To: Daniel P.
From: Geoffrey Klempner
Subject: Hume on the nature of miracles
Date: 1st August 2008 11:46

Dear Daniel,

Thank you for your email of 22 July, with your essay towards the Associate Award, 'An Account of Hume's "Miracles".'

This is a very promising piece of work, which I enjoyed reading. I agree with your conclusion that all Hume has shown -- or could possibly show -- is that the occurrence of miracles defined as breaches of the laws of nature is highly improbable. However, your argument is weakened somewhat as a result of a misunderstanding of Hume's definition of a causal law.

I will start with this point, because it is important. As you relate, Hume argued that there cannot be an a priori defence of induction, concluding that it is 'custom' -- our innate tendency to form beliefs on the basis of the association of ideas -- which is the ultimate bedrock of all empirical knowledge. We believe what we believe. The only room for rational argument is within the context of the evaluation of experiences and reports of experiences, and the rules under which we subsume experience, all of which presupposes that belief formation is itself merely a matter of cause and effect.

Talk of what we 'ought' to believe (a point that seems to have bothered you) has to be understood within the context of the observation that it is part of the very concept of 'belief' that belief is not subject to the will: all that we have the power to do is attend to the arguments and the evidence to the best our ability, and form our beliefs accordingly.

Hume also, famously, attacks the traditional metaphysics of causation, arguing that there is no ultimate metaphysical link in virtue of which a statement of the form, 'A caused B' is true. A caused B, if true, is true by virtue of the fact that it instantiates a universal generalization. It is in this sense, that Hume provides a causal analysis of the process of belief formation itself.

What does this all add up to? No-one can ever be certain of the truth of a proposed 'law of nature', e.g. Newton's Law of Gravitational Attraction. The law can only be believed with a degree of probability which increases the more positive evidence we have for the law, and the continued absence of observations which appear inconsistent with that law.

However -- and this is the point which I think you have not seen -- the law, if true, is true for all times and places, in other words, it is 'an exceptionless regularity'. That is what Hume substitutes for the metaphysical link. The idea of a universal law is indeed extremely powerful. As a Humean about causation, I would be perfectly happy to accept that the only sense we can make of the notion of a 'cause' is in terms of the truth of universal laws.

There are still very considerable problems with this, the most notable being the problem of defining 'ceteris paribus' conditions. You can't derive a causal law from a statement like, 'The stone broke the window' because not all windows break when stones are thrown at them. If you tried to state the law in question, it would be very difficult if not impossible. The stone has to be such-and-such a weight, thrown with such-and-such force, against a window of such-and-such strength, in such-and-such atmospheric conditions, and so on.

Consider now Dr Frankenstein's claim that the bolt of electricity from the lightning strike brought the creature made from exhumed human body parts to life. According to our best scientific knowledge, this can't happen. Could this be a miracle?

Let the experiment be performed in an auditorium, filled with renowned scientists from every country. Hume would simply say, 'Well, it hasn't.' His argument against miracles unashamedly appeals to what we know from experience, not, in the sense that you suggest -- that Hume is merely talking about his own experience -- but rather in the sense that he would expect any reader, including religious believers, to allow: that in fact, the one thing all reports of miracles *to date* have in common is that they have not been performed in such an auditorium.

(Remember that science as an institution has not existed for that long. It is only recently that human beings have devised such things as university departments of physics, learned journals, scientific conferences etc. etc.)

What we have, so far, is an argument about probability. I think we are both agreed that this argument stands up. The question at issue is whether Hume has gone too far, in 'begging the question' over miracles by assuming that any reports of miracles to date must be false.

I don't think that he has. Let's start from scratch, Hume would say. Let's take all the reports of miracles of human beings of being brought back from the dead that have ever been made, or, rather, that we can get our hands on. All this is evidence. Likewise, we have medical evidence for beliefs such as, 'Brain death is irreversible', as well as evidence for underlying explanations for this observation in terms of brain biochemistry. Hume's claim is that the argument from probability overwhelmingly supports scepticism about such miracles.

To repeat the earlier point, IF we suddenly found ourselves in a 'Dawn of the Dead' scenario, we would rapidly change this assessment. But, as a matter of fact, we haven't. (Interestingly, zombie films always throw a sop to science, offering quasi-scientific explanations of how the zombie 'virus' operates. But they can also be taken literally: As the Pastor says in the Dawn of the Dead remake, 'When Hell is full, the dead will walk the earth.')

One issue I haven't addressed is whether Hume's aggressive raises suspicion about his motives for the very reasons Hume gives. Curiously, I didn't hear a 'bitter, caustic' undertone in the quote that you gave. My impression is that Hume's attitude was very different from, say, a contemporary militant atheist like Richard Dawkins. But in any case, Hume would say, 'Ignore my tone, just consider my arguments.'

All the best,

Geoffrey