1 Hunt, A. S., ed., The Oxyrhynchus Papyri, vol. viii (London: Egypt Exploration Fund: Graeco-Roman Branch 11, 1911) 13–14Google Scholar.

2 An image of P.Oxy. viii.1079 is also available online at http://csntm.org/Manuscript/View/GA_P18 (accessed 15 March 2018).

3 Hunt, The Oxyrhynchus Papyri viii, 13. This date is also adopted in NA²⁸.

4 Orsini, P. and Clarysse, W., ‘Early New Testament Manuscripts and their Dates: A Critique of Theological Palaeography’, ETL 88 (2012) 443–74Google Scholar, at 459 and 469 (Table 1).

5 Rev 1.6 (l. 12): του θ̅υ̅ > τω θ̅ω̅.

6 So already Schmid, J., Studien zur Geschichte des griechischen Apokalypse-Textes, Teil 1: Der Apokalypse-Kommentar des Andreas von Kaisareia, Band i: Text, Band ii: Einleitung; Teil 2: Die alten Stämme (Münchener theologische Studien 4; Munich: Karl Zink, 1955–6)Google Scholar ii.171. See also Hunt, The Oxyrhynchus Papyri viii, 13.

7 For Hunt's edition of the Exodus text, see The Oxyrhynchus Papyri viii, 5–6. Hunt noted that the ‘sloping uncial’ hand of P.Oxy. viii.1075 was unlikely to be ‘later than the third century’.

8 Nongbri, B., ‘Losing a Curious Christian Scroll but Gaining a Curious Christian Codex’, NovT 55 (2013) 77–88Google Scholar.

9 Nongbri, ‘Losing’, 79–83. See Charlesworth, S. D., ‘A Reused Roll or a “Curious Christian Codex”? Reconsidering British Library Papyrus 2053 (P.Oxy. 8.1075 + P.Oxy. 8.1079)’, Buried History: Journal of the Australian Institute of Archaeology 53 (2017) 35–44Google Scholar, at 35–6, who has recently called Nongbri's calculations into question, noting that the hypothetical codex would not fit into ‘Turner's Group 8 proper’, though acknowledging that it would probably fit among the sub-group of aberrant cases within that group. In the end, this counter-argument does not have much force, given the flexibility of Turner's groupings in general (as noted below).

10 For a recent re-edition and textual analysis, see Malik, P., ‘Another Look at P.IFAO ii 31 (⁹⁸): An Updated Transcription and Textual Analysis’, NovT 58 (2016) 204–17Google Scholar.

11 For a thorough discussion of the column width in bookrolls, see Johnson, W. A., Bookrolls and Scribes in Oxyrhynchus (Toronto: University of Toronto Press, 2004) 100–19Google Scholar. In prose texts, the widths range from 4.5 to 10 cm, 4.5–7 cm being the normative range; the narrower columns were more common in the second century, whereas the wider ones occur more frequently in the third (p. 113). Hence, the difference between our fragment's column width and the ‘common’ trends in the contemporary bookrolls is not so significant as it might seem.

12 Nongbri, ‘Losing’, 80 n. 11.

13 See the relevant tables in Turner, E. G., The Typology of the Early Codex (Philadelphia: University of Pennsylvania Press, 1977) 14–22CrossRefGoogle Scholar. The attested breadths range from 2.6 to 37 cm, and the height from 2.9 to 41 cm.

14 Nongbri, ‘Losing’, 79. Nongbri's argument is cited with approval by Cate, J., ‘The Curious Case of ⁴³’, Book of Seven Seals: The Peculiarity of Revelation, its Manuscripts, Attestation, and Transmission (ed. Kraus, T. J. and Sommer, M.; WUNT i/363; Tübingen: Mohr Siebeck, 2016) 33–49Google Scholar, at 42 n. 36, who states: ‘Normally, an opisthograph was made from a scroll which had become obsolete or discarded. It is hard to imagine a copy of Exodus being used in such a manner for a copy of the Apocalypse, even if a copy of Exodus scroll had suffered damage.’ In support of this, Cate adduces opisthographs from the Judean desert, none of which included reused biblical texts. The difficulty with this argument, of course, is that the early Christians need not have shared the same attitude concerning the reuse of scriptural manuscripts as the contemporary Jewish communities did. A fitting example are palimpsests found in the Cairo Geniza, some of which involved what were originally New Testament manuscripts. Similar reuse of Old Testament manuscripts, on the other hand, hardly ever took place.

15 Nongbri, ‘Losing’, 83–4.

16 For further details, see Nongbri, B., ‘The Construction of P.Bodmer viii and the Bodmer “Composite” or “Miscellaneous” Codex’, NovT 58 (2016) 394–410Google Scholar; Turner, Typology, 79–81.

17 See e.g. Turner, Typology, 79–81, who instances the Coptic miscellaneous codex BL MS Or. 7954 (LDAB 107763), which contains Deuteronomy (145 pages), followed by Jonah (8 pages) and Acts (112 pages). The difficulty with this example, however, is that Turner dated it ‘before ad 350’ whereas the dating adopted in the manuscript's LDAB entry is in the range of 350–450 ce. The LDAB dating follows Orsini, P., ‘La maiuscola biblica copta’, Segno e Testo 6 (2008) 121–50Google Scholar, at 133–4.

18 The Exodus–Revelation sequence seems difficult to account for on literary grounds, and hence is more likely to have been motivated by economic factors. But see Epp, E. J., ‘The Oxyrhynchus New Testament Papyri: “Not without Honor Except in their Hometown?”’, Perspectives of New Testament Textual Criticism: Collected Essays, 1962–2004 (NovTSup 116; Leiden/Boston: Brill, 2005) 743–801CrossRefGoogle Scholar, at 758–9 for a discussion of possible (if somewhat stretched) thematic links between the two adjacent texts. The possibility of intertextual links is also acknowledged by T. Nicklas, ‘Christliche Apokalypsen in Ägypten vor Konstantin: Kanon, Autorität, kontextuelle Funktion’, Book of Seven Seals, 96–117, at 101, who, however, ultimately rejects the notion that such factors played any role in the production/reuse of our papyrus.

19 In the case of single-quire construction (the most common kind of papyrus codex in this period), one would have to reckon with some sixty vacant pages in a pre-bound codex that would have to be filled up by the text of Revelation – a rather extraordinary scenario. Just as extraordinary, however, would be if the scribe appended scores of leaves (arranged in quires) to contain the entire Apocalypse whose copying was begun on the final few vacant pages of the Exodus codex.

20 In addition, Charlesworth, ‘Reused Roll’, 38–9, notes the unevenness of text blocks between the respective sides of our papyrus. In this vein, he states: ‘Ordinarily, a scribe copying a codex, even a second scribe as here, would want to maintain the uniform appearance of the codex by producing a leaf with text blocks that were as complementary as possible’ (p. 38). This, however, need not have been so in the case of an informally produced miscellany, which after all could have been produced in multiple settings and over a period of time. In the absence of firmer data, which could be provided only by the recovery of further portions of the manuscript – an unlikely scenario – it is impossible to press this argument too far.

21 Nongbri, ‘Losing’, 79, 84–8. Nongbri gives several instances of reused rolls among early Christian papyri (i.e. British Library Pap. 1532, PSI viii.921, P.Lips. i.97, P.IFAO ii.31, P.Mich. inv. 44 and P.Oxy. lxix.4705), noting that he was aware of only one early Christian manuscript deviating from this pattern – P.Oxy. iv.654, a fragment of the Gospel of Thomas.

22 From my survey of the P.Oxy. volumes, unfortunately only a minority of editiones principes have yielded relevant information. Among those that do, however, the upside-down pattern is, as one might expect, more prevalent.

23 Minnen, P. van, ‘From Possidipus to Palladas: What Have Literary Papyri Done for Us?’, JJurPap 42 (2013) 243–61Google Scholar, at 245. Incidentally, Van Minnen's article is also cited in the LDAB database, which classifies P.Oxy. viii.1079 unequivocally as a roll. In addition, Blumell, L. H. and Wayment, T. A., eds., Christian Oxyrhynchus: Texts, Documents, and Sources (Waco: Baylor University Press, 2015) 91Google Scholar and Mugridge, A., Copying Early Christian Texts: A Study of Scribal Practice (WUNT i/362; Tübingen: Mohr Siebeck, 2016) 279CrossRefGoogle Scholar also regard the manuscript as a roll, but curiously omit a reference to Nongbri's article from the bibliography.

24 Van Minnen, ‘From Possidipus to Palladas’, 245.

25 Van Minnen, ‘From Possidipus to Palladas’, 245. Van Minnen's assertion is reinforced by Charlesworth, ‘Reused Roll’, 40, who observes that ‘several vertical crevices are visible … one of which runs down the length of P.Oxy. 8.1079. In addition, it is clear that the ink was applied after the crevices had formed.’

26 Van Minnen, ‘From Possidipus to Palladas’, 245.

27 J. Chapa, ‘The “Jewish” Septuagint Papyri from Oxyrhynchus’ (Paper presented at the ‘Papyri, Septuagint, Biblical Greek’ conference, Strasbourg, 29–30 September 2017) 14 n. 62. I am grateful to Juan Chapa for sharing a pre-publication version of this work with me (personal correspondence, 7 October 2017).

28 J. Chapa, ‘Septuagint Papyri’, 13–14.

29 J. Chapa, ‘Septuagint Papyri’, 12–13.

30 Roberts, C. H., Manuscript, Society and Belief in Early Christian Egypt (The Schweich Lectures 1977; London: Oxford University Press for The British Academy, 1979) 9Google Scholar. A similar line of reasoning is also adopted by Charlesworth, ‘Reused Roll’, 37–40 and Mugridge, Copying Early Christian Texts, 279. It needs to be said, however, that informality in writing need not indicate the private nature of the manuscript's use (intended or otherwise). Particularly problematic in this respect is Charlesworth's argument from scribal practice (p. 36–7): singular error, imperfectly executed correction, or ‘unusual’ nomina sacra may be observed in many of the highly formalised manuscripts, most notably so in Codex Sinaiticus (GA 01; LDAB 3478; TM 62315). For a detailed analysis, see Jongkind, D., Scribal Habits of Codex Sinaiticus (TS 3.5; Piscataway: Gorgias Press, 2007)Google Scholar.

31 T. J. Kraus, ‘“When Symbols and Figures Become Physical Objects”: Critical Notes about Some of the “Consistently Cited Witnesses” to the Text of Revelation’, Book of Seven Seals, 51–69, at 59.

32 Roberts, Manuscript, Society and Belief, 9.

33 For a perceptive discussion of the economics of book production, see Bagnall, R. S., Early Christian Books in Egypt (Princeton: Princeton University Press, 2009) 50–69Google Scholar. As regards ‘private’ ownership of books, particularly perceptive is Bagnall's comparison (p. 62) of income vis-à-vis ability to buy books between lower-level clergy such as readers on the one end of the spectrum and priests and bishops on the other. Note, however, that the figures used for those calculations come from the sixth century; thus, the ability to acquire books in the pre-Constantinian period would have been even more limited. As for the possible book-owning Christians in the early period, Bagnall tentatively suggests that they may have belonged to the ‘urban elite formed in the aftermath of the creation of of the city councils of the metropoleis of the nomes after 200’ (p. 67).

34 In this vein, see also G. Bazzana, ‘“Write in a Book What You See and Send It to the Seven Assemblies”: Ancient Reading Practices and the Earliest Papyri of Revelation’, Book of Seven Seals, 11–31, at 17, who, though accepting Nongbri's proposal, nonetheless rightly suggests that P.Oxy. viii.1079 reflects a ‘heuristic inadequacy’ of the ‘public/private’ binary. Even so, his counter-proposal, namely that our manuscript ‘points towards a (small) circle of intellectuals who had some personal interest in Exodus and Revelation’, begs the question; indeed, it rests on the assumption that a church could not have used a manuscript produced so ‘economically’. As regards the ‘public/private’ binary, see the pertinent remark (made in the context of miniature codices) by Choat, M. and Yuen-Collingridge, R., ‘The Egyptian Hermas: The Shepherd in Egypt before Constantine’, Early Christian Manuscripts: Examples of Applied Method and Approach (ed. Kraus, T. J. and Nicklas, T.; TENT 5; Leiden/Boston: Brill, 2010) 191–212CrossRefGoogle Scholar, at 199 n. 53: ‘To our mind, the association between Christian miniatures and “private use” is as insecure as the “public/private use” dichotomy is unhelpful.’ Just like reused copies, miniature codices have been regarded as ‘private’ products, due to their presumed limited facility for public reading. On miniature codices in general, see the recent article by Kraus, T. J., ‘Miniature Codices in Late Antiquity: Preliminary Remarks and Tendencies about a Specific Book Format’, EC 7 (2016) 134–52Google Scholar.