more plausibly read as relying on PR: If A is constituted by B, A is grounded in reasons and is itself a reason, and someone has reasons, r, for bringing about B, then A is constituted by r. Flannagan, in his response, provided a couple of examples that he claims are counterexamples to PR. So, in this post I will discuss his examples and argue that they are not counterexamples to PR.

Consider a second example drawn from Stephen Sullivan. Certain laws are constituted by an Act of Parliament. Laws constitute reasons for action, they tell us what we can and what we cannot do. There are reasons why citizens are under laws. Parliament has reasons for passing particular Acts of Parliament. However, it is clearly not the case that Parliament’s reasons can be informatively identified with a law. The reasons existed prior to that law. Even if Parliament irrationally ignored the reasons and did not make a particular law, the reasons for that particular law would still exist while the law would not.

Consider another example. Under certain circumstances a decision made by a Family Court Judge constitutes a Protection Order, judicial decisions constitute reasons for action and there are reasons why we have court decisions of this sort. Consequently PR entails that if a judge, reflecting on the evidence submitted, decides that he or she has good reasons for ruling a particular way then those reasons themselves constitute the Protection Order, prior to, and independent of, any judicial decision. But,that is clearly false. These reasons existed before the judge’s decision and would exist, even if the Judge ruled the other way, yet no Protection Order exists unless and until the Judge makes the ruling. So the reasons cannot be identical with the Protection Order.

While I think that these are excellent examples and very revealing of the underlying issues involved, I do not think that they are relevantly parallel to the case of moral obligation.

Let me take a step back and locate the current discussion in the larger debate concerning the divine command theory (DCT). As I indicated in my previous post, I am inclined to think that AA provides a good objection to DCT.  A defender of DCT must maintain that the reasons that ground God’s commands (to use the Flannagan’s helpful terminology, that motivationally explain God’s commands) are not sufficient, on their own, to ground moral obligations. But I think it is more plausible to maintain that the reasons that would motivate a maximally perfect being to issue commands ground moral obligations. And this entails that the commands themselves are superfluous, since, even if God does not issue commands (because, for example, God does not exist), the reasons that would motivationally account for his commands would still exist. [3]

Essentially, what I am suggesting is that it is plausible to identify a maximally perfect God with an ideal observer; and thus we can articulate a plausible ideal observer theory that identifies moral obligations with what would be commanded by a maximally perfect being. On such a view, a moral obligation is grounded in the reasons that would motivate an ideal observer/maximally perfect God to issue a command. Thus the actual commands of the ideal observer are superfluous; the reasons alone are sufficient to ground moral obligations. A defender of DCT, such as Flannagan, must reject such a view, because on DCT, the reasons that motivate God to issue commands are not enough to ground obligations; the commands themselves are required. Thus Flannagan uses examples in which a person or persons have reasons to bring about some state of affairs but in which the reasons cannot plausibly be thought to constitute the state of affairs brought about. Thus, by analogy, he will maintain that even though God commands on the basis of reasons, and in commanding brings about moral obligations, his reasons should not be thought of as constituting the moral obligations themselves.

I think that the examples that he offers are significantly disanalogous to the case of moral obligation and I will attempt to show that the examples are not counterexamples to PR. However, given my ultimate aim, that is, defending the Arbitrariness Argument, I am not very committed to defending PR. My point in the earlier post was to show that Flannagan’s criticism of AA was unsuccessful because AA does not depend upon PI . I suggested that AA could be defended by appealing to a different principle, namely PR, and that PR is not falsified by Flanngan’s (nor Sullivan’s) examples. I continue to maintain that none of the examples offered thus far falsify PR, but I don’t know that it ultimately matters whether PR is true. What is important is that, given the role that God’s reasons play in Adams’ DCT, it is reasonable to think that it is God’s reason, rather than his commands, that ground moral obligations. I’m not convinced that defending PR is necessary to show that.

On to the examples: It is significant that the two cases concern institutional facts. This is important because, according to the theory of institutional facts articulated by John Searle, institutional facts depend upon collective intentionality. [4] In other words, institutional facts are such that their existence depends upon collective beliefs and/or attitudes about them.

Searle’s point is best illustrated by an example: That certain slips of paper are US dollars is an institutional fact. But this fact depends on collective intentionality since a slip of paper counts as money only because we collectively agree to treat it as money. Now, a piece of paper must satisfy other conditions in order to count as money (it must be issued the US treasury, for example). But the point is that something only has the status of a US dollar only given that we collectively agree to treat it as a medium of exchange. To see this, consider the absurdity of an alien civilization (from Omicron Persei 8, say) coming to Earth to steal all our paper money. An alien civilization might be interested in our natural resources, but they would have no interest in our money since dollar bills, for example, could have no value in the alien society. [5] The aliens could not take our money back home and use it to buy all of the things that they ever wanted. Dollars have no value on Omicron Persei 8 because dollars only have value given our collective belief that they have value, and this is a belief that the Omicronians will not share. The upshot of this is that a dollar is partly constituted by the collective belief that it can be used as a medium of exchange. A piece of paper is worth one dollar, in part, because of this collective belief, and in the absence of the belief, the piece of paper has no monetary value.

Many of the same points apply to laws and judicial decisions. A law counts as a law only given the existence of a society that collectively agrees to treat it is a law. A judge’s decision counts as a legal requirement only because we treat it as such. We can see the role that collective intentionality plays in the case of laws when we consider the laws of dead civilizations. No visitor to modern Athens would ask why Solon’s laws are not being enforced. Apparently Solon’s laws forbade the export of all produce except olives. It would be silly to suggest, on that basis, that modern Greece needs to be punishing its grape exporters. The reason is that Solon’s laws are no longer laws and the reason for this is that Greek society has long ceased regarding them as laws.

This fact allows us, I think, to claim that Flannagan’s examples are not counterexamples to PR. Since laws and judicial decisions are institutional facts, they are partially constituted by the collective beliefs of the society in which they have the status of legal requirements. Given this, we can reject both of Flannagan’s examples as irrelevant to PR. PR is a conditional statement whose antecedent is the conjection “A is constituted by B and A is a reason and A is grounded in reasons and someone has reasons for bringing about B.” The first conjunct, “A is constituted by B” is the relevant part here. As a matter of fact, this is not satisfied in the Act of Parliament example nor is it satisfied in the Judicial Decision example because in both cases, due to the role played by collective intentionality, the A term is not fully constituted by the B term. A law is only a law given the collective belief to treat it as a law; that belief is thus partly constituitive of the law. It is therefore, strictly speaking, false to say that a law is (fully) constituted by an Act of Parliament. A judge’s decision is only a legal requirement given the collective belief to treat it as a legal requirement. Therefore it is false to say that a judge’s decision (fully) constitutes a legal requirement.

Notice that neither moral obligations nor divine commands are institutional facts nor are they constituted by collective intentionality. A genuine moral obligation is a binding requirement regardless of what any person or persons believe about it. And the same is true of divine commands; divine commands do not depend for their existence on collective beliefs about them.

Though I think that this is a fine and interesting point, I do not intend to lean too heavily on it. I don’t think that this difference, while a real one, gets at the heart of the important disanalogy between Flannagan’s examples and moral obligations. The important disanalogy concerns the role that reasons play in the examples. In the case of law and judicial decisions, reasons are not relevant to the legal status of the law or decision, but in the case of moral obligations, at least on Adams’ DCT, God’s reasons are essential to his commands’ status as moral obligations.

A law’s status as a law does not depend upon the reasons that the legislative body had for passing the law. Indeed, the reasons for passing a given law may no longer be endorsed by the legislative body at any given time subsequent to the passage of the relevant act, but that does not change the fact that it is still a law. The current US Congress in fact does not endorse the reasons supporting passage of the Affordable Care Act of 2010, but that is not at all relevant to its status as a law. There are probably many reasons for this divorce of reasons and status, but the one that I want to focus on is this: for pragmatic reasons, societies empower individuals or groups to make all kinds of decisions that have the force of law. While we want and expect these decisions to be reasonable, what matters, as far as the efficient and smooth functioning of society is concerned, is that the decision be made. Once there is a decision, it is the content of the decision (the content of the law, for example) that matters; the deliberative process that led to the decision is not relevant.

Now, that is not to say that, even in institutional contexts, the deliberative process might not become relevant at some future point. My claim is only that, in terms of what matters for legal purposes, what matters is only the content of a law (or judicial decision) not the reasoning in favor of the decision to pass the law (or issue the legal decision). A duly passed act of parliament has, in relevant contexts, the status of a law regardless of the reasons behind its passage. Another relevant point is that, especially when it comes to laws, which in democracies are typically established via actions of collective bodies, different individual members of a legislative body may have differing reasons for wanting a particular law. That those reasons are irrelevant to the status of the law as a law is, therefore, very important. The upshot is that, in the case of laws and judicial decisions, the reasons supporting a law or a decision are not relevant to their status as legal requirements.

This sort of thing is not the case with respect to the role that God’s commands play in the divine command theory, at least not Adams’ version of the theory. According to Adams, God must be perfectly loving in order for his commands to ground moral obligations. In Finite and Infinite Goods, for example, he says:

It is only the commands of a definitively good God, who, for example, is not cruel but loving, that are a good candidate for the role of defining moral obligation. (250)

That God is good and loving entails that he will be motivated by certain reasons and not others. His command that we not kill, for example, will be motivated by his love for us rather than, say, his dislike of the smell of decaying bodies. It seems to me that given Adams’ requirement that God is loving, God’s reasons are not irrelevant to the status of his commands as moral obligations in the same way that parliament’s reasons are irrelevant to the status of their acts as laws, or a judges reasons are irrelevant to the status of the judges’ decisions as legal requirements.

It therefore seems to me that, on Adams’ own view, God’s reasons are essential to the moral obligations that are created via his commands in that, if God acted on different reasons (that is, his reasons that didn’t flow from essential goodness), then his commands could not constitute moral obligations. Again, this is very different from the act of parliament and judicial decision examples. It is this significant disanalogy that I think makes these two examples offered by Flannagan irrelevant to the arbitrariness argument.

The upshot is this: in the case of God’s commands, God’s motivational reasons are essential to the status of his commands as moral obligations. But then, if that is right, why not think that the reasons themselves are doing the metaphysical work; why not think that it is the reasons that ground the obligations rather than the commands?

---

[1] Flannagan, “Is Ethical Naturalism more Plausible than Supernaturalism? A Reply to Walter Sinnot-Armstrong” Philo 15, no. 1 (2012): 19-37.

[2] Technically, what I should say is that what I called Version 2 in my previous post is a compelling version of AA and that it (Version 2) does not depend on PI. In this article, when I speak of defending AA, I am talking about Version 2.

[3] Robert Adams is certainly aware of this possible response to his view. See pp. 255-257 of Finite and Infinite Goods for his response to such a view.

[4] For details on Searle’s views, see his The Construction of Social Reality.

[5] Technically what we should say is that US dollars have no value as a medium of exchange in a hypothetical alien society. I suppose it is possible that they could have some other value (as rare and interesting artifacts, for example).