Meta-Ethics and Moral Arguments for God

I haven't talked much on this blog about moral arguments for the existence of God, at least not explicitly. In a way this is a bit odd, because my first book can be viewed as offering one kind of moral argument--not for the existence of God, but for the legitimacy of having faith in God's existence.

Moral arguments for God's existence, or for religious faith, fall right at the intersection of my main philosophical interests. In this post, I want to consider one such intersection: The need for moral arguments for God to engage with the diverse range of ideas that fall under the heading of what is called "meta-ethics."

Off and on I've posted things on this blog pertaining to what is called "meta-ethics," although I've tended to eschew that term. The field of meta-ethics is perhaps best understood in terms of the questions it asks, which in turn are best understood as question about the answers to lower-level ethical questions. Such lower-level questions are usually categorized as either "applied" or "normative." Applied ethics asks questions like the following: "Is it always wrong for a person to terminate an unwanted pregnancy?" Normative ethics, meanwhile, asks more general-level questions about the nature of morality, such as,"Are there some things that are wrong regardless of the consequences, or is the moral status of an action always a function of its outcomes?"

Meta-ethics steps back, looks at the answers to such lower-level questions, and asks about the nature of these answers. When you say that deliberately terminating a pregnancy is always wrong, are you asserting something of pregnancy-terminations that you take to be true of them, namely that they always have the property of wrongness? If so, what kind of property, exactly, is "wrongness"? And if not, what are you doing when you say these words?

When I say that the end does not always justify the means, am I asserting a fact which is either true or false apart from how I feel? Or am I simply expressing, say, a general-level feeling about actions that prioritize ends over means? Or am I, perhaps, simply voicing my personal commitment never to let consideration of ends trump consideration of means? If I'm asserting a fact, is it morel like a mathematical fact ("2+2=4"), an empirical one ("There are two beers in the fridge"), or a socially constructed one ("Barack Obama is President of the United States")? Or are moral facts unlike any of these?

When I've wandered into meta-ethics on this blog, it has generally been in relation to the question of whether moral claims are "objective" or "subjective." This is a classic way of attempting to get at one of the most crucial meta-ethical questions: Does a moral judgment have a truth value that is independent of the preferences/attitudes/beliefs of the individual making the judgment?

Although I haven't explicitly pursued the connection here, these meta-ethical questions have bearing on issues in the philosophy of religion. More precisely, there's one especially widespread kind of moral argument for God's existence whose soundness depends on some very precise answers to a number of questions in meta-ethics. The species of argument I have in mind has something like the following form:

1. Moral claims have an objective truth value, and some moral claims are true (they're not all false).
2. In order for moral claims to have this sort of objective truth value, theism must be true (or at least metaphysical naturalism must be false).
3. Therefore, theism is true (or metaphysical naturalism is false).

(Clarifying note: What is premise 1 saying? Here's a rough elaboration: A moral claim attributes a property of a certain kind--what we might call a normative property, such as "right" and "good"-- to such things as persons, actions, character traits, and states of affairs. Premise 1 is saying, first, that some of these attributions are correct and others incorrect; and second, that what makes the correct attributions correct is something apart from anyone's (individual or group's) actual approval or disapproval, acceptance or rejection, of that which is being called good or right, etc.)

Typically, the case for the first premise of such an argument rests on an appeal to basic moral intuitions. We are invited to consider a moral claim such as "Torturing children for fun is immoral." We are invited to think about what it would mean for such a claim to be treated as lacking in objective truth. If the implications of such thought experiments are deeply counter-intuitive, it follows that we intuitively accept the first premise. Once this is established, an effort is made to defend the second premise--that is, to show that in order to remain true to our intuitions--in order to underwrite the meta-ethical position we intuitively embrace--we must suppose there is a God. How this is done will vary greatly according to the version of the argument being considered.

A weaker form of this strategy of argument would hold that theism does a better job of making sense of our moral intuitions than does any form of naturalism; hence, theism is the best way (at least given our current understanding of matters) to underwrite those intuitions.

One crucial weakness of any such strategy of argument is this: Even if it works, the reasoning can work in both directions. As one of my friends from graduate school once quipped, "One person's modus ponens is another person's modus tollens." (I won't explain the terms. Google them if you care.) In a nutshell, if you agree that given your intutions premise 1 should be accepted, and you accept 2, you actually have two choices: accept theism or reject your intuitions. And there are some who find theism so implausible (or naturalism so obvious) that they are more than ready to take the second option.

But this is hardly the only way to resist the conclusion of a moral argument along these lines. In fact, there are meta-ethicists who question whether our intutitions really speak as strongly in favor of premise 1 as defenders of this sort of argument suppose. Furthermore, there are meta-ethicists who deny premise 2, having offered quite sophisticated defenses of naturalistic accounts of objective moral truth

Put simply, both premises 1 and 2 make meta-ethical claims that can be and have been challenged in the philosophical literature.

Let's start with the first premise. I have formulated this premise so that it would be true if either of two important meta-ethical positions is correct: (i) moral realism and (ii) objectivist versions of constructivism. Moral realism is, roughly, the view that there are moral facts "out there" whose truth is independent of what any individual or group, even an "ideal" one, does or would think/feel/endorse. Objectivist versions of constructivism hold, roughly, that moral truth is determined by the judgments that would be made by a person or community under certain ideal conditions--for example, under the condition that the individual were being perfectly rationally consistent in the presence of complete knowledge of all relavant facts. In other words, what makes some moral claim true is that the moral claim would be endorsed by the ideally-situated person or group (not by what anyone actually endorses--actual endorsements are correct or incorrect based on their correspondence to this ideal). This view is to be distinguished from subjective and relativistic versions of constructivism. Subjectivism holds that a moral claim is true if it correctly reports the actual contingent thoughts/feelings/attitudes of the individual making the report. Relativism holds that a moral claim is true if it correctly describes that actual agreements reached by a given group, such as a culture or society.

The reason I've formulated premise 1 so that it encompasses both (i) and (ii) is simply this: In my judgment the intuitive case for objectivity in ethics (assuming it is convincing) would be satisfied by either of these theoretic approaches. This is not to say that I think both approaches can be philosophically developed or worked out with equal success. It is one thing to say that our intuitions would be satisfied if there were a moral truth "out there." It is something else to give a plausible account of what such moral truth would be like. It is one thing to say that our intuitions would be satisfied if there were some idealized perspective from which these intuitions would be reliably endorsed. It is something else to attempt to describe such a perspective and show that it would actually underwrite our moral intuitions.

In any event, what the first premise rules out are those meta-ethical views that deny that there are objective moral truths. These are: (a) noncognitivist theories (which hold that moral utterances don't have any truth value at all, since they don't assert anything but, instead, merely express attitudes or plans or some such); (b) subjective or relativistic versions of constructivism (sketched out above); and (c) "error theory" (which holds, in effect, that moral claims do have an objective truth value, but that all moral claims are objectively false in the way that all claims about the properties of non-existent things, like unicorns, are objectively false).

Premise 1 of the argument above effectively rejects each of these theories. But there are, of course, sophisticated meta-ethicists who have rigorously defended these alternative meta-ethical views, even in the face of the intuitive challenge. Simon Blackburn, for example, has been an importand defender of (a). Another, with whom I had dinner a couple of weeks ago, is Allan Gibbard. Gilbert Harman has defended (b). And John Mackie has defended (c). These philosophers have endeavored to account for the intuitions that seem to support an objectivist meta-ethical stance either by explaining the intuitions away or by showing that there remains a way to preserve the intuitions (perhaps in a modified form) within these alternative meta-ethical frameworks.

As to the second premise, there are some sophisticated moral realists who, in recent years, have attempted to stake out a version of moral realism that is thoroughly naturalistic (David Brink and Richard Boyd are good examples). Furthermore, there are a range of sophisticated theories--most of them inspired by Kant--that purport to provide an objective foundation for moral truth that is "constructivist" in the technical sense and which makes no appeal to God in the course of offering that foundation (John Rawls, Alan Gewirth, and Christine Korsgaard offer examplars of such approaches). In general, constructivist accounts of morality, whether objective or subjective or relative, are consistent with naturalism (even thought at least some constructivist theories may also be consistent with theism or other forms of supernaturalism). As such, there are a wide range of theories that would need to be discredited in order for the second premise of the above argument to remain anything more than controversial.

In short, moral arguments for the existence of God that have anything like the above structure need to tackle the entire range of meta-ethical literature. It's not enough that they be able to show how moral objectivity could be defended on theistic assumptions. They also need to show (a) that naturalistic forms of moral realism don't work as well as the favored theistic theory; (b) that objective forms of constructivism don't work as well; (c) that contrary to claims made by some noncognitivists, subjective constructivists, etc., these theories really do fly in the face of our moral intuitions; and (d) if one must choose between one's moral intuitions and a metaphysical naturalism, there is good reason to jettison the latter.

This is quite a project, to say the least. It's unlikely that even with a hefty book one could adequately pursue each elements of it (Robert Merrihew Adams has gamely attempted something along these lines with his Finite and Infinite Goods; but while I regard that books as a great intellectual achievement, it is hardly the final word on the subject--as is evidenced by the rich exchange, following the book's publication, between Adams and the naturalistic moral realist Richard Boyd).

So, even if I were inclined to defend this sort of moral argument for God's existence, I wouldn't be inclined to do so in a blog post. And the fact is, I'm not sure what to think about this moral argument. I'd love to find a version of it that convinces me when I put on my critical philosopher's hat, but I haven't yet.

There are, however, other kinds of moral arguments. Some I find myself more drawn to than others.

In fact, one way to read Is God a Delusion? would be to treat it as an extended defense of a moral argument, not for God's existence as such, but for the decision to have faith in the existence of God. Here's how I'd formally reconstruct the argument along these lines that is nascent in the book:

1. Some things are objectively morally good.
2. If naturalism were true, then reality at a fundamental level would be indifferent to what is objectively morally good (the basic constituents of and principles governing reality would, given naturalism, neither embody such goodness themselves nor reliably promote its expression or preserve/perpetuate that which expresses it).
3. There is a worldview, opposed to naturalism, according to which reality at a fundamental level would not be indifferent to what is objectively morally good but, on the contrary, would embody such goodness, promote its expression, and preserve/perpetuate that which expresses it. A worldview of this sort is what unites theists who regard God as a proper object of devotion, trust, and worship (as opposed to fear and fawning subservience), and so can be described as the worldview of theistic religion (as opposed to theistic superstition, following Plutarch's distinction).
4. Both naturalism and the worldview of theistic religion are compatible with reason and our overall body of experience, even though both exceed what reason and evidence can establish.
5. Given the definition of theistic religion offered in (3), it would be objectively morally good if the worldview of theistic religion were true.
6. Living as if the worldview of theistic religion were true is a morally benign choice--that is, doing this would not produce outcomes or behaviors that are objectively morally bad, but would produce outcomes or behaviors that are objectively morally good.
7. Faith in one important sense of the word involves the decision to live as if a worldview is true based on the hope that it is true, and this decision is reasonable and moral if (a) the worldview is compatible with reason and experience even if not uniquely supported by it; (b) it would be objectively morally good were the worldview true (hence making the worldview a suitable object of hope); and (c) living as if it were true is morally benign.
8. Hence, faith in the worldview of theistic religion is reasonable and moral.

Framed in this way, it's obvious that my overarching line of argument in Is God a Delusion? depends on a meta-ethical premise, one which I never explicitly defend. But notice that this premise--roughly, that there are objective moral goods--would be true were any form of moral realism or objective constructivism correct. In other words, this argument does not depend on discrediting naturalistic forms of moral realism and every version of objective constructivism. It does, however, suppose that noncognitivism, error theory, and subjective and relativist forms of constructivism are mistaken.

In short, a unifying argument in Is God a Delusion? presupposes a meta-ethical position, but a far broader one than the more narrow moral argument requires.

There are other moral arguments for God's existence, and I suspect that each such argument probably depends for its soundness on some kind of meta-ethical position being correct. If so, then criticisms and defenses of moral arguments for God, to be complete, will inevitably have to consider work being done in meta-ethics.