“What Did Josephus Mean by That?” A Case Study in the Relationship between Evidence and Probability

I recently engaged a friendly debate with Dennis MacDonald on the MythVision podcast regarding the historicity of Jesus. Much of which was the usual stuff expected of a serious, expert, secular discussion of that question (follow the link to watch). I might blog about it in future as well, as I did last time in Is Jesus Wholly or Only Partly a Myth? The Carrier-MacDonald Exchange, setting aside just one issue I want to explore here today instead: MacDonald’s strange theory about what Josephus thought he was telling us in Jewish Antiquities 20.200, that being the place where our manuscripts today have Josephus telling us a story about a certain “James the brother of Jesus (who was called Christ).”

My goal here is to kill two birds with one stone: first, to put a long, singular digression into its own separate treatment, explaining why MacDonald is probably wrong about this one obscure passage; and second, in the process, illustrate something he and I also discussed: how surrounding evidence relates to determining the probabilities of different explanations of a specific piece of evidence. In this case, the evidence to be explained (the explanandum) is the text of Josephus: everything he wrote encompassing this single story (which occupies § 200-203 in the 20th and final book of the Antiquities). I will describe two competing hypotheses to explain this evidence, mine and MacDonald’s, and then how, methodologically, we determine which is the more probable (and therefore the explanans). The same principles exhibited in this one instance can (indeed must) be generalized to any and all competing claims about history.

Setting Up the Competing Hypotheses

I’ve covered this subject before, not only here (see Josephus on Jesus? Why You Can’t Cite Opinions Before 2014 and Mason on Josephus on James, as well as More Asscrankery from Tim O’Neill), but also at Biblical studies conferences, as well as under formal peer review (in “Origen, Eusebius, and the Accidental Interpolation in Josephus, Jewish Antiquities 20.200” in the Journal of Early Christian Studies 20.4, reproduced within Hitler Homer Bible Christ; and in “Josephus and the Testimonia Flaviana” in On the Historicity of Jesus, pp. 332–342). As you’ll see throughout, I’ve adduced considerable evidence that the phrase “who was called Christ” was never in this text until an accidental scribal interpolation placed it there in the mid-to-late 3rd century (Josephus wrote in the late 1st century).

Once we accept that, the point Josephus was making with this story becomes quite clear—and it has nothing to do with Christians. The High Priest Ananus the Younger illegally convened a court to execute some rivals, including a certain James, and was then punished for that crime by being replaced in the office of High Priest by the very brother of the James he killed: Jesus ben Damneus—in Greek, “Jesus [the son] of Damneus.” Indeed, when “called Christ” was accidentally inserted into the text (probably from an interlinear note deriving from Origen), the phrase meaning “son of Damneus” may well have been replaced by it, by a scribe believing it to be a duplication-error from a few sentences later when we learn how Ananus was punished: by being replaced by the very same “Jesus son of Damneus.” But even were that not the case, and Josephus forgot to include the patronymic (which means a “father’s name”) in the first instance (which he did sometimes do) or it dropped out of the text in copying (which was also a common occurrence in manuscript transmission), it would still be obvious to anyone reading this story what was going on here: Josephus is explaining why Jesus ben Damneus ascended to the High Priesthood, by showing how Ananus’s attack on the Damneus family backfired into exactly that result. Simply remove the words “called Christ” and there is no other way to read this story; and indeed it’s just how any ancient reader of it would have understood it.

But Dennis MacDonald argues that there is evidence in the text of this passage that Josephus was talking about Christians, and not the family of Damneus—and in particular, that this is a story about James being executed for preaching the abandonment of the Torah law (or in some radical fashion “changing” that law), which would be too coincidentally evocative of Christianity to mean anything else. MacDonald insists there are peculiarities in the Greek that indicate this, and that surely Josephus intended us to understand a back-reference to some previous passage about Christianity—a passage MacDonald supposes to be lost, as he agrees the Testimonium Flavianum (the only explicit passage about Christianity in the extant manuscripts of Josephus, which is currently in the 18th book of the Antiquities) is entirely a fake.

How We Determine a Prior Probability

Let’s suppose these are the only two possible explanations of the James passage worth considering, such that the sum of any and all other explanations lies below even 1% likely at the start, and therefore are so unlikely they won’t even show up in our math (if we always round off any probability to the nearest whole percent, and accept a margin of error of, say, +/-5%). And let’s suppose we start without considering any evidence regarding which of our theories is the more likely—in which case, by definition, each starts out equally likely. The entire probability space of 100%, which thus includes all possible explanations of the content of this passage (such that “there is a 100% chance that one of those explanations is true”), is thus divided equally: so my theory is 50% likely; and his theory is 50% likely. After rounding off the tiny fraction of that probability space occupied by every other possibility (being below even 1%, of course, as it then dissolves within our margin of error).

Then let’s ask what happens when we start letting some information affect these probabilities. Prior probabilities are supposed to be the probability of a hypothesis being true given all human background knowledge—everything we know about human nature, the laws of causality, how evidence usually gets generated or destroyed; and everything about the ancient world, ancient Greek; and so on. So these “priors” do not just arise from thin air. We are not ignorant of all those things, so the prior probability distribution for our competing theories does not sit at “50/50” as I just supposed.

One of the things that can quickly ruin a prior is a theory requiring a bunch of assumptions not in evidence. For every thing you have to suppose to be true (as opposed to prove true with evidence confirming that it is reasonably likely, if not indelibly certain), the improbability of that thing, whatever it is, must necessarily reduce your theory’s prior probability—relative to any theory that does not require that supposition. And that’s why you can’t just stack up a bunch of “if, if, ifs” to force your theory to fit the evidence. Every one of those ifs reduces the probability of your theory, by exactly as much as those ifs are improbable. Unless you can adduce evidence that they are not mere suppositions but probably true. And of course that evidence can’t be your theory itself; and it can’t circularly be the evidence you are trying to explain. You need some independent evidence that it is the case that what you suppose to be true is indeed true.

Now, every theory requires some set of suppositions not in evidence; so the only ones that matter are the ones that make a difference between two theories. If our theories each start out equally likely, and each depend on the same set of suppositions, our theories remain equally likely. Any reduction in probability faced by one, is faced by the other, such that in the final equation there is no difference between them. And in that case, we cannot say which of them is true, and thus we cannot use the evidence to be explained as evidence for either theory. So what matters are suppositions another theory doesn’t require, and that aren’t matched in that other theory by some other supposition that’s just as unlikely. Because every one of those you adopt, will cause your theory to become less likely—but doesn’t cause the other theory to become less likely. And when that happens, the other theory starts to far outstrip yours in probability—and all before we even test our theories against the evidence to be explained.

Prior probabilities also change according to reference class, because the latter represents information that affects the probability of a claim. For example, if you say a light in the sky is an alien spaceship and I say it is Venus, we have an extensive database of similar lights being sighted throughout history, from which we can derive an actual observable frequency with which lights like that turn out to be Venus, and how often they turn out to be alien spaceships. If the latter frequency is zero, that does not mean the prior probability that it’s an alien spaceship is zero; but it does mean it is low. It will follow Laplace’s Rule of Succession: (s+1)/(n+2). If a thousand previous times similar lights turned out to be Venus, and never an alien spacecraft, at best the prior probability it even could be the latter is (0+1)/(1000+2) = 1/1002—in other words, less than one in a thousand chance.

It only gets worse when we then also consider the effect of stacking unevidenced assumptions. Whereas we have vast credible evidence confirming Venus exists and “looks like that,” we have literally no credible evidence confirming aliens visit the Earth, much less look like that when they do. So “aliens” as a hypothesis starts with a supposition that is extremely improbable just on background knowledge alone. And on top of that, there is also a vast amount of evidence rendering it quite unlikely that there even would be aliens visiting Earth. So “aliens” requires several highly improbable suppositions, which in turn reduce the probability of that hypothesis even further—before we even get to asking how well it explains what we are observing. That would not be the case if alien space visitors were a well-established fact. For example, in the Star Trek universe, that a light in the sky could be a visiting alien ship would be quite likely, the exact opposite of our actual circumstances now. This is the role played by background information in affecting the prior probability of our hypotheses. All of this, above and below, I cover in detail in my peer reviewed methodological treatise Proving History.

Now let’s dial this analogy back to something more realistic. In Proving History I outline the following example of the role played by background knowledge in constraining the probability of any explanation we might be proposing:

In our personal correspondence, C. B. McCullagh [said] … “consider how the hypothesis that Henry planned to kill William II in order to seize his throne explains the fact that after his death Henry quickly seized the royal treasure. The relation between these events is rational, not a matter of frequency.” But, in fact, if the connection alleged is rational, then by definition it is a matter of frequency, entailed by a hypothetical reference class of comparable scenarios.

To say [that] is rational is thus identical to saying that in any set of relevantly similar circumstances, most by far will exhibit the same relation. If we didn’t believe that … then the proposed inference wouldn’t be rational. … [Hence] the evidence in this case is that Henry not only seized the royal treasure with unusual rapidity, but that his succeeding at this would have required considerable preparations before William’s death, and such preparations entail foreknowledge of that death.

Already to say Henry seized the royal treasure “with unusual rapidity” is a plain statement of frequency, for unusual = infrequent, and this statement of frequency is either well-founded or else irrational to maintain. And if that frequency is irrational to maintain, we are not warranted in saying anything was unusual about it.

Likewise, saying “it would have required considerable preparations” amounts to saying that in any hypothetical set of scenarios in all other respects identical, successful acquisition of the treasure so quickly will be infrequent, and thus improbable, unless prior preparations had been made (in fact, if it is claimed such success would have been impossible without those preparations, that amounts to saying no member of the reference class will contain a successful outcome except members that include preparations). Again, the result is said to be unusual without such preparations, or even impossible; and unusual = infrequent, while impossible = a frequency of zero.

Hence such a claim to frequency must already be defensible or it must be abandoned. Similarly for every other inference: making preparations in advance of an unexpected death is inherently improbable for anyone not privy to a conspiracy to arrange that death, and being privy to such a conspiracy is improbable for anyone not actually part of that conspiracy, and in each case we have again a frequency: we are literally saying that in all cases of foreknowing an otherwise unpredicted death, most of those cases will involve prior knowledge of a planned murder, and in all cases of having foreknowledge of a planned murder, few will involve people not part of that plan. If those frequency statements are unsustainable, so are the inferences that depend on them. And so on down the line.

Proving History, pp. 273-74

The point being: here we are making assumptions about human nature and causal reality, but they are not mere suppositions; they are well-evidenced beliefs, and thus as probable as anything can be. Based on all we know about human beings and the laws of causality, it is all but 100% certain that humans cannot have highly convenient foreknowledge of exactly when someone will die—except when they have arranged that death (or somehow found out about someone else’s having arranged it, but if that arranged death benefits the person we are talking about, rarely will that be a coincidence). It is important that we respect what we have good reason to expect to be the case, based on vast amounts of experience with how humans behave, how they argue, how evidence comes to exist, when it is likely to and when it is unlikely to, and so on.

Here, McCullagh is saying his suppositions are not improbable, but in fact as probable as can be, and in fact they entail the conclusion: Henry probably arranged William’s death. By contrast, suppose someone intent on not accepting such an awful realization tried to propose against McCullagh the following argument: “Maybe Henry had been developing contingency plans for the hypothetical possibility of William’s death, and just got really lucky, such that William’s death occurred at exactly a convenient time that allowed Henry to quickly recruit assistance in carrying out those contingency plans.” The first assumption here can be called the theory itself, but to explain the evidence (to make that evidence “likely”) it requires the second assumption, which is by definition improbable. Because any assertion of remarkably convenient luck simply is asserting that something improbable happened. McCullagh’s theory has no need of such improbabilities; it thus starts out far more probable, does it not?

And lest someone try to evoke the “fine tuning” argument here, remember the existence of a remarkable God requires even more luck than random fine tuning would; it is thus not like McCullagh’s theory, which rests on known facts about human nature, not hypothesized incredible superbeings. And this is precisely my point: mere assumptions carry with them improbabilities, improbabilities you must account for in the final probability of what you are proposing; but assumptions that are actually established facts (such as about how humans can come to know things, what humans must do to arrange things, and so on) do not generate substantial improbabilities to account for.

So now, if we consider McCullagh’s opponent’s theory, even if we deemed such luck to be as likely as a 50/50 shot, that still means that theory starts out half as likely as McCullagh’s (because 50/50 = 50% and 50% of any value halves that value), because none of McCullagh’s assumptions are even improbable, much less equally so. So if both theories started equally likely, then McCullagh’s opponent’s theory would now occupy 25% of the prior probability space, while McCullagh’s theory fills out the remaining 75%. McCullagh’s opponent has immediately rendered his hypothesis three times less likely than McCullagh’s, simply by positing something not in evidence in order to “explain away” the evidence that’s inconvenient for their theory. And in truth, Henry just being “remarkably lucky” somehow is actually much less likely than even that. If to explain what happened you need a lucky coincidence for Henry that’s, say, 20 to 1 against, then you have to divide your theory’s prior probability not by half but by 20. McCullagh’s opponent’s theory would instead go from 50% to (0.5)/(20) = 0.025, a mere 2.5% chance of being true. McCullagh’s theory by comparison shoots up to being 97.5% certain. And this is before we have even looked at the specific evidence for either theory.

To turn that around, McCullagh’s opponent needs to come up with some pretty darned good evidence for “contingency planning” and “luck” having been the cause of the conjunction of events (that being William’s death followed almost simultaneously by Henry’s seizure of the royal treasury), before we as historians can believe such an apologetic. Without that evidence, that remarkable sequence of events is simply improbable—unless Henry killed William. Which would mean he probably did. So rebutting McCullagh’s theory cannot be accomplished by just making stuff up. Stacking up alternative suppositions to explain the evidence does not logically work; you need evidence for those suppositions.

Thus a more apt rebuttal to McCullagh’s theory would be to say that given all we know (everything about the speed of a mounted horse, the ability to call allies to one’s aid unexpectedly, Henry’s cultivating of alliances in every corner of William’s court, actual evidence of Henry repeatedly seizing opportunities with Machiavellian haste on other occasions, and so on) it is not improbable that Henry, surprised by William’s death, could immediately do the math and head straight for the treasury and hastily secure it even without specific planning. Here, none of the things we are supposing is “just being supposed.” Rather, it is all well-evidenced in human nature, political systems, Henry’s particular history and circumstances, and so on, and thus every one of these “assumptions” is so probably true we needn’t even ask the probability of its not being true. This is how actual background evidence, not mere suppositions, wins a theoretical dispute.

Our Prior Probabilities

So let’s look at what MacDonald’s theory requires: that James the brother of Jesus Christ taught the abolition of Torah law or some radical set of changes to it; that Ananus cared so much about that as to have him illegally executed; and that the rest of the Jewish elite were not supportive of that but in fact outraged by it and punished Ananus for it. Already little of this makes logical sense. The Jewish elite persecuted Christians; they didn’t persecute their persecutors. And the only prominent James we know about was part of the pro-Torah faction of Christians, not the Pauline faction who argued for abandoning strict Torah observance. There is no evidence whatever that this (or indeed any) James was teaching radical deviations from Torah law. All evidence we do have suggests he would have been against the Christian factions who did that.

So right from the start, MacDonald’s hypothesis requires inventing ad hoc suppositions, nowhere in evidence, that this brother of Jesus had taken or switched sides with the anti-Torah Pauline sect, and was advocating it so ardently in Judea that the Sanhedrin stoned him for it—an event nowhere recorded, not even in the book of Acts, which mentions no James the brother of Jesus at all, and which only mentions the known Torah-faction James being beheaded, not stoned, and by King Agrippa, not the Sanhedrin (much less Ananus), nor for this reason, and this action being approved by the elite, rather than (as Josephus’s story has it), so ardently opposed by the Jewish elite as to result in Royal and Roman intervention in deposing a High Priest who instigated the murder.

So MacDonald’s theory of this passage requires improbable supposition after improbable supposition even before we get to testing it against the evidence for it. Logically those improbabilities commute to the conclusion. If your theory requires two claims that are each at best 50% likely to be true, their total probability is not 50% but (0.5)(0.5) = 0.25 or 25%. This means your conclusion also must be reduced by that same amount. If we suppose before considering further evidence that each of our hypotheses started out equally likely, then my theory starts out 50% likely to be true, and his starts out 50% likely to be true. But MacDonald requires two suppositions that are not inherently likely, certainly not evidenced to be any more likely than false, which means his theory drops from 50% likely to, at best, (0.5)(0.25) = 0.125 or just 12.5%. Since I require no such suppositions (nor, as we’ll see in a moment, any suppositions comparably unlikely), my theory’s prior probability suffers no such reduction. Right out of the gate, my theory starts out already several times more likely than MacDonald’s. This is the cost of “requiring things to be true” before your theory can even function to explain the evidence—and there is no evidence at all that any of those suppositions is true, or even more likely than not.

We have no evidence any brother of Jesus, much less one named James, joined the anti-Torah faction of Christianity, at all, much less was killed for it. And indeed, that is inherently unlikely. It is also inherently unlikely that Paul, for example, would not have made extensive use of the fact that Jesus’s own brother sided with him; instead, we have no mention of any such oddity in the Epistles. It is also inherently unlikely that a brother of Jesus would have defected from the original teachings of Jesus. Sure, that’s not impossible. But it’s also not something we can say we should expect to have happened—at all, much less happen and there never be anywhere any mention of it ever happening, and it having no visible effect whatever on any of the debates Paul had with the first Apostles or the Torah-observant faction whom he often reports stood against him, and it having no visible effect whatever on the entire history of the church contrived in Acts.

Though now we are getting into how evidence affects relative likelihoods and thus the final probability of our respective theories; more on that shortly. But even before we get to the peculiarities in that other evidence (of what is and isn’t in the Epistles, Acts, other Christian writing about this James and his fate, and so on), already MacDonald’s theory is asking us to believe a ton of things every one of which goes against all expected probability. As there is no evidence for any of them being true, they each can only at best be 50% likely (“as likely as not”); but as each one goes against expected probabilities regarding how human beings behave and how remarkable events and alliances tend to cause observable impacts on the surviving evidence of disputes and developments—such that their not doing so is by definition not probable—we end up with assumptions even less likely than “50/50”; but again that gets to the impact of the evidence on our respective theories, which I’ll look at next.

By contrast, what does my theory require? Actually, really, nothing. Just what Josephus said, and the absence of the phrase “called Christ,” which I have elsewhere shown with evidence is almost a certainty—and thus not a mere assumption. As a mere assumption it would have a probability of at best 1 in 200 (see Hitler Homer Bible Christ, pp. 370-71); but as a well-argued fact, it has a probability well enough above 50% to consider a given. And as the evidence establishes a corruption in the text, we can no longer argue against my thesis on assumptions about what the text originally said. Did it originally say “the brother of Jesus the son of Damneus”? We don’t know. Yet I still adduced evidence that that kind of scribal replacement frequently happened. Or did Josephus assume that association would be obvious from mentioning it three sentences later? We don’t know. Yet I still have adduced evidence Josephus sometimes did that; moreover, we already know he was inclined to do that here, as he gives no patronymic for James either, nor even for the Jesus called Christ if such he wrote.

The answer one might give to that latter point is that by saying this is the Jesus called Christ, that was all the identification Josephus need give. But that’s actually false—as we will see when we start looking at how the specific evidence does not support that supposition at all, but in fact deeply undermines it. Ultimately, we can get no improbable suppositions to be necessary for my theory. It leans solely on what is in the text as we have it, plus the abundant evidence establishing the interpolation, and what would have been obvious to Josephus’s readers or commonplace in scribal practice. It even leans on how Josephus is well known to have composed stories.

And that latter point will become clear when we examine the evidence below. For purposes of relative prior probability, however, one might say Josephus only occasionally didn’t give patronymics at the first mention of a person, so perhaps we should assign his doing that here a lower prior probability, as being an assumption. But that he didn’t originally give the patronymic here is already itself an unevidenced assumption negating that very argument, and that he had less need to than elsewhere is evident in the text already (as it is given three sentences later, and in such a way as to cap the point of the entire story, thus explaining why it is even being told this way). But even more to the last point, it is even less likely that Josephus would substitute omitting a patronymic for inserting a mere two words “called Christ” than that he would fail to explain why he was doing that. And to that point we must now turn.

Josephan Style

As you’ll see when I quote the passage next, in this very same passage Josephus explicitly refers the reader to his previous discussion of Sadducees, so as to explain the actions of Ananus; yet he doesn’t think to refer to any discussion of Christians, or why Jesus was called Christ, or why anyone wanted to kill these people, much less so as to explain the fate of James. It is a well established fact of his consistent style that Josephus would have explicitly referred to his past discussion of this “Christ” (a word otherwise never attested in such a sense in Josephus and that would have been wholly obscure to most Gentile readers), so as to explain any of what is going on here. That he doesn’t, means there was no such passage to refer to.

And that means Josephus cannot have written “the one called Christ,” as that would be a far more obscure thing to say than what he says about Sadducees, yet in this very same story he gives both an explanation of the relevance of “Sudducees” being mentioned and a back reference to his previous discussion of the point. Christianity would have far more required both here, being the far more obscure, particularly in explaining any of this story (why are they being killed? why is the elite opposed to their being killed? indeed, so opposed they actually have the Romans and King Agrippa depose their own High Priest for it? what does a “Christ” mean and what does that explain about this story? what does James have to do with any of this? and so on). Which means it is very unlikely that Josephus would give us both for “Sadducees” and neither for “Christ.” The evidence therefore does not support but in fact wholly opposes any proposition that Josephus was counting the phrase “called Christ” as an explanation and a back reference. It is neither. To the contrary, it is itself the very thing that would require both. Just as the word “Sadducee” did.

So known background facts of Josephan style refute the notion that “called Christ” somehow “is” a back reference to some (evidently lost) other passage, a passage that conveniently—and here again MacDonald stacks unevidenced assumptions—explained James being a Christian, why some Jews persecuted Christians, what laws they broke, why most of the Jewish elite sided with Christians as this story requires, and so on, none of which is in the current Testimonium. That’s the difference between a supposition lacking any evidence rendering it probable and empirical facts rendering a supposition highly improbable, and thus unusable in any argument for any conclusion about this passage. And this is how evidence relates to establishing the plausibility of any claim about history, as with this claim here in particular. “Called Christ” simply isn’t a back reference or an explanation; rather, it is what would require such. So we cannot resort to that excuse here. It simply doesn’t apply. Evidence rules it out.

Without unevidenced suppositions, what we have is simply what Josephus gives us: the specific violations James was accused of were wholly uninteresting and unimportant to him, because he never mentions them. The only thing Josephus stylistically shows he was interested in is the mere fact that James shouldn’t have been killed, that it was too trivial a technicality of Jewish law to warrant assembling an illegal court for, a conclusion with which the Jewish elite all agreed—a strange thing if the Jews are supposed to be persecuting, even trying to outlaw or expel or even kill, Christians (as Paul attests in his Epistles and Acts depicts in its narrative of early Christianity). So we cannot assume Josephus had some ulterior unstated interest here. His stated interest is clear: Ananus should not have illegally killed James; and for that crime he was punished by being replaced in office by James’s brother. That is what Josephus depicts. That is his story. No suppositions need be added to see any of this. It’s all plainly what is depicted. The only thing I am supposing is the hypothesis itself, that this is what he meant, which explains literally everything in the text. MacDonald, by contrast, requires an additional stack of half a dozen assumptions even to get to his hypothesis, not a single one attested in this passage, or backed by any prior evidential probability.

If Josephus had written this passage as about the persecution of Christians, he would have explained things, as is his style consistently in all his historical writing; only a Christian scribe emending the text this way would just assume all those obscure things were already known to the reader—like what a “Christ” was; that James was a Christian; that Jews sought to kill Christians; or even why the Jewish elite and Roman authorities opposed the killing of James if he was a Christian. On the Jewish side, the Christians often touted their “self-guilt” as motivating such contrition, exhibiting divine knowledge that what they were doing was wrong; and on the Roman side, the Christians often touted their “right wisdom” in opposing Jewish villainy and actually supporting the Christians (the consistent narrative of Acts and many other apologetic narratives afterward). But only Christians would have “assumed” all these things were obvious. Josephus absolutely would not. It’s not even likely he’d have known of such notions and ideas; and certainly no evidence he did; and, more to the point, there’s absolutely no way he would “assume” all his readers did.

This is how we know the line “called Christ” was never written by Josephus. It is not how he would compose this passage, at all, if such were his intended meaning. All evidence regarding his stylistic practices and narrative discourse techniques rule against it. By contrast, once we accept the interpolation, this is exactly how we would expect Josephus to have described Ananus being replaced by Jesus ben Damneus. And that is plainly what he is describing. At most one might have expected Josephus to “explicitly” say this; but there is no evidence that in a mere one paragraph account of priestly succession he would feel the need to. To anyone not distracted by what he didn’t write here (the simple phrase “called Christ”), the intent of this story is amply self-evident, requiring no further exposition. Only one Jesus is in it: Jesus ben Damneus. Only one event is described: the murder of Jesus’s brother. And only one consequence is related: the murderer is replaced by the very Jesus whose brother he killed.

Evidence in Respect to the Carrier Theory

There is no evidence against my theory of the text. It is entirely supported by what Josephus says and by other actual evidence establishing what he didn’t say (for example, as the phrase “called Christ” is empirically more likely an interpolation than original, it no longer functions as evidence against my theory). And it is supported by or comports with all evidence regarding Josephan style and scribal practice. Likewise the evidence for my theory of the text is abundant, as I demonstrate in the peer reviewed journal article reproduced in Hitler Homer Bible Christ:

  • Origen had no knowledge this passage was about Christians—despite his explicit search of Josephus for material relating to Jesus in his answer to Celsus. This is quite improbable on MacDonald’s thesis, yet 100% expected on mine.
  • Instead, Origen mistook a story about James in Hegesippus as being in Josephus (a kind of mistake I document Origen sometimes made), and in fact that, our only actual Christian narrative of the stoning of James the brother of the Lord, even in the few details repeated by Origen, does not agree with any detail whatever in this account in Josephus. This is quite improbable on MacDonald’s thesis, yet 100% expected on mine.
  • We know Acts used Josephus as a source text for historical color, yet the author of Acts never noticed this passage as being about Jesus Christ. Which is inexplicable, given that if it was, then it shows Jews being punished for persecuting Christians, exactly the kind of thing the author of Acts strove to include; instead, Acts never mentions this James even being martyred, much less seizes on the rhetorical coup such a Josephan account would have provided. This is quite improbable on MacDonald’s thesis, yet 100% expected on mine.
  • As I already explained, if Josephus had written this passage about the persecution of Christians, he would have explained that, as is his style consistently in all his historical writing; only a Christian scribe would assume those half dozen obscure things are already known to a reader. But Josephus wouldn’t, yet doesn’t give any explanation, not even a back reference to one. This is quite improbable on MacDonald’s thesis, yet 100% expected on mine.
  • Finally, the words tou legomenou christou are more typically Christian (appearing both in the New Testament and in Origen, even in unrelated contexts) than Josephan. That Josephus would contrive such a construction is not as improbable as any of the preceding evidence, but it is still less probable on MacDonald’s thesis than on mine, being that it is exactly what we expect to find in a text corrupted by contact with the work or thought of Origen (as my evidence shows this was), whereas it is not exactly what we’d expect from the hand of Josephus (who otherwise assiduously avoids ever calling any of the many other messianic pretenders he describes a “Christ,” i.e. “Anointed,” i.e. messiah).

That’s a pretty strong case for my being right about what Josephus meant. That’s five improbable things. Even if we drop the last (as being less improbable and thus less significant in shifting probabilities), that leaves four quite improbable things in the evidence on MacDonald’s theory, each of which is, by contrast, exactly what we expect on my theory. We found his theory already starts with a dismally low prior probability, because it requires a larger stack of improbable assumptions than mine does. I calculated that even being very favorable to his assumptions, more even than human experience should warrant, we end up with a prior probability of merely 12.5% for his theory, entailing approximately 87.5% for my theory. But let’s suppose we ignored all that (even though we methodologically never should) and simply gave each of us a 50/50 prior again. Or let’s even allow a middle-ground distribution of 25% to 75%. What then do we end up with when we consider the evidence I just listed?

I cannot reasonably believe any of the four most weighty facts I listed have even so much as a 1 in 4 (or 25%) chance of being true on MacDonald’s thesis; and that’s being extraordinarily generous, even accounting for any mutual dependency they might have. Whereas they are all 100% expected on my thesis. So each one of those things is four times more likely on my theory than his. Being four such items together, that’s 1/4 x 1/4 x 1/4 x 1/4 = 1/256: the evidence I listed is therefore in combination well over two hundred times more likely—as in, more expected—on my theory of what Josephus meant, than on MacDonald’s.

If we start with a prior probability distribution of 12.5%/87.5% or 1/7, we get a final odds of 1/7 x 1/256 = 1/1792, meaning after all evidence is considered, and all background knowledge about Josephus, his audience in the 90s A.D., human nature generally, and so on, my theory of what Josephus meant is nearly two thousand times more likely to be true than MacDonald’s, and well over a thousand times more likely. If we are even more generous, and assume a prior probability distribution of 25/75, which is 1/3, then we get a final odds of 1/768; and if we absurdly ignore all the unevidenced assumptions MacDonald relies on, and pretend they have no effect on his theory’s plausibility, we would start at 50/50, or 1/1, ending up with a final odds of 1/1 x 1/256 = 1/256. Either way, even then, my reading of Josephus is still hundreds of times more likely.

MacDonald’s Evidence

MacDonald mostly had no evidence, but rather a system of unevidenced assumptions that he treated like evidence, which is invalid reasoning. Unevidenced assumptions reduce, they do not increase, the probability of a hypothesis that requires them. They therefore are the exact opposite of “evidence for” a theory. But I’ve already explained that point. What actual evidence did MacDonald try to add? Really only one thing: a claim about the specificity or oddity of Josephus’s choice of words in Greek. In Bayesian terms, MacDonald means that there is something odd about the Greek that is improbable on my theory of this passage but very probable on his. That is at least a formally valid argument—if the premise is true (that the probabilities really do diverge that way). But that wouldn’t even be enough; not only does this one item of evidence have to be more probable on his theory than mine, but to overcome all the evidence I just presented, it would have to be thousands of times more probable.

That’s some extremely strong evidence. It would require an effective certainty, not any ambiguity. So if the Greek MacDonald is talking about is at all in fact ambiguous, it could not possibly be that unlikely on my reading of the text—so even if it were unlikely at all, it wouldn’t be unlikely enough to establish his theory is even close to as probable as mine, much less more so. And yet I’m pretty sure the Greek of Josephus is not even unlikely on my theory, as MacDonald avers—much less so very much more so as his theory requires.

Here is the text in Greek (which I extracted from the industry’s digital standard, the TLG):

(199) ὁ δὲ νεώτερος Ἄνανος, ὃν τὴν ἀρχιερωσύνην ἔφαμεν εἰληφέναι, θρασὺς ἦν τὸν τρόπον καὶ τολμητὴς διαφερόντως, αἵρεσιν δὲ μετῄει τὴν Σαδδουκαίων, οἵπερ εἰσὶ περὶ τὰς κρίσεις ὠμοὶ παρὰ πάντας τοὺς Ἰουδαίους, καθὼς ἤδη δεδηλώκαμεν. (200) ἅτε δὴ οὖν τοιοῦτος ὢν ὁ Ἄνανος, νομίσας ἔχειν καιρὸν ἐπιτήδειον διὰ τὸ τεθνάναι μὲν Φῆστον, Ἀλβῖνον δ’ ἔτι κατὰ τὴν ὁδὸν ὑπάρχειν, καθίζει συνέδριον κριτῶν καὶ παραγαγὼν εἰς αὐτὸ τὸν ἀδελφὸν Ἰησοῦ τοῦ λεγομένου Χριστοῦ, Ἰάκωβος ὄνομα αὐτῷ, καί τινας ἑτέρους, ὡς παρανομησάντων κατηγορίαν ποιησάμενος (201) παρέδωκε λευσθησομένους. ὅσοι δὲ ἐδόκουν ἐπιεικέστατοι τῶν κατὰ τὴν πόλιν εἶναι καὶ περὶ τοὺς νόμους ἀκριβεῖς βαρέως ἤνεγκαν ἐπὶ τούτῳ καὶ πέμπουσιν πρὸς τὸν βασιλέα κρύφα παρακαλοῦντες αὐτὸν ἐπιστεῖλαι τῷ Ἀνάνῳ μηκέτι τοιαῦτα πράσσειν· μηδὲ γὰρ τὸ (202) πρῶτον ὀρθῶς αὐτὸν πεποιηκέναι. τινὲς δ’ αὐτῶν καὶ τὸν Ἀλβῖνον ὑπαντιάζουσιν ἀπὸ τῆς Ἀλεξανδρείας ὁδοιποροῦντα καὶ διδάσκουσιν, ὡς οὐκ ἐξὸν ἦν Ἀνάνῳ χωρὶς τῆς ἐκείνου γνώμης καθίσαι συνέδριον. (203) Ἀλβῖνος δὲ πεισθεὶς τοῖς λεγομένοις γράφει μετ’ ὀργῆς τῷ Ἀνάνῳ λήψεσθαι παρ’ αὐτοῦ δίκας ἀπειλῶν. καὶ ὁ βασιλεὺς Ἀγρίππας διὰ τοῦτο τὴν Ἀρχιερωσύνην ἀφελόμενος αὐτὸν ἄρξαντα μῆνας τρεῖς Ἰησοῦν τὸν τοῦ Δαμναίου κατέστησεν.

The old Whiston translation renders this as (emphasis mine):

[The] younger Ananus, who, as we have told you already, took the high priesthood, was a bold man in his temper, and very insolent; he was also of the sect of the Sadducees, who are very rigid in judging offenders, above all the rest of the Jews, as we have already observed; when, therefore, Ananus was of this disposition, he thought he had now a proper opportunity. [The Roman prefect] Festus was now dead, and Albinus [his replacement] was but upon the road; so [Ananus] assembled the Sanhedrin of judges, and brought before them the brother of Jesus, who was called Christ, whose name was James, and some others; and when he had formed an accusation against them as breakers of the law, he delivered them to be stoned: but as for those who seemed the most equitable of the citizens, and such as were the most uneasy at the breach of the laws, they disliked what was done; they also sent to the king [Agrippa], desiring him to send to Ananus that he should act so no more, for that what he had already done was not to be justified; nay, some of them went also to meet Albinus, as he was upon his journey from Alexandria, and informed him that it was not lawful for Ananus to assemble a Sanhedrin without his consent. Whereupon Albinus complied with what they said, and wrote in anger to Ananus, and threatened that he would bring him to punishment for what he had done; on which king Agrippa took the high priesthood from him, when he had ruled but three months, and made Jesus, the son of Damneus, High Priest.

Notice the explanation and back reference for the term Sadducees, explaining the relevance of bringing that term up here. No such explanation or back reference is given for the term Christ or its pertinence to the story here. Notice Josephus says James and gang were killed “as breakers of the law,” in other words simply for breaking the law, not for preaching its abandonment or radical revision or for preaching anything at all. What they may have taught simply isn’t relevant to the story—had it been, Josephus would have said so. Such is his established style as an author and a historian.

The contended points then follow: Whiston renders Josephus as saying that the citizens who were “the most equitable” and “the most uneasy at the breach of the laws” are the ones who “disliked” that Ananus did this, believing it “was not to be justified,” and thus sought his punishment. What did they dislike about it? What was unjustified about it? What did they want him punished for? Josephus tells us, as per his usual style of narrating: “it was not lawful for Ananus to assemble a Sanhedrin” without Roman consent (Judea was at this time under direct Roman rule). That’s it. No mention of MacDonald’s theory that they disliked something else about it, some obscure thing nowhere stated here. No other thing they disliked about it is ever described by Josephus. We therefore ought not assume he meant anything else than what he said they disliked about it. Likewise when Josephus says Ananus was punished, the only punishment he then lists is Ananus being replaced by Jesus. Why Jesus? Well, that’s what this entire story explains. Right? Otherwise, where else is Josephus’s explanation?

This is where we see MacDonald’s theory stands on unevidenced assumptions nowhere in and even contrary to the evidence, in the end reducing its probability to at best hundreds to one odds against being what Josephus meant. My theory, by contrast, employs no assumptions of the kind; instead, it relies solely on what is actually in the text (and what we can prove probably wasn’t)—in other words, my theory of what Josephus said is simply what Josephus actually said. No more. No less. That’s why our priors diverge. Then we see that his theory leaves a lot of evidence highly improbable (from Acts, Origen, Hegesippus, and the content of this passage in light of Josephus’s established discourse style), all of which evidence is instead highly probable on my theory. That’s why our likelihoods diverge. And the two combined (priors and likelihoods) leads to a posterior probability far in favor of my reading of Josephus—and far against MacDonald’s. By over a thousand to one, or at least hundreds to one, even with estimated probabilities far in MacDonald’s favor.

Against all of this, MacDonald claims there is something about the underlying Greek of the phrases “seemed the most equitable” and “most uneasy at the breach of the laws” (and maybe “disliked” or “not to be justified” or perhaps even the general remark about the Sadducees being “very rigid in judging offenders”) that renders my theory less probable than his. What in Heaven or Earth could that be? Here is my own direct translation of each of these phrases, hyperlinked where useful to the Perseus digital edition of the Liddell & Scott Lexicon:

  • οἵπερ εἰσὶ περὶ τὰς κρίσεις ὠμοὶ, “they are, regarding court judgments, [more] cruel” than all other Jews.
  • καὶ παραγαγὼν εἰς αὐτὸ τὸν ἀδελφὸν Ἰησοῦ τοῦ λεγομένου Χριστοῦ, Ἰάκωβος ὄνομα αὐτῷ, καί τινας ἑτέρους, “and he brought into [the court] the brother of Jesus (the one called Christ) the name for whom was James, and some others” and then…
  • ὡς παρανομησάντων κατηγορίαν ποιησάμενος, “he had made an accusation of having committed a crime” and had them stoned.
  • ὅσοι δὲ ἐδόκουν ἐπιεικέστατοι τῶν κατὰ τὴν πόλιν εἶναι, “as many citizens who seemed reasonable,” in other words not strict or cruel.
  • καὶ περὶ τοὺς νόμους ἀκριβεῖς, “and accurate regarding the laws” (as in accurate, exact, precise, strict).
  • βαρέως ἤνεγκαν ἐπὶ τούτῳ, “grievously bore against this,” i.e. took it very ill, i.e. disliked what had been done, and disliked it quite a lot.
  • μηκέτι τοιαῦτα πράσσειν, Ananus “should cease doing things like that” (so, like what?).
  • μηδὲ γὰρ τὸ πρῶτον ὀρθῶς αὐτὸν πεποιηκέναι, “for it wasn’t right that he should do it in the first place” (and what is “it”).

That’s it. Nowhere in the Greek do we find anything MacDonald claims is there. In context, this is about Jews protesting Ananus deploying an illegal court to excessively enforce the law.

For example, ἐπιεικής is an antonym of dikaios in the sense of “favoring strict justice,” and as such refers to people who are “reasonable, fair, good,” and to judgments that are “fair, equitable, not according to the letter of the law.” Josephus is unmistakably juxtaposing these Jews against the Sadducees he just described as “cruel” in their application of the laws. In other words, these are simply people who aren’t cruel; they think the law should be reasonable, and interpreted in respect to its community intent, not insolently hewing to the very letter of the law, much less convening illegal courts to enforce. This could mean Josephus had in mind that they thought James and gang didn’t deserve to be executed, but that entails no distinctive relationship to Christianity; they could have thought that about anyone. And indeed, that is Josephus’s point: he explains the Sadducees are just excessive; everyone else is not. That’s all that matters to the story. The particulars are irrelevant to him, and to his readers.

Moreover, if Josephus meant that there was something peculiar about Christians that warranted this judgment (that they deserved leniency or tolerance, for example), he would have said so. That he didn’t say so therefore means Josephus didn’t think it even mattered what this dispute was; solely that there was one, motivating the second point: that they abhorred his illegal assembly and wanted that crime never to be repeated. This is the law about which they are being more “accurate” than Ananus, and which finally motivates them to file protests with the higher authorities. They might have let that go if they supported the execution; but as they didn’t (at least not so fervently), they went after him for what they were more worried about: illegal assemblies. Josephus doesn’t think the crime James was accused of even mattered; that’s why he never bothers telling us what crimes James and gang are killed for. If it mattered to the story, he would say so. He doesn’t, therefore it didn’t.

Indeed, in no way is it likely Josephus would have meant “James and his fellow Christians” and instead said “James and some other people,” not even telling us they had any connection to James (or Jesus), much less what it was or how it related to the charges. On MacDonald’s thesis, Josephus simply would have written all of this differently; all human probability entails it, as does all evidence regarding how Josephus usually tells stories. In other words, all evidence pertaining to what is likely here. Probability is thus again against MacDonald, not in favor. He has to “read things in” to the text that not only aren’t there, but that would have been there had they been what Josephus meant. MacDonald’s thesis thus makes the text as we have it improbable, not likely.

And that’s even if we suppose MacDonald only means this to be some sort of tolerance of Torah-observant Christianity; but he argues the even more implausible thesis that this is about James supposedly preaching against adherence to the laws. But there isn’t anything in here about preaching, whether that or anything else. James is simply convicted “of a crime.” Period. The people are simply upset at Ananus being too strict in enforcing the law and engaging an illegal court to do it. Period. There is no more to this story here. That’s all Josephus tells us, because that’s all that mattered. It just makes things worse that there is no reason to believe, and ample reason not to believe, any James ever preached any such thing, much less was killed for it, even less by Ananus. Which is just one stone in that whole stack of implausible assumptions MacDonald leans on, remember?

So even MacDonald’s sole item of actual purported evidence, isn’t evidence at all. That Josephus would say most people were less cruel in enforcing the law than Sadducees is already 100% expected: because he already told us that is what this story is about—not “Christians.” Whereas that Josephus would omit any mention of why anyone is being executed here (James or any of “the others” whom Josephus doesn’t even explicitly link to James), or exactly why anyone would think that was excessive, is simply improbable if he meant to be talking about Christianity here. Yet this is 100% what we expect if it wasn’t. What Josephus says instead is that the elite replaced Ananus with Jesus because they didn’t like illegal courts being used for any excessive enforcement of their laws, and replacing him with the brother of the man he thereby murdered is the most pertinent way to denounce that crime. What Josephus never says is this having anything to do with Christians, Christianity, or Jewish opinions about either. There is nothing in the Greek even implying that.


Sound methodology requires (1) that we attend to the effect of our unevidenced assumptions on the probability of our theory; (2) that we attend to how likely our theory makes all the evidence (including what isn’t in evidence, not just what is); and (3) that we compare our theory’s merits in these regards to any other viable theory, especially the most viable competing theory you can find or contrive. MacDonald’s theory of what Josephus meant here fails all three tests.

The first task means we can’t rest our theory on a bunch of assumptions not in evidence; only if we can produce evidence for them (and enough evidence as makes them quite likely) or show that the competing theory also requires assumptions (assumptions yours does not) that are collectively just as unlikely, can we rest a theory on any assumptions at all. Otherwise, we should simply concede our theory is less plausible than the other. MacDonald relies on a ton of unevidenced assumptions, far exceeding any mine relies on in turn. Indeed, many of MacDonald’s assumptions are worse than unevidenced; many contradict the evidence. Whereas none of the assumptions my theory requires are even unlikely; in fact all comport with or even have support in evidence.

The second task means we can only claim the evidence supports our theory over competing theories if the evidence altogether is more likely to exist as we find it on our theory than on any competing theory. I’ve shown how all the evidence, in Josephus and outside Josephus, is highly improbable on MacDonald’s thesis—all of it is very unexpected and not at all well explained that way—whereas every single item of evidence we have is entirely expected or not at all improbable on my thesis: this is a story about how—and why—Jesus ben Damneus ascended to the High Priesthood; it was never a story about Christians. In the final analysis, even being as generous as possible, we cannot but find that MacDonald’s theory is hundreds of times less likely than mine. We therefore ought not hew to it. We certainly cannot use it as evidence “Jesus existed.”

Indeed, as I pointed out in our MythVision debate, this passage is not evidence Jesus existed even if Josephus wrote “the one called Christ” here—as we have no evidence Josephus did or even would know the difference between an actual brother of “Jesus” and a cultic one. We know for a fact all baptized Christians considered themselves “Brothers of the Lord” (On the Historicity of Jesus, Chapter 4, Element 12), an obscure theological detail Josephus was unlikely to be privy to, much less know how to discern from real kin in any tale that chanced to have reached him. And no, Josephus cannot have gotten any of this from Acts; not a single detail he relates is in Acts. Besides, the evidence shows Acts used Josephus, not the other way around. Which makes this story’s absence in Acts all the more a proof that Josephus never wrote this passage about Christians.


  1. Dr. Carrier wrote:

    So right from the start, MacDonald’s hypothesis requires inventing ad hoc suppositions, nowhere in evidence, that this brother of Jesus had taken or switched sides with the anti-Torah Pauline sect, and was advocating it so ardently in Judea that the Sanhedrin stoned him for it—an event nowhere recorded, not even in the book of Acts, which mentions no James the brother of Jesus at all, and which only mentions the known Torah-faction James being beheaded, not stoned, and by King Agrippa, not the Sanhedrin (much less Ananus), nor for this reason, and this action being approved by the elite, rather than (as Josephus’s story has it), so ardently opposed by the Jewish elite as to result in Royal and Roman intervention in deposing a High Priest who instigated the murder.

    Specifically concerning your reference to the “book of Acts”, you’ve indicated before that it is a forgery/fake.

    So why you would reference and rely on it here as some sort of authoritative text if that were the case?

    Just curious.

    1. Acts cannot be a forgery as when written it was not attributed to anyone. Forgery is a specific term that relates to claiming a false authorship. Anonymous works therefore cannot be forgeries. As to its being nevertheless “fake” that would depend on what you mean.

      Insofar as Acts is a historical fiction in aid of effecting a revisionist history (i.e. it was meant to be taken as a real history, but rarely actually relates the truth of history), you can say it is “fake,” but I don’t regard it as wholly so, i.e. I think Acts is revising a more or less known history, not inventing one entirely out of whole cloth (although the significance of the latter possibility is also addressed in my book On the Historicity of Jesus near the end of Chapter 9).

      As such, Acts would have included or adapted material that served its rhetorical purposes. We know Acts used the Antiquities of Josephus as a source for color context; its author therefore would have known if that work contained discussions of Christianity, and would have used any such discussions that were entirely in line with their aims, as this passage would be (for the reasons I explain) if the text as Luke read it contained the phrase “called Christ” at that time. Therefore, since this golden gift was not employed, it therefore can’t have had those words in it at the time; Luke would only have known of the irrelevant story of Jesus ben Damneus, which would be of no use to him. The only reason he wouldn’t use it.

      Likewise, MacDonald’s brief claim in our debate that Acts corroborates this passage did contain those words because it also describes the death of a James, is refuted by the fact that Acts shows no awareness of getting any detail of its James story from this passage in Josephus or any source for it Josephus would have used. They share not a single detail in common; they aren’t even about the same James. This is the case even if Acts wholly invented that tale of a James being beheaded by Agrippa; which is why I mention it.

      But though it isn’t relevant to any point you quote me making, it is also true that I do believe this tale derives from some genuine historical lore about the church. I think that other James really was executed by Agrippa; and we can infer things therefrom, e.g. a beheading indicates this James belonged to the upper class, and thus was no peasant as Christianity, even Acts, attempts to claim, an incongruous detail that suggests there may be some real history here that the author of Acts is glossing over. For example.

  2. Possible typo: “it having no visible effect whatever on any of the debates Paul had with the first Apostles or the Torah-observant faction whom he often reports stood against him,”
    Should this read “… who, he often reports, stood against him,” the pronoun “who” being the subject not object of the verb “stood”?

  3. Typos:
    In “alien space visitors were a well-established fact. For example, in the Stark Trek universe, that a light in the sky…” “Stark” should be “Star.”
    Should “for” be inserted after “argued” in “Christians, not the Pauline faction who argued abandoning strict Torah observance. There is no evidence whatever…”?

    1. I don’t know which argument you mean, but I assume you mean Element 40 in Chapter 5 of OHJ, establishing that the first Christians were equating Jesus with an angel already known in Jewish angelology (who was already the “Son of God,” the “image of God,” the logos, the paraclete, the high priest of God’s celestial temple, God’s first creation, and God’s agent of creation, and might have even already been known under the name “Jesus,” as one of “many” names it was given).

      This is not an argument for ahistoricity. It is solely background information, which is entirely and equally compatible with both historicity and ahistoricity. It is also hierarchical (some of it is a statement of plausibility or probability rather than certainty, and people tend to conflate which is which; they are not reading my book carefully), but that doesn’t matter, since I never derive any change in the probability that Jesus exists from anything in this element. Hence it is not evidence for or against either hypothesis. It is, rather, evidence against an argument against one hypothesis, a distinction lost on many who don’t pay attention to how I use this material.

      As such, it is a distant, esoteric digression that would needlessly consume clock time in any debate. It would therefore be counterproductive to even bring it up in that context, unless the opposition made an argument that it refutes—and so far, I have not encountered that in live debates.

      Although see The Difference Between a Historian and an Apologist for an exception (though not a live debate).

        1. Not really. Lots of gods and real people share names; indeed, if there was an angel called Jesus (it is in fact the name Joshua), it was for the same reasons as human men were named Jesus (i.e. Joshua): it means “God’s Savior.” So merely that there was a divinity named “God’s Savior” no more argues for someone of that name being divine than human, as plenty of humans were likewise so named.

          It is actually more telling that the name was chosen at all. That God’s savior would also coincidentally have been named by his parents God’s Savior is already argument enough to be suspicious; you don’t need to show this was already done for another angel to reach that conclusion. I discuss this point and its mathematical weight in Chapter 6 of OHJ. An angel already being so named adds nothing to the point.

          At most, the evidence that there was already such an angel knocks down arguments against mythicism, such as that the name can’t have come from anywhere else but being born a regular human (that argument doesn’t hold for a lot of reasons, but this one among them). But that just leaves us back at square one: was this just a coincidence (his parents just happened to name him God’s Savior and he just by chance ended up being proclaimed God’s Savior), or was a historical man equated with the angel Jesus and thus given the same name after his death, or was the belief that he was the same angel encouraged by his already happening to have that name, or did he assume that name himself early on specifically to evoke that comparison, etc.; or was this only ever a revelatory being so named, which just by coincidence is also a common ordinary human name? These are all plausible theories of historicity, and some of them are just as plausible as the contrary mythicist thesis that this being was always only ever the angel so named, and only met in visions or dreams. So we can’t decide which it is on this evidence alone.

          Hence, as I find in OHJ, it ultimately has no effect on the probability of mythicism vs. minimal historicity. It only forestalls some naive arguments against mythicism. But as they were already naive, we shouldn’t have been giving them weight anyway.

  4. Thank you for the detailed and thoughtful discussion, Dr. Carrier.
    Jesus is described in the Gospels as being a descendant of King David of the tribe of Judah, which is important to Jesus’s eligibility to be a christ/messiah. So how could James, if he were an actual brother of Jesus, be alleged by Dr. MacDonald to be even qualified to become any priest, let alone the High Priest, since such positions presumably would have had to be filled by descendants of the priestly clan, Kohanim (of the Levite tribe)? I readily acknowledge that i may be way off base here and I am far from familiar with Second Temple Jewish history and Josephus.

    1. This is a valid point to raise. I didn’t bring it up because it was so tangential. But you are right, the apocryphal Acts of James quoted in Hegesippus portrays James the brother of the Lord as an actual Levitical priest.

      It is actually unclear whether Hegesippus assumes this is that James or if the Acts of James he is quoting also said that; but in any event, MacDonald alluded to his believing it said that this James in that fictional story was the Brother of the Lord, and that this meant biologically, even though there is no evidence to support that assumption either.

      But after we get passed all those layers of unevidenced assumptions, we are still left with the oddity you remark upon: there is no possible way a Nazarene could be a temple priest. So something has to give. Generally, IMO, it’s the obvious: the Acts of James is complete fiction with no historical sources or credibility. It is just making everything up, and commits several errors of anachronism and ignorance of Judean context, for that reason. Ergo, it is of no use as a historical source.

      I cover that point at least in my peer reviewed article on this, which is accessible in my book Hitler Homer Bible Christ.

  5. You made a solid case. I don’t think it’s typical for historians to use math to calculate historical probabilities. And I didn’t know that Acts used Josephus.

    Moreover, If I could describe Dr. MacDonald’s presentation in 3 words it would be “interesting”, “odd” and “irrational/weak”. It was interesting because he used points that I’ve never heard historicists use before, such as the fact that the Gnostic gospels don’t deny the existence of Jesus and that Q was potentially the oldest Christian document that places Jesus in history. He didn’t rely on the former but on the latter he did for a while and if he could have provided evidence for Q and that it was as old as Paul’s writings, that would be good evidence for historicity. But I was surprised when he retracted that point (probably because he couldn’t support it) since he was consequently left with only the Josephus passage (maybe I’m misremembering, but I believe his presentation basically relied on 2 points: Q and Josephus. He didn’t really focus on the Gnostic gospels).

    Unlike the other two points, I’ve heard historicists use the Josephus passage as evidence for Jesus but they don’t usually rely heavily on it. So that made me think, if Josephus is such a good piece of evidence for historicity, why is it that other historicists, particularly Christian apologists who are extremely biased and whose careers depend on the popular belief that Jesus existed, don’t focus more on that passage? I found it quite odd. But Dr. MacDonald didn’t provide much evidence that the passage is authentic and that the word “brother” indicates a biological relationship between James and Jesus. He then said that the probability that Jesus existed was 80%, which seemed pretty random. To be fair maybe that number wasn’t based solely on his presentation he gave that day, nevertheless, I think he could have done a better job supporting his theory during that time frame.

    (Let me just say that I think Dr. MacDonald is a very good scholar. His idea that the authors of the Gospels used Homer directly is very interesting and he seems to be the only scholar who identified specific passages in the Iliad and the Odyssey that were used as templates to construct some of the Gospel narratives).

    Richard, you consistently do something in your debates which I find very convincing. When you are faced with passages that are ambiguous, such as “the brother of the Lord”, or weak evidence, like the 5th point in this article, you either don’t count them as evidence or don’t rely heavily on the them to make your case. Relying primarily on strong evidence gives your case a strong foundation.

    1. I think all your observations are apt.

      IMO, when he and other historians pick a probability, they are simply reporting by benchmarking feelings of confidence. So when MacDonald says “80% chance Jesus existed” this roughly means “given my own personal confidence in these evidence and conclusions, I feel like there is as much as a 1 in 5 chance I’m wrong.” Which is a low level of confidence, albeit still a net positive confidence.

      This in other words captures his gut feelings, after considering all his uncertainties and how many they are and how strong they are (and so on), and how that relates to other things in his life in a comparable state of felt uncertainty. For instance, “usually when I feel this uncertain, I have found I’m wrong maybe 1 in 5 of those times” might be a correct characterization of the psychology generating his number. This is typical (historians rely on unvetted gut instincts rather than careful logical analysis), but something IMO historians should be recognizing is deeply unreliable. That’s why I wrote Proving History.

  6. Dr. Carrier, Tim O´Neill made this huge blog post last year about the “Testimonium Flavianum”.


    He goes to great length of defending the “partial authenticity” – many times in the post he just repeats the argument from consensus -, even to the point of suggesting that the co-incidence hypothesis on the simmilarity with the Emmaus Narrative should not be dismissed! (When refering to the article “The Coincidences of the Emmaus Narrative of Luke and the Testimonium of Josephus
    ” by Gary J. Goldberg.

    And of course, there are things he does not address: he does not the address the fact that the next passage says how another disaster befell the Jews, thus refuting the idea of either a positive or neutral “original Testimonium” and, also he does not address the use of the word “Christ” or “Christians” without any context for it´s Gentile readers to understand what was he talking about.

    Will you ever respond, or are you done with dealing with him?

    1. Tim O’Neill is a liar and a crank. So I rarely spend any more time on him. He is too frequently dishonest and resorts to ad hominem too much to be worth engaging. I think his rants are a waste of everyone’s time.

      For what I mean, see:

      But as to your specific questions, yes, I think your analysis of his argument there is correct. His methodology is typically garbage, and you have identified several examples of why. Compare with my analysis in Josephus on Jesus? Why You Can’t Cite Opinions Before 2014 (which summarizes an official talk I gave to an audience of experts for the Society of Biblical Literature).

    2. P.S. I should also mention my discussion of the TF in OHJ (Ch. 8.9) I go through every single line of the TF and show why it cannot come from Josephus, based on well-established facts about Josephus’s discourse style (as analyzed by Hopper and others). Which refutes all attempts to get “some” of the sentences to be original. Meanwhile positing a completely different TF is the fallacy of special pleading. I show that there is no evidence of such a thing ever being there; yet that we’d expect there to be.

      1. Thanks for you response. It always shocks that scholars – inclunding Goldberg – just miss that part of “another disaster” befelling the Jews after the TF. I say this because Goldberg claims that the Emmaus Narrative is more in line with the Arabic version of the Testimonium, which was believed (and still believed by many) to be the original passage by Josephus (according to the Pines hypothesis), and thus, that makes the “common source” used by both Josephus and Luke as the more “simple explanation” for the similiarities in the writing. But of course, Goldberg was writing years before the most recent scholarship on the issue, so his claim is outdated.

        I am gonna use this opportunity to make a queston about the “TT” (“Testimonium Taciteum”). I red someone in the past claiming that it is more probable that Tacitus used Jewish people living in Rome at the time of his writings, – which according to that author/writer, it would include the writings of Josephus – , as a source for “Christ being executed under Pontius Pilate” rather then Pliny The Younger. What do you think of this claim?

        1. Yes, the Pines hypothesis has been refuted. The Arabic TF is derivative of Eusebius we now know. So that component of Goldberg is to be struck as inapplicable. But that wasn’t a significant component of Goldberg, so that has no effect on his thesis.

          On the TT, we have less evidence Tacitus used Josephus as a source for that section than we have that he used his friend and correspondent Pliny the Younger. So we can’t use it as evidence of anything being in Josephus at the time.

          In formal probabilistic terms, given all we know about the TF and the TP (Testimonium Plinianum), P(TF|J) < P(TP|P), so we should sooner assume the material comes from TP, not TF. But even if we disregarded that (and disregarding things is invalid), a fortiori it is still the case that P(TT|J) = P(TT|~J), ergo the TT does not increase the probability of the TF. It could not do so. Because the material in Tacitus is equally (in fact more) likely to come from P.

Add a Comment (For Patrons & Select Persons Only)

This site uses Akismet to reduce spam. Learn how your comment data is processed.