Doctrine of Christ (part 23)

April 15, 2012     Time: 00:33:39

[Due to technical difficulties with the audio recording equipment, the audio for the Defenders 2 “Doctrine of Christ Lecture 23” session is not available. The audio here is borrowed from the Defenders 1 class, specifically its “Doctrine of Christ Lecture 18” session.

Fortunately, we do have a transcript of the actual Defenders 2 “Doctrine of Christ Lecture 23.” It follows immediately below. After that transcript, we include the transcript of the audio from the Defenders 1 series which you may listen to here.]

In our lesson this morning we want to continue our discussion of the case for the resurrection of Jesus as a literal historical event. You will remember I said there are two steps in constructing a case for the historicity of the resurrection. First is to establish what are the facts to be explained and then, second, is to determine what is the best explanation of those facts. We saw that there were a couple of objections to the Resurrection Hypothesis as the best explanation of the facts that would cut off the Resurrection Hypothesis at its knees and, therefore, these need to be dispatched immediately.

The first of these is Hume’s objection against the identification of a miracle. Remember Hume argued that a miracle, by definition, is an event so improbable that no amount of historical evidence could ever serve to establish it. Therefore, a miracle is inherently unprovable. We saw that this sort of reasoning has been repeated in our own day by a New Testament scholar like Bart Ehrman, who also says that it is a contradiction to say the resurrection is the most probable thing that happened because that would be to say that the most improbable is the most probable thing, which is impossible.

The other objection to the Resurrection Hypothesis would be Plantinga’s Problem of Diminishing Probabilities, where in order to establish the resurrection as the best explanation you have this sort of cascade of probabilities that get lower and lower and lower until you may well come to the point that, when you get to the resurrection, it turns out to be less than 50% probable and, therefore, if you assess it on a historical basis, you would not come to believe that it is the true explanation.

Hume’s Objection against Identification of a Miracle

Let’s talk about each of these problems now in turn. First, Hume’s argument against miracles. Probability theorists from the time of Condorcet1 to John Stuart Mill in the 19th century discussed very thoroughly what sort of evidence it would take to establish a highly improbable event. What they found was that Hume really didn’t understand probability theory, which hadn’t been developed, actually, at his time. He wrote his essay in about 17382, so he could hardly be blamed for not understanding the so-called probability calculus. But contemporary scholars, like Bart Ehrman, have no such excuse. The probability calculus is now well known, and they should not be taken in by Hume’s fallacious argument. What theorists came to realize is that holding that extraordinary events would require extraordinary evidence would rule out many natural events. For example, take the report on the evening news of the winning pick in the lottery. When you think about it, that is a report of a highly, highly improbable event. It is just extraordinarily improbable that that number, of all the numbers that could have been picked, would be the one that was chosen. So this is an enormously improbable event.

But do we need to have extraordinary evidence in order to believe the evening news when it says that here was the winning number of the lottery? Obviously not! It is simply not true that extraordinary events require extraordinary evidence. Otherwise you would be led to scepticism about many perfectly natural events. What probability theorists saw is that you also need to consider the probability that the evidence would be just as it is if the event had not taken place. What is the probability that the evidence would be just as it is if the event had not taken place? Well, what would be the probability of the evening news’ reporting just that number, if in fact that number had not been chosen? That would be very, very low. This could off-balance the improbability of the pick of the number itself.

Probability Calculus

So in the case of the resurrection of Jesus, for example, what Hume needs to assess is not simply the intrinsic probability of the resurrection relative to our background information and the laws of nature; he would also have to consider what would be the probability that we should have the empty tomb, the postmortem appearances, and the transformation of the disciples if the resurrection had not occurred. If there were no resurrection of Jesus, what is the probability that the evidence would be exactly as it is? That seems extremely remote. So that could balance out any increase in improbability in the resurrection itself. If Jesus had not risen, then the evidence might be a totally different range of evidence than what in fact we have.

Let’s put on the board the complete probability calculus and see what factors Hume neglected:

Pr(R|E&B)   Pr(R|B)   Pr(E|R&B)
----------------------  = ----------------  x -----------------------
Pr(not-R|E&B)   Pr(not-R|B)   Pr(E|not-R&B)

What we want to ask ourselves about is the probability of the Resurrection Hypothesis, which we will abbreviate R, the probability of R given our background information B and the specific evidence E [i.e., Pr(R|E&B)]. R is the Resurrection Hypothesis, which we said is “God raised Jesus from the dead.” The background information B is our general knowledge of the way the world is apart from the specific evidence, and E is the evidence that we talked about, namely, the empty tomb, the postmortem appearances, and the transformation in the disciples. So the probability of R given the background information B and the specific evidence E. Let’s compare that, then, to the probability of no-resurrection given B and E [i.e., Pr(not-R|E&B)].

This is called the Odds Form of the probability calculus, or Bayes’ Theorem, because we are constructing the odds of a hypothesis being true rather than not:


So what are the odds that the Resurrection Hypothesis is true as opposed to false? Well, this is going to be the product of two factors. First, will be the intrinsic probability of the resurrection on the background information [i.e., Pr(R|B)] over the probability of no-resurrection on the background information [i.e., Pr(not-R|B)]:


This is sometimes called the intrinsic probability of the hypothesis. Given just the background information – the laws of nature, what we know about the world – apart from our specific evidence, what is the probability of the resurrection, that God raised Jesus from the dead? Or, what is the probability that God did not raise Jesus from the dead relative to the background information of the world? That is the first factor that needs to be considered.

This will then be multiplied by the second factor, and this is going to be the explanatory power of the hypothesis:


This is going to be the probability of the evidence given the resurrection and the background information [i.e., Pr(E|R&B)]. How probable is it that if the resurrection did occur, then we would have the evidence of the empty tomb, the postmortem appearances, and the transformation of the disciples? I think we would all agree, “Well, that probability is going to be pretty high!” Given that the resurrection occurred, these facts are going to be well-explained. This is called often the explanatory power of the hypothesis – how well does the hypothesis explain the evidence? This will then be contrasted with the probability of the evidence given no resurrection and the background information [i.e., Pr(E|not-R&B)]. If the resurrection did not take place, if God did not raise Jesus from the dead, then what is the probability of the evidence? That is going to be probably very, very low.

Now what Hume did, basically, focus just on the question: what is the intrinsic probability of the resurrection on the background information? He said that relative to the laws of nature, a resurrection is enormously improbable. It is improbable that somebody would rise from the dead. Therefore, the probability of the resurrection on the background information and the evidence is very, very low. Well, that just doesn’t follow! He neglects these other factors. Even if you take into account the intrinsic probability of the evidence given the resurrection and the background information, that is still not going to be enough. What you have got to consider especially is the ratio here of these factors:


What is the probability of the evidence if the resurrection had not occurred? Any intrinsic improbability in the resurrection itself can be offset by this ratio of the explanatory power of the hypothesis. So Hume’s argument is simply demonstrably fallacious. It is mathematically fallacious because he doesn’t consider all the factors. He only considers the intrinsic probability of the resurrection on the background information, but he neglects the rest of the probability calculus. Therefore, it is simply a fallacious argument.

Now, in addition to that, the other thing that I would want to question is – why think that the intrinsic probability of the resurrection on the background information is low? Hume just assumes that the resurrection is highly improbable relative to the background information. That doesn’t seem to me to be at all obvious. What is improbable relative to the background information is that Jesus rose naturally from the dead. Yes, that is highly improbable! The idea that all of the cells in Jesus’ body spontaneously came back to life again is enormously improbable. Conspiracy theories, hallucination theories, and apparent death theories would all be more probable than that! So the idea that Jesus just came back to life and rose naturally from the dead is highly improbable relative to the background information. But that is not what R is. R is the hypothesis “God raised Jesus from the dead.” And that isn’t rendered improbable by the fact that Jesus couldn’t rise naturally from the dead. In fact, this probability is going to be equal to a more complicated probability. It is going to be – let’s let G represent God’s existence – it is going to be the product of the probability of God’s existence on the background information times the probability that God would raise Jesus from the dead.

Pr(G|B) x Pr(R|G&B)

What is the probability of God’s existence on the background information times the probability that the resurrection would occur if God exists? Well, now, that all depends on what you include in the background information, doesn’t it? If the background information includes things like the cosmological argument, the ontological argument, the teleological argument, the moral argument, and so forth, then the probability of God’s existence, I think, is very good. So that probability can be very high. What about the probability that if God exists, he would raise Jesus from the dead? Well, again, if we include in the background information everything we know about the historical Jesus up through his crucifixion – things like his radical personal claims, his ministry of miracle working and exorcisms – I don’t see how anybody can say it is improbable that God would raise this man from the dead. For this would be a great way of ratifying those radical personal claims that he had made and that sent him to the cross.

So if Hume is going to say that the Resurrection Hypothesis is very improbable relative to the background information, he has got to show that either that God’s existence is highly improbable or that, if God exists, it is highly improbable he would raise Jesus. If he is going to say the Resurrection Hypothesis is highly improbable relative to the background information, then either God’s existence has to be highly improbable or else it is highly improbable that, if God exists, he would raise Jesus. Hume can’t demonstrate that; there is absolutely no reason to think those are improbable. Quite the contrary, I think those are both probable. So I don’t see that there is any great intrinsic improbability in the resurrection of Jesus.

So it seems to me that Hume’s argument suffers from two faults. To summarize: First, it is demonstrably fallacious because it doesn’t take into account all the factors in the probability calculus. To compute the probability of the Resurrection Hypothesis, given the background information and the evidence, compared to the Non-Resurrection Hypothesis, you have to consider all of these factors, and he simply doesn’t do it. Second, Hume just assumes that the resurrection is highly improbable relative to the background information, and I don’t think there are any good grounds for thinking that.


Question: Can you explain the difference between the background information and evidence?

Answer: When I use “background information,” what I am talking about is what we know about the world apart from this specific evidence that we are including in E. What I am including in E is the empty tomb, the postmortem appearances, and the origin of the disciples’ belief of Jesus’ resurrection. So B would include all of natural theology, all the arguments for God’s existence, and it would include all our knowledge of the historical Jesus up through the crucifixion.

Question: (inaudible)

Answer: Yeah, that would be taken into account – what is the probability of God on the background information?

Question: What’s the typical response you get from sceptics for this argument here from Bayesian probability? I can’t imagine skeptics would say, “Yeah, you are right, I can see your point.”

Answer: I think, probably with regard to my point about the resurrection’s not being intrinsically improbable – they would probably deny that God’s existence is probable. Then you are going to be thrown back to natural theology. This reinforces the point we made earlier that it is helpful to have your natural theology in place before you come to Christian evidences. They will probably want to go back to that. With respect to this [Dr. Craig points to the formula Pr(E|not-R&B) on the board], they may try to propose alternative theories of not-R that would make the evidence not so improbable. So not-R here could include things like the Apparent Death Theory, the Conspiracy Theory, the Hallucination Theory, and the Twin Brother Theory. All of those would be different versions of not-R. So some will try to defend something like “Jesus had a twin brother,” and even though that is improbable, it is not as bad as the resurrection itself. So that would be the kind of moves I think they will make. Then we will simply engage them in argument.

Plantinga’s Problem of Diminishing Probabilities

I think we have time to cover Plantinga’s Problem of Diminishing Probabilities, so let me say something about that. Remember Plantinga said that if you try to argue for the resurrection historically, the probabilities are going to get smaller and smaller and smaller as you get more specific, so that you can’t in the end, perhaps, show that the resurrection is the more probable. Someone in the class remarked, there has got to be something wrong with this argument because we reason like this in science all the time. This would undermine modern science if this were correct.

Let me give a very good illustration of this fact. There is an astronomical object by the name of Cygnus X-1 that many astronomers think is a black hole. In fact, it is widely regarded as the most probable object to be identified as a black hole. Now the evidence that would establish the identity of Cygnus X-1 as a black hole can be analyzed in this same way. First, you have to assume the Copernican Hypothesis. The Copernican Hypothesis is that the same laws of nature that operate here on Earth operate in distant galaxies and in outer space. If there were totally different laws, then we could not make any inference. So you have got to start by assuming some sort of Copernican principle. Secondly, the General Theory of Relativity needs to be true. In order for black holes like this to exist, you would have to assume Einstein’s General Theory of Relativity. Thirdly, you need to think that it is probable that the x-ray eclipse of the companion body of Cygnus X-1 is caused by Cygnus X-1 itself. There is a kind of x-ray flux in and out that goes on in its companion body, and you have to say that there must be an invisible object that is causing this. Fourth, the orbit of this companion object indicates that the object is three to four times the mass of the sun – it is about three to four solar masses. And this is an inference based on its orbit – it is the evidence of the orbit that makes you think that Cygnus X-1 is about three to four times the mass of the sun. Fifth, the flickering x-ray output of the object gives it a size of about nine-miles in diameter. So you must infer that this is about the size of the object based on these flickering x-ray outputs. Sixth, no other object can be this small and yet this massive. For example, a neutron star might be an alternative to a black hole, but a neutron star cannot be so small and yet so massive. So, on the basis of these lines of inference, the inference is that Cygnus X-1 is most probably a black hole.

I think you can see this six step pattern follows exactly the same sort of pattern as the inference to Jesus’ resurrection if you constructed it that way. You might say, “But then when you multiply the probabilities together, they get less and less and less until you conclude Cygnus X-1 is not a black hole,” which would be contrary to what modern science says!

So what is the problem here in Plantinga’s argument? Where does the fallacy lie? Very simply this: Plantinga forgets that at each successive step, you add in new evidence. You are not considering the additional hypotheses relative to the same evidence again. New evidence is added in at each step so that the probability can actually increase. So, for example, the evidence for the General Theory of Relativity isn’t the same evidence as the evidence for the x-ray eclipse that this object produces. That is new evidence. That is additional evidence. And that doesn’t include the evidence concerning its orbit. That is new evidence that is added in and indicates that the mass is three to four times that of the sun. Then you compute the size on the basis of new evidence. So there is new evidence that is being added in at every step. Or another way to think about it is the way I think someone suggested the other day when we talked about this – you can take all of the evidence and fold it into the original body of evidence and then say, “OK, relative to that total body of evidence, what is the probability that Cygnus X-1 is a black hole?” And relative to that total body of evidence, it is very good; it probably is.

Similarly, with respect to the resurrection, you don’t need to unfold the evidence step by step; you can just take all of the evidence, put it into the original body of evidence, and say, “OK, what is the probability of the resurrection relative to that body of evidence?” And it can well be over 50%. So in either case, you don’t encounter the problem of dwindling probabilities. You can either add in new evidence at each step so the probabilities don’t dwindle, or you can fold all of the evidence into the original body of evidence to be explained and simply infer immediately to the best explanation of that evidence. That conquers the problem of diminishing probabilities, which saves not only the Resurrection Hypothesis but also the hypothesis that Cygnus X-1 is a black hole.


Question: Does Alvin know that you disagree with him on that?

Answer: I don’t know if I have actually said it to him, but what I have shared here is nothing original. This comes from Tim McGrew, who is a professor of philosophy at the University of Western Michigan and an expert in probability theory. McGrew wrote, frankly, just a withering critique of Plantinga’s Problem of Dwindling Probabilities, and I simply am sharing it with you in a very simplified version.

[A comment that these things occurred in a relationship rather than as individual pieces. They occurred in relationship with each other and should be analyzed together as such and not as disparate pieces.]

Question: (inaudible)

Answer: For those that could not hear, he wanted to emphasize that what this shows is that we can’t divorce the existence of God from evidences for Christianity because how you assess the Resurrection Hypothesis is going to depend on how probable you think God’s existence is relative to the background information. That also shows again why science is so important to Christians and why we can’t be indifferent about science, since that will be included in the background information when we assess how probable it is that God exists. I agree entirely, and I think, to repeat, this shows the relationship between natural theology and Christian evidences. Christian evidences build upon the foundation established by natural theology, on which we spent literally weeks in this class, laying a solid foundation for belief in God through the various types of philosophical arguments for God existence.

Having dispatched those two objections, next time we will consider what is the best approach to determining the best explanation for the data.3


1 Dr. Craig is referring to Marquis de Condorcet (1743-1794) who was a French philosopher, mathematician, and early political scientist.

2 Its date of publication is 1748.

3 Copyright © 2012 William Lane Craig

Transcript of Audio of Lecture 18 of Defenders 1, Doctrine of Christ

We have been talking in the last several weeks about the resurrection of Jesus as the completion of the work of Christ. The work of Christ consists on the one hand of the cross – his substitutionary atonement – and then also of his resurrection from the dead. We looked at all of the various ways in history that sceptics have tried to account for the resurrection and found them deficient.

So we then finally turned to a discussion of the original interpretation of the resurrection, namely, the literal interpretation. We examined the problems that a literal interpretation of the resurrection of Jesus faces. The main problem was the problem of miracles. Namely, is it rational to believe in events which are beyond the power of nature to produce, like a resurrection from the dead? I argued last time that the old argument that extraordinary events require extraordinary evidence really doesn’t hold water and that this violates the probability calculus that probability theorists have developed, and, in fact, there really is no reason to think that the Resurrection Hypothesis is improbable.

We finally looked last time at the so called Problem of Dwindling Probabilities; that is to say, as your hypothesis becomes more refined, more specific, it becomes less probable on that original body of evidence. We saw that this problem is to be overcome by augmenting the evidence as you make your hypothesis more specific, so that the evidence that is relevant, say, to the broad hypothesis that God exists, will just be evidence, say, about the beginning of the universe or the design in the universe or the existence of moral values. But it won’t include anything about the historical Jesus. But then as you begin to talk not just about God’s existing but about the Christian God’s existing and about Jesus and his radical claims and his Resurrection, you pile up more evidence as the hypothesis becomes specific, and as a result the probabilities do not dwindle. In fact, in the end, the probability of your Christian hypothesis, given all the evidence, may be actually higher than the probability of just, say, God’s existence given general information in the world.

But then I closed by saying that in fact I don’t think this is how we ought to do an apologetic for the resurrection of Jesus. In fact, the probability calculus, interesting as it is, really isn’t very applicable in this area. That is what I want to talk about today.

Historians, when they test for historical hypotheses, don’t really use the probability calculus. You won’t find, I think, any history book that you would pick up that would talk about any secular event in history that would begin to plug in the values to these various elements in the probability calculus to try to figure out what is the best explanation of the evidence for the past. Why not? I think part of the reason is – the main reason is – these probabilities are often inscrutable. That is to say, you just can’t put any kind of numerical value to them. Therefore, you really can’t use the probability calculus to any sort of degree of helpfulness in estimating the probability that some event occurred in the past.

In particular, take the Resurrection Hypothesis. I think here we confront an immediate problem – what is the probability of the Resurrection Hypothesis on the background information alone? You remember that was one of the elements in the probability calculus that you would have to assign a value to. Take your background information, represented by B, which is just our general knowledge apart from any specific evidence concerning Jesus’ resurrection – just our background information of the world. What is the probability relative to that background information that God would raise Jesus of Nazareth from the dead? Well, I don’t think really we have any way of assigning any value to that with any confidence. I think we can’t say it is improbable because God is free to do what he wants; but how can we say that it is probable that he would do that?4 Maybe what would be more probable, given Jewish thinking, would be that God would wait until the end of history and then raise Jesus from the dead along with everybody else, and then everybody would see that he is God’s Son and would go into the kingdom of God and he would be vindicated. How do we know what God would do just given the background information (no evidence concerning the empty tomb or the appearances)? So it seems to me that this probability approach using the calculus isn’t really that helpful in doing an apologetic for the resurrection of Jesus and with that most historians would agree because historians don’t use this.

So how do historians arrive at the best explanation for things in the past? What criteria do they use? I think that the model that can be used here is called inference to the best explanation. By that I am talking about a technical term. Inference to the best explanation is a kind of reasoning in which you have a certain body of data to be explained – certain facts to be explained. These, in the case of the resurrection, would include things like the crucifixion of Jesus, his burial by Joseph of Arimathea, the discovery of the empty tomb, the postmortem appearances to different individuals and groups of people, and the very origin of the Jesus movement itself. All of these would be various data to be explained. Then what the historian or the scientist does – this is really the same model of explanation – is construct a pool of live options for explaining this data. These various live options in the pool of explanations will be his hypotheses. These will constitute hypotheses to try to explain the data. So, with respect to the resurrection, various live options would be things like: the women went to the wrong tomb, or Jesus wasn’t really dead, or the disciples had hallucinatory experiences after Jesus’ crucifixion, or that this is all just mythology. Or one of them would be that God raised Jesus of Nazareth from the dead. That would be one of the live options for explaining the data. Then, in order to determine or infer which is the best explanation, the inquirer (the scientist or historian) will say, “Which one of these options, if true, would provide the best explanation of the data?” So you are not looking for the only explanation – you are not saying there aren’t any other live options (on the contrary there are) – , but you are inferring to the best of these explanations, and you are saying “Which explanation, if true, would best account for the data?”

Now how do you determine what counts as a best explanation? What are the criteria going to be that you use to determine the best explanation? In his book, Justifying Historical Descriptions, the historian Behan McCullagh, lists some of the criteria that historians use in assessing historical hypotheses.5 He is not a Christian. This is a secular historian in a book on historiography – about how you determine what are the best explanations for historical hypotheses. Let me just list what some of these criteria would be for the best explanation.

(1) The first would be explanatory scope. Explanatory scope means that it will imply a greater variety of observable data. It will have a wide range of observable data that it will explain. Let’s apply this to, say, the various resurrection hypotheses. One of the weaknesses of the Hallucination Hypothesis is that it has narrow explanatory scope.6 The Hallucination Hypothesis is offered as an explanation for the postmortem appearances of Jesus, but it doesn’t do anything to explain the empty tomb. So in order to account for the empty tomb, you have got to conjoin some independent hypothesis to the Hallucination Hypothesis to explain all the data. So that would be one of the shortcomings, say, of the Hallucination Hypothesis as an explanation of the data. It has narrow explanatory scope. It doesn’t imply all of the facts – all of the data.

Number (2) would be explanatory power. This means that it makes the observable data more probable than it would have been otherwise. It has the power to explain the data. It makes the data more probable than it would have been otherwise. So you want to pick the explanation which has the greatest degree of explanatory power. Let’s pick on the Hallucination Hypothesis again because I think this one also is wanting in explanatory power. How well does it explain the postmortem appearances of Jesus, wholly apart from the empty tomb? How well does it explain what it purports to explain – those postmortem appearances? Well, not very well. For example, one of the things we saw was that if the disciples were to have hallucinated visions of Jesus, they would have projected visions of Jesus in glory, in Abraham’s bosom, where the righteous dead went to await the resurrection at the end of the world. But in that case, hallucinations of Jesus in glory would never have lead to belief in his resurrection. It would at best have led to belief in Jesus’ assumption into heaven – that he had been assumed into heaven – but not that he was literally risen from the dead, which is a different category in Jewish thinking than assumption to heaven. So the explanatory power of the Hallucination Hypothesis isn’t very great because it doesn’t explain why the original disciples came to believe in and proclaim the resurrection of Jesus, which is one of those data to be explained – namely, the origin of the belief in the Resurrection.

(3) A third criterion would be plausibility. Here we are asking about the degree to which accepted knowledge implies the hypothesis. What is the degree to which accepted knowledge implies the hypothesis? Here, by accepted knowledge, we mean both the background evidence as well as the specific evidence in this case – the data that we have in this case – compared to the degree to which the data would imply the falsity of the hypothesis. So plausibility means you take all of your information and ask “What is the degree to which it implies the truth of the hypothesis as opposed to implying the falsehood of the hypothesis?” Which one is more plausible given the data – that the hypothesis is true or that it is false? Let’s apply this again in the case of the resurrection. Take, for example, the Wrong Tomb Theory – the women went to the wrong tomb and found a gardener there who says, “Oh, you are seeking Jesus of Nazareth? He is not here, see the place where they laid him,” pointing to some other tomb. That was one of the theories for explaining the empty tomb. Is that very plausible? Not at all when you think about it; because if the women had made that mistake and they began to proclaim the resurrection in Jerusalem, all the Jewish authorities would have had to do was just point to the right tomb! They must have been very amused when the disciples went around saying he has risen from the dead, when Joseph of Arimathea could say, “No, here is where he is laid, and there is the corpse in the tomb.” So you see there is not much plausibility to that hypothesis – the data tend to imply its falsehood rather than its truth.

(4) Fourth would be the degree to which the hypothesis is what’s called “ad hoc.” You want a hypothesis which is less ad hoc. What does ad hoc mean? Literally, in Latin, it means “to this.” It means the degree to which the theory is concocted just to explain this. It is contrived – there is a sort of artificiality about it, such that you wouldn’t have good reason to believe it otherwise.7 So if the theory is ad hoc, it means that there really isn’t much independent reason to believe that it is true; instead, you have got to invent all kinds of new hypotheses just to explain this evidence that you don’t have any grounds to believe is true. Let me illustrate this with the Twin Theory. You remember that fellow I debated at the University of California, Irvine, who said that the best explanation for the resurrection was that Jesus of Nazareth had an unknown twin brother.8 Jesus really wasn’t born to Mary and Joseph – he and his twin brother were born to some other Jewish peasant lady, and some way or another he got switched with Mary and Joseph’s baby. So Mary and Joseph now thought that Jesus was their baby, when in fact he was the twin of somebody else born to some other Jewish woman. His brother grew up independently from him in some other town in Judea, and, just at the time of the crucifixion, his brother came back to Jerusalem, saw that Jesus was crucified, stole the body out of the tomb, and presented himself to the disciples, who then mistakenly thought Jesus was risen from the dead. Now that is a theory which is ad hoc in excelsis! We don’t have any reason to believe that Jesus had a twin brother; we don’t have any reason to believe that he was switched with some other baby; we don’t have any reason to believe this fellow came back to Jerusalem just at this time. It is so ad hoc and contrived that it fails to be the best explanation. During the Q&A that I had at this debate, one person in the audience got up and captured it nicely, this ad hoc-ness – he said, “Dr. Cavin [my opponent] there is only one thing that remains unexplained on your theory – who was piloting the helicopter at the time of the ascension?” Well, that was just one more ad hoc hypothesis that needed to be added to complete the theory!

Number (5) is that it needs to be disconfirmed by fewer accepted beliefs. What do we mean by that? To say it is disconfirmed by fewer accepted beliefs means that if you add the hypothesis to accepted truths, fewer falsehoods would result. By adding our hypothesis to the body of accepted beliefs you will have fewer falsehoods if it is disconfirmed by fewer accepted beliefs. So to apply this to one of the theories of the resurrection, take the Apparent Death Theory – that Jesus was taken down from the cross alive, placed in the tomb, where he revived and got out to convince the disciples that he was risen from the dead. That is hugely disconfirmed by what we know about medical science and crucifixion. I think Mike Licona talked about that – studies of what happens to a person when he is crucified show that it is preposterous to think that Jesus could have been still alive when he was taken down from the cross, much less, if he were alive, not to have died almost immediately from exposure in the tomb without extraordinary medical care and medical attention. So the Apparent Death Theory is hugely disconfirmed by accepted beliefs.

One just might say here, “But isn’t the hypothesis that God raised Jesus from the dead disconfirmed by accepted beliefs? Isn’t that also disconfirmed by the fact that dead men do not rise? Doesn’t that disconfirm the hypothesis?” I think not. Notice how the Resurrection Hypothesis is worded. The Hypothesis is not that Jesus came back naturally from the dead – that is disconfirmed hugely by what medical science and biology tell us! The Hypothesis is: God raised Jesus from the dead. That is not in any way disconfirmed by what medical science tells us about what would happen naturally to someone who is dead. It would be a miracle for somebody to come back from the dead! Well, that is exactly what this hypothesis is! It says that God raised Jesus from the dead, and that hypothesis is not disconfirmed by any accepted beliefs about what happens naturally to people after they are dead. That would be again this old problem of miracles: people who have a prejudice against miracles don’t want to accept a supernatural explanation.9

(6) Finally, the sixth criterion would be that it exceeds its rival hypotheses in meeting criterion (1) through (5) – that there is little chance that future evidence or future discoveries will show that one of these other hypotheses turns out to meet the five criteria better than the accepted hypothesis. So, in order to be the best explanation, it should exceed its rival hypotheses in meeting conditions (1) through (5) to a good degree. What I have tried to do in my published work, as well as in my debates, is to show that the best explanation of the data with respect to Jesus’ resurrection is the hypothesis God raised Jesus from the dead. If you are interested in seeing these criteria applied in some detail, look at my book with Gerd Lüdemann called Jesus’ Resurrection: Fact or Figment?10 where Lüdemann defends the Hallucination Hypothesis and I take that hypothesis through each one of these six criteria. In assessing rival, live options, very often what you will find is that one option will, say, meet some criteria better but another option will meet other criteria better, so that one hypothesis will have great explanatory power and explanatory scope but it might be very ad hoc or it might be more implausible, whereas some other theory might be very plausible but would have narrow explanatory scope. What I find in the case of the resurrection, at least with respect to the Hallucination Hypothesis, is that in every single one of these areas the Resurrection Hypothesis far outstrips the Hallucination Hypothesis in terms of its scope, power, plausibility, degree of ad hoc-ness, and so forth. If you are interested in seeing these applied specifically to a rival hypothesis look at the book Jesus’ Resurrection: Fact or Figment?

So, in conclusion, I think this is the way a successful apologetic for Jesus’ resurrection should be structured. As interesting as the probability calculus is, the fact is that some of those probabilities are probably inscrutable for us today. But that is not how historians work anyway. The way historians work is by inference to the best explanation. You have a pool of live options to explain a body of data which has been established. What the Christian apologist does is he establishes a body of data that needs to be explained concerning the fate of Jesus of Nazareth. These will include things like the discovery of the empty tomb, the postmortem appearances, and the very origin of the disciples’ belief in his resurrection. Then you will assess the various hypotheses for explaining that data in terms of these criteria and the theory which provides the best explanation is going to be then the preferred theory and will be the theory that you ought to accept and regard as true. The Christian apologist will maintain that that will be the Resurrection Hypothesis.


Question: Will this be similar to historical sciences like the theory of evolution?

Answer: Yes. You could also say geology or paleontology, which are also historical sciences. For example, one of the arguments that one will hear against creationism is that it doesn’t have explanatory scope. For example, there are facts concerning what scientists call biogeography which are very interesting. For example, why is it that all of the marsupials tend to be found in Australia? Marsupials are mammals that bear their infants in a pouch, like an opossum or a kangaroo. The creationist could say, “Well, God just chose to create all the marsupials in Australia. That was just his choice; he didn’t want to do it anywhere else.”11 But the evolutionist will say it is much more plausible that all of these marsupials are interrelated and all descended from one common ancestor and, being isolated on that continent, that is why they are all there. Now, of course, he has to deal with possums, right? – which are North American marsupials – it is not true that they are all confined to Australia. There are marsupials elsewhere. But nevertheless that is just an application of the criteria that I am talking about. That would be an illustration where someone would indict the theory by saying its scope is too narrow. By contrast, on the other hand, creationists and intelligent design theorists are constantly blasting Darwinism because of its weak explanatory power. The mechanisms of genetic mutation and natural selection don’t have the power to explain the degree of biological complexity that we have today. So these are exactly the kind of criteria that would be applied in those kinds of debates.

Question: (inaudible)

Answer: That is right. When you are dealing with agents you have that additional problem that you are not dealing with random processes. But with things like natural selection and mutation there you are talking about purely non-intelligent, non-directed processes that would be more amenable to a probability-type approach, if that could be done.

Question: (inaudible)

Answer: What you are raising is the important issue that we talked about when we spoke about the ID movement. That is, what goes into the pool of live options? You see, you can’t consider every hypothesis. That would just be uncontrollable. So everybody has to limit the pool to live options. So, for example, no historian, despite what you hear in the sensational press, takes seriously the idea that, say, Jesus was an alien from outer space who had come down from a UFO and learned the Hebrew language so he could convince people to follow him – Chariots of the Gods kind of thing. Nobody takes seriously something like that – it is not a live option. Now, determined naturalists will say that any kind of supernatural explanation is excluded from the pool of live options. Of course, that is Philip Johnson’s point about evolutionary theory. If you make the live options only naturalistic options, he says, “Yeah, I think Darwinism is the best explanation. I would infer to Darwinism, too.” But he says, “Why exclude supernatural explanations?” Here, you see, we are going to demand a rationale for excluding it. What you will find is it will typically come back to the problem of miracles that we have just been talking about for the past few weeks. Or maybe even the question of theism. If God exists, this will raise the probability of a supernatural explanation, and so in one sense maybe one will need to retreat all the way back to the question of “Is there a God?” and then go through the arguments of natural theology that we have talked about on why we think there are good reasons to think that God exists. Otherwise, the exclusion of supernaturalistic hypotheses from the pool of live options becomes arbitrary. You need to have some kind of rationale for excluding it.

Question: (inaudible)

Answer: Very nice! Yes, N. T. Wright says to the sceptic, you have to first recognize that your problem with miracles is not a historical problem – it is not a question of the evidence – it is a philosophical problem. And when you are ready to discuss it on that level, then we can; and I say, “Yeah, let’s do it!”

Question: (inaudible)

Answer: Yeah, that is a good point. I think there are arguments for the existence of God that don’t appeal to Intelligent Design. For example, there are versions of the cosmological argument that asks philosophical questions like “Why is there anything at all rather than just nothing?” Why does anything exist? You could argue that the best explanation is that there is a necessary being who exists. Or the moral argument from moral facts. We sense the existence of objective moral values and duties in the world, and those could not exist without God. Therefore, that gives good reason to believe that God exists as a locus, or source, for moral values.12 There are non Intelligent Design type arguments. If you are interested in looking at those, take a look at J. P. Moreland and my book called Philosophical Foundations for a Christian Worldview, which has two chapters in it on arguments for God’s existence.13 But if the person is not even going to allow you to give arguments for God’s existence because of his naturalism, then he simply presupposes atheism to be true! He is begging the question. So it is not you that is reasoning in a circle; it is he! He is reasoning in a circle in presupposing atheism in order to preclude theism. You have got to say, “If we are going to talk about this, we have to do so with an open mind.”

Question: (inaudible)

Answer: I don’t know the answer to that. I am an open-minded inquirer myself about these things. I find the facts of biogeography to be interesting and disturbing, and I don’t know. It is something I have not looked into. Don’t take my ignorance here as indicative of the fact there is not a good answer; rather, take it as a fact that this isn’t my area of specialization, so I never explored it. I just heard this raised in debates; I’ve never looked into it because it is just not on the front burner for me, so I don’t know the answer. I do know there are marsupials that don’t live in Australia – we had one on our back porch some time ago! So I don’t know. I am just saying that any good creationist theory of origins is going to need to address those facts and will need to have some explanation, and so I am leaving that up to my colleagues like Paul Nelson and Michael Behe and Doug Axe and others to work on that area because that is an area I am not working on.

Question: (inaudible)

Answer: Your point is that they may see that, given this data, the Resurrection Hypothesis is the best option, so they are going to go back and deny the data. They are going to say that Jesus never existed or the tomb is a legend or they never had experiences of seeing Jesus alive from the dead. You are absolutely right! People like John Dominic Crossan, Gerd Lüdemann, and so forth, deny the fact of the empty tomb because once you grant the empty tomb, you are in trouble! So they deny the data. But – and here is the thing that I find so intriguing – the majority of New Testament historians today accept this data. They accept the historicity of the empty tomb. You saw it in the “20/20” show – those Jewish scholars that were interviewed: “Yes, the tomb probably was found empty.” It is the sceptic here who is on the defensive, not the Christian. The data is largely accepted by New Testament historians today, and so we stand comfortably within the mainstream. The question, then, is, what is the best explanation?14


4 5:00

5 C. Behan McCullagh, Justifying Historical Descriptions (Cambridge: Cambridge University Press, 1984), p. 19.

6 9:57

7 15:00

8 This was a 1995 debate with R. Gregory Cavin titled “Dead or Alive?”
Audio of this debate (minus the Q&A session) can be found at (accessed April 17, 2012) or you can purchase an audio CD of this debate at

9 20:05

10 William Lane Craig and Gerd Lüdemann, Jesus' Resurrection: Fact or Figment?, ed. Paul Copan and Ronald Tacelli (Downer's Grove, Ill.: Inter-Varsity Press, 2000).

11 25:03

12 30:08

13 William Lane Craig and J. P. Moreland, Philosophical Foundations for a Christian Worldview (Downer's Grove, Ill.: Inter-Varsity Press, 2000). Chapters 23 and 24.

14 Total Running Time: 33:39 (Copyright © 2012 William Lane Craig)