Skip to main content

P45 and the Problem of the ‘Seventy(-two)’: A Case for the Longer Reading in Luke 10.1 and 17

  • Zachary J. Cole (a1)

At Luke 10.17, most modern critical editions incorrectly cite the wording of P45 as ἑβδομήκοντα δύο (72) instead of ἑβδομήκοντα (70). As this is one of the two oldest witnesses to the verse, this revision of external evidence calls for a fresh examination of the textual problem as a whole. Previous discussions have focused almost exclusively on the perceived symbolic values of ἑβδομήκοντα (+ δύο) to identify the ‘more Lukan’ wording, but this essay argues on the basis of new transcriptional evidence that the earlier reading is more likely ἑβδομήκοντα δύο.

Hide All

1 Thanks to Ms Fionnuala Croke, Director of the Chester Beatty Library, and her curatorial staff for granting direct access to P45, and to Professor Larry Hurtado for organising the visit. I would also like to thank Professor Paul Foster, Elijah Hixson and Jesse Grenz for their helpful feedback on earlier drafts of this paper.

2 Throughout this paper, I refer to ‘scribes’ and ‘copyists’ as the primary agents who transmit and alter texts, but I am aware that readers, users and owners of texts were just as likely to introduce changes.

3 Metzger B. M. and Ehrman B. D., The Text of the New Testament: Its Transmission, Corruption and Restoration (New York: Oxford University Press, 2005 4) 341. Earlier editions contain the same wording (e.g. the first edition (19641) 244).

4 For more on Greek numerical shorthand, see Turner E. G., Greek Manuscripts of the Ancient World (London: Institute of Classical Studies, rev. edn 1987) 1516 .

5 Kenyon F. G., ed., The Chester Beatty Biblical Papyri, vol. ii: The Gospels and Acts: Text (London: Emery Walker, 1933); and for plates, Kenyon F. G., ed., The Chester Beatty Biblical Papyri, vol. ii: The Gospels and Acts: Plates (London: Emery Walker, 1934).

6 For example, the apparatuses of Nestle22 (1956) and Souter2 (1947).

7 Roberts C. H., ‘An Early Papyrus of the First Gospel’, HTR 46 (1953) 233–7, at 236 n. 14. It is worth noting that, in antiquity, this character was referred to as a γαβέξ/γαμέχ.

8 Metzger B. M., ‘Seventy or Seventy-two?’, NTS 5 (1959) 299306 , at 299, reprinted in idem, Historical and Literary Studies: Pagan, Jewish, and Christian (NTTS 8; Leiden: Brill, 1968) 6776 , at 67–8.

9 The designations of recto and verso follow those given in Kenyon's publications.

10 One question that remains is why the line-filler has a supralinear stroke above, the signature indicator of an abbreviation. However, it is not uncommon for the scribe to extend the supralinear stroke far beyond nomina sacra contractions and over the following letters. See e.g. θ̅υι (Mark 7.9, fol. 5v), θ̅υ̅ ϊ̄ (John 11.4, fol. 16v), π̅νι ε̅ (Acts 16.18, fol. 29v), etc.

11 The stigma/digamma was not differentiated from the lunate sigma (ϲ) in manuscripts such as P47 and P115.

12 See Verheyden J., ‘How Many Were Sent according to Lk 10,1?’, Luke and his Readers: Festschrift A. Denaux (ed. Bieringer R. et al. ; BETL 182; Leuven: Leuven University, 2005) 193238 , at 201–2. Also, this is upheld in the most recent edition of Metzger and Ehrman, Text of the New Testament, 340–1 n. 65.

13 See the Center for the Study of New Testament Manuscripts website ( and that of the Institut für neutestamentliche Textforschung (

14 Metzger and Ehrman, Text of the New Testament, 340–1 n. 65. No changes were made in NA25 (1963), UBS1 (1966) or UBS2 (1968).

15 Intervening editions show no difference here: NA27 (1993), UBS3 Corr (1983) and UBS4 (1993). However, this change was not made in the IGNTP volume of Luke, published a few years later; American and British Committees of the International Greek New Testament Project, eds., The New Testament in Greek: The Gospel according to St. Luke (2 parts; Oxford: Clarendon, 1984–7) Part i, 227.

16 Metzger B. M., A Textual Commentary on the Greek New Testament (Stuttgart: Deutsche Bibelgesellschaft/United Bible Societies, 1994 2) 126.

17 For more discussion on patristic sources, see Metzger, ‘Seventy or Seventy-two?’, 300–2.

18 Marc. 4.24.1. Evans E., ed., Tertullian: Adversus Marcionem (2 vols.; Oxford: Oxford University Press, 1972) ii.390: ‘He chooses other seventy apostles also, over above the twelve: for to what purpose twelve, after that number of wells in Elim, without adding seventy, after that number of palm-trees?’

19 For the text of Marcion, see Roth D. T., The Text of Marcion's Gospel (NTTSD 49; Leiden: Brill, 2015) 215–17, who gives the following reconstruction: ἀνέδειξεν … ἑτέρους ἑβδομήκοντα … ἀπέστειλεν … εἰς … πόλιν.

20 Haer. 2.21.1; 3.13.2.

21 Comm. 8.1a. For the text, see L. Leloir, ed., Saint Éphrem: commentaire de l’évangile concordant texte syriaque (manuscrit Chester Beatty 709) (Chester Beatty Monographs 8; Leuven: Peeters, 1963; repr. 1990). Additional confirmation is found in Codex Fuldensis (sixth century), the earliest Latin witness to a text thought to be related to the Diatessaron, which reads LXX duos (Luke 10.1) and septuaginta duo (10.17). For the text, see E. Ranke, ed., Codex Fuldensis (Marburg etc.: Elwert, 1868) 66, and for images, see (at fol. 64r).

22 Strom. 2.20.116; GCS 15.176.

23 Hom. Exod. 7.3; Hom. Num. 27.11.1. Actually, Rufinus’ Latin translation of Hom. Num. 27.11.1 is inconsistent on this point: ubi sunt duodecim fontes aquarum et septuaginta duae arbores palmarum … sed et alios septuaginta … sed et alio septuaginta . Due to (a) the fact that the LXX text of Numbers here is consistent with the value ‘seventy’ (as far as we can ascertain), and (b) the near-verbatim similarity to the text in Hom. Exod. 7.3, it seems most likely that the first value is corrupt, and, here at least, Origen originally had septuaginta. For the text and further discussion, see SC 461.316 n. 2.

24 Comm. Rom. 10.21; PG 14.1280 (on Rom 16.7).

25 GCS 1/2.104–5.

26 Adam. Dial. 1.5; 2.12. For the text of books 1–2, see K. Tsutsui, ed., Die Auseinandersetzung mit den Markioniten im Adamantios-Dialog: Ein Kommentar zu den Büchern iii (Berlin: de Gruyter, 2004).

27 Const. ap. 2.55.2.

28 Comm. Rom. 8.30; 9.13; Comm. 1 Cor. 12.31.

29 Cons. 2.23.54.

30 Hist. eccl. 1.10.5, 12.1, 13.4; 2.1.1; 3.24.5; Dem. ev. 3.2.24; 3.4.37.

31 Mor. 70.1.

32 Hom. Luc. on Luke 10.1–17 (TU 34/1.99–104).

33 Quaest. in oct. 49.110; Quaest. in Psal. 2.8.

34 Exp. Luc. 7.44 (at Luke 10.3).

35 Epist. 78.6.

36 An example of ‘seventy’: ἦσαν γὰρ καὶ ἄλλοι ἀπόστολοι, ὡς οἱ ἑβδομήκοντα (Hom. 1 Cor. 38.4; PG 61.326 (on 1 Cor 15.7); see also Hom. Matt. 38.1; Hom. Jo. 85.3; Hom. Act. 3.2). And of ‘seventy-two’: τί γὰρ ὄφελος ἐκ τοῦ μαθεῖν ἐκείνου τὴν προσηγορίαν, ἐπεὶ οὐδὲ τῶν ἑβδομήκοντα δύο λέγει τὰ ὀνόματα; (Hom. Jo. 18.3; PG 59.117 (on John 1.40)).

37 Metzger, ‘Seventy or Seventy-two?’, 302.

38 Some doubt that Luke intended any symbolism in the numeral, but they nonetheless rely on considerations of symbolism to explain the rise of the variation; e.g. Prieur A., Die Verkündigung der Gottesherrschaft: Exegetische Studien zum lukanischen Verständnis von βασιλεία τοῦ θεοῦ (WUNT ii/89; Tübingen: Mohr Siebeck, 1996) 212–20.

39 For many of the following examples (and more), see Metzger, ‘Seventy or Seventy-two?’, 302–4.

40 Jellicoe S., ‘St. Luke and the “Seventy(-Two)”’, NTS 6 (1960) 319–21, and idem, St. Luke and the Letter of Aristeas’, JBL 80 (1961) 149–55. Similar is Bovon F., Luke (3 vols.; Minneapolis: Fortress, 2013) ii.26.

41 See the editorial note by Aland in Metzger, A Textual Commentary, 127. Similar is Elliott J. K. and Moir I., Manuscripts and the Text of the New Testament: An Introduction for English Readers (Edinburgh: T. & T. Clark, 1995) 52.

42 Cf. Wilson S. G., The Gentile and the Gentile Mission in Luke-Acts (SNTSMS 23; Cambridge: Cambridge University Press, 1973) 45–7. For more on Moses in Luke-Acts, see Mánek J., ‘The New Exodus in the Books of Luke’, NovT 2 (1957) 823 .

43 Cf. Leaney A. R. C., The Gospel according to Luke (London: A & C Black, 1966 2) 176.

44 Cf. Nolland J., Luke 9:21–18:34 (WBC 35B; Dallas: Word, 1993) 549.

45 Would, therefore, seventy-two – which is divisible by twelve – call to mind the twelve tribes of Israel and in effect suggest an Israel-oriented Gospel mission? Cf. Wolter M., Das Lukasevangelium (HNT 5; Tübingen: Mohr Siebeck, 2008) 377.

46 Verheyden, ‘How Many Were Sent?’, 233.

47 NA28, 54*.

48 I should credit Timothy J. Cole for pointing out a possible echo of this passage in Acts 23.23: while Jesus sends out seventy/seventy-two disciples ‘two by two’ (ἀνὰ δύο δύο), Paul is escorted to Caesarea by 70 horsemen, 200 soldiers and 200 bowmen.

49 For example, in criticism of Verheyden's suggestion above, Wolter remarks: ‘Ob dieser Sprachgebrauch jedoch geeignet ist, das textkritische Problem zu entscheiden und den Ausschlag zugunsten der Ursprünglichkeit der Zahl 70 zu geben (so Verheyden 234ff), ist alles andere als sicher, denn man kann mit seiner Hilfe auch begründen, warum 72 in 70 geändert wurde’ (Wolter, Das Lukasevangelium, 377).

50 Metzger, ‘Seventy or Seventy-two?’, 305.

51 Not in view are orthographical variation, bare omission of numerals or the substitution of values.

52 It would seem reasonable to assume that numerical shorthand was more easily corrupted than full number-words by copyists, but this is not apparently the case. See Develin R., ‘Numeral Corruption in Greek Historical Texts’, Phoenix 44 (1990) 3145 . In addition, my own investigation of singular readings involving numerals in the papyri revealed no significant results that would suggest this.

53 In addition to that of the NA28, the following apparatuses were used (where relevant): Tischendorf C., ed., Novum Testamentum Graece (2 vols.; Leipzig: Giesecke & Devrient, 1869–728); von Soden H., ed., Die Schriften des neuen Testaments, Part ii: Text mit Apparat (Göttingen: Vandenhoeck & Ruprecht, 1913); Hoskier H., ed., Concerning the Text of the Apocalypse (2 vols.; London: Bernard Quaritch, 1929); Legg S. C. E., ed., Novum Testamentum Graece secundum Textum Westcotto-Hortianum: Evangelium secundum Marcum (Oxford: Clarendon, 1935); Legg S. C. E., ed., Novum Testamentum Graece secundum Textum Westcotto-Hortianum: Evangelium secundum Mattaeum (Oxford: Clarendon, 1940); American and British Committees of the International Greek New Testament Project, eds., The Gospel according to St. Luke; Elliott W. J. and Parker D. C., eds., The New Testament in Greek iv: The Gospel according to St. John, vol. i: The Papyri (Leiden/New York: Brill, 1995); and Schmid U. B. with Elliott W. J. and Parker D. C., eds., The New Testament in Greek iv: The Gospel according to St. John, vol. ii: The Majuscules (Leiden/Boston: Brill, 2007).

54 Witnesses are given in Gregory–Aland identifications; conversions for Tischendorf and von Soden can be found in Aland K., ed., Kurzgefasste Liste der griechischen Handschriften des neuen Testaments (ANTF 1; Berlin/New York: de Gruyter, 1994 2) 377427 , and for Hoskier, see Elliott J. K., ‘Manuscripts of the Book of Revelation collated by H. C. Hoskier’, JTS n. s. 40 (1989) 100–11 (reprinted in idem, New Testament Textual Criticism: The Application of Thoroughgoing Principles. Essays on Manuscripts and Textual Variation (NovTSup 137; Leiden/Boston: Brill, 2010) 133–44).

55 The IGNTP Luke volume also lists minuscule 349 here, but images of the manuscript indicate that this must be an error.

56 Minuscules 664 and 1094 part of the same family; see Hoskier, Text of the Apocalypse, I.330–7.

57 This form does in fact occur in many manuscripts; for examples, see BDF §63(8).2.

58 A. DeGuglielmo, ‘Emmaus’, CBQ 3 (1941) 293–301; see also Metzger, Textual Commentary, 158; R. Reisner, ‘Wo lag das neutestamentliche Emmaus (Lukas 24, 13)?’, ZAC 11 (2007) 201–20; S. Reece, ‘Seven Stades to Emmaus’, NTS 47 (2001) 262–6; and especially the discussion of Lagrange M.-J., Évangile selon Saint Luc (Paris: Lecoffre, 1927 3) 602–3, 617–22, which, though dated, is still worthwhile.

59 See also Develin, ‘Numeral Corruption’. In a similar study of classical manuscripts, he lists δέκα for ἕνδεκα (Herodotus 6.14.2), δέκα for δώδεκα (Xenophon, Hell. 5.1.6 and again at 7.5.10), μύρια for τρισμύρια (Xenophon, Hell. 7.8.26), δέκα for πεντεκαίδεκα ([Aristotle], Ath. Pol. 36.1); Develin cites no examples of a second digit being erroneously added to a single-digit round number.

60 Wevers J. W., ed., Septuaginta: Vetus Testamentum Graecum auctoritate Academiae scientiarum Gottingensis editum, vol. i: Genesis (Göttingen: Vandenhoeck & Ruprecht, 1974). For omissions, see Gen 5.9, 12, 16; 9.29; 11.14, 18, 24, 32; 17.24, 25; 18.28; 50.22, 26. And for additions, see Gen 21.5; 24.10; 25.20; 31.7.

61 Colwell E. C., ‘Method in Evaluating Scribal Habits: A Study of P45, P66, P75 ’, Studies in Methodology in Textual Criticism of the New Testament (NTTS 9; Leiden: Brill; Grand Rapids: Eerdmans, 1969) 106–24; Head P. M., ‘Observations on Early Papyri of the Synoptic Gospels, especially on the “Scribal Habits”’, Bib 71 (1990) 240–7; Hernández J. Jr, Scribal Habits and Theological Influences in the Apocalypse: The Singular Readings of Sinaiticus, Alexandrinus, and Ephraemi (WUNT ii/218; Tübingen: Mohr Siebeck, 2006); Jongkind D., Scribal Habits of Codex Sinaiticus (TS 3/5; Piscataway, NJ: Gorgias, 2007); Royse J. R., Scribal Habits in Early Greek New Testament Papyri (NTTSD 36; Leiden: Brill, 2008); and for a summary of this development, see Royse J. R., ‘Scribal Tendencies in the Text of the New Testament’, The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis (ed. Ehrman B. D. and Holmes M. W.; NTTSD 42; Leiden: Brill, 2013 2) 461–78.

62 Royse, Scribal Habits, 197 (also 134, 377, 368–9).

63 Royse, Scribal Habits, 134. Hoskier called these omissions ‘“the shorter text” with a vengeance’; Hoskier H., ‘Some Study of P45 with Special Reference to the Bezan Text’, Bulletin of the Bezan Club 12 (1937) 51–7, at 53.

64 Colwell seemed to view such omissions as intentional (Colwell, ‘Scribal Habits’, 118–19), an opinion echoed by Royse, Scribal Habits, 141, 197. See also Tarelli C. C., ‘Omissions, Additions, and Conflations in the Chester Beatty Papyrus’, JTS 40 (1939) 382–7, at 384; Lagrange M.-J., ‘Les papyrus Chester Beatty pour les Évangiles’, RB 43 (1934) 541 , at 12. Others prefer to see them as accidental: e.g. Couchoud P.-L., ‘Notes sur le texte de St Marc dans le Codex Chester Beatty’, JTS 35 (1934) 322 , at 8; Aland B., ‘The Significance of the Chester Beatty Papyri’, The Earliest Gospels: The Origins and Transmission of the Earliest Christian Gospels – The Contribution of the Chester Beatty Gospel Codex P45 (ed. Horton Charles; JSNTSup 258; London/New York: T&T Clark, 2004) 108–21, at 113; Min K. S., Die früheste Überlieferung des Matthäusevangeliums (bis zum 3./4. Jh.): Edition und Untersuchung (ANTF 34; Berlin/New York: de Gruyter, 2005) 130–5.

65 Verheyden, ‘How Many Were Sent?’, 211.

66 The only alphabetic numerals present in P45, aside from that in Luke 10.17, are ι̅β̅ (Mark 8.19), ι̅η̅ (Luke 13.11, 16) and μ̅ (Acts 7.36).

67 I should credit the anonymous reviewer for this helpful suggestion.

68 Metzger and Ehrman, Text of the New Testament, 341.

69 This tendency towards omission might shed light on the similar textual problem that follows almost immediately in Luke 10.1: ἀπέστειλεν αὐτοὺς ἀνὰ δύο °[δύο] (‘he sent them two by two’). The evidence here is similar: δύο δύο is read by B, K, Θ, f 13, 565, l2211 etc., while δύο is read by א, A, C, D, L, W, Ξ, Ψ, 0181, f 1, 33, 𝔐 (P45 and P75 are lacunose). Though not properly a two-digit numeral, the effect would have been similar with the repetition of δύο, inviting the possibility that the tendency for scribes to omit from two-digit numbers also contributed to the initial loss of the second (or first?) δύο.

Recommend this journal

Email your librarian or administrator to recommend adding this journal to your organisation's collection.

New Testament Studies
  • ISSN: 0028-6885
  • EISSN: 1469-8145
  • URL: /core/journals/new-testament-studies
Please enter your name
Please enter a valid email address
Who would you like to send this to? *



Full text views

Total number of HTML views: 7
Total number of PDF views: 102 *
Loading metrics...

Abstract views

Total abstract views: 373 *
Loading metrics...

* Views captured on Cambridge Core between 6th March 2017 - 14th December 2017. This data will be updated every 24 hours.