I am often enough asked what evidence there is for the historical existence of Paul that a summary write up would be handy to refer people to. This also has use as some scholars ignorantly claim that any standard that would deny the historicity of Jesus would entail denying the historicity of Paul (like that renowned fool James McGrath). Such a statement can only be uttered by someone who stalwartly doesn’t know (or is stubbornly refusing to hear) why the historicity of Jesus is said to be improbable.
The best formal attempt to argue for the non-historicity of Paul is that of Hermann Detering (see The Fabricated Paul). I cannot ascertain his qualifications in the field. But his writings are well-informed. They just trip over logic a lot. His case is not sound. Nor is anyone else’s I’ve examined. They falter on basic methodology (like ignoring the effect prior probability must have on a conclusion, or conflating possibility with probability) and sometimes even facts (e.g., Detering seems to think self-referencing signatures commonly appear only in forgery; in fact, they are commonly found on real letters—I’ve seen several examples in papyrological journals).
By contrast, the following is a basic run-down on why the historicity of Paul is actually, unlike Jesus, highly probable…
The Prior Probability
Jesus belongs to several myth-heavy reference classes. He is a worshipped savior deity. He is a legendary culture hero. He is a Rank-Raglan hero. And he is a revelatory archangel (already as early as the earliest writings we have, granting the letters of Paul are such). All of those classes of person already start with a high prior probability of being mythical, because most members of them are mythical (or for culture heroes, about even). And these are beings all of whom are claimed to be historical, yet are usually in fact mythical. Just like Jesus.
Paul does not belong to any such class. Paul thus falls into the class of ordinary persons who wrote letters and had effects on history. In ratio, most of such people claimed to exist, actually existed. By far. The mistake being made then is that people assume the starting prior for anyone claimed to exist is “50/50” (agnosticism) but we know for a fact that that is not true. Examine thousands of cases, and you will find persons claimed to exist, overwhelmingly actually existed. Only a small proportion didn’t. That entails that for any random person claimed to exist that you pick out of a hat, the prior odds are quite high they actually existed (On the Historicity of Jesus, Chapter 6.2 and 6.5).
To challenge that starting point, you need to show that Paul belongs to a reference class whose members are most often mythical (or at least half of whose members are). No one has done this. And it won’t be easy to. Because it’s shaky if you can only collect two or four members in the set you intend to use, because with such low counts, statistical anomalies are highly likely, and therefore any ratio you generate is going to have margins of error so great as to negate any utility of the set. This is why I used the Rank-Raglan set for Jesus: for that we have fifteen members. As far as ancient history and history of religions go, that’s pretty good. It still entails a large margin for error, but not so much as to erase its utility (as I demonstrate in Chapter 6 of On the Historicity of Jesus). And that is only reinforced by the fact that Jesus also belongs to several other reference classes that typically contain mythical people (see Part 3 of my commentary of the Covington Review.). So Paul is very much not analogous to Jesus.
Now, this is just the prior probability. And a prior probability can be overcome and reversed by a lot of evidence to the contrary. And you have to include that step. So, do we have “a lot of evidence to the contrary”?
The Differential Evidence
James Tabor has provided a very brief summary of the evidence for Paul (in The Quest for the Historical Paul). And perhaps the best monograph for the purpose is Gerd Lüdemann’s Paul, The Founder of Christianity (2002). But neither deals with this question thoroughly.
I will concur that, as I have found, the book of Acts is near useless fiction (On the Historicity of Jesus, Chapter 9). And all extra-biblical evidence for Paul, which is not based on the letters attributed to him, derives from Acts and no other source (either by using Acts as a source, or embellishing it’s tales with more mythology about Paul). So on those counts, Jesus and Paul are in the same evidential boat: there is nothing attesting them that counts as independent of Acts, and Acts is wholly unreliable as a source of historical facts in this matter.
Except in Paul’s case. Because we actually do have letters claiming to have been written by Paul. We do not have this for Jesus, except centuries later, when one ridiculous and patently obvious forgery appeared, Christ’s Letter to Abgar (which bears several hallmarks of forgery that Paul’s letters lack, including its naive and simplistic character, reflection of later church assumptions, and historical implausibilities). So the historicity of Jesus and Paul are actually not in the same evidential state.
We know there are many letters forged in the name of Paul. In fact, most letters attributed to Paul are forgeries (several in the canon, but a great many more fabricated centuries later, including a fake Third Corinthians that is actually still in some lesser canons, and a whole fake chain of letters between Paul and the Roman philosopher Seneca). So why do we regard any of the letters to be authentic?
Because unlike the forgeries, we can say several things about what are regarded as the six authentic letters of Paul (usually there are said to be seven, but I think arguments for Philemon being a forgery are sufficient to warrant at least agnosticism, as it is such an unusual letter in the corpus, and does have telltale signs of later invention: see Robert Price, The Pre-Nicene New Testament, pp. 467-68).
- First, they all cohere in style (idioms of vocabulary, connotation, grammar, punctuation, sentence length). The forged letters do not. They neither cohere with each other (except when produced as a unit, like the Seneca correspondence), nor with the style features of the authentic six. So one person did write those six (even if, as the letters openly state, they also reflected the views of a co-worker whom Paul sometimes names in each case).
- Second, they are stitched together from pieces of other letters. Each full letter named in the New Testament actually contains pieces of several letters, whose full content and original destination are now lost (see OHJ, p. 511). Sometimes so badly connected up as to be nearly unintelligible (e.g. the transition between 1 Cor. 8 and 9: OHJ, pp. 582-83). One does not forge letters that way. Which makes this another good indicator that these are not forgeries. Rather, someone tried to semi-reverently keep an original collection, but just the parts they liked, and assembled them together into a new whole in the most logical way they could. Their meddling after that was small and nitpicking, as the manuscript evidence shows, or blatant and obviously un-Pauline, as some of the interpolations made before 150 A.D. show (see Pauline Interpolations).
- Third, they all make arguments and interact persuasively in a context where the Jewish temple was still standing and its cult operating. And in a context where views of Jesus and the Church that appear in the Gospels have not yet come to exist (not even to denounce or counter or rebut, much less use or co-opt or transform). This is very unlikely unless the letters were written before the year 66 A.D. (when the Jewish War began, an event wholly unknown to the author), and before the Gospels were written (which could be as early as 70 or 75 A.D. for Mark).
That third point is important, because the letters explicitly present themselves internally as having been written in the 50s A.D. So the congruence of that fact with their content totally ignoring later existing doctrinal and tradition battles in the Church is very likely if the 50s is indeed when they were written. It is very unlikely otherwise: not impossible, but very unlikely, and that produces a strong Bayes factor favoring authenticity.
I can’t even think of a single example of an ancient forger successfully ignoring all the central doctrinal and tradition disputes of their own day merely to produce a convincing period-accurate but thereby contemporarily-irrelevant document. The temptation to support or attack the then-going views (usually by fabricating early support for them, e.g. 2 Peter) is simply too strong, and in fact is the usual motivation for forging documents in the first place. In short, the letters of Paul make no sense in the second century (in precisely the way that, for example, 2 Peter actually does).
These six letters also claim to be written by Paul and cohere in references to and descriptions of himself, and if they were indeed written when claimed, there is no reason to doubt that the author was Paul. No motive existed then to invent that name. Nor would such an invention at that point in history matter, since the exact name of this apostle is irrelevant to any argument one might construct about this man’s and these letters’ role in the history of Christianity. Paul may well have been the apostle’s adopted pen name, the Mark Twain to his Samuel Clemens. That makes no difference to anything (other than the fact that Paul is a distinctively Roman name that would be unlikely for anyone to have who was not a Roman citizen, which may be where Luke got the idea).
All that considered, that a Paul wrote those letters in the 50s A.D. makes all the evidence very likely. Otherwise, not at all likely. Therefore, it’s the far more probable hypothesis. And that makes Paul far better attested than Jesus: because we have some things written by Paul himself! (Would that we had such evidence for Jesus.) You can only assume “the evidence for Paul’s historicity is equivalent to that for the historicity of Jesus” if you assume the letters are forged. But then you can’t use the fact that “the evidence for Paul’s historicity is equivalent to that for the historicity of Jesus” as an argument for the letters being forged (without arguing in a circle).
Evidence for Forgery?
Robert Price has on occasion argued that Galatians, for example, was forged as a rebuttal to Acts. One can sort of make that fit, since indeed one is definitely responding to the other. But for it to be that way is far-fetched. The evidence is far more likely (far more what we would expect) if it was the other way around. In fact, Acts tries very hard to contradict Galatians (OHJ, Chapter 9.1), and has a clear motive to do so (to whitewash the divisions in the early Church and fabricate a tale of harmony), whereas Galatians does not seem aware of how hard it needs to contradict Acts, and its only evident motive is to respond to opponents of the writer claiming he didn’t really get a vision from Jesus but was conning his way in by actually getting his gospel from human informants (the whole point of Galatians 1).
And that theory requires no heaped up collection of weakly attested assumptions about later hypothetical ideological battles. Most Detering-style arguments are based on claiming hundreds of interpolations in these letters that conveniently and circularly support Detering’s conclusions, all based on a series of ad hoc assumptions about the second century history of the Church, when in fact almost everything we know about that is speculation, not established fact. The more assumptions you have to rely on, and the more conveniently complex they are, the lower the prior probability of your thesis. Speculation in, speculation out. Detering does not seem aware of this logical fact. He thus falls into the common trap of all bad historians: any theory you can gerrymander to fit all the evidence must be true. Because look how well it fits! Sorry. Illogical. (See Proving History, pp. 27 and 61, and the index, “gerrymandering (a theory)”).
Other arguments attempted include the one about personal signatures I mentioned at the start (sorry, no, not evidence of forgery); the fact that the letters of Paul are crafted by the literary and allusionary standards of Greek rhetoric (sorry, no, this would be expected for doctrinal correspondence intended to persuade; comparing this to letters with wholly different functions is simply a fallacy of false analogy; closer analogs are ancient political and legal speeches; 1 Clement, BTW, fits better the charge of not looking like a letter, yet it is obviously a homily, one that was simply delivered by post, to be orated to the congregation—delivering pre-written speeches like that was common in antiquity; in Classical Athens, Lysias made a living at it, and Paul’s letters are closer to that practice than to love letters or requests for mittens; in fact, these letters simply do not look like fictional letters from anywhere else in antiquity—I’ve read a lot of examples of the latter, and they have little in common in style and content with letters like Paul’s); and the fact that the letters are weirdly long (sorry, no, that is only because the editors created them by sticking a bunch of smaller letters together in mash ups sold as single letters).
Is Paul the Persecutor an Anachronism?
Another argument attempted to support a mythical Paul is to claim anachronisms in the letters, such as that Paul couldn’t have been allowed under the Roman Empire to prosecute Christians as the letters claim Paul had done. But that is not an anachronism. Until the 60s A.D., by imperial decree the Jews were allowed their own laws (in fact even Romans had to obey some of them in Judea; otherwise, only Jews were subject to them). This was because they allied themselves with the winning side in the civil war (supporting Julius Caesar and then Augustus, in the 50s to 30s B.C.). That was only no longer the case after the Jewish War (66-70 A.D.). I discuss this in some detail, with references, in my chapter on burial law in The Empty Tomb.
There is some dispute whether this agreement was already being altered before the war. For example, the Talmud and the Gospel of John both claim the Jews had lost the right to execute without Roman permission 40 years before the war, but that number, coming only from the Talmud, is suspiciously theological, and the evidence of Jewish trials and executions in the 30s and 40s is more than extensive enough to disprove such a legend; although they may have gotten the date wrong (or Roman permission may have been so easy to get it never even got mentioned). Crucially, Josephus, far closer to the events and the data than the compilers of the Talmud, makes no mention of this development, which is a significant strike against it. He may have been inclined to conceal this. But it doesn’t look so. Even his story of Ananus getting into trouble for executing James refers only to his assembling the court without imperial permission, and it’s unclear which step in that process was a violation, e.g. Josephus may mean that Ananus had not yet received the endorsement of the Roman authorities to be high priest or chair of the Sanhedrin, and in any event the implication is that a court assembled with imperial permission would be, even then, authorized in issuing death sentences.
But that only relates to death sentences, not other sanctions. And Paul does not say he executed anyone. Roman law certainly once allowed Jewish authorities to arrest, try, and execute blasphemers, and limits on specifically executing them were either merely procedural or would have simply defaulted such trials to meeting out lesser sentences. Roman citizens would have been exempt, of course, and also citizens of other non-Jewish polities, and probably any Gentiles (except in certain cases, such as defiling sacred objects). For instance, a Gentile could probably appeal to any court (maybe even a Jewish court) with the defense that they are not even a Jew.
Moreover, Paul actually implies his base of operations was Damascus (Gal. 1:17; 2 Cor. 11:32), which was not always part of the Roman Empire. Paul is otherwise vague as to where he persecuted (he was never in Judea when he was a persecutor, according to Gal. 1:22). Citizens of Damascus may have had the power to appeal to the Damascene authorities and Damascene law to exempt themselves from Jewish arrest warrants, although (a) Jewish inhabitants, like many others, in Damascus, did not necessarily have Damascene citizenship (you did not have it merely by living there), and so would be subject to Jewish law unless the Damascenes banned that and there is no evidence they did; and (b) any Jew who used that tactic to escape would likely be shunned as an apostate, betraying Jewish law, and could no longer associate with fellow Jews—unless their fellows agreed the warrant was unjust. So in Damascus what Paul could or couldn’t do would be a complex political question, and not a cut-and-dried matter of law. Paul implies he was hunted by the Damascene authorities once he converted, which suggests the Damascenes were actually endorsing the enforcement of Jewish law there (against, as they would see it, troublemaking Jews), and thus annoyed when Paul switched sides. The Romans, meanwhile, wouldn’t care, as long as Roman citizens weren’t being arrested, and the Jewish court didn’t overstep its bounds.
So, in general, in this case and others, I find nothing in Paul’s letters to be anachronistic. It is either too vague to be claimed such. Or it is perfectly consistent with the conditions of the 50s A.D.
Is Paul’s Christianity Too Sophisticated?
A commenter once argued these five claims against the letters’ authenticity:
1) The complexity and depth of the theology and ethics of Paul,
2) The virulent advocacy and opposition of Pauline doctrine,
3) the pronounced post-Jewish Christianity of Paul, enigmatic in a time pre-70,
4) anachronism of concerns addressed (about [being] celibate or [the] criteria of true apostleship)
5) post-apostolic gnosis.
But none of this works as historical logic.
1. There is no reason Christian theology would not be well developed within years (or indeed less than even one). All founded philosophies begin highly sophisticated. In fact, all religions in history have accumulated sophisticated theology in just years from their founding. And indeed, in this case the Christian theology Paul inherited and improved on would have developed from an already-existing Jewish theology, and thus most of its complexity is simply a carry-over of that, with sectarian tweaks (e.g., OHJ, Chapter 4, Element 18, pp. 143-45).
2. There is nothing unusual at all about religious innovators being virulently contrary and forceful right out of the gate—in fact, that is historically typical.
3. There is nothing even unique about the Torah-free Judaism Paul advocates. There were already pre-Christian sects experimenting with that (The Empty Tomb, pp. 107-10). And these letters depict it as new to Christianity, an innovation of a latecomer (rather than trying to fabricate it as original and an idea coming from the original founders, like Luke attempts to do in Acts). Such an innovation is also more likely to occur very early, otherwise institutionalized forces against it would be much stronger. In fact it makes particular sense in Paul’s context, where Diaspora Judaism was becoming increasingly disconnected from Palestinian Judaism, and indeed increasingly compromised and embarrassed by it. Temple nationalism was creating ill will toward Jews everywhere (OHJ, Chapter 5, Elements 23-29, pp. 153-63) and the “exclusivism” of Judaism was a constant cause of pagan anti-Semitism, and thus an obvious thing to get rid of, for anyone who wanted better pagan-Jewish integration…and a more successful growth in congregations (and thus, money: see J.D.M. Derrett, “Financial Aspects of the Resurrection,” in The Empty Tomb)
4. There is nothing anachronistic about Paul’s notions of celibacy or apostleship. If anything, Paul’s concepts make no sense in a post-Gospel Christianity, so there is no plausible way those letters were written after the Gospels became the frontline weaponry of inter-sectarian debate.
5. And to suggest Paul’s ideas of gnosis are post-apostolic is to simply make a circular argument. One has to presume it’s post-apostolic, in order to declare it post-apostolic. Because we have no other apostolic writings. Period.
So, no, there is no argument against Paul’s letters being authentically of the 50s A.D. here, either.
An Argument from Silence?
But can we deploy a valid AfS against the historicity of Paul? Doesn’t the fact that no one seems to mention him for almost a century seem odd? No. To be valid an AfS requires the actual expectation of mentions (Proving History, index, “Argument from Silence”). But between 30 A.D. and 150 A.D. there are no documents (none whatsoever) that would likely have mentioned Paul, except those that would—and they do!
In fact, there are almost zero documents, period. The history of the Christian Church before 150 A.D. has been scrubbed almost entirely clean. Tons of documentation and books that would have existed, were destroyed (OHJ, Chapter 8.4). All we have left is what’s in the canon, and a few stragglers beyond. Among those stragglers, several do in fact mention Paul. The letter of 1 Clement attests to Paul’s existence (and his body of letters)—and is ignorant of later legends placing Paul’s death in Rome; Clement places it in Spain. That letter is traditionally dated to the 90s A.D. Although I and some other scholars concur the evidence actually places it before the year 66 (OHJ, Chapters 7.6 and 8.5). And the Ignatian epistles attest to Paul’s existence (and his body of letters). Those are traditionally dated to the 110s A.D. Although I and several other scholars concur the evidence actually places them decades later (OHJ, Chapter 8.6).
What else could have mentioned him? We have nothing. The apologies of Aristides and Quadratus don’t concern themselves with ancillary actors in history like Paul. Nor do documents like the Didache (which sets rules for “apostles” but does not concern itself with listing who they are). So we can’t expect them to mention Paul. We don’t have the books of Papias. So we can’t say what “wasn’t in” Papias. And in that same period non-Christian sources are all oblivious even to the existence of Christians—or at best are only vaguely aware of Christ, and that’s lifetimes after Paul. So they can hardly be expected to know about Paul.
So there is no sound argument from silence against the historicity of Paul. Any document we would expect to mention him either does mention him, or no longer exists (and therefore can’t be checked). As I explained in my analysis of Bart Ehrman’s use of Pilate as a false analogy (whom Ehrman claimed was no better attested than Jesus, a claim that is factually and egregiously not true), most historical persons have as little evidence for their existence as we have for Jesus. Thus, we cannot infer non-historicity from just that (as I explain in Proving History, pp. 26-29 and 117-19). We need a prior favoring non-historicity. Or a state of evidence favoring it. Or both. For Paul, we have neither. For Jesus, we have both.
Desperate to avoid this fate for the argument, defenders of a mythic Paul will resort to insisting the Gospel of Mark should mention Paul. I cannot fathom a single reason why, or where on earth in the narrative that fact would ever come up. And that’s not just because Mark is a symbolic fiction that communicates nothing literally to begin with (see Mark 4 and OHJ, Chapter 10.4). Even taken literally, there is no point in it where a mention of Paul would be expected. But it does defend the version of Christianity that in his letters Paul says he invented. Yet Paul’s letters show no awareness of any of the stories told in Mark (apart from too vaguely to match the specifics introduced by Mark, e.g. OHJ, Chapter 11.7). The best explanation of that fact is that Paul’s letters precede Mark.
That argument failing, they will resort to saying that Justin Martyr never mentions Paul, therefore Paul didn’t exist. But that’s absurd. Justin wrote around 160 A.D. To suppose the Pauline letters were not circulating by then is ridiculous. Justin’s near contemporary Tertullian reports that Marcion was already circulating those letters by 140 A.D., decades before Justin wrote. Whereas when Justin wrote the anti-Marcionite canon did not yet exist. It was being created during Justin’s life or shortly thereafter. So there is no reason Justin would talk about Paul’s any more than any other letters. There were just a bunch of letters by bunches of different people. Not relevant to anything Justin argues. Only the Gospels were, and Justin was working from, it appears, the post-canonical Gospel of James (the first to place the birth of Jesus in a cave, the only birth story Justin mentions). Which doesn’t argue for the non-existence of Mark, by the way. So neither would not mentioning Paul.
Similarly other late authors (e.g. Theophilus, writing c. 180 A.D.). It is simply not reasonable to expect everyone to mention Paul, as we know for a fact even people using the canon with Paul in it often didn’t, and they rarely wrote anything in which mentioning him would be relevant. So that some authors don’t mention him is 100% expected. True, that some authors would mention him is also expected to be highly probable—except for the fact that we have almost no authors from the relevant historical period to check from. And yet, lo and behold, we still have some references to Paul (1 Clement and Ignatius; and one might also, albeit weakly, count the forgery of 2 Peter and the historical fiction of Acts, etc.). Everything else was destroyed. And we can’t argue from the silence of documents we don’t have.
There is therefore no valid argument from silence to work from.
To be fair, I don’t think there is an argument from silence against the historicity of Jesus, either—provided we accept he was an uninfluential nobody (OHJ, Chapter 2; although that is a tough pill for historicists to swallow: see OHJ, pp. 574-75 n. 82 and p. 557 nn. 55 and 56). Hence I assign a Bayes factor to that of even steven: no effect (OHJ, Chapter 8.13). And if so for Jesus, definitely the same goes for Paul.
What lowers the probability of Jesus existing is the kind of evidence we don’t have for Paul. Evidence placing Jesus in several myth-heavy reference classes, but not Paul. And evidence the non-existence of Jesus explains better than his existence does. There is nothing comparable for Paul.
In particular: Mistakes in Acts make the existence of Jesus less likely, because they are more likely to have occurred if his existence was being fabricated and hamfistedly inserted into a real outline of early Christian history (OHJ, Chapter 9.3-5). We have nothing of comparable effect for Paul. And evidence that Jesus was originally known only through revelations and hidden messages in scripture (OHJ, Chapter 11; e.g. Romans 16:25-26), which we also don’t have for Paul. And though that last requires agreeing the Pauline letters were written before the Gospels, that is a highly probable fact. Even apart from what the dates were, arguing the reverse requires swallowing too many improbabilities to credit. The sequence of Epistles, then Gospels, simply explains the evidence far better than the reverse can.
So unlike Jesus, Paul has a high prior probability of existing. And unlike Jesus, the Bayes factor for Paul not only does not reduce that prior, it actually raises it, because there is no evidence Paul’s non-existence explains better than his existence, and some evidence his existence explains better than his non-existence. The logically correct conclusion is that Paul probably existed and wrote the six authentic letters in the 50s A.D. (albeit we are now seeing only fractions of the actual letters he wrote, and somewhat meddled with). And that conclusion follows even when the same logic and methodology leads to the conclusion that, unlike Paul, Jesus probably did not exist. Except, originally, in the fevered imaginations of the revelators of a schizotypal cult (OHJ, Chapter 4, Elements 15 and 16, pp. 124-41).