Thursday, April 10, 2014

Jesus’s Wife Resurrected from Dead



Eight of the eleven articles in the most recent issue of the Harvard Theological Review discuss the authenticity of the so-called Gospel of Jesus Wife (GJW), which Karen King publicized through a shrewdly-orchestrated media frenzy in September 2012. The core relevant articles include a survey of the papyrus scrap by King, a refutation of authenticity by Leo Depuydt and a response by King. Five supporting articles detail two spectroscopy examinations of the ink (Yardley and Hagadorn; Azzarelli, Goods, Swager), two radiocarbon datings of the papyrus (Hodgins; Tuross), and a paleographic evaluation (Choat).

Karen King’s initial argument that this fragment demonstrates a fourth century literary manuscript of the “the Gospel of Jesus Wife” is now officially dead, by her own admission. We are left with a deflated seventh to ninth century semi-literary scrap ... or a fraud. We have no plausible direct literary evidence for a new non-canonical gospel. The question remains as to whether we should recognize this scrap as an ancient semi-literary document or a modern fraud. According to King, the arguments concerning fraud are highly problematic, and the scientific and linguistic evidence repeatedly affirm authenticity. 
“The scientific testing completed thus far consistently provides positive evidence of the antiquity of the papyrus and ink, including radiocarbon, spectroscopic, and oxidation characteristics, with no evidence of modern fabrication.” (King, “Jesus said,” 2014, 154)


According to the results, the ink used is indeed the most obvious choice for a modern forger — carbon ink. The ink is composed of soot. “The inks used in this manuscript are primarily based on carbon black pigments such as ‘lamp black.’” (Yardley etal., 164) King attempts to paint the resultant test as proving the implausibility of fraud, arguing that “their research to date shows that details of the Raman spectra of carbon-based pigments in GJW match closely those of several manuscripts from the Columbia collection of papyri dated between 1 B.C.E. and 800 C.E., while they deviate significantly from modern commercial lamp black pigments.” (King, “Jesus said,” 2014, 135) However, no one would suggest that this was forged with modern commercial pigments. Someone would have mixed soot with a solvent, producing the obviously low quality and uneven writing medium on the papyrus.


Using two labs, the GJW fragment and a Sahidic John fragment associated with the same papyri lot were carbon dated. The rounded 2-sigma ranges for the manuscripts are as follows:

640–800 CE
650–870 CE
680–880 CE
410–200 BCE

The second test (14 March 2014) was apparently ordered after the extremely early date arrived from Arizona (June–July 2013). Whatever the case, if one of the two GJW 14C dates were to be accurate, it would probably be the Harvard range (650–870 CE), which is corroborated by the related GJohn manuscript (chart above). Having said this, the result remains somewhat inconclusive. (δ13C levels were also higher than expected, suggesting contamination in all samples.)

So does this confirm the authenticity of the GJW? Such a late dating bulldozes King’s first appraisal of the manuscript as a fourth century witness. The GJW fragment under question is broken on all sides except the top, where apparently the modern forger cut the empty section off of a larger fragment which was in fact ancient. Carbon dating has no value for authenticating such a manuscript, although if the Ptolemaic date (410–200 BCE) offered by the Arizona AMS lab were accurate (of which I am not convinced), fraud would be certain.


Choat’s assessment of the scribal hand is hardly an enthusiastic endorsement of its authenticity:
“Overall, if the general appearance of the papyrus prompts some suspicion, it is difficult to falsify by a strictly paleographical examination. This should not be taken as proof that the papyrus is genuine, simply that its handwriting and the manner in which it has been written do not provide definitive grounds for proving otherwise.” (162) 
His article surveys the oddities of the scribal hand, noting the lack of clear literary or documentary parallels. Choat states, “[w]hile I cannot adduce an exact parallel, I am inclined to compare paraliterary productions such as magical or educational texts.” (Choat, 161)


Leo Depuydt presents the argument which is accepted by most specialists who are familiar with the GJW. The modern forger (1) created the text by rearranging several sentences from the Gospel of Thomas and (2) unintentionally left evidence of the fraud through two grammatical infelicities ("blunders"). The first is the omission of the object marker ⲙ- in line one (ⲧⲁⲙⲁⲁⲩ ⲁⲥϯ ⲛⲁⲉⲓ ⲡⲱ̣[ⲛϩ]). The second is the awkward construction ⲣⲱⲙⲉ ⲉⲑⲟⲟⲩ (more correctly ⲡⲣⲱⲙⲉ ⲉⲑⲟⲟⲩ or ⲣⲱⲙⲉ ⲉϥϩⲟⲟⲩ). Depuydt also mentioned a third serious error, which I believe to be the most damning evidence against authenticity (186); in line 6, the forger has combined a positive habitual from GThomas with a negative habitual to create the nonsense chimera verbal phrase ⲙⲁⲣⲉⲣⲱⲙⲉ ⲉⲑⲟⲟⲩ ϣⲁϥⲉ{ⲓ}ⲛⲉ (“Evil man habitually does not he does habitually bring” sic). Notably, Francis Watson, Alin Suciu-Hugo Lundhaug, and Andrew Bernhard have popularized many of these arguments, detailing how Depuydt’s first "blunder" seems to derive from a typo in Michael Grondin’s 2002 online PDF of the Gospel of Thomas.


In Karen King’s mind, if one can not exhaustively prove the inauthenticity of the GJW fragment, then it must be accepted as authentic. The results from spectography, radiometric dating and Choat’s paleographic analysis all leave the door open, therefore the fragment is undeniably authentic. Karen King maintains the problematic infinitive form ϣⲁϥⲉ “swell,” and ignores the persuasive reasoning behind the reconstruction of the damning error above. I encountered no serious discussion of this in her original article. In my opinion, this argument alone inauthenticates the GJW fragment, yet King is unconcerned, instead positing an unattested verbal form. I could imagine why someone might differ with me on various issues here, I can not identify with the stiff-necked concluding statement of King: “In conclusion, Depuydt’s essay does not offer any substantial evidence or persuasive argument, let alone unequivocal surety, that the GJW fragment is a modern fabrication (forgery).”


If a husband were to genetically test his children to determine whether his wife had been faithful, and the tests returned indicating that the children could not conclusively be proven to not be his, would this assure him of his wife’s fidelity? Could he then, based upon these tests, be confident that he had indeed fathered the children? Karen King has produced no new evidence to authenticate this fragment. On the contrary, her prior contentions that the GJW fragment was (1) part of a literary codex and (2) was fourth century are now indefensible. Her method of argumentation was not self-critical or objective, but will doubtlessly be sufficient for those who already want to believe.


One has to ask why Karen King has not published the notorious handwritten note. A typed 1982 note signed by Peter Munro accompanied the fragments which indicated that a Coptic John fragment was among the manuscript group (cf. King, “Jesus said,” 2012, 2). The second notorious handwritten note reads as follows:
“Professor Fecht believes that the small fragment, approximately 8 cm in size, is the sole example of a text in which Jesus uses direct speech with reference to having a wife. Fecht is of the opinion that this could be evidence for a possible marriage.” (King, “Jesus said,” 2014, 153)
Odd, is it not, that Munro mentioned a dime-a-dozen Sahidic manuscript in the typed note, but detailed the GJW in a handwritten note separately?! This handwritten note potentially bears the hand of the forger, who cut the papyrus, falsified the text, and aided its journey with the convenient handwritten note. King’s failure to publish this handwritten note conveniently eliminates a clear avenue for identifying the perpetrator.


  1. An excellent summary,
    thank you!

  2. Rob Marcello4/11/2014 2:59 am

    Thanks Christian, very helpful analysis.

  3. I also find this Coptic manuscript of John that was used for comparative testing quite interesting. Little is said about it, and I don't think it's been published. From n.107 in King's HTR article, it sounds like it was dated paleographically to II-V cent CE but when tested at the same time as GJW it came up VII-IX cent at both Arizona and Harvard.

  4. Waitwaitwait. Click that little chart.

    It looks like one lab says GJW-materials were produced sometime in 680-880 AD. Another lab says GJW-materials were produced sometime in 410-200 BC.

    I must be missing or misinterpreting something. Please explain.

  5. James,
    This is the Ptolemaic date which I mentioned. My initial assessment is that we have here a contaminated sample. Alternatively, if the test were accurate, this results would absolutely inauthenticate the fragment, demonstrating that the forger had reused papyrus from an era predating both written Coptic and Christianity.
    Unfortunately, I had to be brief, and could not go into detail on most issues here! Sorry! Please let me know if I should say more.

  6. Wouldn't contamination give a late date, not an extremely early one? In other words, if a fly landed on the papyrus 500 years after production, died, and somehow became 'part' of the papyrus, it could contaminate the sample, and give a date 500 years later than the real date. It would seem difficult to have 500 year older material than the object being studied accidentally dropped onto it.

    Andrew Wilson

  7. Andrew,
    A small amount of modern organic contamination (say 1% of total specimen) would throw the result toward the modern age by a very small amount. The same amount of radioactive dead contamination (chalk, petroleum, or other subterranean material) could falsely age the sample by a couple of centuries. Therefore, the large gap would be more easily explained by radioactive dead material than by 'modern' contaminates. 14C dating is as much an art as it is a science. The question that remains is whether one or both results for GJW are incorrect. We don't really know the history of the fragment, so the problematic dating for GJW could derive from someone having cleaned the fragment or having stored in a smoke-filled room or who-knows-what. To successfully remove contaminates through cleaning, one ideally should know what exactly needs to be removed.

  8. Thanks for this helpful up-date Christian. It is good to get the whole thing published, even if there are many questions unanswered. It was helpful of Malcolm Choat to address some of the points raised in this blog ( and I accept that photographs can be misleading. (I also note that the whole section on why King was persuaded of the antiquity of the papyrus has been thoroughly re-written.)

  9. I was afraid that the scientific results would be less than conclusive (for either side). I think this shows that scientific results require careful interpretation as well as the philological. Myself, I am struck by the unexpected medieval / Islamic dating.

  10. Christian, could the grammatical error in the text be accounted for by this late medieval dating? I know in other languages common errors emerge when a particular tongue falls into disuse. Is the error and in the text typical of late Coptic texts? Is it possible that this error was made twice by two different individuals?

  11. Excellent summary. Thank you for this!

  12. Stephan,
    Great question! The bread and butter corpus of Coptic manuscripts is the White Monastery of Shenute. The remains of perhaps 300-400 codices now reside at a number of European libraries and holding institutions in a fragmentary condition. We have a smaller number of codices from the Archangel Michael Monastery in Hamuli. In this second case, we have only 47 parchment codices, but they are in perfect condition. Colophons in both collections allow us to date them as groups to the X-XII and IX-X centuries respectively. These libraries basically employ standard Sahidic; the language is intact as a spoken language at least up to the eleventh century. This is to say that ninth century is not late for Coptic, and that the language has not fallen into disuse. In fact, the six to eighth centuries are the heyday for Coptic literature, and this is roughly contemporaneous with the 2-sigma Harvard date (650–870 CE).
    To better answer your question, I should more thoroughly explain the grammatical error which is better understood as an inaccurate cut-and-pasting of Grondin's GThomas PDF. I will try and do this within the next week in a post. If you want to read about this now, read the watershed 2012 article by Andrew Bernhard.

  13. Well, yeah, if one lab says 680-880 AD and another lab says 410-200 BC, we either have a contaminated sample, or two contaminated samples. (Or some other factor like a contaminated machine, but let's go with the usual suspects.) That's a variation, at least, of 880 years! These numbers are completely worthless.

    If the second lab's numbers are right, then the "GJW" is from the late 680's at least. If the first lab's numbers are right, then the thing is even more of a demonstrable forgery than it already is.

    Seems to me like the headline should be, "One of two laboratory dating-tests proves that the fragment is a fake. The other one proves that no one at Harvard knows how to date Coptic script."

    Having laid an egg, so to speak, King & Co. are acting as if it is plaid. Must be Eastertime.

  14. Dr. Depuydt is right. Without a doubt the text is a crude forgery which was carried out on a fragment of ancient papyrus and with ink made from the same ancient components; something very easy to play, because there is a lot of information posted on the components of carbon inks used from the early centuries of Christianity up to the Byzantine times. But the paleographic and grammatical analysis shows, beyond any reasonable doubt possible, that it is a forgery and the more grotesque.

    Here’s the first palaeographic report just a few days since the news came out to the media in 2012:

    Kind regards,

  15. Many thanks again, Christian, for this helpful analysis. Quick comment:

    "Only the Harvard report indicates the date of the test (14 March 2014)"

    King gives the date "June-July 2013" for the Arizona test (132 and 135), unless I am misunderstanding your point.

  16. TBH I am no expert on the authencity of proposed Bible texts or extra Bible Texts. That being said I think its intensely important not to presume any ancient manuscript is real and to study any such finding in fine detail and not presume. It seems that Depyupt did just that leaving nothing for granted. This includes possibility of contamination, the ink and wording Now if the manuscript asked Jesus to friend "his wife" on Facebook you know its a forgery.

  17. Mark,
    Thanks for your correction! Hodgins did not include this date in his article, where I would have expected to find it. I have corrected the main post, and present the original here, "Only the Harvard report indicates the date of the test (14 March 2014); one might surmise that the second test was ordered after the extremely early date arrived from Arizona."

  18. Thanks, Christian. Yes, Hodgins made a mistake in not including the date. There is a much fuller report on the Harvard website at:

    [The PDF link sometimes takes an age to load; if so, here's the linking page: ]

    This report does make clear the dates. The report is dated 19 July 2013, modified 10 March 2014; samples sent to the lab on 29 May 2013.

    Your surmise "that the second test was ordered after the extremely early date arrived from Arizona" appears to be corroborated at least partially by this fuller report. Note in particular this section:

    "The reliability of the Gospel of Jesus Wife radiocarbon measurement is questionable. Three things are at play in the assessment of the current measurement: 1, the abbreviated cleaning protocol; 2, the odd carbon stable isotope measurement; and 3, the early GJW papyrus radiocarbon date." [emphasis added]