This is the second of three posts covering news in the historicity-of-Jesus debate (for the first see Thallus et Alius). I recently finished reading the latest books by John Crossan and Dennis MacDonald. They inadvertently support the mythicist case with their latest arguments (despite making some weak, almost half-hearted arguments for historicity), and are worth taking note of. I don’t have time to write a full review, but here are some observations of interest to the historicity debate…
1. Undermining Historicity
Crossan’s new book, The Power of Parable: How Fiction by Jesus Became Fiction about Jesus (Harper 2012), is superb, well-written for laymen, yet astute enough for scholars. And I am not generally a fan of Crossan (even when he agrees with me, I usually find his arguments ill-informed), so this is high praise coming from me. In Power he argues that the Gospels are essentially extended parables about Jesus, based on things he said or that the authors wanted to say about him, and thus are fundamentally fiction, not history. He demonstrates that this mode of writing was well known within Judaism (and beyond) and thus not novel, and then argues how some of these parables about Jesus were invented and why. This ranks right up there with Randel Helms’ Gospel Fictions. I recommend it for all readers.
MacDonald’s new book, Two Shipwrecked Gospels: The Logoi of Jesus and Papias’s Exposition of Logia about the Lord (Society of Biblical Literarure 2012), argues that the whole Q hypothesis is demonstrably wrong, and that in fact there was a previous lost Gospel called The Logoi of Jesus [“The Words/Tales of Jesus”] that lacked a nativity, passion, or empty tomb narrative, and that was used by all the Synoptics, not just Matthew and Luke. Mark is therefore a redaction of this lost Gospel, and so are Matthew and Luke, who also incorporate and redact material added to it by Mark (such as the passion and empty tomb narratives). He also provides more than convincing evidence that Luke knew and redacted Matthew as well (and thus was not written independently).
MacDonald’s case is not entirely convincing to me. Contrary to the promise it showed when I saw earlier versions of it, he commits a number of fallacies in applying his own criteria to establish his case. I think all his evidence can be explained just as well by the standard Farrer hypothesis (see On Dispensing with Q), much more easily than MacDonald maintains. But even if, like me, you remain unconvinced by his central thesis, his book remains essential reading for experts in three respects:
(1) This is now the most recent and thorough translation and scholarly treatment of the fragments of Papias in English, and in fact anyone who wants to write or theorize about Papias simply must read this book’s chapters on him (mainly chapter 1 and Appendix 5, which provides a complete text and translation of all the fragments of Papias);
(2) MacDonald’s evidence nevertheless abundantly proves Luke’s dependence on Matthew (so anyone who wants to maintain otherwise simply must interact with this book’s evidence, in addition to that already detailed by Mark Goodacre in The Case Against Q [with its associated website and the famous article by Michael Goulder, “Is Q a Juggernaut?“]); and
(3) if MacDonald is right about the lost Gospel he recovers from Mark and Luke and Matthew, then his ancillary argument is also true, that this lost Gospel is essentially a rewrite of Deuteronomy (in fact, the Septuagint text of Deuteronomy), casting Jesus in the role of Moses and reversing or altering much of its message.
That last is a fact that does not bode well for historicity advocates, although MacDonald does not pose it that way–just like his previous work showing that Mark is a retelling of Homer casting Jesus in the role of Odysseus (in The Homeric Epics and the Gospel of Mark). These two theses together explain pretty much the entire content of the Synoptic Gospels. Once you identify the whole primary content of the Gospels to be intentional literary fiction, the Gospels go out the window as evidence for Jesus; just as Mickey Spillane’s novels go out the window as evidence for the historicity of Mike Hammer.
However, unlike Crossan’s book, MacDonald’s book is very dense and advanced. Clocking in at over 700 pages, most of it is a detailed textual commentary on the Greek text of the Gospels (and the merits of many of his arguments require a knowledge of Greek). I recommend it only for hardcore readers in Biblical studies.
2. Supporting Historicity
Both Crossan and MacDonald are aware of the danger their books pose, and thus make a conspicuous point of inserting brief, half-hearted sections arguing for the historicity of Jesus. These really serve little purpose other than to reassure their colleagues and peers that they are still rubbing the totem and thus haven’t defected to the other side…which their colleagues and peers might definitely fear after hearing out their arguments for the character of Jesus being a literary creation. I’ll address each of their arguments for historicity.
Crossan’s argument is more concise and organized, but wholly unsound (pp. 247-50). It consists of concluding historicity from two premises: one based on external evidence, the other on internal evidence. His “external” premise is the testimony of Josephus (93 AD) and Tacitus (116 AD). Of course, the testimony in Josephus is fabricated (I am certain, and can prove to a strong probability, that Josephus never mentioned Jesus, in either passage where he now appears), and the testimony in Tacitus most obviously comes from Christians deriving their claims from the Gospels, which Crossan just got done arguing are fiction.
But even Josephus’ testimony, if granted authenticity, derives from the Gospels (this has been demonstrated for the main passage by G.J. Goldberg, in “The Coincidences of the Testimonium of Josephus and the Emmaus Narrative of Luke,” The Journal for the Study of the Pseudepigrapha 13 : 59-77; the allusion to James later on can just as easily be an invention of Christian legend gullibly reported by Josephus, but I demonstrate that in fact it is an accidental interpolation in Richard Carrier, “Origen, Eusebius, and the Accidental Interpolation in Josephus, Jewish Antiquities 20.200,” Journal of Early Christian Studies 20.4 [Winter 2012]).
Therefore, even at their best, the material in Tacitus and Josephus cannot logically “corroborate” the Gospels. Crossan’s external premise therefore falls.
The internal case, Crossan says, begins with the question, “If you are inventing a nonhistorical figure, why invent one you cannot live with, but must steadily and terminally change into its opposite?” (p. 250), a question he can’t answer so he finds it “much more likely that Jesus was an actual historical figure whose radical insistence on nonviolent distributive justice was both accepted and negated by the tradition it engendered.” But this is a non sequitur. If Jesus could start a movement with such radical ideas, then anyone could. And they could do that by inventing a Savior character (“Jesus” essentially means “Savior”) as its celestial or mythic mascot. Then, as the movement grew and these radical ideas became less popular (and more obviously unlivable), the movement had to start changing its central message, and thus its central character.
Therefore the fact that this happened (even granting that it did; not all scholars buy Crossan’s version of the original Jesus) provides zero evidence that it started with a historical Jesus. Because that same evidence is just as expected if Christianity began with (let’s say) a historical Peter touting a revealed Jesus telling him these same radical things. And then just as Paul got rid of Peter’s Torah-insistent Jesus, so did later generations gradually get rid of Paul’s egalitarian Jesus, eventually ending up with John’s Jesus, the most uncharitable asshole version of Jesus ever devised…until someone invented the Infancy Gospels (which finally turned Jesus into that horrid boy from The Omen).
Therefore, the fact that Christian ideas changed (and Christians changed their stories of Jesus to match) does not support historicity any more than nonhistoricity. Crossan’s internal premise therefore falls.
Crossan therefore has no logically sound case for historicity to offer.
MacDonald’s argument is a bit of a shambles but at least gets more into the evidence (pp. 543-53). He starts with the claim that Paul attests to his “extensive interaction with Jesus’ family and followers” (p. 543), although that begs every question (whether these people were his followers and actual family, as opposed to a metaphorical brotherhood), and even MacDonald is forced to admit that somehow “little of this information” [that Paul should have gotten from them] “has seeped into Paul’s letters,” a curiosity readily explained by there being no such information (Jesus being solely a revealed figure; after all, Paul himself seems to know of no other).
MacDonald then says his reconstructed “lost Gospel” (the Logoi) had accurate knowledge of Galilee and came from a bilingual (Semitized) environment, which are mutually consistent with a Galilean origin. But each is a non sequitur–as is the whole inference “Galilean author => historical Jesus” (as if Galilean authors can’t invent; and yet MacDonald himself admits the Logoi is substantially fiction).
MacDonald acknowledges that’s a problem. But if such an argument were pressed, it is just a cascade of fallacies: first, even granting the premises, the conclusion of historicity doesn’t follow; and second, the premises themselves are already questionable. Even granting that the Logoi does show a reliable knowledge of Galilee (which is more debatable than MacDonald claims), that was available to almost anyone (especially from Palestine or Syria or Arabia or Egypt or anywhere in the Diaspora, where pilgrims and people familiar with the Holy Land were always available). Likewise, Semitic loanwords are not uniquely indicative even of Palestine, much less Galilee (see Proving History, pp. 185-86, with my remarks on Ehrman [A] and [B]), a point with which, notably, MacDonald agrees (p. 544).
MacDonald thus does not lean on this, realizing it’s a weak thread. He instead admits that even in his reconstructed Logoi genuine facts about Jesus are “buried” and “difficult to mine,” requiring the standard “criteria” to extract, a method I and others have already thoroughly refuted–so he is still behind the times here. In addition to my book Proving History, which cites many other scholars agreeing with me on this, see the latest salvo: Jesus, Criteria, and the Demise of Authenticity (T&T Clark 2012). MacDonald’s attempts to apply these criteria fail in all the ways I already document in Proving History (e.g., the claim that the authors would not invent a baptism by John the Baptist is refuted on pp. 145-48; and so go all other such arguments, e.g., pp. 155-57).
Likewise, in any sound Jesus Myth theory the “sayings” of Jesus (or the ideas that inspired their formulation, p. 545) would have come by revelation, and therefore proving their originality cannot establish the historicity of Jesus (PH, pp. 123-24) even if their originality could be proved, which it rarely can, owing to our tremendous ignorance of first century Judaism (PH, 129-34). Because again, if Jesus could innovate, anyone could. Therefore innovation is not evidence of Jesus. Confusingly, yet again, MacDonald confesses this is true, concluding “Because of the author’s debt to rhetorical invention [in the Logoi], it is impossible to attribute any of these sayings with confidence to the earthly Jesus” (p. 552). So he knows no case can really proceed from this argument, either.
His next argument is that all the Gospels contain incidental details that serve no literary or religious purpose, and those must surely then be historical. For example:
There is no reason to challenge the accuracy of the following information: Jesus’ home was in Nazareth of Galilee; he traveled to Judea, was baptized by John, returned to Galilee, conducted a ministry in [specifically named] towns and villages there…and traveled with several male disciples; he was considered a teacher, exorcist, and wonder worker…met hostility from Torah-observant Jews, and was crucified by Romans with encouragement of the Jewish authorities in Jerusalem. … [B]ecause it is not religiously weighted, [this summary] probably reflects reliable traditions about him.
He makes no argument either for these being “not religiously weighted” or for that being any indication of their being historical. So already the argument is logically unsound, for having premises nowhere established. In fact, the premises are insupportable.
MacDonald’s claim that there is no religious purpose for any of these details is false. Nazareth and the baptism by John have identifiable origins in mythmaking (see Proving History, pp. 142-48); Christian missionaries themselves “met hostility from Torah-observant Jews” and were teachers, exorcists, and wonder workers, and thus needed a mythic model to follow and validate those roles (see PH, p. 174); fiction would be no more likely to attach female disciples to Jesus than actual history would; and the whole Romans-encouraged-by-Jews crucifixion narrative (which MacDonald admits originates with Mark, being absent from his reconstructed Logoi) is unintelligible as history (see PH, pp. 139-41). That leaves just one element: situating the narrative in Galilee. But any fictional narrative has to be situated somewhere.
In actual fact, incidental but geographically and culturally accurate details (like the names of towns in an area) naturally accumulate in myths and legends, as demonstrated by Jan Harold Brunvand, in his famous study of urban legends, The Vanishing Hitchhiker: American Urban Legends and Their Meanings (1989). He found that in fact the more specific and incidental details there are in a strange story, the more likely it is to be non-historical. That’s right. Because storytellers add and accumulate just those kinds of things; the story doesn’t feel true without them, so they get added (but to a legend-hunter like Brunvand, they’re a dead giveaway).
There is no rhyme or reason for the details added, other than that they add verisimilitude (they are thus called “validating details”), except for one thing: they do tend to match a specific geographically localized “color,” wherever the story is crystallized, which can often be in different places for different versions of the legend. It is thus notable that, just like urban legends, a completely different version of the Jesus narrative arose, placing Jesus’ execution, by Jewish stoning, in Lydda and dying a hundred years earlier. Even the Gospel of Peter has Jesus crucified by Herod, not the Romans. So the story could be told in all manner of ways. Just like any other legend. For perspective, just look at all the “incidental details” in the many Apocryphal Acts…yet no one would claim those details must then be historical. Those are works of total fiction.
The fact is, the Christians had a scriptural reason to set their messianic fable in Galilee: Isaiah 9:1-7 says to. And Jesus being a “Nazorian”(Nazôraios) provided an obvious basis for putting him in the Galilean town with the nearest-sounding name: Nazareth (Nazareth). Even though those words are not in fact related. So MacDonald’s argument here does not hold up. He recognizes this is true for many other arguments, pointing out, for example, that there are far fewer independent sources than most scholars claim (so “multiple attestation” is not an effective criterion: see PH, pp. 172-75; note how Ehrman tried, illogically, to make exactly the opposite claim: see [A] and [B]), and that multiple attestation doesn’t confirm historicity anyway (p. 551), and that the Gospels are full of examples of “the expansion of sayings [of Jesus] into narratives” and thus they routinely fabricate stories (note that I have pointed out that this was a typical way of fabricating biographies generally, a fact well established in classical studies).
MacDonald then cites Josephus. Which really should be a non-starter as an argument (as I already noted above). He even exposes why when he admits any attempt to reconstruct a “genuine” Testimonium Flavianum can only be hypothetical (p. 547), and it should be obvious how that negates it as evidence (because a hypothetical premise only gets you a hypothetical conclusion: garbage in, garbage out; likewise, even if authentic, the supposition that any of its content derives independently of the Gospels can only be hypothetical–and in fact, as I noted above, is most improbable).
MacDonald even admits that the phrase “the one called Christ” in the second Josephan passage about a certain executed James “may be another Christian gloss” (p. 548). He tries to salvage it anyway by saying no other Jesus could be meant, but as there are many men named Jesus in Josephus, and indeed one in particular in this very narrative (the high priest Jesus ben Damneus, who is obviously the Jesus here meant), that argument is simply invalid. My peer reviewed demonstration of that (in JECS, cited above) will appear this Winter.
Lastly, MacDonald cites corroboration in Paul of some few details and sayings, but we already know that sayings of Jesus came to Paul by revelation (likewise to other apostles of his generation), and the only other details Paul gives of Jesus are unwitnessed mythic facts that Paul fully confesses are derived from scripture (e.g., 1 Cor. 15:3-4) or revelation (e.g., 1 Cor. 11:23; cf. Gal. 1:8-12 and Rom. 16:25-26). In other words, without addressing alternative hypotheses for these details in Paul, no argument for historicity can logically proceed from them. Only a proper comparative argument can attain logical validity here. MacDonald gives none.
And that’s the sum of his case. And it’s probably, really, the best case anyone could make (although it could be organized better). Perhaps you can see why more and more of us are finding historicity doubtful.