The Doctrine of Christ (part 17)

July 06, 2008     Time: 00:34:42


Craig continues on the Doctrine of Christ (Christology). Probabilities continued.

[Opening Prayer][1]

We have been talking about the literal interpretation of the resurrection of Jesus and what the principal objection is to a literal understanding of Jesus’ resurrection. I said that I think, as I look at the landscape philosophically and on the scene of biblical studies, the principal objection to belief in the resurrection as a literal historical event is the so-called problem of miracles. Namely, it is always more probable to believe in some naturalistic explanation of the data than to believe in a supernaturalistic explanation. Last time we saw that this is simply a rehearsal of an old argument by David Hume from his essay On Miracles back in 1738 in which he argued that it will always be more probable to think that a miracle has not occurred than to think that a miracle has occurred.

What I shared last time with you was that probability theorists have shown that Hume’s argument is fallacious. It is a failure, in fact. I laid out for you the probability calculus for calculating the probability of an event like the resurrection of Jesus. Letting R stand for the resurrection hypothesis (that God raised Jesus from the dead) and E stand for the specific evidence for that event (such as the empty tomb, the postmortem appearances of Jesus, the transformed lives of the disciples), letting B stand for our basic background information (that is to say, everything else we know about the world) and A stand for alternative theories to the resurrection such as Conspiracy theory, the Hallucination theory, the Apparent Death theory, the Wrong Tomb theory, then we can calculate the probability (which is abbreviated Pr) of the resurrection given the evidence and the background information as equal to this complicated formula here.

Basically what Hume did was he didn’t understand probability theory – it hadn’t been developed in all fairness by then – and all Hume calculated was the top part of the line. He didn’t realize you had to divide by what is under the line there.[2] So he basically said, well, the probability of a resurrection based on our background information is almost next to nothing because it is a miracle, and given what we know of the laws of nature dead men do not rise. So the probability of the resurrection on the background information is very, very small. Therefore, he said even if we let the probability of the evidence given the resurrection and the background evidence be very high, still the improbability of the resurrection will outbalance any reliability of the witness. He interpreted this to be the reliability of the witness. How probable is it given the resurrection and background information that the witness would testify to it? That the witness would give that evidence given the truth of the resurrection? He said even if you allow that to be very high, this probability is so low that they basically balance each other out and so nobody can give any credence at all to a miraculous event like the resurrection.

What Hume failed to realize is that you not only need to calculate that, but you also need to calculate these values – what is the probability of the alternative theories given the background information and what is the probability that if the resurrection had not occurred (if one of these alternative theories were true) that we would have the evidence that we do. The testimony to the empty tomb, the postmortem appearances, and the transformed lives of the disciples. When you compute in these factors as well, if these are very improbable, that can balance out any improbability that you might see in the resurrection itself. So in fact it might be that the resurrection is very probable given the historical evidence and the background information. Or, at least it is more probable than any of its alternatives.

We also saw that Hume neglected to take into account that multiple witnesses augment the probability exponentially. For example, if you have a witness who is correct 99% of the time (that is to say, he only makes a mistake 1 time out of 100) and you have another witness who is accurate 99% of the time (he only makes a mistake 1 time out of 100), the odds that these two witnesses would both be mistaken is not 1 chance out of 200, rather it is 1 out of 100 times 1 out of 100 which is 1 out of 10,000. So if you have two very reliable witnesses, the odds of their both being wrong is only 1 chance out of 10,000. If you have three very reliable witnesses then the chance of their both being wrong are 1 out of 1,000,000. If you have six then it gets into the billions. In the case of the resurrection of course one would want to argue that we do have independent witnesses to this event which augments the probability of their testimony. In fact, because of this cumulative power of multiple witnesses, it could be the case that even if every witness considered individually is wrong more often than not (that is to say, he has a lower than 50% chance of probability of being right), nevertheless the cumulative force would make it overwhelming probable that in this case what they reported was indeed correct.

That is the first problem that we reviewed with Hume’s argument. He didn’t understand probability theory. In fact, we saw that if you apply Hume’s argument even to many natural events, like the pick in last night’s lottery, that it would force us to disbelieve in many natural events because they would be excessively improbable; not just supernatural events. So Hume’s argument, if taken seriously, would be a positive impediment to science. It would halt the progress of science because we could never learn anything that contradicted what we already take to be the laws of nature or our background information.[3]

But then the second basic point that I wanted to make was that I disagree with Hume’s probability assignments. For example, Hume regards this probability as extremely low – the probability of the resurrection given the background information. What I wanted to say there was I don’t see any reason to think that the hypothesis “God raised Jesus from the dead” is improbable on our background information. Think about it. What our background information makes improbable is the hypothesis that Jesus rose naturally from the dead. That is enormously improbable – that all of the cells in Jesus’ body would just naturally and spontaneously come back to life again. That contradicts everything we know about medical science and cell necrosis (the way cells die). The naturalistic theory that Jesus just came back alive naturally is enormously improbable. But I don’t see anything improbable about the hypothesis that God raised Jesus from the dead. I pointed out that you cannot say that a miracle is improbable simply because it is rare. Rarity, or infrequency, doesn’t translate to improbability because it could be precisely because of its rarity or its infrequency that it would be highly probable that God would choose to do such a thing. To give you an ordinary illustration. Suppose I go into a car lot to buy an automobile – a new car – and 99 out of 100 cars in the lot are black, and there is 1 bright red car in the lot. If you just take this as a random probability you would say the chances are I would drive out with a black car. But suppose I don’t like a black car. In fact, I choose the red car precisely because it is rare. It is precisely because it is unusual that I want to get the red car. In that case the rarity of the color doesn’t say anything at all about the probability that I will drive out with a black car. On the contrary, it is the very rarity of the red car that increases the probability that if I had my pick I am going to drive out with the red one. In general, I think that illustration shows that when you are dealing with a free agent rather than just with a random process, really all bets are off. A free agent can do as he chooses as he pleases. We are not dealing with just a randomizing process here like picking balls blindly out of a lottery or one card blindly out of a deck. One of my Christian philosopher friends put it very nicely when he said, “When there is a card shark at the table then all bets are off.” Think about that. I think that really encapsulizes it nicely. With ordinary shuffling and dealing it is just chance – random distribution. You could say, “What are the chances that my opponent would be holding four aces?” It would be very slim. But then when you learn there is a card shark sitting opposite you, all bets are off. You are not dealing with a randomizing process anymore. For that reason I think Hume is just completely unjustified in assigning a low probability to the resurrection given the background information.

That is why I think this Humean argument doesn’t go through, and that it is therefore demonstrably false. What you will often hear said by unbelievers about things like the resurrection, “Extraordinary events require extraordinary evidence.” That is just false. That is not true that extraordinary events require extraordinary evidence. Once you understand the probability calculus you can see that that is not true.


Student: [inaudible]

Dr. Craig: That is a little bit different question. I would want to ask somebody who said that, “Why can’t you as a historian include miracles?” What would you say then?

Student: [inaudible]

Dr. Craig: Alright. I think it is important to see that even on this project here, one isn’t just arbitrarily appealing to miracles. You are not just punting to God to say, “Oh God did it.” That would be unjustified every time you confront a mystery to say God did it. No, you are calculating the probabilities of this hypothesis against its alternatives.[4] You are saying given the specific evidence and the background information this hypothesis is more probable than these. All you are asking the historian to do is to be open-minded and to consider your hypothesis along with the others. I don’t think that that is a science-stopper or a history stopper at all. What the skeptic would have to show is that there is some alternative theory that would better explain the evidence. If that is the case then obviously we won’t go with a supernatural explanation. If there is a good naturalistic account we will stick with it.

I think that what the secular historian is really doing – you were very nice and open-minded – is saying we are not going to allow any supernatural hypotheses. We are not even going to consider them. Maybe you will remember our Q&A day on intelligent design where we encountered this same attitude with respect to theories about intelligent design. Remember the intelligent design opponent or skeptic doesn’t want to consider the hypothesis that God designed biological complexity. He doesn’t even want to admit that hypothesis as an alternative. Basically what we are saying both in the case of intelligent design (you can let “I” be your hypothesis here – intelligent design – and then the evidence would be all the biological complexity) and you would rework the same formula to try to show that intelligent design is a better hypothesis than all of the various alternatives that are out there. This applies to any hypothesis and any specific evidence that you want to use. All we need to do is just to say let’s be open-minded and allow supernatural hypotheses to be considered along with natural ones and then follow the evidence where it leads. Unless they have some kind of in principle objection for excluding supernatural explanations I don’t see why we can’t include them in the mix.

Student: [inaudible]

Dr. Craig: There is a good book that applies this to the resurrection. The first part of the book is basically all about confirmation theory, which is what this is about. It is by Richard Swinburne, who was a professor of philosophy at Oxford University. I can’t remember the exact title of the book but it has the word “Resurrection” in it. If you look up Richard Swinburne, it is published by Oxford University Press, it would be a title that has the word Resurrection in it and would be very recent like 2003 or 2002, something like that. That would be a really good place to look because Swinburne will give you all the confirmation theory first and then he will apply it to a specific case, namely the resurrection.


Let me move along at this point. Because you will recall I said there was one other objection arising from probability theory against ever believing in an event like the resurrection on the basis of historical evidence. Surprisingly enough, this objection comes from a Christian philosopher, indeed probably the greatest living Christian philosopher, Alvin Plantinga. Plantinga is, as I say, a great Christian philosopher, but he is skeptical about the value of historical apologetics such as Swinburne uses. What Plantinga poses is something he calls the problem of dwindling probability. What does he mean by that? What he wants to say is that if you consider a hypothesis like the resurrection given the evidence, this isn’t just a single hypothesis. Rather, you also are supposing that God exists, and you are supposing that God would want to reveal himself in the world, and that Jesus of Nazareth existed and made certain radical claims about himself, and that God raised him from the dead to vindicate those radical claims. This is not just a sort of simple hypothesis. This is really a series of hypotheses.[5]

He says suppose we calculate the probability given the evidence and background information. First of all, you need to have the probability that God exists. Let’s let that be G. After that there would be not only that God exists but that God would want to, say, reveal himself in the world. We’ll let that be R. Then, assuming that he does want to do that, the probability, say, that Jesus lived and made certain radical claims about himself. And then finally that Jesus rose from the dead. We’ll just let that be S. Plantinga says even if each one of these has a very high probability – say, 90% – nevertheless as you multiply these probabilities together they dwindle. They get smaller and smaller. 0.9 times 0.9 is 0.81. 0.81 times 0.9 would be 0.729. Then 0.9 times 0.9 times 0.9 gets smaller and smaller until finally it may turn out that you actually get to less than 50% - less than 0.5. Your probability turns out to be very small. So you face this problem of dwindling probability that, as a result, the event turns out to be not very probable at all. So Plantinga doesn’t think that we should base our belief in something like Jesus’ resurrection on historical evidence because of this problem of dwindling probability. To recap again – the idea basically here is that the more you refine your hypothesis the less probable it gets. The probabilities just dwindle until they can become very negligible.

Plantinga, as I say, is a great philosopher. But he made a fundamental mistake in his calculation of the probabilities here that has been pointed out by Timothy McGrew who is a professor of philosophy at Western Michigan University. Before I get to that, let me say something about the argument in general. There has got to be something the matter with this argument. I mean, even if you don’t know what it is, you know just hearing it that something has got to be wrong with this argument. Because this argument wouldn’t just destroy belief in, say, the resurrection of Jesus; it would make it impossible to believe almost anything in history that was based upon a series of events. Or something in science, for example, that was based upon a series of probabilities. Let me take an example from science. In the constellation Cygnus, there is an object that astronomers call X-1 – Cygnus X-1. Most astronomers think that it is very probable that Cygnus X-1 is a black hole based upon the scientific evidence. This object is probably a black hole existing in our own galaxy. Think about that in terms of Plantinga’s probabilities. That is based upon the probability first of all of the Copernican Principle. What is the Copernican Principle? That is the principle that we occupy no special place in the universe. That the laws of nature that we experience are the same laws of nature that obtain and hold in the constellation Cygnus. If those laws of nature were totally different than our laws of nature we couldn’t make any kind of conclusion. So first of all we have to presuppose the Copernican Principle. Secondly, we have to have the probability of the General Theory of Relativity, because it is on the basis of this gravitational theory that black holes are predicted. Thirdly, there is the probability that the x-ray eclipse that we observe in Cygnus X-1 is due to a companion object that is rotating around this object X-1 – a kind of other star or other stellar object in this binary system that helps us to identify that.[6] Fourthly, we have to say on the basis of the orbits of this companion object that X-1 has a mass of around three to four times the sun. Number five, we have to calculate the size of this object X-1 is about nine miles across based upon the flickering x-rays that are emitted from X-1. Finally, we have to consider the probability that there is no other object that could cause these sort of phenomena – for example, a neutron star which is a stellar object which is highly compacted but is not a black hole. Even if every one of these had a 90% probability you could see that the probabilities would soon dwindle in exactly the same way so that no scientist could ever conclude that Cygnus X-1 is probably a black hole. And yet most scientists do think it is probably a black hole. There has got to be something wrong with Plantinga’s probability theory here or it would undermine both science and history.

And, in fact, as I say Timothy McGrew, who is a Christian philosopher at Western Michigan, has pointed out what the fallacy is. The fallacy is that Plantinga mistakenly holds the evidence constant while he refines the hypothesis. But that is incorrect. When someone says that the probability of the existence of God is 90% given the evidence, he is talking just about the evidence that is relevant to God’s existence. Remember our arguments for the origin of the universe, or the fine-tuning of the universe for intelligent life, or objective moral values in the world, or the ontological argument. None of those arguments appeal to evidence about Jesus or the Bible. They appeal to certain general facts of experience like moral facts or facts about the origin of the universe and so forth. When we say that there is a probability that God exists and that he would want to reveal himself then we augment that evidence with a little bit more evidence. Similarly, when we say that what is the probability that God exists, and he would want to reveal himself, and Jesus existed, then we add yet more evidence. You see, as long as the evidence keeps increasing it doesn’t matter that the hypothesis is being further and further refined. Plantinga’s mistake was thinking that the evidence is held constant. That is not right. As the evidence adds up, as the evidence mounts, this probability can actually be greater than that probability. The probability that God has revealed himself in Jesus and raised him from the dead could actually be greater than just the probability that God exists, because you added in all this extra evidence that would make it more probable. Therefore, Plantinga’s objection based upon dwindling probabilities turns out to be fallacious.


Student: [inaudible]

Dr. Craig: Not in print, but I was with Plantinga at a Veritas Forum as UCSB in California and we were sitting around talking and he said, “Say, Bill, have you heard of this paper by Tim McGrew on this problem?” My ears perked up. I said, “Yeah, I have in fact. I heard him deliver it at a conference.” He kind of said, “Yeah, it looked pretty good to me.” So he was very impressed with McGrew’s point.[7] I have also heard Swinburne, whom I referred to a moment ago, also say of McGrew’s point that is exactly what I was trying to say but McGrew said it more effectively than I did. So I think that this is correct. Once you understand it, it is obvious. Take my example of Cygnus X-1. You are adding more and more evidence there – as you say, now we are going to look at the General Theory of Relativity, but the General Theory of Relativity is based on one set of evidence. Then the evidence that there is a black hole in Cygnus is based upon additional astronomical evidence concerning the x-ray flux. So you are heaping up the evidence as your hypothesis is refined. I think it is clear that Plantinga just made a mistake here.

This is also a lesson in itself. It is humbling when you realize even the greatest of minds make mistakes. That is a good reminder for us not to be dogmatic as well.

Student: [inaudible]

Dr. Craig: This is actually a very perceptive comment. It may sound weird what he just said, but it is true that sometimes the evidence can be such that you ought to believe something that is very improbable. For example, the probability that you exist is enormously improbable given the sperm and the egg that had to unite to produce just you, or that you should be here. It is enormously improbable. But given . . .

Student: [inaudible]

Dr. Craig: [laughter] OK! But given what you know, namely “I am here and I am alive” you have good grounds for believing in something that is highly improbable.

Student: [inaudible]

Dr. Craig: No, I don’t think I agreed to that. If you look at the New Testament, there is one verse where Jesus says something like “No man takes my life from me. I lay it down of my own accord, and I have the power to take it again.” But apart from that verse, the testimony of the New Testament over and over again is to say God raised Jesus from the dead. That is the proclamation of the early Christians – God raised Christ, or that Christ was raised. I wouldn’t make a big deal out of that.

Student: [inaudible]

Dr. Craig: I am not sure how to respond to that. In this case, we are looking backwards.

Student: [inaudible]

Dr. Craig: It seems to me that that is fallacious to say that because I am here that proves evolution has occurred, because there is a good alternative in that case. You have to say something about . . . remember again, it is the same principle we replied to Hume. What would be the probability of my being here if evolution had not occurred? They might think it is zero. But someone who says, “I believe in divine creation” would say it is not at all improbable that I be here if evolution had not occurred. They’ve got to consider what is the probability of the falsity of their hypothesis given the evidence.


In fact, having said all of this, I want to say that actually use of this probability calculus and probability theory is really not the best way to do an apologetic for the resurrection of Jesus. This is very popular among apologists today, but I don’t think it is really the best way to do it. I don’t think that’s the way historians work. You won’t find historians using this theorem or this probability calculus. Why? Because historians assess their hypotheses in a different way. That is what I will talk about next week – about a different approach to this than the probability calculus.[8]

[1] 2:31

[2] 5:07

[3] 10:04

[4] 15:04

[5] 20:10

[6] 25:21

[7] 30:00

[8] Total Running Time: 34:42 (Copyright © 2008 William Lane Craig)