On (Hume’s Philosophical Views On) Miracles

As an atheist, it is not surprising that David Hume did not believe the miraculous claims made in the Bible or in the sacred stories of other religions. More than not believing in miracles himself, though, Hume argued that it is never rational for anyone to believe claims of miraculous events. The very meaning of the word “miracle” makes it impossible for such claims not to be promptly disproved by the overwhelming body of evidence against them. More interestingly, Hume appears to say it would be irrational to believe in a miraculous event even if the event really did happen. This leads to an epistemological system in which it is sometimes more rational to believe a false proposition than a true one. However, counterintuitive as it may seem, this system is actually quite effective for ensuring that one does not add new beliefs without having sufficient evidence for them.

According to Hume, we can define a miracle as “a transgression of a law of nature by a particular volition of the Deity or by the interposition of some invisible agent.”[1] Whether an event appears to human observers to be a violation of natural laws plays no part in whether it is miraculous: “the raising of a feather when the wind wants ever so little of a force requisite for that purpose is as real a miracle, though not so sensible with regard to us.”[2] The fact that some invisible agent, typically God, is supposed to have caused all miracles claimed to have happened thus far also doesn’t figure prominently in Hume’s argument. We can safely suppose that he would have as much of a problem with the claim that natural laws were broken in a completely random manner as he would with the claim that they were broken by God or some other invisible agent, because the most important part of the definition is simply that miracles are transgressions of a law of nature. To be considered a law of nature, something must be uniformly in accordance with a huge body of experience. The entirety of this experience, in turn, weighs overwhelmingly against any claim that the law in question has been violated.

In general, whenever there are multiple contrary claims, whichever one is most in accordance with previous experience is the one most rationally believed, and it should be believed with a level of certainty corresponding to the superiority of the evidence for it over that for the other claims:

It is experience only which gives authority to human testimony and it is the same experience which assures us of the laws of nature. When, therefore, these two kinds of experience are contrary, we have nothing to do but subtract the one from the other and embrace an opinion either on one side or the other with that assurance which arises from the remainder.[3]

Regarding miracles in particular, it is rational to believe miraculous claims only when the falsehood of all the evidence in favor of a claim would be more miraculous than whatever the claim seeks to establish. Since most miracles are supported by human testimony alone, and since there is nothing particularly miraculous about people lying or being mistaken, in all cases so far, “no testimony for any kind of miracle has ever amounted to a probability, much less to a proof.”[4] Even if you, possibly along with many other people, seemed to witness a miracle with your own eyes, it would in general be far more in accordance with previous experience for (all of) you to be hallucinating than for the event to have occurred in reality. The “uniform experience against every miraculous event…amounts to a proof. [So] there is here a direct and full proof, from the nature of the fact, against the existence of any miracle.”[5]

However, what Hume must mean here is that there is proof against every miraculous event claimed so far to have happened. He can’t mean there is a direct and full proof against the future possibility of any miracle, as he later admits that “there may possibly be miracles or violations of the usual course of nature of such a kind as to admit of proof from human testimony.”[6] The possibility of miracles admitting of proof certainly implies the possibility of miraculous events that go unproven. What is interesting about this is that, taken together with Hume’s argument that we ought not in general believe miraculous claims, it means there are possible situations in which we ought to have false beliefs. If a miraculous event actually did happen, but the only evidence a particular person has for it is the testimony of someone else, then it is almost always more rational for that person to believe the other is lying than to believe the event really occurred. In fact, it is usually most rational for the person to believe with the highest possible certainty that the other is lying. The uniform experience behind the natural laws supposedly broken so completely overwhelms the statement of one person, even a generally trustworthy person, as to remain almost undiminished when we “subtract the one [piece of evidence] from the other.”

On the face of it, this appears to be a fairly reasonable injunction: Don’t believe something just because one person told you it was true. At the same time, we generally think of reason as being a good way to increase the amount of true beliefs we have, and yet here reason tells us to believe in something false. Reason furthermore tells us in this case to believe the false thing with the utmost certainty. This is not restricted to claims about events everyone in the world would consider miraculous, either. Rather, it applies to claims about any event that transgresses what one’s own experience has established as an apparent law of nature. For a contemporary of Hume who lived his or her entire life in southern India, and who had not received any education in physics or chemistry, experience about the phases of water would have been universal enough to establish as a “law” of nature, in that person’s mind, that water can only be liquid or gaseous. If one British visitor tells the Indian that water can, in fact, also be solid, the latter’s universal experience would so overwhelmingly outweigh the Briton’s testimony as to amount to a “proof” against the perfectly true claim that ice exists. Hume seems to be saying we should prefer to have false belief more justified by our personal experience over a true belief that is not justified by our personal experience, but shouldn’t we rather have true beliefs than false ones, regardless of the amount of justification?

In a word, no. Without justification, beliefs, however true they may be, do not count as knowledge, and our (or at least Hume’s) real goal is not just to increase the number of true beliefs we have, but to increase the number of things we know. But false beliefs, however justified they may be, do not count as knowledge, either. If we shouldn’t prefer true beliefs over false ones, regardless of the amount of justification, then why should we prefer justified beliefs over unjustified ones, regardless of their truth? The reason Hume would most likely give is that past experience tells us that beliefs more justified by experience tend to be true more often than those that are not so justified. This isn’t always the case, of course,[7] especially considering that everyone’s set of experience is finite. Many propositions we would reasonably doubt on some level of experience ought to be believed once we have more of a certain kind of experience. For instance, if the southern Indian visited the Himalayas he or she would certainly be provided with enough experiences of ice to render the belief that water can be solid far more justified than its negation. Similarly, if Hume was told by one friend the truly miraculous claim that the seventy-two hours preceding his birth were marked by constant daylight, he would be completely reasonable to doubt this friend. But if Hume then asked other acquaintances and had the story confirmed, and found even upon traveling abroad that everyone alive at the time had accounts of three days of sunlight, he would likely consider this to be testimonial proof that the event had really occurred.

Even if we agree that one ought not to believe a claim until there is sufficient experience backing up the claim (and most people do seem to agree with this, at least implicitly), we seem to be in the unfortunate position of just having produced two contradictory “proofs”. Hume, in our hypothetical scenario, first has a body of previous experience sufficient to prove false his friend’s claim that 72 hours of daylight preceded his birth. Then, after hearing accounts from a number of people all over the world confirming this claim, he has a body of experience sufficient to prove true the very same claim. Clearly the two proofs cannot both be sound and valid, and since both are of essentially the same form, this seems to mean that neither are really deserving of being called proofs. But we must remember that even the idea of necessary connection, for Hume, comes only from experience. As such, we cannot expect proofs for him to have the same level of necessity, in the sense that valid arguments necessarily preserve truth, with which we today generally associate mathematical or geometrical proofs. Proofs do not necessarily produce true conclusions, they merely produce conclusions we can accept with as much assurance as we can have about anything based in experience.

[1] Hume, David. An Inquiry Concerning Human Understanding. Note on p. 537 c. 1 of Modern Philosophy. (emphasis in original)

[5] Ibid. p. 536 c. 2. (emphasis in original)

[7] Hume writes, on p. 534 c. 2: “Though experience is our only guide in reasoning concerning matters of fact, it must be acknowledged that this guide is not altogether infallible, but in some cases is apt to lead us into errors.”