How far can a textbook go? The case of Metzger’s Text of the New Testament

This afternoon I was reading in Leonard Whibley, ed., A Companion to Greek Studies (Cambridge: University Press, ¹1905), and more in particular R. C. Jebb’s contribution, “Textual Criticism,” (pp. 610–623). When I came to p. 621, § 695, “Modern use of conjecture,” I was in for an unpleasant surprise, for parts of the text were already familiar to me.

It turns out that Metzger, in his The Text of the New Testament. Its Transmission, Corruption, and Restoration (Oxford: Clarendon, ¹1964), pp. 182–183, took over key sentences from Jebb's section, as can be seen below (in blue and green). I do not think this practice was acceptable or common in 1964, so I suspect something went wrong here. The general question is how much borrowing can be expected and accepted in introductory texts.

Two further points are of interest as well: (1) in between stands a paragraph on Bentley, for which Metzger refers to another publication by the same Jebb; Jebb’s section itself is listed on p. 156 (n. 1); (2) in the example of conjectures on Shakespeare, a footnote does warn the reader that the “example is taken nearly verbatim” from another source (see the grey passages below for an impression of what verbatim means). Here at least is a disclaimer, and though I do not think we would like to accept such use of other sources, it demonstrates how problematic the use of Jebb’s words actually is.

Finally, the passages are still found in the same form in the fourth edition by Metzger and Ehrman (Oxford: Oxford University Press, ⁴2005, pp. 226–228). I would suggest a revision for the fifth edition, and perhaps a more thorough check of various other passages as well.

Metzger, Text, 11964 (= Metzger & Ehrman, 42005)	Source: Jebb in Whibley, Companion, 11905, p. 621
182 Modern Methods of Textual Criticism VII. CONJECTURAL EMENDATION The classical method of textual criticism regularly involves, as was mentioned at the beginning of this chapter, the exercise of conjectural emendation. If the only reading, or each of several variant readings, which the documents of a text supply is impossible or incomprehensible, the editor’s only remaining resource is to conjecture what the original reading must have been. A typical emendation involves the removal of an anomaly. It must not be overlooked, however, that though some anomalies are the result of corruption in the transmission of the text, other anomalies may have been either intended or tolerated by the author himself.1 Before resorting to conjectural emendation, therefore, the critic must be so thoroughly acquainted with the style and thought of his author that he cannot but judge a certain anomaly to be foreign to the author’s intention. This aspect of criticism has at times been carried to absurd extremes. In his later work Richard Bentley, for example, largely disregarded the evidence of manuscripts in determining the correct readings, and depended chiefly upon his own instinctive feeling as to what an author must have written. He justified such a procedure in the magisterial phrase, nobis et ratio et res ipsa centum codicibus potiores sunt, which may be rendered ‘for me both reason and the subject-matter are worth more than a hundred manuscripts’. In following this bold principle he did much that was rash and indefensible as well as much that is brilliant and convincing. The reductio ad absurdum of such a subjective method is found in Bentley’s edition of Milton’s Paradise Lost, in which he offers more than 800 emendations, restoring what in his opinion Milton must have really said (or meant to say) while dictating the poem to his daughters.2	695. If the only reading, or each of several readings, which our documents supply is seen to be impossible, then the remaining resource is conjectural emendation.
Before a conjecture can be regarded as even probable, it must satisfy the two primary tests which are customarily applied in evaluating variant readings in manuscripts: (1) it must be intrinsically suitable, and (2) it must be such as to account for the corrupt reading or readings in the transmitted text. There 1 For a discussion of the paradoxical possibility of a textual critic’s ‘improving’ on the original, see G. Zuntz’s article on 1 Cor. vi. 5 entitled ‘The Critic Correcting the Author’, Philologus, xcix (1955), pp. 295–303. 2 See James Henry Monk, The Life of Richard Bentley, D.D. , 2nd ed., ii (London, 1833), pp. 309–23, and Richard C. Jebb, Bentley (London, 1889), pp. 18o–91.	Before a conjecture can be regarded as even probable, it must satisfy the two primary tests which we apply to doubtful readings of mss.: (1) it must be intrinsically suitable: (2) it must be such as to account for the corrupt reading or readings in the transmitted text. There
Modern Methods of Textual Criticism 183 is, however, an important difference between the method of applying these tests to a conjectural emendation, and that of applying them to variants in manuscripts. We accept the variant which best satisfies the tests; but we require of a successful conjecture that it shall satisfy them absolutely well. The conjecture does not rise from a certain level of probability (‘a happy guess’) to the level of certainty, or approximate certainty, unless its fitness is exact and perfect. The only criterion of a successful conjecture is that it shall approve itself as inevitable. Lacking inevitability, it remains doubtful.	is, however, one important difference between the method of applying these tests to a conjectural emendation, and that of applying them to variants in mss. We accept the variant which best satisfies the tests; but we require that the conjectural emendation shall satisfy them absolutely well. The conjecture does not rise from probability to certainty, or approximate certainty, unless its fitness is exact and perfect.
	Source: Gow, Companion, pp. 65–66
An example from English literature will illustrate the wide differences of merit among proposed conjectures.1 Since the early printers in England were often foreigners, who made quite as bad mistakes as their predecessors the scribes, the text of Shakespeare contains almost as many problem passages as that of Aeschylus. In the folio editions of Henry V, Act ii, scene iii, the hostess says of the dying Falstaff, ‘his nose was as sharp as a pen and a table of Green Fields’. The words ‘a table of Green Fields’, which appear with trifling variations of spelling in the folio editions but which are omitted in the quarto editions, have been the subject of numerous conjectural emendations. Pope suggested (perhaps ironically) that this was a stage direction to bring in one of Greenfield’s tables, Greenfield being supposed to be the furniture-dealer who supplied props for Shakespeare’s theatre. Collier proposed ‘on a table of green frieze’, and another critic suggested ‘or as stubble on shorn fields’. The conjecture which today is adopted by editors is ‘and a’ babbled of green fields’, being a modification by Theobald of a happy proposal made by an anonymous annotator who corrected ‘a table’ to ‘a’ talked’.2 The fault most often committed in the use of conjectural 1 This example is taken nearly verbatim from James Gow’s Companion to School Classics, 2nd ed. (London, 1889), pp. 65 f. 2 Several passages in Shakespeare are corrupt beyond the ingenuity of palaeographer and textual critic to propose a cure. Apart from lucky coincidence, what lay behind the hodgepodge of nonsense set by the compositor of the first quarto of King Lear in iii. iv. 118 ff. is probably unattainable: ‘swithald footed thrice the old a nellthu night more and her nine fold bid her, O light and her troth plight and arint thee, with arint thee.’ On the special problems involved in the textual criticism of Shakespeare’s works, see Madeleine Doran, ‘An Evaluation of Evidence in Shakespearean Textual Criticism', English Institute Annual, 1941 (New York, 1942), pp. 95-114, and F. P. Wilson, ‘Shakespeare and the “New Bibliography” ’, in The Bibliographical Society, 1892–1942, Studies in Retrospect (London, 1945), pp. 133–4.	In England also the early printers, who were mostly foreigners, made quite as bad mistakes as their predecessors the scribes, and the text of Shakspere contains almost as many hopeless difficulties as that of Aeschylus. One example will suffice to illustrate this fact and to show the wide difference of merit in conjectures. In Henry V., act ii. sc. 3, the hostess says of the dying Falstaff, ‘his nose was as sharp as a pen and a’ babbled of green fields.’ The words italicised are omitted in the quarto editions, but are printed in the folios (with trifling variations of spelling) ‘and a table of Green Fields.’ Pope suggested (perhaps ironically) that this was a stage direction to bring in one of Greenfield’s tables, Greenfield being supposed to be the furniture-dealer who supplied Shakspere’s theatre. Mr. Collier proposed “on a table of green frieze,” another critic suggested “or as stubble on shorn fields.” The reading “a’ babbled,” which is now universally adopted, is Theobald’s, but it was first suggested by an anonymous annotator, who corrected “a table” to “a’ talked.” The emendation is a very beautiful example of the critical art.

Update 20 September 2022

I found some further coincidences in one of the passages shown above, namely in Benjamin Breckinridge Warfield, An Introduction to the Textual Criticism of the New Testament (London: Hodder & Stoughton, 1886).

Metzger, Text, 11964 (= Metzger & Ehrman, 42005)	Source 1: Jebb in Whibley, Companion, 11905, p. 621 Source 2: Warfield, Introduction, p. 209
182 Modern Methods of Textual Criticism Before a conjecture can be regarded as even probable, it must satisfy the two primary tests which are customarily applied in evaluating variant readings in manuscripts: (1) it must be intrinsically suitable, and (2) it must be such as to account for the corrupt reading or readings in the transmitted text. There 1 For a discussion of the paradoxical possibility of a textual critic’s ‘improving’ on the original, see G. Zuntz’s article on 1 Cor. vi. 5 entitled ‘The Critic Correcting the Author’, Philologus, xcix (1955), pp. 295–303. 2 See James Henry Monk, The Life of Richard Bentley, D.D. , 2nd ed., ii (London, 1833), pp. 309–23, and Richard C. Jebb, Bentley (London, 1889), pp. 180–91.	Before a conjecture can be regarded as even probable, it must satisfy the two primary tests which we apply to doubtful readings of mss.: (1) it must be intrinsically suitable: (2) it must be such as to account for the corrupt reading or readings in the transmitted text. There
Modern Methods of Textual Criticism 183 is, however, an important difference between the method of applying these tests to a conjectural emendation, and that of applying them to variants in manuscripts. We accept the variant which best satisfies the tests; but we require of a successful conjecture that it shall satisfy them absolutely well. The conjecture does not rise from a certain level of probability (‘a happy guess’) to the level of certainty, or approximate certainty, unless its fitness is exact and perfect. The only criterion of a successful conjecture is that it shall approve itself as inevitable. Lacking inevitability, it remains doubtful.	is, however, one important difference between the method of applying these tests to a conjectural emendation, and that of applying them to variants in mss. We accept the variant which best satisfies the tests; but we require that the conjectural emendation shall satisfy them absolutely well. The conjecture does not rise from probability to certainty, or approximate certainty, unless its fitness is exact and perfect.
	No conjecture can be accepted unless it perfectly fulfil all the requirements of the passage as they are interpreted by intrinsic evidence, and also perfectly fulfil all the requirements of transcriptional evidence in accounting for the actual reading, and if variants exist also for them (either directly or mediately through one of their number). The dangers of the process are so great that these rules are entirely reasonable, and indeed necessary. The only test of a successful conjecture is that it shall approve itself as inevitable. Lacking inevitableness, it remains doubtful.

Stendahl and the End of Romans 7

While preparing a course, I came across Stendahl’s complaint (in 1963) that the Nestle editions treat Rom 7:25b as belonging to Rom 8. Which made me wonder how that looks on the page, and whether all Nestle editions do so. Stendahl of course used NA25 (1963) or an earlier edition, not NA26 (1979) or later.

It turns out the subdivision was introduced in the 10th Nestle edition (1914), as the bottom of p. 405 may show:

For comparison the same portion in the 9th edition (1912):

Interestingly, NA26 and NA27 (1993) have a major division between 7:25 and 8:1, but still set 7:25b apart as a subparagraph.

In doing so, these two editions draw even more attention to the conjecture according to which 7:25b is an interpolation. This conjecture was first mentioned in the 16th edition (1936), and attributed to the Dutch pastor Michelsen (1879), but there happens to be an earlier author, namely Peter Aloys Gratz in 1814. It has found support among a good deal of Pauline scholars.

In NA28 (2012), the conjecture is no longer mentioned, and there is no subdivision any more between 7:25a and 7:25b. The only trace still left of these typographical wanderings is the capital letter of the first word of 25b, Ἄρα.

See Krister Stendahl, “The Apostle Paul and the Introspective Conscience of the West,” HTR 56 (1963): pp. 199–215, p. 213 n. 30 (= “The Apostle Paul and the Introspective Conscience of the West,” in Paul among Jews and Gentiles and Other Essays [Minneapolis, MN: Fortress, 1976], pp. 78–96, p. 94 n. 20).

A slightly different wording

As happy owner of a first edition of Metzger's The Text of the New Testament (1964), I sometimes come across text-critical phenomena in the book itself.

A few years ago, at the ETC blog, Tommy already mentioned the beautiful error "electric" for "eclectic," in "Its text … is curiously electric, …" (p. 103).

Today I spotted another intriguing error, on p. 193: "The assimilation of the wording of one passage to the slightly different wording form in a parallel passage, …"

Obviously "form" is wrong here (though English is a strange language), but what is the correct reading? I must be a lousy conjectural critic, for I guessed "wording found in." After all, what did compositors not do?

The true reading emerged in the second edition (1968): "wording in," simply omitting "form." So "form" must have been a gloss on "wording," that should never have been part of the text. In conclusion, the 1964 sentence offers an nice example of conflation.