I was going to do a news roundup of several new developments in ancient manuscript studies, until one of them turned out to be a roller-coaster ride down a rabbit hole filled with all manner of twists and turns. The subject? The Gospel of Jesus’s Wife. The other news I’ll post on separately. Because this one. Boy. It needs an article all unto itself.

I wrote about this in 2012. But everything has since changed. The intrigue is at its peak. And I don’t even know where to begin. It’s especially frustrating for me because, at the heart of this, I barely give even a fraction of a shit. This is a single almost useless tiny fragment of a medieval Coptic fabricated Gospel. It’s historically useless. It’s vastly too late to be of any interest to the origins of Christianity. It says almost nothing. Hence I figured, a quick news item at most, since so many people seem interested in it. Particularly kooks and cranks who think it proves Jesus was married, because a fringe heretical medieval sect of Christians made up a story in which Jesus says something illegible about “my wife” (possibly, from the text, meaning Mary, arguing either that Mary should be accounted a Disciple or was Jesus’s wife or both…or neither; so much is missing not a single complete sentence can be discerned—see Wikipedia for translation and backstory, although you might want to finish my article first, since I’ll be isolating some salient facts).

Almost right away suspicions this was a forgery arose (with the probable aim of scamming the anonymous owner out of tons of cash; unless the owner is in on the scam and wants to get it validated so s/he can sell it themselves). Evidence mounted. Documentaries and publications were pulled until further studies could be done. Scientific tests were done…which were claimed in the press to prove its authenticity, even though the critics kept explaining those tests were incapable of doing so (and the critics are right). Ad hoc excuses were made for all the evidence of forgery, to try and dismount the forgery hypothesis. People were dicks to each other. Flame wars ensued. Subtle sexism may have been involved. And articles have now been written about the chaos for The Boston Globe, Bible History Daily, LiveScience, and The Atlantic. It’s become a crazy mess.

All over what is, to be honest, an essentially useless shred of papyrus.

But there are two bigger issues in all this: One is the ease with which clever forgers can waste vast scholarly resources, all just to make a buck (or even just for the LOLs). The rampant threat of forgery in this field is huge. It needs to be taken seriously—not dismissed as unlikely. It’s actually very likely indeed. More so than in almost any other field. And the secret antiquities market (where this thing came from) drives a huge demand for forged artifacts like this. The other issue is how probability theory is crucial to resolving forgery debates; and how the mistakes forgers make are key to catching them. That part I find fascinating. And it’s what I’ll have the most to say about.

Rounds One and Two

My interest this week was initiated by someone discussing the April 2014 issue of the Harvard Theological Review that contains several papers analyzing the fragment. Harvard is the school housing and studying the papyrus and that has nearly staked its reputation on its authenticity, under the auspices of their renowned expert Karen King. Those papers were mostly glowing insistances the fragment is authentic. But all the arguments are fallacious. And their own evidence in some cases only raises suspicions further. Only one article made a case for forgery, and it sucked (indeed, it almost seemed like they picked the worst candidate to make the case—more on that shortly). Which is one of the most elegant ways of deploying a straw man fallacy: actually pick a critic who sucks, and have them build the straw man for you!

The scientific tests only showed things no one doubted. (1) That the ink used was chemically like ancient ink. It did not date the ink; tests are being designed that can—but, even though no one at Harvard seems willing to admit this, dating the ink can’t rule out forgery, because ancient ink can be bought on e-Bay, where ancient inkwells often go for sale (I’ve seen some sold for maybe fifty bucks), from which ancient ink can be reconstituted. Obviously a forger knows tests will be run. They are going to use the best material fabrication methods they can. And (2) that the papyrus itself is not a modern creation. Articles keep saying it was verified ancient (even Harvard keeps saying that), though in fact it was verified medieval—a problem Harvard might not want you to notice, which I’ll get to. But this, too, critics already predicted. Again, obviously forgers were going to use an authentic scrap to write their text on. Blank scraps of ancient papyrus are easily acquired. As are scraps from which existing ink can be washed away from (ink which, perhaps even, can be distilled and reused). So the tests were useless. They were not capable of ruling out a forgery. Only if the forgers were the dumbest fools in history could they have caught them with these methods. To the contrary, forgers know scholars love to rest on the fallacy “ancient carbon date = authentic,” just as Harvard is doing. Harvard may well be getting played like a fiddle.

No, the blunders of forgers are going to be harder to detect, and they are going to be found where the skills needed to make no mistakes are the most difficult skills to acquire. Everyone agrees the scribe sucked. Just like a forger would (not having written that script on papyrus with pen or brush thousands of times, in a language they speak). Everyone agrees the scribe messed up some of the spelling and grammar. Just like a forger would. And so on. The rebuttal is that shitty scribes who messed up spelling and grammar are also known from antiquity. So if the mere fact that the scribe sucked was the only evidence we had for forgery, that would not really be strong evidence of forgery. But the evidence does not consist of the scribe merely sucking at their job. I’ll explain below. But for now, the point is, ruling out forgery requires more than what Harvard is doing.

Critics weren’t impressed. More evidence was dug up and added to the rest and the case for forgery was re-presented, this time properly and soundly, in July of 2015, this time in several papers in New Testament Studies (which is published by Cambridge University, out of the hands of Harvard). And there has not been a significant response. Not from Harvard. Nor from King. Indeed King’s most recent comment on the debate (a letter to BAR, on page 8 of the September-October issue of 2015) doesn’t even attempt a response at all. And Harvard’s one official webpage on the forgery question, curiously fails to even mention (much less answer) any of the actual arguments of these forgery advocates. It instead converts their real arguments into sham straw men and dispatches the result. Which is, to me, practically an admission they’ve lost, and just can’t admit it. If you can’t face the real argument, if you can’t even mention that it exists, you’re done. Time to fold.

Here are the real arguments of the pro-forgery case…

Mistake Number One

It has now been pointed out that even what Harvard had done has increased, not decreased suspicion. Because the carbon tests came up with an unexpected result.

Karen King dated the manuscript, on solid grounds of script and dialect, to the 4th century A.D. (and proceeded to argue it could preserve a text originally composed, perhaps in another language, in the 2nd century—I should mention that’s pure conjecture; there is zero evidence supporting it, and I think King would agree). But. Um. Ooops. The carbon in it dates centuries later. This is a huge problem. Because the reason King dated the GJW fragment to the 4th century is that it is written in a dialect of Coptic that died out after the 4th century. No other examples of its Lycopolitan dialect exist in any known manuscript after that (beyond only some dwindling few examples in the 5th century). (And that dialect isn’t attested before the 4th century, either. Hence, 4th century.)

This looks like a fuck up. The forgers couldn’t do their own carbon dating (that requires skills and equipment very difficult to acquire and pretty hard to deploy without anyone noticing), and they are unlikely to have the very rare and advanced skills archaeologists and paleographers deploy to date recovered papyri, so they didn’t know the exact age of the papyrus scrap they were using. They figured it had to be ancient, from wherever they got it. They didn’t realize it was medieval. Mistake number one.

Noreen Tuross is the scientist who completed the only successful carbon test of the GJW so far, and it found a median expected date for it of 741 AD, with a more than 95% chance the papyrus was made between 659 and 869 AD (“Accelerated Mass Spectrometry Radiocarbon Determination of Papyrus Samples,” HTR 107.2 = April 2014). In other words, it could be as late as the 9th century. And is almost certainly not earlier than the 7th century. Lycopolitan died out hundreds of years before even that earliest date, much more so the median or latest date. That is extremely odd.

King argued that maybe some scribes were still copying Lycopolitan manuscripts from centuries before, even though no one used that dialect anymore. But that we have never seen an example of this ever happening before makes her ad hoc excuse improbable. Not impossible. But to claim something happened, for which we have no other examples of it happening—especially when we ought to have some by now—is a fallacy of special pleading. In probability theory, it’s called gerrymandering the hypothesis: adding yet more unevidenced claims onto the hypothesis to get it to fit the evidence. Which reduces the prior probability of your hypothesis (see Proving History, index, “gerrymandering”). It is in a sense a violation of Ockham’s Razor. The forgery hypothesis makes this incongruity somewhat likely (it’s the kind of mistake forgers are likely to make). The authenticity hypothesis does not (based on all past examples, and human behavior generally, continuing to copy out Lycopolitan manuscripts in the Middle Ages is not something we expect to happen).

So, that looks bad.

It’s all the worse that, it turns out, the anonymous owner (yes, they won’t go public with who they are) had a bunch of these Coptic manuscript fragments and lent King two of them. One is the GJW. The other is a fragment of the otherwise known Coptic Gospel of John. Also in Lycopolitan Coptic. And guess what? That papyrus has the exact same carbon date (within an expected random variance; Tuross found its median date is 718 A.D. with more than a 95% chance of it dating between 648 A.D. and 800 A.D.). Curious, don’t you think? That there is evidence that that is also a forgery only piles on the uh-oh. I’ll discuss that evidence below. But now we have two bizarre oddities: two medieval manuscripts written in a dialect that died out before the Middle Ages, written on papyri of essentially the same date (they could well come from the exact same papyrus sheet). Odds on this being a bizarre coincidence? Low. Odds on this being because the forgers thought this papyrus was ancient enough to contain Lycopolitan? Not so low.

Mistake Number Two

The forgers also made a mistake in how they constructed their fabricated text of the GJW. Some of this involved getting spelling and grammar wrong, although as a general phenomenon, so did ancient and medieval scribes, so that’s pretty inconclusive by itself. This is essentially what King argued in the 2014 Harvard Theological Review in rebuttal to the one author who submitted a case for forgery in that issue, Leo Depuydt (in “The Alleged Gospel of Jesus’s Wife: Assessment and Evaluation of Authenticity,” answered by Karen King in “Response to Leo Depuydt”). Depuydt’s piece was terrible. It was verbose, disorganized, confusing, and full of snark and outrageous overconfidence. King easily dispatched (most of) it. But as I mentioned, there are far better versions of the argument, made by far more capable and careful scholars. And King’s rebuttal is impotent against those. So including only him as dissent looks like an elaborate straw man.

It’s only worse that when Depuydt then answered King (online), he was such an unprofessional dick to her (I kid you not—read it; its shittiness is correctly assessed by Jim Davila) that I was genuinely appalled. He even failed to address most of her arguments. He quibbles over only one, and with snark dialed so far past 11 even champion internet trolls would balk. And that seems solely because his ego was severely ballslapped by her essentially (though not explicitly) calling him incompetent in Coptic. In the end, his manner of addressing the debate, IMO, disgraces him as a scholar. I don’t think we should bother listening to him further. There are better folk to pay attention to in this debate. He needs to be benched.

Even so, King’s reply to Depuydt did contain some fallacies indicative of her own motivated reasoning.

For example, King says the “easy dismissal of the relevance” of the ink and radiocarbon testing results “is simplistic.” But she never says why. Or even, in fact, why that matters. How does “being simplistic” affect the conclusion? And in what way is it simplistic to point out that obviously forgers in the 21st century should be expected to use ancient ink and papyrus for their craft? In part I think she is badly trying to articulate the fact that she actually didn’t understand Depuydt’s point. Since his delivery sucked, I can sympathize.

I’ll try to make his point better than he did: the probability that a forger would use ancient materials precisely to pass such obvious tests is very, very high (particularly in this ultra-high-tech period of history), therefore confirming the materials ancient does not reduce the probability of forgery to any significant degree. The forgery hypothesis already predicts that observation. And it does so without ad hoc suppositions. To the contrary, this is a suppositions well backed by current background knowledge (about the behaviors, knowledge, and requirements of forgers today—where it would be well known that using modern materials would never pass obvious tests), unlike King’s completely ad hoc supposition that maybe some medieval scribes were still copying Lycopolitan manuscripts for some inexplicable reason even though we’ve never seen evidence of that before.

King also admitted “the meaningfulness of statistical analysis for such a limited data sample is problematic for me to assess,” but again Depuydt did not make a clear or wholly defensible case on the matter of probability. He did a lousy job of explaining why the agreements between this one fragment of GJW and material in the Coptic Gospel of Thomas is highly unlikely by chance, but entirely expected if the forgers had raided a modern edition of the Coptic Gospel of Thomas for almost all of the Coptic words they constructed the GJW from. In fact only two words in this one business-card-sized snippet do not come from the GThom; everything else does, even down to the phrasing in several places. This is weird. And King does not seem to understand that it’s too weird to explain away, as she does, by just saying medieval authors would “also” know the GThom.

Because there are two things about this that don’t support that explanation. And King responded to neither of them. Understandably, as Depuydt buried these points in dense and flippant prose. But they are the only two points worth making: (1) the number of lifts from GThom is highly suspicious in so short a span of text in GJW; and (2) the lifts not only match a digital interlinear edition from 2002, they do so even to the point of repeating a typo (and other features) unique to it! The odds of (1) and (2) resulting from a merely “coincidental” conjunction of a medieval scribe being lousy and a fan of GThom are simply not good. This is resting on a lot of amazing luck. Whereas the forgery hypothesis already predicts both observations. And this is, IMO, damning. (It gets worse when we see the evidence that the owner’s other fragment, of the Coptic Gospel of John, appears to have been forged using nearly the same technique. More on that later.)

On the first point, what King does not appreciate (and her admission that she doesn’t understand the math explains why) is this: Her explanation, that the authors of GJW borrowed a lot of words and phrasing from GThom, would predict a lot of such matches across the whole Gospel, but not in one tiny little section of seven lines, which doesn’t even contain a single complete sentence on it. How likely is that? That you randomly grab just seven lines, and not even the whole lines but just fragments of them, and nearly every single word or phrase in that tiny snippet comes from the Gospel of Thomas? The odds are surely quite low, if not astronomical. If we were to suppose this random sample were typical (as her argument requires), then we’d have the bizarre situation where 95% of all the words and phrases in the entire GJW match GThom! In other words, that they almost never used any other words or phrases in their whole Gospel, than words or phrases in that other Gospel. Come on. Add to that the fact that the only few differences there are all exactly match what makes this papyrus financially valuable on the modern market (the addition of “Mary,” “my wife,” and re-gendering a few words to match).

On the second point, what King does not appreciate (and again her admission that she doesn’t understand the math explains why) is this: Her explanation, that the scribe who penned the GJW just “by accident” made a mistake exactly matching a mistake unique to a modern digital edition of the GThom, and then made several other mistakes that just “by accident” match up with physical features of that modern digital edition, requires assuming a series of coincidences that is certainly very unexpected, and thus intrinsically improbable, whereas the forgery hypothesis makes it far more probable. Indeed, it explains the oddities well. Without requiring a convenient supposition of numerous coincidences.

King’s argument thus requires that we just by coincidence got the one tiny bit of the GJW that contains not one but two of the weirdest statistical anomalies imaginable: nearly everything derivative of another Gospel, with numerous errors exactly matching those in a modern digital edition. Yes, it’s “possible” we got this bizarrely lucky, and got just the most statistically bizarre part of the GJW, both entirely matching words from another Gospel and also containing scribal errors that just by coincidence match exactly a modern digital edition of that other Gospel. But is it probable? I cannot imagine even remotely. And this is why it’s crucial for historians to understand how probability works; to grasp basic concepts of frequency and random distribution; to be able to tell the difference between common instances of randomness, and evidence of design.

King did correctly discern “Depuydt’s hypothesis that GJW was fabricated from patchwork extracts from [The Gospel of Thomas], basically comes down to an argument for literary dependence. Two questions are pertinent: 1) How close are the parallels? 2) Are they better accounted for by modern fabrication or ancient compositional practices?” Those are really one and the same question. Because the question amounts to: How likely are the established coincidences? It is hard to see how one can argue they are likely to have happened. Unless the text was forged by someone using that modern digital edition. Then every coincidence is explained, by a single assumed fact, rather than a panoply of already unlikely accidents.

Enter Bernhard

But I chafe at having to defend Depuydt. A far better proponent of forgery is Andrew Bernhard, who brilliantly summarizes the evidence in “The End of the Gospel of Jesus’ Wife Forgery Debate,” a guest post on New Testament scholar Mark Goodacre’s blog. He gathers and summarizes the discoveries of several scholars on the point. And it’s so well and clearly, and economically argued that you need read little else. This is the article King should have been asked to respond to. And still should. (Key elements of his argument were also formally published in the 2015 issue of NTS.)

Bernhard summarizes the evidence thus:

As I researched the textual relationship between the Gospel of Thomas and the Gospel of Jesus’ Wife, I began to collect evidence that ultimately convinced me that the Gospel of Jesus’ Wife was indeed prepared by someone relying directly on the PDF edition of Grondin’s Interlinear Coptic/English Translation of The Gospel of Thomas posted online in November 2002. I discovered that the textual similarities between the Gospel of Jesus’ Wife and the Gospel of Thomas were overwhelming. Basically, to create the Gospel of Jesus’ Wife, all a forger would have had to do was “cut and paste” text from the Gospel of Thomas, switch a few masculine pronouns to feminine (a single letter change in Coptic), and place two key Coptic words (meaning “Mary” and “my wife”) into the “patchwork” text to create its “sensational” content. The only other change that would have been needed was the simple deletion of the two-letter Coptic word meaning “not” in line 5.

The figure [above—ed.] illustrates the relationship between the Gospel of Jesus’ Wife and the Gospel of Thomas. It presents the Coptic text of the Gospel of Jesus’ Wife papyrus fragment: text that appears to have been copied verbatim from the Gospel of Thomas is underlined (double underlined if it might easily have differed). Parallels to the Gospel of Thomas (with their manuscript page and line numbers in parentheses) are noted beneath the Coptic text. The Coptic pronouns that appear to have been changed from masculine to feminine are printed in green italics. The Coptic words not copied verbatim from the Gospel of Thomas that look like they have been specifically inserted into the “patchwork” text are printed in bold red: “Mary” (line 3) and “my wife” (line 4).

You’ll notice every single word and phrase is underlined. Other than just the two words he says. Odds on that being a chance accident? No fool would bet on it.

So, that’s problem number one. King has no valid response to this. Claiming it’s all a coincidence is not a rebuttal, but actually a concession to the point. Coincidences are by definition less probable than design. And when design doesn’t have a low prior (as in fields where forgery is a frequent risk), that’s a problem. Which is why this is evidence of forgery. And it looks like really strong evidence, too. Even before adding the scandal of the carbon date and dialect.

And then we add the errors…

Bernhard goes on:

Line 1 [of the GJW] and the corresponding passage in Grondin’s Interlinear both unexpectedly omit the required direct object marker…before the final word visible on the line. This Coptic grammatical error might reasonably be compared to writing “She played the dog for me” rather than “She played with the dog for me.” A few other ancient manuscripts do contain an analogous mistake, but … [so does] Grondin’s Interlinear. The 2002 PDF version of Grondin’s Interlinear omitted [that same] direct object marker by accident as the result of a typographical error (unlike any other version of Grondin’s Interlinear).

Thus we have a text that appears entirely cobbled together using Grondin’s Interlinear also repeating a mistake in Grondin’s Interlinear unique to the 2002 digital edition of Grondin’s Interlinear! Odds on that? I wouldn’t take that bet either.

Bernhard also finds other strange features of the GJW that match up exactly with the line breaks in Grondin’s Interlinear. One being an accidentally dropped word, right where a line break in Grondin’s Interlinear could cause such a confusion. Another being the fact that the first line of the GJW “unexpectedly shares a line break” with the matching text of GThom in Grondin’s Interlinear. The latter is not an error, but it is yet one more amazing coincidence. Evidently the forgers also borrowed a relative clause from Grondin’s Interlinear, without realizing it is supposed to follow a correct form of noun (as it does in Grondin’s Interlinear, but not in GJW, producing a weird and unprecedented violation of Coptic grammar). In a similar error, the forgers appear to have borrowed several words from Grondin’s Interlinear in order to construct a sentence in GJW without realizing two of the words they borrowed were verb conjugations, thus creating a bizarre sentence with two verbs. As Bernhard puts it, “the [resulting] text could be compared to an English statement something like, ‘Let no wicked man does bring’.”

That’s a lot of mistakes and oddities the forgery hypothesis explains. The alternative is, again, just leaning on a whole bunch of coincidences. Which add up to one truly amazing coincidence. Which, really, defies credulity at this point. As Bernhard says, “it is startling that so many suspicious textual features appear on a papyrus fragment so small that it contains just seven lines of text with more than a single word.”

It Gets Worse (You Knew It Was Going To Get Worse, Right?)

By the time I had completed all the above research for what was going to be a single paragraph of news on my blog, I had blown half a day. I realized I was going to have to write a whole article. And that I had to exhaust all the avenues of inquiry by making sure I read all the latest on this, from everyone on both sides of the issue (or at least confirmed they hadn’t responded). So I trudged on. And found even more.

That’s when I nearly fell out of my chair.

Twice.

Everything I’ve surveyed so far I pretty much knew already, back when I first blogged about this in 2012. We didn’t know what the carbon dates would be, but we knew they’d confirm antiquity, because we knew that any forgers would know that they had to use ancient materials. But that the papyrus came from the Middle Ages was a surprise. So was the fact that the other manuscript the same owner lent to Harvard had basically the same carbon date. And a few other things. But the fact that the text of the GJW came almost entirely from a modern digital edition of GThom, complete with typos, line breaks, and errors, I already knew. I just figured you could use a more detailed summary this time. Because you need to get all that, to get what follows—and why it nearly knocked me out of my chair.

Again on Goodacre’s blog, Andrew Bernhard made another observation, in “The Forger and the ‘Translator’.” Because there had been a development since all of that. Critics had been pressuring Karen King to release the complete text of the original translation of the GJW that had been handed over to her from the (still anonymous) owner (who made that translation has also not been revealed). They’ve been asking for other documents she mentions having been given as well (more on that shortly). But King finally published the entire translation that came with the fragment (which is not her translation; as it was shoddy, she had more competently produced her own). And that’s when the case really exploded.

What Bernhard saw, and demonstrates in another article in the same venue, is a real howler: guess what most of that translation exactly matches? Grondin’s Interlinear. D’ope!

As Bernhard rightly points out, “nobody would ever attempt to translate a papyrus fragment of unknown content using an interlinear translation of another text. First of all, anybody with the ability to decipher and translate a Coptic papyrus fragment would use a dictionary to look up unknown words; a translator would want to determine the precise meanings of the words on the papyrus fragment in their actual context, not in the context of another text.” In other words, the translator magically knew to use the same digital edition of the Gospel of Thomas that everyone suspected was used to create the Coptic of the GJW, to also create the translation of the GJW! That’s simply impossible. Unless the forger wrote the translation. Which entails the Coptic was forged.

The translators even recreated an error in their source, proving they were using it: at one point the English in Grondin’s Interlinear has “(for)” to indicate a missing preposition. The translator (and forger) evidently didn’t realize that that had to be thus indicated with parentheses. So they produced the Coptic of the GJW with the missing preposition (not realizing it was missing in the Coptic of the GThom, but assuming what they copied could be translated as the English in Grondin’s Interlinear indicated). That’s one of those weird coincidental errors Bernhard noted previously. But then when they generated their English translation, they kept the “for” (assuming it was actually in the Coptic, because it was in the English of Grondin’s Interlinear), but didn’t put it in parentheses (as they should have, to indicate the word was conjectured and not actually in the text). Thus demonstrating not only that they were using Grondin’s Interlinear (and there is no reason they would have, unless the text was also forged using Grondin’s Interlinear), but also that they thought the Coptic they borrowed from Grondin’s Interlinear to create the GJW included the word “for,” when in fact it didn’t!

There are many other odd coincidences like that which give away the game. Such as, they incorrectly translated a word as “this” from Grondin’s Interlinear that they accidentally left out of the GJW, because Grondin’s Interlinear had so translated that word (sort of an error in Grondin’s, though more like interpretive license). But how did they know that that word was “there” when they had forgotten to include it in the Coptic? It’s in Grondin’s Interlinear, and grammatically is required, but because it starts a new line in Grondin’s, the forgers overlooked it, and forgot to put the Coptic word it corresponds to in the GJW. Yet magically remembered that it was there when they copied out the translation from Grondin’s for the GJW!

The creators of the English evidently had to be the forgers of the Coptic. Because the probability of all this being a random coincidence is so low it’s really at the earth’s core by now.

This matches one prediction I mentioned earlier: the one place where forgers can be caught, is where the skills required to succeed are extremely rare and difficult to acquire. That’s where mistakes will arise. Getting ancient ink and papyrus is easy. Learning a sloppy scribal script is easy. But being experts in composing and translating Coptic? The evidence makes clear the forgers were barely competent to do this. And being an obscure language even by the standards of ancient history, which is itself already a very obscure field, relatively speaking, this is to be expected. It’s one of the least likely skills the forgers could have adequately mastered for this project. And as one would thus expect, the Coptic of the GJW is repeatedly hosed in its spelling and grammar. Yet it all matches nearly exactly what would have been borrowed from Grondin’s Interlinear, even, in various ways, in their errors, and other coincidental matching features.

And now their English translation shows the same dependence on Grondin’s Interlinear, including repetition of errors they created in using Grondin’s Interlinear to compose the Coptic.

Really.

Oh. Right. Remember That Other Stuff?

Also on Goodacre’s blog, Bernhard provides an excellent summary in “The Gospel of Jesus’ Wife: A Call for Closure” of all the external evidence of forgery, by surveying the suspicious information and documentation that the owner of these fragments has so far provided.

This includes: (1) the discovery that the guy the owner claims he bought it from, Hans-Ulrich Laukamp, never had any interest in collecting antiquities, and never to anyone’s knowledge did—according to those who knew Laukamp—and would not likely have hazarded going to East Germany at the time to fetch it, whereas by contrast, conveniently, none of the owner’s claims about Laukamp can be independently verified; and (2) the fact that Gerhard Fecht, a renowned expert who supposedly described the fragment to a witness according to documentation the fragment’s new owner provided Harvard copies of, would never have said the things attributed to him (e.g. he would never have said it dated to the second century, because he full well would have known what Karen King did, that it was written in Lycopolitan and thus can’t have predated the 4th century). This last also shows the forgers were trying to pass this off as a second century text. It must have greatly ruined its market value when King pegged it to the 4th. And now that it’s carbon dated to the 8th, their financial hopes are not turning out well. (Though maybe they already made their windfall off its naive new owner.)

There is a lot that looks shady there, telltale signs that a fake provenance was created using forged documents implicating people who were conveniently now dead and with no surviving relatives or estates to verify any of it. Plus the evidence that both the manuscript fragments the secret owner lent to Karen King are forgeries (the fragments of both the GJW and the Coptic Gospel of John). And the fact that the English translation provided to her matches (even in its mistakes) the otherwise unlikely source the forgers had to have used to create the Coptic of the GJW. I concur with Bernhard. This really should be a closed case by now.

Which brings me back to that other manuscript fragment the same owner lent to King: a piece of the Coptic Gospel of John. Another expert, Christian Askeland, exposed this, too, as a forgery, in that same 2015 issue of NTS, I mentioned earlier (in “A Lycopolitan Forgery of John’s Gospel“). Askeland had already begun building a case earlier in The Tyndale Bulletin in 2014 (“A Fake Coptic John and Its Implications for the ‘Gospel of Jesus’s Wife’“). And notably, his discovery shows similar methods were used. Although this time, they didn’t have to invent anything.

As Askeland summarizes it:

First…all seventeen of the John’s fragment’s line breaks coincide with those of the [known] Qau codex exactly. The forger skipped every other line of [a modern edition of that] text when copying it onto his papyrus fragment. Perhaps, the forger’s goal was to create a single-column codex from what he wrongly perceived to be a double-column codex due to [the modern edition’s] page layout. Together with his alteration of [one] Lycopolitan [word] for [a] Sahidic [one], one might assume that his goal was somehow to confuse later scholars by dissimilating his forgery from Codex Qau. However, the forger failed to skip a line when he had to turn two pages of [that same modern] edition, and thus the last line of the verso breaks the pattern and creates an impossible textual variant. Such an exact textual relationship as there is between the new John fragment and the Qau codex has no parallel in the extant Sahidic Bible tradition and can only reasonably be explained by the John fragment having been copied directly either from [that modern printed] edition of the Qau codex, or else from that codex itself, before it was damaged (since the fragment includes some of [the modern editor’s] restorations of lacunas in the Qau codex).

That last point let me make clear: the forgers even used a modern scholar’s conjectured reconstruction of the text! Yes, maybe that scholar was right and the text really did read as they surmised. But it’s at least a bit slightly odd that the text matches their conjectures exactly. And the other evidence only compounds the improbability, including again an improbable exact match with a modern edition of the text being reproduced, and again an error that exactly matches a page break in that modern edition.

One of the same Coptic experts who endorsed the GJW originally, Malcolm Choat, also came to examine this fragment as well. And this time, he found the scripts of both fragments appear to be by the same hand. Though much is made of whether the ink itself is the same or different, I should note the ink needn’t be the same; expecting these tests as they would, the forgers could have used two different inkwells. But Choat also found other suspicious mistakes in the inking of the John manuscript, which only add to the coincidental exactitude of it matching a modern edition. His conclusion is: suspect forgery.

And if that is a forgery, very likely so is the GJW. As if we didn’t already know that. From all the other evidence.

And Then It All Gets Shanksed

This scandal of the owner-provided English translation almost made me fall out of my chair when I read it. Seriously. I was reeling, and literally saying out loud, “Oh you’ve got to be fucking kidding me!” And then I typed notes furiously. Wagging my head. But just when I thought I’d reached the climax of this adventure, I nearly fell out of my chair again.

I was trying to track down the latest remarks of Karen King (had she replied to the articles in the 2015 issue of NTS for example? Or to the other findings of Bernhard and Askeland?). I was told she had said something in the pages of BAR, and in the effort to track that down, I came across an article on the whole debate by Hershel Shanks (those familiar with biblical archaeology might be rolling their eyes at this point). Which almost knocked me out of my chair as soon as I started reading it.

Now, I think BAR is a crap publication, of increasingly dubious reliability. I think it shills to a religious audience for money. And that it is essentially a propaganda mill for whatever far-fetched opinions and financial interests are lately obsessing Hershel Shanks. I stopped my subscription long back when I realized Shanks isn’t exactly honest, and BAR all but complicit. So I don’t have a subscription. And I couldn’t view the whole article right away. But what I could see is that his article on the Gospel of Jesus’s Wife starts like this (and later with help from colleagues I acquired the whole article). It launches immediately into a glowing, eight-paragraph, page-and-a-half commercial for that shameless money-grubbing crank Simcha Jacobovici (those familiar with biblical archaeology might be retching at this point).

Wait. What?

No, really.

Somehow, Shanks made the story about Simcha Jacobovici. And then used that pretense as an excuse to say this barf:

Simcha Jacobovici (pronounced Yacobovitch) has made some remarkable archaeological discoveries.

For example, the first plague: When pharaoh refused to let the people go, Moses held out his rod, and the waters of Egypt, including the Nile and even the waters in vessels, turned into blood (Exodus 17:14–21). Simcha discovered how it was done: Earthquakes triggered gas leaks that did it; Simcha found a lake that turned reddish brown from such a gas leak.

Before the Israelites left Egypt, they “borrowed” silver and gold from the Egyptians and then took it with them (Exodus 11:2–3; 12:35–36). Simcha has found some of this—in Greece. As Simcha discovered, some of the Israelite slaves did not go to the Promised Land but boarded ships and sailed to Greece. There some of the gold that the Israelites brought with them from Egypt turned up in the tombs of Mycenae.

And this is only the beginning. Simcha’s discoveries relating to Jesus and early Christianity are breathtaking. Let’s begin with the nails that were used in Jesus’ crucifixion; he found them in the lab of an Israeli academic at Tel Aviv University.

Simcha has also discovered the true tomb of Jesus and much of his family.

In another Jerusalem tomb, Simcha found one of the earliest Christian symbols—Jonah emerging from the whale after three days, just like the resurrection of Jesus from the tomb after three days. Elsewhere was an engraving of the earliest Christian symbol—a fish.

Most recently, Simcha has been arguing that Jesus was married to Mary Magdalene and together they had a son. “The proof of the historical marriage is … overwhelming,” he says.

Let me say it straight out: Simcha is a friend of mine. I like him. He’s very smart. And I admit that I’m not enough of a scholar to refute his claims. But I think it fair to say that the reaction of the academy to Simcha’s claims, by and large, has been a thunderous silence. Simcha is simply not taken seriously. The major academic reaction, where there has been one, is a rolling of the eyes. Harvard’s Lawrence Stager has characterized Simcha’s archaeology as “fantastic archaeology.” And BAR does not publish Simcha’s far-fetched claims. Simcha has his own reaction to the academic response—or non-response—to his claims, this one regarding the fish: It has been “largely ignored because … [it] upsets too many theological apple carts.” As his astounding finds pile up, he admits he is met “only [with] more derision.”

I recite all this in contrast to Karen King’s announcement that in an ancient Coptic gospel fragment Jesus, speaking in the first person, refers to “my wife.” At first glance this would seem to be of a piece with Simcha’s “discoveries.” In fact, it is far different.

Then Shanks goes into a puff piece about Karen King. And sort of builds the case for forgery. He never mentions Jacobovici again or explains why he ever mentioned him at all.

WTF?

Stranger even than that, even though Shanks just gave Jacobovici a huge glowing commercial, and totally endorsed all his crazy claims in a top biblical studies magazine, and merely mentioned that others won’t take him seriously, fans, and Jacobovici himself (and his friends, like James Tabor) responded with angry letters to the editor (in the same later issue as Karen King’s letter) as if Shanks had actually insulted him. And then took that opportunity to list all the amazing things they’ve done and all the support they have from scholars they then name.

Does anyone notice anything fishy about all this? Jacobovici is Shanks’ friend. And Shanks says BAR won’t publish Jacobovici’s claims. Then Shanks published all of Jacobovici’s claims in BAR. Giving Jacobovici (and his minions & sidekicks) an excuse to publish even more in BAR in defense of their claims, which now BAR has to publish because they, after all, morally “deserve” the opportunity to protect their good name from this “vicious smear.” See what just happened? Shanks just got the equivalent of an entire article about Jacobovici’s claims published in BAR. Nice trick. Yeah. I’m sure Jacobovici was so insulted. I’ll bet they are scowling at each other ruthlessly over their God-priced bottle of single malt scotch while sunning themselves in the Keys, high-fiving over how well they just played the suckers

I suppose it’s possible Shanks is just a crotchety old man, now on the outs with Jacobovici for some arcane reason, and instead used this as an excuse to “mock” him (as Jacobovici claims he was). On that interpretation (in which there is still no reason whatever to have burned even one paragraph on Jacobovici in this article, much less eight), you should read everything Shanks wrote as snearing sarcasm. But I think it was written that way on purpose. To sound like sarcasm. I think BAR just got scammed. Shanks just got his ignored genius of a friend published in BAR.

Yay science.

Conclusion

Is that last discovery in any way relevant to the Gospel of Jesus’s Wife? No. But it capped my evening of research with such an appalling and bizarre conclusion I just had to close with sharing that.

The rest, though, is conclusive on the forgery question. The weight of evidence and suspicions is overwhelming. The evidence of provenance is suspicious. The texts are suspicious. The English translation provided by the owner is suspicious. The carbon date is suspicious. And more than just suspicious.

The conclusion that seems highly probable at this point is that the Gospel of Jesus’s Wife and its associated fragment of the Coptic Gospel of John were forged with antique materials and a practiced rudimentary scribal technique by some poor students of Coptic who attempted to use modern printed editions as their source texts, and botched it just enough to leave so many clues to this fact, that the alternative explanation, of a vast array of remarkable coincidences, simply has no credible probability to tout.

Really, we should be done with this.