“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…” – David Hume, Of Miracles.
The above statement summarizes David Hume’s essay on miracles. There has been a fierce debate over what exactly Hume meant with his arguments, but I choose not to focus on that debate because I see it as a red herring. Who cares what Hume thought? What we’re really interested in is whether it is ever acceptable to believe a claimed miracle, and if so, how much evidence we should require for such a claim.
I define a miracle as an event caused by a supernatural agency. What is the probability that miracles occur at all? Without calling on the evidence that they do not occur, there are good arguments for the impossibility of supernatural agents (see argument 1 here), and at least one good argument for the impossibility of a supernatural agent causing anything in the material world even if such a being did exist (The apparent impossibility of an immaterial thing causing a material event without any point of contact between the two; That is, an immaterial entity, since it is not material, would not in anyway whatsover be touching or physically connected to any matter or energy, and without such a point of contact it is quite plausibly not even possible for one to affect the other). Ockham’s razor would suggest that we keep our view of the world simple, and as such we shouldn’t add supernatural beings into our account of how the universe and all of existence works unless proven otherwise. But set all of these arguments aside. Let’s begin by being generous and assuming right out of the gate that there is a fifty percent chance that miracles happen occasionally (that probability will be raised or lowered based upon our examination of the evidence).
First of all, how do we tell when a miracle has happened? It is not so easy to tell when an invisible and undetectable agent has caused something, after all. We need to look for events that are unlikely to occur unless a supernatural agent was involved. On the other hand, so many events happen in the world that even super-unlikely events will happen every once in a while (for example, lottery winners who won multiple times). How to find our way out of the maze? Well, if seemingly supernatural events happened more often than was likely if there were no miracles, the problem would be solved. It’s hard to say exactly how often such events would need to happen in order to justify a belief in the supernatural, but obviously if these were common and repeatable, no one would doubt that they occurred. Fortunately, I don’t think the real world is ambiguous enough to throw us into debating how much is enough.
So we know what a miracle is and how to identify it. I haven’t observed such an event myself, but there are people throughout history who have claimed to have observed such an event. How do we treat their testimony?
Let’s look at this with an analogy I’ve used in my book Extraordinary Claims, Extraordinary Evidence: Lots of people have claimed to be able to create perpetual motion machines, that is, machines, that work without losing any energy at all. Of course, this violates the second law of thermodynamics (or the first, if it is claimed that the machine creates its own energy for free). We have two possible theories: either the laws of thermodynamics cannot be broken or they can be broken occasionally. If the laws cannot be broken, what should we expect? We ought to expect that the vast majority of human beings will attest that it has been constant in their experience, but we may have a few people who say otherwise because they are liars (we all know people lie sometimes) or just plain mistaken about what they thought they saw. If the laws of thermodynamics are not truly laws, but just general rules which have some exceptions, we ought to expect a similar pattern (most people will say it holds good in their experience, a few will attest to an exception). For believers in perpetual motion, we are already knee-deep in trouble. They cannot trust someone who told them about perpetual motion because the testimony here is not nearly as trustworthy as it would be about an ordinary event: we expect a few people might attest to perpetual motion even if it is a myth, and thus the word of their perpetual-motion-machine-selling uncle provides no good grounds for belief. On the other hand, there might be a key difference in the kind of testimony we would have if the laws of thermodynamics were not constant: exceptions could be found and repeated, hundreds of scientists could investigate and decide the matter. If they were in unanimous (or even near-unanimous) agreement that they had indeed found an exception to the law, and had no reason to lie about the event (their palms weren’t greased by a savvy and well-financed perpetual motion entrepreneur) I would judge such an outcome to be so unlikely on the hypothesis that the law was constant that I would immediately believe in the exception. However, that such a well-recorded exception has never happened strikes me as a pretty good argument that no exception exists. On this evidence, the constant-law hypothesis may not be one hundred or even ninety nine percent likely, but it is a great deal more likely to be true than false.
But a fan of perpetual motion is in even worse shape than I have let on: a huge number of claimed perpetual motion machines have been investigated, and all of them have turned out to be phonies, and the US Patent Office refuses to even consider them anymore. So, the odds that the next perpetual motion claim is right is pretty slim. So far, such claims have been correct exactly zero percent of the time. Maybe we’ve just never learned out to assemble one properly, maybe on the edge of unseen physics awaits a perfectly workable one, but the track record we have gives solid reason to actively disbelieve any such claims unless and until extremely strong evidence is provided to outweigh the evidence we have from our running tally of failures.
Miracle claims are in a similar category: only a few people ever attest to anything that could definitely count as a miracle (in my experience, most people who make such claims tell stories of seemingly unlikely events or vague phenomenon, neither of which is conclusive because the improbable will happen occasionally regardless of whether the supernatural exists) which is just what we would expect if no miracles ever occurred (after all, a few will lie, hallucinate, be mistaken, and so on). Just like perpetual motion machines, most claimed miracles that have been investigated have been found false or of questionable veracity. J. C. Tierney of the International Marian Research institute says that “Out of the 386 apparitions, the Church has decided that ‘yes’ there is a supernatural character only in eight cases.” Even the eight cases deemed genuine by the church are probably not real, in my opinion. Or, as philosopher Matt McCormick once wrote: “[Lourdes, France recieves] over 80,000 pilgrims a year for over a century… Suppose, charitably, that half experienced something they took to be supernatural. [Sixty-six] miracles have been declared real by the official investigating body of the Catholic church.” (Update: see addendum for further discussion of this point). Doing the math, it works out that only about one miracle claim in every hundred thousand is genuine, and again, the church’s judgement is probably not to be accepted so uncritically and thus more realistically we ought to deem accurate miracle claims even less frequent than one per hundred thousand.
Hence, if you agree with my conclusion about perpetual motion machines, you must agree with my conclusion about miracles, because the two are exactly the same, and therefore any particular miracle claim must be backed by strong, powerful, airtight evidence to be believed. The vast majority of human testimony, even testimony by more than one person, just isn’t enough. Still, what follows from this? Will we never be justified in believing that a miracle has occurred? I think not. All this means is that testimony of average (or even a little above average) credibility cannot prove a miracle even if several other people corroborate the testimony. But there are theoretical circumstances in which we could judge a miracle likely. As Hume himself put it: “[S]uppose all authors, in all languages, agree, that, from the first of January 1600, there was a total darkness over the whole earth for eight days: suppose…that all travelers, who return from foreign countries, bring us accounts of the same tradition…it is evident, that our present philosophers…ought to receive it as certain, and ought to search for the causes whence it might be derived. The decay, corruption, and dissolution of nature, is an event rendered probable by so many analogies, that any phenomenon, which seems to have a tendency towards that catastrophe, comes within the reach of human testimony, if that testimony be very extensive and uniform.” I agree with Hume, though I don’t think the analogy to the decay and corruption of nature makes much difference either way. The scenario Hume describes would be super-improbable unless a miracle had occurred, regardless of what that miracle was (even if it was of the Buddha going around the world and healing people of all nations). I think it’d be fair to say that this type of testimony would have less than one chance in a million of existing if no miracle had occurred, and the miracle itself would arguably have a prior probability of one in a million, which would arguably make the miracle believable or would at least put it within the realm of rational debate (if you have trouble understanding the stuff about prior probability please learn about Bayes’ theorem with this blog post or, even better, this video.)
The same reasoning would apply to other alleged supernatural events. For example, many people have claimed that they have continued to have experiences after their heart has stopped. I believe them. Not because of the sheer number (in a world of six billion people, we surely have thousands and thousands of liars) but the percentage. Though I don’t have an official estimate, it’s my understanding that a fairly significant percentage of those who have flatlined have claimed to have had experiences during that time. It’s improbable that a large fraction of people would report this unless it were true, and without smuggling in any double-standard for judging these claims, it would be the height of irrational to disbelieve these people. Of course, I believe that these “near death experiences” are not glimpses of an afterlife but are generated by the mind of the patient (See Keith Augustine’s article on the subject or The NDE Delusion by PZ Myers). However, if we actually had a huge fraction (better yet, 100%) of people who not only testified to having an NDE but testified to it when we could confirm that their brain and nervous system had stopped functioning, then I would take this as powerful evidence of the supernatural. Further investigation (i.e. ruling out alternative hypotheses such as false memory creation) could even convince me this was real, as sure as stars are real. Why? Because unless we think the prior probability of a supernatural world is uber-tiny, that sort of evidence would be so unlikely on any other hypothesis that we’d have to believe it was true. But we don’t have that evidence, and my guess is we never will. What do you think the correct explanation for that is?
Addendum: On the Lourdes Cures.
Jayman has posted a critical commentary on McCormick’s argument from Lourdes. Though Jayman’s points merit consideration, I still feel that Lourdes examples aren’t very supportive of miraculous occurrences. See my comments and discussion with him here.