Thursday, May 12, 2016

135 Years Ago Today

Today marks the 135th anniversary of the publication of Westcott and Hort’s momentous edition The New Testament in the Original Greek. The date was May 12, 1881 (see FB discussion here). Later in the same year they would publish their introduction and appendix justifying and explaining their text critical method.

Though certainly dependent on the work of many pioneers before them, it was especially the clarity and power of the argument in their introduction that established Westcott and Hort’s work as not only the high water mark of 19th century NT textual criticism, but, in many ways, the tipping point for the textus receptus. In 1904 the British and Foreign Bible Society would stop publishing the textus receptus and, even before Westcott and Hort went to press, their edited text was being used by the translators of the Revised Version, itself published just days later. Since then, most English translations of the New Testament have followed an eclectic Greek text.

Today, our critical editions typically vary less from the text of Westcott and Hort than they do from the textus receptus. To some, this might suggest that we have made little progress. From another perspective it is a testament to Westcott and Hort’s skill. As Gordon Fee has said, “If all of this means that we still appear to be crossing the Atlantic in an 1881 ship, it may be that they built them better in those days.”*

In all, work on their Greek New Testament took 28 years. Initially they were much more optimistic about the time such an edition would take. On April 19, 1853 Hort wrote to tell Westcott that the publisher, Macmillan, had agreed to print their edition. Here’s part of that letter:
I promised to let you know Macmillan’s answer about the Greek Testament. It has been slow in coming, but is quite favorable. He fully approves of the plan, but leaves all to our own discretion, saying that they will be delighted to “do their part, that is to say, to take all risk and publish, and push.” Besides describing our plan, I told him of my own schemes of examining and restoring (if possible) the texts of the several versions (beside the Peshito, in which we both hoped to effect something), so soon as I should have learned the languages sufficiently: and likewise my intention to go through such of the earlier Fathers as have been examined but cursorily: and likewise to see whether anything can be gleaned from the Latin or cursive Greek MSS at Cambridge. I told him at the same time that these were schemes which might very probably not be realized; and indeed several times urged that there must be no definite agreement, as we might wish to desist for various reasons; I referred in particular to the change in the Greek Professorship which (if Thompson were elected) might cause a good text to be published by authority, and to swamp our edition. He says that no text is published by the Pitt Press [= Cambridge University Press] under the Greek Professor’s authority; but that the Pitt Press merely reprints Lloyd, and Scholefield published a text of his own. He adds that, if I carry out my scheme, our book will be worth publishing, whatever any one else may do. He likewise want to know whether a MS of Photius on St Paul’s Epistles, which Hardwick quotes in a note to his Church History, would be of any use to you. If I were at Cambridge, I would examine it myself, and probably shall do so when I get there. But meanwhile I must tell him that the chances are very small—certainly not enough to make it worth your while to get a Grace of the Senate. He reports that Lightfoot is very much pleased with the proposal to him, and will evidently do the Lexicon and Grammar, though he has not formally undertaken them: he insists on the utmost silence till the books are ready: and Macmillan promises us the same in the strongest terms.
*Gordon Fee, “The Myth of Early Textual Recension in Alexandria,” in Studies in the Theory and Method of New Testament Textual Criticism, p. 273.


  1. PG: "Since then, most English translations of the New Testament have followed an eclectic Greek text."

    Has anyone made a statistical determination of how "eclectic" the WH compilation actually is, in terms of results? What percentage of its readings are not Alexandrian (or, in Hort's jargon, "Neutral")? More than 1%?

    Or to put the question another way: setting aside the matter of Western Non-Interpolations, if someone set out to create the archetype of the Alexandrian Text, how different would it be from the WH compilation?

  2. May the 12th be with you....

    P.Gurry: "In 1904 the British and Foreign Bible Society would stop publishing the textus receptus"

    And yet the American Bible Society (at least) continued to promote and sell in their catalog at least one TR edition until the appearance of UBS1 in 1966 (I know, since I have a copy that I ordered from the ABS around 1965).

    Snapp: "If someone set out to create the archetype of the Alexandrian Text, how different would it be from the WH compilation?"

    I would suggest at those many places where Aleph or B or both depart from the bulk of the remaining Alexandrian witnesses that the latter in such cases would represent a more likely Alexandrian archetype. Thus there would be some significant differences from the WH text at those locations.

    1. Maurice,
      I rephrase: regarding Matthew-Jude: if one were to set aside all parts of NA27's text where it agrees with Byz, and then set aside all parts of NA27's text where its text is supported by B, or Aleph, or both, then, how much text would remain? (Perhaps some Logos-user has an easy way to find out.)

    2. Ok, those data are easy enough for me to access, since I tabulated them long ago. The answer is that not much would be left (no surprise there, right?).

      My totals show that in the NA27 edition, the main text lacks precise support from either Aleph or B in 150 variant units (Byz being regarded as either the unified Gothic M or as "pm" divided among 2 or more readings; or in Revelation as divided between Ma and Mk).

      Not surprisingly, nearly half of these (73) reflect variant units in Revelation where Aleph does not support the NA27 main text (B in that book being totally absent).

    3. Addendum as a caveat: as I noted previously, those data do not imply that all remaining readings supported by Aleph and/or B necessarily would represent the Alexandrian archetype, particularly in situations where the bulk of the remaining Alexandrian witnesses depart from Aleph and B.

  3. Maurice,
    Do I understand correctly that, if your calculations are correct, the total number of readings in NA27's text that deviate from both B and Aleph in Matthew-Jude is 77? That seems impossible.

    What an eclectic text!

    1. Some clarifications are in order: the data I tabulated includes only those variant units where the NA27 main text differs from the unified Byzantine (Gothic "M"). And yes, already included in those tabulations are those variant units where B is not extant but where NA27 does not follow Aleph.

      More importantly,there definitely are other variant unit locations where NA27 and Byz are in agreement, but where Aleph and B jointly differ from such -- these have not been tabulated as yet, but also likely represent a small number of variant units.

      Further, locations exist where the Byzantine text is divided ("pm" in the NA27 apparatus) in which Aleph and B might both depart from the NA27 main text -- once more, these have not yet been tabulated.

      As a result, I would expect there to be perhaps another hundred or so instances where NA27 does not follow Aleph or B.

      Eventually I will rosin up my bean-counter machine, and see what results; right now I can only surmise.

  4. That 77 number, apparently does not include the reduction for the places where Vaticanus is lacuna.

    And in many of those, Sinaiticus could be singular or so oddball that there really was no choice but to go to other manuscripts. So the 77 figure is going to be subject to significant reduction after what I would see as a proper adjustment for the places where there is no Vaticanus reading.

    Steven Avery

  5. To be more precise, Revelation was covered, so the reduction would be the places in the latter third, approximately of Hebrews, 1 and 2 Timothy, Titus and Philemon where there is no Vaticanus reading.

    The most important number would be the total number of such places where Sinaiticus is not the NA27 reading, which would be a part of the 77.

    On a secondary basis it might be nice to examine the group for the nature of the divergence of Sinaiticus and NA27 in those spots, however simply seeing how many of the 77 are in that spot would be interesting.

    One other point would be interesting. In the remaining group (now we are getting to a small number) of true divergences of Sinaiticus and Vaticanus from NA27, how many of them represent spots where the NA27 text differs from the original Westcott-Hort recension text.

    Thanks for your research and numbers on this, especially to Professor Maurice Robinson, and also the astute comments of James E. Snapp, Jr.


    Outside of all this, may I suggest we reconsider whether, in looking at the anniversary, there really is much "clarity and power of the argument in their introduction". Or in the logic of the presentation. Maybe some quotes with support would be helpful. Personally, I have missed this "clarity and power" but am open to having it attempted to be represented.

    Steven Avery