We discussed two problems for Hume’s discussion of miracles.
Lewis accused Hume of both winning by defining miracles in a peculiar way (as Vivian noted) and of arguing in a circle (as Abi noted). The charge is that his argument amounts to the assertion that there is “uniform experience” against miracles. But we can only know that this is so if we have already dismissed all reported experiences of miracles.
I think that Hume would have been guilty of arguing in this way if he had done two things. First, he would have had to have defined miracles as contradicting the uniform experience of mankind. Second, he would have had to have meant something very specific by the word “uniform”, namely, that the relevant experiences are utterly without exception.
However, I don’t think Hume is guilty of arguing in that way. He asked us to compare all of the observations of the natural world on one hand with the reported observation of a miracle on the other. Our experiences of the natural world are the “uniform experience” because, well, they’re uniform.
Against the uniform experiences, we put the reported experience of the miracle.
We weigh the two and decide whether to believe that the reported experience of a miracle was genuine or, for whatever reason, false.
Hume’s claim is that this weighing process will always work against the reported miracle.
That is not the same thing as arguing that since our experiences uniformly, meaning completely and without exception, include no miracles, there have not been any miracles.
Lewis’s second argument is the second point I noted above, in the first section of this page.
I think it’s a genuine problem for Hume. I also think that we hit on the best way that Hume might try to answer it (sorry, I only remember Kimbia’s contributions here, though I know that a number of you chipped in). This attempts to distinguish between the skeptical attack on our reasons for drawing inferences that presume the uniformity of nature and our normal standards for assessing probabilities. The argument against miracles is conducted in terms of our normal standards for assessing probabilities. The skepticism about causal inferences and the uniformity of nature, on the other hand, maintains that we don’t have any reasons for adopting those normal standards.
It helps that, as Tommy pointed out, we have to use causal reasoning. But that is coming dangerously close to the kind of wishful thinking that, we all know, isn’t a very good guide to the truth.
We spent relatively little time on Lewis’s point that there can be a supernatural explanation for the uniformity of nature. We can’t explain that uniformity using the purely natural processes that Hume describes. But we could explain it if we supposed that a supernatural being preferred a uniform natural world.
I think it’s fair to say that the sentiment in the room was that this was not a fully satisfying explanation. We were less clear about why. Could it be that we think there should be a reason to believe in the supernatural that is independent of our need to solve a particular intellectual problem like this one? It’s hard to put the point.