Where the world appears, as it does to the vulgar, capricious and uncertain it is natural for the imagination to construct a number of deities, whose differing attributes and personalities can be invoked as appropriate. The ideas of the various gods arise from our ignorance by a process in which emotion, mediated by the imagination, leads to beliefs in deities that are constructions of the mind:

We are placed in this world, as in a great theatre, where the true springs and causes of every event are entirely concealed from us; nor have we either sufficient wisdom to foresee, or power to prevent those ills, with which we are continually threatened. We hang in perpetual suspense between life and death, health and sickness, plenty and want; which are distributed amongst the human species by secret and unknown causes, whose operation is oft unexpected, and always unaccountable. These unknown causes, then, become the constant object of our hope and fear; and while the passions are kept in perpetual alarm by an anxious expectation of the events, the imagination is equally employed in forming ideas of those powers, on which we have so entire a dependance 5

Thus polytheism rather than monotheism is the first form of natural religion.

It should be noted that in the Natural History, and elsewhere in his writings on religion, Hume is prepared to adopt a distinction between ‘true’ and ‘false’ religion. The origin of polytheistic beliefs in fears and hopes exemplifies what he calls ‘superstition’, and he is willing to regard this together with another form of religious belief originating in emotion, ‘enthusiasm’, as ‘the corruptions of true religion’, as ‘two species of false religion’. There is thus a similarity here with the deists, who also regarded many forms of devotion and ritual as superstitious. Hume’s list is:

… ceremonies, observances, mortifications, sacrifices, presents, or… any practice, however absurd or frivolous, which either folly or knavery recommends to a blind and terrified credulity.6

But this similarity should not obscure from us his opposition to deism. The Natural History undermines the deist claim that ‘true religion’, in the form of something like Herbert’s ‘common notions’, can be recommended on the grounds that it has always been accepted in all places. There is no argument for true religion based on the ‘universal consent of mankind’.

A distinction between the true religion of the wise and the superstitions of the vulgar is found already in Cicero’s dialogue on which Hume’s work is modelled. There Cotta, the spokesman for the Academic school, says:

With the ignorant you get superstitions like the Syrians’ worship of a fish, and the Egyptians’ deification of almost every species of animal; nay, even in Greece they worship a number of deified human beings… and with our own people Romulus and many others, who are believed to have been admitted to celestial citizenship in recent times, by a sort of extension of the franchise! Well, those are the superstitions of the unlearned; but what of you philosophers? How are your dogmas any better?

Hume, too, considers that to demonstrate that superstition and enthusiasm originate in ignorance, fear, hope, elation, flights of fancy and so on, is not to discredit true religion. The wise, the philosophers, may yet have grounds for a rational belief in a supreme creator and in the world as ordered and governed by divine providence. Are these dogmas any better? Is it possible to construct a rational natural theology, even if men and women are not generally led to religion by reasoning?

It is this question which is debated between the three characters in the Dialogues. Following Cicero’s model, each character represents a different position on the central questions of natural theology. Can reason establish the existence of God? What can be known by reason of the nature and attributes of God? Can there be a reasoned solution to the conflict between divine goodness and divine power, given the existence of moral evil and natural suffering? What answer a philosopher or theologian gives to these questions is determined in part by his or her epistemology. It is necessary now to give an outline of a central aspect of Hume’s own epistemology, as it is found in his earlier writings.

In the Treatise he argues that the fact that one phenomenon, or ‘object’ as he usually says, is the cause or effect of another is never something which is discoverable simply by reflecting on what is contained in our idea of it.7 In order for us to know that fire burns us or bread nourishes us, we must experience the effect of fire or bread on our bodies. He expresses this by saying that we cannot have a priori knowledge of the causes and effects of objects. This contrasts, he thinks, with what holds in, say, mathematics. Here we can know that one object is related to another simply by abstract reflection. We can see that 16 is the square of 4 just by examining our ideas of these numbers. The necessary relations between numbers are thus relations of ideas; they are purely conceptual, and can be known a priori. One mark of truths which are relations of ideas is that the negations of such truths (e.g. that 16 is not the square of 4) are inconsistent and lead to a contradiction. In contrast, however certain we may be that fire burns us, we can imagine without contradiction a possible state of affairs (different from what actually obtains) in which we are not burned by fire. That fire burns us is thus simply a matter of fact, not an a priori necessary truth.

In the Treatise Hume therefore distinguishes between two domains, relations of ideas and matters of fact. Using terms in a technical, philosophical sense, he calls the domain of relations of ideas knowledge, and the domain of matters of fact probability.8 In this technical sense, it is a matter of probability that fire burns us. This does not mean, for Hume, that we cannot be certain that fire burns. Certainty is a matter of the degree of conviction that we feel about a thing, and we can be certain of something whose negation is still quite conceivable.

That we have beliefs about the causal properties of objects is itself a matter of fact. The question why fire burns us is a question for natural sciences – physics, chemistry, physiology – to answer. The question why we believe that fire burns us, however, is of a kind which, Hume thinks, has not previously been successfully answered. Evidently, that something is the case is not itself an explanation of why we believe that it is the case.