Doctrine of Creation (Part 18)December 28, 2012 Time: 00:34:17
We have been talking about the doctrine of providence and in particular God’s acts of extraordinary providence which are called miracles. Last time I dealt with the objections of the 17th century philosopher Benedict de Spinoza to the possibility and identification of miracles.
Question: [The person has a hard time expressing his question here. The question seems to be in regards to a supernatural event that could be attributed to some unknown natural phenomena.]
Answer: I’m not sure I understand the question. If I understand Spinoza right, what he argued was that when we see a purported miracle – or hear of one – that we have no way of knowing whether this was indeed a genuine miracle or something that was simply due to an unknown law of nature. I think that is the reason why most of us would be skeptical about certain miracle reports. We think, well, maybe there is some unknown natural cause. What I tried to do is give some criteria for the detection of a miracle that would enable us to say that this is, in fact, a genuine miracle. In particular I tried to apply that to the resurrection of Jesus and argued that if the resurrection actually took place then in all probability this was a miraculous event. That is, an event that is naturally impossible and therefore wrought by God. I don’t understand how your question interacts with that.
Followup: If we think about the supernatural event actually occurring – for example you saying Jesus being observable by his disciples in a physical body – wouldn’t you say that it could still be some sort of phenomena that God just hasn’t revealed?
Answer: That’s what I was arguing against. I was arguing against that view. I argued in general that when the miracles in question have occurred in a momentous time, when they are numerous and various and do not recur regularly throughout history, then the chances of them being the result of an unknown natural cause are minimal. With regard to Jesus’ resurrection in particular I argued that everything that we know about medical science and biology says that it is naturally impossible for a body which is truly dead to come back to life spontaneously. There is nothing in the causal capacities of nature that could do that. Secondly, the miraculous interpretation is given in the religio-historical context in which the resurrection occurs. This isn’t a bald anomaly without a context. It comes as the climax to Jesus’ own unparalleled life and teachings. It is that religio-historical context, I think, that tips us off to saying that this is indeed a miracle. So those two factors, I think, combine to make it in all plausibility the case that if Jesus rose from the dead this was an act of God.
Question: This sounds suspiciously similar to the design argument. You have complexity conforming to an independently specified pattern. It sounds like your contra-Spinoza has a lot of that in it. You have these unusual events but conforming to an independently specified religio-historical context.
Answer: That is interesting. I have never thought of it that way. That wasn’t at all deliberate. I actually think it is more parallel to cosmological arguments for God’s existence. As someone remarked to me once, the cosmological argument for God based on the creation of the universe is sort of an argument from miracles writ large. It is the supreme miracle. So here I don’t think that the religio-historical context is meant to rule out the chance hypothesis the way the independently given pattern is designed to rule out chance. Rather, here it is intended to provide some sort of interpretive framework in which it makes better sense to say this is a miracle than this is a product of physical necessity in this case.
Now we want to turn to the very influential objections of the 18th century Scottish skeptic David Hume. Spinoza argued against the possibility of miracles; Hume, by contrast, argues against the possibility of the identification of a miracle. In his essay Of Miracles, he presents a two-pronged attack upon the possibility of identifying any event as a miracle. This two-pronged attack takes the form of an “even if . . . but in fact” argument. That is to say, in the first part of the argument, he argues under certain conditions that he concedes for the sake of argument: “even if such and such is the case.” And in the second part of the argument, he argues on the basis of what he thinks is, in fact, true. We can call these two parts of his argument his in principle argument and his in fact argument. He will first argue that even if you concede certain points, it is in principle impossible to prove that a miracle has taken place. But, in fact, the evidence is not very good and therefore we should not believe in miracles.
Let’s start by examining his in principle argument against the identification of a miracle. Hume begins by noticing that a wise man proportions his belief to the evidence. If the evidence makes a conclusion virtually certain then we may speak of a full proof in such a case and the wise man will give wholehearted assent to that conclusion. On the other hand, if the evidence simply makes a conclusion more likely than not then we will speak of a probability rather than a proof and the wise man will proportion his belief to the degree of probability of the conclusion. If it is highly probable then he will give a strong assent to that conclusion. If it is just slightly more probable than not then he will give a kind of tentative and light assent to that conclusion. Now, Hume argues, even if we concede that the evidence for a particular miracle amounts to a full proof for a miracle, he says it is still in principle impossible to identify that event as a miracle. Why? Because standing against that testimony is an equally full proof for the unchangeable laws of nature which are incompatible with that event being a miracle. So Hume seems to imagine, as it were, a scale in which the evidence is being weighed. On one side of the scale is the evidence for a particular miracle which (he is willing to grant for the sake of argument) amounts to a full proof. So on one side of the scale is the evidence for some miracle which he says is a full proof. The problem is on the other side of the scale stands the evidence of all the people in all the ages of the world for the regularity of nature’s laws. And that also amounts to a full proof. He says,
A miracle is a violation of the laws of nature, and as a firm and unalterable experience has established these laws, a proof against a miracle, from the very nature of the fact, is as entire as any argument from experience can possibly be imagined.
So proof stands against proof and so the scales are equally balanced and therefore the wise man cannot give any assent to either conclusion and therefore the wise man will not believe in miracles on the basis of the evidence. In fact, Hume says to prove that a miracle has taken place, you would have to show that it would be an even greater miracle for the testimony in support of it to be false. So with respect to the resurrection, Hume says, which would be a greater miracle? That a man should rise from the dead or that the witnesses should be mistaken or lying? Hume has no doubt as to which one of those he thinks is the greater miracle. He says even if all historians agreed that on January 1, 1600, Queen Elizabeth publically died and was buried and her successor installed but that a month later she reappeared, resumed the throne and ruled England for three more years, Hume says he would not have “the least inclination to believe so miraculous event.” He said he would accept the most extraordinary hypothesis for her pretended death and burial rather than admit such a striking violation of the laws of nature. So even if the evidence for a miracle constituted a full proof, the wise man should not believe in miracles because opposed to that evidence is an equally full proof for the laws of nature which would be violated by that miracle. That is Hume’s in principle argument.
What about his in fact argument? Well, in fact, Hume says the evidence for a miracle doesn’t amount to a full proof. In fact, the evidence for miracles is so poor it doesn’t amount even to a probability and therefore the decisive weight lies on the side of the scale containing the evidence for the laws of nature. The evidence for miracles is so negligible it can’t hope to outbalance the full proof for the laws of nature which the miracle would allegedly violate.
Hume gives four reasons as to why he thinks the evidence for miracles is negligible. First, he says, no miracle in history is attested by a sufficient number of educated and honest men who are of such social standing that they would have a great deal to lose by lying. Secondly, he says people crave the miraculous and they will believe the most absurd stories as the abundance of false miracle stories attests. Thirdly, he says miracles occur only among barbarous peoples. This is your good Enlightenment Englishman speaking here – miracles only occur among barbarous peoples. And, fourth, he says miracles in any case occur in all religions and therefore they cancel each other out; all religions have their favorite miracles and since they support contradictory doctrines they all cancel each other out. Therefore the evidence for miracles doesn’t even amount to a probability much less a proof.
So Hume concludes that miracles can never be the foundation for any system of religion. He says, speaking as a nominal Christian, “Our most holy religion is founded on Faith, not on reason.” He says,
. . . the Christian Religion not only was at first attended with miracles, but even at this day cannot be believed by any reasonable person without one. Mere reason is insufficient to convince us of its veracity: And whoever is moved by Faith to assent to it, is conscious of a continued miracle in his own person, which subverts all the principles of his understanding, and gives him a determination to believe what is most contrary to custom and experience.
In other words, Hume is saying it is a miracle that anybody could be so stupid as to believe in Christianity! So, that is Hume’s argument against miracles.
Question: Is his assumption that the laws of nature are never violated?
Answer: Yes, that is his assumption. He says that a firm and unalterable experience has established these laws. Many people have accused Hume of therefore simply begging the question; that is, to assume that no miracles have occurred. So, in assuming that a firm and unalterable uniform experience has established the laws of nature, he is assuming that, yes, no violations have occurred. Now, I think that his argument could be recast so as not to make that question begging assumption. I think you are quite right that his argument, as he presents it, does beg the question. But I think you can recast the argument so as to not make it too easily dismissed. You could just say we have very good evidence for the laws of nature; even if it is not exception-less, still there is a tremendously powerful amount of evidence for the laws of nature. Then you are going to need to deal with the argument. But you are quite right in saying that, as he presents it, it is really just question begging.
Question: Isn’t Hume’s thinking along the line of the atheist who would say, “as long as your explanation for, let’s say, the Big Bang or all the things that would point to a creator doesn’t conclude in a God then I’ll listen.” In other words, you can point to all the evidence that would indicate that God exists but, in an atheist’s mind, there is no God therefore, no matter what evidence you might point to, that conclusion cannot be reached.
Answer: I don’t think that Hume’s argument is presupposing atheism because he is not denying that miracles are possible. In fact, Hume is really writing in a deist frame of mind. Where the deists are quite willing to grant that there is a creator of the universe – there is a God – they don’t think he has acted in history to reveal himself in any special way – there is no special revelation from God, no particular miracles performed by this God. So it is not really atheistic. It is simply saying that we can never have good enough evidence to believe that God, if he is there, has acted in a miraculous way in history.
Followup: I guess I was aligning it to that way of thinking because he is saying if there is a God who set in place the laws of nature then there could be no violations of those laws of nature.
Answer: That is more Spinoza. That is Spinoza who thinks that God has established these laws of nature and therefore there could be no violations of them. Hume’s argument isn’t like that. He is willing to grant that God has acted miraculously in history. All the time he could be doing miracles. But what Hume is saying is that you could never know it. You could never know, on the basis of the evidence, that this event is a miracle wrought by God. It is an argument against the identification of a miracle. It is an argument against discerning what is a miracle. How is this relevant to Christian apologetics? It doesn’t disprove the resurrection of Jesus or the Gospel of miracles but it would undercut Christian evidences where someone would say, “The evidence for the resurrection or the evidence of Jesus’ miraculous life is evidence that the God of nature has revealed himself in history.” So this is an attempt to undercut the project of Christian evidences.
[Q&A: Just asks to repeat Hume’s four reasons as to why he thinks the evidence for miracles is negligible.]
Question: It seems to me there are two types of miracles. One involves timing that doesn’t violate the laws of nature and the other violates the laws of nature.
Answer: You will remember, I stoutly rejected that definition of miracles in favor of saying that a miracle is a naturally impossible event. I think you are right in saying you could have events that are truly miraculous in that they are naturally inexplicable and then you have these, what I prefer to call, special providences. For example, that the landslide occurs blocking the Jordan River and drying it up just as the tribes of Israel are to cross into the Promised Land. I would say that, technically, that is not a miracle. A miracle technically is a naturally impossible event. But these timing events I would call special providences. They are naturally explicable but by, as you say, their coincidental timing it is evident that this is a special providence that God has brought about. I would say he can do that through his middle knowledge. If you remember, middle knowledge enables him to set up the natural conditions so that just at the time the Israelites are ready to cross in the Promised Land, the erosion and so forth produce naturally the landslide that blocks the river.
Followup: So curing leprosy and all of the things that have happened, it is just curious that it happened right then.
Answer: Well, sometimes of course miracles may occur at propitious times as well. Being at a coincidental time isn’t a sufficient condition for saying this is merely a special providence. Miracles can occur at special times, too. It would be simply the difference between the two would be whether or not it is naturally explicable. We are focusing on events that are properly and strictly miraculous. That is to say, they don’t have natural causes that account for them.
Let me then give some response to Hume’s argument. First let’s talk about his in principle argument. Despite the influence of Hume’s in principle argument, particularly on biblical scholars like Bart Ehrman for example, Hume’s argument is generally recognized by philosophers today to be, in the words of the philosopher of science John Earman, an “abject failure.”  John Earman is an agnostic philosopher of science at the University of Pittsburg. He is an extremely imminent philosopher of science. He wrote the book Hume’s Abject Failure on Hume’s argument against miracles. By an abject failure, what Earman meant was this isn’t just a mistake. We all make mistakes as philosophers. Rather, this is an irredeemable mistake. The argument is an irredeemable failure; a failure that cannot be corrected. It is an abject failure. So even Hume’s admirers today, for example Peter Millican whom I debated at the University of Birmingham and the former editor of the journal Hume Studies, recognize the argument doesn’t work. They will, at most, try to salvage some nugget from Hume’s convoluted discussion, typically Hume’s so-called maxim: “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 endeavors to establish.” But even that maxim, it turns out, is trivial if true or, if it’s significant, it turns out to be false. So Hume’s argument, among philosophers, despite its influence is generally recognized to have failed. Why is that? Well, let’s look at it more closely.
His in principle argument actually falls into two more or less independent claims. On the one hand, there is this claim that miracles are by definition inherently and utterly improbable. By definition, a miracle is just an utter improbability. That is the one claim. On the other hand, there is this claim that no amount of evidence for a miracle can overcome that intrinsic improbability. A miracle is utterly improbable in and of itself – that’s the first claim – and the second claim is no amount of evidence can ever suffice to overcome that intrinsic improbability. As it turns out, both of these claims are mistaken.
Let’s talk first about the second claim that no amount of evidence could ever establish a miracle. The so-called probability calculus or modern probability theory didn’t exist in David Hume’s day. So he was ignorant of it. But stimulated by his discussion of miracles, probability theorists from Condorcet to John Stuart Mill wrestled with the question of what sort of evidence would it take to establish a highly improbable event. What probability theorists soon realized was that if you simply weigh the probability of the event against the reliability of the witnesses then you are going to be led into denying the occurrence of purely natural events which, though they may be very improbable, we reasonably know to have actually happened. For example, suppose on the morning news you hear a report that the pick in last night’s lottery was 7492871. This is a report of an extraordinarily improbable event – one out of several million. Even if the morning news’ accuracy is known to be 99.99%, still the improbability of that event will simply swamp the reliability of the witnesses’ credibility so you should never believe such reports on television. In order to believe the report, Hume would say you have to have enough evidence in favor of the morning news’ reliability to counterbalance the improbability of the winning pick. And that is just absurd. You would never have that sort of evidence in favor of the morning news’ reliability. Therefore, we would never be able to believe such reports which we rationally believe all the time.
So probability theorists came to understand you can’t simply weigh the improbability of the event against the reliability of the witnesses. Rather, they saw that you also need to consider the probability that if the reported event had not occurred that the witnesses’ testimony would still be just as it is. What is the probability that, if the event had not occurred, you would have the evidence that you do, in fact, have? This is what John Stuart Mill said:
. . . to know whether a coincidence does or does not require more evidence to render it credible than an ordinary event, we must refer, in every instance, to first principles, and estimate afresh what is the probability that the given testimony would have been delivered in that instance, supposing the fact which it asserts not to be true.
You have to consider: what is the probability that you would have the evidence you do if the event had not taken place? So, to go back to the example of the winning pick in last night’s lottery: the probability that the morning news would announce the pick as 7492871 if some other number had been chosen instead is incredibly small. Given that the newscasters had no preference for that number, the probability that they would pick that number and announce it is just incredibly tiny. On the other hand, the announcement of that pick is much more probable if 7492871 were the actual number chosen. So the announcement of the pick is vastly, vastly more probable given that that was the number that was picked than if it were not the number that was picked. This comparative likelihood easily counterbalances the high prior improbability of the event reported. So again, what you have got to consider is: what is the probability that the evidence would be just as it is if the event had not occurred? If that is a very low probability then that can outbalance any intrinsic improbability in the event itself.
The realization that other factors had to be considered in estimating the probability of highly improbable events came to be codified in a theorem called Bayes Theorem which is the modern probability calculus. Let’s let R represent some miraculous event, say, the resurrection of Jesus. And we will let E equal the specific evidence for that event, such as the empty tomb, the postmortem appearances, the origin of the Christian faith and so forth. Then we will let B represent our general background information. This is our knowledge of the world at large without the specific evidence – you just subtract the specific evidence out of that and that gives you your general background knowledge. So R will be the resurrection, E will be the specific evidence for that event and B is your general background information without the specific evidence being included in it. What Bayes Theorem states is that we can compare the probability of R given the evidence and background information [Pr(R|E&B)] with the probability of not-R given the evidence and background information [Pr(not-R|E&B)]:
This is the so-called “Odds Form” of Bayes Theorem, where you compare the odds of the event given the evidence and background information with the denial of the event given the background information and the evidence. This is called the total probability of the event. It is total because it considers not only the background information but also the specific evidence. We want to compare the total probability of R with not-R. This will be computed as the product of two other factors that go to make up the total probability.
The first will be the probability of the miracle on the background information alone [Pr(R|B)] compared to the miracle not occurring given the background information alone [Pr(not-R|B)]. So you look at our general knowledge of the world and you ask how probable is the resurrection of Jesus on that background information compared to how probable is it that he did not rise given the background information?
This is called the intrinsic probability of the hypothesis. It is the probability of the hypothesis independent of any specific evidence for it.
So the total probability will be made up, or computed by, the intrinsic probability of the hypothesis. Then it is multiplied by another ratio and that will be the probability of the evidence given the resurrection and the background information [Pr(E|R&B)] compared to the probability of the evidence given that there is no resurrection – that it did not occur – and the background information [Pr(E|not-R&B)]:
And you can see this is the factor that the probability theorist said we need to consider. What is the probability that we would have the evidence we do if the event had not occurred? This is the explanatory power of the hypothesis. It tells us how well the hypothesis explains the evidence. Is the evidence more probable given the hypothesis or is the evidence more probable given the negation of the hypothesis? How well does the hypothesis explain the evidence?
So the total probability, in this case, of Jesus’ resurrection will be computed by comparing the intrinsic probability times the explanatory power of R and not-R:
Now Hume’s mistake, being unaware of the probability calculus, is that the only factor he considers is the intrinsic probability. He says because a miracle is enormously, utterly improbable given our background information that no amount of evidence can ever go to establish a miracle as probable. That is simply mathematically demonstrably fallacious. It is wrong. Imagine, say, the odds here [Dr. Craig is referring to the intrinsic probability factor] are something like 1-to-100 in favor of not-R. But suppose the odds here [Dr. Craig is now referring to the explanatory power factor] are 100-to-1 in favor of R. Then they just balance each other out and the odds are even. So Hume’s argument, by neglecting the probability of the evidence on the hypothesis or its negation, is simply fallacious. Hume never discusses this other ratio. He simply concludes that because the intrinsic probability of a miracle is so low therefore the total probability of the miracle is low. That is simply mathematically demonstrably fallacious.
There is a slogan which is beloved in the free thought culture: “extraordinary events require extraordinary evidence.” That sounds so common sensical, doesn’t it? Yet, what Bayes Theorem reveals to us is that is demonstrably mistaken. It is simply not true that in order to establish some highly, highly improbable event you need to have extraordinary evidence in any sort of acceptable sense. Think again of the illustration of the pick in last night’s lottery. So even if the event is intrinsically, highly improbable, that can be easily outbalanced by the hypothesis having greater explanatory power. What Bayes Theorem shows us is that believing in a highly improbable event on the basis of the evidence doesn’t always require an enormous amount of evidence. What is critical is that the evidence should be more probable, given the hypothesis, than it is if the hypothesis is false. So the bottom line is that establishing a miracle doesn’t always take a huge amount of evidence.
Well, so much for the second claim of Hume’s in principle argument that no amount of evidence can go to establish a miracle. As John Earman demonstrates, that is demonstrably fallacious. What we will do next time is look at the first part of Hume’s claim where he says that the intrinsic probability of a miracle is very, very tiny. Hume just assumes that the probability of a miracle on the background information is almost infinitesimal. But is that true? Is it really true that the probability of R on B [Pr(R|B)] is very small? That will be the question that we will take up next time.
 This essay is part of a larger work by David Hume titled An Enquiry Concerning Human Understanding which you can find in the public domain at http://www.gutenberg.org/ebooks/9662 - you can find the essay “Of Miracles” at
 David Hume, An Enquiry Concerning Human Understanding, Section X, “Of Miracles,” Part I.
 “When anyone tells me, that he saw a dead man restored to life, I immediately consider with myself, whether it be more probable, that this person should either deceive or be deceived, or that the fact, which he relates, should really have happened. I weigh the one miracle against the other; and according to the superiority, which I discover, I pronounce my decision, and always reject the greater miracle. If the falsehood of his testimony would be more miraculous, than the event which he relates; then, and not till then, can he pretend to command my belief or opinion.” (Ibid.)
 Ibid., Part II.
 These four reasons are found at the beginning of Hume’s “Of Miracles,” Part II.
 John Earman, Hume’s Abject Failure: The Argument against Miracles (Oxford: Oxford University Press, 2000).
 Hume, “Of Miracles,” Part I.
 John Stuart Mill, A System of Logic, 2 vols. (London: 1843), bk 3, chap. 25, §6, cited in S. L. Zabell, “The Probabilistic Analysis of Testimony,” Journal of Statistical Planning and Inference 20 (1988): 331.
 Total Running Time: 40:19 (Copyright © 2012 William Lane Craig)