In Paul Russell's masterful analysis of the irreligious nature of Hume's Treatise of Human Nature entitled, The Riddle of Hume's Treatise, Russell recounts an argument for theism by Samuel Clarke and Hume's refutation. Russell shows how, in his efforts to defend a natural science of human knowledge, Hume is also attacking common theistic proofs. Clarke's argument is as follows:
Something existed from all eternity. If something had not existed from all eternity then something would have come from nothing. That which has existed from all eternity is unchangeable and independent. If there had not existed from all eternity something unchangeable and independent then everything that has ever existed would be changeable and dependent. If everything is changeable and dependent then it is equally possible that everything exist and everything not exist. Since everything does exist there seems to be no determining factor that makes it so apart from nothing, and nothing can come from nothing.
Having established this first principle, Clarke moves to argue that the immutable, unchangeable something must be a necessarily existing intelligent being. Since anything immutable and unchangeable cannot be caused by anything it must be "self-existent." It is clear that material things are not immutable or unchangeable since we have seen that all material things might not have existed (since all material things are changeable and dependent). Since material things do exist they require a necessarily existing thing as their determining factor for existence.
Such a necessarily existent thing is an intelligent being. This, says Clarke, can only be argued a posteriori. Everything in the world, according to Clarke, demonstrates the effects of an intelligent cause. Clarke writes:
In general there are manifestly in things various kinds of powers, and very different excellencies and degrees of perfection; it must needs be, that, in the order of causes and effects, the cause must always be more excellent than the effect; And consequently, the self-existent being, whatever that be supposed to be, must of necessity (being the original of all things) contain in itself the sum and highest degree of all the perfections of all things... 'tis impossible that any effect should have any perfection, which was not int he cause. For if it had, then that perfection would be caused by nothing; which is a plain contradiction.Since, among the effects, we find intelligence we must conclude that the original cause of all things is intelligent.
Hume's reply is to suggest that the first principle--that whatever begins to exist must have a cause--cannot be demonstrated. He then produces his famous maxim that "any thing may produce any thing" and that this can be arrived at a priori.
Hume suggests that it is not a contradiction to suppose that something could begin to exist without a cause. If it is possible to conceive of something not existing one minute and existing the next then there can be no contradiction in denying Clarke's causal maxim. Furthermore, argues Hume, there is nothing observable about any object that can be called an effect or a cause. Since nothing about an object is identifiable as a cause or an effect we must derive the idea from the relation between objects. All we can discover is that one thing happens after another and it is only from experience that we derive any idea of cause and effect. Yet cause and effect cannot be demonstrated conclusively from experience either. All that is possible is that we observe many things happening in succession, but we cannot conclude from this that there is a cause and an effect. If cause and effect cannot be demonstrated Clarke cannot reach the conclusion that a necessarily existent being exists.
Hume then argues that, given the problem of demonstrating cause and effect, it is possible that "any thing may produce any thing." There is no way to demonstrate particular effects from particular causes. On this basis we cannot conclude that material in motion cannot produce intelligence. As Russell summarizes:
In rejecting the principle of causal adequacy, Hume removes the entire causal foundations of all attempts to demonstrate the being and attributes of God on the basis of the assumption that "nothing can be efficiently caused or produced by that which hath not in it at least equal (if not greater) perfection, as also sufficient power to produce."Hume's rejection of the causal maxim found in many enlightenment philosophers led to skepticism over natural religion. The price of his denial is Hume's skepticism In denying that it is possible to demonstrate the existence of God he ends up denying that it is possible to demonstrate the existence of anything.
Throughout Hume's career, he continued to undermine theistic proofs, most notably in his Dialogues Concerning Natural Religion.
Whereas, at the time, Clarke's argument was popular and widely used, it is now seen in the light of Hume's (and Kant's) skepticism. Yet, if such theistic arguments are so undermined, even at a high price, then what, if any, are the benefits of Clarke's argument?
One possible answer is to declare them hopeless, useless, and abandon them entirely. Karl Barth, in his denial of natural theology, suggested that all theistic proofs that rely on experience or reason without reference to revelation cannot make any headway in demonstrating the existence of God. However, Barth, in order to deny their usefulness is forced to assume somewhat of a Kantian framework.
Another, more recent, suggestion comes from C. Stephen Evans. Evans, in his book, Natural Signs and Knowledge of God, suggests that arguments such as Clarke's serve to point to the likely existence of God. Signs, says Evans, are universally available and easily resistible. Hume's resistance, therefore, is also indicative of the sign itself.
However, I'm inclined to think that Clarke's and other similar arguments could be used in order to produce skepticism. It seems to me that what one wants to "prove" is that in the denial of God one must begin to deny everything. Hume's denial not only leads him to a skepticism about the existence of any object, but a skepticism about human knowledge per se. In other words, in denying God, the price, one should suggest, is not merely the loss of one part of demonstrable human knowledge, but the whole jolly lot!