Wednesday, April 18, 2018

Two important, shorter Byzantine readings in 1 John

In reading through 1 John with my Greek students this semester, I noticed two unexpected variants. They are both places where the Byzantine majority preserves a shorter reading that is easily explained as an accidental omission.

The full list of Greek data for 1 John 2.23 is in Text und Textwert, but the evidence from ECM is:
  1. πᾶς ὁ ἀρνούμενος τὸν υἱὸν οὐδὲ τὸν πατέρα ἔχει, ὁ ὁμολογῶν τὸν υἱὸν καὶ τὸν πατέρα ἔχει.
    01. 02. 03. 04. 025. 044. 5. 33. 61. 94. 104. 206. 218. 252. 254. 307. 321. 323. 326. 378C. 398. 429. 436. 442. 453. 459. 467C. 468. 522. 614. 621. 623. 630. 720. 808. 918. 996. 1067. 1127. 1243. 1292. 1359. 1409. 1448. 1490. 1505. 1523. 1524. 1563. 1611. 1661. 1678. 1718. 1735. 1739. 1751. 1799. 1831. 18372. 18382. 1842. 1844f. 1852. 1881. 2138. 2147. 2200. 2298. 2344. 2374. 2412. 2464. 2541. 2544. 2652. 2805. 2818. L596. L1281. Ath. Cyr. CyrH. Or. K:S>BV>. S:P>H. A. G:A1. Sl.Si. Ä
  2. πᾶς ὁ ἀρνούμενος τὸν υἱὸν οὐδὲ τὸν πατέρα ἔχει.
    6. 81. 88. 181. 378*. 467*. 629. 642. 915. 945. 1241. 1875. 2186. 2243. 2492. Byz [424*. 424C2]. PsOec. K:Bms. Sl:ChMS
While the minuscules are not unified here, there is still a clear Byz text identified by the ECM. Because of this unity, you will not find this variant in Robinson-Pierpont as a Byzantine variant though it is in the apparatus as an NA27 reading. The obvious explanation for the second reading is, of course, homoiteleuton (ἔχει ... ἔχει).

By way of illustration, here is the correction of the text in 424 adding the text back in followed by a second correction expunging it.

The double correction in 424. See in VMR
The second such omission is just a few verses later in 1 John 3.1. There the main evidence is
  1. Ἴδετε ποταπὴν ἀγάπην δέδωκεν ἡμῖν ὁ πατήρ, ἵνα τέκνα θεοῦ κληθῶμεν, καὶ ἐσμέν.
    01. 02. 03. 04. 025. 044. 5. 6. 33. 81. 94. 104. 206. 307. 321. 323. 378. 398. 4242. 429. 436. 442. 453. 459. 467. 522. 614. 621. 623. 629. 630. 918. 945. 996. 1067. 1127. 1243. 1292. 1409. 1490. 1505. 1523. 1524. 1611. 1735. 1739. 1799. 1831. 1838. 1842. 1844. 1852. 1881. 2138. 2147. 2200. 2298. 2344. 2374. 2412. 2464. 2541. 2652. 2805. 2818. L596. L:VT. A. G:A1. Sl:ChMSi
  2. Ἴδετε ποταπὴν ἀγάπην δέδωκεν ἡμῖν ὁ πατήρ, ἵνα τέκνα θεοῦ κληθῶμεν.
    61. 88. 181 . 218. 254. 326. 642. 808. 915. 1359. 1448. 1563. 1718. 1837. 1875. 2186. 2243. 2492. Byz [424T]. PSOeC. L:Vms. K:Sms>. Sl:S
Once again, we have the earliest evidence, several dozen minuscules, and most of the versions in favor of the longer reading and the Byzantine manuscripts in favor of the shorter. This variant won’t show up in the RP as an intra-Byzantine variant either. Again, the simplest explanation for the Byzantine reading is homoioteleuton, the eye skipping from -μεν to -μεν.

Klaus Wachtel (Der Bzyantinische Text, 302–303) also suggests that the shorter reading would be preferable because it removes the abrupt shift from subjunctive (κληθῶμεν) to indicative (ἐσμέν) following ἵνα. Confirming this as a possible motive is the fact that we find the subjunctive ὦμεν in 2544 and this appears to be what is translated by the Harklean Syriac and some Coptic witnesses. 

Here is this variant again in 424 showing another correction.

1 John 3.1 in 424. See in VMR
Both variants are pretty easy to deal with for reasoned and thoroughgoing eclectics and pretty difficult for Byzantine prioritists. It may be surprising to see the Byzantine tradition preserve such obvious mistakes, but in this, it also shows how careful the Byzantine scribes often were. It also suggests that, in some cases, the Byzantine text goes back to a single exemplar that is not the autograph and not in agreement with our earliest extant Greek witnesses. These two cases also illustrate well the reality that no single text-type or manuscript has a corner on the original text all the time. In other words, why I’m a reasoned eclectic.

Update (11/29/18): here is further info on the interesting way that the 1611 KJV handled 1 John 2.23.


  1. Great examples! Homoioteleuton seems like the simplest explanation for the shorter readings, and it's interesting to see that in both cases the longer reading has the support of a good handful of minuscules from diverse backgrounds in addition to the majuscules.

    The "double correction" in 424 is a bit confusing to me. The corrective addition does not result in the longer reading ("the one confessing the Son also has the Father"), but simply rephrases the first half ("the one not confessing the Son neither has the Father"). Is the expunging of the correction indicated by the dots all over the addition in the margin?

    1. Yes, the marginal correction in 424 has a negated form of the longer reading. The ECM marks that as a distinct reading here but I left it out for sake of space. And yeah, the dots around the marginal note are marking expunction.

  2. Of course, NA28 also preserves many readings "easily explained as an accidental omission." Are we to appeal finally to transcriptional error everywhere this is possible, or are there mitigating factors?

    1. I don’t see any mitigating factors here. But yes, I generally prefer the longer reading when homoioteleuton is the obvious explanation for the shorter reading. And I am willing to side with Byz when that happens. I’ve discussed some of these before (e.g., Matt 18.29 and Mark 11.26; Matt 20.16; Eph 5.30). And I could add Matt 19.9 and maybe Rom 3.22.

    2. The first instance can be explained as an explication consciously or unconsciously assimilated to John's pronounced style. While dealing in counterfactuals is risky, the long reading here would I think have been rejected or a subject of dispute by eclectics were it not for its presence in certain MSS. The second long reading could also be an explication, not wishing sonship to be a mere potentiality. You did mention another line of evidence beyond transcriptional error here, which I must acknowledge. My point is that we should not appeal only to transcriptional error only in the places where such an appeal falls in line with out favored theory.

    3. To quote Maurice Robinson quoting a favorite philosopher, “What if there were no hypotheticals?” Seriously though, I don’t follow you. Who are you saying is appealing to transcriptional error only when it falls in line with their favored theory?

    4. No one here, since you say that there are other factors. Initially this post seemed to be saying, "These Byz readings seem to be transcriptional errors, therefore they are."

  3. In 1885, Spurgeon broke from his usual practice of preaching from the Authorised Version to preach a sermon on καὶ ἐσμέν in 1 John 3:1: "'And We Are’: A Jewel from the Revised Version" MTP 32.673–84.

    1. Oh nice. Didn't know about that one. It is interesting to me that I've never seen these two variants, especially the 1 Jn 3.1, discussed in books about how theologically corrupt the NIV is. Here it seems that the KJV is the translation leaving out something quite theologically important.

    2. Or translating a reading which makes something implicit explicit. I do, of course, grant that KJV advocates cherrypick examples that could be taken as evidence of their argument.

    3. Elijah H,
      That sermon of Spurgeon's is at .

  4. Still good critics (such as Mill) have held the addition in 1 John 2:23 to be an explanation that crept into the text, and even Mill's uncharitable adversary Whitby, agreed on this point, suggesting its origin in John 8:19, to which Origen also links this passage. For 1 John 3:1, Bengel and others have suggested that the addition arose from the following verse. The Byzantine text may well preserve some "obvious mistakes," but the above-mentioned critics, while certainly aware of homoioteleuton and the patristic and versional evidence, did not judge most mss to be obviously corrupt in those places. And, I suppose, if one were to follow the above critics' decisions, one could "suggest" that those earlier mss, versions, and fathers that have the additions there actually all derived from a single corrupt source that is posterior to the source of most manuscripts. But then again it would only be a suggestion, right?

    1. I’m not surprised at all that Mill and Bengel would defend what was the traditional reading for them. But their explanations aren’t very satisfying. John 8.19 is not a good verbal parallel and an addition in 1 John 3.1 would lead us to expect the subjunctive not the indicative. The simplest transcriptional explanation by far is homoioteleuton. We may not like the implications of this evidence but that obviously isn’t a good reason for rejecting it.

    2. Whitby probably only mentioned John 8:19 in relation to 1 John 2:23 because Origen himself did so. And so whether one judges it a fitting parallel or not, when perhaps the greatest ancient practitioner of textual criticism does something, that is enough to suggest that others even earlier may have done so as well.

      By the way, I would not call the absence in NA 1 Pet 4:14 of κατα μεν αυτους βλασφημειται κατα δε υμας δοξαζεται an "obvious" but a possible error (h.t. -ετατ...-εται), and thus any implication drawn from that would be measured. When some of the greatest ever in the discipline disagree on the direction of change in some passage, it may be premature to draw implications from it and phrase them as seemingly incontrovertible. That's all.

    3. Paolo Trovato4/24/2018 6:28 am

      “When some of the greatest ever in the discipline disagree on the direction of change in some passage, it may be premature to draw implications from it and phrase them as seemingly incontrovertible.”
      I would like to start from this sentence to put in common a couple of very general questions, which have a very low connection with their starting point. 1.Generally speaking, I like the medieval idea that we are dwarfs on the shoulders of giants. As a consequence I think that nowadays we can see more things far removed from the point we are looking from (at least the scholars with more expertise master powerful updates on editorial methodology). 2.I also think that the search for “consensus omnium” can not paralyse a discussion for centuries. In astronomy, most of us think that the geocentric model of the Ptolemaic system is a superseded description of the universe with Earth at the center. Do we agree? Still, and rightly so, Ptolomeus is considered one of the greatest in his discipline. If not the laymen, but the best scholars are uncertain and the rate is 60 % against 40 % it is certainly “premature to draw implications”, but if after the spreading of a new “theory” the most influent scientists who favor it are 95 % against 5 %, I incline to think that 5% is not so relevant. In hard sciences libraries they simply sell off the books of 25 years ago. I don’t like at all selling off old books, but I think (with Kuhn) that paradigms change.

    4. Paolo,
      If I read you correctly, the Textual Critics, particularly the most informative, of the past should not be so easily dismissed just because they are ‘out of date’. In fact, even when a new ‘theory’ has won the day, we still need to recognize their achievements. As always, your words express much needed wisdom. It is an easy thing to view the latest method as a ‘new and better way’ without remembering the shoulders upon which it was built.

    5. Paolo Trovato4/24/2018 8:12 pm

      Thank you, Tim. Indeed, for me both matters are complicated. I am curious to know how other colleagues feel about them

    6. For what it's worth, I would partly agree and partly disagree. Partly agree: Just because there is not unanimity on a point does not mean that research cannot proceed. Partly disagree: Textual Criticism is not a hard science.

    7. Paolo Trovato4/27/2018 9:44 pm

      Thanks, Stephen. I am looking forward to getting colleagues' opinions

    8. Matthaei's editions of the Greek NT (1782-1788, 1803-1807) are still valuable for modern TC. His printed text is "of little value because it is based on manuscripts if recente date, but his apparatus is valuable" (so, Metzger, Text). Nevertheless Mathaei presented the longer text in 1 John 2.23. His comment on 23b in both editions still deserves to be read.

  5. Certainly not impressed by the Greek text form of the Textus Receptus was Libertus Fromondus (Libert Froidmont, 1587-1653) in his "Commentaria in omnes B. Pauli Apostoli, et septem canonicas aliorvm Apostolorvm Epistolas", on 1 John 2,23.
    See Migne's "Scripturæ Sacræ cursus completus", vol. 25, col. 900. :
    Hanc particulam, 'qui confitetur', etc., Graeca hodiè non habent; sed tamen in quibusdam vetustis codicibus Graecis reperitur, et Syrus interpres eam legit. Unde Beza ex quatuor manuscriptis, ut fatetur, illam in textum suum restituit; ut hinc pateat Latinum nostrum textum esse in quibusdam Graeco vulgari sinceriorem; contra id quod haeretici clamare solent de sinceritate moderni Graeci textûs supra Latinum nostrum vulgarem.

    1 John 2,23 is one of Reuss's 'Teststellen' in his "Bibliotheca". Beza introduced the longer text in his folio of 1582. For Beza's influence see the text of the English AV of 1611 and the note in the Dutch States version of 1637.

  6. Paolo Trovato4/29/2018 6:48 pm

    Thanks, Teunis. (By the way, I greAtly appreciated your thoughts on PA and lectionaries: "Once again: floating words, their significance for textual criticism," New Testament Studies, 1995).

  7. What's interesting to me are all the exceptions to what one would expect, that any given manuscript would be consistent in transmitting both readings, or neither.

    MS that include the first long reading, but omit the second:
    61. 218. 254. 326. 424*C2 808. 1359. 1448. 1563. 1718. 1837C.

    MS that include second long reading, but omit the first:
    6. 81. 378*. 467*. 629. 945.

    MS hiant for the second variant:
    252. 720. 1241. 1490. 1661. 1678. 1751. 2544.

    MS that omit both long readings: 88. 181. 642. 915. 1875. 2186. 2243. 2492.

    MS that include both long readings:
    01 02. 03. 04. 025. 044. 5. 33. 94. 104. 206. 307. 321. 323. 378C. 398. 424C1, 29. 436. 442. 453. 459. 467C. 468. 918. 996. 1067. 1127. 1243. 1292. 1409. 1490. 1505. 1523. 1524. 1611. 1735. 1739. 1799. 1831. 1838C. 1842. 1844. 1852. 1881. 2138. 2147. 2200. 2298. 2344. 2374. 2412. 2464. 2541. 2652. 2805. 2818

    I imagine this is a feature of only listing 'interesting' mss in the apparatus while consigning all the 'boring' (thus consistent in omission) mss to 'Byz'.