1 Elliott, J. K., ‘The Biblical Manuscripts of the John Rylands University Library of Manchester’, Bulletin of the John Rylands University Library 81 (1999) 3–50Google Scholar, at 6–8. The same author has written more recently in a similar fashion that P.Ryl. iii.457 ‘is the famous papyrus fragment containing four verses of John 18 and known to New Testament scholars as P52. Experts are generally agreed that it was written by the mid-second century, and thus this tiny fragment is not only our first witness to this Gospel, but the oldest example of any New Testament text in the world – and probably the earliest Christian writing extant.’ See J. K. Elliott, ‘Manchester Bibles’, Times Literary Supplement 5613 (29 October 2010) 15.

2 Roberts, C. H., An Unpublished Fragment of the Fourth Gospel in the John Rylands Library (Manchester: Manchester University Press, 1935)Google Scholar; republished with minor changes under the same title in the Bulletin of the John Rylands Library 20 (1936) 44–55. The Rylands fragment of John's gospel is inventoried as number 2774 in the Leuven Database of Ancient Books (hereafter LDAB).

3 B. Nongbri, ‘The Use and Abuse of P52: Papyrological Pitfalls in the Dating of the Fourth Gospel’, HTR 98 (2005) 23–48. For basic details about the papyrus and a discussion of scholarship on the fragment prior to 2005, I refer the reader to this article.

4 Published accounts of the acquisition details are vague; see Henry Guppy's ‘Prefatory Note’ to Roberts, An Unpublished Fragment, 5–9.

5 See Choat, M., ‘Lord Crawford's Search for Papyri: On the Origin of the Rylands Papyrus Collection’, Actes du 26^e Congrès international de papyrologie (ed. Schubert, P.; Geneva: Librairie Droz, 2012) 141–7Google Scholar.

6 For details, see R. Mazza, ‘Graeco-Roman Egypt at Manchester: The Formation of the Rylands Papyri Collection’, Actes du 26^e Congrès international de papyrologie, 499–507.

7 For more detailed discussion of Grenfell's 1920 purchases, see T. Hickey's introduction to his forthcoming editions of papyri from the British Library.

8 Letters in the archive of the John Rylands Library written by Grenfell on these dates mention the purchases (JRL/4/1/1/1920/Grenfell).

9 JRL/4/1/1/1920/Grenfell.

10 Elsewhere, Grenfell mentions that he had ‘damped out’ the papyri he purchased during this trip after his return, suggesting again that he had looked with some care over each individual papyrus (British Library Add. Ms. 59511, f. 114, letter from Bernard P. Grenfell to Harold Idris Bell, 4 April 1920). On this process, see the memorandum of Hunt, A. S., ‘The Damping out and Flattening of Papyri’, The Oxyrhynchus Papyri, vol. l (London: Egypt Exploration Society, 1983) viGoogle Scholar.

11 The prominent letters οι ϊουδαιο̣ι̣ on the top line of the recto would have clearly marked the piece as ‘theological’, to use Grenfell's preferred term for material relating to Christians or Jews.

12 It is not surprising that Roberts would assign much earlier dates to Christian material than would Grenfell and Hunt. This phenomenon would recur in coming decades; see Nongbri, B., ‘Grenfell and Hunt on the Dates of Early Christian Codices: Setting the Record Straight’, BASP 28 (2011) 149–62Google Scholar.

13 The latest letter is dated 17 July 1920, and in it Grenfell expressed hopes to return to Egypt in the coming winter.

14 See Hunt, A. S.'s obituary for Grenfell in Aegyptus 8 (1927) 114–16Google Scholar, at 115.

15 I have not been able to determine the exact date that Hunt received custody of the papyri.

16 It is not clear what part of the material Hunt saw. In a letter to Guppy dated 14 November 1933, Hunt stated that he had ‘been looking through the later (unpublished) documents this afternoon’ (JRL/4/1/1/1933/Hunt). In 1938, Guppy stated that Hunt had done no more than ‘a little preliminary sorting’. See Guppy's preface to Roberts, C. H., Catalogue of the Greek and Latin Papyri in the John Rylands Library Manchester, vol. iii (Manchester: Manchester University Press, 1938) ixGoogle Scholar.

17 For biographical details on Roberts, see his obituary by Russell, A. F. M. and Parsons, P. J., ‘Colin Henderson Roberts 1909–1990’, Proceedings of the British Academy 84 (1994) 479–83Google Scholar.

18 JRL/4/1/1/1935/Roberts.

19 In a letter of 25 July 1935, Roberts wrote to Guppy, ‘It is important to get it photographed as soon as possible (for one thing, I should like to send prints of the photograph to several scholars both in England and abroad) and so I will send the ms. to you by registered post on Saturday next, together with a transcript’ (JRL/4/1/1/1935/Roberts).

20 See the literature cited in Nongbri, ‘The Use and Abuse of P52’, 30 n. 19. A notebook of press clippings at the Rylands Library gives more evidence of the stir that the publication of the fragment caused in the popular press (JRL/7/7/7, which covers 1 October 1931through 13 March 1936).

21 Images of all of these papyri can be found collected in Nongbri, ‘The Use and Abuse of P52’, 32–40.

22 There were exceptions that reported dates both narrower and earlier than Roberts had reported. One example, which also highlights the dimension of civic pride involved in the publication, comes from the Manchester Evening Chronicle (18 November 1935). Under the headline ‘Important Bible Find for City: May Dwarf Previous Discoveries’, the article begins: ‘World-wide interest has been aroused by the announcement that Rylands Library, Manchester, has acquired a fragment of papyrus dating about ad 100–120’ (JRL/7/7/7, p. 170).

23 A. Deissmann, ‘Ein Evangelienblatt aus den Tagen Hadrians’, Deutsche allgemeine Zeitung 564 (3 December 1935). The piece was quickly translated into English as ‘The New Papyrus Fragment of the Fourth Gospel’, The British Weekly (12 December 1935). Deissmann's fuller statement in regard to the date was this: ‘I have chosen the date of Hadrian's reign (117–138 ad) for the title of this article, and in so dating I have intentionally kept within the bounds of prudence, for it is by no means out of the question that we might go back to the time of Trajan, who died in 117 ad.’

24 Deissmann's ungrounded assessment of the date of P52 is echoed, for example, in W. F. Albright's discussion of ‘The New Testament and Archaeology’, in which he asserts that P52 ‘is written in a hand attributed to the time of Trajan or more probably Hadrian’. See Albright, W. F., The Archaeology of Palestine (Harmondsworth: Penguin Books, 1949) 239Google Scholar. Roberts himself became fond of invoking Deissmann's earlier dating. See, for example, Roberts, C. H., ‘The Rylands Collection of Greek and Latin Papyri’, Bulletin of the John Rylands Library 36 (1953) 97–110CrossRefGoogle Scholar, at 98.

25 Colwell, E. C., review of Roberts, An Unpublished Fragment, The Journal of Religion 16 (1936) 368–9Google Scholar, at 368.

26 See the literature cited in Nongbri, ‘The Use and Abuse of P52’, 30–1 n. 22.

27 Comfort, P. W. and Barrett, D. P., The Text of the Earliest New Testament Greek Manuscripts: A Corrected Enlarged Edition (Wheaton: Tyndale House, 2001) 367Google Scholar; and Jaroš, K., Das neue Testament nach den ältesten griechischen Handschriften (Mainz: Verlag Franz Philipp Rutzen, 2006) 273Google Scholar: ‘Eine Datierung ist daher etwa ab 80 möglich und nach 125 unwahrscheinlich.’

28 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 (Grand Rapids: Eerdmans, 1989 ²) 85Google Scholar. A handful of scholars dissented from this view, most notably Andreas Schmidt. In a two-page article published in 1989, Schmidt suggested a date around 175 on the basis of comparison with (the palaeographically dated) Chester Beatty biblical papyri x (LDAB 3090, the Daniel leaves) and iii (LDAB 2778). See Schmidt, A., ‘Zwei Anmerkungen zu P. Ryl. iii 457’, Archiv für Papyrusforschung 35 (1989) 11–12Google Scholar. Yet, since the Beatty papyri are themselves palaeographically dated, this argument carries little weight. Others who have taken a more cautious approach to P52 include Pickering, S. R., ‘Short Notes’, New Testament Textual Research Update 2 (1994) 5–6Google Scholar; Ehrman, B. D., ‘The Text as Window: New Testament Manuscripts and the Social History of Early Christianity’, The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis (ed. B. D. Ehrman and M. W. Holmes; Grand Rapids: Eerdmans, 1995) 371–2Google Scholar and n. 49; and Hurtado, L. W., ‘P52 (P. Rylands Gk. 457) and the Nomina Sacra: Method and Probability’, Tyndale Bulletin 54 (2003) 7Google Scholar n. 20.

29 For a more detailed discussion of some of these difficulties with evidence from the papyri themselves, see Nongbri, B., ‘The Limits of Palaeographic Dating of Literary Papyri: Some Observations on the Date and Provenance of P.Bodmer ii (P66)’, Museum Helveticum 71 (2014) 1–35Google Scholar, esp. 19–20.

30 Comfort and Barrett, The Text of the Earliest New Testament Greek Manuscripts, 367. See the criticisms laid out in Nongbri, ‘The Use and Abuse of P52’, 32 n. 26.

31 Nongbri, ‘The Use and Abuse of P52’.

32 Turner, E. G., The Typology of the Early Codex (Philadelphia: University of Pennsylvania Press, 1977) 100CrossRefGoogle Scholar.

33 Bagnall, R. S., Early Christian Books in Egypt (Princeton: Princeton University Press, 2009) 12Google Scholar. For similar reactions, see e.g. Parker, D. C., An Introduction to the New Testament Manuscripts and their Texts (Cambridge: Cambridge University Press, 2008) 324CrossRefGoogle Scholar; and Barker, D., ‘The Dating of New Testament Papyri’, NTS 57 (2011) 571–82CrossRefGoogle Scholar, at 573–5.

34 Porter's paper did not appear in the proceedings of the congress, but it was recently published as part of a volume of collected essays: Porter, S. E., ‘Recent Efforts to Reconstruct Early Christianity on the Basis of its Papyrological Evidence’, Christian Origins and Greco-Roman Culture: Social and Literary Contexts for the New Testament (ed. S. E. Porter and A. W. Pitts; Leiden: Brill, 2013) 71–84CrossRefGoogle Scholar.

35 Problems surrounding the dating of the Egerton gospel (LDAB 4736) will be discussed in some detail below.

36 See the discussions in Nongbri, ‘The Use and Abuse of P52’, 34–5 and Bagnall, Early Christian Books, 11–15.

37 Porter, ‘Recent Efforts’, 79.

38 Thompson, E. Maunde, An Introduction to Greek and Latin Palaeography (Oxford: Clarendon, 1912) 144–7Google Scholar and Turner, E. G., Greek Manuscripts of the Ancient World (BICS 46; London: Institute of Classical Studies, 1987²) 2Google Scholar. Such tables can of course have some use for dating purposes (as long as the tables themselves are derived from securely dated manuscripts), although considering letters in isolation can cause one to lose the important perspective of letters’ relationships to each other and their layout on the page. Of the ‘second century’ alphabets to which Porter appeals in Maunde Thompson and Turner, only the so-called Hawara Homer (LDAB 1695) is relatively datable. Although Maunde Thompson himself had earlier assigned the manuscript to the fifth century (see Petrie, W. M. Flinders, Hawara, Biahmu, and Arsinoe (London: Field & Tuer, 1889) 24Google Scholar), Frederic G. Kenyon made a case for a second-century date based on, among other things, the presence of marginal notes he attributed to the third century (see Kenyon, G. G., The Palaeography of Greek Papyri (Oxford: Clarendon, 1899) 101–3Google Scholar). In any event, these considerations are distinct from the question of whether the writing of the Hawara Homer or the other alphabets actually bears much resemblance to that of P52 (Porter simply asserts rather than arguing the point).

39 Porter, ‘Recent Efforts’, 83–4. Porter is undeterred by Michael Gronewald's redating of the Egerton Gospel to the beginning of the third century; see Gronewald, M.et al., eds., Kölner Papyri, vol. vi (Opladen: Westdeutscher Verlag, 1987) 136–45Google Scholar and Plate v; and also Foster, P., ‘Bold Claims, Wishful Thinking, and Lessons about Dating Manuscripts from Papyrus Egerton 2’, The World of Jesus and the Early Church: Identity and Interpretation in Early Communities of Faith (ed. C. A. Evans; Peabody, MA; Hendrickson, 2011) 193–211Google Scholar.

40 P. J. Parsons, review of G. Cavallo, Libri scritture scribi a Ercolano, CR n.s. 39 (1989) 358–60, at 360.

41 Porter, ‘Recent Efforts’, 74. See the similar comments at 72 and 78.

42 Bell, H. I. and Skeat, T. C., Fragments of an Unknown Gospel and Other Early Christian Papyri (London: Trustees of the British Museum, 1935) 1Google Scholar.

43 Bell and Skeat, Fragments of an Unknown Gospel, 2.

44 Bell and Skeat, Fragments of an Unknown Gospel, 7 (emphasis added).

45 Bell's preface to Fragments of an Unknown Gospel, vi.

46 The existence of the Egerton gospel was made public on 23 January 1935 in an article for the Times by Harold Idris Bell. Bell described the manuscript as being ‘written in a literary hand which appeared to date from a period not later than the middle of the second century’ (‘A New Gospel: Fragments from Egypt’, Times (23 January 1935) 13–14, at 13).

47 ‘Die Handschrifte ist ohne Frage alt, nur weiss ich nicht, ob man sie mit Sicherheit vor 150 n. Chr. setzen darf. Ähnlichen Stil habe ich bei datierten Urkunden und Briefen, die allein weiter helfen können, sowohl in der Zeit Hadrians wie des Marcus und Commodus gefunden. Deshalb möchte ich die Zeitgrenzen etwas weiter setzen. Es bleibt nun einmal eine der schwierigsten Aufgaben, solche Buchschrift zu datieren’ (British Library Add. Ms. 59519, ff. 78–9, letter from Wilhelm Schubart to Harold Idris Bell, 11 February 1935). Schubart continued in a more sanguine way, expressing hope that the accumulation of more data in the future might make more confident palaeographic dating possible (‘Hoffentlich kommt man später durch umfassende Bearbeitung des gesamten Materials zu festeren Ergebnissen, möglich wird es sein, das glaube ich auch’). What has instead occurred is a recognition that a variety of styles were in simultaneous use and individual types or styles of writings remained in use for longer periods of time than previously suspected. For documentary papyri, this observation can be most easily confirmed by spending a few hours browsing the PapPal website (http://pappal.info).

48 Given Schubart's hesitance to exclude dates later than 150 in the letter, I wonder how broadly he understood the words ‘middle of the second century’, the date that he is said to have pronounced ‘as good as certain’. For the phrase could be interpreted as ‘about 150, give or take a few years’, which seems to be the way Bell and Skeat use the term, with their emphasis on the first half of the second century. But the phrase could also be taken as meaning something like ‘115–85’ or something similar, which would bring us nearer to Schubart's stated range of Hadrian to Commodus. Bell and Skeat were a bit more specific in a more popular treatment of the Egerton gospel published later in 1935: ‘the manuscript was written not very far from the middle of the second century, with perhaps some preference for a date slightly before over one after ad 150. If the upward and downward limits of date be fixed at respectively 130 and 165 we shall probably not be far wrong as to the period within which the manuscript is most likely to have been written’; see Bell, H. I. and Skeat, T. C., The New Gospel Fragments (London: Trustees of the British Museum, 1935) 10Google Scholar.

49 It is possible that Schubart may have had other communication with Bell pertinent to this discussion that is unknown to me, but I did not see other relevant materials in Bell's correspondence in the British Library.

50 Porter's chapter contains other similar inaccuracies. Consider his appeal to the authority of Ulrich Wilcken. Porter claims that while Roberts opted for a fifty-year span of possible dates for P52, ‘Wilcken restricted the date [of P52] … to around ad 117–120’ (‘Recent Efforts’, 73 and 76). Porter again gives no specific citation in his support of this claim. I know of only one place in which Wilcken mentioned the Rylands gospel of John, a brief aside in a short article, ‘Die Bremer Papyrus-Sammlung’, Forschungen und Fortschritte 12 (1936) 89–90 (cited in Nongbri, ‘The Use and Abuse of P52’, 30 n. 22). What does Wilcken actually say? ‘Das Johannesevangelium, von dem C. H. Roberts soeben ein kleines Fragment veröffentlicht hat, das er mit Recht in die erste Hälfte des 2. Jahrhunderts setzt, könnte vom paläographischen Standpunkt aus gleichaltrig mit den Bremer Papyri sein’ (90; emphasis added). Wilcken's point is not disagreement with Roberts, but rather agreement! He affirmed Roberts’ assigned date in the first half of the second century and offered further (unspecified) comparanda that fell within that range. On the archive of Apollonios strategos, of which these Bremen papyri are a part, see now the entry in the Trismegistos database, www.trismegistos.org/arch/detail.php?tm=19.

51 Crisci, E., ‘Riflessioni paleografiche (e non solo) sui più antichi manoscritti greci del Nuovo Testamento’, Oltre la scrittura. Variazioni sul tema per Guglielmo Cavallo (ed. D. Bianconi and L. Del Corso; Dossiers Byzantins 8; Paris: Centre d’études byzantines, néo-helléniques et sud-est européennes, 2008) 53–93Google Scholar, esp. 68–77.

52 Interestingly, P.Flor. i.1 was the latest dated palaeographic parallel cited by Roberts in his original edition of the fragment. Yet, he was not at all confident in its overall similarity: ‘In this text the upsilon, the omega and sometimes the alpha are similar to those in our text, but other letters are radically different and its general style is not very close to that of P. Ryl. Gk. 457’ (An Unpublished Fragment, 16).

53 Crisci claimed that P52 was ‘riferibile al primo quarto del ii secolo’ and that its writing ‘mostra una chiara impronta burocratico-amministrativa … Ove si vogliano individuare, in ambito documentario, termini di confronto per questa particolare tipologia grafica, il primo reperto a dover esser preso in considerazione è il PSI v 446, editto del prefetto Marco Petronio Mamertino del 133–136, esempio di scrittura cancelleresca rotonda e occhiellata con la quale il P. Ryl. 457 ha molte affinità di fondo … Accanto a questo, si può segnalare pure P. Flor. i 1, mutuo ipotecario del 153 d.C., documento di ambiente notarile la cui scrittura principale, pur se declinata sul versante di una maggiore propensione alla corsività delle forme, esibisce le stesse caratteristiche di fondo’ (‘Riflessioni paleografiche’, 68–9).

54 Orsini, P. and Clarysse, W. ‘Early New Testament Manuscripts and their Dates: A Critique of Theological Palaeography’, ETL 88 (2012) 443–74Google Scholar. They describe the hand of P52 both as a ‘round majuscule’ (462) and as a form of writing that ‘originated in bureaucratic and chancery practices … used in the main central and peripheral offices in the second and third centuries’ (458). They describe this writing as ‘round, unimodular and looped, and the strokes end in apices (in the lower parts) and small hooks (in the upper parts); sometimes curves and flourishes are added at the end of letters’ (458).

55 Orsini and Clarysse, ‘Early New Testament Manuscripts’, 466 and 470.

56 See e.g. Cavallo, G., La scrittura greca e latina dei papyri. Una introduzione (Rome: Fabrizio Serra, 2008) 89–91Google Scholar, where the latest dated example of this class of writing provided is P.Oxy. l.3593, a record for the sale of a slave from 238–44 ce.

57 For a fuller palaeographic description of P.Fay. 87, see Turner, Greek Manuscripts of the Ancient World, item no. 48.

58 Grenfell, B. P., Hunt, A. S. and Hogarth, D. G., Fayûm Towns and their Papyri (London: Egypt Exploration Fund, 1900) 220–1Google Scholar.

59 Much of Turner's library and a small collection of his personal papers were purchased by the University of Western Australia in Perth in 1985. Eight boxes of Turner's papers are kept in Special Collections, while his books have been integrated into the library's general collection.

60 The notes all appear to have been written at the same time, and they mention P.Oxy. xlii.3016, which was published in 1974.

61 In particular, the second hand visible in column 16, which was republished as Rom.Mil.Rec. 76.46; see Fink, R. O., Roman Military Records on Papyrus (Cleveland: American Philological Association, 1971) 297Google Scholar.

62 Turner, The Typology of the Early Codex, 100.

63 It is not clear how long after 200 CE P.Oxy. 51.3614 was copied. The editor of this piece wrote that the handwriting ‘looks as if it belongs still to the third century’.

64 And it is even less advisable to describe it as having been copied ‘circa’ any given year.

65 Don Barker (‘The Dating of New Testament Papyri’, 574) has suggested that the ‘graphic stream’ of P52 extends into the late third century, citing P.Oxy. xliv.3183, a registration of children dated 26 July 292, a duplicate document written in two different hands (presumably Barker refers to Hand ‘A’).

66 Nongbri, ‘The Limits of Palaeographic Dating of Literary Papyri’.

67 Furthermore, the best course would also be to assume, with Bagnall, that a later date within that widened spectrum of possible palaeographic dates is more probable than an earlier one, simply given the relative lack of evidence for Christians in the Egyptian chora prior to the third century. But it is important to note that preference for this somewhat later date is not an issue of palaeography.

68 During the conference at which this paper was originally presented, a question arose about the possibility of using radiocarbon analysis on P52. The current technology of accelerator mass spectrometry requires sample sizes for papyrus of as little as 20 milligrams (personal correspondence with staff at the Oxford Radiocarbon Accelerator Unit), which is less than 1 cm² in terms of surface area. A small uninscribed portion of the upper margin of P52 could be removed for this purpose with little cosmetic damage to the fragment. While such testing would still result in a fairly wide range of probable dates, it would be beneficial to have this added data in order to find out with which part of the spectrum of palaeographic dates the radiocarbon dates might overlap.

69 I give the text of the label as I saw it in September 2014. The label was in need of correction on other grounds as well. It described the contents of the fragment as containing part of John 17, when in fact it contains verses of John 18.

70 The website of the Rylands Library did a somewhat better job, by virtually quoting Roberts’ 1935 assessment: ‘The importance of this fragment is quite out of proportion to its size, since it may with some confidence be dated in the first half of the second century ad, and thus ranks as the earliest known fragment of the New Testament in any language.’ This was the text on the website in August 2014. Now (in 2019) the website description reads as follows: ‘The Fragment is widely regarded as the earliest portion of any New Testament writing ever found. It provides us with invaluable evidence on the spread of Christianity in the provinces of the Roman Empire in the first centuries of our era. The first editor dated the Fragment to the first half of the second century (between 100–150 ad). The date was estimated palaeographically, by comparing the handwriting with other manuscripts. However, palaeography is not an exact science – none of the comparable Biblical manuscripts are dated and most papyri bearing a secure date are administrative documents. Recent research points to a date nearer to 200 ad, but there is as yet no convincing evidence that any earlier fragments from the New Testament survive. Carbon-dating is a destructive method and has not been used on the Fragment.’

71 The issue of the fragment's lack of clear provenance would seem especially pressing. Even competent biblical scholars are prone to speak of P52 and other artefacts bought on the antiquities market as though they were uncovered in secure, datable archaeological contexts. See the various discussions cited in Nongbri, B., God's Library: The Archaeology of the Earliest Christian Manuscripts (New Haven: Yale University Press, 2018)Google Scholar.