[1] Metzger, Bruce M., A Textual Commentarv on the Greek New Testament (London and New York: United Bible Societies, 1971).Google Scholar

[2] References here are to the third edition (London and New York: United Bible Societies, 1975), edited by Kurt Aland, Matthew Black, Carlo M. Martini, Bruce M. Metzger, and Allen Wikgren.

[3] See, for instance, the comments on John 20. 11, Rom. 12. 11, Col. 1. 12 (ίκανώσαντι), Heb. II. 11, and Rev. 14.6.

[4] Such leaps can usually be classified as homoeoarcta or homoeoteleuta, but these are in fact just special cases of the general error of a scribal leap from the same to the same (or to the similar). Cf. the comment by Louis, Havet, Manuel de critique verbale appliquée aux textes latins (Paris: Hachette, 1911), §456, and our note 39 below.Google Scholar

[5] Commentary, p. xxvii.

[6] E.g. Matt. 10. 37, Mark 8. 38 (λόγονς), Luke 1. 68, Acts 17. 13, Heb. 9. 19, 1 John 2. 23 (where no apparatus is given in the edition), and Rev. 6. 2.

[7] The low rating here might be due in part to the fact that A has a further addition, even though Metzger easily explains that as assimilation to Col. 3. 16. J. K. Elliott has argued that the shorter text is correct here, and that the UBS text has also arisen by assimilation to Col. 3. 16, a possibility which Metzger mentions; cf. ‘The United Bible Societies’ Textual Commentary Evaluated’, N.T. XVII (1975), p. 145.

[8] The longer text has been defended by Kilpatrick, George D., ‘The Greek New Testament Text of Today and the Textus Receptus’, The New Testament in Historical and Contemporary Perspective, ed. Hugh, Anderson and William, Barclay (Oxford: Basil Blackwell, 1965), p. 195. Incidentally, the reading τη¯ γυναικι clearly results from assimilation to the parallel at Matt. 19. 5. Conversely, the usual (as in UBS) text at Matt. 19. 5 has been assimilated in many manuscripts to the longer (and surely original) reading at Mark 10. 7, as found in D Κ W Θ Byz. The readings in both Gospels are adequately explained by the tendency to harmonize Synoptic parallels, and one need not bring in Gen. 2. 24 at all.Google Scholar

[9] Cf. Elliott, , op. cit, p. 146Google Scholar, and Ross, J. M., ‘The United Bible Societies' Greek New Testament’, J.B.L. 95 (1976), p. 119.Google Scholar

[10] Ross, op. cit., pp. 120–1 discusses Mark 11. 26, as does Elliott, op. cit., p. 146.

[11] See further the remarks of Elliott, op. cit., pp. 145–6.

[12] C D W 565 have καινη¯ς at Matt. 26. 28 but not at Mark 14. 24. It may be noted that the presentation of the evidence at these two places is misleading in the edition itself, since the variation involving τό before τη¯ς is ignored at the first text but included at the second. In fact, the evidence for τό at Matt. 26. 28 is somewhat stronger (only ???³⁷ ℵ B D L Z Θ 33 102^vid 544 omit) than at Mark 14. 24 (here ℵ B C D^b E L V X Θ Ψ 11 22 565 579 892 al pc omit). The UBS text omits τό at both places, but cites the textual evidence only for the second, more strongly supported, omission.

[13] Tischendorf reports that a Syriac codex omits καινή at Luke 22. 20; this could be the result of a leap from one ??? to the next within Syriac.

[14] Metzger also includes Mark 7. 16, where the entire phrase is omitted from the UBS text, based on its absence in ℵ B L δ* 28 al. If the longer text is correct at Matt. 11. 15 et al., there exists a remarkable example of coincidental agreement in omission: the άκούειν has fallen out (with no manuscript support) of Tischendorf's text at Luke 14. 35 (see the correction in Volume III, p. 1270).

[15] Von Soden suggests this origin of the B D Θ reading in his apparatus by prefixing ‘(του???του?)’. (He prints the longer reading.) The reading of A δ al could, moreover, have arisen from the Byz text by this same leap, followed by the ‘correction’ of restoring του¯ δικαίον out of place. The omission of τούτον 1010 1515 is the result of a leap from the final two letters of τικαίον to the same letters at the end of τούτον. The Byz reading here has been cited as having Semitic word order; cf. Kilpatrick, , ‘Some Problems in New Testament Text and Language’, Neotestamentica et Semitica, ed. Ellis, E. Earle and Max, Wilcox (Edinburgh: T. & T. Clark, 1969), p. 198.Google Scholar

[16] See Metzger's judicious observation at Commentary, p. xxiv.

[17] This is clear enough, it seems from the Commentary, p. v. Explicit remarks to the same effect are made by Wikgren in his response to Kilpatrick's ‘A Textus Receptus Redivivus?’ (Berkeley: Center for Hermeneutical Studies [Colloquy 32], 1978), pp. 43–5.

[18] Many reviewers have commented on the editors' preference for ℵ B; cf. Elliott, op. cit., pp. 131–2, and Eldon, Jay Epp, C.B.Q. 37 (1975), p. 135. Sometimes, however, the editors give undue weight to other witnesses: at Rom. 6. 16 (cited by Ross, op. cit., p. 119) omission by ‘unintentional oversight’ in a few witnesses leads to a ‘C’ rating for words (found in ℵ B) which really are certain.Google Scholar

[19] This example of a scribal leap is cited by Metzger, , The Text of the New Testament (second edition; New York and Oxford: Oxford University Press, 1968), p. 189, and referred to in the Commentary, p. 152.Google Scholar

[20] Cf. the related remarks by Fee, Gordon D. concerning the reading by ???⁷⁵ A B at John 5.11: ‘The textual relationship between ???⁷⁵ and B means that theirs is a single witness’ (‘Rigorous or Reasoned Eclecticism - Which?’ Studies in New Testament Language and Text, ed. Elliott, J. K. [Leiden: E. J. Brill, 1976], p. 196).Google Scholar

[21] Westcott, B. F. and Hort, F. J. A., The New Testament in the Original Greek, Volume 2, Introduction [and] Appendix (Cambridge and London: Macmillan, 1881), §54. The point is, of course, recognized in the Commentary at p. xxvi (under ‘C’).Google Scholar

[22] For example, Günther, Zuntz has argued in The Text of the Epistles (London: The British Academy, 1953), p. 15Google Scholar, that a primitive error (caused by a leap) is present at 1 Cor. 6. 5. Ross postulates an early error at Mark 11. 26 (op. cit., pp. 120–1). And Streeter, B. H. notes in The Four Gospels (fifth impression; London: Macmillan, 1936), p. 171: ‘If there are 48 examples in the Gospels of omission through homoioteleuton in ℵ alone, it would be odd if there were none in the first copy of Mark which went to Antioch’ (and thus was the source of the copy used by Matthew).Google Scholar

[23] Cf. Hort, Introduction, §§57–58; I believe that at least some examples of scribal leaps fit Hort's description of errors which are ‘obvious and tempting’ or ‘very plausible and tempting’. Hoskier, H. C., Concerning the Text of the Apocalypse (London: Quaritch, 1929), Volume 1, p. xxxixGoogle Scholar, says that ‘omissions or reduplications owing to homoioteleuta may occur and reoccur without malice prepense’ (that is, as ‘coincidences’). Finally, see Colwell's, Ernest C. ‘Method in Evaluating Scribal Habits’, Studies in Methodology in Textual Criticism of the New Testament (Grand Rapids: Eerdmans, 1969), p. 123, item III.Google Scholar

[24] See, e.g. the comments ad Matt. 13. 35 (διά), ad John 10. 29 (παγπός), and ad John 11. 51. It may be that the editors assume that coincidental agreement plays a role in the significant support given to the omission of Luke 14. 27, where the recurring portion of text (at least in some manuscripts) would have been ????-NATAIEINAIM-???-MAΘHTHC, surely a temptation to any scribe.

[25] See, e.g. the argument ad Luke 23. 17: ‘Although homoeoarcton (ANAI'KHN … ANEKPAΓON) might account for the omission in one family of witnesses, such a theory is unable to explain its widespread omission and its presence at two different places.’ But about half the support for the omission is Alexandrian (???⁷⁵ B L T 892* 1241 cop), and of the rest most (A K п 1079 1546) is part of one textual line, namely family п and the closely related Alexandrinus (cf. Jacob, Geerlings, Family п in Luke [Salt Lake City: University of Utah Press, 1962]; as Geerlings notes, besides K 1079 1546 support among the descendants of п for the omission is found in 265 489* 1219 1313 1346*). The ‘presence at two different places’ may be due to a leap followed by a ‘correction’ out of order. This sort of scribal act in an ancestor of the Western text would account for the transposition in D it^d syr^cs, at least; the Ethiopic support is in fact limited to eth^ro (ethPP has vs. 17 in the usual order), and may be coincidental. By the way, the homoeoarcton here is slightly more tempting than Metzger says, since the similar texts are ANAΓKHNδE and ANEKPAΓONδE.Google Scholar

[26] Scrivener, Frederick H., A Full Collation of the Codex Sinaiticus with the Received Text of the New Testament (Cambridge: Deighton, Bell, and Co., 1864), p. xv.Google Scholar

[27] Hermann, Freiherr von Soden, Die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt hergestellt auf Grund ihrer Textgeschichte, I. Teil, Untersuchungen, II. Abteilung, Die Textformen, A. Die Evangelien (Berlin: Arthur Glaue, 1907), pp. 917–18: ‘Um so häufiger sind “geniale” Leichtfertigkeiten. Manchmal schreibt er völlig gedankenlos Unsinn nieder. Sprünge bei homoioteleuta oft über mehrere Zeilen weg passieren ihm ebenso leicht, wie sonstige Auslassungen von Worten.’ See also pp. 920–1 on ℵ's ‘Omissionen’ in the Gospels. Hort, Introduction, §326, understates the situation.Google Scholar

[28] Ibid., §312. Cf. also von Soden, op. cit., p. 908: ‘Durchaus in dieser Linie der versehentlichen Auslassungen von Buchstaben und Silben liegen auch eine ganze Anzahl von Wortauslassungen, und zwar nicht nur solche, die nur eine Silbe oder gar nur einen Buchstaben umfassen, sondern selbst ganze Wortkomplexe oder Satzteile. Als versehentliche Auslassungen dokumentieren sie sich dabei teils durch ihre Unentbehrlichkeit in der Gesamtstruktur des Satzes, teils durch das ein Abirren des Auges veranlassende Homoioteleuton.’

[29] See the figures (which need some revision) in Colwell, op. cit., p. 112.

[30] There are, in fact, 167 singular omissions in ???⁴⁶; in a few of the 50 cases mentioned there might be some doubt as to whether a leap was involved. Such figures give, of course, a conservative count of the number of leaps a scribe made, since coincidental agreements are ex hypothesi excluded. Von Soden evidently used a similar criterion in isolating the peculiarities of ℵ, when he says (op. cit., p. 918): ‘In den folgenden Listen sind alle in δ2* erscheinenden Lesarten, die sicher nicht H zuzuschreiben sind, verzeichnet.’

[31] Wilhelm, Streitberg, Die gotische Bibel (sixth edition; Heidelberg: Carl Winter Universitäts-verlag, 1971), pp. 198–9, gives the Greek equivalent of the Gothic as simply μητέρα. However, his Gothic text has aipein seinai, and thus (as Tischendorf notes) agrees with ℵ al.Google Scholar

[32] This is pointed out by Kilpatrick, ‘A Textus Receptus Redivivus?’, p. 14. Wikgren's response to Kilpatrick on this issue is that ‘in Mk 10:7 the alternative to the long addition for the Committee was μητέρα and not μητέρα αύτου¯’ (Ibid., pp. 45–6). But this misses the point. The (prior) question is not which variant is authentic, but what transcriptional errors are possibilities to be considered.

[33] Of course, the omission in ℵ B Ψ 892* may again involve an earlier omission in a common Alexandrian ancestor, but in this case too it is at least possible that we should postulate μητέρα αύτου¯ in that ancestor, since ℵ 579 1241 cop among the Alexandrians have that reading.

[34] Comments along these same lines may be found in Kilpatrick's ‘The Greek New Testament Text’, pp. 195–6.

[35] Commentary, p. xvii.

[36] Cf., e.g. the remarks of Kilpatrick, ‘A Textus Receptus Redivivus?’, p. 7.

[37] Compare, for example, the canon of Griesbach cited by Metzger, Text, p. 120.

[38] The fact of the scribe's losing his place led Colwell to dub such a leap ‘the case of the misplaced scribe’ (op. cit., p. 112).

[39] At any rate, this seems to be the simplest way of looking at the leap. It is however, possible that the scribe of ℵ wrote αντι, let us say, of the first occurrence of άντιπαρη¯δθεν, then lost his place by returning to the second occurrence of άντιπαρη¯δθεν, and thus continued by writing παρηδθεν of this second occurrence, followed by vs. 33. Note that the scribe would write exactly the same letters if he made this mistake, or indeed if he made the leap from any sequence of letters within the first occurrence of άντιπαρη¯δθεν to the same sequence within the second occurrence. Especially within scriptio continua there is really no reason why such a leap has to be confined to the beginning or end of a word. (Of course, leaps omitting entire words probably have a better chance of going unnoticed by the scribe himself or by later correctors or copyists.)

[40] See note 3 above and the Commentary's references to ‘scribal inadvertence’ (ad Matt. 12. 31), ‘scribal oversight’ (ad Matt. 18. 21), and ‘scribal blunders’ (ad Matt. 20. 17).

[41] Text, p. 189. Metzger is not alone in his misstatement of the result of a leap. C. R. Gregory, for instance, reports Tischendorf as saying that an omission by homoeoteleuton occurs ‘scriba a syllaba vel verbo ad similem syllabam vel simile verbum transiliente, quae intersunt omittente’ (my emphasis; cited in Tischendorf, Volume III, p. 58). Similarly, Eberhard, Nestle, Introduction to the Textual Criticism of the Greek New Testament, tr. William, Edie, ed. Allan, Menzies (London: Williams and Norgate, 1901), p. 235Google Scholar, comments: ‘the eye of the scribe jumps from one word or group of letters to another the same or similar to it, either before or after it. In the former case the intervening words will be repeated, in the latter they will be omitted’ (my emphasis). In fact, just as the intervening words and the second occurrence of the group of letters will be omitted in a leap forward, so the intervening words and the second occurrence will be repeated in a leap backward.

[42] The majority of witnesses omit και έν ται¯ς χερσιν, but none omits only και έν ται¯ς. All the manuscripts which omit καιναι¯ς have και έν ται¯ς χερσιν following.

[43] Metzger occasionally appeals to the corrections within a manuscript as evidence for the direction of scribal activity: cf. his remarks ad Mark 9. 24 (A C), Luke 12. 31 (D), John 4. 25 (???⁶⁶ ℵ), and Heb. 8. 8 (ℵ D). On the other hand, correction in the manuscript itself may be an indication that the original reading is a mere ‘scribal oversight’, as we find in the note ad John 7. 37 (???⁶⁶* ℵ*), and as is presumably the case with ℵ*'s omission of Luke 10. 32, corrected by ℵ^c.

[44] A few other comments may be found in Kilpatrick's ‘A Textus Receptus Redivivus?’, p. 12, and in my response to Kilpatrick's paper, p. 29.