1 Colwell, E. C., ‘Method in Establishing the Nature of Text-Types in New Testament Manuscripts’, in Studies in Methodology (NTTS; Leiden: Brill, 1969) 45–55Google Scholar; cf. Bousset, W., ‘Zur Textkritik der Apokalypse’, in Textkritische Studien zum Neuen Testament (TU 2/4; Leipzig: Hinrichs, 1894) 42–4Google Scholar; Schmid, J., Studien zur Geschichte des griechischen Apokalypse-Textes, vol. 2: Die Alten Stämme (Münchener theologische Studien 1, Historische Abteilung Ergänzungsband 1; Munich: K. Zink, 1956) 127–9.Google Scholar

2 The discussion about text types has advanced considerably since the publication of Colwell's study on this topic. The most up-to-date treatment of the subject can now be found in Epp's, Eldon Jay essay, ‘Textual Clusters: Their Past and Future in New Testament Textual Criticism’, in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis (2nd edn; ed. Ehrman, B. D. and Holmes, M. W.; NTTSD 42; Leiden and Boston: Brill, 2013) 519–77Google Scholar. Colwell's discussion, however, is far from passé. His assessment of Schmid's study and, in particular, his discussion of the Andreas text type represent the last serious engagement of the issue. In fact, Colwell raised a number of concerns that are still to be addressed. The central concern of this essay, however, is not the debate over text types but Schmid's claim about the existence of a fourth-century witness to the Andreas text type. By clarifying this matter, it is hoped that the way will be cleared for further study in this area.

3 Aland, K. and Aland, B., The Text of the New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism (2nd edn; trans. Rhodes, E. F.; Grand Rapids, Mich.: Eerdmans, 1986)Google Scholar 107.

4 Colwell, ‘Method’, 53 n. 1.

5 Colwell also called for further clarification regarding Schmid's claim that the later text types (i.e. Byzantine and Andreas) should not be understood as ‘later, revised forms of their elders’, AC and P⁴⁷א. The only thing that Schmid had proven was that the later text types are not entirely derived from the older ones. See Colwell, ‘Method’, 52; cf. Schmid, Studien, 2.147–8.

6 Uncovering select readings from the late Byzantine tradition in an earlier manuscript, for example, does not move the Byzantine tradition ‘as an entity’ back to the date of the earlier manuscript according to Colwell. See Colwell, ‘Method’, 52, 55.

7 Schmid himself never actually used א^a but, as will become clear, his S^a is the equivalent.

8 Tischendorf lists these as א^a (contemporary or nearly contemporary with the fourth-century scribe); א^b (probably fifth/sixth century); א^c, א^ca, א^cb, א^cc and א^cc* (seventh century); and א^e (twelfth century). See Tischendorf, Novum Testamentum Graece (3 vols.; 8th edn;Leipzig, 1869–4) 3.346Google Scholar; and Scrivener, F. H. A., A Full Collation of the Codex Sinaiticus with the Received Text of the New Testament (Cambridge: Cambridge University Press, 1864)Google Scholar xxiii; cf. Jongkind, D., Scribal Habits of Codex Sinaiticus (TS 3.5; Piscataway, NJ: Georgias Press, 2007) 10–11Google Scholar.

9 Tischendorf, Novum Testamentum Graece, 3.346.

10 The corrections made in the scriptorium are more clearly (and fully) laid out by Tischendorf in his Novum Testamentum Sinaiticum.

11 The use of S^a (א^a) for the Apocalypse's fourth-century corrections is unattested prior to Schmid. Hoskier actually deploys א^a before Schmid but not for the fourth-century corrections. Hoskier's use of the siglum accords with Tischendorf's seventh-century corrector C^a (in his editio princeps of Codex Sinaiticus) or א^c (in the apparatus of his Editio octava critica maior). See Hoskier, H. C., Concerning the Text of the Apocalypse: Collations of All Existing Available Greek Documents with the Standard Text of Stephen's Third Edition, together with the Testimony of the Versions, Commentaries and Fathers. A Complete Conspectus of All Authorities (2 vols.; London: Bernard Quaritch, 1929)Google Scholar; cf. Tischendorf, A. F. C., Novum Testamentum Sinaiticum cum Epistula Barnabae et Fragmentis Pastoris (2 vols.; Leipzig: F. A. Brockhaus, 1863)Google Scholar; idem, Novum Testamentum Graece, 3.346.

12 Schmid, Studien, 2.127–9.

13 Bousset, ‘Zur Textkritik der Apokalypse’, 42–4.

14 So Schmid, Studien, 2.127–8 n. 6.

15 Milne, H. J. M. and Skeat, T. C., Scribes and Correctors of the Codex Sinaiticus (Oxford: Oxford University Press, 1938)Google Scholar.

16 In particular, Milne and Skeat took note of corrections Tischendorf had missed and examined obscurities under ultraviolet light. For particular improvements, see Milne and Skeat, Scribes and Correctors, i–xii, 13, 18, 20, 28, 40 n. 2, 41, 47, 50, 66, 70, 89.

17 Schmid, Studien, 2.127.

18 So Colwell, ‘Method’, 52.

19 Bousset had argued that the large number of א^c corrections that accord with K (which is where he places Andreas) demonstrates that the codex was corrected against a manuscript belonging to the Andreas tradition. Corrections that deviate from K do not alter these results; they merely show that one or more additional manuscripts were also used for the corrections. Bousset, ‘Zur Textkritik der Apokalypse’, 44.

20 The claim is so extraordinary that it is clear that most of Schmid's readers have been either unaware of it or unaware of its implications. With the well-known dearth of early manuscript attestation for the Apocalypse (from the fourth century and earlier), Schmid's claim should have been a ‘bombshell moment’ for the study of the book's textual history, garnering widespread attention and generating vigorous debate over its implications.

21 א^c is labelled C^a in Milne and Skeat, conforming to Tischendorf's use of the siglum in his editio princeps. This is clear in their chapter ‘The Correctors’, which – ironically – Schmid cites in support of his reconstruction. The chapter is divided into two sections: ‘The A and B Correctors’ and ‘The C Correctors’. The use of A and B in the heading is derived from Tischendorf's classification of correctors, which he dates to the scriptorium. Milne and Skeat accept Tischendorf's judgement on the date but simplify considerably his system (a system which was elaborated by Lake to include A¹, A², A³, A⁴, A⁵, A^obliqu, A^Herm, B, and B^a!); Milne and Skeat assign the entire batch of A and B corrections to the fourth-century scribes: A and D. Tischendorf's classification was considered too complex, ‘resembling a committee of revisers’ rather than a ‘business establishment’. Milne and Skeat believed that the A and D scribes could write with ‘several different hands’. As for the C corrections, again, Milne and Skeat follow Tischendorf's classification with some refinements. Nowhere, however, are the C corrections assigned a date at variance with Tischendorf (with the exception of C^c*, which they consider possibly eighth-century), nor are any of the C correctors identified with the A and D scribes. See Milne and Skeat, Scribes and Correctors, 40–50; cf. Schmid, Studien, 127 n. 2.

22 This appears to be the most plausible explanation. While we can only speculate about the precise origin of Schmid's misunderstanding, it is quite possible that Schmid misread a particular English construction in their discussion. For example, the phrase ‘by the original scribe or C^a – read in isolation – can be understood in two ways: as referring either to two mutually exclusive entities (i.e. ‘the original scribe’ + C^a) or to a single entity (i.e. ‘the original scribe’ = C^a). This is precisely the kind of construction one encounters in Milne and Skeat's treatment of the correctors (Scribes and Correctors, 49). It is therefore possible that Schmid understood a statement of mutual exclusivity as one of inclusivity. The broader context, however, makes it clear that the two are to be understood as distinct entities from different periods. If any doubt remains about C^a's seventh-century provenance, one need only turn to Milne and Skeat's opening discussion of the manuscript's colophons, where C^a is referred to as the ‘near contemporary’ of C^b1 – an undoubtedly seventh-century corrector (see Milne and Skeat, Scribes and Correctors, 2). Unless this is explained as a translation error of some sort, it is difficult to see how the seventh-century correctors could have been equated with the fourth-century scribes: the distinction between the two groups is fairly clear.

23 Most textual critics have relayed Schmid's classification of the Apocalypse's four text types or textual streams/forms with the qualifying phrase ‘according to Schmid’ (or its equivalent), creating some distance between themselves and Schmid. Challenges to his use of א^a as the basis for a fourth-century Andreas text type, however, are nowhere to be found in the literature: Aland and Aland, Text of the New Testament, 247; Birdsall, J. N., ‘The Text of the Revelation of Saint John: A Review of its Materials and Problems with Especial Reference to the Work of Joseph Schmid’, EvQ 33 (1961) 228–37Google Scholar; Elliott, J. K., ‘The Distinctiveness of the Greek Manuscripts of the Book of Revelation’, in New Testament Textual Criticism: The Application of Thoroughgoing Principles, Essays on Textual Variation (NovT Suppl. 137; Leiden and Boston: Brill, 2010), 149–50Google Scholar; Hernández, J. Jr, Scribal Habits and Theological Influences in the Apocalypse: The Singular Readings of Sinaiticus, Alexandrinus, and Ephraemi (WUNT 2.218; Tübingen: Mohr Siebeck, 2006)Google Scholar 4, 26; Karrer, M., ‘Der Text der Johannesapokalypse’, in Die Johannesapokalypse: Kontexte-Konzepte-Rezeptio /The Revelation of John: Contexts – Concepts – Reception (WUNT 287; Tübingen: Mohr Siebeck, 2012) 43–78Google Scholar; Kilpatrick, G. D., ‘Professor J. Schmid on the Greek Text of the Apocalypse’, VC 13 (1959) 1–3Google Scholar; Nicklas, T., ‘The Early Text of Revelation’, in The Early Text of the New Testament (ed. Hill, C. H. and Kruger, M. J.; Oxford: Oxford University Press, 2012) 228–9Google Scholar; Metzger, B. M. and Ehrman, B. D., The Text of the New Testament: Its Transmision, Corruption, and Restoration (4th edn; New York and Oxford: Oxford University Press, 2005)Google Scholar 79; Parker, D. C., An Introduction to the New Testament Manuscripts and their Texts (Cambridge, 2008) 240–1CrossRefGoogle Scholar; Royse, J. R., Scribal Habits in Early Greek New Testament Papyri (NTTSD 36; Leiden and Boston: Brill, 2008) 359–60Google Scholar. The exception appears to be Colwell, but – as already noted – Colwell's reservations stem from his understanding of text types, rather than from questions over Schmid's redating and relabelling of the seventh-century א^c to the fourth-century א^a. See Colwell, ‘Method’, 53 n. 1; cf. Hernández, J. Jr, ‘The Relevance of Andrew of Caesarea for New Testament Textual Criticism’, JBL 130.1 (2011)Google Scholar 196 n. 66.

24 Actually, this is not unexpected. Milne and Skeat denied that their work was anything close to a ‘collation’. Their task consisted simply of comparing the manuscript against the critical notes of Tischendorf's 1862 editio princeps. See Milne and Skeat, Scribes and Correctors, xi.

25 See Milne and Skeat, Scribes and Correctors, Figure 15.

26 Note ‘Ich schreibe dafür S^a (und S^c für S^cc).’ See Schmid, Studien, 127 n. 3.

27 Schmid, Studien, 2.127.

28 Schmid, Studien, 2.129.

29 As noted above, Schmid himself never actually used א^a, but his S^a is the clear equivalent. The conversion of S^a to א^a is to be found in Birdsall, ‘The Text of the Revelation’, 235; Colwell, ‘Method’, 53 n. 1; Kilpatrick, ‘Professor J. Schmid’, 3; and Metzger, B. M., A Textual Commentary on the Greek New Testament (2nd edn; Stuttgart: German Bible Society, 1994)Google Scholar 681, 685. It is worth noting that Schmid – and Hoskier for that matter – might have been more judicious in their selection of sigla had their studies of the codex's corrections included more than just the text of the Apocalypse. They might then have anticipated that the use of א^a would conflict with the use of the very same siglum by Tischendorf for other biblical books in the codex. Conversely, if textual critics with a general knowledge of Codex Sinaiticus's correctors had familiarised themselves with the particularities of these in the Apocalypse, they might have spotted Schmid's idiosyncratic (and ultimately discrepant) usage and avoided the perpetuation of an error.

30 See Appendix i for sigla comparisons across critical editions, Appendix ii for the scribes and correctors of the Apocalypse in Codex Sinaiticus, and Appendix iii for a schema outlining Schmid's misidentification and dating error.

31 Colwell, ‘Method’, 52.

32 See Kilpatrick, ‘Professor J. Schmid’, 3.

33 Schmid, Studien, 2.127–9.

34 Schmid, Studien, 2.45–9, 51, 57, 60, 72–3, 75, 81, 83, 91, 95, 98–9, 102, 104–5, 114 n. 4, 115–16, 125, 134, 142–3, 154–6, 157, 161, 165, 172.

35 Schmid, Studien, 2.180, 183.

36 Schmid, Studien, 2.196, 199–200, 206, 208, 214, 226–7, 231 n. 3, 246.

37 The omission of the final καί in Rev 11.8 is instructive. The reading is attested (in Schmid's collation) by P⁴⁷, S^a and Αν. If S^a is dated to the fourth century, then there are two early Greek witnesses that join the later Andreas tradition in support of the omission; but if S^a is dated to the seventh century, there is only one early Greek witness to the reading, joined subsequently by S^a and Andreas. With Schmid's S^a correctly dated to the seventh century, the early attestation of the omission becomes all the more remarkable and increases the value of S^a as a witness to this variant. The up-to-date apparatus of NA²⁸ reinforces this impression, showing the support of tenth-century 1611, as well as a handful of versions (ar* sy^ph and bo) for the omission. See Schmid, Studien, 2.115–16, 125.

38 See Epp, ‘Textual Clusters’, 519–77.

39 No discussions of the Apocalypse's textual history are included in the most recent publications of contemporary text-critical research, accurately reflecting the ‘state of the question’. See Wachtel, K. and Holmes, M. W., eds., The Textual History of the Greek New Testament: Changing Views in Contemporary Research (Atlanta: Society of Biblical Literature, 2011)Google Scholar; Ehrman and Holmes, The Text of the New Testament in Contemporary Research.

40 The new edition was unveiled at the Society of Biblical Literature meeting in Chicago in 2012.

41 Colwell, ‘Method’, 53.

42 Indeed, Colwell left us with a list of items that could have been attended to in the intervening years. Issues that Colwell identified as meriting further consideration include the following. First, Schmid did not fully realise the implication of the data for the interrelationship of the major text types; according to Colwell, he overrated agreement in the original reading as evidence of common lineage and underrated the possibility of coincidental agreement of error. Second, Schmid mistakenly thought of text types as frozen blocks rather than as processes. Third, he did not demonstrate that the later text types are not revised forms of their elders; he only demonstrated that they were not entirely derived from the older ones. Colwell also called for clarification on Schmid's use of ‘later, revised forms’. Intriguingly, Colwell also appears to suggest that the Apocalypse may preserve the equivalent of a ‘Western’ text – despite the absence of Latin support – in some of the readings of P⁴⁷ א, especially if ‘Western’ is understood as ‘non-Neutral’. After all, he adds, ‘Western has long since ceased to be a geographic term’. See Colwell, ‘Method’, 52–4.

43 So Schmid, Studien, 2.7–8, 8 n. 1; cf. Birdsall, ‘The Text of the Revelation’, 229; Hernández, Scribal Habits, 24 n. 63.

44 Schmid, Studien, 2.x; cf. Kilpatrick, ‘Professor J. Schmid’, 8–10, 12; Birdsall, ‘The Text of the Revelation’, 236–7; Hernández, Scribal Habits, 24 with n. 61.

45 M. Karrer and U. Schmid, ‘Report on the Apocalypse Project’, SBL Annual Meeting, Chicago, 19 November 2012.

46 The variants listed are from Bousset's tally of א^c corrections that accord with the Andreas tradition (see Bousset, ‘Zur Textkritik der Apokalypse’, 42). The comparison illustrates the departure of Schmid (and in part NA²⁷) from the dating established by Tischendorf and followed by Bousset. Hoskier and Milne and Skeat are not included in the comparison but, as already noted, these also follow Tischendorf in their dating and identification of א^c.

47 Milne and Skeat, Scribes and Correctors, 18–21.

48 Milne and Skeat regard all subsequent corrections of the Apocalypse assigned to C^c as really belonging to C^c*. See Milne and Skeat, Scribes and Correctors, 50.

49 Milne and Skeat, Scribes and Correctors, 50.