The heart of my argument is the following assumption. The conventional, theistic religions with which most of us are most familiar—Judaism, Christianity, and Islam—have two parts: a science part and a value part. The science part offers answers to important factual questions about the birth and history of the universe, the origin of human life, and whether or not people survive their own death. That part declares that an all-powerful and all-knowing god created the universe, judges human lives, guarantees an afterlife, and responds to prayer.
Of course I do not mean that these religions offer what we count as scientific arguments for the existence and career of their god. I mean only that this part of many religions makes claims about matters of fact and about historical and contemporary causes and effects. Some believers do defend these claims with what they take to be scientific arguments; others profess to believe them as a matter of faith or through the evidence of sacred texts. I call them all scientific in virtue of their content, not their defense.
The value part of a conventional theistic religion offers a variety of convictions about how people should live and what they should value. Some of these are godly commitments, that is, commitments that are parasitic on and make no sense without the assumption of a god. Godly convictions declare duties of worship, prayer, and obedience to the god the religion endorses. But other religious values are not, in that way, godly: they are at least formally independent of any god. The two paradigm religious values I identified are in that way independent. Religious atheists do not believe in a god and so reject the science of conventional religions and the godly commitments, like a duty of ritual worship, that are parasitic on that part. But they accept that it matters objectively how a human life goes and that everyone has an innate, inalienable ethical responsibility to try to live as well as possible in his circumstances. They accept that nature is not just a matter of particles thrown together in a very long history but something of intrinsic wonder and beauty.
The science part of conventional religion cannot ground the value part because—to put it briefly at first—these are conceptually independent. Human life cannot have any kind of meaning or value just because a loving god exists. The universe cannot be intrinsically beautiful just because it was created to be beautiful. Any judgment about meaning in human life or wonder in nature relies ultimately not only on descriptive truth, no matter how exalted or mysterious, but finally on more fundamental value judgments. There is no direct bridge from any story about the creation of the firmament, or the heavens and earth, or the animals of the sea and the land, or the delights of Heaven, or the fires of Hell, or the parting of any sea or the raising of any dead, to the enduring value of friendship and family or the importance of charity or the sublimity of a sunset or the appropriateness of awe in the face of the universe or even a duty of reverence for a creator god.
I am not arguing, against the science of the traditional Abrahamic religions, that there is no personal god who made the heavens and loves its creatures. I claim only that such a god’s existence cannot in itself make a difference to the truth of any religious values. If a god exists, perhaps he can send people to Heaven or Hell. But he cannot of his own will create right answers to moral questions or instill the universe with a glory it would not otherwise have. A god’s existence or character can only figure in the defense of such values as a fact that makes some different, independent background value judgment pertinent; it can only figure, that is, as a minor premise. Of course, a belief in a god can shape a person’s life dramatically. Whether and how it does this depends on the character of the supposed god and the depth of commitment to that god. An obvious and crude case: someone who believes he will go to Hell if he displeases a god will very likely lead a different life from someone who does not have any such belief. But whether what displeases a god is morally wrong is not up to that god.
I am now relying on an important conceptual principle that we might call “Hume’s principle” because it was defended by that eighteenth-century Scottish philosopher. This principle insists that one cannot support a value judgment—an ethical or moral or aesthetic claim—just by establishing some scientific fact about how the world is or was or will be. Something else is always necessary: a background value judgment that shows why the scientific fact is relevant and has that consequence. Yes, whenever I see that someone is in pain, or threatened with danger, I have a moral responsibility to help if I can. Just the plain fact of pain or danger appears to generate, all by itself, a moral duty. But the appearance is deceptive: the pain and danger would not generate a moral duty unless it was also true, as a matter of background moral truth, that people have a general duty to relieve or prevent suffering. Very often, as in this case, the background principle is too obvious to need stating or even thinking. But it must still be there, and it must still really connect the ordinary judgment with the more concrete moral or ethical or aesthetic judgment it is supposed to support.
I agree that the existence of a personal god—a supernatural, all-powerful, omniscient, and loving being—is a very exotic kind of scientific fact. But it is still a scientific fact and it still requires a pertinent background moral principle to have any impact on value judgments. That is important because those background value judgments can only themselves be defended—to the extent they can be defended at all—by locating them in a larger network of values each of which draws on and justifies the others. They can only be defended, as my account of the religious attitude insists, within the overall scheme of value.
So a god’s existence can be shown to be either necessary or sufficient to justify a particular conviction of value only if some independent background principle explains why. We might well be convinced of some such principle. We might think, for instance, that the sacrifice of God’s son on the Cross gives us a responsibility of gratitude to honor the principles for which He died. Or that we owe the deference to the god who created us that we owe a parent, except that our deference to that god must be unlimited and unstinting. Believers will have no trouble constructing other such principles. But the principles they cite, whatever they are, must have independent force seen only as claims of morality or some other department of value. Theists must have an independent faith in some such principle; it is that principle, rather than just the divine events or other facts they claim pertinent, that they must find they cannot but believe. What divides godly and godless religion—the science of godly religion—is not as important as the faith in value that unites them.