Hume on miracles, part 1 Notes for October 26

Main points

We started with some background to Hume’s discussion of miracles. He was showing how an argument Protestants used against Catholicism undermined all versions of Christianity. Whoops!

Then we turned to Hume’s argument in part I. Part I treats the testimony on behalf of miracles as a “proof.” This is confusing since the whole point of the argument is that it is no such thing. Here is my (disputable) interpretation. In Part I he meant to grant for the sake of argument that we have no reason to doubt the testimony. Suppose the testimony were accepted as a proof; it will never overcome the balance of evidence against it. (As we will see, he withdraws this supposition in part II on the grounds that other instances of testimony about miracles are deeply flawed.)

What did he claim?

There are three things to note about what Hume claimed for his conclusion.

First, he claimed that we should dismiss reported miracles out of hand. This is much stronger than saying we should set a high standard of proof or rigorously investigate reported miracles. Hume claimed this would be misguided and that any investigation of miracles is a waste of time. Whether he showed that is a different matter, of course.

Second, as Daniel pointed out, Hume’s argument is about what we have reasons to believe. He did not try to show that miracles are impossible. Given what he said about causation, he could not show that. His standard there was “if it is conceivable, then it’s possible.” Miracles are clearly conceivable, so they are possible by his standards.

One question that we will have to confront eventually concerns what difference, if any, there is between:

  1. We have no reason to draw inductive inferences, inferences about cause and effect.
  2. We have no reason to believe reported miracles.

Hume claimed that the evidence for the miraculous is outweighed by the contrary evidence for the laws of nature. So there had better be evidence for the laws of nature. It’s not easy to see how to square that with Hume’s skepticism about induction, though.

Finally, as Harley noticed, the argument probably doesn’t stop with the testimony of others. There’s no reason why it wouldn’t apply to what you (think you) see with your own eyes. Hume didn’t apply it in that way himself, but it’s hard to see how he could have distinguished the two cases or why he would have wanted to do so.


Does Hume’s argument prove too much? That is, did he just say that we should always dismiss any novel or unexpected experience, whether we have it or others do?

That’s the essence of the questions that Daniel and Camille asked. If that’s his conclusion, then it seems way too strong. We would never learn anything! In any event, we’ll begin next time by talking about this problem.

This page was written by Michael Green for Problems of Philosophy, Philosophy 1, Fall 2009. It was posted October 27, 2009.
Name of website