Home » Library » Modern Library » Richard Carrier Indef 4e

Richard Carrier Indef 4e

Craig’s Empty Tomb & Habermas on Visions (1999, 2005)

Richard Carrier


[Part 4E of a larger Review of In Defense of Miracles.]


The Resurrection of Jesus

Two chapters are devoted to the resurrection of Jesus. First, Craig argues that Jesus was miraculously resurrected and that we can know this from an analysis of the sources. In particular, he argues that “miracle” is the best explanation for the combined facts of an “empty tomb” and the rise of the Christian belief in a resurrection. Second, Habermas argues that “after his death Jesus appeared alive to his followers” (262) and this constitutes proof of a miracle. I will only address particular analytical and methodological problems in these chapters, since I have already dealt exhaustively with the evidence in another collection of articles, Why I Don’t Buy the Resurrection Story (6th ed., 2006), and in more precise detail in three chapters I contributed to the book The Empty Tomb: Jesus Beyond the Grave (2005). I will occasionally refer and link to specific sections of the former as I go. Such links are identified by the abbreviation “WIDBRS” in brackets. Some of Craig’s arguments involve things I have already discussed in my reviews of the other chapters of this book (see Table of Contents), so I will not repeat them here.

To get things going, there is a humorous detail in Craig’s essay which will forever come back to haunt him: Craig approves of the principle that “too many scholars think it sufficient to show that the evidence can be interpreted in accordance with their hypothesis rather than that their interpretation is required by the evidence” (255). If only Craig followed his own advice! Indeed, this principle undermines Craig far more than anyone else, for if we can show a genuinely possible natural explanation, then, even if it is not a necessary explanation, Craig still cannot establish that a miracle occurred even if one did (see my review of Purtill and Corduan). For if “miracle” is not a “required” explanation of any given report, then belief in Christianity is not warranted on the grounds of such a report. Even so, Craig is overreacting to historical theories: rarely is only one theory “required” by the evidence. In reality, there are any number of viable theories for almost any event, and the one with the greatest evidential support generally wins the day–but it is rare that a winning theory for any historical event is certain enough to bet your life on, a lesson Christians really need to learn.

Would Jesus Have Been Buried in a Tomb?

I will begin by agreeing with Craig that, if Jesus was crucified in Judaea in the first century, then we have at least some reason to believe Jesus was buried in a tomb, and not tossed into a pit or left to rot. Although it is indeed possible that the entire tomb story is an invention (and I have argued elsewhere that in fact it is), there is no obvious dogmatic or rhetorical reason to invent this fact, nor any evidence directly to the contrary, and tomb burial was the usual method in Palestine at the time, even for condemned criminals, as I explain in my article Jewish Law, the Burial of Jesus, and the Third Day (2002).

Even so, the evidence of a tomb burial is not as good as Craig makes it out to be. For example, he cannot use Acts (13:29-30) to support the Gospels, because Acts was written by the author of the Gospel of Luke, who in turn could have been depending on Mark for this information. Craig is also wrong that 1 Corinthians 15:3-5 confirms the tomb story, since it neither mentions a tomb nor any surrounding details like the involvement of Joseph of Arimathea. It doesn’t challenge the story, either. It’s simply too vague. The entire reference consists of a single word, etaphê, which is the past tense of thaptein which simply means to bury, in either a tomb or a grave. One must also realize that even if we can show that Jesus was buried in a tomb, we still cannot be sure it was a known tomb, much less that of the possibly-fictional Joseph, or that the burial occured as quickly after death as the Gospels claim, etc.

Likewise, Craig’s argument that Mark’s omission of the name of the high priest (14:53-4, 14:60-3) implies that the tomb story was written while Caiaphas was still holding that office (hence before 37 A.D.) is not really as secure as he would like. Mark is uniquely short and brief and lacking in historical details of all kinds. That Mark is not specific thus does not entail that the author wrote when Caiaphas was still the high priest–it may merely be another example of this characteristic brevity and lack of detail. In fact, Mark might have omitted the name because he did not know it, which would actually suggest a late date for the story, not an early one. The later Gospels show historical details that could have been added after historical research, or by more knowledgeable authors (as lists of priests, for example, would have been publicly inscribed or accessible in records or histories).

Ultimately, I believe if there was a historical Jesus he would have ended up in someone’s tomb eventually, regardless of what anyone knew, but Craig is far overstating the case when he says that we have “extremely early, multiple attestation” of his burial in a tomb. We essentially have only one source from half a century later: the Gospel of Mark, which was used, directly or indirectly, by all the other relevant authors, and which very likely invented both the tomb and burial story, as I argue in The Empty Tomb (2005), pp. 155-97, while a similar point is made in my review of The Homeric Epics and the Gospel of Mark (2000).

There Is a Possible Anachronism

There is another reason to doubt the tomb burial: the tomb blocking stone is treated as round in the Gospels, but that would not have been the case in the time of Jesus, yet it was often the case after 70 C.E., just when the gospels were being written. Amos Kloner, in “Did a Rolling Stone Close Jesus’ Tomb?” (Biblical Archaeology Review 25:5, Sep/Oct 1999, pp. 23-29, 76), discusses the archaeological evidence of Jewish tomb burial practices in antiquity. He observes that “more than 98 percent of the Jewish tombs from this period, called the Second Temple period (c. first century B.C.E. to 70 C.E.), were closed with square blocking stones” (p. 23), and only four round stones are known prior to the Jewish War, all of them blocking entrances to elaborate tomb complexes of the extremely rich (such as the tomb complex of Herod the Great and his ancestors and descendants). However, “the Second Temple period…ended with the Roman destruction of Jerusalem in 70 C.E. In later periods the situation changed, and round blocking stones became much more common” (p. 25).

Why is this significant? Three of the four Gospels repeatedly and consistently use the word “roll” to describe the moving of the tomb’s blocking stone (“rolled to” proskylisas, Matthew 27:60; “rolled away” apekylisen, Matthew 28:2; “rolled to” prosekylisen, Mark 15:46; “roll away” apokylisei Mark 16:3; “rolled away” apokekylistai Mark 16:4; “rolled away” apokekylismenon Luke 24:2). The verb in every case here is a form of kyliein, which always means to roll: kyliein is the root of kylindros, i.e. cylinder, in antiquity a “rolling stone” or even a child’s marble. For example, the demon-possessed boy in Mark 9:20 “rolls around” on the ground (ekylieto, middle form meaning “roll oneself,” hence “wallow”). These are the only uses of any form of this verb in the New Testament.

Kloner argues that the verb could just mean “moved” and not rolled but he presents no examples of such a use for this verb, and I have not been able to find any myself, in or outside the Bible, and such a meaning is not presented in any lexicon. His argument is based solely on the fact that it “couldn’t” have meant rolled because the stone couldn’t have been round in the 30’s C.E. But he misses the more persuasive point: if the verb can only mean round, then the Gospel authors were not thinking of a tomb in the 30’s C.E. but of one in the later part of the century. If the tomb description is flawed, this would also put Mark as being written after 70 C.E., and it would support the distinct possibility that the entire tomb story is a fiction.

Even so, there is nothing decisive about this. There could still be a core truth about a tomb burial, with the details being added out of the imaginations of the authors or their sources, as often happened when even reliable historians described scenes in such vivid detail. There was a kind of acceptable license when painting scenes this way, provided the historian did not contradict any known facts or propose the implausible. So the fact that the story was told in terms familiar to the writers, though historically inaccurate, would not entail that the story did not originally involve sliding a square stone instead. But the incongruity would still lend some support to an overall case against the authenticity of the story, since Glenn Miller’s claim to the contrary is not well founded.[1]

Was There an Empty Tomb?

Regardless of what we conclude about the tomb burial, I doubt the empty tomb was an early Christian belief. It finds no mention, directly or indirectly, in any of the epistles, and there is some good evidence that Paul did not believe in the resurrection of a buried corpse, but resurrection into an entirely new body.[WIDBRS] Craig asserts that it is “clear” that Paul believed in a physical resurrection of a corpse, but the only evidence he offers entails just the opposite conclusion (252). And when he tries to argue the point further, none of his evidence really convinces (253): the phrase “died…was buried…was raised” (apethanen…etaphê…egêgertai) can just as easily be a metaphor as an indication of an actual physical raising; the “concept of the resurrection itself” does not entail any more a physical than a spiritual idea (I demonstrate this extensively in The Empty Tomb, pp. 105-54); Paul’s “Pharisaic background” is irrelevant, since he is, intellectually, no longer a Pharisee when he converts to Christianity, and a conversion entails fundamental changes of beliefs; the expression “on the third day” likewise has no necessary connection with a physical resurrection, since it was believed to be required by scripture and by Jesus’ own promises;[2] the phrase “from the dead” (Romans 4:24) does not entail a physical return from the dead, for ek nekrôn literally means “out from among the dead people” and a spiritual resurrection would entail this just as a physical one would; and Paul’s “belief in the personal return of Christ” (1 Thessalonians 4:13-17) includes the idea that we will all be sucked up into the sky to live with God, which could be a spiritual existence just as well as a physical one, and the rest of the passage is certainly compatible with such an idea.

An even worse argument is the claim that a verbal parallel between Paul (1 Corinthians 15:4) and Mark (16:6) confirms that Paul was “familiar with the empty-tomb tradition.” But how does that follow? Mark is adopting Paul’s language, and thus Mark may have added any new ideology to the simpler core belief of Paul. This therefore tells us nothing about what Paul believed. It is also silly to suppose that Paul read Mark or any gospel, considering that he never mentions them, much less quotes them, and he himself was converted by direct revelation on the Damascus road, and further indoctrinated by conversations with Peter and others years later, and not by any writings, much less any Apostle Mark (who probably did not write the book anyway). In the end, Craig is again far overstating his case when he says “it seems nearly certain, then, that Paul believed in the empty tomb” (253).

Since there would in fact be a rhetorical reason to invent the story of the empty tomb after Christian beliefs changed to a physical concept of resurrection, we have here a reason to doubt the empty tomb story. After all, it is not implausible that the Christian beliefs could change from a spiritual to a physical conception.[WIDBRS] However, a belief in a spiritual resurrection would not necessarily exclude an empty tomb, an important point to consider when examining Craig’s argument above for Paul’s belief in an empty tomb.[WIDBRS] But the fact that the empty tomb only appears in late writings (the gospels, Mark being the first and written in the 60’s at the earliest), and the fact that Christianity was unpopular in Jerusalem and only succeeded in distant locations, also fits the theory that the empty tomb only comes up after the death of Paul, since at such a late date and at such a distance, checking the tomb was not an option. It is certainly peculiar that it would not be mentioned before. The story may have been a symbolic creation of the author of Mark (as noted above). All this throws enough doubt on the story to reject a miraculous explanation, for as long as a reasonable doubt remains, it follows that we cannot establish God as a necessary cause of the account by appealing to an empty tomb. God remains only a possible explanation, and that is insufficient for identifying a miracle by Purtill’s definition (see my discussion of Purtill’s Chapter).

Natural Explanations for an Empty Tomb

I also think there are several plausible natural explanations for an empty tomb.[WIDBRS 1 & WIDBRS 2] Craig thinks “that most alternative explanations for the empty tomb are simply incredible” (259) but I wonder how he figures that. They may be unusual, but they are certainly not beyond belief–a great many unusual things have actually happened in history. Moreover, whereas Craig claims that there are no “plausible historical explanations” which “fit the facts without bruising them” he is on quite a rhetorical high horse. All of the explanations that I provide in the book The Empty Tomb fit and do not bruise any ‘facts’ any more than Craig’s theory does. Indeed, Craig claims that “medical knowledge of the pathology of crucifixion makes the apparent death theory intrinsically improbable,” yet it is not so improbable that it could not have happened. I’ve estimated the odds to be better than 1 in 7000, and things that improbable happen all the time.[WIDBRS] Likewise, Craig says that “the disciples’ evident sincerity and willingness to be martyred for their faith is quite improbable on the assumption of the conspiracy theory,” but this is not necessarily true, either.[WIDBRS]

Finally, “grave clothes” being left in a tomb is not evidence against random theft (260), since necromancers want body parts, not clothes. Nor is this evidence against planned theft, since the clothes would be deliberately left behind. Moreover, the very mention of grave clothes is a natural embellishment to such a narrative and thus cannot be trusted as historical. In just the same way, any historian’s description of a historical scene will include plausible details that actually have no source except the historian’s imagination. For example, when battles were described in antiquity, vivid details might be given of sword blows and conversations, which the author invents–not to lie, but to paint an engaging yet plausible scene. We have to be wary of such license even in religious literature, for embellishments are especially common when one or two generations of oral tradition have intervened between the events and their written record, and when scenes are being described in vivid detail, rather than as bare historical facts.

Ultimately, Craig does not make an adequate case here for an empty tomb really being found, nor does he make an adequate case for such a find necessarily requiring a miracle. We simply have too little information, and only from sources whose reliability or sincerity, even their connection to the supposed events, are all too uncertain. Even if there was an empty tomb, we cannot be certain of it, nor can we be certain what caused it. We just don’t know.

The Mighty Circular Argument

After Craig, Habermas strains to show that 1 Corinthians 15:3-8 “meets all the demands of historical reliability” (264) and thus demonstrates that miraculous appearances really happened. But as we have seen with the other chapters, he is grossly incompetent when it comes to historical method, and his entire chapter essentially elaborates a monstrous circular argument, using dependent sources as if they were independent, and creating the appearance of abundant “corroboration” when in fact very little really exists. He then concludes with a surprisingly feeble effort at refuting the theory of hallucination. He also spends some ink attacking the idea that Jesus was indeed resurrected spiritually, and that some related psychic phenomenon explains his appearance to the disciples–something I won’t bother defending. He also dismisses the possibility of an actual survival without much argument (packing what objections he has into an endnote), but his complaints I address elsewhere.[WIDBRS] Although I do agree that this is the least likely possibility, I do not believe it is impossible, and with the meagre evidence that we have, it cannot really be ruled out–at least not with sufficient force to establish God as a necessary cause of his reported survival.

Stretching Two Sources into Nine

Habermas presents four Pauline and five non-Pauline examples of corroboration for the historicity of the physical appearances (see his table on p. 266), which really only amounts two sources, not nine. The first point is that Paul claims to have gotten the Gospel from eye-witnesses (Peter and James), although I’ve noted above that Paul shows no clear signs that what he received was a belief in a physical resurrection, so this hardly establishes such a claim. But Habermas also omits to mention the fact that Paul was converted by direct revelation three years before he even spoke to anyone among the supposed witnesses (Galatians 1:11-12, 1:16-18). This does not increase the reliability of his report, but undermines it. Historians always look for a critical mind as a crucial element in a reliable witness–Tacitus and Polybius and Thucydides are trusted in part because they demonstrate a critical mindset, and thus we can be assured that they are less likely to have been duped or led by prior expectations than other witnesses. No New Testament author displays any such mindset. And here we have a man who is convinced by a powerful emotional experience well before he even receives the formulaic gospel. Moreover, if we can explain Paul’s experience naturally,[WIDBRS] then we might do the same with the others, like Peter and James.

Ultimately, I believe it is credible that the credo of 1 Corinthians 15:3-7 was developed before Paul, but my concern is whether this actually improves its historical verity. In fact, it does not. It bothers me that Paul, a second-hand witness, is our first and only nonanonymous source that can be securely dated before 90 A.D. That we have no writings of Peter (apart from the questionable epistles in his name), or Joseph of Arimathea, or Caiaphas, or even of Jesus for that matter, throws grave doubts on whether we can really reconstruct what happened, or sift out what are actually the unique interpretations or changes or additions of Paul, whose interpretation would inevitably color all later literature, including the Gospels and Acts. Of course it is quite possible that Paul made up a lot of what he says–all of his known congregations are outside Palestine, so he could have gotten away with almost anything.[3].

Habermas’ second point is that Paul saw Jesus himself. But this is precisely the problem: he saw Jesus in what was unmistakably a vision (Galatians 1:11-12, 1 Corinthians 15:8, Acts 9:1-9, 22:1-11, 26:9-19), so this has a ready and plausible explanation in ordinary religious experience, making this a psychological rather than a divine phenomenon. For we have abundant evidence that powerful and convincing visions occur among deeply religious people, validating every religion on earth, from Santeria to worship of Asclepius, from Buddhism to New Age spiritualism. And we know that visions were a cultural force in antiquity, more common and accepted then than today (see my review of Beckwith and my discussion in The Empty Tomb, pp. 182-88). There are also historical contradictions in the accounts of his vision that leave us uncertain what actually happened, or whether we are being told the truth at all.[WIDBRS]

The third point begins to show us Habermas’ flawed methods in full relief: seventeen years after Paul began preaching the gospel, Peter, James, and John expressly told him that his gospel was accurate (Galatians 2:1-10). But we only have this report from Paul himself (it does no good to cite Acts 15:1-35, which was written well after Paul’s letters and may well have had Paul as its source), in a letter whose obvious rhetorical aim is to prove he has authority, written to a community that is a thousand miles from Jerusalem (Galatia was in the middle of modern-day Turkey). For instance, the letter begins with a peculiar emphasis on the fact that he is sent by god and not men (1:1, 1:10), he is trying to bring doubters back into the flock (1:6) and thus has a strong motivation to exaggerate or invent or at least color the truth in his favor, he is worried about people changing his system into something new (1:7), he curses twice (another peculiar emphasis) those who interpret things differently than him (1:8-9), and he goes out of his way to swear he is not lying (1:20). Anyone knows that this is often a red flag for deceit, but it at least proves he was worried that people are questioning his authority.

This is not evidence we can count on without reservation. But there is something more interesting here: to prevent doubt, he does not reiterate all the physical proofs, like the resurrection or empty tomb, but rather emphasizes his authority. This shows that belief was more encouraged by assertions of authority than by evidence. This also undermines the historical certainty of the beliefs in question. And the fourth point Habermas makes does not improve things, for Paul is the only one who reports that he was preaching the same message as the other apostles (1 Corinthians 15:11-15), so we can’t know if that’s true, since we do not have any genuine works from any of the other apostles. Even if we buy the improbable claim that some of the non-Pauline epistles were written by the actual Apostles, these still say little that allows us to determine the exact nature of their conception of the gospel. Thus, Habermas is essentially saying “Paul is telling the truth because he said so,” which is an assertion, not an argument.

Habermas commits the same error when he attempts to offer four of his five “non-Pauline” corroborations, which in fact all come from the exact same source (Acts). But you cannot use one source as if it were four independent sources! And Acts does not even count as one independent source, because it is not independent: at best it has Paul as its principle source, and is written not only long after Paul’s epistles, but even well after the first Gospel (Mark), possibly even after the Gospel of Matthew, and was written by the author of the Gospel of Luke. Habermas also claims that the repeated “Kerygmatic Creeds” in Acts are independent because they show indications of a Semitic, maybe even Aramaic background (268). But it does not matter if there are feasible theories of Semitic or Aramaic origin, since throughout the first century a common convert to Christianity was the Diaspora Jew (Paul, for example). Thus, Semitic or Aramaic origins tell us nothing about how early the creed is, nor that the creed is independent of Paul, nor what portions of it may be pre- or post- Pauline. Moreover, even if the Greek text “could” have originated in an Aramaic original, the fact is that we have no Aramaic sources of any kind to corroborate this, and that is most peculiar. Why should all the earliest Christian literature appear only in Greek?

Last but not least, it is useless to point to Acts as evidence of ‘appearances to groups’ (268) when Paul had already made such a claim decades earlier. This is thus not a corroborating account, but a secondary development from the primary source of Paul’s letters and post-Pauline oral history. There is greater trouble when we realize that the earliest versions of Mark lack the appearances, since 16:9ff. was tacked onto the end of Mark at a later date, probably in the 2nd century after the other three gospels were united with Mark in a concordance by Tatian in 160 A.D., although the text actually does not appear until the 3rd century, and then only in Coptic, while Greek texts are first found in the 4th century, around the time when Jerome composed his Latin Vulgate. Since Mark is the earliest Gospel, it is most curious that the appearances should be omitted from the earliest versions.

The fifth “non-Pauline” corroboration offered is what Habermas calls “literary evidence” from 1 Corinthians 15:3-8. Of course, if the whole Christian creed is based solely on a vague, subjective literary analysis of one text, this is not something to be proud of. Indeed, consider what this “evidence” (266) consists of:

  • “technical terms delivered and received

    Habermas gives no explanation of why this should increase the reliability of Paul’s letter. The term “delivered” (paredôka, from paradidômi: 15:3) is an extremely common Greek word, and is not even remotely technical. It means “give over” or “hand over” and was used frequently to mean deliver, betray, entrust, pass on, commend, risk, or permit, and is the most common word used to mean “pass on information.” And since Paul uses it to refer to a direct revelation from God (1 Cor. 11:23), we cannot assume it implies some sort of witness tradition passed on to him, rather than something he learned in a vision. Likewise, the term “received” (parelabete, from paralambanein: 15:1; also found in 15:3 as parelabon) is also a common word that would be natural for Paul to use in such a context–it means to learn or accept or take on. Again, Paul uses this very word to refer to direct revelations from God (Gal. 1:12). Hence we cannot assume it implies some sort of witness tradition passed on to him, rather than something he learned in a vision. So why would Paul’s use of these words make his account more reliable?

  • “parallelism and stylized account”

    Again, Habermas gives no explanation of why this would make a claim more reliable. Grammatical techniques like style and parallel structure were what every student of Greek was taught to use. Thus, that Paul should use them only shows us that he was formally educated in the standard Greek literacy of his day. It does not improve the reliability of what he says.

  • “proper names Cephas and James”

    I’m not sure why this matters, either. I have no doubt that Paul knew James and Peter. That does not tell us how reliable these men are, or even what they taught. All this does show is that Paul is familiar with the Aramaic word for “rock” (Peter is the Greek version), but why should that be surprising? As a high-ranking, well-educated Jew (a “Hebrew of Hebrews,” Philippians 3:5) from a bilingual city (Tarsus, populated by Syrians, who knew Greek but mainly spoke Aramaic), Paul certainly spoke Aramaic, even had an Aramaic name (Saul), and no doubt addressed Peter as Cephas on many occasions. If we believe Acts, then we know for a fact not only that Paul spoke Aramaic, but that he received his Damascus revelation in Aramaic (Acts 26:14). This, then, can explain the entire Aramaic origin of the credo: it comes from Paul’s vision. So this gives no support at all to the “physical” resurrection and appearances of Jesus.

  • “non-Pauline terms”

    Habermas gives no examples, nor any arguments. I have no idea which words are supposed to be “non-Pauline” or how we are supposed to establish them as non-Pauline when we have only a dozen or so epistles to draw from. Seneca wrote over a hundred letters, and he still uses some words on occasion only once throughout, thus even if we could find words here nowhere else used by Paul, that would not tell us that they were non-Pauline.

  • “triple ‘and that’ (hoti) clauses”

    To be exact, he means kai hoti, not just hoti. This is not a new element, but is actually a repeat of the second “evidence” (parallelism and stylistic composition), and is thus equally useless. Indeed, the repetition of connectives was a standard rhetorical device used for emphasis, and thus proves only that Paul intended to emphasize these doctrines. It tells us nothing about whether those doctrines are true or even where they came from.

  • “two references to scripture being fulfilled”

    This is only evidence that the credo was believed to be scripturally anticipated. But this is obvious already. So there is nothing in this that supports the reliability of the credo or that proves any particular origin (except a Jewish religious background, which is not disputed). Since Paul was a Jew and clearly conceived of the Christian creed as a reform of Judaism, we would already expect him to cite scripture in support of his beliefs or teachings. Indeed, for all we know, scripture was the actual source of those beliefs and teachings.

  • “possibly Aramaic original text (‘Cephas’, etc.)”

    This amounts to the same thing I discussed above about Peter’s name. Habermas presents no other examples of Aramaic words or influences, and even if he did it would not prove an early date since Paul and many Christians after him no doubt spoke Aramaic. Indeed, in Paul’s case, Aramaic is probably his native tongue, and thus we should expect him to slip into his native language on various occasions, and Habermas has not endeavored to show that there are more Aramaicisms in this section than elsewhere in Paul’s letters. But even demonstrating that would show us little, since it is almost certain that Paul would imagine his God speaking to him, and delivering the Gospel to him (as in the Damascus road vision), in Aramaic, not Greek (and Acts 26:14 reports that very thing). Likewise, many of the early Christians whom Paul persecuted, and from whom he would may have heard the Gospel, may not have even spoken Greek. So this again gives no support to the physicality of the appearances or of the resurrection.[4]

As you can see, this amounts to no historical evidence of any kind. Habermas asserts that “we need to judge the texts by the same criteria as those used in ancient historiography” (269). But Habermas never explains what those criteria are, and it seems that he thinks that all historians do is read and analyze literary content. He has tried to use dependent sources as independent, and tried to treat one document as if it were several, something no real historian would do. He has also not examined the actual historical problems–such as the absence of an appearance narrative in the earliest versions of Mark, or the fact that 1st century Christian converts were Jews who spoke Aramaic and may not have even spoken Greek–and thus he has exerted no historical effort in this chapter at all. All he has done is use literary analysis to discover what certain texts (Acts, Galatians, and 1 Corinthians) are saying, and what their cultural background may be. He has done nothing to show that what they say has historical support or verity. Instead, he merely cites a few historians, selectively chosen, as if we were obliged to agree with them even though we cannot check their arguments.

It’s not that Habermas could not have demonstrated some historical verity for many elements of these texts–he simply has not done so in the informed, researched manner that an actual historian would. For some of what this actually entails, see my review of Beckwith, though I discuss historical method in far more detail in my book, Sense and Goodness without God (2005), § II.3.5 (p. 57) and § IV.1.2 (pp. 227-52).


Moreover, none of the evidence Habermas cites from Paul confirms any notion that Jesus was seen in any other way than in subjective visions. And yet, tn personal correspondence to me, Steven Carr pointed out another human reality which Habermas seems to ignore:

Ask 500 Catholics after Mass if they have received the real body and blood of Jesus and they will say yes. Are they hallucinating? No. Does this mean a non-Catholic would have seen the real body and blood of Jesus during Mass? No. What people say they have seen is conditioned by what they want to say they have seen. Talk of ‘hallucinations’ is beside the point.

In other words, it is not even necessary for people who claim to have seen something to have actually seen it–it only matters that they believe it, or have any other reason to assert it. And human belief (or loyalty to a doctrine) can be secured in any number of ways, such as suggestion, indoctrination, desire, trust in authority, persuasion, even loyalty to a connected ideal, such as belief that the new creed will be a necessary good for the Jews or for all mankind. Indeed, it is possible that religious leaders like Paul and Peter would invent visions specifically for the purpose of gaining authority for their doctrines, which they could believe to be worthy enough on their own that anything should be used to persuade others to adopt them. Moreover, since belief in having encountered Jesus was (and still is) said to secure eternal life in heaven, anyone who had a strong enough desire to believe they will live forever in paradise would in turn have a strong enough motive to convince themselves that they have seen Jesus, or to voice the doctrine in the hopes that verbal assent to it might be enough to ensure God’s favor.

Nevertheless, I think it more probable that Peter and James, and certainly Paul, maybe several others, saw something that inspired their faith. I do recognize the possibility that the whole story could be an invention of Paul and others, but even in that remarkable case the behavior of Paul would make more sense if he had some kind of convincing vision (which by all accounts was not of a physical Jesus), and likewise for all other scenarios of invention.[5] I think it most likely that others had these visions earlier than Paul, and that Paul’s letters give more or less a correct version of his own experiences (such as his persecution of the early believers), albeit with all the usual rhetorical deviations that are found and expected of all autobiographical material written in antiquity. But since the only nonanonymous, first-person record of a visitation by Jesus is that of Paul, we have no historically solid ground for supposing that the earlier experiences were any different. Certainly, an experience like that of Paul would obviously not be any less convincing than a physical appearance, since we certainly see how convinced Paul was (and there are many other cases of obviously nonphysical visions being used to justify doctrinal beliefs, cf. Acts 16 and 2 Corinthians 12).

Habermas asks whether “hallucinations (or other subjective phenomena) adequately explain the facts” (271). He begins by citing some old sources on hallucination (his most recent source is a private letter to himself written in 1977 by a “clinical psychologist”) to the effect that “hallucinations are private events” and thus cannot explain how Jesus appeared to groups of people. Of course, that Jesus appeared to groups of people is the least well-attested fact in the tradition, and is a likely thing to invent for its rhetorical power–these two facts combined make the accounts suspect. Even Paul’s claim that “500” unnamed people saw Jesus at one time is curious–are we to suppose he interviewed 500 people? Indeed, is it even feasible that everyone in a gathering of 500 would be able to confirm that some person they saw was actually Jesus?

But Habermas’ source (at least the one he quotes) is not telling the truth. Although this correspondent claims that it is not possible for one person to induce hallucinations in another, it has long been known that hypnosis can in fact induce hallucinations by suggestion.[6] The power of suggestion, and the influence of a similarity in socialization and cultural expectation and background, can also contribute to groups sharing, or believing they are sharing, the same experience. Dr. Louis Jolyon West, editor of Hallucinations: Behavior, Experience, and Theory (1975)–which Habermas lists as a reference in an endnote–writes in the Encyclopedia Britannica:

If some external object is present but inadequately recognized, an incorrect perceptual engram [i.e. a stored perceptual expectation] may be activated to be experienced as an illusion; in the absence of an external stimulus, such an engram is perceived as a hallucination. This may account for the specificity of collective visions (i.e., those shared by more than one person). Among lifeboat survivors at sea, for example, several people who share similar expectancies (mental sets) may see a nonexistent ship projected against the blank screen of empty sea and sky. Such an experience may persist in some of the people even after a logical belief in its impossibility has been communicated to all.

Sarbin and Juhasz contribute a chapter to West’s anthology entitled “The Social Context ofHallucinations” where this idea is partially explored. The general consensus in psychology appears to be that the circumstantial as well as cultural or social expectations and experiential background of a group, if held in common, can lead their brains to produce similar delusions. Although this phenomenon is difficult to study, it seems that collective visions are indeed possible under the right conditions.

If this is possible in any circumstance which ancient witnesses may have found themselves in, it follows that it would have been possible to create a common hallucination–each being entirely subjective, but sharing enough features among them that all those involved think they have seen the same thing. People are also known to alter their memories as a result of social influence and suggestion, especially under the anchoring guidance of a charismatic leader or the social pressure of a group, so that even different experiences could be remembered later (even very soon after the event) as having been the same, and this can be expected to happen in situations where there is a known authoritative suggestive influence, as would be created by a group of fellow believers and by figures such as Peter.[7]

Habermas addresses only one book which presents evidence of mass delusions, and yet his critique only appears in a lengthy endnote and is not very sound (317).[8] His first objection is that since most such delusions are religious in nature, they are probably real visions of the divine and not hallucinations at all. That requires argument, and he presents none–he merely says that assuming they aren’t real is “begging the question” but so is assuming that they are real. He ought to spend much more time discussing this, and addressing all the leading literature, but he does not. Instead, he tells us virtually nothing about this one book that he claims to have read, or what evidence it presents.

He also claims that whereas the evidence shows these hallucinations only occur after “expectation” and “emotional excitement,” the appearances of Jesus do not fit this pattern. But it seems that they do, since the Gospels claim that Jesus repeatedly predicted and thus created the expectation of his resurrection. Moreover, every group experience for which we have any kind of details begins with one or two individuals seeing something and then creating the expectation in others, who then “exuberantly gathered for the explicit purpose of seeing something” (see my analysis of the Gospel accounts below). It is, of course, difficult to rely on any specific details of these stories because they contain elements of drama which may or may not come from sources other than the author’s imagination, but we have so little reliable information about any of the mass experiences we simply cannot rule out the possibility that expectation and excitement did not play a factor, and if we cannot rule that out, then we cannot establish a miracle as a necessary explanation. One thing I will add: the argument that hallucinations would not inspire radical transformations of character is absurd, since the very nature of hallucinations is such that you rarely know you are hallucinating. Because of its nature, a hallucinated experience will easily be believed real, and will thus have exactly the same effect as a real experience.

As far as individual hallucinations, there are reasons for believing that hallucination may have been a factor in the formation of the Christian beliefs. Peter Slade and Richard Bentall, in Sensory Deception: A Scientific Analysis of Hallucination (1988), survey the evidence for hallucination, and examining several studies in 1894, 1948, 1968, and 1983, found hallucination to be rather commonplace (pp. 69-71)–between 7% and 14% of those surveyed who did not exhibit any mental illness reported having experienced hallucinations, and this sample naturally did not record those who had hallucinations but did not know it. Of these identified experiences, over 8% were multi-sensory hallucinations, and 5% involved entire conversations. Surveying this and much more evidence, the authors conclude that “many more people at least have the capability to hallucinate than a strictly medical model implies should be the case” (p. 76).

Moreover, social and cultural factors can increase the frequency and acceptance of hallucinations. Of 488 societies surveyed, 62% accepted some form of hallucinated experiences as real (such as being visited by the dead, or talking to animals or trees), and the majority of these accepted experiences were not induced by drugs (p. 77). In a particularly interesting case, one study found that 40% of Hawaiian natives have reported veridical encounters and conversations with dead people, usually after violation of a tribal taboo (p. 78). This study was inspired by a few clinical cases of such hauntings, which the therapists could not cure, and in seeking a cure they investigated the cultural influences behind the experiences. After their findings, they resolved to “cure” the problem by leading the victims to engage in culturally-established atonements, which were “expected” to end the visits, and they did. Habermas would have this stand as proof that the Hawaiian native religion was genuine, but clearly he cannot have that–for if Christianity is true, then violating Hawaiian tribal taboos could hardly cause the dead to rise and chastise people. After all, why would God arrange or allow for such a vindication of their non-Christian beliefs?

The survey demonstrated another important point: visual hallucinations are rare in Western cultures, but not in many others (especially developing countries). Moreover, “the folk theory of visions and voices adopted by a culture may be important in determining whether a hallucination is viewed as veridical or as evidence of insanity” (p. 80). As a historical example, “medieval writings on insanity make few references to hallucination and instead take overt evidence of disturbed behavior (e.g. babbling, wandering aimlessly, thrashing, biting) as diagnostic of madness” and yet many medieval reports of visions which were regarded as real match modern visions reported by those with a psychotic disorder (p. 80). As the authors conclude, “we must seek the causes of hallucination, at least in part, in the social and historical environment of the hallucinator” (p. 81). When we look at the cultural situation in antiquity, we see exactly the same circumstances: hallucinations are rarely mentioned as evidence of insanity, but visions of the deceased and of gods and all sorts of other things are accepted as real (see my review of Beckwith and my discussion in The Empty Tomb, pp. 182-88, for more on the cultural status of visions in antiquity). Thus, Habermas, as a Westerner with little acquaintance with ancient culture, finds hallucinations hard to swallow–but this is only his own cultural bias leading him to the wrong conclusion. In ancient times, hallucinations were readily believed as real, and were, as in the case of the Hawaiians, far more common and culturally distinct from hallucinations in our society today.

The authors also found that “hallucinations involving bereavement” are common–and, for example, visits by the dead to the bereaved are culturally accepted as genuine in Hopi Indian culture (p. 86-8). Finally, they found evidence that hallucination plays a role in reducing anxiety, and this anxiety-relieving property in turn has a reinforcing effect on the believability and frequency of hallucination (p. 108). These two factors fit the situation of the disciples after the crucifixion incredibly well. They were primed for hallucination by their bereavement, their anxiety-filled circumstances (cf. John 20:19), their cultural predisposition to see and believe things that confirmed their deepest expectations in religious terms, and the opportunities for social influence and suggestion (as one or two individuals prepare the rest for the possibility that Jesus is risen–see below). The only better circumstances for hallucination than these are drugs, hypnosis, fatigue, hypnagogia, or sensory deprivation.

There is also the consideration of benevolent mental disorders. Claridge McCreery, in “A Study of Hallucination in Normal Subjects” (Personality and Individual Differences 2.5; November, 1996: pp. 739-747) found that there is a kind of ‘happy schizotype’ who is “a relatively well-adjusted person who is functional despite, and in some cases even because of, his or her anomalous perceptual experiences.” This brings together the concepts of hallucination as an anxiety-reducing benefit and of the socio-cultural acceptance and guidance of hallucination, and explains two other features of antiquity: why there were almost no reported cases of psychosis in antiquity (and why hallucination was not regarded as an index of insanity), and why miracles were so frequently reported in that period (not just Christian, but pagan as well). It is entirely possible that cultural support lead schizophrenics into comfortable situations where their visions were channeled into “appropriate” religious contexts.

Thus, psychotics could live entirely normal lives, even benefit from their condition by the relief of anxiety, and by social acceptance, all the while experiencing amazing things on a frequent basis. These people could then be principally responsible for the origin of miracle reports, which are then believed out of credulity, or in reaction to the sincerity of the witnesses, who would not doubt the reality of their visions, by the general populace. We would expect these “happy schizotypes” to find their most accepted place in religious spheres, and they would naturally gravitate into the entourage of miracle-workers. This would in turn explain why so many disciples have hallucinatory experiences, since there would be a higher-than-normal proportion of schizoid personalities in any wonder-working religious group. Evidence of possible past mental aberration, such as in the case of Mary who was previously possessed by demons, only confirms this further–although possession does not correspond to psychosis, Mary also is reported to have seen angels and other odd things.

In fact, the historian must explain where these “schizoids” are, since this particular class of person does not appear anywhere in the literature–in contrast, only actual schizophrenics are referred to (and regarded as insane, since they are so broken from reality that they are impaired by their visions). Even modern churches must recognize schizoidal personalities in their congregations, and other excessively religious people which evoke concern or referrals to psychiatric care, yet we find nowhere in ancient Christian literature any description of or reaction to such personalities in their midst (the demon-possessed are all so-described for their aberrant physical and verbal behaviors, not for hallucinations). The fact is that they must have been there, and this means the only explanation for their absence in the literature is that they were accepted as sane and devoted members, and may even have been given respect as particularly godly and inspired–just the sort of people who would end up in the circle of the first disciples. Habermas has not dealt with these facts at all.

Looking at the Gospel Details

Both Habermas and Craig claim that the Gospels are “lacking in typical mythical tendencies” (Habermas, 268) and “theological motifs that a late legend might be expected to incorporate are wholly lacking” (Craig, 254). But that is far from true. Matthew’s record contains many details that are fantastical and fail to be corroborated anywhere else: the earthquake and the guards at the tomb (28:2-4), the masses of living dead after the crucifixion (27:52-3), secret conversations among non-Christians (27:62-5, 28:11-5), and Herod’s killing of the babies (2:16).[9] Moreover, the constant use of scriptural references, such as in John especially, is in itself a theological motif.

Instead, the earliest account, in Mark (Paul records none of these events, not even vaguely), contradicts Matthew: there are three women, not two, and they go to anoint the body, not just “to look at the grave” (Mark 16:1 v. Matt. 28:1), and Matthew adds these fantastic details: an angel who looks like lightning descends from heaven and rolls away the stone, all right before their eyes, and an earthquake makes guards immobile with terror (28:2-4)–guards nowhere mentioned in Mark–whereas in Mark the tomb is already open when the women arrive, and there is no angel, but a man wearing white, who is sitting in the tomb–he does not descend from heaven–and this man creates the expectation in the women that Jesus will appear (Mark 16:4-7). There is nothing supernatural here. Given this version alone, it is easy to see how any later visions of Jesus have been set up with emotional expectation, even by naming the place where the visions are to be expected (16:7). And Matthew has clearly added legendary, theological motifs–angels from heaven, terrified guards, etc.–and this leaves us to wonder what else he has added or changed. Matthew clearly cannot be regarded as a reliable historical source for the first appearances of Jesus.

Examine, in turn, the added ending on Mark: in this account, Jesus first appears to one person, who has a history of psychological problems (16:9), and she then creates the expectation in others that Jesus is alive and will appear to them. John also reports that only one woman visits the tomb (20:1), and omits most of the other fantastic details, placing the two angels at a different point in time (20:12-13), but once again only one woman is there to see this vision. In his version, Mary finds only an empty tomb and runs to tell Peter that the tomb is empty, then he runs in an excited state to see for himself (20:4), and when he sees, it is he who deduces that Jesus was raised (20:8-9), an obvious origin for the creation of an expectation of appearances–and Peter’s authoritative role in the group would carry a lot of influence. Likewise, Mary later sees a vision of angels when by herself, and then mistakes a gardener for Jesus (20:14-18), easily a natural psychological event. The Thomas episode follows, but this is a story that appears nowhere else and has obvious rhetorical motivation, for it is there to counter skeptical arguments, and also other Christian groups who are preaching a spiritual resurrection–John is the last gospel to be written, and thus is most prone to adding details like this.

Then there is Luke’s account (24:1-10), which contradicts the other two: it follows Mark’s less-fantastic version, but records that there are two men, not one, in “dazzling apparel” (rather than simply wearing white). Of course, this shows the legendary embellishments growing, and Matthew simply takes this license a step further. But what is important is that these men create the expectation in the women, who in turn inspire the others to expect an appearance–Peter again runs to see, demonstrating emotional expectation (24:12). Then the appearance on the road (24:13ff.) shows obvious signs of delusion: it is not Jesus they see until much later when they “realize” the stranger is him (24:31). Then, he appears just when the expectation is being created again (24:36). There are other details here, although it is almost certain that Luke is adding embellishments (as we’ve seen by comparing Luke to Mark). There could easily be a true core story of mass delusion behind this tale.[WIDBRS]


Habermas says that of all the evidence that Jesus physically appeared after death, “particularly impressive” is the “literary evidence for the resurrection creed in 1 Corinthians 15:3ff. and the creedal confessions from Acts” (275). This is telling, because these are in fact the worst evidences in his entire chapter. Is this the best he can do? Since nothing he has presented is sufficient to establish God as a necessary cause of the events in question, he has simply failed to establish that anything miraculous was involved. Certainly, there may have been miracles here, but we are unable to recognize them. Plausible natural explanations remain to account for all of the data. Arguments about improbability are moot, since improbable things happen–so being improbable does not make something implausible. With this, the last chapter, the fact remains that this book has presented no evidence sufficient to warrant belief in any miracle, and this essentially means the book is a failure.

Return to this review’s Table of Contents to read more detailed critiques of specific chapters in In Defense of Miracles.

[1] In his rebuttal to this chapter, Is the Tomb Story Flawed because the Term ‘Rolled’ is Used?, Glenn Miller adduces some evidence that the word can be used to describe rolling nonround things, which is worth examining, but there are fatal flaws in his case. He circularly assumes his own conclusion by constantly importing unproven assumptions regarding the shape of the things he refers to. For example, he inexplicably imagines a barleycake as not being round, yet square loaves or cakes were certainly not the norm in antiquity. I am also assuming, perhaps wrongly (you be the judge), that the tomb of Jesus was not a vast monument on a royal scale, yet those were the only sort of tombs bearing round stones before 70 A.D. Likewise, I am not aware of any transport techniques like beds of logs ever being used in Judaea for moving tomb doorstones, and any such elaborate mechanism would not have been at all likely in the case of Jesus.

[2] See Richard Carrier, Jewish Law, the Burial of Jesus, and the Third Day (2002).

[3] Romans, of course, was written to a congregation in Rome; Corinthians to Corinth (in southern Greece); Galatians to Galatia, and Colossians to Colossae (both in middle Asia Minor); Ephesians to Ephesus (on the Western coast of Asia Minor); Philippians to Philippi (in Macedonia, well north of Greece); Thessalonians to Thessaly (in middle Greece); Timothy is in Ephesus and travelling to Macedonia (1 Timothy 1:3), and Paul only talks to him about congregations in Rome, Greece, or Asia Minor (2 Timothy 1:15-18, 4:10-13); Titus is in Crete (1:5) while Paul is in Nicopolis (3:12) which is north of Corinth; Hebrews does not name any recipients (though Clement’s first letter to Corinth mentions it many times, suggesting that Rome or Corinth was its audience), and only notes Paul’s location in Italy (13:24); all the other letters mention only congregations in all these same places and regions, none in Palestine (cf. Romans 15:26, 16:1ff.; 1 Corinthians 16; etc.). Paul mentions disciples in Jerusalem in Galatians, but that’s all; beyond that, he mentions “donations for the poor among the saints in Jerusalem” which he delivers (Romans 15:25-6), but this does not indicate a congregation. The point is that none of his letters are written to anyone in Palestine, and never even allude to any such congregation except when Paul is trying to prove his authority in Galatians. Hence what Paul taught might not have resembled what Peter and James were actually preaching.

[4] Craig does not mention it in this chapter, but he has his own version of the bogus “Aramaic is old” fallacy that others have pointed out to me. In his book Reasonable Faith (p. 275), Craig writes:

As E.L. Bode explains, if the empty tomb story were a late legend, it would almost certainly have been formulated in terms of the accepted and widespread third day motif. The fact that Mark uses “on the first day of the week” confirms that his tradition is very old, even antedating the third day reckoning. This fact is confirmed by the linguistic character of the phrase in question. For although “the first day of the week” is very awkward in the Greek, when translated back into Aramaic it is perfectly smooth and normal. This suggests that the empty tomb tradition reaches all the way back to the original language spoken by the first disciples themselves.

What Craig forgets to mention here is that this is the exact same language spoken by Paul and by numerous Christian converts throughout the first century, thus it does not entail an origin with the first disciples. Moreover, Craig’s contention that “on the first day of the week” is “very awkward in the Greek” is not relevant–it is a Hebraic form commonly used by Greek-speaking Jews in Hellenistic times. It was not awkward to them. Indeed, the exact same phrase, sometimes a very similar one, appears in Luke 18:12, Acts 20:7, and 1 Corinthians 16:2, none of which pertain to the resurrection. In fact, the last passage is advice Paul is giving to the Corinthians, which shows this phrase to be perfectly ordinary in written Greek among Paul’s correspondents, having no connection with some sort of “early tradition.” In fact, Mark almost certainly chose the phrase himself, because it derives from a relevant Psalm that fits his particular narrative, as I explain in The Empty Tomb (2005), pp. 155-65.

[5] Yet I do not think all invention scenarios to be impossible or even implausible, and since these authors must establish miracle as a necessary cause of the accounts, even an improbable but plausible explanation prevents miracle from being necessary, and reduces it to the status of merely possible. The question of whether miracle is more likely than all other natural explanations requires information, such as a known likelihood of miracles, that no one has. See my review of Purtill and Corduan, and also of Clark and Nash.

[6] cf. John Edgette, The Handbook of Hypnotic Phenomena in Psychotherapy (1995).

[7] There is so much literature on this, I will simply list all the most recent works: David Bjorklund, ed., False-Memory Creation in Children and Adults: Theory, Research, and Implications (2000); and Daniel Schacter and Joseph Coyle, Memory Distortion: How Minds, Brains, and Societies Reconstruct the Past (1997). For primary research on the subject see: Josh Landau, Repeated Testing of Eyewitness Memory: The Influence of Misleading Information (1993); Martha Brown, False Memory Production: Effects of Self-Consistent False Information and Motivated Cognition (1996); Lori Canfield, Predictors of Suggestibility and False Memory Production in Young Adult Women (1997); Karen Mitchell, An Examination of the Effect of Contextual Overlap on Eyewitness Suggestibility (1997); Sarah Drivdahl, The Role of Imagery and Individual Differences in the Creation of False Memories for Suggested Events (1997); Lisa Fellner, The Effects of Repeated Exposure to Suggestions: Examining the Parameters (1997); Mary Devitt, The Effects of Time and Misinformation on Memory for Complete Events (1995).

[8] Leonard Zusne and Warren Jones, Anomalistic Psychology: A Study of Extraordinary Phenomena of Behavior and Experience (1982), which has been updated since: Anomalistic Psychology: A Study of Magical Thinking, 2nd ed. (1989).

[9] That Philo, Josephus, and Tacitus would all fail to mention this, or that a war would not have been started by such an abominable atrocity, is far too incredible to believe. Moreover, the story of an evil ruler trying to kill the “chosen one” who then has to be hidden away has many parallels in other hero and king legends (Moses, Oedipus, Romulus and Remus, Cypselus of Corinth, Krishna, etc.).