1 For a survey of recent contributions in this area, see the discussion and bibliography in K. Haines-Eitzen, ‘The Social History of Early Christian Scribes’, The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis (ed. B. D. Ehrman and M. W. Holmes; Leiden: Brill, 2013²) 479–95.

2 The label nomina sacra derives from Ludwig Traube, whose 1907 study Nomina Sacra: Versuch einer Geschichte der christlichen Kürzung (Munich: Beck; repr. Darmstadt: Wissenschaftliche Buchgesellschaft, 1967) brought this feature to the attention of the scholarly community. Traube's work was updated and supplemented by A. H. R. E. Paap in his 1959 work Nomina Sacra in the Greek Papyri of the First Five Centuries (Leiden: Brill). Other important studies of the topic include: José O'Callaghan, Nomina sacra in papyris Graecis saeculi iii neotestamentariis (Rome: Biblical Institute, 1970); C. H. Roberts, Manuscript, Society and Belief in Early Christian Egypt (London: Oxford University, 1979) 26–48; P. W. Comfort, Encountering the Manuscripts: An Introduction to New Testament Paleography & Textual Criticism (Nashville: Broadman & Holman, 2005) 199–254; L. W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006) 95–134; A. Luijendijk, Greetings in the Lord: Early Christians and the Oxyrhynchus Papyri (Harvard Theological Studies 60; Cambridge, MA: Harvard University Press, 2008) 57–78. For additional studies on the nomina sacra, consult the references in Hurtado, L. W., ‘The Origin of the Nomina Sacra: A Proposal’, JBL 117 (1998) 656 n. 6Google Scholar; Luijendijk, Greetings in the Lord, 58 n. 5; and Haines-Eitzen, ‘Social History’, 490–1.

3 Hurtado, ‘Origin’, 655–60. Hurtado clarifies: ‘This distinguishes them from the kinds of abbreviations in non-Christian Greek manuscripts, ostraca, and inscriptions, which are usually abbreviated by “suspension”, the first letter or two written and the rest omitted, with varying marks to indicate an abbreviated word’ (‘Origin’, 658).

4 Hurtado, ‘Origin’, 655. It is helpful to remember, however, that there is a wide range of texts that employ nomina sacra (Hurtado, ‘Origin’, 657–8). In her examination of documentary papyri from Oxyrhynchus, Luijendijk astutely observes that ‘nomina sacra appear in all sorts of Christian manuscripts and cross the rhetorical territories of “orthodox” and “heretical” writings’, for example in the Gospel of Thomas, the Gospel of Luke and some magical texts (Greetings, 60).

5 Nomina sacra for πνεῦμα (plus derivatives) are visible, for instance, in the following early manuscripts: P1 (Matt 1.18, 20), P4 (Luke 1.57, 80; 3.16, 22; 4.1), P5 (John 1.33; 20.22), P15 (1 Cor 7.40), P20 (Jas 2.26), P24 (Rev 5.6), P27 (Rom 8.14), P37 (Matt 26.41), P38 (Acts 19.1–2, 15), P45 (Mark 7.25; 9.20, 25; Luke 9.39; 10.20–1; 11.13; 13.11; Acts 4.31; 5.32; 6.10; 7.55; 8.15–19, 39; 9.17; 11.12; 13.9; 16.16, 18; 17.16), P47 (Rev 11.11; 13.15; 14.13; 16.13–14), P49 (Eph 4.23), P50 (Acts 8.29), P66 (John 1.32; 3.5, 8 (where a scribe changed a plene form to a nomen sacrum); 3.34; 4.23–4; 6.63; 7.39; 11.33; 13.21; 14.17, 26; 19.30), P72 (1 Pet 1.2, 11–12; 2.5 (which preserves unusual forms for the adjective); 3.4, 18–19; 4.6; 2 Pet 1.21; Jude 1.19), P75 (Luke 3.22; 4.36; 6.18 (the previous two examples have unusual forms of the nomen sacrum for ‘unclean spirits’); 8.29 (again for an ‘unclean spirit’); 8.55; 9.39, 42 (for an ‘unclean spirit’); 10.21; 11.13, 24, 26; 12.10, 12; 13.11; 23.46; 24.37, 39; John 1.32–3; 3.5–8, 34; 4.23–4; 6.63; 7.39; 14.17), P101 (Matt 3.11, 16; 4.1), P106 (John 1.32–3), P113 (Rom 2.29), P115 (Rev 11.11), P123 (1 Cor 14.32); see also Uncial 0189 (Acts 5.3). For those, like Hurtado, interested in tracing the origins of the nomina sacra and in demonstrating early standardisation of special forms that reflect proto-orthodox piety, the abbreviations for πνεῦμα do not play a large factor because they seem to have emerged slightly later and display less consistency than the four earliest forms. In his book Encountering the Manuscripts, Comfort does devote significant attention to the nomina sacra for πνεῦμα in early papyri, but both his analysis and conclusions are misguided as a result of two faulty assumptions: (1) that there is basically a fixed pattern to the usage of nomina sacra for forms of πνεῦμα, and (2) that this enables us to infer from their presence or absence whether or not a scribe interpreted πνεῦμα as a reference to the divine Spirit. While Comfort and others have highlighted (as anomalous) the few places where πνεῦμα is written in full and yet still refers to the divine Spirit, what most have failed to appreciate fully is the converse. One exception is Paap, who observes this phenomenon but whose study does not focus on interpreting its significance (see his tables in Nomina Sacra, 8–9, 82–3 and his limited discussion on 102–3). Another, more recent, exception is Edgar Battad Ebojo, who correctly notes the inconsistency in the manuscript tradition concerning the contraction or non-contraction of πνεῦμα relative to its referent (‘A Scribe and his Manuscript: An Investigation into the Scribal Habits of Papyrus 46’ (PhD diss., University of Birmingham, 2014) 352–4. eTheses (4838), accessed 30 October 2014, http://etheses.bham.ac.uk/).

6 In contrast to earlier studies of nomina sacra (e.g. those of Traube and Paap), which have mainly surveyed a broad range of papyri (some very fragmentary), there are several methodological advantages to focusing on one extensive manuscript, including the following: (1) tracing patterns across fragmentary papyri may produce distorted results, while attending carefully to patterns in a single extensive witness paints a more accurate portrait of developing trends; (2) since so few early witnesses are extant, attempting to relate them to each other can be ‘like trying to set up a molecule with the spheres but not the rods to connect them’ (D. C. Parker, Codex Bezae: An Early Christian Manuscript and its Text (Cambridge: Cambridge University Press, 1992) 2); (3) examining nomina sacra within the context of a single manuscript allows for a more nuanced investigation into their ‘sacral’ or ‘non-sacral’ status relative to their referents. For more on the benefits of this methodology, see Parker, Codex Bezae, 1–4; Ebojo, ‘A Scribe and his Manuscript’, 324–7.

7 Contra Comfort, P. W. (‘Light from the New Testament Papyri concerning the Translation of Pneuma’, Bible Translator 35.1 (1984) 130–3)CrossRefGoogle Scholar, who goes so far as to suggest that the presence or absence of nomina sacra in early manuscripts can aid modern Bible translators in their decisions about whether or not to capitalise the word ‘S/spirit’.

8 The date of the papyrus has been the subject of some debate, though a date around 200 ce claims wide support. For a lucid discussion of the various proposals, see J. R. Royse, Scribal Habits in Early Greek New Testament Papyri (NTTSD 36; Leiden: Brill, 2008) 199–201.

9 Comfort claims that P46 is one of only four early Christian manuscripts that write πνεῦμα both as a nomen sacrum and plene, the others being P45, P66, P75, which all date to the late second or early third century (Encountering, 234). However, it should be noted that many of the earliest manuscripts do not preserve enough text to allow us to discern whether both forms of the term were originally present. Moreover, as Paap's chart of all the occurrences of πνεῦμα in the first five centuries amply demonstrates (Nomina Sacra, 82–3), the presence of both the nomen sacrum and plene forms of πνεῦμα, while uncommon, is not quite as exceptional as Comfort asserts.

10 Considering its age, P46 is remarkably well preserved. Of the original 104 folios, portions of 86 folios survive, usually with significant loss only to the bottom few lines. Originally, the whole manuscript formed a single quire, with the verso side of the leaf preceding the recto in the first half and the recto preceding the verso in the second half (H. A. Sanders, A Third-Century Papyrus Codex of the Epistles of Paul (Ann Arbor: University of Michigan, 1935) 2). The codex is now divided into two collections. Fifty-six folios are housed in the Chester Beatty library in Dublin, Ireland (Chester Beatty Papyrus ii) and thirty at the University of Michigan (Inv. 6238). Both of these collections have been digitised and are now available online as high-resolution images. The thirty Michigan leaves are available through the Advanced Papyrological Information System (APIS) at the University of Michigan (http://www.lib.umich.edu/papyrology-collection/advanced-papyrological-information-system-apis, accessed 4 November 2014), and the fifty-six Chester Beatty leaves are now available at The Center for the Study of New Testament Manuscripts (CSNTM) (http://www.csntm.org/Manuscript/View/GA_P46, accessed 4 November 2014). A facsimile edition of the entire codex was published by Frederic G. Kenyon in 1937, and the photographic plates in this volume, in comparison with the digital images now available online, provide the basis for my study (The Chester Beatty Biblical Papyri, fasc. 3, supp. 3.2: Pauline Epistles, Plates (London: Emery Walker, 1937). In several places, Kenyon's plates actually preserve text that is now missing from the physical manuscripts that remain due to deterioration (compare e.g. the outer edge of f. 84. r. in Kenyon's plates with the digital image on APIS: http://quod.lib.umich.edu/cgi/i/image/image-idx?rgn1=apis_inv;q1=6238;size=20;c=apis;subview=detail;resnum=57;view=entry;lastview=thumbnail;cc=apis;entryid=x-3617;viewid=6238_164.TIF/, accessed 4 November 2014; see also n. 3 in the description of the Chester Beatty images from CSNTM: http://images.csntm.org/Manuscripts/GA_P46/P46%20(CBL%20BP%20II).pdf, accessed 4 November 2014). To facilitate comparison with Kenyon's plates, I have opted to employ his numbering system throughout this study (e.g. f. 8. v. = folio 8, verso).

11 Since seven leaves (the final fourteen pages) are missing at the end of the codex, it is likely that the codex originally contained 2 Thessalonians. If so, the remaining pages would not have contained enough space for the Pastorals. It is possible that several pages were simply left blank, or that some extra pages were added to the codex to accommodate these letters, but these hypotheses remain speculative. See the discussion in Royse, Scribal Habits, 202–3; D. C. Parker, An Introduction to the New Testament Manuscripts and their Texts (Cambridge: Cambridge University Press, 2008) 252–4.

12 See F. G. Kenyon, The Chester Beatty Biblical Papyri, vol. iii: Supplement: Pauline Epistles, Text (London: Emery Walker, 1936) xiii; Sanders, Third-Century Papyrus, 12; G. Zuntz, The Text of the Epistles: A Disquisition upon the Corpus Paulinum (London: Oxford University, 1953) 18. Royse counts a total of 183 corrections in the papyrus, the majority of which he says appear to be by the scribe himself (Scribal Habits, 211). Of the 183 total corrections in P46, he ascribes 38 corrections definitely to ‘man 1’ (i.e. the scribe) and another 71 to what he believes to be ‘possibly man 1’ (see his chart in Scribal Habits, 223). Following Zuntz (Text, 252–62), Royse also observes that corrections were made by at least three other later hands (Scribal Habits, 213–14).

13 This need not imply, however, that the scribe worked within the context of a scriptorium, as Zuntz suggested on the basis of the correcting activity (Text, 18, 273). See K. Haines-Eitzen, Guardians of Letters: Literacy, Power, and the Transmitters of Early Christian Literature (Oxford: Oxford University Press, 2000) 83–91.

14 Rom = 1000; 2 Cor = 1000; Eph = 316; Phil = 225 (or) 222 (depending on whether the text reads or ⲥⲕⲉ or ⲥⲕⲃ; the former seems more likely); cf. Sanders, Third-Century Papyrus, 21–2. For Philippians, P. W. Comfort and D. P. Barrett, eds., The Text of the Earliest Christian Manuscripts (Grand Rapids: Baker, 2001²) 327 has ⲥⲕⲉ (= 225 lines), as does Kenyon, Pauline Epistles, Text, 9.

15 Royse, Scribal Habits, 358.

16 Royse, Scribal Habits, 225.

17 Royse, Scribal Habits, 263.

18 These spatial divisions were also observed by Kenyon, leading him to say, ‘They suggest at any rate some perception by the scribe of the sense of what he was writing’ (Pauline Epistles, Text, xiv). For more on the meaning and function of space divisions in P46, see Ebojo, E. B., ‘When Nonsense Makes Sense: Scribal Habits in the Space-Intervals, Sense-Pauses, and Other Visual Features of P46’, The Bible Translator 64.2 (2013) 128–50CrossRefGoogle Scholar.

19 Royse, Scribal Habits, 259.

20 See Ebojo, ‘A Scribe and his Manuscript’, 329–30, 355–6.

21 Both Royse and Haines-Eitzen assume that errors involving nomina sacra in P46 are the result of the scribe's inability to reproduce accurately his Vorlage (Royse, Scribal Habits, 259, cited approvingly in Haines-Eitzen, Guardians, 93). However, it seems more likely to me that πνεῦμα and its derivatives were mostly, if not always, written plene in the scribe's exemplar and that the variation of forms for this term indicates the scribe's attempt to alter these forms to appropriate nomina sacra. As Sanders notes, the irregularities of the forms of the nomina sacra, and also the absence of certain forms, indicate an early date for the manuscript (Third-Century Papyrus, 16).

22 Comfort, Encountering, 237.

23 See also the discussion of scribes as readers in B. D. Ehrman, The Orthodox Corruption of Scripture: The Effect of Early Christological Controversies on the Text of the New Testament (New York: Oxford University Press, 2011²) 336–7.

24 Guardians, 16.

25 These numbers (and the remaining statistics in this study) reflect only those forms that are visibly discernible from the manuscript evidence that remains. In this way, my discussion of these terms is grounded firmly on the evidence that is indisputably apparent in the manuscript itself, rather than in reconstructions of it, however reliable they may be. All of the statistics in this section are tabulated based on my own evaluation of the facsimile edition of P46 (Kenyon, Chester Beatty, 1937), along with the high-resolution digital images available from the Advanced Papyrological Information System at the University of Michigan and The Center for the Study of New Testament Manuscripts (see n. 10 above). In an appendix, I have compiled a list of verse references where all of the terms discussed in this study appear in P46, whether as nomina sacra or plene. Compared with modern critical editions of the Greek New Testament (e.g. NA²⁸), there are only two differences concerning the appearance of ‘spirit’ language in P46: (1) in Eph 5.19 the critical text reads ᾠδαῖς πνευματικαῖς (‘spiritual songs’) whereas P46 lacks the adjective πνευματικαῖς; (2) in 1 Cor 15.47, where the critical text reads ὁ δεύτερος ἄνθρωπος ἐξ οὐρανοῦ (‘the second man [is] from heaven’), P46 has ⲟ ⲇⲉⲩⲧⲉⲣⲟⲥ ⲁⲛⲑⲣⲱⲡⲟⲥ ⲡⲛⲕⲟⲥ ⲉⲝ ⲟⲩⲣⲁⲛⲟⲩ (‘the second spiritual man [is] from heaven’).

26 As noted previously, this is the contention of Comfort (Encountering, 234).

27 Due to space constraints, I have restricted my survey of nomina sacra to this core set of names/titles. However, in addition to the terms discussed here, four other words appear as nomina sacra in P46: σταυρός, υἱός, πατήρ and ἄνθρωπος. According to Ebojo, σταυρός appears 19 times and is always contracted, except in Rom 6.6; υἱός, πατήρ and ἄνθρωπος, however, are much more irregularly contracted, indicating that the status of these terms as nomina sacra are, like πνεῦμα, still in the process of refinement (‘A Scribe and his Manuscript’, 323–66).

28 These totals differ somewhat from those in Paap (Nomina Sacra, 8–9, 82–3), who lists only 311 occurrences of θεός. He does, however, rightly indicate only 3 appearances of the term plene. Since Paap only lists statistics about frequency and not actual verse references, it is impossible to verify his figures without independently checking the manuscript itself, as I have done (see n. 10 above). Ebojo's thorough study of P46 also tallies all occurrences of nomina sacra and their plene counterparts (‘A Scribe and his Manuscript’, 323–66). My results generally agree with his analysis, with a few slight statistical differences. While Ebojo indicates that he has listed the locations for all the nomina sacra in his Appendix P (see ‘A Scribe and his Manuscript’, 324 n. 2), this appendix has not yet been made available (the version of his dissertation accessible online does not include his appendices). Once Ebojo's complete study becomes available, it will be possible to compare our analyses and to confirm, or perhaps revise, the tabulations offered here.

29 In 1 Cor 8.5 the plene is used in the plural to refer to many false ‘gods’ over against the one true God, written as a nomen sacrum in the subsequent verse. That the scribe is using nomina sacra intentionally in order to differentiate meaning is confirmed by the contrast between many false ‘lords’ (plene) and the one true Lord, Jesus Christ (all nomina sacra) in the same two verses. Thus, the nomina sacra are used consistently here to bolster the rhetorical argument of the passage and accentuate its central contrast; they are absent in v. 5 where the referents are ‘profane’, but punctuate every line of v. 6 where the referents are ‘sacred’. In this passage, then, it seems obvious that the scribe has a good sense of the difference between the forms and uses them consistently to clarify the meaning of the text. The same could be said of Gal 4.8, where, once again, the presence or absence of a nomen sacrum signals an important contrast between the true God (nomen sacrum), and ‘beings that by nature are not “gods” at all’ (plene).

30 One other possible exception is 2 Cor 4.4, but damage to the manuscript makes it hard to say for sure. In P46, 2 Cor 4.4 is split across f. 64. v. and f. 64. r. and several lines are missing. This verse speaks of the ‘god of this age [nomen sacrum?]’ who has ‘blinded the minds of those who do not believe so that they would not see the light of the glorious gospel of Christ [nomen sacrum?], who is the image of God [nomen sacrum?]’. The difficulty with this verse is that it is almost indecipherable in P46 due to damage to the bottom few lines of folio 64. On the last visible line of the page, only the very tops of some nine letters remain from the end of v. 3 ([ⲗⲩⲙⲉⲛⲟⲓⲥ] ⲉ̣ⲥ̣ⲧ̣ⲓ̣ⲛ ⲕⲉⲕⲁ̣[ⲗⲩⲙⲙⲉⲛⲟⲛ]), and the rest of the transcription must be worked out based upon a deduction about how many letters would fit between the remaining text on the recto of folio 64 and where the text picks up again at the top of the opposite side (in the middle of v. 4: ⲉⲓⲥ ⲧⲟ ⲙⲏ ⲁⲩⲅⲁⲥⲁⲓ ⲧⲟⲛ ⲫⲱⲧⲓⲥⲙⲟⲛ …). While on the basis of spacing it is reasonable to speculate that all of the occurrences of θεός in the missing text were written as nomina sacra, since the nomen sacrum and plene forms of this word only involve a difference of two letters, and since the length of lines in P46 can vary considerably even on the same page (cf. Sanders, Epistles of Paul, 6), it is impossible to know for certain.

31 The scribal distinction between ‘sacral’ and ‘non-sacral’ referents, observed above for 1 Cor 8.5 and Gal 4.8 (see n. 27), is also evident in Eph 6.8–9, where the nomen sacrum and plene forms are used to contrast human ‘masters’ (ⲟⲓ ⲕⲩⲣⲓⲟⲓ) with the ‘Master who is in heaven’ (ⲟ ⲕⲥ ⲉⲥⲧⲓⲛ ⲉⲛ ⲟⲩⲣⲁⲛⲟⲓⲥ, f. 80 v.).

32 For Heb 4.8, Ebojo (‘A Scribe and his Manuscript’, 339–40) points out a blog entry by Peter Head, which suggests that the use of the nomen sacrum form here might indicate that P46 is ‘interpreting this verse in terms of “Jesus” rather than “Joshua”’ (http://evangelicaltextualcriticism.blogspot.com/2010/06/if-jesus-had-given-them-rest-heb-48-in.html/, accessed 3 November 2014). Ebojo also notes a similar suggestion in Comfort, Encountering, 222 (see ‘A Scribe and his Manuscript’, 340). While such an explanation is possible, it seems more likely, given the complete absence of the plene form in P46, that the scribe simply contracted every occurrence of Ἰησοῦς as a matter of course, without careful consideration for its referent (Ebojo, ‘A Scribe and his Manuscript’, 340; cf. Roberts, Manuscript, Society and Belief, 37).

33 Cf. BDAG 1091.

34 Cf. Royse, Scribal Habits, 248; Haines-Eitzen, Guardians, 93.

35 It is not always easy to discern when the original author/text intends to signify the divine Spirit versus some other kind of spirit. For this reason, in the figures below I differentiate between ‘clear’ and ‘possible’ instances of exceptions. For many of the examples listed in the ‘possible’ category a strong case could be made that they, too, clearly break the ‘pattern’, but I have taken a conservative approach to include only the most convincing examples in the ‘clear’ category. In making these judgements, immediate context is my primary guide, including such criteria as the presence of direct contrasts (e.g., flesh/spirit; letter/spirit) and descriptive modifiers (e.g. holy, eternal, Spirit of God/grace, etc.). In each case, I have also checked my decisions (for heuristic purposes, recognising, of course, the differences between P46 and the eclectic critical text of the NT) against several standard commentaries (e.g. ICC, AB, WBC) and modern translations (e.g. NRSV, NIV), as well as G. D. Fee, God's Empowering Presence: The Holy Spirit in the Letters of Paul (Peabody, MA: Hendrickson, 1994), and have included in the ‘clear’ category only those instances in which the referent of πνεῦμα is generally beyond dispute. For ‘possible instances’ I have listed a sampling of English translations that either capitalise or do not capitalise ‘S/spirit’ in order to indicate its ‘sacrality’. As a result, my judgements are more cautious than those of Ebojo, who offers statistics about the ‘sacrality’ or ‘non-sacrality’ of the nomina sacra for πνεῦμα with more confidence than I feel the evidence permits (see e.g. the discussion and Table 4-C4 in ‘A Scribe and his Manuscript’, 361).

36 In the case of 2 Cor 13.13, most other ancient manuscripts include the adjective ‘holy’ before ‘Spirit’ (τοῦ ἁγίου πνεύματος), thus removing any doubt about the referent (see the critical apparatus of NA²⁸). Without the adjective, P46 is slightly more ambiguous and, admittedly, it is conceivable to interpret the fellowship τοῦ πνεύματος as the connection between human spirits rather than the fellowship of the Holy Spirit (similar to how Paul can describe being with people in spirit but not in body). Nevertheless, it seems much more plausible to understand the referent of πνεῦμα in this passage as the divine Spirit, in line with the majority of ancient witnesses.

37 See Hurtado, Artifacts, 120–33.

38 See Royse, Scribal Habits, 266 (and his citations in n. 366); also see his helpful summary of the copying activity of the scribe of P46 on pp. 357–8.

39 Rom 9.1; 14.17; Heb 2.4; 3.7; 6.4; 9.8; 10.15; 1 Cor 6.19; 2 Cor 6.6.

40 Rom 15.19; 1 Cor 2.11; 3.16; 6.11; 12.3; Phil 3.3.

41 Royse, Scribal Habits, 263–4.

42 Royse, Scribal Habits, 264.

43 Unlike, for example, Codex Vaticanus (Matt 4.1), where πνεῦμα is contracted at the end of the line to conserve space; in this case, Paap says that ‘we may infer that the text he copied had the word in full and, as a rule, he followed his model, using the contracted form as an abbreviation, i.e. where he was short of space’ (Nomina Sacra, 121). Space considerations also occasionally play a role in the use of nomina sacra in Codex Sinaiticus (see D. Jongkind, Scribal Habits of Codex Sinaiticus (Texts and Studies, Third Series 5; Piscataway, NJ: Gorgias, 2007) 83).

44 Comfort, Encountering, 234.

45 However, see Fee's discussion of the possible ‘sacral’ meaning of the adjective in such passages (Presence, 28–32, 107–12, 267–9; cf. my notes in the Appendix).

46 Moreover, the fact that this singular reading was not corrected by a later hand indicates that it was palatable to subsequent users of the manuscript. Earlier in P46 a second hand felt compelled to correct a nomen sacrum for πνεῦμα that was out of place; in Heb. 9.14 the text that appears originally to have been copied, ⲧⲟ ⲡⲛⲁ ⲧⲟⲩ ⲭⲣⲩ (‘the spirit of Christ’), was corrected by a later hand to ⲧⲟ ⲁ̣ⲓ̣ⲙ̣ⲁ̣ ⲧⲟⲩ ⲭⲣⲩ (‘the blood of Christ’ – interestingly, the suprascript line was retained, creating an anomalous nomen sacrum for ‘blood’). However, through several waves of scribal corrections, the singular variant in 1 Cor 15.47 remained untouched, evidently because it did not present a problem for later readers.

47 Haines-Eitzen, Guardians, 93.

48 Haines-Eitzen, Guardians, 112.

49 It is widely acknowledged that during the first few centuries ce precise formulation of the person and work of the Holy Spirit developed slowly, lagging behind emerging Christology, which occupied the forefront of early Christian debates (Cf. V.-M. Kärkkäinen, The Holy Spirit: A Guide to Christian Theology (Louisville: Westminster John Knox, 2012) 10–11, 27).

50 A. I. C. Heron, The Holy Spirit (Philadelphia: Westminster, 1983) 64.

51 S. M. Burgess, The Spirit and the Church: Antiquity (Peabody, MA: Hendrickson, 1984) 19, 24.

52 B. D. Ehrman, trans., The Apostolic Fathers, vol. ii (LCL 25; Cambridge, MA: Harvard University, 2003) 336–7. Note that here and in the quotations that follow decisions about where to capitalise πνεῦμα reflect the translator's interpretation.

53 H. B. Swete, The Holy Spirit in the Ancient Church: A Study of Christian Teaching in the Age of the Fathers (London: MacMillan, 1912) 28.

54 B. D. Ehrman, trans., The Apostolic Fathers, vol. i (LCL 24; Cambridge, MA: Harvard University, 2003) 179, 187–8.

55 Ibid.

56 For helpful surveys of the developing reflections on the Spirit in the early church, see Swete, Holy Spirit in the Ancient Church; Heron, Holy Spirit, 63–86; G. Verbeke, L’évolution de la doctrine du pneuma du Stoïcisme à S. Augustin: étude philosophique (Paris/Louvain: L'Institut Supérior de Philosophie, Université de Louvain, 1945); Bolgiani, F., ‘La Théologie de l'Esprit Saint. De la fin du 1er siècle après Jésus Christ au concile de Constantinople (381)’, Quatres Fleuves 9 (1979) 33–72Google Scholar; J. P. Burns and G. M. Fagin, The Holy Spirit (Message of the Fathers of the Church 3; Wilmington, DE: Michael Glazier, 1984); Burgess, Spirit and the Church; J. C. Elowsky, ed., We Believe in the Holy Spirit (Ancient Christian Doctrine 4; Downers Grove, IL: IVP Academic, 2009); Kärkkäinen, Holy Spirit, 10–28; A. Thiselton, The Holy Spirit – in Biblical Teaching, through the Centuries, and Today (Grand Rapids: Eerdmans, 2013); Cf. C. W. Griggs, Early Egyptian Christianity: From its Origins to 451 ce (Leiden: Brill, 1990).

57 In her discussion of nomina sacra, Haines-Eitzen observes that there is ‘an apparent transition from an inconsistency and lack of uniformity in the second and third centuries to a consistent, standardized, and uniform use that emerges only in the fourth and later centuries’ (Guardians, 94).

58 Such caution is warranted by the fact that several other terms are also treated irregularly as nomina sacra in P46, including πατήρ, which is treated at least as idiosyncratically as πνεῦμα (see e.g. the fluctuating use of plene and nomen sacrum forms in Gal 1.1, 3; 2 Cor 1.2–3; Eph 5.31), without necessarily indicating any theological ambiguity about the ‘sacrality’ of God the Father (Ebojo, ‘A Scribe and his Manuscript’, 348–50).

59 To be more specific, in Codex Sinaiticus all but four occurrences of πνεῦμα (or 99%) appear as nomina sacra, and the four exceptions are all the product of a single scribe (Jongkind, Sinaiticus, 67, 80). Likewise, the adjectival form is always rendered as a nomen sacrum, and θεός, κύριος, Ἰησοῦς and Χριστός almost universally (98–9% of the time) appear as nomina sacra (Jongkind, Sinaiticus, 67). In Codex Bezae, there is less consistency, with only about 81% of occurrences contracted, and ‘no thorough attempt … to confine the use to places where the text refers to the sacred’ (Parker, Codex Bezae, 98–100, 106). By contrast, the NT portion of Codex Vaticanus contracts πνεῦμα in only about 2.9% of the occurrences (T. Bokedal, ‘Notes on the Nomina Sacra and Biblical Interpretation’, Beyond Biblical Theologies (WUNT 295; ed. H. Assel, S. Beyerle, C. Böttrich; Tübingen: Mohr Siebeck, 2012) 277). Bokedal's essay as a whole helpfully lays out developments in the system of nomina sacra across the first several centuries of their use. See also his broader treatment of the topic in The Formation and Significance of the Christian Biblical Canon: A Study in Text, Ritual and Interpretation (London: Bloomsbury T&T Clark, 2014) 83–123, and especially his suggestion about the relationship between nomina sacra and the development of the regula fidei on p. 120.

60 Guardians, 20.

61 Orthodox Corruption, 3.

62 Indeed, after analysing nomina sacra for πνεῦμα language across the first five centuries of early Christian manuscripts, Paap concludes: ‘One can hardly notice any distinction in meaning between the word contracted and fully written’ (Nomina Sacra, 102).

63 Cf. Luijendijk, Greetings in the Lord, 61.

64 Verse references in regular type indicate occurrences of nomina sacra; those in bold type indicate plene forms; verses with a double underline represent clear exceptions between the form and referent (i.e. nomina sacra used with a ‘non-sacral’ referent or plene forms used with a ‘sacral’ referent); verses with a single underline represent possible exceptions between the form and referent. References are listed in the order in which the text appears in P46. The total number of occurrences for each individual book (including both nomina sacra and plene forms) is placed in a rectangular box. All tabulations are based on my own independent observation of Kenyon's facsimile of P46 (Chester Beatty, 1937) and the digital images available through APIS and CSNTM (see n. 10), in consultation with Comfort and Barrett's transcription (Text, 202–334). Only the occurrences clearly visible in the extant manuscript (or at least visible enough to discern whether or not the forms in question are nomina sacra) are included in these lists. Footnotes clarify where only a small part of the word is visible (or only the suprascript line above the word, i.e. the overbar).

65 All three plene forms are plural (1 Cor 8.5: ⲑⲉⲟⲓ twice; Gal 4.8b: ⲑⲉⲟⲓⲥ).

66 Overbar is just visible at the bottom of f. 26. v.

67 Overbar is just visible at the bottom of f. 50. v.

68 Overbar is just visible at the bottom of f. 59. r.

69 Form is just visible at the bottom, outside of f. 70. r.

70 Overbar is just visible at the bottom of f. 88. r.

71 Form is just visible at the bottom of f. 88. r.

72 ⲟ ⲑⲥ ⲏ ⲕⲟⲓⲗⲓⲁ, f. 89. r.

73 All four plene forms are plural (1 Cor 8.5: ⲕⲩⲣⲓⲟⲓ; Eph. 6.5: ⲕⲩⲣⲓⲟⲓⲥ; Eph 6.9a: ⲕⲩⲣⲓⲟⲓ; Col. 3.22: ⲕⲩⲣⲓⲟⲓⲥ).

74 Overbar is just visible at the bottom of f. 13. r.

75 ⲧⲱ ⲓⲇⲓⲱ ⲕⲱ, f. 17. r.

76 The first of these occurrences is just visible at the bottom, outside f. 17. r.

77 Only the overbar is visible at the bottom of f. 17. r.

78 The overbar is just visible for two occurrences in 15.12 and 13, at the bottom, outside corner of f. 20. r.

79 Form is still visible despite damage to the manuscript at f. 29. v.

80 Form is just visible towards the bottom of f. 39. r.

81 Form is just visible at the bottom of f. 42. r.

82 Final letter and overbar are just visible at the bottom, outside edge of f. 91. v.

83 Final letter and overbar are just visible at the bottom of f. 13. r.

84 A nomen sacrum may also be present in Rom 15.5, where a final ⲛ̅ is visible at the edge of f. 18. r. Comfort and Barrett reconstruct the end of this line as ⲕⲁⲧⲁ ⲭⲣⲛ ⲓⲏ]ⲛ ⲓ̈ⲛⲁ (Text, 220). However, since a number of early manuscripts reverse the order of the relevant terms at this point (i.e. ⲓⲏⲛⲭⲣⲛ; see NA²⁸), and only the final ⲛ̅ is extant in P46, there is no way to be sure whether the existing letter represents the end of ⲓⲏⲛ or ⲭⲣⲛ. Thus, I have taken the cautious approach to leave Rom 15.5 off the list above.

85 See article, n. 30.

86 In Heb 10.10, a later scribe corrected the text from ⲓⲏⲩⲭⲣⲥ to ⲓⲏⲩⲭⲣⲩ (f. 31. r.); a similar correction takes place in Heb 12.21 (f. 38. v.), indicating either that the initial scribe inaccurately abbreviated the full form, or that this initial scribe was hesitant to alter the nomina sacra already existing in his exemplar, necessitating a correction by a later scribal hand.

87 Only the last two letters of the form are visible at the bottom corner of f. 64. v. (ⲓ]ⲏⲛ).

88 2 Cor 11.4 clearly uses a nomen sacrum with a non-sacral referent (ⲁⲗⲗⲟⲛ ⲓⲏⲛ, f. 71. v.).

89 Damaged, but still visible, on the outside corner of f. 77. v.

90 Form is just visible on the outside edge of f. 87. v. ([ⲕⲥⲓⲏ]ⲥ̣ⲭⲣⲥ). Although only the final sigma remains, we can be confident that it belongs to ⲓⲏⲥ and not ⲕⲥ for four reasons: (1) no extant early Christian manuscript attests anywhere the combination/word order of ⲓⲏⲥⲕⲥⲭⲣⲥ; (2) space considerations seem to demand the presence of four to five letter slots before ]ⲥ̣ⲭⲣⲥ and after the ⲟⲧⲓ on the previous line; (3) in all early Christian manuscripts, ⲕⲥⲭⲣⲥ nowhere appear directly together in this order without ⲓⲏⲥ between them, except in P46 Col 3.24 (f. 93. r.); (4) by contrast, the combination ⲕⲥⲓⲏⲥⲭⲣⲥ is well attested in P46, occurring besides this passage in Rom 10.9; 1 Cor 6.11; 8.6; 2 Cor 1.2; 13.13; Eph 1.2; 6.23. Thus, even though some manuscripts have only κύριος Ἰησοῦς at Phil 2.11, and one has Χριστὸς κύριος, the most plausible reading of P46 is ⲕⲥⲓⲏ]ⲥ̣ⲭⲣⲥ, as it is reconstructed in Comfort and Barrett, Text, and listed in the critical apparatus of NA²⁸.

91 Overbar is just visible at the bottom of f. 89. r.

92 Col 4.11 (ⲓⲏⲥ ⲟ ⲗⲉⲅⲟⲙⲉⲛⲟⲥ ⲓ̈ⲟⲩⲥⲧⲟⲥ, f. 93. v.) clearly uses a nomen sacrum with a non-sacral referent.

93 First letter and overbar are just visible on the inner edge of f. 8. r.

94 Barely visible on the outside edge of f. 14. r.

95 Overbar and portions of two letters (ⲭ]ⲣ̣ⲥ̣) are just visible at the bottom of f. 43. r.

96 Only the overbar is visible at the bottom of f. 57. v.

97 Overbar and portions of two letters are just visible at the bottom of f. 80. r.

98 Only the final letter and overbar are visible on the bottom, outside f. 82. v.

99 Form is damaged but still visible on the outside margin of f. 86. v.

100 Only the overbar and a portion of the first letter are visible for the second occurrence at the bottom of f. 88. v.

101 Only the overbar is visible on the bottom, outside f. 89. r.

102 Form is visible at the top of f. 93. v., despite some damage to the initial letter.

103 The adjectival forms of πνεῦμα (and sole occurrence of the adverb in 1 Cor 2.14) are indicated by [square brackets]. See the article (esp. n. 35) for a discussion of clear and possible ‘exceptions’ (indicated here by a double underline or single underline respectively).

104 The adjective ‘spiritual’ and the adverb ‘spiritually’ are, in many cases, arguably instances of πνεῦμα with a ‘sacral’ referent, carrying the sense, ‘that which belongs to, or pertains to, the Spirit’ or ‘by means of the Spirit’ (cf. Fee, Presence, 28–32). These words, in brackets above, are all underlined to indicate the possibility that a plene form could carry a ‘sacral’ meaning (even where this possibility is more doubtful, as in 1 Cor 10.3–4; 15.44; cf. Fee, Presence, 141–4, 263), or conversely that the nomen sacrum could be interpreted in sense that does not refer obviously to the Holy Spirit (as in e.g. 1 Cor 2.15; 3.1; 15.46, 47). The only adjective not underlined is Eph 6.12, where the plene form explicitly refers to things pertaining to evil, not the Holy Spirit (ⲧⲁ ⲡⲛⲉⲩⲙⲁⲧⲓⲕⲁ ⲧⲏⲥ ⲡⲟⲛⲏⲣⲓⲁⲥ, f. 80. v.).

105 Only the overbar is visible at the bottom of f. 55. v.

106 Singular reading: ⲟ ⲇⲉⲩⲧⲉⲣⲟⲥ ⲁⲛⲑⲣⲱⲡⲟⲥ ⲡⲛⲕⲟⲥ ⲉⲝ ⲟⲩⲣⲁⲛⲟⲩ, f. 59. r.

107 Variant reading: ⲟⲓ ⲉⲛ ⲡⲛⲉⲩⲙⲁⲧⲓ ⲗⲁⲧⲣⲉⲩⲟⲛⲧⲉⲥ, f. 88. v.