“Jesus never really existed as a historical person. He originated as a mythical character in tales symbolically narrating the salvific acts of a divine being who never walked the earth. Later this myth was mistaken for history, or deliberately repackaged that way, and then embellished over time… The odds Jesus existed are less than 1 in 12,000 [.008%]. Which to a historian is for all practical purposes a probability of zero. Even when I entertain the most generous estimates possible, I find I cannot by any stretch of the imagination believe the probability Jesus existed is better than 1 in 3 [32%].” (Richard Carrier, On the Historicity of Jesus: Why We Might Have Reason to Doubt, pp xi, 600)
“As a prophet of the Arabs who taught a vaguely defined monotheism, Muhammad may have existed. But beyond that, his life story is lost in the mists of legend, like those of Robin Hood and Macbeth. As the prophet of Islam, who received (or even claimed to receive) the Qur’an, Muhammad almost certainly did not exist… Muhammad the messenger of Allah came into existence only after the Arab Empire was firmly entrenched and casting about for a political theology to anchor and unify it. Muhammad and the Qur’an cemented the power of the Umayyad Caliphate and then that of the Abbasid Caliphate. That is the most persuasive explanation for why they were created at all.” (Robert Spencer, Did Muhammad Exist? An Inquiry into Islam’s Obscure Origins, pp 214-215)
I used to dismiss mythicists as agenda-driven cranks, and frankly many of them still are. The two authors under review are different, however, and seem to have been leery about mythicism before becoming advocates for it. In his decades-long study of the Qur’an and hadiths, Robert Spencer was aware of how unreliable the sources are; yet he concluded that Muhammad probably existed (Did Muhammad Exist?, pp 6-7). Richard Carrier associated Christ-myth theorists with crackpots until finally examining the issue (On the Historicity of Jesus, pp 1-3). A mythicist camp isn’t exactly something you have a burning desire to join if you care about your credibility.
Both Spencer and Carrier are credible advocates, and I should address certain objections that have been, and might still be, raised against them before examining what they have to say about Muhammad and Jesus, respectively.
1. Neither Spencer nor Carrier have degrees in Qur’an or Biblical Studies. Spencer has an MA in religious studies from the University of North Carolina at Chapel Hill, and Carrier holds a PhD in ancient history from Columbia University. But the claims of Spencer and Carrier stand or fall like any claims: on their merits, not on the kind of degrees that back them up. Amateurs can be right and experts can be wrong (Secret Mark a case in point, exposed as a hoax by two non-specialists independently of each other); being tied to a guild can bring its own baggage. Credentials obviously matter in plenty of ways (if you want a teaching position at a respectable institution you’d better have the right kind), but in general, “pulling rank” on someone by appealing to a superior degree is no argument at all. It’s a sign of insecurity, in some cases even incompetence.
2. Spencer and Carrier are controversial figures. Spencer is a political conservative who runs the Jihad Watch blog, and is viewed by some as an Islamophobe. Claiming that Islam is inherently militant — that jihadists aren’t perverting the Qur’an but rather understanding it too well — doesn’t invite a warm reception these days. But Spencer gets way more flak than he deserves. I’ve explained how political correctness impedes our understanding of religions, and Spencer is to be commended in this regard. Carrier is an atheist crusader for “sense and goodness without God”, urging that the religion of Christianity is a delusion. But he displays enough competence in the field of early Christianity. Yes, he blunders from time to time, but he also gets plenty right where it counts.
3. Spencer and Carrier can be off-putting. They generally avoid ad hominem attacks (unlike some of their critics, BTW), though not always. Ridiculing one’s sexuality or looks (Spencer: “the metrosexual Reza Aslan has no good arguments”) is poor form, as is calling the mental state of other scholars into question (Carrier: “I suspect Maurice Casey is insane”). It’s no surprise that people find Spencer and Carrier to be unrewarding conversation partners.
These elements are irrelevant and ignored in what follows. Did Muhammad Exist? and On the Historicity of Jesus are not only well researched and fairly argued, they show that mythicist positions are defensible in the right hands. I haven’t been moved to either conclusion, but I’ve been moved to greater doubt, and to revise some of my historicist judgments.
I’ll start with Spencer’s book, since it’s the more straightforward. Simply put, we don’t have any early sources, let alone multiple or independent ones, to build a case for Muhammad’s existence. There’s no mention of Muhammad or the Qur’an by the Arabs who were supposed to have been inspired by them in the first six decades. That’s sixty years of complete silence — a long time. In the eighth and ninth centuries we get stories of Muhammad in sudden immense detail, but in the early expansion following the 630s, the prophet and holy book are completely unheard of. They’re absent from surviving Arab communications, absent from the writings of those they conquered, absent period. These Arabs are said to have come and laid waste, but they are not called Muslims, just Hagarenes, Ishmaelites, and Saracens.
On coins from this period, we do find the word “Muhammad” inscribed, but the inscription comes under kingly figures bearing a cross, which is obviously a symbol of Christianity (Did Muhammad Exist?, pp 43-4). “Muhammad” can mean “the chosen/praised one,” and so the coins could be conveying the idea that the ruler is praised or chosen in God’s name (p 45). Alternatively, they could be referring to Jesus — at a time when the religion of the Arab conquerors was still a vague monotheism — or a proto-Muhammad figure unlike the latter-day messenger of Allah. Even the inscriptions on the Dome of the Rock (completed in 691 AD and often thought to be the first elaborations on traditional Islamic theology) could be referring to Jesus, explaining how he, “Muhammad”, is a mere messenger and not divine as orthodox Christianity held (pp 56-7).
The reconstruction is simple: The earliest Arab rulers were Hagarenes, monotheists whose chief heroes were Abraham and Ishmael, and they were Christian-friendly enough that they minted coins with crosses on them. It was in 691+ that Islam as we know it began to emerge in defiance of this Hagarene faith and Christianity, amongst the Umayyads (who ruled from 661-750). The Dome of the Rock’s inscription referring to the “praised one” could no longer refer to Jesus, but a new militant prophet. The Umayyads were supplanted by the Abbasids in 750, who regarded their predecessors as irreligious, and who began massively rewriting the past with the Qur’an and hadiths, and filling in historical gaps, demonizing the Umayyads, who in turn created their own hadiths blaming the Abbasids.
Given the 60-year deafening silence and lack of early sources, the only defense for Muhammad’s existence can come by using the criterion of embarrassment in the late sources. Spencer finds this problematic, and I will return to the issue at the end. Richard Carrier finds the criterion equally useless in assessing the figure of Jesus, and in this they both agree with an increasing number of experts.
In the case of Jesus, we don’t have a 60-year stretch of silence. Nor even 40 years (when the gospels start appearing). The letters of Paul give us a 20-year window, and so the question hinges on what Paul tells us. Carrier argues that Paul and other apostles worshiped a purely mythical figure into which the gospels later pumped historical life.
In other words, there was never a man named Jesus who acquired followers in his life, and who was executed, or believed/claimed to be executed, which in turn led to his status as a divine Christ (On the Historicity of Jesus, pp 33-34). That’s how Carrier defines the minimal historicist position.
Instead, Jesus was originally thought to be a celestial deity, who communicated with his subjects through dreams and visions. Like other celestial deities, he was at first claimed to have gone through an ordeal of incarnation, death, burial and resurrection in the supernatural realm. Also like other celestial deities, he was later placed on earth, in history, as a divine man, with an earthly family, friends, and enemies, complete with deeds and sayings (p 53). That’s how Carrier defines the minimal mythicist position. (The Ascension of Isaiah paints a picture like this: a preexistent divine Jesus descending below the moon to be killed by sky demons in outer space.)
As a result, there’s a constant dialectic running through Carrier’s work, as he examines evidence to see whether it lends credence to the minimal historicist position or the minimal mythicist position. He assesses gospels and epistles alike, but the real weight hinges on Paul:
“When we approach the Epistles of Paul we must look at each passage with the assumption that mythicism is true and then estimate how likely it would be that the passage would look like that. And then look at that same passage with the assumption that historicity is true and do the same. The latter probability may even be high. But is it as high as the probability on mythicism? (Or higher?) That’s the only question that logically matters?” (p 514)
This brings us to Carriers methodology. By use of Bayes’ Theorem, Carrier estimates the odds Jesus existed less than 1 in 12,000 (.008%). That’s what he genuinely believes the odds are. Playing devil’s advocate — that is, being as generous as possible to the historicist camp — he estimates a 1 in 3 (32%) likelihood. This is all based on his own series of judgments made throughout the course of his book, and as he noted in his prior volume, Proving History, the formula is only as good as what you put into it. Needless to say, the reliability of his .008% – 32% spectrum depends on how one accepts his judgments.
For the most part he doesn’t become victim of his aggressive claims. So for instance, in his assessment of the gospels, he finds nothing at all which can verify Jesus’ existence, but also nothing which proves mythicism. “As evidence, the gospels simply make no difference to the equation.” (p 509) That might be objectionable, but he’s not stacking the deck in his favor as I expected.
He grounds the Christian movement in the phenomenon of cargo cults (pp 159-163) against the backdrop of Roman occupation (pp 153-159), which itself I take to be accurate. If the Jewish people couldn’t retake control of their temple by armed revolt, some sectarians could conceive a replacement of it altogether, which Carrier sees as effected by Christ’s atoning death:
“The centrality of the temple was a continual problem for the Jews. A physical location requiring political control entailed military domination. So long as the Romans had the latter, the Jews would never have the former. The Zealots took the logical option of attempting to remove the Romans and restore Jewish control. The Christians took the only other available option: removing the temple from their entire soteriological scheme. Christians could then just await God’s wrath to come from heaven, while in the meantime, God’s promise could be delivered unto the kingdom they had spiritually created, first in an anticipatory way, and then in the most final way (the apocalypse).” (pp 158-159)
Carrier notes that many cargo cults worship saviors or follow messiahs that never existed, which is true, though the opposite is also true. Historicists like Dale Allison have also framed Christian origins as a cargo cult response to the socio-political realities of ancient Palestine (see Millenarian Prophet, pp 78-94).
But back to Paul, who gets the key penultimate chapter. It’s well presented, and many of Carrier’s arguments give you pause if you forget you know anything about the New Testament outside Paul’s authentic letters; but he’s not always convincing.
For instance, he rightly dismisses the historicist evasion that Paul doesn’t mention Jesus’ earthly life because he didn’t care about it (pp 517-518). That doesn’t necessarily count as a strike against the historicist position, however. Paul did care about the historical Jesus, but that figure was no help to him, sometimes against him, and so he had to avoid mentioning him. Carrier says that “letters about persons almost always contain historical references to them” (p 523) — but not if those references are a problem. Not if that historical person was wrong about the issue at hand. The reason what Jesus said and did in life isn’t relevant in Gal 1-2 (as Carrier objects, p 526) is that Jesus never spoke against circumcision, as Peter and James would have obviously known and already thrown in his face. Paul needed an anti-circumcision gospel (on the basis of a continually delayed parousia), which only the (convenient) heavenly Christ could provide. All of this explains why Paul shuns the term “disciple”, which Carrier makes much of (p 524). There is nothing improbable about an apostle who never knew Jesus, and was at loggerheads with those who did, and who wanted to avoid any reference to his earthly business.
With regards to the eucharist, Carrier finds that Paul’s use of the Lord’s Supper is “too easily explained on minimal mythicism and too unusual and sparsely detailed on minimal historicity, producing at best a 50/50 fit either way. It thus argues for neither historicity nor mythicism.” (pp 562-563) That conclusion is fair enough, but more needs saying about the way Paul understood Jesus’ death.
When you weigh all of Paul’s Jesus-death metaphors, the scales tip in favor of minimal historicity. Carrier says (pp 143-145) that early Christians saw Jesus’ death has having replaced the sacrifices of both Passover and the Day of Atonement (passover is implied by the eucharist account, and Yom Kippur by Rom 3:24-26), which I take as correct, but they also saw him as a martyr. Some of the texts Carrier cites (like I Cor 15:3) evoke martyrdom more than passover/atonement. The catalog of texts include I Cor 8:11, I Cor 15:3, II Cor 5:15 (x2), Rom 5:6-8 (x2), Rom 14:9, Gal 2:20-21, I Thess 5:9-10. “Christ dying for our sins” parrots the “X dying for Y” standard (see Jeffrey Gibson’s “Paul’s Dying Formula”), signalling real flesh-and-blood mortals who died so that others could follow their example. Jews believed that copying a martyr gained victory over a tyrant (IV Macc 1:11; 18:4), and pagans thought copying a philosopher gained victory over fortune (Seneca, Ad Lucilium Epistulae Morales 24:4); likewise, Paul thought copying Christ gained victory over sin and death (Rom 6:1-11; 8:10).
If martyrdom isn’t the most important death metaphor for Paul, it is the one he most frequently invokes. Jesus is martyr (I Cor 8:11, I Cor 15:3, II Cor 5:15 (x2), Rom 5:6-8 (x2), Rom 14:9, Gal 2:20-21, I Thess 5:9-10), paschal lamb (I Cor 5:7, I Cor 11:25), mercy seat of faith (Rom 3:25), sin-bearer/scapegoat (Gal 3:13, II Cor 5:21, Rom 6:6, Rom 7:4, Rom 8:3), and redeemer (I Cor 6:20, 7:23) all in one, and those metaphors are conceptually at odds with one another. It’s easy to understand a cargo cult leader who martyred himself and then later became deified on the cultic level, based on the weight of these death metaphors. That’s what happened elsewhere (on which see David Seeley’s The Noble Death): The Maccabeans were exemplary, but their blood also served as “an atoning sacrifice” (IV Macc. 17:21-22); the blood of Thrasea’s model suicide was sprinkled on the ground as a libation to the gods (Tacitus, Annals 16:35).
Carrier covers the noble death theme on pp 209-211, and gives it good treatment (especially noting how the concept wasn’t embarrassing: “the more awful and shameful the manner death, the more heroic and powerful it was”), but he doesn’t register the full implications. It’s at least as likely (and I think slightly more so, again based on the distribution of death metaphors in the Pauline corpus) that Jesus’ historical martyrdom took on a heavenly atoning function, than that the starting point was a celestial atoning Christ.
Finally, a word about Gal 1:19, which in Carrier’s view is “the only real evidence” historicists have from Paul’s letters. He argues that James “the brother of the Lord” is fictive kinship language rather than biological, as Paul wants to distinguish Christians generally from apostles specifically (p 590), which means of course that he’s not referring to James the pillar. It’s not a convincing argument. Paul would have little reason to bring up a lesser non-apostolic James in the context Gal 1-2, as such a figure would be beneath mentioning. Paul is referring to the apostle James who in fact is the biological brother of Jesus, and who has to be acknowledged, because he’s a thorn in Paul’s side being a rival authority in the Antioch incident as it now bears on the Galatian situation.
This was a point made by Zeba Crook in his recent debate with Carrier, to which Carrier later responded online:
“Crook claimed Paul ‘wished’ James wasn’t the brother of Jesus (because that made James a greater authority than Paul). There is no indication of that anywhere in the Epistles, at all (this is the same error I caught Mark Goodacre in). That is a Christian faith doctrine, that Crook has sublimated from having been taught ‘mainstream assumptions’ in his field inherited by its progenitors, who were not analyzing the evidence objectively in the first place.”
Carrier is being a bit obtuse here. No one, least of all Crook and Goodacre, is leaning on Christian faith doctrine; this is a scholarly construct based on objective assessments of Paul’s relationship to James and the other pillars. Even if you know nothing of Acts 15, it’s not hard to see the power struggles implied in Gal 1-2. (As an aside, I even suggest that James used his authority treacherously.) This is a feeble swipe on Carrier’s part and one of his least persuasive arguments.
But even though Carrier thinks the kinship language of Gal 1:19 (and I Cor 9:5) is twice as likely on the mythicist assumption, he allows that it also might be twice as likely on the historicist assumption (p 592). So again, fair ball.
The criterion of embarrassment has undergone something of a reassessment in the past decade. Other criteria (like dissimilarity) have been long acknowledged to be flawed in premise, leaving embarrassment as the trump card. If you have no early sources to work with (in the case of Muhammad), or if your early sources are stingy and oblique (in the case of Jesus), then the way to salvage late narratives is with the luring promise of scandal: that which mocks your beliefs, or undermines your credibility, or somehow makes you look bad, can’t have been invented by you, because you don’t cut your throat. You don’t embarrass yourself unless you are forced to acknowledge, perhaps even defend, something that really happened. Thus Muhammad falling under the power of black magic; thus Jesus baptized by the inferior John, implying he had sins that needed washing away.
The principle is fair, the application problematic. Mainstream scholars like Dale Allison and Mark Goodacre would both agree with Spencer and Carrier that (1) what appears to be embarrassing to us often wasn’t (Spencer, Did Muhammad Exist?, pp 111-117; Carrier, Proving History, pp 129-134); (2) even material that was embarrassing may not have been so at an earlier stage, when it could have been invented (Carrier, pp 126-128); (3) by virtue of the fact that it’s there at all, how embarrassing could the material really have been? the ancient authors had creative license to omit whatever bothered them (Carrier, pp 134-137).
The first two points are valid, the third is somewhat lame. The gospel/hadith writers had creative license, but they didn’t write in a vacuum, and could be kept in check by entrenched traditions. Beliefs remain cherished despite nervousness owing to other evolving beliefs, and there are cases where the apologetic process is obvious. Even Spencer concedes the Satanic verses were embarrassing (that Muhammad received revelations from the devil), and a genuine case from the gospels would be Jesus’ mistaken prophecy about the apocalypse.
The problem involves point (2). Embarrassment at a late stage doesn’t imply the same for an earlier one. Jesus’ promise that some disciples standing next to him wouldn’t die before the apocalypse could have been created in a first-generation church that was getting impatient for his return (as its members were starting to die off). Mk 9:1 would have then served as an “assurance” text (somewhat like I Thess 4:13-18 or I Cor 15:51-53), the message being, “Don’t worry, Jesus is coming again and some of you will still be alive when it happens.” Only at the point when all first followers died off would the saying become embarrassing.
For detailed examples of where the criterion succeeds and fails in the hadiths and gospels, I refer the reader to my post, The Embarrassing Existence of Muhammad and Jesus, but the upshot is that I agree it has very limited value. There aren’t many examples where embarrassment is obvious; in cases where it is, it implies the material probably originated earlier, but where on the 40-60 year spectrum is hard to estimate. On whole, I take it that scholars like Allison and Goodacre, and now Spencer and Carrier, have struck a significant blow against the criterion’s utility. In the few cases where it applies, it could have some impact on Carrier’s Bayesian probabilities pertaining to the gospel narratives; not much, granted, but some.
Did Muhammad Exist? and On the Historicity of Jesus represent levelheaded arguments for mythicism. Their authors may have controversial personas, but that shouldn’t be confused with whacky theorizing.
It isn’t far-fetched to suppose that Muhammad never existed from a 60-year stretch of silence. If not for a couple of sticking points, I could be moved to Spencer’s conclusion.
Nor is it a crackpot theory that Jesus began as an apostolic fantasy until historicized 40 years later. Especially if you deny the existence of Q, as I dislike admitting that I do, and allow that the meager amount of early evidence — Paul’s eight letters, possibly Hebrews — evince high Christology and do little to hint at an historical Jesus. I believe Paul does this more than Carrier grants, but not so as to leave me supremely confident.
Scholars haven’t taken mythicism seriously enough to respond to it appropriately, and frankly you can’t blame them. But I think it’s time to recognize mythicism as a viable alternative. Bart Ehrman’s Did Jesus Exist? has some good points, but on whole it’s a rather superficial rebuttal to mythicism; Maurice Casey’s Jesus: Evidence and Argument Or Mythicist Myths? is dire. (The scholarship of Ehrman and Casey are usually impressive.) Critiques of Spencer’s case against Muhammad have been worse.
I enjoy seeing Bayes’ Theorem in action; it’s a wonderful integration of math and the social sciences. The ’90s showed that any Jesus theory can fit the evidence, and Carrier’s Bayesian strategy starts by asking how likely mythicism is to start with; then asking how probable all the evidence is on its minimal version; then also how probable it is on historicism’s minimal version. Some of Carrier’s judgments are objectionable, but not greasy; I never felt he was abusing our trust. He grounds early Christianity in the proper framework of cargo cults. He dismisses testimony that was never strong to begin with, like Josephus. He pegs the likelihood Jesus existed at a putrid .008%, but also concedes as high as 32% to his opponents. I’m not sure what my spectrum would end up looking like (it would be higher), but it would be an interesting project to work on at some point.
For anyone who teaches courses on the historical Jesus, I stand by my previous list of recommendations, though I have added Carrier’s work to it (I’d already included Arnal’s The Symbolic Jesus, which is a hop away from mythicism). If I were an instructor, I might use Allison’s trilogy and Carrier’s duology, to provide students with the best of what historicism and mythicism have to offer.
UPDATE: Richard Carrier replies.