Hume on Miracles, part 1

Class notes for 28 February

Main points

After beginning with Hume’s reference to Tillotson’s argument, we discussed the following points.

  1. Hume’s discussion of belief based on experience, especially his distinction between proof and probability [handout]
  2. his discussion of our reasons to believe in what others tell us (a.k.a. ‘testimony’)
  3. his definition of miracles as violations of the laws of nature
  4. his apparent argument that we cannot have reason to believe someone’s testimony of having witnessed a miracle.

I say that he apparently argued that we cannot have reason to believe in miracles because it is not clear to me that he meant the first part to constitute an independent argument against our reasons for believing in miracles.

I think it’s at least possible that he used the first section to set up standards for believing in miracles based on testimony and the second section to show that the testimony supporting Christian miracles fails to meet this standard. We’ll talk about this next time.


We began with the argument from Tillotson that Hume refers to in the first paragraph. The whole sermon is 6 pages long; the argument is on p. 3.

I said that I thought Hume’s reason for citing Tillotson was Tillotson’s argument that the reason for believing Christian doctrine depends on believing the testimony of the New Testament’s authors and sources, especially concerning Christ’s miracles. I think this was as close as Hume dared to come to pointing out that the arguments in this section of the Enquiry strike at the foundations not just of the Catholic wing of Christianity but at those holding up the Protestant one as well.

The split between Catholics and Protestants is evident in Tillotson’s sermon. More generally, the two main branches of Christianity were divided over miracles, at least in the seventeenth century (the Enquiry was published in 1745, by which time things might have been different; I only know about the earlier period). Protestants generally held that God used miracles to establish the church and that once this was accomplished God ceased producing them. You can see this view in Hobbes’s chapters on miracles and prophets in Leviathan, for instance, as well as in more devout Protestants of the time. Catholics held the opposite view, that miracles continue to occur.

So there was a tradition of casting doubt on reports of miracles in Protestant countries. I think Hume meant to push the argument one step further. The kinds of doubts that Protestants raise about miracles now can also be applied to all of the miracles cited in the Christian scripture.

What’s a law of nature?

A miracle is defined as a violation of the laws of nature. But, as Emily and Nathana asked, what’s a law of nature for Hume?

I said that I thought the most a law of nature could be is a well-established constant conjunction between two events. But there are still problems. Hume has already insisted that we don’t really know how any event or thing could make another event or thing necessary. That has to apply to these ‘laws’ as well. They don’t make things happen or even illuminate why things happen. The most they can be is really good generalizations.

So when a miracle is reported, the problem isn’t that it contradicts the way we know things have to be. Nathana made this point quite forcefully: for Hume, almost anything is possible. Rather, the point has to be that the evidence in favor of the law of nature is stronger than the evidence provided by the person testifying to the miraculous violation of the law of nature.

I asked whether this gets Hume into trouble concerning a different set of beliefs: those that we arrive at as a result of making observations that contradict what we had thought was a law of nature. For example, people observed for millions of years that unsupported inanimate (note: that handles birds — hah!) objects move towards the center of the Earth. But at some point, someone noted that this wasn’t true of planets. Their movements are primarily controlled by the gravitational pull of the sun.

But why was the first person to observe this justified in believing he had correctly observed something that was contrary to the laws of nature, as they were then understood? Didn’t the prohibitive weight of evidence favor the old, false putative law that unsupported inanimate objects move towards the center of the Earth? How could one observation outweigh the evidence provided by the billions of contrary observations?

Garrett and Joanna proposed two different ways that this might work. Both involved reinterpreting the old observations: people in the past weren’t really observing the movement of objects towards the center of the Earth, though that’s what they thought they were observing. Rather, they observed the operation of gravitational force.

Diana added that one thing distinguishing observations that support revisions to the laws of nature from observations of miracles is that the former can be repeated while the latter are (usually) singular. That would fit quite nicely with Hume’s point of view.

What’s a miracle?

Nathana suggested that Hume’s definition of a miracle as a violation of the laws of nature cooks the books in his favor. Perhaps it does.

There are certainly things that people find miraculous that don’t violate the laws of nature because the laws of nature don’t explain them. For instance, Augustine argued that the creation of the universe is a miracle (City of God, Book 10, Ch. 12). And Malebranche could be taken as saying something along the same lines about causation: we can’t understand how a cause makes its effects necessary except by supposing that a supernatural force does the work.

The laws of nature, as Hume understood them, wouldn’t apply in either case. We don’t have observations of the creation of universes, so we don’t have any laws that apply to that sort of thing. And, notoriously, our best efforts don’t reveal anything to the connection between cause and effect beyond constant conjunctions of things or events.

Second thoughts

Having thought about it a bit, it isn’t clear to me that the putative miracles I mentioned in the previous section are relevant.

Hume was trying to show that we have no reason to believe the Biblical stories of miracles that are used to prove Christ’s divinity. Christ’s rising from the dead is an example of such a miracle.

Those miracles are violations of the laws of nature, or, at least, they seem to be. If Hume’s aim was to undermine our reasons for believing them, then there may well be mysteries concerning the origin of the universe or causal necessity, but there would be no reason to believe the story of creation in Genesis as opposed to any other culture’s creation story.

Brad’s head

If he said he took his head off and handed it to a skeleton, you wouldn’t believe it. Brad Sugarman's head in a skeleton

But would you believe a picture? Oddly enough, he had already made such a thing, which he was kind enough to share.

Thus did Brad anticipate what I had thought was a perfectly good example of testimony to something so outlandish that no sensible person would believe it.

All of which goes to show that you should never assume anything.