Why Hume's Argument against Miracles Fails

David Hume famously argued against miracles, or more strictly against the rationality of believing in miracles, two and a half centuries ago. And while it was subject to objections immediately -- most notably that it is question-begging -- it is still influential, though, ironically, less in philosophy than in other fields. He presents it in section 10 of his Enquiry Concerning Human Understanding (click the link and scroll down to page 55). My impression is that his argument is very widely rejected among philosophers -- even the Hume scholar Antony Flew, in his atheist days, argued that, as written, Hume's argument was unsuccessful -- but there are some who defend it. John Earman wrote Hume's Abject Failure: The Argument against Miracles, but then Robert Fogelin wrote A Defense of Hume on Miracles. I'm afraid I haven't read either book. Yet.

Roughly, the argument is as follows: the more frequently something occurs the more probable it is that it will occur again, and the less frequently it occurs the less probable it is that it will occur again. A miracle would be the rarest of all events, totally unique, and would go contrary to our universal experience. Therefore it would be maximally improbable: virtually any explanation would be more probable than that a miracle took place. The usual example of an alleged miracle that enters into the discussion is Jesus' resurrection. According to Hume, regardless of the evidence, it would be more rational to accept any theory, even a conspiracy theory, than to accept the resurrection. So the idea that everyone just went to the wrong tomb, or that Jesus' survived the crucifixion, or that Jesus had an evil twin (these are all real alternative explanations that have been given) are more rational than that a miracle took place.

It seems to me that Hume's argument makes three assumptions that are not justified: that the universe is mechanistic, static, and closed. The mechanistic part essentially means that there is no event in the universe that is not completely determined by the physical events preceding it. The problem with this is that we have reason for thinking this is not the case. On one side we have the apparent randomness of quantum events. I say "apparent" because it's not necessary to believe that these events really are random. Bohmian mechanics explains just as much data as the Copenhagen interpretation, but Bohmian mechanics is strictly deterministic. Nevertheless, it seems to be more ad hoc than explanations which grant that some quantum events are not inevitably produced by preceding events. So many scientists would deny that the universe is mechanistic in the way that Hume's argument requires. On the other side we have the possibility of free agents with free will whose actions are not completely determined by the physical events preceding them. Naturally, free will is an enormous and enormously controversial subject, but at least we can say that most people intuitively believe in it, the arguments against it are indecisive, and the arguments for it, also indecisive, are, nevertheless, reasonably strong. Of course neither of these issues (quantum indeterminacy and free will) speak directly to the possibility of miracles, but they do speak directly to Hume's argument against miracles, since his argument only makes sense within a mechanistic framework.

Hume's second assumption is that the universe is static. This is not the same thing as being mechanistic: a system could be mechanistic without being static, but a static system would automatically be mechanistic. Here the idea is that even if the universe is mechanistic, it is not necessarily static in the way Hume's argument requires. New things may occur, first time events may take place, even though everything is mechanistic and determined. Take a system in which a drop from a river flowing through a forest is deposited in a large basin at a higher altitude every minute or so. This system is completely mechanistic. Accounting for evaporation, say that after 10,000 years, the amount of the water is so immense that the basin splits and disintegrates, dumping the huge mass of water on the forest and river below, obliterating the whole area and completely changing the environment. Ex hypothesi that had never happened before. Using Hume's criteria, we should assess the probability that it would happen as maximally improbable. But clearly it's not. In fact, we could probably predict that something like that would happen. Again, this doesn't speak directly to the possibility of miracles but to Hume's argument against them. If the universe isn't static then there is no reason to accept his probability account.

The third assumption is the big one: that the universe is closed. This is basically a challenge to the idea that the more something happens within a system the more probable it is that it will happen again. But if the universe is open, then there may be forces that exist independently of the universe which can produce effects within it. In fact, these could potentially even be mechanistic, static supernatural causes, but due to their transcending the universe do not happen predictably within our space-time frame.

Of course, the meaning of "miracle" is "sign": it's a sign of a transcendent agent acting in order to bring about an effect within the universe. An unprecedented supernatural event that just seemed random would not bespeak of an agent who intended to bring it about for a reason, but miracles do precisely that. With Jesus' resurrection, it's pretty obvious. To paraphrase Wolfhart Pannenberg, the significance of the resurrection is not that some guy came back to life, it's that this particular guy did -- a guy who acted and spoke as if he had the authority that only God can have, and was arrested for blasphemy and put to death for sedition. For that guy to rise from the dead amounts to his divine vindication. Of course, this is not a logical absolute, it's a common sense assessment. If there were a mechanistic, static, yet transcendent cause that brought it about, it would be an enormous coincidence that a guy who was executed for making radical claims about God and the resurrection of the dead would effectively have his execution reversed; that by sheer coincidence a resurrection from the dead just happened to that one guy who said death could not hold him. That would be incredibly improbable. But of course no one is making that claim.