Consider the following passage from J. J. C. Smart:
It looks as though the theistic hypothesis is an empirically refutable one, so that theism becomes a refuted scientific theory. The argument goes: (1) If God exists then there is no evil, (2) There is evil, therefore (3) It is not the case that God exists. Premiss (1) seems to follow from our characterization of God as an omnipotent, omniscient and benevolent being. (2) is empirical. We can hardly reject (2). It seems therefore that the theist has to find something wrong with (1) and this is not easy. (J. J. C. Smart and J. J. Haldane, Atheism and Theism, Blackwell 2003, 2nd ed, p. 60)
Smart's argument from evil is plainly valid, being of the form modus tollens. But for an argument to be probative, other conditions must be met. One of these conditions is that the premises be true. Another is that the argument involve no 'informal fallacy' such as equivocation.
So let us ask: how would 'evil' in (1) have to be construed so that (1) comes out true? I say that 'evil' must be short for 'gratuitous evil.' But then, to avoid equivocation, we would have to replace 'evil' in (2) with 'gratuitous evil.' The result would be:
1*. If God exists, then there is no gratuitous evil.
2*. There is gratuitous evil.
—
3. It is not the case that God exists.
The resulting argument is valid, and (1*) is plainly true, unlike (1) which is not plainly true, but false. That (1) is false can be seen from the fact that an omni-qualified God could easily permit the existence of an evil that was necessary for the attainment of a greater good. Such an evil would not be gratuitous. So it is just false to say, "If God exists, then there is no evil."
But (1*) is plainly true. Now it may be — it is epistemically possible that — (2*) is also true. The reformulated argument would then be sound. A sound argument, by definition, is a deductive argument that is both valid in point of logical form and whose premises are all of them true. As for epistemic possibility, a proposition p is epistemically (doxastically) possible for a subject S if and only if p is logically consistent with what S knows (believes).
But note that a sound argument will be probatively worthless if it begs the question, if it is such that one cannot know a premise to be true without already knowing the conclusion to be true. So let us ask a very simple question: How does one know that (2*) is true? How does one know that there is gratuitous evil? Smart tells us that (2) is empirical. 'Empirical' is a term of epistemology. It is applied to those propositions that are known from experience, by observation via the senses and their instrumental extensions (microscopes, telescopes, etc.) Now I am willing to grant arguendo that (2) — There is evil — is an empirical truth. (2), however, is not what Smart needs to make his argument work. He needs (2*). But is (2*) an empirical truth? Can one know from sensory experience (whether inner or outer) that there is gratuitous evil? Is gratuitousness an empirical attribute of the evils one experiences?
Consider the evil of intense pain. I am acquainted with pain by 'inner sense.' And I am willing to grant arguendo, though it is not quite obvious, that I am acquainted empirically with the evil of intense pain. But I am surely not acquainted empirically with the gratuitousness of experienced evils. Gratuitousness is no more an empirical attribute than the createdness of the natural world. It is not evident to the senses that nature is a divine creation. Similarly, it is not evident to the senses that instances of evil are gratuitous. Is it not epistemically possible that they are all non-gratuitous?
To say that an evil is gratuitous is equivalent to saying that it is an evil inconsistent with the existence of the omni-qualified God. It is to say that it is an evil that no such God could have a morally sufficient reason for permitting. Clearly, one cannot 'read off' such a complex relational attribute from any instance of evil. Even if it is empirically obvious that there are evils such as the intense pain both physical and psychological of a fawn being incinerated in a forest fire, it is not empirically obvious that this evil is gratuitous.
The conclusion I am driving towards is that Smart's argument supra is question-begging. For in order to know that premise (2*) is true, I must know that the conclusion is true. That is, to know that there are gratuitous evils, I must know that God does not exist. For if God exists, then then there are no gratuitous evils. At most, there would only seem to be such.
Smart tells us above that the theistic hypothesis is empirically refutable. But I say Smart is mistaken: he needs (2*) for his argument to work, but this proposition — There is gratuitous evil — is not empirical. It may be true for all that, but it is not knowable by experience. You may be convinced that it is true, and I won't blame you if you find it much more plausible than the truth of 'God exists'; but it is not an empirical truth, if it is a truth. It is an interpretation imposed upon the data. It is as metaphysical as 'God exists.'
One last observation. There is something strange about referring as Smart does to theism as an empirical hypothesis. This is not the place to pursue this. I will merely suggest that the existence of God is more like the existence of truth than it is like the existence of something empirically confirmable or disconfirmable. It would be easy to show the absurdity of the question, Is the existence of truth empirically refutable? So too, perhaps, with the existence of God. The existence of truth is an ultimate presupposition of all inquiry and all assertion. The existence of God may be in the same logical boat.
If you are thinking of God along the lines of Russell's celestial teapot, you have missed the boat.
Leave a Reply to Vito B. Caiati Cancel reply