Historian Mark Bilby is proposing a new thesis about dating the Gospels (and other NT texts) in an experimental study that, as of this publication at least, is at Zenodo (an open research archive), titled Neologismata, Volume 1: Proofs of New Testament Redactions between 138 and 249 CE (he also promotes it in a video). In it he asks for help vetting its results. He asked me privately for my own thoughts and I obliged. But he dismissed everything I said and ceased communications. I had examined his first three examples and found common flaws fatal to all three. So I asked him if he had any examples not subject to those flaws. He did not identify any.

Perhaps you can? I won’t spend any more time on this until someone does. So what I shall do here is explain the problems I found with his first three examples (plus a few other alarming errors), and leave it to others to explore his remaining examples (he claims some forty or so) to see if any survive the same critique. If you think there is any value in doing that.

General Merits and Flaws of the Bilby Study

The actual thesis of Bilby’s monograph is a bit muddled. His introductions and conclusions conflate a bunch of different objectives. But the overall gist is something like this: Bilby finds numerous words or phrases in the Gospels and Acts that appear nowhere else in Greek literature, papyri, or inscriptions, until after the New Testament is supposed to have been written, which he claims proves a few different things, such as (less controversially, 153–54) that Christian authors were astute, erudite, and creative literary auteurs (which indeed the trend in scholarship now has been to confirm, and with which I concur: see Robyn Faith Walsh and the Gospels as Literature and Obsolete Paradigm, 26–34) and (more controversially, 8) that the Gospels and Acts (even the Epistles) were all written after 138 AD, or indeed in the 160s AD. And that’s the theory that does not hold up. His approach appears to be methodologically incapable of supporting it. It is also, I suspect, a product of motivated reasoning (he seems keen to vindicate something like the Vinzent thesis).

In his own words, Bilby says his conclusion:

[S]quarely confronts the status quo scholarly dates assumed and/or argued for New Testament texts, dates ranging from the mid-1st to early 2nd century. If our most reliable chronological witnesses point to new Greek words being coined and/or first popularized after Hadrian [i.e. after 138 AD], and these words are found throughout the books of the canonical New Testament, then this should move the needle decisively on the outer bound dates of NT documents.

Indeed, if we could prove that direction of influence, he’d be right. The problem is, we can’t. At least not with his first three examples. And again, when I asked him for which of his forty or so examples could do it, he blanked. So I suspect none do. The overall problem is that he does not rule out a competing explanation of the same data that already enjoys higher priors and likelihoods: that what he thinks are NT authors emulating post-Trajanic literature are in fact NT authors and post-Trajanic literature emulating lost literature (or slang) in common—all of which statistically is surely Trajanic or pre-Trajanic, often even pre-Roman.

Bilby even acknowledges this as a problem (14):

Many texts are fragmentary, many more entirely lost. Many neologisms may have started out as oral custom before being written down. Many works that may be relevant (such as the Greek loanwords found in Coptic Nag Hammadi Codices) are not yet available as a digital corpus to search. We must acknowledge the limitations, arbitrariness, and survival biases inherent in the data that have been preserved, digitized, and made available to query.

But he never applies the impact of this admission to any of his first three examples. And that’s fatal. Because so many books are lost, and ancient authors routinely emulated literature they had but we do not, that it remains generally, already, more likely that any example Bilby has is not evidence of the origin of any word or clause, but of the emulation of earlier sources unknown to us—and thus undatable.

By not ruling out what is already the more likely hypothesis, his hypothesis remains, simply, unlikely. So does he ever overcome this problem in any of his forty plus examples? He could not show me an instance. Can you? Post any results you get in comments below. But first let me walk you through why this is a problem for his first three examples—and thus, I suspect, every example he might have—as well as other problems I encountered you should also be on the look out for.

Did Acts Use the Strategems of Polyaenus?

Bilby’s first example is developed in his Introduction rather than chapter one: the peculiarity that an extremely rare word, κατείδωλος, an adjective meaning “full of idols,” first appears in Acts 17, while the next time it ever shows up in any (extant) Greek source is an obscure book called The Strategems by Polyaenus, which was indisputably published in the 160s AD (it effectively says so in its introduction). Bilby therefore concludes Acts used The Strategems as a source here and therefore must have been written after 160 AD. Otherwise, “If the traditional 1st century CE date for Acts is to be believed, then at face value, the results suggest that the author/editor of Acts coined this term, or otherwise picked it up from an unknown source or from unattested, broader cultural usage” (6; cf. 4–8). The current mainstream dating of Acts is not the “traditional” date (of 60s–90s) but closer to 115–130 (as two volumes, Luke-Acts). But “after 160” is not currently an accepted range. So this is an extraordinary claim.

Bilby’s implicit argument here is that it is improbable that Polyaenus in the 160s would even know about much less be consulting Acts as a source for a military history. And he’s right. Though that’s not impossible, it is improbable, and empirically so (we can adduce ample evidence for why Polyaenus would be unlikely to know or use Acts here). Bilby’s explicit argument is that there are too many coincident features of these two passages (beyond just the one rare word) to argue that it’s just a chance accident. And he’s right (sort of). Bilby thus concludes Acts used Polyaenus. Q.E.D.

But he’s made two fatal mistakes here. First, he doesn’t consider that the reasoning for rejecting Polyaenus using Acts also holds for Acts using Polyaenus: Acts is also very unlikely to have even considered a book on military strategy much less used it as a source for a narrative about Athens—the author would sooner have cracked open some books about Athens to look for such material and ideas. So, really, we should consider that thesis as doubly unlikely: all the evidence already rendering such a late date for Acts unlikely counts against it as well as the same logic about unlikely source awareness and usage as applied to Polyaenus using Acts. But the second error is worse: Bilby mentions but then for no reason dismisses the likelihood that Polyaenus and Acts are both emulating a common source earlier than both.

There are many reasons that that is the most likely explanation of his data. But first let me pause on the underlying logic required here. The following diagram (that I just penned and edited) exhausts all logically possible explanations of this data (or indeed any other data Bilby could ever present for the same thesis):

There are four possible models here: coincidence (the texts match simply by accident, completely unaware of each other), standard derivation (like Polyaenus using Acts), inverse derivation (like Acts using Polyaenus), or use of a common source (like Acts and Polyaenus using the same source—like some history of Greece that covered the third century BC, as Polyaenus indicates).

Model 1

In his introductory example, Bilby does an okay job of ruling out Model 1 (coincidence), or seems to (more on that soon), but we’ll act like what he shows us is correct for a moment: Bilby rules out Model 1 by finding too many similarities between the two passages. First, Bilby says that “when describing Brennus, king of the Gauls” parading into “Athens” (sic), Polyaenus wrote that “after descending into Greece and observing the city was full of idols,” Brennus “asked those of the captives being held if the gold of the idols was solid” (where this sentence supposedly is in Polyaenus is a problem I’ll get to, likewise whether it referred to Athens; but just suppose for now that this is what Polyaenus wrote). And then, Acts says, “Now while Paul awaited them in Athens, his spirit was provoked within him as he beheld the city was full of idols” (this being my translation; I found Bilby’s too awkward).

But the relevant point here is as Bilby explains (7):

Both passages associate the term with a “city”, indeed the same city, Athens … [and] the word form (accusative singular), syntactical parent (πόλιν / “city”), and trailing neighbor[ing words] are identical, except for one intervening word (οὖσαν / “being”): “full of idols […] the city” / κατείδωλον […] τὴν πόλιν. Both passages use the word κατείδωλος to describe the main character as a figure from the past who observes the city he visits, apparently for the first time. Though the verbs of seeing differ, both happen to be participles with identical opening and closing letters (cp. θεασάμενος and θεωροῦντος).

To be precise, the word “full of idols” describes the city, not the main character (Paul / Brennus). But most of the rest is correct as stated, and though this is not an exact quote, there are too many similarities for chance accident to be likely: θεωροῦντος κατείδωλον … τὴν πόλιν (Acts) looks too much like θεασάμενος κατείδωλον τὴν πόλιν (Polyaenus) to be a coincidence. At least, enough so to render Model 1 the least probable of the four. Except that Bilby is not actually quoting Polyaenus. Nor was the city in this sentence actually Athens, but in fact Delphi. Which are mistakes suggestive of Bibly not researching this example very carefully, which complicates his entire thesis and project—because if he made these mistakes here; then probably, elsewhere. His every example thus requires careful fact-checking before simply trusting it as given.

A Sideline on Bilby’s Errors

The problem that this isn’t in Polyaenus, and the city wasn’t Athens, kind of tanks Bilby’s case in an unfortunate way, because these mistakes are not the actual problem with his argument that defeats (I suspect) all his examples. But I have to get through these incidental errors before returning to the universal problem all his examples I suspect fall to. So bear with me for a moment.

The actual text of Polyaenus (Stratagems 7.35.2) reads:

Βρέννος Γαλάτας ἀγαγὼν ἐπὶ τὴν Ἑλλάδα, θεώμενος τοὺς ἐν Δελφοῖς χρυσοῦς ἀνδριάντας, προσ καλεσάμενος Δελφοὺς αἰχμαλώτους διὰ τοῦ ἑρμηνέως ἤρετο, εἰ στερεὸν εἴη τὸ χρυσίον τῶν ἀναθημάτων.

Brennus, a Galatian, marching into Greece and seeing the golden statues in Delphi, summoned to [him his] Delphian captives, and through their interpreter, asked if the gold of those votives was solid.

The city is clearly Delphi, not Athens. I cannot figure out where Bilby got the idea it was Athens; that’s not in the Excerpts that Bilby (actually) quotes either, nor did Brennus ever reach Athens in his invasion: he was defeated at Delphi. Bilby also confuses this Brennus with a different one who sacked Rome a century before. So there’s a lot of bad homework going on here. But even more notably, Polyaenus never uses the word “full of idols” that Acts is supposed to have borrowed from here. Or even the word “city.” So all of Bilby’s evidence of an improbable coincidence doesn’t even exist.

What went wrong? Bilby didn’t check his source for the Greek text of Polyaenus. I found it odd that he cites the Excerpta Polyaeni for this (“excerpts,” i.e. quotations or paraphrases of Polyaenus in later authors), not The Stratagems itself, when obviously this is supposed to be in The Stratagems (and Bilby even says it was). I didn’t query this when I first corresponded with him because I assumed he would have done proper diligence here and not have made this mistake. But then when I took the time for this article to figure out why his text is coming from the Excerpta and not the Stratagems, I tracked the text down that he was using: it comes from an unknown Medieval Christian redactor (maybe of the 6th century) via manuscripts of the 10th and 11th centuries (see Melber & Woelfflin, Polyaeni strategematon libri viii, xix, 346, 449).

Modern editions declare the Excerpta Polyaeni to be unreliable redactions by a medieval Christian author. Which means, odds are, the entire phrase “city full of idols” crept into the text from a Byzantine editor riffing on Acts. So, this whole example must be removed from his study. But I want to make a point that applies to all his examples, not just this one. So let’s pretend this didn’t happen and that his mistaken version of the text of Polyaenus were the real one:

In that event, Bilby indeed would have ruled out Model 1. And though we now know it’s probably Model 2 (the medieval redactor is, most likely, using Acts to rewrite Polyaenus), if Polyaenus had written this, we’d be comparing Model 2 with Models 3 and 4. As I just noted, there is then as much against Model 2 as 3—and more against 3 if we count all the evidence for Acts not being as late as that model requires. Though I agree that that evidence carries only a small weight because it’s not very solid, it does carry some weight. Whereas we don’t have any evidence against Model 4. In fact, not only is there no evidence against Model 4, as I’ll explain, we have moderate evidence for Model 4. It would therefore be the most likely explanation of the coincident text between Acts and Polyaenus…had that text actually been in Polyaenus.

Model 4

Stratagems was a genre in antiquity. It’s one I’m especially familiar with in my dissertation studies of ancient scientific and technical literature. I cite several examples in my books on ancient science. They are all basically the same in structure: the author writes an introduction explaining what he’s doing (and sometimes for whom or why), and then collects a bunch of random examples from ancient literature, often organized by theme or something. Sometimes they collect literal quotations but often they paraphrase their sources, but they still keep material, especially the most pithy turns of phrase, and make small changes to make it their own. But always they are redacting sources. Nothing in a military strategy collection is original to the author. It’s all borrowed—explicitly. If you hop over to the introduction Polyaenus wrote for Emperor Aurelius, you’ll see he outright says he collected examples from prior authors; he is not coming up with anything in this book on his own.

So we know for a fact Polyaenus is using sources. And we know for a fact we don’t have whatever source he is using for Brennus. And so we cannot say what century that source was even written in, or that it didn’t originate “city full of idols.” But most histories of Greece that he would likely select would be a century or more old (these authors more often relied on respected classics than new publications). And a story about this Brennus could be in anything written from the 3rd century B.C. to the early 2nd century AD. We know for a fact a lot of histories of Athens existed that we don’t have (it was a whole genre called Atthidography); likewise other regional histories. I already noted that the odds are near certain that the author of Acts had at least one historian of the Aegean that he used for color background there just as he used Josephus for color background in Palestine (see How We Know Acts Is a Fake History where I explain that, as well as how Acts used sources and how authors used sources generally). We just don’t have that historian. Just as if we didn’t have Josephus, we’d never know where and how Acts used Josephus for material.

This means we have considerable evidence that Polyaenus and Acts shared sources: historians of Greece. When the author of Acts wrote a whole chapter about Athens, he surely looked for a decent title to give him fluff and detail to describe scenes in Athens with. And for that he would specifically look for books that focused on Athenian history. As the excerpt pertains to religion, and (we now know) Delphi, he might have used a collection of passages about religion or a local history of Delphi, but a history of Athens or lower Greece (either of which would for that century cover stories of the Aetolian League and thus Brennus’s invasion), would be more likely. And there were a ton of them then that we don’t have. Polyaenus would have been doing the same, only instead of specifically looking for things about Athens, he’d be skimming books about the history of Greece, wars in Greece, etc., for examples of military strategy. It is thus quite likely, certainly likely enough that we can’t rule out, that Acts and Polyaenus would have mined the same text for neat stuff to say—in Acts’ case, about Athens; in Polyaenus’s case, about military matters.

Since we know for a fact these authors used sources like this, and we know for a fact lots of lost sources existed that they would both be apt to use, and we know for a fact Polyaenus is even explicitly doing nothing but redacting passages straight out of histories of regions like Athens or Greece, we have here the most probable explanation for how Acts and Polyaenus could share a line. Model 4 is thus the most likely model. And that kills all hopes of Bilby getting Model 3 to be more likely. To do that he would need good evidence against Model 4—not the lack of evidence for it, but actual specific evidence against it. Remember this. Because it is absolutely crucial to the logic Bilby needs to justify his thesis. It cannot help him here to say “but we don’t know what source that was or that there was one,” because that doesn’t matter. We know there likely was. And since we’ve ruled out Model 1 (its probability collapses below rounding error), and we have evidence against Models 2 and 3 and no actual evidence for them, Model 4 wins, by a sufficient enough margin as to entail Bilby’s Model 3 is simply improbable.

And that’s even if Bilby’s evidence existed. As noted, it doesn’t. The actual text of Polyaenus shares nothing distinctive with Acts. It shares the vague facts of being about seeing statues in a city, but that is not an improbable coincidence. Probably lots of passages across ancient literature shared those elements. So Model 1 actually suffices here. But as well, since the only way Polyaenus could know this much about Brennus is if someone wrote it down in the 3rd century BC, his material could well come from then, even if through literary intermediaries. That is, even if Polyaenus used a later source than that, even that source had to have a source from the time of Brennus. And so therefore could Acts have had any of those sources. So Model 4 still works there too.

Sideline on Probability and Word Histories

That’s the main but not the only problem collapsing probably every example Bilby has. I asked him for any examples he had that do have evidence against Model 4. He adduced none. So, probably Model 4 kills all his examples and he has no case to make at all. But there is a second problem that likely plagues them all: Bilby’s implausible assumption that he has the first instance of the use of each of over forty rare words. That’s a bizarre claim to make on no evidence. And the Polyaenus/Acts example provides a particularly apt way to explain this problem, and how Bilby never addresses it in any of his first three examples, and couldn’t produce to me any example that did—so again, I suspect none do.

Given all the above, we don’t need to explain why the adjective “full of idols” first appears in Acts. It could have come from anywhere. It could even be an invention of that author. Bilby has no evidence it wasn’t. But it could appear in many prior lost sources and Acts got it from there. It could even have been common in oral discourse as Bilby also admits. Not every spoken word will have ended up in literature, and certainly once it did, that is very unlikely to be the same century the word started being used. We have no statistical reason to believe that we “just by chance” have the first-ever appearance of a word, when our extant sources are scant compared to all ancient literature there was, and largely selected at random with respect to this variable, and mostly high dialect rather than low. So how can Bilby establish this premise? He cannot, so far as I could discern. He just got angry with me that I said he had to.

But it’s basic math: if a word appears in ten books but only one in ten books survive, what is the probability that the book we have is going to be the first of the ten books it appeared in? Not good. Bilby never accounts for this. His premise is actually improbable. It therefore is not by itself able to make any conclusion probable. To the contrary, the improbability of a premise commutes to the conclusion. And this is even to account for all the lost books the word may have appeared in before Acts. We then have to also account for how long the word was in private use, use on the street, use at parties, use in speeches, before any author just by chance had an occasion to use it in a written text. Again, the statistical probability that this is going to be even once “in the same exact century,” much less forty straight times, is low, not high.

To prove that “full of idols” was never even a word until Acts invented it, or (as Bilby wanted) Polyaenus invented it, requires specific evidence for that odd claim that I don’t see Bilby presenting. He just “assumes” this premise is true, forty plus times. That’s illogical. This cannot get his conclusion to be probable. To the contrary, the more assumptions you have to embrace that are improbable, the less and less probable your conclusion becomes—not the other way around. In logical reality, Model 3 requires not just one extraordinary coincidence, but a whole long string of extraordinary coincidences. Otherwise, it’s far more probable these words predate the texts we just happen to have. Model 4 is just always more likely in cases like this. Until you can produce specific evidence against it. So, where is that evidence here?

It is also not improbable that Acts invented the word. These are, by Bilby’s own admission, brilliant and creative writers. And I had noted to Bilby that his discussion of this adjective left out important information: that while Acts is the first known surviving example of κατείδωλον as an adjective, it is not the first instance of an ancestor phrase, κατ’ είδωλων, which in ancient manuscripts (written in scripta continua without punctuation) would look identical to the adjective. Bilby tried to claim to me that it had a different morphology in the manuscripts, but in saying that he only confirmed he is inexpert at this: he was confusing medieval script with the ancient script that would have been in use when the Bible was written. He confused medieval manuscripts with ancient manuscripts. But never mind that gaffe. The relevant point here is that the history of κατ’ είδωλων provides a very plausible explanation for Acts’ invention of κατείδωλον—that the author of Acts was intentionally riffing on κατ’ είδωλων.

The clause κατ’ είδωλων is ancient. It’s found in literature of every century, going all the way back to Democritus. It was well known and often discussed: it referred to the atomist theory of vision, that “idols” (images) constantly bombard the eyes, shed by the objects emitting them, thus disturbing the soul, producing vision. Epicurus famously used this to explain gods as distant extraterrestrials, whose television signals could find their way into our brain when we slept or were in a trance, thus explaining how we have dreams and visions of them. In atomist theory it was also through the bombardment of idols (images) upon the eyes that human passions were incited. One could research this and build it out and explain it. But my point here is that this would produce an obvious play on words for the author of Acts…

Converting the well-known atomist idiom for “bombardment of images” on the eyes that can disturb the mind into “bombardment of idols” on the eyes to disturb Paul’s “spirit,” in order to comment on the evils of idol worship and its reduction to visual spectacle (and perhaps even a dig at public nudity and therefore lust, as some statues of divinities displayed bare breasts or genitals) provides us with a very plausible historical background to how and why the author of Acts could innovate this word, explaining its relationship to its visually identical ancestor phrase, and the specific choices of the author: to emphasize the sight of these “idols,” punning on the different sense of the word between atomist physics of vision and public cult, “disturbing” the “spirit inside of him.” An astute reader would get the pun—and the critique of paganism (and, indeed, of atomism) that it entailed.

And as best I can tell, the word was only ever used by Christians. Which supports the conclusion (at least mildly) that indeed it was a Christian invention and thus most likely invented by the author of Acts. Bilby does not identify any non-Christian use of the word except what he mistook as Polyaenus—which we saw was also from a medieval Christian, not Polyaenus, Which really here leaves us with Model 0: this is simply a word Acts invented to a distinct and explicable purpose, and that only Christians ever used after that. That does not rule out Model 4. It simply establishes that Model 0 is just as available. So whether 0 or 4, there is simply no case for Model 3 (while Model 2 explains the medieval Christian redaction of Polyaenus).

Hence, when we add up all these gaffes and fallacies and alternatives, this is literally the worst possible example Bilby could ever have attempted to argue his thesis with (yet his video leaned entirely on it).

Did Mark Use Arrian?

So much for Bilby’s capital example. It’s shredded. What about his second and third examples, occupying his whole first chapter? Sometime in the 140s Arrian wrote the Anabasis, an account of Alexander the Great’s conquest of Persia. As I noted in On the Historicity of Jesus (22–23):

Arrian, who though he wrote five hundred years later, nevertheless employed an explicit method of using only three eyewitness sources (two of them actual generals of Alexander who wrote accounts of their adventures with him). He names and identifies these sources, explains how he used them to generate a more reliable account, and discusses their relative merits. 

This is important because it means the same thing we learned about Polyaenus: Arrian is explicitly entirely constructing his text out of prior sources, in this case sources we know because he names them, and all of them are pre-Christian (they are all eyewitness accounts written in the 4th century BC). But we have none of them. So we cannot confirm, for example, that Arrian isn’t getting some rare word from them. So again, we cannot rule out Model 4. So if, for example, Arrian got something from Ptolemy, so could Paul or the author of Mark have done. We therefore cannot rule in, again, Model 3. There is no evidence, and Bilby presents no reason to believe, that Paul or Mark were using Arrian’s Anabasis to color their discourse or hagiography. And all the evidence that Paul or Mark wrote even before the 2nd century is also evidence against Model 3 that Bilby would then need evidence to overcome. And he doesn’t have any.

We can still Rule out Model 2—it’s very unlikely that Arrian would even know much less use Mark or Paul for ideas when writing a book about Alexander. But can we even rule out Model 1? Much less Model 4. After all, someone writing about a hero (like Jesus) intended to replace the likes of Alexander can be expected to have perused histories of Alexander, and used or absorbed what they liked. So Bilby is facing an uphill battle here, before we’ve even seen his examples.

His actual examples then only make it all worse.

Bilby’s second example is a colloquial interjection, οὐά (“Wah!” or “Ha!”). That’s it. It appears only once in the NT, in Mark 15:29, as a voice from the crowd mocking Jesus on the cross. It was commonly used as an idiom by the pop philosopher Epictetus in his public street lectures, which also happened to be recorded by Arrian in the first decade of the 2nd century. But Epictetus was an ex-slave and orated in a popular style, so his speeches can be expected to incorporate the people’s tongue, which Arrian clearly understood well enough to transcribe correctly. In fact it is obvious from its familiarity to his audience that Epictetus was using an expression everyone already knew. So it was clearly common dialect at the time. Epictetus was probably raised in the 50s and 60s AD, so he would likely have acquired his popular vocabulary and idiom then—well in time for Mark to have appealed to the same popular idiom in his similarly low-dialect tale, particularly when depicting riffraff mocking a street preacher. Indeed, Epictetus was street preaching himself in the late 60s in Rome, so likely even already using the expression before Mark wrote. But as I already noted, there is no reason to think Epictetus “invented” this expression or that Mark “needed” to hear it from him to have learned it. Mark would have learned it the same way Epictetus did, and their audiences did: from the street. And we have no relevant evidence by which to know for how long it had been in use by then.

We can’t even rule out Model 1 here, since this is just one word, not a conjunction of them. Bilby tries to claim Epictetus used the phrase sometimes in connection with vaguely similar themes as Mark is having the waggers spout at Jesus, but that cannot establish an unlikely conjunction. Epictetus used the word a lot. Mark talked about (and had Jesus talk about) those themes a lot. And those themes are super vague, with wide circles of similarity. Moreover, Epictetus clearly assumed the word was already widely familiar. So there is no oddity in Mark just happening one time to use this popular street expression. There is nothing particular enough to entail anything but a coincidence here.

Nor is there any evidence to support Models 2 or 3. There is even less reason to believe Mark had to wait for Arrian to publish speeches of Epictetus to know a street way to say “Ha!” than there is to believe Epictetus needed to read Mark to pick up the word and start confusing audiences with a sound they didn’t know, pretty much the opposite of how an orator is supposed to select their words (and even if he did, that would actually date Mark to the 1st century, not after 138). To the contrary, Model 1 is obviously the most likely explanation here: Mark and Epictetus are independently showing us first century street slang, which they both learned from common talk all around them, possibly even at the very same time.

Bilby needs evidence against this conclusion before he can get Model 3 to be at all probable. And yet Model 3 is stymied by evidence of its own improbability right from the start; while Model 1 is bolstered, not stymied, by that very same evidence. Model 4 is meanwhile moot here. With just one casual word, there is nothing they needed a common source for.

Bilby’s third example is προεπαγγέλλω, “forepromised” or “preannounced,” which Bilby says appears first in Paul (2 Cor 9:5, Rom 1:2), and never elsewhere until Arrian’s Anabasis 6.27, “therefore” Paul’s letters were forged in or after the time of Hadrian (Model 3). But we know Arrian is repeating three pre-Christian sources, at least two of whom would have covered the passage in question, so we know Arrian had two sources where he could be lifting this word from. We have no evidence he “invented” the word, or that someone in his time did and he by coincidence used a hot new word to redact his source material, rather than simply used a word already in either or both of his sources. And since those sources long pre-date Paul, we can no longer say the author of Paul’s letters had to learn of it from Arrian.

There just is no way to get to Model 3 here. We can say Model 2 is unlikely (no way Arrian is reading Paul). But as just one word, with no other correspondences, Model 1 is also likely again. But there is no way to rule out Model 4, either. We have enough evidence to suspect the word comes from popular pre-Christian sources available to Arrian (since that is far more likely given all the evidence against the letters of Paul even being second century). So when Bilby speculates that Paul is somehow creating “an antithesis to Apollophanes” losing a satrapy in an Alexandrian administrative move (a speculation I see no evidence for no matter how hard I squint), we can just as well say Paul is riffing on Ptolemy or Aristobulus just as Arrian was. We have no evidence to rule that out. Nor can we rule out the even more likely event that the word was used in many more places than Arrian’s sources, we just don’t chance to have examples that survived. So we have no evidence to rule in the far less credible Model 3 that Bilby wants.

But it gets worse. I trusted Bilby when he said this word “did not turn up in searches of major databases of Greek papyri or inscriptions.” But again, his promise to have checked is, well, false. Because I immediately (in less than sixty seconds) found six examples of this word’s regular pre-Christian use in inscriptions in the Packard Humanities collection, including an 84 B.C. Priene inscription, and in an inscription from the 2nd/1st BC, and another, and another of the 1st century BC. I can only speculate how Bilby missed these. Best guess is, he just didn’t check the PHC. But that was unwise, as the Prienne inscriptions are famous for giving us an early appearance of the word “gospel” in relation to a “savior,” Caesar Augustus. And Paul ministered in that region. He would have been exposed to that word there.

So Bilby’s third example is also hosed by botched homework, illustrating that you can’t trust any of his examples to have been diligently researched. You should assume, for safety’s sake, that anything you can think of, he didn’t—and thus fact-check him every time.

Conclusion

Bilby’s first three examples are so botched, and all fall to the same methodological errors that I don’t think it likely any of his examples could recover from, that I haven’t burned any more time on this. This was a day of labor as it was. And I wasn’t paid to do this. I just did it on spec for my Patreon supporters. But do feel free to check his open source document and his remaining examples, and let me know if you find any where he actually presents good evidence ruling out Model 4, actually checked his sources and facts correctly, and actually produces evidence for Model 3 (other than suppositions or assertions). I am sure Model 1 is often moot and Model 2 usually unlikely. But I would be astonished, after the above fiasco, if any of his examples actually produce evidence that both survives review and gets a Model 3 explanation more probable than a Model 4 explanation, or even (especially in single word cases) Model 1, or even 0.

But after having read this article, you are now ably armed to find out. If you have the mettle and spoons. I’m just too weary to continue with it myself.