The Self-Eliminating Nature of Miracles

There are two ways to live: You can live as if nothing is a miracle; you can live as if everything is a miracle. —Einstein

This post is inspired by a short paper from the pen of John Earman (2002), which I came across 20 years after its publication. This is a pity because I have long been intrigued by what I took to be David Hume’s (1748) devastating critique of the grounds by which many believe in miracles. Hume’s view is epitomized by his so-called maxim, which states that:

No testimony is sufficient to establish a miracle, unless the testimony be of such a kind, that its falsehood would be more miraculous than the fact which it endeavours to establish (1748, pp. 115-16).

In paraphrase, for a presumed miracle to be accepted, the probability that the witnesses’ reports are mistaken—either unwittingly or by cunning design—must be smaller than the probability of the factuality of the miracle in light of everything else we have learned about the workings of Nature.* This is a high bar, particularly in light of two facts: First, the empirical evidence for the unreliability of good-faith human perceivers and reporters is very strong—and so is the evidence for humans’ willingness to deceive. Second, there is no agreed-upon threshold the probability of an event must fall beneath so that the event qualifies as a miracle, but it is clear that this threshold must be very low, for otherwise we’d just be talking about unusual, rare, weird, or "random" events, but not miracles. The rare event must rather be regarded as "impossible" by both, those who would welcome a miracle and those who reject the possibility of miracles out of hand. Miracles, in other words, must violate laws of Nature. Miracles do this, according to the believers, because a higher power intervenes and suspends the lawful operations of Nature. The divine intervenor uses Her powers sparingly, lest miracles degenerate into something commonplace and lose their spiritual force.

David Hume "meets" Thomas Bayes

Earman’s (2002) essay is, in my opinion, brilliant because it has David Hume meet the Reverend Thomas Bayes, another Titan of the British Enlightenment—although as best we know the two never met in person. Earman links their thoughts through Richard Price, who was a friend and student of Bayes’s, and the publisher of his famous essay on probability. And Price had read Hume.

Earman suggests that Bayes developed his "apparatus" of probability after reading and doubting Hume’s skepticism of induction. Bayesian conceptions of probability now hold sway in many sciences. Earman seeks to show that with Bayesian statistics we find that ultimately, with enough evidence (i.e., witnesses), even the most outlandish claims eventually gain credence, unless, of course, the prior probability of the claim at stake (e.g., that a kiss can turn a prince into a frog) is exactly zero. And the belief that p = 0,, Earman suggests, we should not have because it reveals the closed mind of a zealot. The mind of a scientist must be at least a crack open to "anything."

What is a law of Nature?

The key assumption appears to be the idea that what we accept as laws of Nature is really only the result of innumerable converging observations. If the apple always falls down when we look, and never up, we presume the presence of a law that says apples must fall down, forever. But this is the sort of naïve induction by enumeration that Hume had exposed as logically unjustified. There might come an apple that falls up.

Really?

It seems to me that certain physical events must be so and can be no different, and such events can be regarded as lawful (Fiedler & Krueger, 2013). The laws of thermodynamics come to mind. Think of the effect of temperature on the state of aggregation of H20. Do we need more observations to know that higher temperatures will not turn water into ice? No matter how large the number of witnesses that say otherwise, the Bayesian conditional probability that heat produces ice will not come up from zero.

Earman’s sympathies lie with Hume—with his conclusions, that is, if not with the cogency of his method. As for me, I am happy to accept the folk psychological desire for miracles in the sense of good or relieving news in distressing times. A sudden and unexpected recovery may colloquially be called a miracle; a person rising after three days in the morgue is another story.

Why are miracles always good?

We may ask about bad miracles. Hume’s or Bayes’s arguments and calculations are value-free. Their "apparatuses" apply with equal force to negative events that violate laws of Nature. No one wants those, and what even are they? It is easy to imagine freak accidents and the like, but what is a dark miracle? A haunted house perhaps? Might there thus be a miracle-asymmetry? If we cannot conceive of something bad that "miraculously" materializes in opposition to the laws of Nature, what do we make of it? We might conclude that Nature is ultimately benign because she does not allow such things, or we might conclude that things are harsh enough given the Nature we know, and that’s why we want miracles.

Note: In this post, I have capitalized Nature because I consider Her divine. And: I asked Hoca Camide, "Hoca, what is the defining feature of a miracle?' "It doesn't replicate," Hoca replied, "and forget about pre-registration."