Evangelical Textual Criticism

Tuesday, November 29, 2005

Internal evidence in a text-critical edition of the GNT

Quote: "I think the aim of the NA28 in deleting such notes [i.e. on conjectural emendations - GV] would be that it is a text based on the documentary evidence for the NT text, not for the history of the study of the NT." (Peter Head, under "Conjectural emendation")

Should the printed text (including text in apparatus) of a text-critical edition be based solely on the documentary evidence (= external) or on internal evidence (intrinsic + transcriptional) as well? If internal evidence is to be excluded, references to parallel passages (as in the NA app. with Col. 1:14) should also be left out.


  1. Of course internal evidence is included, but a textual conjecture is not evidence.

  2. A parallel passage is not evidence either. It is drawing the attention to what may have happened.

  3. Of course we may differ in our understanding of 'evidence', but a parallel passage may be understood as a constituent part of circumstantial evidence that may be adduced in support of a hypothesis. A conjecture is a hypothesis itself.

  4. But at the end of the day I do think that the registering of parallel passages in an apparatus may distort evaluation of variants if editors have an excessive preference for harmonistic explanation of variants.

  5. Pete, Your formulation "excessive preference for harmonistic explanation of variants" is interesting. However, you're using the term "excessive" with your own idea about an acceptable measure of harmonistic explanation of variants as your norm.

  6. I like as many data as possible in a Greek NT. Parallels that may be considered by some to be the source of a reading, should be provided, even if the editor has another idea about the genesis of that reading. It's like listing the options. A conjectural emendation may be (in a few exceptional cases) another option.

  7. I have no particular definition of 'excessive', though I could envisage situations in which the information supplied in an apparatus led readers to prefer particular options in a way lacking warrant.

    Yes, parallels are certainly more useful than conjectures. I like as much information as can be fitted into the space, but would prefer to have information from antiquity over information from after the printing of the GNT. I see the editor's tasks as to present information from the manuscript transmission period to the readers. Parallels also existed in this period. However, look at the various guesses that NA27 records for 2 Peter 3:10. There is no evidence that any of them existed before their comparatively recent invention by modern authors. This is one of the places in the NT said most to need conjectures, yet none of the conjectures even approaches the readings in the mss in likelihood.

    Space in the apparatus could better be given to telling us whether there are other places where the Sahidic adds a negative or whether there are other texts where P72 adds a word.

  8. "Space in the apparatus could better be given to telling us whether there are other places where the Sahidic adds a negative or whether there are other texts where P72 adds a word."
    I agree.
    But not the parallels.
    In an edition I look for information that is not readily available already.
    The relevant parallels most scholars can find quickly, but who will find these emendations (or is informed about addition of a negative in the Sahidic)?

  9. The real function of the reference is to explain the editor's opinion that the variant arose through harmonisation to a parallel text. In the gospels it is enough just to give "p)"; elsewhere when it may not be obvious the reference is given.