I just recently got my copy of John Loftus’ new edition of Why I Became An Atheist: A Former Preacher Rejects Christianity, and I was reading his chapter on miracles. Loftus believes (as do I) that Hume was basically right about miracles, and although Hume’s case may need a little bit of modification or qualification in some areas, he was on the right track. Loftus takes on the counter-arguments of Christian apologist and New Testament scholar Michael Licona in order to show that Hume’s case still stands, since Licona’s counter-arguments are really very typical of those used by Christian apologists, the evangelical community in particular.
Loftus does a fine job of responding to Licona, but I think there are even more problems with what Licona says than Loftus discusses, and so this post will discuss those problems.
Apologists often accuse skeptics of being “dogmatic” and “close-minded.” I think if you look closely enough you’ll find quite a bit of those same flaws not only among skeptics, but amongst believers as well. I would really like to see a Christian taking Hume seriously and attempting to interpret him charitably, to look at modern reconstructions of his arguments, and to give a thoughtful response. Licona’s response to Hume, unfortunately, misses this mark. This is not an attack on him personally. I know for a fact that Licona is an honest guy, and he has shown a willingness to correct his mistakes in the past. I’m hoping that this will prompt him to take Hume more seriously and to do a better job responding to Hume the next time he writes about the subject.
The pages of Licona’s work can be read on google books, and the key section is his chapter on David Hume which begins on page 135. I’ll also be referring frequently to Hume’s essay “Of Miracles.”
Can a miracle ever be established?
Although Hume’s remarks are often interpreted to mean that one can never establish the occurrence of a miracle, even in theory, there are tell-tale signs that this isn’t quite what Hume meant, see Robert J. Fogelin’s A Defense of Hume on Miracles. In any case, it’s perfectly possible to adopt the position that miracles, while not impossible nor impossible to establish as true, must meet a tall burden of proof before they can be believed, and this is the position I hold.
Hume’s Four Arguments
Hume first argues that:
“There is not to be found, in all history, any miracle attested by a sufficient number of men, of such unquestioned good-sense, education, and learning, as to secure us against all delusion in themselves; of such undoubted integrity, as to place them beyond all suspicion of any design to deceive others; of such credit and reputation in the eyes of mankind, as to have a great deal to lose in case of their being detected in any falsehood; and at the same time, attesting facts, performed in such a public manner, and in so celebrated a part of the world, as to render the detection unavoidable: All which circumstances are requisite to give us a full assurance in the testimony of men.”
Licona unwittingly puts words on Hume’s lips by saying that, “According to Hume, the testimony of any reliable event must have the [above] credentials in order to qualify as a historical event.” (p.138) The trouble is, Hume never says that “any reliable event” must meet those standards. Indeed, it isn’t very plausible to suppose Hume believed anything like that, since, as Licona himself points out, most everything we know about the past wouldn’t meet those standards (and, I’d add, neither would a large number of our everyday, common-sense beliefs). Let’s try a different and more charitable interpretation of Hume. We can interpret Hume’s argument into Bayes’ theorem: Let’s look at the whole of human history and test the hypothesis that miracles occur against the hypothesis that they do not. If miracles do occur, it is completely possible that Hume’s stringent standards would be met, and perhaps even met hundreds of times. On the other hand, if miracles never happen, it would be more probable to look at human history and find that these standards have never been met even once. In fact, it would very improbable under the thesis that “miracles don’t happen” for such stringent conditions to be met even a single time, but somewhat more probable if miracles do take place from time. After all, if miracles are real then this certainly could happen. It’s a bit like perpetual motion machines: if perpetual motion machines were real and actually worked, we might well expect that a large number of reliable witnesses of the utmost education would come to believe in them and that machine would never be debunked. On the other hand, if perpetual motion machines are completely bogus then the absence of such airtight testimony is much better explained. Now all of this doesn’t count as a conclusive refutation of miracles (or a conclusive refutation against perpetual motion machines) but it is a powerful argument that adds weight to Hume’s case.
Licona gives a fairly accurate summary of Hume’s next argument when he writes, “Hume’s argument concerning the intelligence and integrity of the witnesses of miracles makes three claims: Testimonies of miracles abound among the ignorant and uneducated; they do not occur in modern times; and deceitful witnesses abound.”
Licona rebuts Hume’s first point by noting that intelligent and educated witnesses to miracles also exist. This is hardly relevant, as Hume’s exact words are “[Miracles] are observed chiefly [Not exclusively, but chiefly or mostly] to abound among ignorant and barbarous nations…” Licona is no doubt correct that there have been intelligent and educated folks who claimed to have seen a miracle, but so what? The same could be said about many fantastic claims, including UFO reports. That a person is smart or well-studied is not an airtight guarantee that they will not mistake a normal phenomenon for a paranormal one, that they will not hallucinate, or that they will not lie. Hume’s point is that if miracles don’t happen, then the reports of them ought to be much less frequent among the educated and intelligent than among the uneducated and unintelligent. I personally am not sure about the strength of this argument, as one might be able to give other reasons for the relative frequency differences of miracle claims between educated and uneducated people even under the assumption that miracles do occur.
Licona rebuts Hume’s second contention, that miracles “do not occur in modern times” by comparing it to the principle of analogy, which one book describes as “causes used describe past events must be similar to or analogous to causes of events in the present.” I would agree with Licona that this reasoning ought not to be seen as absolute. Antony Flew made a mistake in saying that all relics of the past must be interpreted by assuming the uniformity of natural law observed in the present (he once said that, for example, that to make sense of a text one must assume that texts are produced the same way back then as they are now, by human penstrokes). While this view is common sensical, I would weaken it a little bit and say that it works just as well if we assume that causes tend to operate almost always the way we observe them to operate presently. We don’t need to assume that the words inscribed on a statue are 100 percent likely to be the work of human hands, we can assume they are over 99 percent likely to be the work of human hands and get the same historical results in most cases while at the same time avoiding dogmatism. Licona seems to think that the principle of analogy would lead one to conclude that dinosaurs never existed because they don’t exist nowadays, but that’s silly: something like the principle of analogy (or at least my weakened version of it) is the only thing that could lead you to conclude dinosaurs exist! Sets of bones are always left behind by previously living animals in our experience. If one refuses to conclude from this statement that it is therefore likely that any bones we find are very probably the remnants of a corresponding animal that once lived, then you would have a really hard time coming up with a way to reason from the existence of fossils to the past existence of dinosaurs! You would also be illogically violating Bayes’ theorem and rejecting the validity of inductive reasoning, which all people use, and must use. The cases discussed here are parallel to that of miracles: true miracle claims, if they even exist, are rare birds among a vast sea of false miracle claims. It is therefore the case that really strong evidence must be adduced for them before they can be believed. Please see the addendum here on miracles and dinosaurs for a more thorough refutation of this claim.
Licona says that the principle of analogy might rule out new scientific phenomena. My version of the principle doesn’t, since I only think that new phenomena ought to be viewed as improbable prior to finding sufficient evidence for them.
After telling us that the principle of analogy assumes metaphysical naturalism (it doesn’t, see definitions given above), Licona contradicts himself by saying that miracles might even be consistent with the principle of analogy, since he believes miracles occur today. Whether miracles occur today or not is a topic for another day, but my view is that they don’t, and so it should be beyond dispute that historical miracles have a low prior probability from the context of holding a belief that they don’t happen nowadays.
Licona quotes C.S. Lewis’ rather silly interpretation of Hume which stated that Hume was arguing in a circle (by assuming that there is uniform testimony against miracles, which can be the case only if all the reports of miracles are false, which we can only know for sure if there are no miracles). Look at what Hume actually says:
“It is no miracle that a man, seemingly in good health, should die on a sudden: because such a kind of death, though more unusual than any other, has yet been frequently observed to happen. But it is a miracle, that a dead man should come to life; because that has never been observed, in any age or country. There must, therefore, be a uniform experience against every miraculous event, otherwise the event would not merit that appellation. And as an uniform experience amounts to a proof, there is here a direct and full proof, from the nature of the fact, against the existence of any miracle; nor can such a proof be destroyed, or the miracle rendered credible, but by an opposite proof, which is superior.”
Notice that Hume specifies ‘uniform experience against’ an event as a prerequisite for showing that such an event is a miracle, meaning the experience must be uniform outside that particular example of the miraculous happening. Though Hume labels ‘uniform experience against’ as a proof, we must understand that by ‘proof’ Hume meant only a very strong probability, see Peter Millican’s 20 Questions about Hume and Miracles, page 2. I myself don’t think that Hume’s prerequisite is a necessary one in all cases, as I can imagine that there might be routine happenings that are best explained by a supernatural power (like if prayer were demonstrated to work better than chance on a consistent basis). I further doubt that human testimony can be doubted on the sole basis of whether it attests to an unlikely or infrequent event. We’ll discuss this again later on.
Licona adds that he doesn’t think skeptics would be able to disprove each and every miracle claim ever made, and that even one demonstrates the supernatural exists. One wonders why Licona doesn’t show that at least one or two of these miracle claims meets a stringent burden of proof, considering the vast number of claims that have been made. Though Licona is right that even a single miracle would demonstrate the existence of the supernatural, he is no doubt committing the virgin Mary fallacy by apparently thinking that simply pointing to a large number of claims does the trick. On that, see the third and fourth paragraphs of this blog post.
Hume says, “We ought to give preference to such as are founded on the greatest number of past experience.” Licona disputes even this, saying that it would rule out improbable (I think he means explanations with a low prior probability) and unique explanations (which would also tend to have low but non-zero prior probability) as being the best ones. Even if he were right about this, logic can never be wrong, and probability considerations by definition are wrong only on occasion, so any explanation that doesn’t live up to such considerations is mistaken, period. More to the point, he’s wrong because it is possible to have enough evidence to show that a unique or rare explanation has a high final probability. Licona says that we could never conclude that one specific lottery player won under such considerations of probability, but that’s sorely mistaken. Having a ticket with all the right numbers on it counts as extraordinary evidence, since winning tickets themselves are as rare as lottery winners.
Licona’s lottery argument would gain some strength if he used the following example (which is similar to one that has been discussed in the journal literature on the subject):
“Suppose that there is a newspaper that reports a certain lottery sequence winning, and the odds of this exact sequence turning up is one in ten million. Suppose also that the paper is known to make a mistake one time in ten thousand (that is, one in every ten thousand facts the paper reports is false). Would it be thousands of times more likely that the paper was mistaken than that the exact lottery number was drawn?”
No, sure wouldn’t. Pay very close attention to what follows. I mean it. Focus, because there is a very subtle distinction that I need to get across: It is illogical to compare the probability of a certain number coming up with the probability of the paper making a mistake in order to figure out the probability that the paper in fact made a mistake. Instead, one should compare the probability of the paper being mistaken with the probability of the paper being correct. If we do that, we will find that (as implied in the information given above) the chances that the paper is wrong concerning the lottery number is one in ten thousand. Likewise, miracles shouldn’t be doubted because they are rare occurrences, but because the claims are rarely true. This is not a distinction without difference: it might conceivably be the case that miracles are rare but that the claims of miracles are almost always true. Theoretically, we could’ve verified that 95% of all miracle claims are true, but we haven’t. In fact, we’ve verified exactly the opposite, as I discuss extensively in my post Of Miracles.
Similar considerations apply to the much-discussed “Indian Prince” objection, on which I disagree with Hume. Hume thought that an Indian prince who had never seen ice would be justified in disbelieving reports of it from other people, for similar reasons that we are justified in disbelieving miracle reports. But Hume fails to realize that a report of ice would have a high probability of being true, since most people tell the truth most of the time. We can’t use the general reliability of testimony for miracles, though, because miracles fall under a special category of human testimony that is known to be unreliable, whereas reports of unusual substances and states of matter do not. I’ve laid out my case against the reliability of miracle testimony in the above-linked essay ‘Of Miracles.’ To give a good analogy of what I mean: a wife wouldn’t trust her husband’s word that he has stuck to his diet, even if the husband’s words were generally true (he doesn’t lie about money, or what household work he has done, or a hundred other things), if the husband were known to cheat on his diet and lie about it. That’s because the husband’s statement about his diet falls under a special category in which he is not reliable. The same thing goes for miracles: while most human testimony can be trusted, testimony about miracles can’t be, because it is a type of testimony that is provably unreliable whereas other kinds of testimony are not.
Licona claims that Hume’s argument “would only be legitimate” if natural processes caused Jesus’ resurrection, but not if God caused it. None of the arguments that either I or Hume have given specify or logically imply only “natural processes” rendering Licona’s point moot. Even if we grant up front that God exists and can interact in the world, we still aren’t entitled to believe an improbable claim like “God raised Jesus from dead” since we know that most reports like that are false. We all believe that Marilyn Monroe existed and that she had the ability to write letters, but that doesn’t mean that the Marilyn Monroe letter possessed by your friend’s grandpa is authentic; we know for a fact that most of the “letters from Marilyn” sent to fans were in fact written by secretaries, and without an expert’s examination of the letter (or similarly stringent verification) we would all believe that Grandpa’s letter was probably just the same.
Hume’s Fourth Argument goes like this:
“It is impossible the religions of ancient Rome, of Turkey, of Siam, and of China should, all of them, be established on any solid foundation. Every miracle, therefore, pretended to have been wrought in any of these religions (and all of them abound in miracles), as its direct scope is to establish the particular system to which it is attributed; so has it the same force, though more indirectly, to overthrow every other system. In destroying a rival system, it likewise destroys the credit of those miracles, on which that system was established; so that all the prodigies of different religions are to be regarded as contrary facts, and the evidences of these prodigies, whether weak or strong, as opposite to each other…This argument…is not in reality different from the reasoning of a judge, who supposes that the credit of two witnesses, maintaining a crime against any one, is destroyed by the testimony of two others, who affirm him to have been two hundred leagues distant, at the same instant when the crime is said to have been committed.”
Licona ventures four arguments against this one:
(a) The evidence for opposing religious claims is not as good as the resurrection of Jesus.
This is not uniformly true. There are lots of miracle claims that have even better evidence for them than Jesus’ resurrection. The alleged witchcraft in Salem, Massachusetts and the appearances of the Virgin Mary in the last century come to mind.
(b) Plausible natural explanations exist for the claims of at least some other religions, whereas this “may not be the case” with Jesus.
This comment only makes sense if one believes that this argument rules out miracles completely, whereas Hume only seems to be using it to argue against the probability of miracles, and as such it stands unrebutted. Licona ought to admit the argument is successful in establishing the improbability of the resurrection, and his only response to it should be that the resurrection can overcome this improbability.
(c) The miracle reports of other religions might be true, and might have been done by the Christian God.
First, a great many miracle claims are so specific to certain religions that it isn’t feasible to suggest that the same God did them all. The Baptists can’t suppose that the miracles of Joseph Smith, Sai Baba, Uri Geller, or likewise figures are due to the God of Protestant Christianity. Second, if Licona can claim that his God did the miracles experience by Buddhists or others, then why can’t they claim that their supernatural forces were responsible for the resurrection (which would empty the resurrection of any power to persuade disbelievers)? Although Licona’s suggestion is logically possible, it isn’t believable. Incidentally, I think this is the weakest of Hume’s arguments since it cannot provide a purely logical/empirical case against miracle belief; the most that can be said for it is that the only way for religious believers to avoid contradiction is that they go for the implausible and highly arbitrary suggestion that their God is the one doing everybody’s miracles (or argue that the devil or some other spirit does them, which again raises the specter of how they can know that Jesus wasn’t raised by Satan). A similar and much better argument is this: since appearances of the Virgin Mary usually only happen to Catholics, since Asclepius only appears to his ancient Greek worshipers, and so on, it is highly probable that these things are usually generated by the human mind, and this again entails a low prior probability of miracle claims.
(d) Parallel reports do not imply an underlying cause. The point Licona makes is that, for example, parallel stories of the fictional Titan ship and the real-life Titanic. It’s difficult to see why Licona thinks this is an important point. Human error, hallucination, and deception produce most miracle claims, and it is, of course, possible in theory that Jesus’ resurrection, although parallel in many ways to other miracle reports, has different underlying cause. So what? Licona has not demonstrated that the probability is against this, and the probability must be assumed in favor of it, at least initially, unless we’re going to toss out inductive reasoning.
Conclusion
Licona’s numerous arguments against Hume miss the mark. While it is still possible in principle to prove the resurrection of Jesus (or any other miracle, for that matter), Licona cannot escape the fact that the resurrection has a very low prior probability, and as such Licona would need remarkable evidence in order to outweigh this prior improbability. Whether he can do this is a topic for another day, although I have written about it here and in my book Extraordinary Claims, Extraordinary Evidence, and the Resurrection of Jesus.
Addendum: Dinosaurs and Miracles
I’d like to take a step back and compare the hypothesis that “dinosaurs used to exist” with the hypothesis “Jesus was raised from the dead,” just to show how different they really are. For the sake of argument, we will look at the dinosaur hypothesis from the point of view of someone who has never encountered any evidence of dinosaurs, just to make the two cases as parallel as possible. Someone who with no knowledge of dinosaurs would know that dinosaurs are at least possible, since large reptiles don’t violate the inverse square law, like insects of that scale would. We have abundant scientific evidence that living things can vary in their characteristics greatly, witness the great and numerous differences between different breeds of dogs or cabbages. That means that varieties or species of reptiles might have existed in the past with different characteristics than are presently observed (John Loftus made a similar point in his book). On the other hand, we don’t know if miracles are possible. Miracles require the existence of disembodied minds which many philosophers doubt are possible for a variety of reasons. Miracles also presuppose that there is a way for non-material entities to interact with the material world without a point of contact, which seems at least dubious. Miracles require presuppose that these disembodied minds actually do meddle in the material world, but maybe they don’t, as the Deists believe. We have plenty of knowledge entailing the existence of dinosaurs is at least possible, but with miracles not only do we lack such knowledge, we in fact have knowledge showing the very possibility is dubious. Even if the guy who lacked knowledge of dinosaurs wished to argue that the prior probability of the dinosaur hypothesis was low (which is doubtful), the evidence we have for dinosaurs surely overcomes even a really small prior probability. In fact, you would only need to show him a tiny fraction of the evidence we have for dinosaurs to prove this. A dozen well-preserved specimens that had been documented in peer-reviewed journals and subjected to thorough analysis and re-analysis by museum curators and paleontologists ought to do the trick. Such evidence even passes Hume’s criteria that we have “a sufficient number of men, of such unquestioned good-sense, education, and learning, as to secure us against all delusion in themselves; of such undoubted integrity, as to place them beyond all suspicion of any design to deceive others…”
