Craig on Objective Morality (Part Two)

(Part One)

You should probably read part one before attempting this, but if that seems like too much to ask, then be reassured that I’ll try my best to summarise the important parts of that discussion when it seems important to do so.

Anyway, I’ve been trying to work out exactly what Craig means when he says that God makes objective moral values and duties possible. As noted in part one, this is an important exercise since imprecision in what is meant by objectivity can be a way to both confuse and be confused.

Craig dedicates two paragraphs in his book Reasonable Faith to clarifying what he means by objectivity. Last time out, I subjected the first of these paragraphs to a close analysis and concluded that Craig seems to mean at least two things when he claims that an entity or property is "objective": (i) that it is mind-independent; and (ii) that it is non-relative. It is important to keep these two concepts distinct since they need not always go together.

In this entry, I’ll first look at paragraph two in more detail and then consider moral platitudes and metaethical theories in general.


1. Paragraph Two: Cognitivism
Again, for ease of reference, I’ll quote the paragraph in full:

“To say that there are objective moral values is to say that something is good or bad independently of whether any human believes it to be so. Similarly to say that we have objective moral duties is to say that certain actions are right or wrong for us independently of whether any human being believes them to be so. For example, to say that the Holocaust was objectively wrong is to say that it was wrong even though the Nazis who carried it out thought that it was right, and it would still have been wrong even if the Nazis had won World War II and succeeded in exterminating or brainwashing everybody who disagreed with them so that it was universally believed that the Holocaust was right. The of premise (1) [of Craig’s moral argument] is that if there is no God, then moral values and duties are not objective in this sense.”

In many ways, this paragraph just further elaborates on the distinctions and clarifications we discussed in part one. For instance, the concept of mind-independency is once again clearly to the fore, as is the idea that moral truths are not relative to particular sets of cultural beliefs and practices (e.g. those of the Nazis).

What is interesting about this paragraph is what it has to say about the truth value of moral propositions. It is clear that Craig thinks, in addition to being mind-independent and non-relative, moral propositions such as “the Nazis were wrong to carry out the Holocaust” have a truth value and, indeed, that that truth value is positive in this instance.

In some ways this observation probably ties in better with a discussion of premise (2) of Craig’s argument (in which he affirms the existence of objective moral values and duties), but I think it is worth making here for a couple of reasons.

First, not all metaethical theories accept that moral statements like “X is good” or “I have a duty to do A” have truth value. Such theories are non-cognitivist in nature. Examples include, universal prescriptivism, expressivism, and emotivism. These theories usually argue that a non-cognitivist approach is more plausible on semantic and ontological grounds, and that it can actually give us a good deal of what we want from morality in return for sacrificing the possibility of truth value.

Second, it is possible to believe that moral propositions are intended to refer to mind-independent phenomena, to be non-relative in their scope, and to have truth value, and still deny that they have positive truth value. In other words, it is possible to be a moral nihilist or error theorist and still accept Craig’s other claims about what is needed to have an objective morality. This would really be a rejection of premise (2) of Craig’s arugment. We call theories that accept the positive truth value of moral propositions “realist”.

Third, and most importantly, although it might be correct to say that cognitivism follows straightforwardly from mind-independence and non-relativity -- and so, one who accepts those claims automatically accepts cognitivism -- it is not necessarily true to say that one who rejects these properties rejects cognitivism. In other words, it could be argued that moral propositions are mind-dependent but still capable of having a positive truth value.

This is a significant observation since it seems to me that Craig’s argument as whole is intended to appeal to the fact that people really want to be able to affirm that certain things are good and bad ( or right and wrong). Like Craig, I think we all really want to be able to say that what the Nazis did was just wrong and that no amount of cultural brainwashing will make it otherwise. But if that’s what we want to be able to say, then it seems like all we really care about is positive truth value - mind-independency and non-relativity might be purely incidental.


2. What are the key moral platitudes?
So far I’ve been analysing what Craig has to say about objectivity. I now want to step back from this analysis and offer some more general comments on metaethics and successful metaethical theories.

Let's ask the question: What is so important about moral objectivity? Why does Craig latch onto objectivity in his moral argument? I hinted at this above, but to be more explicit now, I think Craig’s discussion is only interesting to the extent that it tells us something about the kinds of things we demand from a successful moral theory. These are sometimes referred to as the "platitudes" of moral discourse.

Craig clearly demands objectivity, but as we have seen his use of this term is imprecise. He could be referring to mind-independence, non-relativity, or cognitivism, or some combination of the three. It could also be that he has other demands that are more implied than explicit.

So let's not be too wedded to what Craig wants, let's focus instead on the kinds of things we all might want from a successful moral theory. I don’t know if I can do justice to all the demands we might place on a successful moral theory, but the following list seems to me to be fairly typical. I’ll define the terms as I go along (there’s going to be some overlap with what I’ve already said):

(a) Mind-Independence: We hope that moral properties exist independently of the desires and beliefs of human beings. If we wish, we can be even more discriminating in how we define mind-independency. We could, for example, distinguish between desire-independent and belief-independent theories. I won’t do that here, but it’s worth noting anyway. We could also distinguish between theories that make moral properties independent of our own personal minds and theories that make moral properties independent of all minds. 

(b) Absolutivity: We want moral propositions and claims to hold true for all people, at all times and all places. How absolute is “absolute”? That depends. We may include a number of contextual factors in our moral theories (e.g. “X is morally obliged not to kill another human being (Y) provided that Y is not trying to kill X), but we could hold that anyone who find themselves in a similar context is bound to affirm the truth of such a proposition. This would be a kind of absolute morality. To distinguish it from an even more absolute morality, we might like to refer to "strong" and "weak" forms of absolutivity.

(c) Practicality: We want moral properties to actually affect how people behave. In other words, if someone recognises that they have a moral obligation not to kill another human being (unless certain contextual factors apply), then we would expect them to acknowledge that they must act in a way that is compliant with this obligation. We could add an “overridingness” condition to this if we wish. This condition would stipulate that we expect the recognition of a particular moral status to supply a reason for action that overrides any other reasons for action that a particular agent may think he has. 

(d) Positive Truth Value: We want it to be the case that at least some moral propositions are true and at least some are false. We don’t want to be moral nihilists. 

(e) Non-trivial truth value: This might be a slightly unusual condition and so requires some explication. I’m adding it in here to exclude a certain type of moral theory. The theory I have in mind is one which satisfies our demand for positive truth value, but does so in a manner that is trivial or vacuous. As it turns out, this might be one ground upon which to object to a theistic metaethics, but it is not prejudicial against this position. It can apply equally well to non-theistic theories. Here’s an example: 

• In Plato's Republic, Cleitophon proposes at one point (340b) that “justice” is whatever the strongest believes to be to his advantage. Suppose we drop the “whatever the strongest believes...” clause and apply this to all individuals? This would mean that a proposition like “X is just” has a positive truth value. But that truth value is surely trivial in nature: the individual could never be mistaken about what is just since it is relativised to their occurrent beliefs about what is to their advantage.

(f) Non-Scepticism: We want it to be possible to find out which moral propositions are true and which are false; we don’t want to be moral sceptics. This is an epistemic constraint and Craig seems to think such constraints are irrelevant in metaethical debates (at least in the ones he has had). I disagree (see here for more). It seems to me that a moral theory which satisfies the other conditions but which forces us to be moral sceptics would be unwelcome.

(g) Fit with Background Knowledge: We want the theory to be consistent with other parts of our worldview. Obviously this might be difficult to apply in some cases: theists are likely to think that morality must be consistent with God’s existence, and naturalists are likely to think that it would be nice if moral properties were somehow reducible to natural properties. Despite its difficulty, it still seems like a reasonable condition since it applies to all theories, not just moral ones.

These are the platitudes or conditions of success I’ll work with for the remainder of this entry. I could probably add more, and refine them a little bit, but I don’t want this to become too tedious. If you have any complaints about what I’ve said, or think certain refinements are necessary, feel free to comment below.

I doubt very much that any metaethical theory can satisfy all of these demands. The history of moral philosophy would seem to support this observation. Still, I think it is possible for some metaethical theories to satisfy a good number of these demands and so get us a good deal of what we want from morality. Indeed, I think the way to assess competing theories is on the basis of which ones get us the most of what we want.

This suggests that when deciding which is to be our preferred metaethical theory, we may be forced to compromise. We may have to give up one of the things we want in order to have more of the others.

All of which begs the question: which of these platitudes would we be willing to give up?

I’ll lay my own cards on the table at this point. I’d be willing to give up mind-independence. If I could formulate a metaethical theory which allowed for some degree of absolutivity, practicality, non-triviality, positive truth value, non-scepticism, and is consistent with my background knowledge, I’d be delighted. It would be like all my Christmases come at once. It would be bordering on petulance if I were to hold out for mind-independence after getting all of these things. Wouldn’t it?


3. The Scorecard Exercise
Now that we have some idea of what we want from morality, we can put it to use in the analysis of Craig’s theory. As was the case in part one, I am going to propose an exercise here. Its completion is left to the reader.

First, I suggest that we construct a scorecard with which we can evaluate different metaethical theories. On the "row" side of the scorecard we will have the moral platitudes outlined above. On the "column" side of the scorecard we will align all the different metaethical theories. Obviously, the goal is to see how well the different theories do with respect to the platitudes.

To perform this evaluation properly, we will have to have some idea of the available competitor theories. This is one area in which Craig’s discussion in Reasonable Faith is sorely lacking. He only seriously considers non-natural realism (or Platonic realism, as he calls it) as an alternative to theistic metaethics. I think we need to be far more expansive in our search of the available theoretic space.

Here is an abbreviated list of the available moral theories to get the ball rolling (I’m going to exclude non-cognitive theories):

• (1) Non-natural realism: Moral properties are real, but they exist in some abstract, non-natural realm.

• (2) Theological voluntarism: Moral properties are real, but they depend for their existence on one or more of God’s voluntary acts (willings, desirings, or commands).

• (3) Paradigmatic Exemplarism: Moral properties are real, but only in virtue of the fact that there exists a being that is a paradigmatic exemplar of those properties (I confess, you probably won’t find this theory mentioned in the available textbooks. As far as I know, the name is of my own invention. Nevertheless, it comes close to Craig’s own views with respect to how God grounds the existence of certain moral properties; and it is distinct from theological voluntarism).

• (4) Naturalistic Realism: Moral properties are real, but they are *reducible* to natural properties. I use *reducible* somewhat reluctantly since there are different theories and some of them object to the label of reductionism (Luke Muehlhauser has been doing a series on these recently, which is well worth reading).

• (5) Extreme Moral Subjectivism: Moral properties are real but their content is directly determined by the beliefs and desires of each individual.

• (6) External Moral Subjectivism: Moral properties are real but they depend for their existence on the subjective attitudes of one or more individuals external to oneself.

• (7) Constructivism: Moral properties are real, but they depend for their existence on some appropriately construed constructive procedure (e.g. a social contract). This comes in a variety of forms (Kantian, Humean and Decision-theoretic) and has been discussed on this blog before.

• (8) Nihilism: Moral properties are not real. We might think they are, but they’re really not.

Here then is the proposed scorecard (click to embiggen).

Now all we have to do is determine how each of the available theories does in relation to each of the proposed platitudes. We could place a tick or an “x” in each box for comparative purposes, and whichever theory satisfies the most of these demands “wins” (gets the most ticks), in the sense that it is the best we can hope for. I've kick-started the process by marking the score for nihilism.

I suggest that if we perform this exercise, Craig’s theory might not do nearly as well as some of its competitors. Furthermore, even if it does equally as well as some of its competitors, Craig’s moral argument is undercut. Why? Because he claims that God’s existence is essential for getting what we want from morality, but if there are other theories that get the same amount (or more) then this is wrong.

We must careful when doing the necessary comparison. This is for two reasons. First, it is probably best to construct a separate scorecard for each of the moral properties under consideration, i.e. one for moral values and another for moral obligations. Second, we should only reach a determination on a particular criterion after having evaluated the associated argument.

I’m not going to go ahead and actually fill in the remainder scorecard here since it requires a good deal more careful analysis to do this properly. Still, I think this is a good methodology to adopt when addressing Craig’s (and others’) arguments about metaethics.



4. Conclusion
The discussion in this post has taken us away somewhat from the original concerns with Craig’s definition of objectivity. I’ll try to tie it all back together.

As we have seen, Craig thinks the existence of God is the only way in which to secure the objectivity of morality. The problem is that it is not entirely clear what he means by objectivity. He might mean mind-independence, or non-relativity or positive truth value, or some combination of all three.

In light of his lack of clarity, I suggest we are better off interpreting Craig’s claims about objectivity as being akin to those of someone placing demands or constraints on successful moral theories. I further suggest that there are many such demands and that those employed by Craig may not even be that important.

In order to properly evaluate proposed metaethical theories, I think we should construct a kind of scorecard on which we determine how many of our moral demands are satisfied by the different theories and compare them with one another.

That’s it for now.