The role of manuscript corrections in studying textual transmission of the New Testament has been long recognised by textual critics. And yet, the actual witness of corrections may at times be difficult to interpret. A case in point is Josef Schmid's seminal work on the text of Revelation. Following Wilhelm Bousset, Schmid argued that a particular group of corrections in Codex Sinaiticus reflected a Vorlage with a text akin to that of the Andreas text-type. By dating these corrections – unlike Bousset – to the scriptorium, Schmid utilised their witness to trace the text of Andreas back to the fourth century. Recently, Juan Hernández has shown that the corrections cited by Schmid were significantly later, hence calling his fourth-century dating of Andreas (among other things) into question. Through an analysis of the corrections cited by Schmid, supplemented by a fuller data-set of Sinaiticus’ corrections in Revelation, this study seeks to reappraise Schmid's claims concerning the textual relations of these corrections, and identify their role in the later transmission of the text of Revelation.
I would like to thank Juan Hernández Jr and the anonymous reviewer for their careful reading of an earlier draft of this article, as well as their helpful comments and corrections. All the remaining shortcomings are my own.
1 See J. Schmid, Studien zur Geschichte des griechischen Apokalypse-Textes, Teil 1: Der Apokalypse-Kommentar des Andreas von Kaisareia, Band 1: Text, Band 2: Einleitung, Teil 2: Die alten Stämme (Münchener theologische Studien 4; Munich: Karl Zink, 1955–6) 2.147.
2 Cf. D. C. Parker, Codex Sinaiticus: The Story of the World's Oldest Bible (London/Peabody, MA: British Library/Hendrickson, 2010) 7, who suggests a date at ‘[a]round the middle of the fourth century’. Conversely H. J. M. Milne and T. C. Skeat, Scribes and Correctors of the Codex Sinaiticus (London: British Museum, 1938) 61–2, suggest the first half of the fourth century.
3 W. Bousset, ‘Zur Textkritik der Apokalypse’, Textkritische Studien zum Neuen Testament (TU 2.4; Leipzig: Hinrichs, 1894) 1–44, esp. 42–4. See also Hernández, J. Jr, ‘The Creation of a Fourth-Century Witness to the Andreas Text Type: A Misreading in the Apocalypse's Textual History’, NTS (2014), 109. A thorough reappraisal of Bousset's theories has been recently undertaken in J. Hernández Jr, ‘The Legacy of Wilhelm Bousset for the Apocalypse's Textual History: The Identification of the Andreas Text’, Studien zum Text der Apokalypse (ANTF; ed. M. Sigismund, M. Karrer, U. Schmid; Berlin and New York: De Gruyter, forthcoming). Note that Bousset does not explicitly align the corrected readings of Sinaiticus with Andreas, but rather with K, the archetype of the text of Revelation that is preserved in the text of Andreas’ commentary. See further Hernández Jr, ‘The Legacy of Wilhelm Bousset’, 22–3; id., ‘Creation,’ 110 n. 19.
4 Milne and Skeat, Scribes and Correctors.
5 Schmid, Studien, 2.127.
6 Hernández Jr, ‘Creation’, esp. 109–10.
7 Cf. Bousset, ‘Textkritik’, 42; C. Tischendorf, ed., Novum Testamentum Graece (vols. i–ii, Leipzig: Hinrichs, 1869–18728; vol. iii: Prolegomena (scripsit C. R. Gregory), Leipzig: Hinrichs, 18948) iii.346.
8 Hernández Jr, ‘Creation’, 110–13.
9 Hernández Jr, ‘Creation’, 115.
10 The earliest layer of corrections in Revelation is a subject of my forthcoming study. On the earliest corrections of the Marcan portion, see my ‘The Earliest Corrections in Codex Sinaiticus: A Test Case from the Gospel of Mark’, BASP 50 (2013) 207–54.
11 Milne and Skeat, Scribes and Correctors, 65.
12 T. C. Skeat, ‘The Codex Sinaiticus, The Codex Vaticanus and Constantine’, Collected Biblical Writings of T. C. Skeat (introduced and edited by J. K. Elliott; NovTSup 113; Leiden: Brill, 2004) 200.
13 Cf. A. C. Myshrall, ‘Codex Sinaiticus, its Correctors, and the Caesarean Text of the Gospels’ (Ph.D. diss., University of Birmingham, 2005) 91: ‘The date suggested by Milne and Skeat as between the 5th and 7th centuries can thus be seen as reasonable, although I would tend to place Ca towards the first half of this period.’
14 See Hernández Jr, ‘The Legacy of Wilhelm Bousset’, 30–1 nn. 50–1. Incidentally, NA28 continues to date these corrections (designated as ℵ2) to the seventh century. Cf. Hernández Jr, ‘Creation’, 116, 118–19.
15 When using the ‘Sa’ siglum, I refer to Schmid's manner of citing these corrections rather than my own classification. Note that I exclude corrections listed in Schmid, Studien, 128 n. 2, as they play no role in Schmid's argument.
16 See E. J. Epp, ‘Toward the Clarification of the Term “Textual Variant”’, Perspectives of New Testament Textual Criticism: Collected Essays, 1962–2004 (NovTSup 116; ed. E. J. Epp; Leiden/Boston: Brill, 2005), 47–61.
18 Lembke, M., ‘Die Apokalypse-Handschrift 2846: Beschreibung, Kollation und Textwertbestimmung eines wichtigen neuen Zeugen’, NovT 54 (2012) 369–95.
19 H. C. Hoskier, 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 Versions, Commentaries and Fathers. A Complete Conspectus of All Authorities (2 vols.; London: Bernard Quaritch, 1929).
20 Schmid, Studien, 2.45–9, 51.
21 As will be seen below, a not insignificant number of the Ca corrections result in singular and weakly attested readings, suggesting that the corrector may occasionally have acted without consulting an exemplar. Where appropriate, then, alternative possibilities shall be noted throughout our analysis. Naturally, the weaker the attestation of a corrected reading, the stronger the case for its non-genetic origin.
22 See Schmid, Studien, 2.236–7. On solecisms in Revelation (including 4.1), see BDF, § 136.
23 On 7.9, see Schmid, Studien, 2.245–6.
24 Note esp. the early support of 0207. Incidentally, Schmid, Studien, 2.172 does not cite this correction in his collation of 0207.
25 See H. W. Smyth, Greek Grammar (rev. by G. M. Messing; Cambridge, MA: Harvard University Press, 1956) § 2878.
26 See B. M. Metzger, A Textual Commentary on the Greek New Testament (London/New York: United Bible Societies, 19942) 674: ‘the presence of και … appears to be due to copyists who wished to alleviate the strained syntax’.
27 So D. E. Aune, Revelation 6–16 (Word Biblical Commentary 52b; Nashville: Thomas Nelson, 1998) 715, 745, who translates the phrase as ‘to blaspheme his name and his dwelling, that is, those who dwell in heaven’ (emphasis added). In contrast, the meaning produced by the Ca corrector is: ‘to blaspheme his name, his dwelling, and those who dwell in heaven’.
28 Cf. J. Hernández Jr, Scribal Habits and Theological Influences in the Apocalypse: The Singular Readings of Sinaiticus, Alexandrinus, and Ephraemi (WUNT 2.218; Tübingen: Mohr Siebeck, 2006) 70, 72, 74, 84–5.
29 As far as I am aware, Schmid, Studien, refers only to seven Sc corrections at various points of his analysis: 2.2 (2.72), 2.4 (2.185), 11.1 (2.129, 240), 11.8 (2.129), 12.8 (2.129), 16.10 (2.129), 21.27 (2.70). Schmid's reference to 21.27, however, is incorrect, and he correctly cites it elsewhere as Sa. For further discussion of Cc* (and Cc) corrections, see Milne and Skeat, Scribes and Correctors, 49–50. Incidentally, NA28 cites the correction as ℵ2, which seems misleading as the same siglum is used for Ca corrections. The same applies to 11.4, 11.12, 14.8, 16.10, 16.18, and 19.13a, discussed below.
30 The Cc* corrections would in themselves warrant a separate treatment. Suffice it to say that not all of Cc*'s corrections betray the use of an exemplar. See e.g. 10.10, 11.3, 11.8, 12.6, 12.8b, 16.18, 18.9, 18.21, 19.7, where the corrected readings lack solid external support.
31 Incidentally, of the 24 occurrences of κυριοϲ ιηϲουϲ χριϲτοϲ in NA28, the formula never appears in the vocative. Apart from Acts 11.17; 28.31 and Jas 1.1, all the remaining occurrences are limited to the Pauline corpus: Rom 1.7; 13.14; 1 Cor 1.3; 6.11; 8.6; 2 Cor 1.2; 13.13; Gal 1.3; Eph 1.2; 6.23; Phil 1.2; 2.11; 3.20; 4.23; 1 Thess 1.1; 2 Thess 1.1; 1.12; 3.1; Phlm 3; 25.
32 Expansion of Christological titles is a common tendency in the transmission history. See further G. Zuntz, The Text of the Epistles: A Disquisition upon the Corpus Paulinum (The Schweich Lectures 1946; London: The British Academy, 1953), 182–4; Head, P.M., ‘A Text-Critical Study of Mark 1.1: “The Beginning of the Gospel of Jesus Christ”’, NTS 37 (1991) 627. See also B. M. Metzger and B. D. Ehrman, Text of the New Testament: Its Transmission, Corruption, and Restoration (New York/Oxford: Oxford University Press, 20054) 263–4.
33 Schmid, Studien, 2.57, 60.
34 CSP seems to be wrong in postulating a nonsensical reading τ̣ωγ. Admittedly, the letter is formed slightly differently than the following two gammas in that its horizontal bar lacks a finial. Apparently, though, Scribe A did not write this letter consistently, as both gammas in ϲυναγαγειν a line below are written without finials and resemble the present instance.
35 So also C. Tischendorf, ed., Novum Testamentum Sinaiticum sive Novum Testamentum cum epistula Barnabae et fragmentis Pastoris (Leipzig: F. A. Brockhaus, 1863) lxxv.
36 See L. Havet, Manuel de critique verbale appliquée aux textes latins (Paris: Hachette, 1911) §§ 429–32; D. C. Greetham, Textual Scholarship: An Introduction (Garland Reference Library of the Humanities 1417; New York/London: Garland, 1994) 280; D. Jongkind, Scribal Habits of Codex Sinaiticus (TS 3.5; Piscataway: Gorgias, 2007) 245.
37 Schmid, Studien, 2.72–3, 81–3.
38 Cf. Elliott, J. K., ‘A Short Textual Commentary on the Book of Revelation and the “New” Nestle’, NovT 56 (2014) 93.
39 Incidentally, Schmid, Studien, 2.73 settles the matter based on the witness of the versions, which otherwise do not figure prominently in his discussions.
40 See further Schmid, Studien, 2.24–5.
41 Schmid, Studien, 2.91, 94–5, 98–9, 102, 104–5.
42 Schmid treats 14.8–9 as three distinct readings (cited here as 14.8–9a, b, c), though it is actually only one correction. In discussing Schmid's citations, I have retained this (somewhat artificial) distinction, but I list it as a single reading in my final summary of Ca corrections below. The same applies to 9.13.
43 Cf. Schmid, Studien, 2.98; BDF, § 338(3).
44 Schmid, Studien, 2.207–8.
45 See Schmid, Studien, 2.104–5.
46 Schmid, Studien, 2.91.
47 So Schmid, Studien, 2.94.
48 So Schmid, Studien, 2.95.
49 Schmid, Studien, 2.114 n. 4, 115–16. Schmid (p. 125) also cites 11.8 as an ‘Analog’ to 20.10.
50 Cf. Tischendorf, Novum Testamentum Sinaiticum, lxxv, who only notes: ‘ν erasum’. In the same vein, Hoskier, Concerning the Text of the Apocalypse, 2.298, cites the correction as ‘ℵa vel c’.
51 Although it is difficult to determine on the basis of the digital image alone, it seems that, before the erasure, the deletion may have been marked by an obelus, as there are traces of a diagonal stroke written across the original diagonal of the nu. This kind of deletion is typical of scriptorium corrections. Cf. Malik, ‘Corrections’, 243–4 n. 134.
52 Cf. Elliott, ‘A Textual Commentary’, 98.
53 Schmid, Studien, 2.134.
54 Hernández Jr, Scribal Habits, 65.
55 Schmid, Studien, 2.142–3.
56 F. G. Kenyon, ed., The Chester Beatty Biblical Papyri, vol. iii: Pauline Epistles and Revelation, Text (London: Emery Walker, 1934) 29, edits: πεπτω[κε]ν. Schmid's (Studien, 2.142) citation of 𝔓47 as πεπτω[. .]ν seems unnecessarily conservative. Note that, later on, Schmid remarks that ‘Sa und 𝔓47 ebenfalls πεπτωκαν (oder evtl. πεπτωκεν) lesen’.
57 So Schmid, Studien, 2.143.
58 See e.g. E. C. Colwell, ‘Method in Evaluating Scribal Habits: A Study of P45, P66, P75’, Studies in Methodology in Textual Criticism of the New Testament (ed. E. C. Colwell; Leiden: Brill, 1969) 113–14; J. R. Royse, Scribal Habits in Early Greek New Testament Papyri (NTTSD 36; Leiden/Boston: Brill, 2008) 735, 737–8; Head, P. M., ‘Observations on Early Papyri of the Synoptic Gospels, especially on the “Scribal Habits”’, Bib 71 (1990) 246; id., ‘The Habits of New Testament Copyists: Singular Readings in the Early Fragmentary Papyri of John’, Bib 85 (2004), 407–8; Jongkind, Scribal Habits, 247, 254; Hernández Jr, Scribal Habits, 193.
59 Schmid, Studien, 2.154–5, and also 128 n. 2.
60 Schmid, Studien, 2.154.
61 Incidentally, the numeral in Origen's citation of 14.1 (Comm. Jo. 1.2), too, has an article – apparently without external support.
62 Hence, Hernández Jr, Scribal Habits, 83 classifies the reading as a change in tense. Alternatively, the Vorlage of Sinaiticus may have agreed with Origen (Comm. Jo. 2.5) in reading a pluperfect form εκεκλητο. If so, the scribe would have had to drop the augment and omit the article το by haplography. Less likely still is CSP's transcription κεκλη το.
63 Schmid, Studien, 2.157, 161, 165.
64 See further Schmid, Studien, 2.101.
65 Schmid, Studien, 2.172.
66 Schmid, Studien, 2.172.
67 Schmid, Studien, 2.180, 182–3, and also 128 n. 2.
68 It seems that the original present αφειϲ fits better with the preceding εχω κατα ϲου: Jesus’ rebuke comes precisely at the time when the church was still tolerating Jezebel.
69 Schmid, Studien, 2.196, 199.
70 Cf. Jongkind, Scribal Habits, 244.
71 Note that Schmid elsewhere cites S* (= ℵ*) in support of the omission at 11.16. Cf. Schmid, Studien, 2.135.
72 Schmid, Studien, 2.200.
73 Schmid, Studien, 2.206.
74 Schmid, Studien, 2.214.
75 On interchange of ου and ω(ι), see F. T. Gignac, A Grammar of the Greek Papyri of the Roman and Byzantine Periods, vol. i: Phonology, vol. ii: Morphology (Testi e Documenti per lo Studio dell'Antichità 55; Milan: Istituto Editoriale Cisalpino–La Goliardica, 1976–81) i.208–9.
76 Schmid, Studien, 2.226–8.
77 On this type of scribal behaviour, see Royse, Scribal Habits, 157; Colwell, ‘Scribal Habits’, 116; Hernández Jr, Scribal Habits, 75.
78 Thus, the Vorlage of Sinaiticus most likely accorded with A 0163 1006 1611 1841 1854 2053 2062, and not its usual ally 𝔓47, which is here followed by 051 2329 𝔐K.
79 So Tischendorf, Novum Testamentum Sinaiticum, lxxv: ‘Cc improbavit (licet και intactum reliquerit)’.
80 Cf. Schmid, Studien, 2.226.
81 See our previous discussion, above.
82 Schmid, Studien, 2.231 n. 3.
83 Cf. Hernández Jr, Scribal Habits, 121.
84 Schmid, Studien, 2.246.
85 So Schmid, Studien, 2.246.
86 Cf. Tischendorf, Novum Testamentum Sinaiticum, lxxiv–lxxvi, who assigns Cc, as he does not distinguish between Cc and Cc* correctors. CSP's more precise assignment follows Milne and Skeat, Scribes and Correctors, 49–50.
87 Milne and Skeat, Scribes and Correctors, 50.
88 At 4.1, 𝔐A is joined by 2846, which was not available to Schmid at the time of his writing and, at 7.9, we have the support of 2053 with the text of Oecumenius. The support of 2053 at 13.6 is only partial, as the reading does not appear in the commentary, and so it is classified in the former group. The support of these minuscules with an early text is so treated throughout the following discussion.
89 I also exclude four corrected itacisms at 1.17a, 2.16, 14.11, 20.3 and ten reverted corrections at 2.14b, 3.12d, 9.4, 9.18, 10.2, 11.9a, 15.3b, 16.16, 18.13b, 19.15.
90 As noted in n. 42 above, Schmid artificially divides the single corrections at 9.13 and 14.8–9 into three distinct variation-units. As seems clear from their classification under weakly attested readings, taking the corrections in their entirety may alter their textual character rather significantly.
91 Pace Bousset, ‘Textkritik’, 42.
92 Cf. Schmid's (Studien, 2.127) reference to Bousset, ‘Textkritik’, 42, quoted above. As it is, even Bousset's list of ℵc readings agreeing with 𝔐A (his K) is not without problems. In fact, only fourteen of the twenty-eight readings he lists (1.20, 2.13, 4.1, 4.5, 6.15, 7.9b, 7.9c, 9.12b, 10.1, 11.8b, 13.6b, 16.17, 17.8, 22.20b) are clearly readings of 𝔐A (at 1.20, 7.9c, 9.12b, 11.8b and 13.6b with earlier support). At 1.6a, 9.14, 21.9 and 22.2, the resultant reading is singular and seems better explained as Ca's grammatical improvement rather than his failure to transmit the Vorlage. Further, 3.5 and 21.20c agree with both 𝔐A and 𝔐K; 18.8b is supported by 1 250 424 616 2048 2186 2428 al; 20.9 is supported by P 2053txt 2846; and 21.23 is supported by 051s 2030 2377 pc. Bousset's collation appears to be incorrect at 2.20c, 6.9, 7.10b, 8.11a and 10.2. Cf. Hernández Jr, ‘The Legacy of Wilhelm Bousset’, 23–5.
93 On the role of manuscript corrections in textual transmission, see Royse, Scribal Habits, 75–6; M. W. Holmes, ‘Working with an Open Textual Tradition: Challenges in Theory and Practice’, The Textual History of the Greek New Testament: Changing Views in Contemporary Research (ed. K. Wachtel and M. W. Holmes; SBL Text-Critical Studies 8; Atlanta: Society of Biblical Literature, 2011) 69.
94 Schmid, Studien, 2.146. Ironically, though the text used in Andreas' commentary is replete with grammatical improvements, Andreas himself condemns scribes who atticised the text of Revelation. See Hernández, J. Jr, ‘The Relevance of Andrew of Caesarea for New Testament Textual Criticism’, JBL 130 (2011) 188–91, 194–5.
95 Schmid, Studien, 2.53. Schmid also notes a possibility that a small portion of improvements could have been inherited.
96 See Schmid, Studien, 2.52–3. The Sa corrections assisted Schmid in tracing the Andreas text to the fourth century.
97 E. C. Colwell, ‘Method in Establishing the Nature of Text-Types of New Testament Manuscripts’, Studies in Methodology in Textual Criticism of the New Testament (ed. E. C. Colwell; Leiden: Brill, 1969) 53.
98 So Colwell, ‘Text-Types’, 52. See also Hernández Jr, ‘Creation’, 107–8, 114.
99 Cf. Colwell, ‘Text-Types’, 52–3: ‘Revision almost universally proceeded on a documentary basis. Manuscripts from outside the text-type were used to revise it. This opened a door through which Ur-text readings could be added just as surely as it opened the door to alien corrections and corruptions.’
* I would like to thank Juan Hernández Jr and the anonymous reviewer for their careful reading of an earlier draft of this article, as well as their helpful comments and corrections. All the remaining shortcomings are my own.
Email your librarian or administrator to recommend adding this journal to your organisation's collection.
* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.
Usage data cannot currently be displayed