Divine Command Theory and the Metre-Stick Analogy (Part One)

I’ve written about theistic metaethics and the critiques thereof before. In doing so, I think I’ve explored many of the best contemporary arguments for and against the particular brand of theistic metaethics known as divine command theory (or, more generally, “theological voluntarism”). Given this, and given that this is a debate that has raged on for a long time, one might be forgiven for thinking there is no need to write about it any further. But that’s naive; there is always more to learn and more ways in which to refine and articulate the arguments.

In this regard, Jeremy Koons’s recent paper “Can God’s Goodness Save the Divine Command Theory?” caught my eye. In it, Koons takes up one of the most popular contemporary defences of DCT, and subjects it to a fairly withering critique. Over the next couple of posts, I want to explain this critique, both to further my own understanding and to share it with my readers.

The critique is targeted directly at William Alston’s particularist version of the divine goodness solution to the Euthyphro dilemma. Koons argues that Alston’s defence falls considerably short of the margin because it saves theistic metaethics at the expense of stripping the notion of “goodness” of all its significance, thereby opening the door to a revised version of the Euthyphro dilemma. This is a significant conclusion since Alston’s solution to the Euthyphro dilemma has won some admirers over the years.

To set up the discussion and analysis of Koons’s critique, this first post will outline the basic dialectical thrusts and parries in the literature. It does so by first by looking at the classic Euthyphro dilemma and the divine goodness solution to that dilemma. It then outlines Alston’s specific conceptualisation of divine goodness, and considers some of the problems it raises.


1. Theological Voluntarism, Euthyphro’s Dilemma and Divine Goodness?
The world consists of entities, activities and states of affairs. Any actually existent entity, activity or state of affairs can be referred to as a “fact”. A peculiar feature of the world is that some of facts have moral statuses attached to them. These statuses come in two distinct flavours: (i) value statuses, which ascribe the property of goodness/badness to things; and (ii) deontic statuses, which ascribe the property of rightness/wrongness to things. Metaethics is the branch of moral philosophy concerned with explaining why facts have the morally statuses that they do. In other words, it tries to explain what grounds moral facts.

Proponents of theistic metaethics contend that moral facts cannot be grounded in the absence of God. Theological voluntarism is probably the most widespread and popular form of theistic metaethics. Theological voluntarists argue that moral statuses are explained by reference to one or more of God’s voluntary acts. Thus, for example, they might say that giving money to charity is good because God commands it, or desires it, or wills that it be good. Similarly, they might say that murder is wrong because God forbids it, desires that we not do it, or wills that it be not done.

Theological voluntarism runs into problems when it confronts the Euthyphro dilemma, which was first presented (in a somewhat odd form) in the Platonic dialogue the Euthyphro. The dilemma takes a particular moral status (say, the wrongness of killing) and asks: is killing wrong because God forbids it, or does God forbid it because it is wrong? If it’s wrong simply because God forbids it, then it seems like the wrongness of killing is arbitrary, something that God could reverse by changing his mind (which he has the power to do). Conversely, if God forbids it because it is wrong, then it seems like its wrongness is independent of God, something that is not welcome to the theist.

This gives us an argument against theological voluntarism, as follows:

• (1) Either God commands X because it is right/good, or X is right/good because God commands it. 

• (2) If God commands X because it is right/good, then rightness/goodness is independent from God. 

• (3) If X is right/good because God commands it, then the rightness/goodness of X is purely arbitrary since God could command anything. 

• (4) If theological voluntarism is to succeed, then rightness/goodness cannot be arbitrary and cannot be independent from God. 

• (5) Therefore, theological voluntarism does not succeed.

Many challenge this argument by disputing the truth of premise (4), claiming that it does not matter too much if moral statuses are arbitrary. William of Ockham is famously said to have adopted this view. Others will challenge premise (1) by arguing that it presents a false dilemma because there is a third option that does not lead to the pitfalls mentioned. But by far the most popular contemporary solution to the dilemma is to challenge premise (3) by limiting the explanatory role of theological voluntarism to the deontic realm.

The solution works as follows. Theological voluntarists maintain their commitment to the view that moral rightness is explained by direct reference to one or more of God’s voluntary acts, most commonly his commands. Thus, killing is wrong if and only if God forbids it. But they then skirt the arbitrariness horn of the dilemma by arguing that God could never command anything that was morally abhorrent because God is essentially good. He is perfectly loving, perfectly kind, perfectly just (and so on). Goodness supervenes on these characteristic properties and they prevent him from commanding anything that is wrong. The basic idea is illustrated below.

This resolves the Euthyphro dilemma, at least as it pertains to the explanation of moral obligations. But it seems to raise the possibility of a revised Euthyphro dilemma. It is in responding to this revised dilemma that Alston makes his main contribution to the debate. So let’s look at that next.


2. The Revised Euthyphro and Alston’s Solution
The revised version of the Euthyphro dilemma runs as follows. Those who favour the modified version of theological voluntarism that I just outlined claim that God is essentially good, and that this goodness supervenes on his character traits (lovingness, kindness etc.). But then the question arises: does this mean that God also provides an ontological grounding for moral goodness? Two options present themselves. Either God’s characteristic traits are good because they’re God’s, or God is good because he has those particular traits. If it’s the former, then we run into a significance problem (discussed more below). Contrariwise, if it’s the latter, then it seems like goodness is ontologically independent from God, which is unwelcome to those who seek to ground all moral statuses in God.

If we limit ourselves to the trait of “lovingkindness” we can pose the following argument:

• (6) Either the property of lovingkindness is good-making because God happens to exemplify it, or God is good because he exemplifies the property of lovingkindness. 

• (7) If lovingkindness is good-making because God happens to exemplify it, then any property could be good-making if God happened to exemplify it, which suggests that God’s goodness is an insignificant fact. 

• (8) If God is good because he exemplifies the property of lovingkindness, then good-making properties are ontologically independent from God. 

• (9) In order for there to be a successful theistic metaethics, two conditions must be met: (i) all moral statuses must be explained by God; and (ii) God’s goodness must be a significant fact. 

• (10) Therefore, there can be no successful theistic metaethics.

The most important point to bear in mind is that this argument is subtly different from the preceding one. Both share a concern for the ontological dependence of moral statuses on God, but while the classic Euthyphro dilemma was concerned with the potential arbitrariness of moral rightness, this dilemma is concerned with the significance of God’s goodness. In other words, it is saying that the theistic claim that God is good, or that God serves as the ultimate standard of goodness, must be a significant claim. And that it can’t be a significant claim if good-making properties are simply explained by the fact that God exemplifies them.

What Alston does in his work is offer some clarification of the grounding relationship between God, the good, and the divine properties, before then moving on to show how this relationship does not make God’s goodness an insignificant fact. As we shall see, he is ultimately unsuccessful in this project, but it’s worth identifying exactly where his account goes wrong since it has seemed attractive to many theists (including William Lane Craig who has explicitly endorsed it).I’ll start by just sketching Alston’s account of the grounding relationship. In the next post I’ll present Koons’s criticisms of it.

Alston argues, importantly, that good-making properties such as lovingkindness are not ontologically independent from God. Instead, God is the ultimate standard of goodness, and his being the ultimate standard of goodness explains why the properties he instantiates are good-making. Thus, Alston is clear that the second horn of the dilemma (the independence horn) is avoided by his account.

But then what does it mean to say that God is good? Typically, we say that someone or something is good or bad in virtue of the properties they exemplify; we do not say they are good and bad simpliciter. If we did, we would rob the predicate “good” of all its meaning. In other words, we wouldn’t be offering any deep justification or account of why a person or an act was good or bad, we would simply be assuming that they are and demanding that others accept this view.

Somewhat bizarrely, Alston thinks he avoids this problem. To quote:

Note that on this view, we are are not debarred from saying what is supremely good about God. God is not good qua bare particular or undifferentiated thisness. God is good by virtue of being loving, just, merciful and so on.(Alston “What Euthyphro Should have Said”, 2002 p. 292)

But this just sets up a circle (illustrated below). God goodness is explained by reference to his character traits, but the goodness of those character traits is explained by reference to the fact that they are the ones that God, as the supreme standard of goodness, happens to exemplify. This doesn’t make much sense.

Alston tries to quell these doubts by getting more sophisticated about the nature of predicates (such as “good” and “bad”) and the way they work in different contexts. First, he points out that there are at least two kinds of predicate with two associated criteria of application.

Platonic Predicates: The criterion for the application of a Platonic predicate is some general “Idea” or “essence”. An example might be the criterion for the application of the term “triangle”. Whether it makes sense to call a particular object or representation thereof a “triangle” depends on whether the object in question resembles the Platonic ideal of a triangle.

Particularist Predicates: The criterion for the application of a particularist predicate is determined by reference to one or more concrete individuals. An example might be the criterion for the application of the term “metre”. Whether it makes sense to say that a particular length is one metre long depends on whether it is isomorphic to the standard metre-stick which is kept in Paris.

Now, as it happens, the account of “metre-hood” offered here is historically inaccurate. It was not the case that a particular platinum-iridium bar was chosen to be standard against which all other purported metre-lengths would be judged. Rather, it was the case that a particular platinum-iridium bar was chosen because it most closely approximated the length of one metre (which was a measurement that was already being used). But this doesn’t matter. Suppose it was the other way round, then we would have an idea of what a particularist criterion for predicate application might involve.

The argument Alston then offers for our delectation is the following. It is perfectly coherent to suppose that “goodness” is a particularist predicate which is correctly applied by reference God. In other words, it is coherent to suppose that God is to goodness as the standard metre-stick is to the length of 1m.

But does this rebut premise (7)? Could it be true that God is like the divine moral metre-stick, and yet nevertheless it remains significant to say that God is good? We’ll try to answer that question in part two.