Monday, January 13, 2020

Erasmus, the KJV, and the Order of Matt 23.13–14

If you are dutifully reading your Robinson-Pierpont (RP) Byzantine Greek NT alongside your Nestle-Aland or THGNT (as you should be), you will notice that they diverge at Matt 23.13–14. The Byzantine text here reads
13. Οὐαὶ δέ ὑμῖν, γραμματεῖς καὶ Φαρισαῖοι, ὑποκριταί, ὅτι κατεσθίετε τὰς οἰκίας τῶν χηρῶν, καὶ προφάσει μακρὰ προσευχόμενοι· διὰ τοῦτο λήψεσθε περισσότερον κρίμα. 14. Οὐαὶ ὑμῖν, γραμματεῖς καὶ Φαρισαῖοι, ὑποκριταί, ὅτι κλείετε τὴν βασιλείαν τῶν οὐρανῶν ἔμπροσθεν τῶν ἀνθρώπων· ὑμεῖς γὰρ οὐκ εἰσέρχεσθε, οὐδὲ τοὺς εἰσερχομένους ἀφίετε εἰσελθεῖν.

13. Woe to you, scribes and Pharisees, hypocrites! For you devour widows’ houses and for a pretense you make long prayers; therefore you will receive the greater condemnation. 14. Woe to you, scribes and Pharisees, hypocrites! For you shut the kingdom of heaven before men. For you neither enter yourselves nor allow those who would enter to go in.
Matt 23 in the 1611 KJV
The NA27 omits the first woe and numbers the second as v. 13. The KJV has both verses but inverts the order (and the versification) of the RP. (I’ll follow the RP versification rather than the KJV in the rest of this post.)

Normally, I expect the KJV to match the Byzantine textform, so what’s the deal here? It turns out the RP is indeed following the order of most manuscripts whereas the KJV is following most printed editions, editions that go back to Stephanus and before him to Erasmus. For a full list, see Abbot here). Erasmus’s first edition of 1516 has the KJV order. The Complutensian, however, has the reverse. Both orders are found in the two most important editions of Stephanus. In his beautiful 1550 edition, he has the RP order. But, in his 1551 edition, he reverses course, giving the Erasmian order and adds the versification followed by most since then.

Erasmus’s 1516 edition, with v. 14 before 13 in both Latin and Greek (source)
The Complutensian Polyglot has the order found in most medieval manuscripts (source)
Stephanus 1550, showing the RP order
Stephanus 1551, the first to have verse numbers, showing the KJV order and versification

There are some witnesses that have the Erasmian/Stephanus 1551/KJV order, but not many. According to NA27, they are f13 pc it vgcl syc bomss.

So where exactly did Erasmsus get his order? Did he follow one of these manuscripts or did he give this order independently of them? I think it’s the latter. The reason is that none of the Gospel manuscripts he used have both verses in his order. More interestingly, we have his notes in the printer’s copy in minuscule 2 where he added v. 14 in the margin, placing it before v. 13 (still following the RP versification). The manuscripts that are said to be used by Erasmus in the Gospels are 1, 2, 817, and 07 (although this last one is contested). Let’s look at these.

Minuscule 1 with v. 14 but not v. 13 (source)
Minuscule 2, Erasmus’s printer’s copy with v. 13 added in the margin before v. 14.
The red “crayon” is apparently from Froben (
Minuscule 817, showing v. 13 before v. 14. Commentary is in between (source)
Majuscule 07 (E) which some think Erasmus had access to, also with v. 13 before v. 14 (source)
If none of Erasmus’s Greek sources have his order, where did he get it from? We do know he had access to minuscule 69 in Cambridge, which, as a member of family 13, does have Erasmus’s order. But by the time he was printing his edition, he was in Basel. The simpler theory is that he leaned on the Latin. His two main sources gave him the full reading of the Latin, but couldn’t tell him the proper order since neither witness has both verses. So, he did what he did elsewhere: he let the Latin decide. That simple decision is essentially why we have the order found in the KJV. (I should note that I don’t have access to Erasmus’s Annotationes so it’s possible he explains himself there. Update: he does not.)

In any case, this is yet one more place where the TR not only doesn’t reflect modern critical editions, but it doesn’t even reflect the majority of NT witnesses. And all this goes back to a simple choice of Erasmus, one he had to make with very limited resources.

Sources: Jerry H. Bentley, Humanists and Holy Writ: New Testament Scholarship in the Renaissance (Princeton: Princeton University Press, 1983), 129–130; Kenneth W. Clark, “The Erasmian Notes in Codex 2,” in Studia Evangelica, ed. K. Aland, F. L. Cross, et al., Texte und Untersuchengen lxxii (Berlin, 1959): 749–56; William W. Combs, “Erasmus and the Textus Receptus.” DBSJ 1 (1996): 35–53; C. C. Tarelli, “Erasmus’s Manuscripts of the Gospels.” JTS 44, no. 175/176 (1943): 155–62; idem, “Erasmus’s Manuscripts of the Gospels.” JTS 48, no. 191/192 (1947): 207–8.


  1. It should also be noted that William Tyndale followed Erasmus' order. His translation was more than likely the source of the KJV order too.

  2. Dr. Gurry,

    Thank you for the post. This is an interesting case: here we have a small block of Text misplaced (transposed) by a scribal mishap which was caused almost assuredly by HA/HM/HT.

    You write:

    "In any case, this is yet one more place where the TR not only doesn’t reflect modern critical editions, but it doesn’t even reflect the majority of NT witnesses. And all this goes back to a simple choice of Erasmus, one he had to make with very limited resources."

    Indeed, yet in truth; only *some editions* of the TR follow the Latin order here. Even the most extreme TR supporters can still uphold the correct reading here because *both* variations are present within the TR tradition. I think that this is important and often overlooked by TR advocates and opponents alike. Unfortunately, most within the "TR" camp simply follow the KJV regardless of the evidence.

    1. Matthew, how would such TR advocates choose in such a case as this where the printed editions are divided?

  3. Dr. Gurry,

    That's the "million dollar" question, and I'm not sure I can answer it for anyone but myself (but I'll try).

    In a word, evidence. They must follow the TR reading which is backed by the strongest evidence. Unfortunately, the newest form of TR advocacy is nothing more than KJV-Onlyism in a slightly modified and relabeled form (and therefore immune to reason, logic and common sense when variation arises within the TR tradition).

    When one cannot adjudicate between clear cut variant readings like; Luke 2:22, Ephesians 3:9 or Revelation 16:5, but only falls back upon an exaggerated authority given to the KJV--all hope for progress and growth is essentially quenched. So it seems that until "Confessional Text" advocates are willing to entertain that the KJV has errors, there's really nothing to talk about.

    1. An honest answer. I can appreciate that.

  4. Who is saying that Erasmus used 07???

  5. "We do know he had access to minuscule 69 in Cambridge, which, as a member of family 1, does have Erasmus’s order." I think that you mean family 13.

  6. "In spite of Tarelli’s efforts, it is today admitted neither Erasmus nor his team used it [GA 07]." So Andrist, p. 84:
    See also p. 85 (and note 11) about GA 69.

  7. Thanks, Peter!

    The Aussie writer Gavin Basil McGrath gives us a robust analysis.


    Matt. 23:13,14: ... Component 2 (verse order) {B}. p. 231

    While I thank God for the labours of Erasmus ... Stephanus ... the other great neo-Byzantines of the 16th and 17th centuries. p. 232

    For the purposes of much of the textual analysis of Matt. 23:13,14, my twin concerns are the verse order (Component 2, Variant 3) and presence of verse 14 (omitted in Component 2, Variant 4). p. 232

    ... the majority Byzantine text reading; and Green’s Textual Apparatus (1986) says 95-100% of all manuscripts support it. ... p. 232

    Von Soden’s (1913) ... textual apparatus p. 233

    ... (Matt. 23:13a,14a), the TR’s verse order which places Matt. 23:13b before Matt. 23:14b is a minority Byzantine reading found in Lectionaries 673 (12th century, Athos, Greece) and 547 (13th century, Rome, Vatican City State). It is further found as a minority Latin Vulgate Codices reading in Codices F (Fuldensis, 6th century, Fulda, Germany), Q (Kenanensis, 7th / 8th century, Dublin, southern Ireland), T (Toletanus, 8th century, Madrid, Spain), and Th (Theodulfianus, 9th century, Paris, France); and in the margins of Codices S (Sangellensis, 6th century, St. Gallen, Switzerland) and H (Hubertianus, 9th / 10th century, London, UK). It is also found in old Latin Versions b (5th century), ff2 (5th century), h (5th century), l (7th / 8th century), s (7th / 8th century), and c (12th / 13th century). From the Latin support for this reading, it is manifested in the Clementine Vulgate (1592) ... ancient church Latin writer, Hilary (d. 367). ... Greek Novum Testamentum (New Testament) editions of Erasmus (16th century), Beza (16th century), and Elzevir (17th century).

    Looking in overview at the style of address used by Jesus in Matt. 23:2-39, we find that he repeatedly uses a technique of first stating a principle, and then elucidating on it with one or more examples. p. 236

    But when we come to Matt. 23:13-15 in the representative Byzantine text we find a clear and obvious textual problem. The statement of principle in Matt. 23:13b, is sandwiched between the elucidation examples of Matt. 23:14b before it, and Matt. 23:15 after it. Clearly something has gone wrong at this point of textual transmission! The only way to remedy this textual problem is to place Matt. 23:13b before Matt. 23:14b,15. When this is done we find the following stylistic harmony with the rest of Matt. 23. p.238

    (the elephant in the living room.)

    homoeoteleuton - top of p. 241

    The TR’s Component 2 reading is strongly supported by textual analysis, supra. ... weak support in the Greek ... stronger support in the Latin with a minority of Vulgate Codices, and also half a dozen old Latin versions of which half date from ancient times. of the church father, St. Hilary. ... the Greek-Latin editions of Erasmus’s New Testament (1516 & 1522) have the TR’s reading in verse order 13,14, & 15 in both the left hand Greek column and right hand Latin column. ... the neoByzantines of the 16th and 17th century must have reconstructed the completed form of the Greek of the Textus Receptus from the Latin. (continues) p. 242

    The more perfect form of the TR is here found in the Latin alone. (continues) p. 243

    Component 2 (verse order), that correctly follows the TR’s verse order which places Matt. 23:13b before Matt. 23:14b is found in the Family 13 Manuscripts, which contain Minuscules 788 (11th century, independent text), 346 (12th century, independent), 543 (12th century, independent), 826 (12th century, independent), 828 (12th century, independent), 983 (12th century, independent), 13 (13th century, independent), et al. ... some manuscripts of the Egyptian Coptic Bohairic Version. p. 244

    … spread the infection of Origen’s shorter reading around. p. 245


    A Textual Commentary on the Greek Received Text of the New Testament -
    Volume 3 Commentary on Text: Matthew: 23-25


    1. That’s a bit garbled. But the source isn’t much clearer.

    2. Hi Steven,

      Mr. McGrath repeatedly states, "the TR's verse order" and "the TR('s)"; when in fact the TR tradition reads both ways (as far as the longer reading is concerned). How could one label it as, "the TR's verse order" when the TR is in fact split?

    3. Thanks Peter and Matthew,

      Matthew, my understanding with McGrath, "the TR verse order" refers to the final orders of the TR .. as seen in the AV, Elzevir, the Beza 1598. If the variant has a difference in that limited focus, which is quite rare and usually minor, Gavin likely would be more precise. My understanding is that the TR did normalize on the order we see in the AV, thus explaining Gavin's usage of "the TR verse order".

      And I wonder if anyone would comment constructively and interact with what I call his "elephant in the living room", the internal evidence of the pattern of writing in the Mark chapter. This looks to be the decisive element in what might otherwise be a close call.

      And note his additional ms and ECW information that is not given above, including the specifics of the Old Latin, the Vulgate and the ECW.

      And I did have to remove some text to get under the posting limit of 4096. And I plan to put my longer excerpted version online in a bit.

      To me, McGrath shows the major points missed by Peter Gurry, and missed by the general modern textcrit Greek-primacy-only perspective. While I agree that his writing style is unusual, and can be exhaustive, I highly recommend a careful perusal of his writing.

      McGrath's analysis, however, can be discomfiting to the modern way of looking at Erasmus and Stephanus and Beza, as, e.g. pioneered by Jan Krans. While the modern-textcrit-centric analysis can often be helpful, it can also miss the true Reformation Bible dynamic! As in the thread above.

      Please notice how McGrath discusses the symbiotic and coordinating relationship of the Greek and Latin lines :).

      Here are two quotes from Gavin:


      "The importance of the Latin to this reading means that the neoByzantines of the 16th and 17th century must have reconstructed the completed form of the Greek of the Textus Receptus from the Latin. Thus this was an example of the servant maxim, The Latin improves the Greek, dutifully bowing down low to its master maxim, The Greek improves the Latin, since (unlike the old Latin Papists whose textual analysis was based on the Latin text,) any such reconstruction is based on textual analysis of the Greek text, which here supports the reading best preserved in the Latin." - p. 242

      The more perfect form of the TR is here found in the Latin alone. We here see the servant maxim, The Latin improves the Greek, like a good wife who calls her husband, “lord” (I Peter 3:6), being in humble “subjection” (I Peter 3:1) to the master maxim, The Greek improves the Latin; for our textual analysis is based on the Greek which is given the priority, even though in this instance, it is then found from this Greek textual analysis that it is in fact the Latin that preserves the better overall reading, for which reason we here give all due “honour” (I Peter 3:7) to the Latin. - p. 243


      And I understand that it very hard for modern textcrits to really grasp the Reformation savants approach to the Greek and Latin mss. You almost need a tabula rasa for your patterns of thinking!

      Steven Avery
      Dutchess County, NY

    4. Why not just state the obvious:

      If it's isn't in the KJV you reject it; if it is in the KJV you accept it.

      Nothing more, nothing less. Textual Criticism has nothing to do with it.

  8. Hi Steven, thanks for the response.

    You state; "This looks to be the decisive element in what might otherwise be a close call."

    How do you see the evidence as a close call? It appears very lopsided to me. I think the "elephant in the room" explanation is a stretch, and personally it looks like grasping at straws. If the shoe was on the other foot; would you suffer a critic to propose such reasoning as a primary determining factor against your own position? I would hope not.

    The Latin is indeed a voice of antiquity and a vital part of the versional evidence. Although, I don't believe it should be given the amount of weight that Mr. McGrath gives it in this instance (quaint parallels aside). The overemphasis on these points comes across more as an attempt to jockey for position--than a careful (and unbiased) weighing of the evidence. Even so, the retention of the verse is a much more important matter than the issue of order; And on this point I'm sure we can agree.