I suspect that many professional philosophers, including ones such as myself who have no religious beliefs at all, are slightly embarrassed, or even annoyed, by the voluble disputes between militant atheists and religious apologists. As Michael Frayn points out in his delightful book The Human Touch, the polite English are embarrassed when the subject of religion crops up at all. But we have more cause to be uncomfortable.

The annoyance comes partly because of the strong sense of deja vu. But it is not just that old tunes are being replayed, but that they are being replayed badly. The classic performance was given by David Hume in his Dialogues Concerning Natural Religion, written in the middle of the 18th century. Hume himself said that nothing could be more artful than the Dialogues, and it is the failure to appreciate that art that is annoying.

In the Dialogues, there are three principal characters. The first is Philo, a religious sceptic, whose voice is clearly that of Hume himself. Cleanthes is an apologist whose stock in trade is the argument that design is evidence of the existence of a deity: the familiar argument that the delicate and wonderful adjustments of nature irresistibly point to the existence of a divine architect - all nature declares the Creator's glory.

Finally, there is Demea, who wants the God of the philosophers: infinite, perfect, immutable, eternal or transcending space and time, incomprehensible and mysterious. Hume's art consists first in setting Cleanthes and Demea at each other's throats. Each represents an element in monotheistic religious belief, yet they cannot fit together.

³ÉÈËVRÊÓƵ

In some of the most humorous passages - and it is a very amusing work - Philo sides with Demea in trashing the conception of the deity available to Cleanthes, and indeed calls him little better than an atheist. But then he sides with Cleanthes, who trashes the conception of the deity available to Demea, and in turn calls him little better than an atheist, too. On each front, Philo wins, by two votes to one.

The two wings of theology, one making God immanent, something to be understood as analogous to ourselves, and one making Him transcendent, beyond spatio-temporal physical understanding, cannot be reconciled. The believer has to oscillate incoherently from one to the other.

³ÉÈËVRÊÓƵ

The problems with the divine architect, creating a cosmos in a manner analogous to the way humans design artefacts, are manifold and familiar. Our own creative activities are highly dependent on the delicate adjustments of the physical world. Our ideas are ideas of the things we come across in that world. Human designers are dependent on parents, not self-caused or self-explaining. Finally our aims and passions are adapted to the animal and social lives we lead. None of this is supposed true of the divine architect. But suppose we waive those difficulties, we still have it that human designers work in groups, refine the designs of others, sometimes lose interest in their designs, go on to make improved versions, and so on.

Cleanthes' theology leaves it open that the world, "for aught he knows, is very faulty and imperfect, compared to a superior standard; and was only the first rude essay of some infant deity, who afterwards abandoned it, ashamed of his lame performance: it is the work only of some dependent, inferior deity; and is the object of derision to his superiors: it is the production of old age and dotage in some superannuated deity; and ever since his death, has run on at adventures, from the first impulse and active force which it received from him."

Philo rightly concludes: "I cannot, for my part, think that so wild and unsettled a theology is, in any respect, preferable to none at all." Demea agrees: Cleanthes is little better than an atheist.

Later in the Dialogues, Cleanthes gets another bout in the ring, when the moral attributes of the deity come into play. But this only bruises him further, for it is obviously absurd to advance a perfectly benevolent, all- powerful and all-knowing architect as the best explanation of the spotty and often appalling course of human and animal lives. We cannot infer from the way of the world a deity that has any preference for good over evil, any more than for heat over cold or day over night.

So then we turn to Demea's transcendental conception of the deity. But this outruns any analogy to those things of which we have experience, and which therefore provide the origins of our ideas.

We cannot understand how anything could be necessarily existent, beyond time, immutable yet active. Since we have no idea of what the property of being necessarily existent in this way might be, then for all we can understand it might as well belong to the whole given cosmos as anything else.

Some might suggest that the world of abstract mathematical objects provides an example of the kind of existence needed, but most philosophers hold on to Gottlob Frege's insight that numerals are adjectives rather than nouns. They deny there is a "world" of mathematical objects in any relevant sense. And even if we were to talk that way, it would give us no usable concept of a deity.

The number 4 is not on the face of it a source of moral and political authority, or an actor in the world's affairs, or the target of prayers or the source of consolation, although it has as much claim to be the sustaining ground for the ongoing order of nature as anything else we can try to imagine. Cleanthes agrees: Demea can say nothing intelligible about his deity, and this makes him little better than an atheist.

³ÉÈËVRÊÓƵ

So is Hume himself an atheist? The word does not fit, and he never described himself as such. He is much too subtle. Philo the sceptic says that we cannot understand or know anything about a transcendent reality that explains or sustains the ongoing order of nature, while theists such as Demea say that we cannot understand or know anything about the transcendent reality, which is God, that explains or sustains the ongoing order of nature. Since the inserted clause does not help us in the least, the difference between them is merely verbal. And this is Hume's conclusion.

³ÉÈËVRÊÓƵ

At the end of the Dialogues, the little boy, Pamphilius, who is present as an auditor, says that Cleanthes' arguments appealed to him most, and even Philo, surprisingly, makes some apparently complimentary remarks about the design argument, provided it has a completely undefined conclusion. Some commentators have rather flat-footedly thought that this was some kind of recantation on Hume's part. But of course it wasn't. It was a supreme piece of his habitual irony.

Since by the end neither Cleanthes nor Demea can defend any usable conception of a deity, it matters not in the least whether you are drawn to say that "it" exists or to deny it. There is no inference to be drawn about anything - moral, political, empirical or theoretical - from either the assertion that "it" does or "it" does not. Joining in on either side equally implies that we know what we are talking about, and the right philosophical attitude is just to laugh at persons who suppose that.

Hume elegantly sidesteps the common charge that dogmatic atheism is just as much a "matter of faith" as faith itself. You cannot make that claim against someone whose mocking irony is careful to issue no "ism" at all.

He also escapes the debating point that atheism is "parasitic" on religious belief. A contented absence of belief is no more parasitic on what is absent than the absence of crocodiles in England is parasitic on them being there, although it is also true that you could not laugh at faiths without them being there to laugh at.

But it is also wrong to call Hume an agnostic. That would imply a definite question about which we do not know the answer. But since there is no definite question at stake, that too lapses.

Hume knew that he was unlikely to be understood. He also knew that the interesting questions now shift to the study that he pioneered in The Natural History of Religion: the comparative study of religious practices and the psychological and social mechanisms that give rise to them, and that they articulate.

The interesting questions surround the anthropology of activities such as drama, dance, music, rituals and ceremonies. Here the question of belief subsides, and the focus turns to what Ludwig Wittgenstein called the "stream of life" that issues in these doings. There is no doubt that these doings and sayings have a function, for good or ill.

They may express hope or fear, safety in the universe or unease at its harshness; tribal solidarity and hostility to others; or universal benevolence and brotherly love. Since religious practices are those of ordinary people, they inherit both the best and the worst sides of human nature.

According to Hume, all human beings have "some particle of the dove kneaded into our frame, along with the elements of the wolf and the serpent", and even Christians are human. Some of their music, architecture and poetry is rather good, some parts less so.

Bad things happen when people decorate their bare, inchoate, unstable and inconsistent imaginings with the baser trappings of their culture. They come out of the fog bearing ludicrous beliefs about cosmology or biology, or carrying their envies and fears, their embarrassments about sex in general or certain varieties in particular, their desire to steal some land or make war on their neighbours. Deities then become dangerous, megaphones through which emotions are whipped up and particular moral demands are given a spurious authority. People need prophets and priests to carry the megaphones, and they are often supposed to signal their rapport with the deity by making remarkable things happen.

Hume also completely destroyed the reasons for believing in any such revelation and signal of revelation in the other prong of his scepticism: the devastating argument against belief in the testimony of miracles. This strips away any pretension to special authority, and then we can go on to test the moral injunctions in their own terms, standing on our own feet. The scandal is when the forum for debate, such as the House of Lords, is stacked with one set of devotees, with the kind of result witnessed in its defeat of Lord Joffe's Assisted Dying for the Terminally Ill Bill.

³ÉÈËVRÊÓƵ

The upshot ought to be not dogmatic atheism, but sceptical irony. Of course, the latter is just as infuriating to those making special claims to authority, perhaps more so. Men and women of God may find it invigorating and bracing to meet disagreement, but even benevolent mockery is mockery. They would find that it is much harder to bear the Olympian gaze of the greatest of British philosophers.