Last week I did a series on early reviews of my book On the Historicity of Jesus. If you know of reviews I haven’t covered, post them in comments (though please also remark on your own estimation of their merits).
One of those early reviews began a series by Nicholas Covington. Last week I commented on part 1. Here is my commentary on part 2, which deals with Paul’s reference to James. More to come. Here I’ll just comment item by item. But those who want to can skip all the commentary and go directly to my two-paragraph summary.
- Of course, if “James, brother of the Lord” meant “James the Christian” we must ask why Paul needed to specify that the James in question was a Christian (wouldn’t that have been obvious enough?)
As he points out, I give the reasons why (more than he mentions). But that will be apparent to anyone who reads the book on this point (ch. 11, § 10).
- Under the proposition that Jesus really lived, Jesus had a brother names James who must have later on played a role in the church (perhaps not as an apostle, but as somebody at least) and that explains the two passages reasonably. The probability of the evidence is close to 100% under the historicist framework.
Notice that I do not agree with this reasoning. The evidence is actually to the contrary that James “must have” played a later role. In fact, not even the evidence that he merely “may have” is sound. I make all the following points in the book, although not necessarily in the same place (you can check “brothers of the Lord” and “James” in the subject index to find every discussion):
Mark has Jesus simply disown his brothers. Mark has no evident awareness that any were even in the church later, much less famed leaders of it. Ditto every other Gospel. And Acts (written by Luke at the end of the century) also evinces no knowledge of a James the brother of Jesus ever being a leader in the church. Luke simply assumes generically that his brothers were Christians; but in Acts they disappear from history thereafter (I even have a whole section on this point, ch. 9, § 9). The notion that any brother of Jesus became an apostle (much less a leader) does not appear until a century or more after the cult began, and then only in manifestly absurd legends (e.g., see my discussion of Hegesippus in ch. 8, § 8).
Meanwhile, it is a demonstrable fact that all baptized Christians were brothers of the Lord. This means that if there were biological brothers, Paul would have had to make that distinction (e.g., by saying something like “brother of the Lord in the flesh,” or something akin). Yet Paul shows no awareness of any need to make that distinction. That means Paul only knew of one kind of brother of the Lord. And the only kind we can prove Paul knew of, is the fictive (see “fictive kinship” in the subject index), not the biological. Paul never anywhere shows any awareness of there being biological kin for Jesus; he even in some places conspicuously omits them; and his use of the phrase in 1 Corinthians 9 actually contradicts such a notion (ch. 11, § 10).
All of this I discuss and explain in OHJ. We mustn’t skip over that.
This is relevant, as Covington says “it seems to me that under historicism it is equally likely that either (a) Paul would use ‘brother of the Lord’ to refer to a literal brother or (b) use it to denote someone was a Christian,” but per above I do not believe that is soundly assumed. To the contrary, under historicism, Paul would need to distinguish those two groups from each other. He could not therefore refer to a biological brother of Jesus as “brother of the Lord” and mean distinctly a biological brother rather than a baptized Christian. Thus, under historicity, the probability that he would do so is low, not equal. The evidence therefore does not fit historicity–unless Jesus existed and had no brothers in the church, or his brothers were treated as equals and thus not singled out with any special phrase but treated like, and referred to like, every other baptized Christian, but in either case the Galatians passage ceases to be evidence for historicity.
Likewise, under historicity all the other evidence is very unlikely. Mark would know James became a revered leader of the church (or at least a member) and would write accordingly (barring ad hoc assumptions, which lower the prior); as would all the other Gospels; and Acts would record this as well; and our first evidence of it would not be in absurd, century-late myths and legends.
Covington does not take any of this into account in his calculations. So his final math, if he does not include these facts, won’t be accurate (it will violate the rule of total evidence).
The omission in Acts one could say is already covered by the likelihood ratio for Acts (OHJ, pp. 371-75, 603-05). But the omission in the Gospels and the absurdity of later legends about the brothers of Jesus either become background evidence for estimating probabilities in the Epistles (and thus lower the consequent probability of their contents on historicity) or else by “enhancing” the historicity hypothesis with the a priori assumption (since it is then not otherwise in b) that Jesus had brothers who were leaders in the church, the consequent probability of what’s in the Gospels and extrabiblical evidence goes down (further than already argued in chs. 8 and 10). Either way the mathematical effect is the same.
Also lowering the probability of the extant Epistles is the expectation that Paul, if there really were biological kin of Jesus in the church for Paul to refer to, would have to have made a distinction between them and the adopted brethren of Jesus. That he isn’t aware of any need to is to some degree (I would say a significant degree) less probable on historicity than mythicism (or else requires ad hoc assumptions that lower the prior: OHJ, pp. 584-85). In fact, on the historicist hypothesis (enhanced with the addition “Jesus had brothers in the church”), Paul would more probably have made clearer and more frequent reference to such a remarkable group within the church. Which further entails his not doing so is less probable on historicity than on mythicism.
- It is also my judgement that the fact that Paul identifies this “brother” as someone with the same name as one of the brothers listed in Mark is more probable under the historicist explanation than under the mythicist explanation.
Here I think Covington uses the wrong math. In fact, he makes a similar error to the blogger who argued that Jesus is more likely fictional because he had an unusually large family (see commentary here). Because there would have been dozens of men named James in the church even if historicity is false, therefore mentioning one for Paul is not unlikely. Indeed, as Covington points out, the apostle James whom Paul immediately discusses in Galatians 2, was conspicuously not the brother of Jesus. Likewise, the grammar in Galatians 1 entails the James being referred to there was not an apostle (contrary to every legend claiming the brother of Jesus became a leader in the church). I discuss this in OHJ, pp. 588-92. Paul thus means a non-apostolic Christian. Curiously, as I note in OHJ, it appears that when Paul most needs to distinguish non-apostolic Christians from apostles (due to the required force it has on his argument), he always uses the full term for a Christian, “brother of the Lord,” rather than its abbreviation, “brother.”
So what is the probability that Paul would sometimes refer to a baptized Christian named James as a brother of the Lord, given that all baptized Christians were known to Paul and his congregations as brothers of the Lord, and Paul needed to distinguish between an apostle and a non-apostolic Christian, so as to strengthen his claim not to have spoken to anyone who could tell him the secrets of the gospel until long after he had been preaching that gospel, thus proving he learned the gospel from the revealed Jesus and not human testimony? (Which is the entire argument he is making in Galatians 1.) I point out in OHJ that just saying “brother” would not suffice, since that could be mistaken as referring to a biological brother of Cephas; and not saying anything would not suffice, since that would leave unclear why this person is being mentioned at all (if Paul just met some Jew or catechumen, it would not be clear why he was mentioning them; that he met a baptized Christian in addition to an apostle would be relevant, however, as I there explain).
Covington’s reasoning is that it would be a coincidence (and thus not 100% expected) if Paul called a Christian by the name “brother of the Lord” who just happened to have the same name as what was later claimed to be one of the brothers of Jesus. But this is not the case if all baptized Christians were known by the name “brother of the Lord.” And they were. On mythicism we should expect Paul to not always use the full phrase, because it was a pleonasm (the same reason we don’t say President of the United States Obama every time we talk about President Obama), so we should expect it only occasionally, especially when a particular contrast needed to be emphasized (expecting most of the time just the abbreviated “brother” and “brethren”). Whereas what we should expect if historicity is true is for Paul to be even more specific than he is in Galatians 1, distinguishing the fact that he means not just a baptized Christian, but a biological brother of Jesus–who was not even an apostle, and yet still needed to be mentioned for some reason.
Thus the evidence is actually contrary to expectation on historicity, and not at all unusual on mythicism, wherein there would have been many men (non-apostolic, baptized Christians) whom Paul could sometimes fully refer to as James the brother of the Lord. Note that I am not positing this ad hoc (against which coincidence might be a relevant argument); that all baptized Christians were known to be (and thus could by anyone be called, at any time) brothers of the Lord is an established fact (OHJ, element 12, p. 108), and that pleonasms will appear less frequently than their abbreviations is a universal truth of human language use.
Meanwhile, the notion that Jesus had a brother named James appears only decades later, in a mythical text, in which Jesus disowns said brother (yet not even by name, just all his siblings collectively), and where no knowledge is evinced of said brother being subsequently special in any way at all, much less joining the church (even less leading it). And all subsequent stories concur for several decades on, even the first history of the church–except by then (decades even after the first Gospel is written) the assumption is finally made that those brothers joined the church, but then never do or say anything or appear in history at all–and none become leaders, or are described as meeting Paul. Then, many, many decades later, wild legends are composed that have this James, for the first time, being an apostolic leader of the church.
That sequence of events is improbable on the historicist reading of Galatians 1. And that has to be factored into the math. As does the fact that under historicity, Paul would need to be more specific if he meant to distinguish a biological brother from a baptized Christian. As does the fact that “brothers of the Lord” can’t mean biological kin when Paul uses it in 1 Cor. 9 (as I show in OHJ, pp. 582-88), which decreases the chances that he meant it so in Gal. 1. As does the fact that had the biological brothers of Jesus joined the church, Paul would be more likely to mention this remarkable fact at several other points in his letters (such as I note in OHJ, e.g. p. 524).
These things have to be calculated, in one way or another, as discussed in the preceding section above–depending on whether “Jesus had brothers” is treated as an ad hoc assumption that then generates expectations in other evidence like the Gospels, or a post hoc assumption based on placing this data, e.g. from the Gospels, in b (as Covington effectively does).
- [T]he probability of the evidence in question is close to 100% under the historicist theory whereas it is about 25% probable under the mythicist theory.
Covington’s mathematical assumption here requires that, on historicity, Paul could not refer to another person this way, e.g. “Matthias the brother of the Lord,” but that’s not true (and therefore his prediction of 100% for a legendary brother’s name here is incorrect). That the phrase “brother of the Lord” would mean baptized Christian is in our background evidence (OHJ, p. 108): it is not an enhancement to the mythicist theory, but an established fact, that Paul could refer to another person that way, even on historicity. So contrary to Covington’s assumption, Paul’s ability to do so is not a peculiarity of the mythicist theory. In fact, the probability is essentially the same, since there will be by proportion just as many Jameses etc. who were baptized Christians as would have been biological brothers of Jesus, since the name frequencies in the general population commute to both sets, and Paul’s infrequency of using this particular pleonasm would be the same either way (absent ad hoc assumptions like those I discuss on pp. 584-85).
Indeed, I dare say if Paul wrote “Matthias the brother of the Lord,” historicists would insist this is still evidence of historicity, and that the Gospels just forgot this one guy or got his brothers’ names wrong, thus exposing their assumptions are irrationally dogmatic–because they can never be falsified. But what if indeed Paul had said that? The probability of historicity would drop even further than I estimate, because then we would have a direct mismatch between what historicists need to be true (the assumptions or background data they are relying on to bolster h), and what the evidence actually contains. And all mismatches entail reductions in probability. That there is a match only keeps the estimated probabilities where I generously have them.
I discuss this mathematical effect, using the impact of our having or not having the trial records of Pontius Pilate, in Proving History, pp. 219-24. That our not having them is expected on b (just as it is expected on b that Paul will sometimes refer to any random James as a brother of the Lord) does not significantly reduce the probability that he crucified Jesus (any more than Paul sometimes referring to a James as a brother of the Lord significantly increases the probability that Jesus existed). There would arguably be an effect, but it would be too small to care about mathematically (as in the case of Pilate’s lost records).
On historicity, even if we set aside all the considerations I laid out above, it remains at best only trivially more likely that Paul would refer to a biological brother of a coinciding name using the exact same term as an adopted brother (i.e. a baptized Christian), just as it is only trivially more likely that Pilate didn’t crucify Jesus given that we don’t have Pilate’s trial records (see PH for how this works mathematically). And when we don’t set aside all the considerations I laid out above, it is actually considerably unlikely that Paul would do this even on historicity.
- Moreover, this [passage about James in Galatians 1] is about the only piece of evidence for an historical Jesus of which I know (I’ve looked into the issue before, and have come to the conclusion that most claimed evidence is actually very doubtful).
In the end it is reassuring that someone else sees the same point. Indeed, this is just about all the evidence there is. In my final calculation, when arguing a fortiori in favor of historicity, I only find two pieces of evidence have any strength, the two references to brothers in Paul and the two references to parents in Paul. I actually think these argue the reverse, due to considerations like those above, so in my a judicantiori estimates I find these passages to be very bizarre (= unexpected = improbable) on historicity, contrary to the assumptions of historicists, who don’t actually think very hard about these passages. But even when counting them as evidence for Jesus, I end up finding historicity improbable.
Covington might concur. Although, per above, I think he violates the rule of total evidence when assigning the evidence in Galatians 1 too high a likelihood ratio favoring historicity. Other considerations that he doesn’t assess drag it back down. And I think he uses the wrong math to generate a likelihood ratio of Paul using a legendary brother’s name on either historicity or myth (e.g. after putting the “legendary brothers’ names” data from the Gospels into b). Because due to Element 12 being in b (OHJ, p. 108), P(~e|h.b), that Paul would have used a non-legendary brother’s name in Gal. 1, is not zero, but in fact pretty much the same as P(~e|~h.b). Consequently, P(e|h.b) is pretty much the same as P(e|~h.b). (See PH, pp. 230, 255, 302n13.)
See my commentary on part 3. And my commentary on part 1. I will eventually blog on all of Covington’s entries in this series (his continuing index is here). He has also responded to this commentary on part 2 and I have replied in turn (see comments). And after that and further thought and discussion with other commenters, Covington came to rethink his position on the James passage in Galatians. He discusses his reevaluation in part 9. He now comes out roughly in agreement with my a fortiori estimate in OHJ. But nothing he has so far said has convinced me to alter my a judicantiori estimate. For the same reasons laid out above.
For a complete list of my responses to critiques of OHJ, see the last section of my List of Responses to Defenders of the Historicity of Jesus.