Here’s the link to this article by David Kyle Johnson–December 31, 2021
[This article was originally published in Socio-Historical Examination of Religion and Ministry, Vol. 3, No. 1 (Summer 2021), pp. 26-52. The version presented here has been slightly modified.]
William Lane Craig has argued that it is rational to believe that God raised Jesus from the dead because the hypothesis that such an event happened provides the best explanation of the available evidence (e.g., the biblical witness and the martyrdom of the apostles). According to Craig, if you consider what an explanation must do—the criteria it should meet—the resurrection hypothesis meets them better than its naturalistic competitors. Robert Cavin and Carlos Colombetti, however, very skillfully refuted his argument by revealing not only the problems with Craig’s criteria and his criteria-based approach, but also how the resurrection hypothesis actually fails to meet the criteria Craig proposed.
Later, in the pages of SHERM Journal, Stephen T. Davis tried to defend Craig’s argument, but Cavin and Colombetti (again, in SHERM Journal) not only refuted his argument as well, but fully explained why the resurrection hypothesis cannot be the best explanation. The standard model of physics “entails that God never supernaturally intervenes in the affairs of the universe that lie within its scope”, and thus the resurrection hypothesis is necessarily implausible and has low explanatory power. They even go on to give a rigorous Bayesian analysis of why the legend hypothesis—the idea that “the New Testament Easter traditions that relate group appearances of the Risen Jesus did not originate on the basis of eyewitness testimony but arose, rather, as legend”—is a much better explanation of the available evidence than the resurrection hypothesis.
Their argument is devastating for those who claim to believe in the Resurrection because it is the best explanation of the evidence. From start to finish, however, the debate between Craig, Davis, and Cavin and Colombetti is difficult to follow—most definitely for the lay person, and even for some professional philosophers. Not everyone is familiar with the relevant nuances of inference to the best explanation and philosophy of science, and Bayesian analysis can go over the head of even analytically trained philosophers. It is the goal of this essay, therefore, to explain in more easily understandable terms why the resurrection hypothesis cannot be the best explanation for the evidence that those like Craig and Davis offer in its favor. To do so, I am going to use a method of reasoning, a version of inference to the best explanation, I teach to college students every semester—one popularized by Theodore Schick, Jr. and Lewis Vaughn in their textbook How To Think About Weird Things. They call it the SEARCH method. I will explain the criteria it utilizes and how the method works before going on to apply it to the resurrection hypothesis and its competitors.
In doing so, I do not take myself to be breaking new ground, or even presenting an argument that is immune to objection. Indeed, while Cavin and Colombetti would undoubtedly agree with the conclusion of my argument, they would also point to shortcomings of the SEARCH method and argue for the superiority of the Bayesian approach. I would not presume to argue that they are wrong (although, as I will explain, I do not think the shortcomings prevent my argument from establishing its conclusion). But I believe that the following is an argument that the average person, and average philosopher, can more easily wrap their head around—and having such an argument, clearly articulated, is not only useful, but can do nothing but provide further reason to think that Cavin and Colombetti are right. The resurrection hypothesis not only fails to be the best explanation, but it is not even a good one.
Abduction and Inference to the Best Explanation
The procedure for identifying and embracing the best explanation for something is very appropriately named: “inference to the best explanation” (IBE). When employing IBE, one compares multiple possible explanations for something, and then infers (i.e., accepts) what is determined to be the best one. In the literature, IBE is often equated with “abduction,” but as William H. B. Mcauliffe explained in “How did Abduction Get Confused with Inference to the Best Explanation?,” the person who coined the term abduction, C. S. Peirce, did not originally conceive of abduction as IBE.
As Mcauliffe demonstrates, according to Peirce, abduction is the process by which one generates hypotheses or explanations to be tested, considered, or compared. Now, as Harry Frankfurt points out, the process Peirce describes does not actually generate any new ideas; and what Peirce is most concerned with are the criteria by which one can determine which hypotheses we should bother to consider or test (once they have been generated). Peirce did argue that “the only way to discover the principles upon which anything ought to be discovered is to consider what is to be done with the constructed thing after it is constructed”, so perhaps “the problem of constructing a good hypothesis is … analogous to the problem of choosing a good hypothesis.” But, strictly speaking, the actual formulation or creation of hypotheses to explain certain phenomena does not really follow a pattern; it must truly be creative, like art. Nevertheless, since some of the criteria Peirce described are also useful for determining which explanation is the best one, later philosophers like Harman, Lipton, and Thagard, folded those criteria into their articulations of IBE, and also used the term “abduction” to refer to IBE.
One shortcoming of IBE is the so-called “best of a bad lot” problem; if the list of hypotheses you are considering does not include the true one, determining which among them is the best cannot reveal the truth. This is where recognizing and developing what Peirce called abduction could be very useful; a process aimed at generating hypotheses that are likely to be true would make “bad lots” less of a threat. Fortunately, however, for our purposes, we do not have to worry about this. Not only are the hypotheses we will consider already generated—they are already discussed in the literature and even anticipated by some of the Gospel authors—but to prove that the belief in the resurrection hypothesis is irrational, all we have to do is find one that is better. It need not even be true. If even one possible explanation that we consider is better than the resurrection hypothesis, we will have shown belief in the Resurrection to be irrational.
In my opinion, the clearest, easiest to understand, and most useful articulation of IBE belongs to Theodore Schick, Jr. and Lewis Vaughn. In their book How To Think About Weird Things, they divide IBE into steps that they call the SEARCH method. State the claim. Examine the evidence for the claim. Consider Alternative hypotheses. Rate, according to the Criteria of adequacy, each Hypothesis. Because I take this to be relatively easy to understand, this is what we shall use. Before we begin, however, it will be useful to clarify each step.
When stating a claim, one needs to be as specific and precise as possible. This will make clearer what evidence should be considered. When evaluating the evidence for the claim, one must determine whether the cited evidence is relevant and then bring to bear all the critical thinking lessons that philosophers should know (and that Schick and Vaughn teach in their book). What is the nature and limit of the evidence? Could the shortcomings of our perception and memory be at work? Are there logical fallacies involved? Are there probabilistic or statistical errors? One must be very careful to not be fooled by bad evidence. Considering alternative hypotheses really involves two steps: stating alternative hypotheses and evaluating the evidence for them as well. A person can state one, or many, but of course, it is most efficient to restrict oneself to the hypotheses that are most likely. And when rating the hypotheses, to figure out which one is the best, one must use the criteria of adequacy.
The criteria of adequacy are what a good explanation should be (or do), by definition. A good explanation should:
- be testable: make novel predictions by which the explanation can be falsified.
- be fruitful: get the novel predictions it makes right.
- be wide-scoping: explain a variety of phenomena without raising unanswerable questions or replacing one unexplained thing with another.
- be simple/parsimonious: not make more (especially existential) assumptions than are necessary.
- be conservative: cohere with that which we already have good reason to believe (e.g., established scientific knowledge).
Now a good, or even the best explanation, need not always be or do all these things. Indeed, scientific revolutions happen when nonconservative explanations become accepted because they proved themselves to be more testable, fruitful, wide-scoping, and simpler than their “established” competitor. Einstein’s theory of relativity, for example, was not conservative when it was first proposed because it contradicted Newton’s, which was the established theory at the time. But relativity proved itself by being testable, more fruitful,, wide-scoping, and simple than Newton’s theory. Or take the germ theory of disease. It introduced new entities, and thus was not simple, but it proved itself by successfully predicting things like the efficacy of handwashing, vaccines, and explaining how diseases spread.
That is also not to say that there cannot be ties. In my opinion, this is why it is currently impossible to delineate between the major interpretations of quantum mechanics. For example, neither David Bohm’s pilot wave theory nor Hugh Everett’s multiverse theory make any new novel predictions; and both seem to have the same explanatory power. But whereas Everett’s multiverse interpretation is not simple (because it requires the existence of multiple universes), Bohm’s interpretation is not conservative (because it violates relativity by requiring faster than light signaling). Which is better? It is a matter of preference. Which do you think is more valuable: simplicity or conservatism?
But the fact that there can be ties does not mean that, when there is not, one explanation is not the best. If you compare competing hypotheses according to the criteria, and one clearly turns out to adhere to them the most, then that is the best explanation and it is irrational to embrace the others. You might refrain from endorsing it, in hopes of an even better one coming along, but that possibility cannot make embracing the worse hypotheses rational. Or, alternately, if one hypothesis does not emerge as the best, but others are clearly demonstrated to be inferior, embracing the inferior one(s) cannot be rational. And as we shall now see, among the proposed explanations for the cited evidence of Jesus’ resurrection, the resurrection hypothesis is the worst. Thus, it cannot be rationally accepted.
State the Claim and Evaluate the Evidence
In the case of Jesus’ resurrection, stating the hypothesis that we need to evaluate first, clearly and precisely, seems easy enough to do: “God supernaturally raised Jesus from the dead.” But Craig also believes that Jesus’ postresurrection body was a “soma pneumatikon” (spirit body image) that was immortal, imperishable, and able to de- and re-materialize. So, to align with Schick’s suggestion that one should state the hypothesis as precisely as possible, it should be stated like this:
The resurrection hypothesis: A first-century Palestinian apocalyptic preacher named Jesus was crucified and then raised from the dead; this was accomplished by the supernatural powers of a supernatural being we now call “God,” who gave him a spiritual body that was immortal, imperishable, and able to de- and re-materialize.
In other words, Jesus’ resurrection happened pretty much as the Bible described. Evaluating the evidence for this claim, however, is a bit more complicated than stating it.
The first piece of evidence is the biblical account itself, and its report that Jesus was crucified, died, was placed in a tomb which was later found empty, and then appeared to the apostles (by, for example, just appearing out of thin air in a room they were in). But this piece of evidence must be rated as low, given that the Bible is in no way a reliable historical document. For example, scholars generally agree that the Old Testament is mostly ahistorical; neither Abraham, nor Moses, or even King David, for example, likely ever existed. The New Testament does not fare much better. Even if one sets aside the (often convincing but unpopular) view that Jesus never existed either, the New Testament is not considered by most biblical scholars to be historically reliable. Indeed, it is especially unreliable when it comes to the details of Jesus’ life. It is full of historical contradictions. The two nativity stories in Luke and Matthew, for example, are impossible to reconcile historically. And the same is true for the different gospels’ accounts of Jesus’ crucifixion and resurrection.
The latter is worth elaboration. The Gospels say two things about when Jesus was crucified. Mark (Mark 14:12; Mark 15:25) says he was crucified on the day of Passover at the third hour, but John (19:14-16) says it was the day before Passover at the sixth hour. In the synoptic Gospels (Mark 15:23; Matt. 27:48; Luke 23:36), Jesus refuses to drink while on the cross, but in John (19:29-30) he does not refuse. Historically, the Romans usually did not crucify thieves—crucifixion was reserved for enemies of the state—but the synoptics say the men crucified with Jesus were thieves. Additionally, those the Romans did crucify were not taken down and put into a grave but were instead left up to rot and later thrown into a pit. Indeed, the list of discrepancies in the story is vast and numerous. More importantly, for our purposes, discrepancies also exist in the biblical account of the Resurrection. How many people saw the resurrected Jesus, for example? Paul says it was over 500 (1 Corinthians 15:6), but the Gospels never mention this crucial event. Were the women joyful or sad? Could the apostles touch Jesus, or not? Was he recognizable? Did he appear in Galilee or Jerusalem? Different gospels say different things. If two people told you so such wildly conflicting stories about anything, you would know that one is lying, and would not be able to believe either without corroborating evidence.
Such discrepancies are not surprising given that we now know the Gospels were not written by followers of Jesus named Matthew, Mark, Luke, and John right after the events happened—but instead by noneyewitnesses decades after Jesus would have lived. (The first clue was that the Gospels are written in Greek and the apostles and other followers of Jesus would have spoken Aramaic.) Mark was written first, and whoever wrote Matthew and Luke used it as source material, but John was written independently—that is why it is so different and contradictory. Therefore, the resurrection accounts in the Gospels cannot be considered a reliable account of what happened after Jesus’ death because the people who wrote them were not around when Jesus’ resurrection supposedly happened. The Gospel accounts, therefore, do not provide good evidence that a resurrection occurred.
Of course, some have claimed that the Gospels are simply written accounts of oral traditions that were started by the apostles and perfectly preserved by the Christian community. But as Bart Ehrman proves in his book Jesus Before the Gospels, there is no way that actually happened. Oral traditions are no different than rumors; all research suggests that they change and are added to drastically as they are passed on. Add to that how unreliable our senses and memories are to begin with, and we have every reason to not take this notion seriously. To put it simply, I would not believe my neighbor if he said he saw someone raise from the dead with his own eyes yesterday—and justifiably so. How much less should I believe noneyewitness accounts passed through multiple languages for 2000 years?
Outside the biblical witness, proponents of the Resurrection will often cite other factors—like the tradition that the apostles died as martyrs for their faith. “Why would they be willing to do that,” the argument goes, “if they were not convinced the Resurrection happened?” But there are two things to say in response.
First, the evidence that the apostles actually died for their faith is poor. In his book, The Fate of the Apostles, Sean McDowell lays out the evidence and argues that (at best) Peter, Paul, James (son of Zebedee), and Jesus’ brother (James) were martyred. But even the evidence he mentions for their martyrdom was written down decades after it would have happened—which means the stories would have been preserved by oral tradition and thus are not reliable. It is just as likely that all such stories are Church traditions that were invented to try to convince skeptics. Indeed, as we shall soon see, Christians confabulating stories to convince skeptics was quite common.
Second, and more importantly, the apostles being convinced that the Resurrection occurred would not be good evidence that it happened—even if they were so convinced that they were willing to die for it. Some people are convinced Elvis still lives, and people in cults die for (demonstrably and obviously) false things they are convinced are true all the time. Again, as we shall soon see, there are many (more likely) reasons the apostles could have been that convinced that Jesus rose, even though he did not.
Something similar could be said about the “success” of Christianity. The fact that Christianity spread and is still thriving today is not a testament to the fact that Jesus indeed rose—any more than the fact that Islam is thriving today is testament to the fact that Muhamad was Allah’s prophet and rode a flying horse named Buraq up to Heaven. Religions succeed or fail primarily because of historical accident (e.g., the conversion of Constantine), not because they are (or are not) rooted in historical events. Besides, to think that something is true because it has been believed by a lot of people for a long time is a doubly fallacious argument: one that combines “appeal to tradition” with “appeal to the masses.”
So the evidence for the resurrection hypothesis is not convincing.
Consider Alternative Hypotheses
To consider an alternative hypothesis is to state it and to evaluate the evidence for and against it. Of course, given how far back in the past the Resurrection would have occurred, no evidence we will consider here will deductively settle the matter. But the discussion will be very useful when comparing the hypothesis to the criteria of adequacy in the next step.
The obvious alternative hypothesis is that Jesus did not actually raise from the dead, but people ended up believing that he did anyway. But that is far from specific, as there are a number of ways that could have happened. So, let us consider the four specific hypotheses that seem to be the most likely.
The lie hypothesis: The apostles lied and said that Jesus rose; they knew he did not but made it seem like he did. The disciples stole his body out of his tomb and buried it somewhere else, so that the women who went to the tomb on Sunday morning would find it empty and conclude that Jesus rose from the dead. People believed them, and the story grew.
The best reason to doubt this hypothesis is the suggestion that the disciples were martyred; although it’s not impossible, it seems unlikely that they would be willing to die for a lie that they knowingly told. But, as was mentioned above, the evidence for the suggestion that the disciples died as martyrs is unreliable, and so this cannot be a reason to dismiss this hypothesis.
The best evidence for this hypothesis is the fact that the author of Matthew felt it necessary to try to knock it down.
While the women were on their way, some of the guards went into the city and reported to the chief priests everything that had happened. When the chief priests had met with the elders and devised a plan, they gave the soldiers a large sum of money, telling them, “You are to say, ‘His disciples came during the night and stole him away while we were asleep.’ If this report gets to the governor, we will satisfy him and keep you out of trouble.” So the soldiers took the money and did as they were instructed. And this story has been widely circulated among the Jews to this very day. (Matthew 28:11-15, NIV)
Given the low historical value of the Gospels, and since there is no way the author of Matthew could have been privy to this secret meeting, these verses cannot be good evidence against this theory. Indeed, clearly the idea that this meeting took place was invented by the author to “shoot down” this alternative explanation. As the author makes clear, “the disciples stole the body” was a retort that early Christians often heard when they tried to convince people that Jesus rose from the dead. So this passage actually provides at least some reason to think this hypothesis is true.
At the same time, however, the mere fact people were saying the disciples stole the body is not good evidence that they did. It probably just occurred to people as a better explanation for the stories about an empty tomb. And we will have to wait until later in this process to determine whether it actually is.
One thing this does demonstrate, however, is that elements that seem to have been added to the story—like secret meetings and the guards and stone in front of the tomb—which were clearly designed to “head off” objections or alternate explanations, do not actually provide good evidence against the alternative explanations in question (and thus for the resurrection hypothesis). Yes, it is possible that there was a giant boulder in front of Jesus’ tomb, but it is at least equally possible (indeed it is more likely) that such notions were fabricated by someone in response to people saying that the apostles lied. As Christian origins scholar James Tabor put it, “What is clearly the case is that neither Matthew nor Luke are relating history, but writing defenses against charges that are being raised by opponents who are denying the notion that Jesus literally rose from the dead.”
The second alternative explanation would be along these lines.
The coma hypothesis: Jesus did not die on the cross, but was sent into a temporary coma due to his wounds. This was mistaken for death, and he later awoke and left his tomb, leaving it empty.
This hypothesis is sometimes chidingly called the swoon theory, but the condition described would be much more serious than Jesus simply passing out.
The evidence usually cited against it is actually, once again, some evidence for it: the part of the crucifixion story where the soldiers at the cross do not break Jesus’ legs to hasten his death, but instead stick a spear in his side to confirm he already died. This bit appears nowhere else but in the Gospel of John, the last and least historically reliable gospel to be written. It is, therefore, more likely that it was added to the story as a way to head off reasons people were giving at the time to doubt the resurrection hypothesis. After all, even given the story they were told, people at the time had good reason to suspect that Jesus did not actually die on the cross. It usually took most who were crucified days to die, and even those who were nailed to a cross would likely live up to 24 hours; and even according to those telling the story, Jesus was only on a cross for a few hours. Moreover, it is not like Roman guards usually stuck around and made sure people who were crucified actually died; again, the crucified were usually just left to die (and then tossed into a mass grave if wild animals did not get to them first).
One might argue that comas are never as short as 36 hours (roughly the time that passes between Friday evening and Sunday morning). But (a) I was unable to find any evidence that comas cannot last this long, and (b) this hypothesis does not require that the time between the crucifixion and finding the empty tomb was 36 hours. It could have actually taken much longer, and the “3 days” part of the story was added later.
Evidence that such things happened in first-century Palestine is bolstered by the fact that they still happen today. Consider the many stories from the modern world where even medical professionals thought someone was dead when they were not—like 29-year-old Gonzalo Montoya Jiménez, who was declared dead by 3 separate doctors before later awakening and completely recovering (even after being in cold storage). How much more common must mistaking a coma (or illness) for death have been in the ancient world? How much more unqualified (than modern medical doctors) to tell whether someone was really dead must the illiterate apostles have been? The fact that Jesus’ inerudite apostles thought Jesus had died and then been raised would not be a valid reason to think that he had. (Notice that Christians would be very quick to embrace this hypothesis to explain non-Christian resurrection stories, like that of Mr. V. Radhakrishna.)
The coma hypothesis does raise questions, however. For example, if Jesus did not die, and walked away from the tomb, what happened to him afterwards? Well, it is possible such an event would have convinced Jesus himself that he had risen from the dead. If so, things could have unfolded basically as the Gospels suggest: the women found the empty tomb, Jesus appeared to them and the apostles, he and they believed he had risen, and he sent them to convert the nations. But it is just as likely that Jesus simply went on to die of his wounds somewhere else, and his body was never found or identified. (It would likely have been eaten by animals and left unrecognizable.) And then the merely empty tomb convinced the apostles that he had risen.
But if Jesus did die (whether on the cross or after) but did not rise, this makes one wonder about all those postdeath appearances. How do we account for those? This brings us to our last two alternate hypotheses: the imposter hypothesis and the legend hypothesis. Let us deal with the former first.
The imposter hypothesis: Jesus died but the apostles came to believe that he rose because they came to believe that a different (live) person was Jesus.
There are a number of ways this could be true, so this hypothesis is not as specific as it should be; but the different possibilities are all about equally likely, so that should not matter for our purposes here.
On this hypothesis, the disciples may not have even known where Jesus was buried. Again, victims of crucifixion were usually just thrown in mass graves. If so, perhaps other people later claimed to be the resurrected Jesus, and without a way to prove them wrong with a body, people (including the apostles) believed it. After all, apocalyptic preachers were common at the time; and claiming to be Jesus could have been an easy way to gain a following—like gurus in India today often claim to be reincarnations of Vishnu, or others (like Texas’ David Koresh and Siberia’s Sergey Torop) claim to be Jesus. Such persons could have easily excused away why they looked different. “The Resurrection changed me.” But perhaps they did not need to because Jesus had a doppelganger (or even twin brother) that people mistook for Jesus, after his death. We know that twins exist, and even unrelated doppelgangers are not that uncommon. Any of these possibilities could very easily make sense of the strange biblical story (from Luke 24) where two apostles, on the road to Emmaus, did not quite recognize the person they were talking to “as Jesus” until after he was gone.
On this view, all the subsequent details—like the empty tomb, and doubting Thomas story where Thomas sticks his fingers in Jesus’ side, which were obviously later added to the story to squelch doubts—were fabricated after the fact. But this is not unreasonable at all. As we have already seen (and biblical scholars agree), those who passed them on, whether in written or oral form, embellished the Gospel stories readily, and for this very purpose. And the doubting Thomas story only appears in the latest and least historical of the Gospels: John. But that brings to mind the possibility that the entire story was fabricated, and leads us to our final hypothesis:
The legend hypothesis: The idea that a man named Jesus rose from the dead is a complete fabrication, a legend that arose in 1st-century Palestine—one that people, like the apostle Paul, came to believe, and that later inspired the writings of the Gospels, and the religion of Christianity.
One way that this hypothesis would be true is if Jesus never existed at all, and his entire story is legend—especially the part about his crucifixion and resurrection, which bears certain resemblance to the stories of other resurrected deities. On this view, Christianity was essentially invented by Paul, which accounts for why he (and not any eyewitness) was the first to write about it. Although it is not mainstream, a number of scholars have argued for this idea; as such, all the evidence they present would be evidence for this theory.
The (arguably) more conservative interpretation of the legend hypothesis, however, is that an apocalyptic preacher named Jesus did exist, and was crucified and died—but that his followers were so distraught as a result that they simply began to believe that he had resurrected out of despair. Because of this belief, “sightings” occurred and oral legends emerged—and they, in turn, were believed by Paul, later embellished by the Gospel writers (who, again, were not apostles), and believed by the early Church.
That such a thing can and does happen, especially in cults of personality, is not uncommon. Indeed, there are modern-day examples. It could have been that the disciples, much like fans of Elvis, despite the obvious evidence, simply could not accept his death and began to believe they saw him alive as a result. Such rumors spreading (especially among illiterate first-century Palestinians) would have made “sightings” more frequent, and easily laid the groundwork for a widespread belief that Jesus was still alive. From this, the stories about an empty tomb, his subsequent appearances, and even his early miracles, could have easily emerged.
Direct evidence that any of these things happened is, of course, as impossible to obtain as direct evidence that a resurrection occurred. However, there is one bit of indirect evidence that favors the legend hypothesis over all the others: the amount of time it took for the belief in Jesus’ resurrection to take hold, to be professed in writing (by Paul), and to later be written in the Gospels. On the lie, coma, and imposter hypotheses, belief in Jesus’ resurrection would have risen immediately because the evidence (faulty as it was) would have been obvious and available. The empty tomb, the postcoma Jesus, or the imposter, would have been right there! You would have therefore expected a written account of it within a year, and for the Church to have preserved such an important document. As it stands, we do not get even a bare-bones statement of the Resurrection until the writings of Paul, which were written at best 20 years later (ca. 50 CE), and the subsequent accounts proceed just as legends do: they become more grandiose over time.
This last point is worth some elaboration. From Paul we just get a creed: “That Christ died for our sins according to the Scriptures, and that He was buried, and that He was raised on the third day according to the Scriptures” (1 Corinthians 15:3-4, NASB). He mentions earlier in the letter that Jesus died by crucifixion (1 Corinthians 2:1-2), but that is it. It is not until twenty years later (ca. 70 CE, at least 40 years after Jesus would have died) that the author of Mark adds details to the crucifixion (Ch. 15) and resurrection story (Ch. 16). But even then, we do not get much; after the crucifixion, Jesus never physically appears again. The women just meet a young man who tells them Jesus was raised. Not until still another 20 years later (ca. 90 CE), after all the eyewitnesses would have been dead, do we get Matthew’s and Luke’s accounts about the crucifixion and personal appearances made by Jesus to eyewitnesses, which add even more (divergent) details to the resurrection story. And then, as much as 20 years after that (ca. 110), we get John’s gospel, which not only neglects many details in Luke’s and Matthew’s gospels, but adds many more of its own. This is exactly how legends develop; it decidedly is not how history is accurately preserved.
Indeed, James Tabor has argued that how the stories developed indicates that the earliest Christians did not believe that Jesus was physically raised and physically appeared to the disciples, but instead that Jesus was “‘lifted up’ or ‘raised up’ to the right hand of God,” and that his followers merely experienced “epiphanies of Jesus once they returned to Galilee after the eight-day Passover festival and had returned to their fishing in despair.” Only decades later, after all the original followers had died, did the belief that Jesus physically resurrected become common. This comports precisely with the legend hypothesis.
According to the Criteria of Adequacy, Rate Each Hypothesis
We have five hypotheses: resurrection, lie, coma, imposter, and legend. The fact that there is no way to directly test any of these hypotheses, since they deal with far, distant beliefs and actions of first-century Palestinians, might lead one to think that we are forced to just shrug our shoulders and say we cannot know what happened. “You cannot prove one hypothesis over the other, so it is just a matter of faith.” But there are two important responses here. First, knowledge does not require proof. If one hypothesis can be shown to be much more likely than the others, belief in it will at least be justified, if not be knowledge. Second, knowing what happened is not the issue here. The question is whether the resurrection hypothesis is, as Davis and Craig suggest, the best explanation; and we can determine that even if we are not able to directly test any of the hypotheses, or even know exactly what happened.
We can do so because testability and fruitfulness—the only criteria that deal with whether a hypothesis makes correct, novel, observable predictions—are only two of the five criteria of adequacy. We can still use the other three: scope, simplicity, and conservatism. If I go downstairs to find my TV missing, but have no video surveillance or way to lift fingerprints, I can do no testing… But I can still know that “I was robbed” is a better explanation than “a ghost took my TV.” Why? Because the former is simpler, more conservative, and has wider scope. The former does not require ghosts to exist, does not contradict the laws of physics and facts of neuroscience, and does not raise unanswerable questions about how nonmaterial objects can move TVs. We can do something very similar with the above five hypotheses.
That said, testability and fruitfulness are still somewhat relevant. Although you cannot test any of the hypotheses in a lab—as I mentioned in the previous section—on the legend hypothesis, one would expect the development of Jesus’ crucifixion and resurrection story to develop a certain way: to become more elaborate over decades. That type of development is not what to expect on the resurrection, lie, coma, or imposter hypotheses. Since that is what happened, the legend hypothesis is more fruitful. Indeed, since such hypotheses have been around for centuries, but only until relatively recently (historically speaking) did we date and order the letters of Paul and the Gospels, and discover that the Gospels were written decades later by noneyewitnesses, we could even say that the legend hypothesis made a novel prediction that turned out to be right.
The simplicity of the hypotheses is much easier to compare. The resurrection hypothesis requires the existence of a supernatural entity operating with supernatural powers. That is two enormous, grandiose assumptions that none of the other hypotheses have to make. Thus, it is the least simple. (And the fact that it says Jesus was given a unique spiritual body makes it even less simple.) Yes, the lie, coma, imposter, and legend hypotheses also make certain kinds of assumptions—but they do not assume the existence of an entirely new kind of force, entity, or body. We know that lies, comas, imposters, and legends exist. Indeed, scholars already know that additions were made to Jesus’ story, how legends develop, and that misdiagnoses of death readily occurred before the advent of modern medicine.
Granted, the imposter hypothesis requires the existence of an imposter/doppelganger, and the lie hypothesis requires a bit of a conspiracy; and that does make them less simple than the coma or legend hypothesis. But even the imposter and lie hypotheses are simpler than the resurrection hypothesis. At least we know that doppelgangers, twins, and conspiracies can and do exist. If Princess Diana appeared on TV tomorrow claiming to be back from the dead, saying that she had faked her death or had a long-lost twin would both be simpler explanations than, say, “aliens resurrected her corpse with advanced technology.” None of this entails that the other hypotheses we have considered are contrary to belief in God. Indeed, even if we granted that God exists, since such explanations do not require God or his supernatural powers to exist, by definition, they are simpler than the resurrection hypothesis.
The scope of the hypotheses is also easy to compare. The resurrection hypothesis has little scope because it invokes the inexplicable: an infinite, incomprehensible being who uses unknown, un-understandable powers, to create a body made of an inexplicable substance that has mysterious magic powers. As Schick might point out, this is a bit like trying to explain why a bridge collapsed by saying “a mysterious gremlin zapped it with a magical ray gun.” Such explanations actually explain nothing. (Notice that the alien explanation would not help you build a more stable bridge next time). To paraphrase Plato, to say “the gods did it” is not to offer an explanation, but to just offer an excuse for not having one.
The legend hypothesis, however, has very wide scope because it can explain not only the evidence cited for the Resurrection, but a vast number of other phenomena, like other legends, and things like Elvis and Hitler sightings. It explains the contradictions in the biblical accounts of the Resurrection, and why those accounts were written so much later than the events they purport to relay by noneyewitnesses. It could even explain a host of other religious beliefs about the Resurrection of other supposedly dead persons (although the lie, coma, and imposter hypotheses could also explain those as well).
The legend hypothesis is also monumentally conservative because it conflicts with nothing that we know is true. We know that (and how and why) false beliefs, even in the face of contrary evidence, can arise—even the belief that someone who has died is still living (again, like Elvis and Hitler). It coheres with what we know about how and who the Romans crucified, how long it took those they crucified to die, and how the Romans disposed of their bodies. It even aligns with what Bart Ehrman revealed about how unreliable group memories and “oral traditions” are. (Even the lie, coma, and imposter theories require the unlikely assumption that the stories about Jesus’ “resurrection” were reliably preserved orally for decades.)
Most notably, however, all but the resurrection hypothesis aligns with perhaps one of the most established facts there is: the dead stay dead. And this brings us back around to Cavin and Colombetti. They argue that, regardless of whether God exists, the resurrection hypothesis is contrary to the Standard Model of physics. Defenders of the Resurrection, like Davis and Craig, argue that the Standard Model comes with a proviso: the laws operate as usual unless there is divine intervention. They thus argue that the Resurrection is not scientifically impossible. But, as Cavin and Colombetti very skillfully explain, such a proviso is either superfluous or “renders [the] laws untestable metaphysical pseudo-science.” Consequently, the Standard Model “entails that God never supernaturally intervenes in the affairs of the universe that lie within its scope.” And that would include raising Jesus from the dead.
In other words, the Standard Model of physics—along with all of the research that has established it over the years—is in direct conflict with the resurrection hypothesis. The Standard Model thus entails that the Resurrection did not happen. This means that the resurrection hypothesis is practically as nonconservative as a hypothesis can be—not only less conservative than the lie, coma, imposter, and legend hypotheses, but even less conservative than creationism, geocentrism, and the flat Earth theory. It is contrary to all of science.
This, by all accounts, is one reason Cavin and Colombetti’s paper is so devastating to the resurrection hypothesis; it is its death nail. The unconservative nature of the resurrection hypothesis that Cavin and Colombetti reveal puts an evidential burden on it that it cannot overcome. If the resurrection hypothesis could prove itself—by being vastly more fruitful, simpler, and wider scoping than its competitors—it might have a fighting chance. As we have seen, that is how unconservative scientific theories win scientific revolutions. But (also as we have seen), the resurrection hypothesis is none of those things; indeed, by its very nature, it cannot be. It invokes inexplicable entities, acting in the unobservable and inexplicable ways; it is untestable, cannot be fruitful, has virtually no explanatory power (i.e., scope), and is, by definition, not simple. Even the ridiculous idea that Jesus had a long-lost twin brother who just happened to show up three days after Jesus was crucified would be a better explanation.
But, of course, as we saw above, the most likely explanation is the legend hypothesis. It is not only simpler than the resurrection hypothesis, but is also simpler than the lie, coma, and imposter hypotheses. It was more successful in predicting what we discovered about how Jesus’ story evolved over time, is more consistent with what we know about the reliability of oral traditions, and what the early Christians seemed to have believed. It is therefore, by definition, the best explanation of the evidence that Davis, Craig, and others cite for the resurrection hypothesis. That does not necessarily mean that the legend hypothesis is true; perhaps there is an even better explanation out there that we have not considered. But it certainly is better than the resurrection hypothesis, and thus belief in the resurrection hypothesis is irrational.
As I mentioned in the introduction, Cavin and Colombetti would undoubtedly point out the shortcomings of the SEARCH method and argue for the superiority of their Bayesian approach. For example, the SEARCH method does not show how the criteria of adequacy “fit” together or when and how one criterion should take precedence over another. What do you do in in case of ties, where one hypothesis is simpler, but another has wider scope? By factoring in conservatism and simplicity into prior probability, and scope and fruitfulness being a factor of Bayesian likelihood, the Bayesian approach can potentially answer these kinds of questions. But this shortcoming of the SEARCH method in no way affects the strength of the argument I have presented here because there is not a tie. The legend hypothesis is the best explanation is every respect; and the resurrection hypothesis is the worst in every respect. There is no tie for the Bayesian approach to break.
But none of this should be surprising. Supernatural explanations never fare well against their competitors because, by their very nature, they do not meet the criteria of adequacy—they do the exact opposite of what good explanations must, by definition, do. They invoke inexplicable, extra, supernatural assumptions that are contrary to the laws of science and are thus, by definition, nonsimple, unconservative, and cannot have scope. Indeed, Schick has argued that “God did it” can never be an adequate explanation of anything, and I have argued that the same is true for “a miracle occurred.” Since the Resurrection would have been a miracle caused by God, it is no wonder that it fails so monumentally at being a good explanation.
See article for notes and references.