Reading the Gospel of Thomas in the Third Century: Three Oxyrhynchus Papyri and Origen’S Homilies

BIBLIOTHECA EPHEMERIDUM THEOLOGICARUM LOVANIENSIUM

CCXLII

READING NEW TESTAMENT PAPYRI IN CONTEXT LIRE LES PAPYRUS DU NOUVEAU TESTAMENT DANS LEUR CONTEXTE

EDITED BY

CLAIRE CLIVAZ – JEAN ZUMSTEIN

IN COLLABORATION WITH

JENNY READ-HEIMERDINGER AND JULIE PAIK

UITGEVERIJ PEETERS LEUVEN – PARIS – WALPOLE, MA 2011

994185_Clivaz_BETL242_vwk.indd4185_Clivaz_BETL242_vwk.indd V 44/07/11/07/11 11:3911:39 TABLE OF CONTENTS

PREFACE ...... VII LIST OF ABBREVIATIONS ...... xi INTRODUCTION (Claire CLIVAZ – Jean ZUMSTEIN) ...... 1

I. PAPYROLOGY AND THE NEW TESTAMENT

Claire CLIVAZ (Université de Lausanne) The New Testament at the Time of the Egyptian Papyri: Reflec- tions Based on P12, P75 and P126 (P.Amh. 3b, P. B o d . XIV-XV and PSI 1497) ...... 15

Thomas J. KRAUS (Independent Scholar) Christliche Papyri aus Ägypten – Kleine Facetten des großen Ganzen: Exemplarische Wechselbeziehungen...... 57

J. Keith ELLIOTT (University of Leeds) Recently Discovered New Testament Papyri and Their Signifi- cance for Textual Criticism ...... 89

II. EGYPT, PAPYRI AND CHRISTIANS

Paul SCHUBERT (Université de Genève) Les papyrus en Égypte aux trois premiers siècles de notre ère. 111

Sylvie HONIGMAN (University of Tel-Aviv) Les Juifs dans la société de l’Égypte romaine au croisement des sources documentaires et littéraires ...... 131

Daniel STOEKL (CNRS Aix-en-Provence) De l’arbre à la forêt: Quelques pensées quantitatives sur les papyrus littéraires juifs et chrétiens de l’Égypte ancienne ...... 169

Régis BURNET (EPHE, Paris) Des textes comme les autres: Réinscrire le Nouveau Testament dans les écrits du monde méditerranéen ...... 189

Kim HAINES-EITZEN (Cornell University) Imagining the Alexandrian Library and a “Bookish” Christianity 207

994185_Clivaz_BETL242_vwk.indd4185_Clivaz_BETL242_vwk.indd IXIX 44/07/11/07/11 11:3911:39 X TABLE OF CONTENTS

III. EVERY PAPYRUS TELLS A STORY

Jean ZUMSTEIN (Université de Zürich) Quand l’exégète rencontre le manuscrit: Le P66 ...... 221

AnneMarie LUIJENDIJK (University of Princeton) Reading the Gospel of Thomas in the Third Century: Three Oxyrhynchus Papyri and Origen’s Homilies ...... 241

Michael THEOPHILOS (Australian Catholic University, Melbourne) Multispectral Imaging of Greek Papyrus Fragments from Oxy- rhynchus ...... 269

David PASTORELLI (CNRS Aix-en-Provence) The Chester Beatty I Papyrus (P45) and the Main Greek Manu- scripts of Mark 6 and 9: A Classification Based on a New Quantitative Method ...... 281

Jenny READ-HEIMERDINGER (Bangor University, UK) – Josep RIUS- CAMPS (Facultat de Teologia de Catalunya) Tracing the Readings of Codex Bezae in the Papyri of Acts . . 307

Christian-Jürgen GRUBER (Universität Salzburg) Ein ungewöhnlicher Terminus? Thymiaterion in Heb 9,4 . . . . 339

IV. SOME FURTHER CONSIDERATIONS

Xavier GRAVEND-TIROLE (Université de Lausanne) From Sola Scriptura to Pluralibus Scripturis ...... 355

Thomas J. KRAUS (Independent Scholar) Lacunae in History: How and in What Respect Documentary Papyri Contribute to Remedy “Defective History”...... 383

INDICES

INDEX OF AUTHORS...... 405 BIBLICAL AND OTHER REFERENCES...... 417 PAPYRI AND MANUSCRIPTS ...... 430

994185_Clivaz_BETL242_vwk.indd4185_Clivaz_BETL242_vwk.indd X 44/07/11/07/11 11:3911:39 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY

THREE OXYRHYNCHUS PAPYRI AND ORIGEN’S HOMILIES*

I. INTRODUCTION

The publication of new texts found at Nag Hammadi and elsewhere over the past century has sparked an important re-thinking of early Chris- tianity, emphasizing its diversity. One tractate, the Gospel of Thomas, has especially generated much scholarship. Scholars nowadays even debate whether this gospel should become part of the Christian, or New Testament, canon, the authoritative body of Christian sacred scriptures. For instance, Bruce Metzger asked: “How far … does the Gospel of Thomas (which, of all the tractates in the Nag Hammadi library, seems to be closest to the New Testament) meet the criteria of apostolicity and orthodoxy…?”1. While Metzger decided against adding the Gospel of Thomas to the canon2, Robert Funk, among others, proposed a New Testament that indeed includes the fifth gospel3. The modern discussion about adding the Gospel of Thomas to the Christian canon evokes the question of how early Christians in antiquity perceived this gospel. Did they consider it sacred scripture? Did they hear it recited during a worship service or did they study it privately? In other words: Was the Gospel of Thomas read inside or outside of church?

1. B.M. METZGER, The Canon of the New Testament: Its Origin, Development and Significance, Oxford, Clarendon Press, 1987, pp. 271-272. W.R. FARMER, Reflections on Jesus and the New Testament Canon, in L.M. MCDONALD – J.A. SANDERS (eds.), The Canon Debate, Peabody, MA, Hendrickson, 2002, p. 321, disagrees: “I am convinced that the specific proposals to … add the Gospel of Thomas to the New Testament canon rep- resent a comprehensive misunderstanding of what ‘canon’ means and how canon functions among those who accept it”. Similarly, J.D.G. DUNN, Has the Canon a Continuing Func- tion?, in MCDONALD – SANDERS (eds.), Canon Debate, 558-579, p. 567, finds the Gospel of Thomas “less than adequate as expressions of [the] gospel”. 2. METZGER, Canon (n. 1), p. 272, presumes that “in this case the evaluation of modern readers will no doubt corroborate that of the early Church, namely, that in the Gospel of Thomas the voice of the Good Shepherd is heard only in a muffled way, and that it is, in fact, often distorted beyond recognition …”. He concludes his query about whether the canon is open or closed, stating that: “the canon cannot be remade – for the simple reason that history cannot be remade” (ibid., p. 275). 3. R.F. FUNK, The Once and Future New Testament, in MCDONALD – SANDERS (eds.), Canon Debate (n. 1), 541-557, p. 557.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 241241 44/07/11/07/11 11:5611:56 242 ANNEMARIE LUIJENDIJK

In this chapter, I conclude that in one Egyptian city in the third century, the Gospel of Thomas was read in a private setting and may also have been recited in Christian worship. Obviously, I do not derive this claim from the writings of ancient church leaders4. They answered the question whether the Gospel of Thomas counted as sacred scripture in the negative. As so often in the history of the early Christian movement, their assessment has been influential and definitive. Consequently, the question about the status of the Gospel of Thomas in antiquity has not yet been satisfactorily addressed, as I shall demonstrate. If we want to seriously reconsider the diversity of early Christianity and contribute to the rewriting of an early Christian history that does not immediately and linearly move toward proto-orthodoxy and orthodoxy, we should examine the evidence again with fresh eyes, not solely through the lens of the writings of church leaders. This piece contributes to such a reconsideration of early Christian history by considering a variety of ancient data to be authoritative: that is, instead of looking only at the opinions of early Christian church leaders that happen to survive, I address the antique usage of the Gospel of Thomas through a case study of three third-century papyrus fragments with its text in Greek found at the ancient Egyptian city of Oxyrhynchus (P.Oxy. I 1, IV 654 and 655). The three fragments provide evidence for the Greek version of the text and lower the date ante quem for the composition of that gospel. An aspect of these papyri that has received very little scholarly attention is that they also give indications about its textual transmission, circulation, and use5. In studying these papyri of the Gospel of Thomas, my aim is not to describe how we should understand the original composition of the Gospel of Thomas, although the Greek fragments offers striking insights into earlier

4. For a convenient overview of the (possible) quotations of and references to the Gospel of Thomas in patristic literature, so-called testimonia, see: H.W. ATTRIDGE, Appen- dix: The Greek Fragments, in B. LAYTON (ed.), Nag Hammadi Codex II, 2-7: together with XIII, 2*, Brit.Lib. Or. 4926(1), and P.Oxy. 1, 654, 655, Leiden, Brill, 1989, 96-128, pp. 103-109: “Testimonia to the Gospel of Thomas”. See also H.-Ch. PUECH, The Gospel of Thomas, in E. HENNECKE – W. SCHNEEMELCHER (eds.), New Testament Apocrypha. Vol. 1: Gospels and Related Writings, trans. R.MCL. WILSON, Philadelphia, PA, Westminster, 1963, 278-307, pp. 278-283, and B. BLATZ, Das koptische Thomasevangelium, in W. SCHNEE- MELCHER (ed.), Neutestamentliche Apokryphen in deutscher Übersetzung. Vol. 1: Evan- gelien, Tübingen, Mohr – Siebeck, 1990, 93-113, esp. pp. 93-94: “2. Bezeugung”. 5. With the exception of L.W. HURTADO, The Greek Fragments of the Gospel of Thomas as Artefacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654 and Papyrus Oxyrhynchus 655, in J. FREY – E.E POPKES – J. SCHRÖTER (eds.), Das Thomasevangelium: Entstehung – Rezeption – Theologie, Berlin – New York, de Gruyter, 2008, 19-32.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 242242 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 243

stages of meaning of the text6. Nor do I want to give an improved edition and better reconstruction of the text of these pieces. Rather, I pair and compare the material evidence of these pieces with other papyri and literary sources and so study the transmission and reception of that gospel in the third century, at one particular place – Oxyrhynchus – against the larger questions in our field of the diversity of early Christianity, the use of holy texts, and the development of the Christian, New Testament, canon. The fact that we have multiple early fragments from different books from the same well-known archaeological provenance, the ancient middle- Egyptian city Oxyrhynchus, makes the Gospel of Thomas a great subject for a case study. Not only do these papyri come from a known place – for numerous early Christian texts we lack a specific provenance – but Oxy- rhynchus constitutes a particularly important site: no other place has yielded such abundance of ancient fragments of early Christian documentary and literary texts. Oxyrhynchus therefore forms an excellent site to study early Christianity in a local context7. Scholarship on the formation of the Christian canon has in the past relied heavily on elite sources. The writings of influential ecclesiastical authorities such as Irenaeus, Origen, Eusebius, and Athanasius, important indeed, have been mined for clues on what they considered canonical texts8. The elite

6. On this topic, see D. LÜHRMANN, “Das Reich Gottes ist ausgebreitet auf der Erde”: Die griechische Überlieferung des Thomasevangeliums, in ID., Die apokryph gewordenen Evangelien: Studien zu neuen Texten und zu neuen Fragen (SupplNT, 112), Leiden, Brill, 2004, 144-181. In her commentary on the Gospel of Thomas, April DeConick uses the evidence from the three Oxyrhynchus papyri to establish the earliest possible text of that gospel, see A.D. DECONICK, The Original Gospel of Thomas in Translation: With a Commentary and New English Translation of the Complete Gospel (Early Christianity in Context. LNTS, 297), London, T&T Clark, 2006, esp. p. 9. 7. In A.M. LUIJENDIJK, Greetings in the Lord: Early Christians and the Oxyrhynchus Papyri (HTS, 60), Cambridge, MA, Harvard University Press, 2008, I investigated early Christians at Oxyrhynchus in the documentary papyri from the site; that is, private letters and official papers. With those results as the backdrop, I have now turned towards the literary papyri, the remains of early Christian books. 8. Similarly, C. MARKSCHIES, The Canon of the New Testament in Antiquity: Some New Horizons for Future Research, in M. FINKELBERG – G.G. STROUMSA (eds.), Homer, the Bible, and Beyond: Literary and Religious Canons in the Ancient World (Jerusalem Studies in Reli- gion and Culture, 2), Leiden, Brill, 2003, 175-194, p. 183, proposes: “for a more convincing picture, the horizon of research must be expanded substantially. For example, the traditional attention to great theologians and to the juridical system should be supplemented by the examination of which types of canons and books of contents of canons appeared in the public, liturgical and private lives of ancient Christians”. See also H.Y. GAMBLE, The New Testa- ment Canon: Recent Research and the Status Quaestionis, in MCDONALD – SANDERS (eds.), Canon Debate (n. 1), 267-294, p. 273: “The history of the canon has traditionally been written mainly on the basis of patristic evidence as to the documents known, used, and valued by different writers in different periods and places … It is more and more widely recognized that what we need is a more thorough understanding on how documents were used”.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 243243 44/07/11/07/11 11:5611:56 244 ANNEMARIE LUIJENDIJK

nature of the sources consulted for scriptural writings also pertains to material culture, as mainly spectacular and exceptionally expensive ancient manuscripts such as the Codex Sinaiticus and the Codex Vati- canus were taken into account9. I maintain that we can obtain novel insights about the use of ancient texts by taking an approach grounded in papyrological methods that include the writings of the less than elite10. I apply that approach in this chapter to the papyrus manuscripts of the Gospel of Thomas.

II. MANUSCRIPTS OF THE GOSPEL OF THOMAS

The Gospel of Thomas is known in its entirety from a Coptic translation preserved in a fourth-century papyrus codex found in 1945 near the modern Egyptian town of Nag Hammadi, ancient Chenoboskion11. Yet the manuscript evidence for the Gospel of Thomas includes also the three lesser-known third-century Greek papyri from Oxyrhynchus, P.Oxy. I 1, IV 654 and 65512. The papyri survived only in fragmentary state; the

9. On the elite character of Codex Sinaiticus and Codex Vaticanus, see D.C. PARKER, An Introduction to the New Testament Manuscripts and Their Texts, Cambridge – New York, Cambridge University Press, 2008, p. 71: “from the fourth-century copies, Codex Sinaiticus and its contemporary Codex Vaticanus, we have two parchment manuscripts that are superior to most other parchment manuscripts of any date, in the quality of the materials as well as in the skill of the execution…”. Regarding the evidence of manuscripts with the entire New Testament, Parker (ibid., p. 74), writes: “All these codices were of a de luxe quality and were conceived on a grand scale. … The use of teams of scribes … also reveals the size of the task. The evidence from these extensive surviving witnesses indicates that the single codex New Testament was exceptional in their own times, and unheard of in the succeeding period”. Thomas Pattie has calculated that the 730 folios that made up the codex Sinaiticus, “require[d] the perfect skins of 365 sheep or goats”. He adds: “The labour involved in the scraping, washing, stretching and polishing would have been substantial” (T.S. PATTIE, The Creation of the Great Codices, in J.L. SHARPE III – K. VAN KAMPEN [eds.], The Bible as Book: The Manuscript Tradition, London, British Library – New Castle, DE, Oak Knoll Press, in association with The Scriptorium, Center for Christian Antiquities, 1998, 61-72, p. 65). On labor costs for scribes, see R.S. BAGNALL, Early Chris- tian Books in Egypt, Princeton, NJ – Oxford, Princeton University Press, 2009, chapter 3 (pp. 50-69): “The Economics of Book Production”, esp. pp. 55-60. 10. For similar approaches, see MARKSCHIES, Canon of the New Testament (n. 8), and D. STÖKL BEN EZRA, Canonization – a Non-Linear Process? Observing the Process of Canonization through the Christian (and Jewish) Papyri from Egypt, in ZAC 12 (2008) 193-214. 11. Edition: LAYTON, Nag Hammadi Codex II, 2-7 (n. 4). 12. There is also a later funerary shroud with one saying of that same gospel that I will discuss in another paper. For its publication, see H.-Ch. PUECH, Un logion de Jésus sur bandelette funéraire, in Bulletin de la société Ernest Renan n.s. 3 (1954) 126-129, p. 127, and ID., Un logion de Jésus sur bandelette funéraire (1955), in ID., En quête de la Gnose. II: Sur l’Évangile selon Thomas: Esquisse d’une interprétation systématique, Paris, Galli- mard, 1978, 59-62.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 244244 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 245

Coptic text and parallels in other Greek writings have enabled scholars to produce reconstructions of the Greek text13. Of the 114 Coptic sayings that make the Gospel of Thomas in Nag Hammadi Codex II, the Oxy- rhynchite papyri of that gospel preserve fragmentary sections of 21 sayings in Greek (sayings 1–7; 26–33, 77a, 24, 36–39). Below, I give a brief papyrological description of these papyri, listing them in the order of the sayings (not in order of publication). As we shall see, one of them formed part of a codex, like most early Christian manuscripts, but the other two were part of book rolls, a less common format for early Christian texts. Do the book format and other aspects of the production of these manuscripts give us indications of their use by early Christians? The first piece, P.Oxy. IV 654, a narrow papyrus scrap of 7.8 by 24.4 cm, has writing on both sides in different hands: the recto, written along the fibers, preserves administrative scribbles; the verso, against the fibers, features the first seven sayings of the Gospel of Thomas in 42 frag- mentarily preserved lines14. In its first guise, this papyrus roll contained “a survey-list of various pieces of land”15. Grenfell and Hunt classified the handwriting as “a cursive hand of the end of the second or early part of the third century”16. In the second half of the third century, at least, based on palaeography, a Christian scribe copied the Gospel of Thomas on the back of that piece. Writing in “an upright informal uncial of medium size” with a thin pen, the scribe achieved a legible and elegant result17.

13. See especially ATTRIDGE, Appendix (n. 4), with excellent bibliography of previous scholarship, and J.A. FITZMYER, The Oxyrhynchus Logoi of Jesus and the Coptic Gospel According to Thomas, in ID., Essays on the Semitic Background of the New Testament, London, Chapman, 1971, 355-433. Also: A.D. DECONICK, Corrections to the Critical Reading of the Gospel of Thomas, in VC 60 (2006) 201-208, pp. 201-207. S. VAN DER LAAN’s website The Gospel of Thomas Resource Center (http://www.gospels.net/thomas/ last accessed August 3, 2010) has an extensive bibliography. 14. Ed. princ.: B.P. GRENFELL – A.S. HUNT, New Sayings of Jesus and Fragment of a Lost Gospel from Oxyrhynchus, New York, Oxford University Press, 1903, 1-36, and EID., New Sayings of Jesus, in EID., Oxyrhynchus Papyri IV (Egypt Exploration Society, GRM, 6), London, Egypt Exploration Fund, 1904, 1-22. 15. GRENFELL – HUNT, P.Oxy. IV 654 (n. 14), p. 1. 16. They stated: “The survey-list, which is in a cursive hand of the end of the second or early part of the third century, provides a terminus a quo for the writing on the other side” (ibid.). 17. Ibid. DECONICK, Original Gospel of Thomas (n. 6), p. 9, deemed “the scribal hands on the verso … quite beautiful and legible”. ATTRIDGE, Appendix (n. 4), p. 97, described the handwriting as “a common informal literary type of the third century. It consists of upright, slightly rounded capitals of medium size. It is similar to the hand of P.Oxy. I 1, although not so flattened. It is probably to be dated somewhat later than P.Oxy. 1, viz. in the middle of the third century”. The hand is very similar to that of P.Oxy. LXVI 4499 (P115) Revela- tion (late 3rd or early 4th cent). Other comparanda are: P.Bodmer XXIV Psalms (ed. princ. 3rd or early 4th cent.; others have proposed a 2nd cent. date, see A. PIETERSMA, The Edited

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 245245 44/07/11/07/11 11:5611:56 246 ANNEMARIE LUIJENDIJK

The scribe wrote the nomen sacrum IJS for ˆIjsoÕv, and used several forms of punctuation as reader’s aids18. The next papyrus occupies the prime position in the series of the Oxy- rhynchus Papyri: P.Oxy. I 119. Its rank in publication signals the great importance that excavators and editors of the papyri from Oxyrhynchus, Bernard Grenfell and Arthur Hunt, attributed to this text20. It is a single, incomplete and badly abraded page of 14.5 by 9.5 cm with writing on both sides in the same handwriting, an indication that it once belonged to a codex. The letters iota alpha written in the top right corner of the verso (↓) make it page 11 of the book. The handwriting is characterized by “upright, rounded, yet rather flattened capitals of medium height”21, palaeographi- cally dated “not later than about the middle of the third century”22. Nomina

Text of P.Bodmer XXIV, in BASP 17 [1980] 67-79, p. 67); P.Oxy. 1171 (P20) James (HUNT ed. princ., “late third century”), and P.Mich. XVIII 764 Christian Hortatory Text on a roll (2nd – 3rd cent.). 18. I list here several scribal mistakes. In line 1 an error occurs: whether one reads {oï} to⁄oi oï lógoi or oœtoi oï {oï} lógoi, in either case the scribe made a mistake in the opening line of the text, presuming that this was indeed the beginning of the text. (Mistakes like these are not limited to antiquity; for typographical errors in printed books, see the website Book Errata at http://www.bookerrata.com/). Other scribal mistakes in the papyrus are, for instance, gnws‡e for gnÉses‡e (line 20), the omission of the theta in âpokaluf(‡)ßsetai (line 29), and signs of iotacism, like basileusj instead of basileú- sei (line 8), gnwses‡a[i] instead of gnÉses‡e (line 18), eperwtjse for êperwt±sai (line 23), ‡e‡ammenon for te‡amménon (line 31). In lines 19 and 25, the scribe added a word over the line that was probably omitted at first. All these examples occur commonly in papyri of the period, including in manuscripts of books that are now part of the New Testament canon, see, for instance, J.R. ROYSE, Scribal Habits in Early Greek New Testa- ment Papyri (NTTSD, 36), Leiden, Brill, 2008. L.W. HURTADO, The Earliest Christian Artifacts: Manuscripts and Christian Origins, Grand Rapids, MI, Eerdmans, 2006, p. 189, evaluates corrections as follows: “Even the mistakes of scribes give us hard data of esti- mating attitudes toward the texts copied, their own efforts and those of unknown others reflecting a concern for the wording of these texts and for careful transmission of them”. 19. Ed. princ. B.P. GRENFELL – A.S. HUNT, LOGIA IJSOU, Sayings of Our Lord from an Early Greek Papyrus, London, Henry Frowde, 1897, and EID., The Oxyrhynchus Papyri I: 1 LOGIA IJSOU (Egypt Exploration Society, GRM, 1), London, Egypt Exploration Fund, 1898, pp. 1-3. (Oxford, Bodleian Library MS. Gr. th. e. 7 [P]). Image: P.J. PARSONS, City of the Sharp-Nosed Fish: Greek Lives in Roman Egypt, London, Weidenfeld – Nicolson, 2007, pl. 11. 20. On the reasons for excavating at Oxyrhynchus, see LUIJENDIJK, Greetings in the Lord (n. 7), pp. 6-8. 21. ATTRIDGE, Appendix (n. 4), p. 96. 22. So GRENFELL – HUNT, P.Oxy. IV 654 (n. 14), p. 10, when discussing the date of P.Oxy. I 1 and IV 654: “Both were found on the same site and the papyri are of approx- imately the same date, which is not later than about the middle of the third century, so that both collections must go back at least the second century” (emphasis mine). They initially dated this piece as “within the period 150-300 A.D.” and “probably … not much later than the year 200” (Sayings of Our Lord [n. 19], p. 6), a period that according to Attridge “remains very plausible” (Appendix [n. 4], p. 97). T. BAARDA, Het Evangelie van Thomas:

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 246246 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 247

sacra are used for the words God, father, and human (‡eóv, patßr and ãn‡rwpov). Striving to give the page a neat look, the scribe employed two common techniques to create a straight line in the right margin: using line fillers and abbreviating final nu with a stroke. The third fragment, P.Oxy. IV 65523, consists of six fragments of a papyrus roll24. It preserves logia 24 and 36–39 in Greek25. Written along the fibers, the handwriting consists of a “small uncial of the common sloping oval type, which in most cases belongs to the third century”26. In the sections preserved here, nomina sacra and punctuation lack27.

III. ON A ROLL

I begin with the most striking material aspect, namely the fact that two papyri of the Gospel of Thomas, P.Oxy. IV 654 and 655, are written on a book roll, a feature that may have implications for their use and perceived status. When examining the use of rolls, it is important to state at the outset that throughout antiquity the book roll was the primary vehicle for literature. Its use for a Christian text, however, contravenes strong

Vier korte studies, in ID., et al., Het Evangelie van Thomas (VU-segmenten, 5), Zoeter- meer, Meinema, 1999, 9-45, p. 27, noted that in subsequent publications, Grenfell and Hunt pushed the date up later: “verrassend is dat zij nog eens terugkomen op de datering van het eerste fragment, waarvan zij nu zeggen dat het in de derde eeuw geschreven is, vermoedelijk een decennium eerder dan het tweede fragment”. On this tendency of dating Christian manuscripts early, see BAGNALL, Early Christian Books (n. 9), chapter 1 (pp. 1-24): “The Dating of the Earliest Christian Books in Egypt. General Considerations”. 23. Ed. princ.: B.P. GRENFELL – A.S. HUNT, Fragment of a Lost Gospel, in EID., New Sayings (n. 14), 37-43, and EID., P.Oxy. IV (n. 14), pp. 22-28. See also R.A. KRAFT, Oxyrhynchus Papyrus 655 Reconsidered, in HTR 54 (1961) 253-262. 24. Of the originally eight fragments, two were lost when the papyrus was reframed, see KRAFT, Oxyrhynchus Papyrus 655 (n. 23), p. 257, n. 5. The largest fragment (fragment b), “measures h. 8.2 × w. 8.3 cm; the smallest, g, measures h. 1.0 × w. 0.8 cm” (ATTRIDGE, Appendix [n. 4], p. 98). 25. For differences between the Greek and Coptic text, see ATTRIDGE, Appendix (n. 4), pp. 99-101. 26. GRENFELL – HUNT, P.Oxy. IV 655 (n. 14), p. 22. See also ATTRIDGE, Appendix (n. 4), p. 98. Comparanda are: P.Flor. II 108 (3rd cent.; image: C.H. ROBERTS, Greek Literary Hands 350 B.C. – A.D. 400 [Oxford Palaeographical Handbooks], Oxford, Claren- don Press, 1955, Pl. 22a); P.Oxy. XIX 223 (1st or 2nd cent.; image: W.A. JOHNSON, Book- rolls and Scribes in Oxyrhynchus, Toronto – Buffalo – London, University of Toronto Press, 2004, Pl. 18, foldout 3); P.Oxy. XVII 2078 (2nd cent.; image: E.G. TURNER, Greek Manuscripts of the Ancient World, 2nd rev. ed. P.J. PARSONS, London, University of Lon- don, Institute of Classical Studies, 1987, nr. 33, p. 65). 27. As ATTRIDGE, Appendix (n. 4), p. 99, noted, “No instances of the abbreviation of divine names appear in the surviving fragment, although the abbreviation IS for ˆIjsoÕv must be restored in several places”.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 247247 44/07/11/07/11 11:5611:56 248 ANNEMARIE LUIJENDIJK

Christian preference for the codex as book format28. In the material world of early Christian manuscripts, book rolls stand out. Statistics obtained by searching the Leuven Database of Ancient Books (LDAB) illustrate that at this time there are still more non-Christian than Christian codices29. These data also exemplify how uncommonly rolls with Christian texts occur. In the third century, out of 1218 manuscripts of Greek classical literary texts, 1009, or 83%, are roll format, and only 209, or 17%, take the form of a codex. The situation for the Christian manuscripts is exactly the opposite as early as the third century C.E. already. Out of 182 third-century Christian texts, we find 40 texts copied on rolls, or 22%, and 142, or 78% in codices30. Thus these two rolls with the Gospel of Thomas belong in the minority of 22% of preserved third- century Christian literary texts in roll format (fig. 1)31.

28. For important discussions on the Christian predilection for codices, see: BAGNALL, Early Christian Books (n. 9), pp. 70-90; HURTADO, Earliest Christian Artifacts (n. 18), pp. 43-93; G. STANTON, Early Christian Preference for the Codex, in C. HORTON (ed.), The Earliest Gospels: The Origins and Transmission of the Earliest Christian Gospels. The Contribution of the Chester Beatty Gospel Codex P45, London, T&T Clark, 2004, 40-49; E.J. EPP, The Codex and Literacy in Early Christianity and at Oxyrhynchus: Issues Raised by Harry Y. Gamble’s Books and Readers in the Early Church, in CRBR 10 (1997) 15-37; H.Y. GAMBLE, Books and Readers in the Early Church: A History of Early Chris- tian Texts, New Haven, CT, Yale University Press, 1995; C.H. ROBERTS – T.C. SKEAT, The Birth of the Codex, London, British Academy – Oxford University Press, 1983; E.G. TURNER, The Typology of the Early Codex (Haney Foundation Series, 18), Philadel- phia, PA, University of Pennsylvania Press, 1977, pp. 35-42. See also the interesting observations in R.S. BAGNALL, Jesus Reads a Book, in JTS 51 (2000) 577-588. 29. See also BAGNALL, Early Christian Books (n. 9), p. 73: “It is striking … that even in the fourth century classical literature and other types of non-Christian text make up something like three-fifths of the population of codices, and before that point non-Christian texts make up an even higher percentage of codices. This is part of the reason that it has become impossible to maintain that the codex was a specifically Christian book-form or that the move from roll to codex in the Roman world was primarily driven by Christianity”. 30. The situation for Jewish texts is more difficult to assess. Only 13 manuscripts are classified as Jewish on the LDAB for this period: 6 rolls, 7 codices (August 3, 2010). LDAB catalogs Philo and Josephus manuscripts as Jewish – but were they copied by Jewish scribes? Robert Kraft rightly problematizes a categorization of Jewish and Chris- tian manuscripts, where rolls are Jewish and codices are Christian and concludes that “the presence of Greek codex fragments of Jewish scriptures that do not bear unmistak- able Christian features may be more significant than has usually been admitted”. See R.A. KRAFT, Scroll, Codex, and Canons: The Relationship of Ancient Book Formats to Larger Collections and Anthologies (with Special Reference to Jewish and Christian Scriptures), at http://ccat.sas.upenn.edu/rak/publics/formats/ sub [3], 21 February 2008 (last accessed August 3, 2010). 31. Limiting the search to Christian literary manuscripts found at Oxyrhynchus gives these results: out of 164 (recovered and published) Christian literary manuscripts from Oxyrhynchus, 27 are rolls and 135 codices in that find, or only 16% is in roll format, 82% codex. I conducted a search for “Bookform: roll”; “Culture: literature”; “Religion: Christian”; “Provenance: Oxyrynchos” (sic, without h) and received 27 hits. Changing “Bookform” to “codex” resulted in 135 instances (August 3, 2010).

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 248248 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 249

Fig. 1. Rolls and Codices in the Third Century

What, then, does it mean to find Christian texts copied on a roll? In light of the question of the use of early Christian manuscripts, does the fact that these pieces are written on a roll preclude their use in a worship setting and therefore their status as sacred scripture? Larry Hurtado concluded this for P.Oxy. IV 655: “given the strong general preference for the codex among ancient Christians, especially for texts used as scripture, the choice to copy a text in a fresh roll surely further indicates that this text (or at least this copy of the text) was not used as scripture, i.e., not read publicly in worship settings”32. Indeed, as we shall see, analysis of Christian rolls suggests that they were used primarily in private settings, with only slim hints of public reading. Jews continued to use rolls for their scriptures and the extent to which Jews and Christians formed distinct communities at this time remains open for debate33. As we shall see next, further evidence comes from Christian book rolls themselves.

32. HURTADO, Greek Fragments (n. 5), p. 30. In his Earliest Christian Artifacts (n. 18), p. 60, Hurtado concluded that Christians had “a particularly strong preference for the codex for the texts that they used as scripture”. He noted (ibid., p. 58) that “there is no New Testa- ment text copied on an unused roll among second- or third-century Christian manuscripts”. See also BAGNALL, Early Christian Books (n. 9), pp. 76-78: Christians “put scripture into codices but homilies and the like into book rolls, as if they were normal literary texts. The codex was thus not so much adopted generally by the early Christians for their book production; rather, the Christians adopted the codex as the normative format of deliberatively produced public copies of scriptural texts, but they did not generalize from this adoption to broader use for all books. Or at least they did not do so a great deal sooner than other people did”. 33. On this topic, see, for instance, the essays in A.H. BECKER – A.Y. REED (eds.), The Ways that Never Parted: Jews and Christians in Late Antiquity and the Early Middle Ages (Texts and Studies in Ancient Judaism, 95), Tübingen, Mohr Siebeck, 2003.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 249249 44/07/11/07/11 11:5611:56 250 ANNEMARIE LUIJENDIJK

A survey of the extant early Christian rolls from the second and third centuries indicates that most rolls were meant for private use34. In this category fall texts penned on the back of documentary rolls or badly written ones35, for instance a roll with Hermas, Mandata, penned on the back of a tax register (P.Mich. inv. 44-H)36. Some of these rolls belonged in studious milieus, such as an elegantly written copy of Irenaeus’s Adversus haeresis on a fresh roll (P.Oxy. III 405)37. And we can imagine an educated Christian consulting an onomasticon of Hebrew names with Greek explanations, written in a literary hand on the back of a land register (P.Oxy. XXXVI 2745)38. Several rolls were intended for public reading, that is, texts produced to be easily legible with clear handwriting and/or reader’s aids such as punctuation and spacing39. Examples are fragments of a roll with Genesis in a Biblical Uncial (P.Oxy. IX 1166)40

34. A LDAB search for: “Century: AD02 or AD03”; “Bookform: roll”; “Religion contains Christian” (August 3, 2010) produced 47 results. 35. See C.H. ROBERTS, Manuscript, Society, and Belief in Early Christian Egypt (The Schweich Lectures of the British Academy, 1977), London, British Academy – Oxford University Press, 1979, p. 9. For a discussion of codices with gospels intended for private or public reading, see S.D. CHARLESWORTH, Public and Private – Second- and Third- Century Gospel Manuscripts, in C.A. EVANS (ed.), Studies in Scripture in Early Judaism and Christianity (Library of Second Temple Studies, 70), London, T&T Clark, 2009, 148-75. It should be noted that Charlesworth discusses only gospels that later became canonical gospels, without mentioning the now apocryphal ones. 36. P.Mich. 2.2 130 = P.Mich. inv. 44-H. (Arsinoites, Fayum, 3rd cent.; “about 200 A.D.”, C. BONNER, A New Fragment of the Shepherd of Hermas [Michigan Papyrus 44-H], in HTR 20 [1927] 105-116, p. 107; UMich website: “[Third quarter of the] IInd century A.D.”). On the verso of a tax register from the “latter part of the second century” C.E. (BONNER, New Fragment, p. 107). BONNER (ibid., p. 109) noted: “Our text was written on the verso of a discarded document, and since the hand is not a practised book-hand, it may be that only a few chapters were copied for the writer’s personal use. … we may imagine that the writer of our fragment had before him the modest aim of copying only the Mandates, which might have been contained in a roll of moderate length”. See also J. VAN HAELST, Catalogue des papyrus littéraires juifs et chrétiens (Papyrologie, 1), Paris, Publications de la Sorbonne, 1976, nr. 657; LDAB 1096. 37. P.Oxy. III 405. Oxyrhynchus, 2nd/3rd cent. Irenaeus, Adversus Haereses, on the recto of a roll. VAN HAELST, Catalogue, p. 240, describes the handwriting as: “très belle écriture littéraire”. See also VAN HAELST 671; LDAB 2459. 38. P.Oxy. XXXVI 2745. Oxyrhynchus, 3rd/4th cent. “The verso contains three columns of clear uncial script of a type common in literary papyri” (P.Oxy. XXXVI 2745, p. 1). E.G. TURNER, P.Oxy. XXXVI 2745, p. 2, noted: “The use of a nomen sacrum … demonstrates that this text was copied by a Christian scribe, not a Jewish one, and was probably written in a Christian ambience”. See also VAN HAELST 1158; LDAB 3503. 39. As CHARLESWORTH, Gospel Manuscripts (n. 35), p. 148, observed: “In a public setting where immediacy was called for, punctuation, lectional aids and various kinds of sense breaks in the text could greatly assist the task of the lector”. See also HURTADO, Earliest Christian Artifacts (n. 18), chapter 5: “Other Scribal Features”, pp. 155-189. 40. P.Oxy. IX 1166. Oxyrhynchus, 3rd cent. Gen. 16,8-12 with nomina sacra. HUNT, P.Oxy. IX 1166, p. 1: “The large and upright calligraphic handwriting is apparently an early example of the so-called Biblical uncials and may well fall within the third century”. Hunt

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 250250 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 251

and Isaiah (P.PisaLit. 14)41. At Dura Europos in Syria, the gospel-har- mony Diatessaron was read from a parchment roll42. Often, our attempts to decide a roll’s use are hampered by inconclusive evidence43. Why, for instance, was the Gospel of John copied on the back of a roll, whereas the recto is blank (P.Oxy. X 1228)44? How do these insights in the use of rolls and the inscription of Chris- tian manuscripts affect our understanding of the two rolls with the Gospel of Thomas? Conceivably, the neatly written, freshly used roll of P.Oxy. IV 655 with its short lines functioned as a church piece, although the small size of the letters complicates reading out loud and thus would contradict hypotheses of its use in a larger worship gathering45. It may indeed have been a copy intended for reading at home and display to friends46. The other roll, P.Oxy. IV 654, emits more ambivalent material signs. As mentioned above, this text was penned not just on a roll, but

also points to punctuation: “A medial point, followed, sometimes at any rate, by a short blank space, is used for purposes of punctuation; and a rough breathing is once added. These signs are apparently due to the original writer” (ibid.). Clearly, the text was meant for reading out loud, but in a Jewish or Christian worship? In line 13, kuríou was written as nomen sacrum, but only kappa is preserved (k[). ROBERTS, Manuscript (n. 35), p. 77, concluded: “It is perhaps more likely to be Christian than Jewish”. See also A. RAHLFS – D. FRAENKEL (eds.), Verzeichnis der griechischen Handschriften des Alten Testaments. Vol. 1: Die Überlieferung bis zum VIII. Jahrhundert (Supplement zur Septuaginta Vetus Testa- mentum Graecum), Göttingen, Vandenhoeck & Ruprecht, 2004), nr. 944, p. 296. 41. P.PisaLit. 14 (= P.Alex. inv. 203). Provenance unknown, 3rd-4th cent. Isa 48 (fragment). Nomen sacrum kv. See also VAN HAELST 300; RAHLFS – FRAENKEL 850; LDAB 3127, ROBERTS – SKEAT, Birth of the Codex (n. 28), pp. 39-40, nr. 11. For ROBERTS – SKEAT, ibid., this roll is the only exception in Christian book format. 42. P.Dura 10 = 0212. Dura Europos, 3rd cent. Diatessaron on parchment roll (recto). See also J. JOOSTEN, The Dura Parchment and the Diatessaron, in VC 57 (2003) 159-175; VAN HAELST 699; LDAB 3071. 43. The same applies for codices. 44. ROBERTS – SKEAT, Birth of the Codex (n. 28), pp. 39-40, nr. 13, classify this papyrus as “an eccentric production”, concluding that therefore “for the present purpose we can reasonably leave it out of account” in their overview of Christian rolls. Re P22, K. ALAND, Studien zur Überlieferung des Neuen Testaments und seines Textes (ANTF, 2), Berlin, de Gruyter, 1967, p. 114, gives the ingenious explanation that a scribe copied the text on a reused roll and attached additional pages after reaching the end of the roll. See also VAN HAELST 459; LDAB 2779. 45. HURTADO, Greek Fragments (n. 5), p. 27, interprets the “narrow columns” of P.Oxy. IV 655 as compensation “for the very small size of the lettering”. With a column- width of ca. 5 cm, P.Oxy. IV 655’s format comes very close to William Johnson’s description of the generic ancient book roll, see W.A. JOHNSON, Toward a Sociology of Reading in Classical Antiquity, in The American Journal of Philology 121 (2000) 593-627, p. 609. 46. Johnson’s work allows us to picture such reading groups and their manuscripts: see W.A. JOHNSON, Reading Culture in the High Roman Empire: A Study of Elite Communities (Classical Culture and Society), Oxford, Oxford University Press, 2010, and ID., Sociology of Reading (n. 45). See also his Bookrolls (n. 26), p. 18: “one wonders, but cannot prove, whether the papyri were commissioned for use in some sort of readers’ group”.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 251251 44/07/11/07/11 11:5611:56 252 ANNEMARIE LUIJENDIJK

on a previously used piece. Resourcefully, the ancients often copied writings on the back of earlier used texts, documentary or literary. We find this, for example, also for copies of Hebrews and Revelation47. How should we evaluate the use of such texts? According to Roberts, “any texts written on the back of a roll or sheet discarded as waste declare themselves to be private copies, a view at times borne out by the manner of writing”. He mentions our papyrus, P.Oxy. IV 654, among them48. Yet Roberts warns that not all texts written on the back of other documents should be interpreted solely as “casual or occasional” and gives examples of reused manuscripts that probably belonged to scholars49. Furthermore, he cautions from the example of P.Ryl. I 1, a copy of Deuteronomy: not all texts written on improvised material need have been private. It may have been a paper shortage or just poverty that led one church to economize by sticking together sheets of papyrus already written on one side, fold them, and so form a makeshift codex out of the unwritten sides; this was then used, about the year 300, to take a copy of Deuteronomy whose public character is strongly suggested by the addition of lectional aids – accents, breathings, punctuation, critical signs – to a carefully corrected text50.

I wonder whether Roberts looked with canonical hindsight as he shelved one manuscript made from reused materials, the now biblical text of Deuteronomy, among the public readings, but classified another manuscript with lectional aids of a now apocryphal book, the Gospel of Thomas, among the private copies? If we apply his insights, the Thomas papyrus may indeed have been a private copy but could also have been in the possession of a scholar or even a cash-strapped church51.

47. For Hebrews, see: P.Oxy. IV 657 + PSI XII 1292 = P13. Oxyrhynchus, 3rd/4th cent. Written on the verso of a roll with Livy on the recto (P.Oxy. IV 668 and PSI XII 1291). See also ROBERTS – SKEAT, Birth of the Codex (n. 28), pp. 39-40, nr. 14; VAN HAELST 537; LDAB 3018; For Revelation, see: P.IFAO II 31 = P98. Provenance unknown, 2nd/ 3rd cent. On the back of a document. See D. HAGEDORN, P.IFAO II 31: Johannesapoka- lypse 1, 13–20, in ZPE 92 (1992) 243-247. 48. ROBERTS, Manuscript (n. 35), p. 9. Similarly, BAGNALL, Early Christian Books (n. 9), p. 76, observed: “when making private copies on used papyrus, early Christians behaved just like anyone else, using the blank backs…”. 49. ROBERTS, Manuscript (n. 35), p. 9. 50. Ibid., pp. 9-10. The manuscript, P.Ryl. I 1 (= VAN HAELST 55, RAHLFS – FRAENKEL 920, LDAB 3169), contains Deut 2,37–3,1.3-5.7-13. On the practice of constructing codices from used book rolls and its implications, see also BAGNALL, Early Christian Books (n. 9), p. 59: “They give us clear evidence that in the circles that produced these books, which have not yet been identified with any degree of confidence, the cost of papyrus was a sufficient consideration to warrant the labor of gluing the old rolls together rather than simply buying new ones”. 51. GAMBLE, Books and Readers (n. 28), pp. 80-81, proposed that one late second- century Christian roll, P.Oxy. III 405, a copy of Irenaeus, Adv. haer., “could suggest that

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 252252 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 253

The textual layout in P.Oxy. IV 654 further complicates the picture of its reading environment and also provides interesting glimpses into the possible use of the Gospel of Thomas in an earlier phase. As common for its time, the text is written in continuous script (scriptio continua). In order to guide the reader, the scribe also used multiple forms of lectional aids: diaeresis, coronis, and paragraphus52. The paragraphus, “a horizontal stroke written below the line at the left margin”, occurs five times in this piece and serves to separate the sayings: every paragraphus marks the end of a logion53. This ancient punctuation illustrates how copyists and readers of this text conceived its sense units54. So what do tiny dots and short strokes reveal about this text’s use? In order to fully understand the paragraphus’ function, I recap here an experiment William Johnson conducted. Johnson discovered the function of the paragraphus when he tried “to vivify for undergraduates what it was like for the ancients to read”. He narrates that he wrote his class lecture in an imitation of a book roll. When he read it out loud, he became aware of the paragraph’s performance value: As I practiced and later delivered the lecture, it became clear to me how very useful the paragraphus is when you are reading aloud. Prose columns in ancient book rolls are generally very narrow, with the result that a typical column contains only a few sentences, hence only a few paragraphi. As you look up to your audience, or pause in the reading of the lecture to add some parenthetic remark or entertain a comment, you need only recall, ‘second paragraphus down.’ Returning to your text, the paragraphus immediately reorients you to the start of the next sentence55.

Similarly, diaeresis and coronis guided the readers in pronouncing the words and finding passages. Inscribed with reader’s aids, the Thomas roll appears intended for reciting. Why does this copy have so many lectional aids? I offer several solu- tions. The manuscript may have been used in a liturgical setting, like

works of scholarship, as distinct from scriptural texts, persisted for a while in roll form in Christian scholastic circles”. 52. On layout and punctuation, see TURNER, Greek Manuscripts (n. 26), pp. 8-15. 53. For the definition, see W.A. JOHNSON, The Function of the Paragraphus in Greek Literary Prose Texts, in ZPE 100 (1994) 65-68. In P.Oxy. IV 654, the paragraphus stands between lines 5 and 6, after the first saying; between lines 9 and 10, after the second saying; between lines 21 and 22, after the third logion; between lines 27 and 28, after the fourth logion, and, finally, between lines 31 and 32, after the fifth logion. 54. HURTADO, Earliest Christian Artifacts (n. 18), p. 181, also points to the exegetical implications of readers’ aids such as “sense-unit divisions”: “the early manuscripts in which these devices were deployed are artifacts of early Christian exegesis of these texts, and probably also reflect something of how these texts were read liturgically…”. 55. JOHNSON, Paragraphus (n. 53), p. 66.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 253253 44/07/11/07/11 11:5611:56 254 ANNEMARIE LUIJENDIJK

Roberts proposed for the Deuteronomy manuscript mentioned above. Alternatively, this text may have been intended for reading out loud in a different context, for instance, in an educational setting56. Or the scribe may have copied the punctuation from the Vorlage. Indeed, Johnson found that scribes often replicated the punctuation from their exemplar, concluding that “substantial portions of details like adscript and punctuation seem to be part of what is traditionally copied, part of the para- dosis”57. If that has happened with P.Oxy. IV 654, too, we find a hint of an earlier stage in the transmission of the text, with a copy intended for declamation. Thus a possibility has opened up that if we peeked into an early Christian worship service, we could overhear the Gospel of Thomas being read from P.Oxy. IV 654’s exemplar.

IV. THE GOSPEL OF THOMAS IN CODEX FORMAT

Having examined the two rolls, I now turn to the page from a codex, P.Oxy. I 1. What can we infer from material aspects about this fragment’s use? Can we deduct whether it was intended for private or public read- ing58? Scholars have drawn attention to the Christian predilection for the

56. For the use of the Gospel of Thomas in the 4th century, see Eduard Iricinschi’s recent dissertation on the Nag Hammadi codex that contains that gospel in Coptic (NHC II): E. IRICINSCHI, The Scribes and Readers of Nag Hammadi Codex II: Book Production and Monastic Paideia in Fourth-Century Egypt, Ph.D. diss., Princeton University, 2009. 57. JOHNSON, Bookrolls (n. 26), p. 8 and case studies on pp. 15-37 (for an exception to this practice, see, for instance, ibid., p. 28 re scribe #A33). Johnson noted that for the 2nd and 3rd cent. literary rolls he studied “punctuation was not formulated independently by the scribe”; he even speaks about “some sense of a tradition of punctuation” (ibid, p. 58). Similarly, TURNER, Greek Manuscripts (n. 26), p. 12, observed: “during the Roman period in Egypt (especially in ii A.D.) the view seems to have taken root that if punctuation was present in the exemplar it was the first scribe’s duty to copy it”. See also CHARLES- WORTH, Gospel Manuscripts (n. 35), p. 151. We observe these same practices in Jewish manuscripts, see E. TOV, Scribal Practices and Approaches Reflected in the Texts Found in the Judean Desert (Studies on the Texts of the Desert of Judah, 54), Leiden, Brill, 2004, p. 150: “As a rule, scribes copied the divisions between section units from their Vorlagen, but they sometimes deviated from them…”. 58. CHARLESWORTH, Gospel Manuscripts (n. 35), studied the use of “second- and third- century gospel manuscripts” (again, only manuscripts of the four now canonical gospels). According to that author, the distinction between copies intended for private and public reading originated in different production sites. He argued (ibid., p. 149) that “early gospel MSS intended for public use were produced in controlled settings, while MSS intended for private use were copied in casual settings where quality controls over copying/production were lacking”. CHARLESWORTH, ibid, p. 171, supposes the existence of “Christian copying centres in major cities” that would have “master copies of public gospel MSS. … If such copying centres were attached to larger churches, the ‘master’ copes were probably MSS read in the church”. He argues (ibid., pp. 148-149 and throughout) that most of the

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 254254 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 255

codex format for copies of what were considered scriptural texts59. As Roger Bagnall aptly observed, those are not confined to what later became canonical writings60. Hurtado evaluated P.Oxy. I 1 as private document, “prepared by a scribe of modest literary ability and quite possibly for someone of modest financial resources … or else by/for a reader who wanted only a readable copy of the text”61. His conclusions are grounded in evidence that is open for debate. Hurtado concluded that the codex was not a public manuscript because of the quality of the handwriting, which he char- acterizes as “clear and competent, but workaday and certainly not calligraphic… Other features, such as ligatures confirm a copyist of very limited aesthetic abilities”62. Grenfell and Hunt, however, deemed P.Oxy. I 1 “a leaf from a handsomely-written book, which may well have been

second-century gospels were copied for the purpose of reading in worship. For the third century, his argument goes, the rise in number of Christians brought a growing interest in private study, which necessitated the production of appropriate manuscripts. Unfortunately, we have hardly any information about the production sites of Christian texts in this period. Possibly the church at Oxyrhynchus had one, associated with its bishop Sotas (see LUIJENDIJK, Greetings in the Lord [n. 7], pp. 144-151), where also documents were issued, or at least, where a scribe for the bishop could obtain writing material for a documentary letter. 59. HURTADO, Earliest Christian Artifacts (n. 18), pp. 58-61; BAGNALL, Early Chris- tian Books (n. 9), p. 78. Both define this differently, see BAGNALL, p. 81. 60. BAGNALL, Early Christian Books (n. 9), p. 81: “The fact that gospels that did not make it into the eventual canon and authors like Hermas who also did not made the cut are written in the form of the codex almost as much as Matthew, Luke, John, and Paul, and that the use of the codex for these noncanonical texts is strongest in the earliest manuscripts, suggests that any confidence in an emerging canon as the basis of the use of the codex is misplaced. It was not a specific canon but a type of material that the church – I leave aside for the present the question of who ‘the church’ is – decided to reproduce in codex form”. 61. HURTADO, Greek Fragments (n. 5), p. 24. He continued: “The codex format reflects the general preference for this book-form among early Christians, and the likely page-size is a common one among codices of the period. The absence of the sort of scribal devices that we customarily associate with copies prepared for public reading suggests that this may have functioned as a personal/private copy”. 62. Ibid., p. 22 (emphasis mine). In reaction to Attridge’s description of the handwriting as “an informal literary hand”, HURTADO, ibid., p. 22, n. 12, notes: “I would say that among papyri of this period and page-size, the letters are somewhat on the small side, and that the hand is unskilled at literary quality, rather than ‘informal’” (emphasis mine). In his Earliest Christian Artifacts, Hurtado still described the script as an “informal but competent hand” (HURTADO, Earliest Christian Artifacts [n. 18], p. 239, Pl. 7). Hurtado mentions a repair in the page before writing yet even the most expensive codices, such as Sinaiticus and Vaticanus have such defects. According to PATTIE, Creation of the Great Codices (n. 9), p. 64, “an average books has many flaws in the parchment when wounds in the skin have expanded when the skin was stretched. Even a fine manuscript like Vaticanus has many flaws of this nature, including a significant number in the text area, that force the scribe to write around the holes. Sinaiticus has many fewer flaws and most were repaired before the text was inscribed”.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 255255 44/07/11/07/11 11:5611:56 256 ANNEMARIE LUIJENDIJK

a valuable trade-copy”63. This discrepancy in evaluation demonstrates the well-known subjectivity in palaeographical judgments. The matter of handwriting alone, however, does not suffice to set this piece aside as intended for private reading. Numerous contemporary manuscripts of now biblical books reveal similar hands, written in what Colin Roberts labeled “informal uncial” or “reformed documentary”64. Roberts analyzed this style of script as follows: though the writing is far from unskilled, [these Christian papyri] are the work of men not trained in calligraphy and so not accustomed to writing books, though they were familiar with them; they employ what is basically a documentary hand but at the same time they are aware that it is a book, not a document on which they are engaged. They are not personal or private hands; in most a degree of regularity and of clarity is aimed at and achieved65.

From all we know, such relatively unimpressive manuscripts were also read in worship settings and considered sacred scripture66. Our worn page of the Gospel of Thomas, P.Oxy. I 1, conforms with this evidence67. Moreover, in trying to establish the use of this codex, I find its reported archaeological context intriguing68. Grenfell and Hunt, the excavators at

63. GRENFELL – HUNT, P.Oxy. IV 654 (n. 14), p. 10. 64. ROBERTS, Manuscript (n. 35), p. 15. K. HAINES-EITZEN, Guardians of Letters: Literacy, Power, and the Transmitters of Early Christian Literature, Oxford, Oxford Uni- versity Press, 2000, p. 68, see also pp. 32f., argued for the “multifunctionality” of the scribes of early Christian manuscripts. 65. ROBERTS, Manuscript (n. 35), p. 15. According to Roberts, this style persists into the fourth century (ibid., p. 16). 66. See also E.J. EPP, The New Testament Papyri at Oxyrhynchus in Their Social and Intellectual Context, in W.L. PETERSEN – J.S. VOS – H.J. DE JONGE (eds.), Sayings of Jesus: Canonical and Non-Canonical: Essays in Honour of Tjitze Baarda (SupplNT, 89), Leiden, Brill, 1997, 47-68, p. 67; reprinted in E.J. EPP, Perspectives on New Testament Textual Criticism. Collected Essays, 1962-2004 [SupplNT, 116), Leiden, Brill, 2005, 497- 520, p. 517: “What is relevant, however, is that early Christian books were essentially practical and produced for use in the life of the Christian community” (emphasis original). 67. BAGNALL, Early Christian Books (n. 9), p. 74, also placed this codex fragment, P.Oxy. I 1, among Christian scriptures, noting: “Almost all texts that can be described as scripture appear in codex form. In this category I include the Old Testament, which was effectively canonical for Christians of this period; the books of the New Testament, although not yet fully canonized; those gospels that did not make it into the eventual canon, like Thomas; and the Shepherd by Hermas, a very popular also-ran… Up to the turn of the second to third centuries, the twenty-seven Christian (or probably Christian) codices shown in table 4.1 consist entirely of scripture in this sense, as far as the texts have been identified. By contrast, there are 9 or 10 Christian texts in roll forms, of which only three might be scripture”. 68. HURTADO, Greek Fragments (n. 5), p. 31, commented that he did not know how these Thomas fragments were found in relation to one another: “Unfortunately, I am not aware that we know the precise find-spot(s) for these three copies of GThom in relation to the other Christian manuscripts from Oxyrhynchus…”.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 256256 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 257

Oxyrhynchus, seldom mentioned where they unearthed their papyri. Therefore, it is remarkable that they repeatedly stated that they found P.Oxy. I 2, a page of the Gospel of Matthew “near that containing the ‘Logia’ [= P.Oxy. I 1] a day or two afterwards”69. While I recognize that this scenario fits their agenda of the importance of this particular early Christian fragment70, the fact that these two papyri of the Gospel of Matthew and the Gospel of Thomas apparently were found together could indeed mean that they had belonged to a Christian library, as Gren- fell and Hunt also proposed71, and as such may have been read alternatively in worship. Thus materially and archaeologically, this codex of the Gospel of Thomas matches other Christian papyri from this period, including those that now, in retrospection, are considered canonical. So far a closer look at these three manuscripts has complicated the picture of the Gospel of Thomas’ use. A papyrological investigation opened the possibility that the Gospel of Thomas was declaimed from P.Oxy. I 1’s codex to those gathered for worship. The two rolls, P. Oxy. IV 655’s new roll and P.Oxy. IV 654’s re-used roll, appear more at home in a private, studious milieu. And P.Oxy. IV 654’s reader-friendly Vorlage might have featured in worship. I suspect that it is significant that the earliest copy of Thomas, P.Oxy. I 1, was written in codex format, while the later copies appeared in rolls. Apparently, we witness here a development from the Gospel of Thomas as scripture to document for private study. In the following section, I shall argue that the papyrological evidence as I have laid it out here matches Origen of Alexandria’s perception of the Gospel of Thomas.

V. ORIGEN THE PREACHER AND THE GOSPEL OF THOMAS

In order to get a fuller grasp on the reception of the Gospel of Thomas in the third century72, we make a stop at Caesarea, where we attend a

69. GRENFELL – HUNT, P.Oxy. I 2 (n. 19), p. 4. 70. See also LUIJENDIJK, Greetings in the Lord (n. 7), pp. 6-8, esp. p. 7, n. 18. 71. B.P. GRENFELL, Excavations at Oxyrhynchus: First Season (1896-7), in A.K. BOWMAN, et al. (eds.), Oxyrhynchus: A City and Its Texts, London, Egypt Exploration Society, 2007, p. 348: “It is not improbable that they [i.e. P.Oxy. I 1 and 2] were the remains of a library belonging to some Christian who perished in the persecution during Diocletian’s reign, and whose books were then thrown away”. For a very similar account, see B.P. GRENFELL, The Oldest Record of Christ’s Life: The Story of Its Recent Discovery, in McClure’s Magazine, October 1897, 1022-1030, pp. 1027-1028 (available online at http://www. gospels.net/thomas/grenfell_oldestrecord.pdf). The link with the Great Persecution cannot be maintained, as I have shown elsewhere, see A.M. LUIJENDIJK, Sacred Scriptures as Trash: Biblical Papyri from Oxyrhynchus, in VC 64 (2010) 217-254, pp. 240-241. 72. I do not discuss another third-century source here: Hippolytus’s Refutatio omnium haeresium V 7.20, for this does not present a quotation from the Gospel of Thomas. See

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 257257 44/07/11/07/11 11:5611:56 258 ANNEMARIE LUIJENDIJK

Christian worship to hear premier biblical scholar Origen of Alexandria preach73. In two of his homilies, the famous exegete quotes a saying that is also preserved as logion 82 in the Gospel of Thomas, however, without identifying it as such74. Once, Origen informed his audience: I have read elsewhere as if the Savior was speaking (legi alicubi quasi Salva- tore dicente) – and I question whether it was someone who was a figure for the person of the Savior or if it was appended in his memory or if this may be truly what he said – the Savior there says (ait autem ibi Salvator), ‘Whoever is near me is near fire; whoever is far from me, is far from the kingdom’75.

At another occasion, preaching on Joshua, Origen told the congregation: Blessed are those who deserve to be very close to God. But remember that it is written (scriptum est), ‘Those who draw near to me, draw near to fire,’

S.R. JOHNSON, Hippolytus’s Refutatio and the Gospel of Thomas, in Journal of Early Christian Studies 18 (2010) 305-326. Johnson (ibid., p. 306, see also pp. 314-315) concludes, rightly: “Hippolytus himself probably did not have an actual copy of the Gospel of Thomas at hand”. 73. My aim here is not provide a comprehensive overview of Origen’s (possible) citations and allusions to the Gospel of Thomas, but rather to illustrate his perception of that gospel through selected examples as a parallel to the papyrological record. For fuller treatments of the Gospel of Thomas in Origen’s writings, see M. GROSSO, Osservazioni sui testimonia origeniani del Vangelo secondo Tommaso (in Luc. hom I,1; contra Celsum VIII,15; in Ier. hom. lat. I,3; in Jesu Nave hom. IV,3), in Adamantius 15 (2009) 177-194 and S.C. CARLSON, Origen’s Use of the Gospel of Thomas (forthcoming). I thank Stephen Carlson for sending me his piece in advance of publication. For a source-by-source analysis of Origen’s (and Clement’s) use of non-canonical scriptures, see A. VAN DEN HOEK, Clem- ent and Origen as Sources on “Noncanonical” Scriptural Traditions During the Late Second and Earlier Third Centuries, in G. DORIVAL – A. LE BOULLUEC, et al. (eds.), Origeniana Sexta. Origène et la Bible/Origen and the Bible. Actes du Colloquium Orige- nianum Sextum. Chantilly, 30 août–3 septembre 1993 (BETL, 118), Leuven, University Press – Peeters, 1995, 93-113. On Origen and the New Testament canon, see, for instance, E.R. KALIN, Re-examining New Testament Canon History. 1: The Canon of Origen, in Currents in Theology and Mission 17 (1990) 274-282. 74. In the Coptic, logion 82 reads peãe is ãe pet#yn eroei ev#yn etsate auw petouyu Mmoei vouyu NtmNtero (“Jesus said, ‘Whoever is near me is near the fire, and whoever is far from me is far from the kingdom’”). See, for instance, DECONICK, Original Gospel of Thomas (n. 6), pp. 246-247. 75. Origen, Hom.Jer. 27.3.7. Translation: Origen: Homilies on Jeremiah, Homily on 1 Kings 28, trans. J.C. SMITH (The Fathers of the Church, 97), Washington, DC, Catholic University of America Press, 1998, pp. 254-255, Homily 27 (50). The Greek has not been preserved. The longer section in Jerome’s Latin translation reads: Legi alicubi quasi Salvatore dicente – et quaero, sive quis personam figuravit Salvatoris sive in memoriam adduxit, an verum sit hoc dictum est –, ait autem ibi Salvator: ‘Qui iuxta me est, iuxta ignem est; qui longe est a me, longe est a regno’. Ut enim qui iuxta est, iuxta salutem est, ita et iuxta ignem est. Et qui audiens me et audita praevaricans factus est vas irae paratum in perditionem, cum iuxta me est, iuxta ignem est. Si vero quis cavens, quoniam qui iuxta me est iuxta ignem est, longe factus fuerit a me, ne iuxta ignem fiat, sciat talis quia longe futurus est et a regno. From: Origène: Homélies sur Jérémie, ed. and trans. P. HUSSON – P. NAUTIN (Sources chrétiennes, 238), Paris, Cerf, 1977, p. 324, lines 101-111 (“Homélie L. I (III), 3”).

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 258258 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 259

If you are gold and silver and have drawn near to the fire, you will shine forth more splendid and glowing because of the fire76. An important observation from these two quotations is that Origen does not rely here on oral tradition. In both homilies, he states explicitly that he found the saying in a written source (“I have read…”)77. He even introduces one citation with the words “it is written”; a formula he frequently uses to present scriptural quotations78. Yet he does not draw from the biblical canon as we have it now79. So where did Origen find this saying? Did he indeed quote from the Gospel of Thomas or did he perhaps find the saying in another work80? The latter option remains possible for at Caesarea, and earlier in his career at Alexandria, Origen

76. Origen, Hom.Josh. 4.3. Translation: Origen: Homilies on Joshua, trans. B.J. BRUCE, ed. C. WHITE (The Fathers of the Church, 105), Washington, DC, Catholic University of America Press, 2002, pp. 55-56, Homily 4: “Concerning the crossing of the Jordan” (at 4.3). The Latin reads: Beati, qui merentur proximi esse Deo. Sed memento quod scriptum est: qui approximant mihi, approximant igni. Si aurum vel argentum fueris et igni approximaveris, multo per ignem splendidior et rutilantior effulgebis. From: Origène, Homélies sur Josué. Texte latin, introduction, traduction et notes de Annie Jaubert (Sources chrétiennes, 71), Paris, Cerf, 1960, pp. 154, 156. Jaubert refers to Hom. Jer. 20 and quotes the Greek of Didymus the Blind, Exp. in Ps. 88,8 PG 39, 1488 D: ö êggúv mou êggùv toÕ puróv (ibid., pp. 154-155, n. 1). BRUCE, Hom. on Josh., p. 56, n. 30, notes: “In Homilies on Jeremiah 20.3, Origen gave the full quotation but questions its authority. Following him, Didymus of Alexandria quoted it in his Commentary on the Psalms 88.8. In Homilies on Luke 1 … Origen condemned the Gospel of Thomas, but he evidently felt the saying had a ring of authenticity”. Origen probably preached this homily towards the end of his life, according to BRUCE, Hom. on Josh., p. 19: “the discourses on Joshua … seem to refer to the Decian persecution (249-50), under which Origen suffered, the Homilies on Joshua are usually considered to have been given in the last years of his life”. See also GROSSO, Testimonia origeniani (n. 73), pp. 190-194. 77. Joachim Jeremias observed: “Origen took the quotation from a written (‘I read’) apocryphal (‘somewhere’) source”, see J. JEREMIAS, Unknown Sayings of Jesus, trans. R.H. FULLER, London, SPCK, 1964, p. 66. So also CARLSON, Origen’s Use (n. 73), p. 9: “Origen’s ‘I have read somewhere’ … indicates that his source is written, not oral”. On the citation methods of Origen’s predecessor Clement, see A. VAN DEN HOEK, Techniques of Quotation in Clement of Alexandria: A View of Ancient Literary Working Methods, in VC 50 (1996) 223-243. 78. A TLG search (last accessed August 3, 2010) produced 678 results (and that only for Origen’s works preserved in Greek). Origen applies this formula, for instance, to citations from the Gospel of Matthew, John and “Old Testament” texts. 79. JAUBERT, Origène, Homélies sur Josué (n. 76), p. 154, n. 1, also remarks: “Cette citation ne se trouve nulle part dans l’Écriture”. 80. Origen’s predecessor at Alexandria, Clement of Alexandria, Stromateis, twice quoted a saying similar to that found in GosThom 2. In the first instance, he attributed the quotation to the Gospel according to the Hebrews; in the second instance he did not give a source (resp. Strom. II 9.45 and V 96.3). Or was this saying in GosThom 82, as Keith Elliott phrased it, part of “a floating oral tradition of sayings attributed to Jesus [that] sometimes came to ground in more than one place”? (J.K. ELLIOTT, The Apocryphal New Testament: A Collection of Apocryphal Christian Literature in an English Translation Based on M.R. James, Oxford, Clarendon Press, 1993, p. 129).

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 259259 44/07/11/07/11 11:5611:56 260 ANNEMARIE LUIJENDIJK

had access to the best-stocked early Christian libraries of his time. Conceivably, therefore, he chanced upon this saying in another, now lost, text. As we will see next, however, there are good indications that he cited these words from the Gospel of Thomas. Beyond the main fact that the quotations match saying 82 in the Gospel of Thomas81, I adduce two other pieces of evidence that Origen cited from that gospel. First, the phrases he employs to introduce the quotation, namely “as if the Savior was speaking” and “the Savior, however, says there” (ait autem ibi Salvator), follow the organization of the Gospel of Thomas with its repetition of “Jesus says/said”: légei ˆIj(soÕ)v/peãe i(ysou)s) and therefore makes it very likely that they derived from that text. Secondly, we know from his other writings that Origen had indeed read the Gospel of Thomas. In a homily on the Gospel of Luke, Origen admits that he knew the Gospel of Thomas82. According to the great preacher, the four now canonical gospels alone were written with the help of the Holy Spirit: Matthew, Mark, John, and Luke did not ‘try’ to write; they wrote their Gospels when they were filled with the Holy Spirit. … The church has four gospels. Heretics have very many. … I know one gospel called According to Thomas, and another According to Mathias. We have read many others, too, lest we appear ignorant of anything, because of those people who think they know something if they have examined these gospels. But in all these

81. The citation in Hom.Jer. matches the GosThom 82 verbatim; the one in Hom.Josh. differs slightly. CARLSON, Origen’s Use (n. 73), p. 9, argues persuasively that “the information gleaned from Origen’s disclaimer in quoting this saying of the Savior fits the Gospel of Thomas”. According to VAN DEN HOEK, Clement and Origen (n. 73), p. 103, Origen presented his borrowings from the Gospel according to the Hebrews and Hermas carefully: “Origen introduces all of his three citations [of the Gospel according to the Hebrews] with a conditional clause, ‘if one accepts the gospel …’. This kind of qualifica- tion was also present in some of the previous quotations from Hermas and is totally absent in Clement. It seems Origen’s hallmark not to want to offend anybody in using these texts, and the phrasing of his apology is always similar”. The same applies to this quotation also. 82. So also JEREMIAS, Unknown Sayings (n. 77), pp. 66-67: “There can be no doubt it was from the Gospel of Thomas. He [Origen] certainly knew that work, for in a homily on Luke he stigmatizes it rather emphatically as an apocryphal writing. But if he rejected it on that occasion, he could hardly refer to it in another homily as a reliable source or base any argument upon it”. T. BAARDA, De datering van ‘Thomas’, in ID., Het Evangelie van Thomas (n. 22), p. 30 and n. 61, wonders whether Origen indeed knew the Gospel of Thomas or only had heard about it. He bases his caution on the Greek preserved in the catenae: “For there is also in circulation the Gospel according to Thomas” (féfetai gàr kaì tò katà Qwm¢n eûaggélion). Origen gave these homilies after his move to Caesarea, Pales tine, in ca. 233. J. LIENHARD, Origen; Homilies on Luke (Fathers of the Church, 94), Washington, DC, Catholic University of America Press, 1996, “Introduction”, p. xxiv, dates them between the years 233 and 244, so early in Origen’s Caesarean period. Presumably, Origen already knew the text in Alexandria, Egypt.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 260260 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 261

questions we approve nothing but what the church approves of, namely only four canonical gospels83.

A sincere scholar, Origen did not refuse a book ignorantly; rather, he emphasized that he had read the books he rejected84. The passage’s strong rhetoric is revealing: Origen’s adamant opposition to the Gospel of Thomas only makes sense if some Christians were reading this text on par with the other gospels. As we have seen, this may have happened at Oxyrhynchus, in Origen’s native country Egypt. I interpret Origen’s use of the Gospel of Thomas in his homilies in tandem with the manuscript evidence of that text. In these homilies, Origen displays an ambivalent attitude towards the Gospel of Thomas: on the one hand, he did not consider it part of his evangelical quartet, as evidenced in his homily on the Lukan prologue, yet on the other hand, in other homilies, he quoted from that same Gospel, even introducing the saying with the scriptural quotation formula (scriptum est), as if it were scripture, attributing its words to the savior (ait autem ibi Salvator)85. If this scenario sounds familiar now, I argue that this is because we found a similar dualistic attitude towards the acceptability of the Gospel of Thomas in the three papyrus manuscripts from Oxyrhynchus produced at roughly the same time that Origen voiced his opinions.

83. Translation LIENHARD, Hom.Luke (n. 82), pp. 5-6. Latin: H. CROUZEL – F. FOURNIER – P. PÉRICHON (eds. and trans.), Origène, Homélies sur S. Luc: Texte latin et fragments grecs; introduction, traduction et notes (Sources chrétiennes, 87), Paris, Cerf, 1962, p. 100. 84. As METZGER, Canon (n. 1), p. 141, acknowledged: “It is difficult to summarize the views on the canon entertained over the years by a mind as fertile and as wide-ranging as Origen’s. Certainly it can be said, however, that he regarded the canon of the four Gospels as closed. … At other times Origen, … accepts as Christian evidence any material he finds convincing or appealing, even designating such writings on occasion as ‘divinely inspired’”. 85. Origen may have chanced his position on the Gospel of Thomas, perhaps prompted by his move from Alexandria to Caesarea and the ensuing change of profession from teacher to preacher. As VAN DEN HOEK, Clement and Origen (n. 73), p. 95, suggests, “perspectives can change even during the course of these writers’ lives; particularly in Origen’s case, the move from Alexandria to Caesarea seems to have marked an altera- tion”. Similarly, regarding Origen’s take on Hermas, she remarks (ibid. p. 99): “Origen nonetheless indicates that the booklet of the Shepherd – as he calls it – was not accepted by everyone. In his later years Origen seems to have expressed this reservation rather frequently. It is unclear whether the environment of Caesarea offered more resistance to Origen’s visionary source or if Origen himself eventually grew more doubtful about it”. Also METZGER, Canon (n. 1), p. 141, noted a development in Origen’s thinking on the canon: “There is somewhat greater readiness to make use affirmatively of non-canonical texts while he was a teacher at the catechetical school of Alexandria, as compared with a certain caution and circumspection observable later in the contexts of giving Biblical expositions from the pulpit at Caesarea”.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 261261 44/07/11/07/11 11:5611:56 262 ANNEMARIE LUIJENDIJK

VI. DIVERSE READING PRACTICES: THE GOSPEL OF PETER AND HERMAS

The situation I have sketched here for the reading of the Gospel of Thomas does not constitute an isolated instance. Literary and manuscript evidence shows that in worship gatherings throughout the Mediterranean, early Christians heard texts read in worship that later were considered to be apocryphal. Let me offer some examples. Some decades before Origen and the transcription of the Thomas papyri, the Gospel of Peter was read at the church of Rhossus in Asia Minor. We know about this instance through a conflict reported in Eusebius of Caesarea. Before having studied the text, Serapion, bishop of Antioch (ca. 190-211), approved of the public reading of the Gospel of Peter. However, when he had managed to obtain a copy of it “from others who had studied it diligently” and “read it through”, he changed his mind and considered it heretical86. A sixth- century copy of the Gospel of Peter from Akhmim proves that Greek- speaking Christians in Upper Egypt still held that text in high esteem87. Another case forms the Shepherd of Hermas. Part of the pandect Chris- tian bible of the Codex Sinaiticus, the text also exists in an abundance of early papyrus fragments, clearly an indication of its use and value. While second-century bishop Irenaeus of Lyon counted the Shepherd as scripture, Alexandrian bishop Athanasius in his 39th Festal Epistle of the year 367 stipulated that it was not canonized but rather should “be read to those who now approach [us] and want to be instructed in the word of piety”88.

86. Apparently, there were fractions among the congregation, reading of the Gospel of Peter being one of the reasons for disagreement. Serapion wrote to the church (apud Eusebius, Ecclesiastical History VI 12.4): “For I myself, when I came among you, imagined that all of you clung to the true faith; and, without going through the Gospel put forward by them in the name of Peter, I said: If this is the only think that seemingly causes captious feelings among you, let it be read (ânaginwskés‡w). But since I have now learnt, from what has been told me, that their mind was lurking in some hole of heresy, I shall give diligence to come again to you; wherefore, brethren, expect me quickly…”. See: Eusebius, Ecclesiastical History, Vol. 2, trans. J.E.L. OULTON (Loeb Classical Library, 265), Cambridge, MA, Harvard University Press, 1932; repr. 1994, p. 41. 87. So LÜHRMANN, Die apokryph gewordenen Evangelien (n. 6), p. 18: “Noch im 6. Jh. finden sich in Oberägypten griechisch sprechende Christen, die das Petrusevangelium in hohen Ehren hielten…”. On the Gospel of Peter, see also ibid., chapter 2, pp. 55-104. 88. ânaginÉskes‡ai to⁄v ãrti proserxoménoiv kaì bouloménoiv katjxe⁄s‡ai tòn t±v eûsebeíav lógon, Athanasius, Letter 39. Edition: P.-P. JOANNOU, Fonti. Fas- cicolo ix. Discipline générale antique (ii-ix s.). Les canons des pères grecs. Vol. II: Rome, Tipographia Italo-Orientale S. Nilo, 1963, 71-76, p. 75. See also the translation from the Coptic in D. BRAKKE, Athanasius and the Politics of Asceticism (Oxford Early Christian Studies), Oxford, Clarendon Press, 1995, 326-332, p. 330. See also ID., Canon Formation and Social Conflict in Fourth-Century Egypt: Athanasius of Alexandria’s Thirty-Ninth Festal Letter, in HTR 87 (1994) 395-419 and ID., A New Fragment of Athana-

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 262262 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 263

The Muratorian Fragment, a fourth-century acephalous canon list, recommended the Shepherd for private reading, but banned its public reading: “it cannot be read publicly to the people in church”89. These prohibi- tions in fact suggest that one could still hear this text performed liturgically in fourth-century church gatherings90. Christians maintained such diverse reading practices91. These continued into even the sixth century, as a homily by the Coptic monk John of Parallos (c. 540 to 610/620) proves92.

sius’s Thirty-Ninth Festal Letter: Heresy, Apocrypha, and the Canon, in HTR 103 (2010) 47-66. 89. The relevant passage in the Canon Muratori reads in translation: “(71) We receive only the apocalypses of John and Peter (72) though some of us are not willing that the latter be read in church. (73) But Hermas wrote the Shepherd (74) most recently, in our times, in the city of Rome. (75) while bishop Pius, his brother, was occupying the [episcopal] chair (76) of the church of the city of Rome. (77) And therefore it ought indeed to be read; but (78) it cannot be read publicly to the people in church either among (79) the prophets, whose number is complete or among (80) the apostles, for it is after [their] time” (emphasis mine). The notoriously bad Latin reads: 71Apocalapse etiam Iohanis et Pe72tri tantum recip(e)imus quam quidam ex nos73tris legi in eclesia nolunt. Pastorem uero 74nuperrim e(t) temporibus nostris in urbe 75Roma Herma conscripsit sedente cathe76tra urbis Romae aeclesiae Pio eps fratre(r) 77eius et ideo legi eum quide oportet se pu78plicare uero in eclesia populo neque inter 79profetas conpletum numero neque inter 80apostolos in fine temporum potest. Translation: METZGER, Canon (n. 1), p. 307, slightly modified, based on Hahneman’s work; Latin: G.M. HAHNEMAN, The Muratorian Fragment and the Devel- opment of the Canon (Oxford Theological Monographs), Oxford, Clarendon Press, 1992, p. 7 (with some added punctuation) and chapter 2 “The Shepherd of Hermas”, pp. 34-72. For an argument on the Canon Muratori as a late-second- or early-third-century Western list, see J. VERHEY DEN, The Canon Muratori: A Matter of Dispute, in J.-M. AUWERS – H.J. DE JONGE (eds.), The Biblical Canons (BETL, 163), Leuven, Leuven University Press – Peeters, 2003, 487-556. 90. Or as MARKSCHIES, Canon of the New Testament (n. 8), p. 192, concluded regarding Athanasius’s 39th festal letter: “it seems to me absolutely clear that not only did the simpler Christians and simpler priests of Egypt know nothing about the exact number of Biblical books, but also that the Alexandrine Bishop himself had ‘to investigate the matter (i.e., the canon) from the beginning …’ before he was able to enumerate the exact number of canonical books. Or, obversely, without serious research, a learned theologian and bishop of the metropolis of Alexandria was not able to tally the 39 books of the Old Testament and the 27 books of the New Testament”. 91. Also the writings of Didymus of Alexandria (310/13 – 398) prove important in this regard, see B.D. EHRMAN, The New Testament Canon of Didymus the Blind, in VC 37 (1983) 1-21. 92. Edition of the homily: John of Parallos, Vienna K 9831v, ed. A. VAN LANTSCHOOT, Fragments coptes d’une homélie de Jean de Parallos, in Miscellanea Giovanni Mercati I (Studi e Testi, 121), Vatican City, Biblioteca Apostolica Vaticana, 1946, 296-326. See also D. FRANKFURTER, Beyond Magic and Superstition, in V. BURRUS (ed.), Late Ancient Chris- tianity: A People’s History of Christianity, vol. 2, Minneapolis, MN, Fortress, 2005, 255- 284; 309-312, pp. 259, 271, 309; and ID., The Legacy of Jewish Apocalypses in Early Christianity: Regional Trajectories, in J.C. VANDERKAM – W. ADLER (eds.), The Jewish Apocalyptic Heritage in Early Christianity (Compendia Rerum Iudaicarum ad Novum Testamentum; Jewish Traditions in Early Christian Literature), Assen, Van Gorcum; Min- neapolis, MN, Fortress, 1996, 129-200, esp. pp. 198-200.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 263263 44/07/11/07/11 11:5611:56 264 ANNEMARIE LUIJENDIJK

VII. THE GOSPEL OF THOMAS AT OXYRHYNCHUS

At the end of our journey, let us go back to Oxyrhynchus and consider what we can deduct about the readership of the Gospel of Thomas in that city. Although two of the three copies are written on rolls, a format associated with Jewish texts, it is not very likely that Jews owned these manuscripts of the Gospel of Thomas. Jews were familiar with Christian gospels, that much we know93, but we have no evidence for Jewish reading of this particular gospel. Furthermore, following the devastation of Jewish communities in Egypt after the Jewish Revolt in 115-117 C.E., evidence for Jews at Oxyrhynchus surfaces only from the end of the third century, when the synagogue, that is, the community, paid for the manu- mission of a female slave and her children (P.Oxy. IX 1205, dated April 14, 291 C.E.)94. Some ancient traditions associate the Gospel of Thomas with Manichae- ans. The fifth-century Decretum Gelasianum, for instance, informs about “a Gospel under the name of Thomas that the Manichaeans use” (evangelium nomine Thomae quibus Manichei utuntur)95. Several literary and documentary papyri attest to the presence of Manichaeans at Oxyrhynchus96. The mid-third-century Thomas papyri, however, copied during Mani’s

93. See P. SCHÄFER, Jesus in the Talmud, Princeton, NJ, Princeton University Press, 2007, esp. pp. 8-9 (8: “the Diatessaron and/or the Peshitta”) and pp. 124-129. 94. CPJ III 473. See R. SCHOLL, “Freilassung unter Freunden” im römischen Ägyp- ten, in H. BELLEN – H. HEINEN (eds), Fünfzig Jahre Forschungen zur antiken Sklaverei an der Mainzer Akademie 1950-2000. Miscellanea zum Jubiläum (Forschungen zur antiken Sklaverei, 35), Mainz, Akademie der Wissenschaften und der Literatur, 2001, 159-169, esp. pp. 159-160, 165, 168. On Jews at Oxyrhynchus, see E.J. EPP, The Jews and the Jew- ish Community in Oxyrhynchus: Socio-Religious Context for the New Testament Papyri, in T.J. KRAUS – T. NICKLAS (eds.), New Testament Manuscripts: Their Texts and Their World (TENTS, 2), Leiden, Brill, 2006, 13-52 (on this papyrus, pp. 34-37) and A. KASHER, The Jewish Community of Oxyrhynchus in the Roman Period, in Journal of Jewish Studies 32 (1981) 151-157, pp. 153-157. 95. Text from ATTRIDGE, Appendix (n. 4), p. 107, nr. 11. BLATZ, Das koptische Thomas- evangelium (n. 4), p. 94, cautions that it remains unclear whether this refers to the Gospel of Thomas or the Infancy Gospel of Thomas: “Unklar ist hierbei, ob es sich um das (häretische) Thomasevangelium oder um das dem Thomas zugeschriebene Kindheits- evangelium handelt”. 96. For the literary fragments, see S.N.C. LIEU, Manichaeism in Mesopotamia and the Roman East (RGRW, 118), Leiden, Brill, 1994, “5.1 Fragments in Syriac from Oxyrhyn- chus and others”, pp. 62-64. There are two Manichaean private letters from Oxyrhynchus: P.Oxy. XXXI 2603 (3rd-4th cent.) and P.Oxy. LXXIII 4965 (4th cent.). For the identification of P.Oxy. XXXI 2603 as Manichaean, see LIEU, Manichaeism in Mesopotamia (n. 96), p. 98, n. 316 and I. GARDNER – A. NOBBS – M. CHOAT, P.Harr. 107: Is This Another Greek Manichaean Letter?, in ZPE 131 (2000) 118-124, p. 118, n. 1. On the spread of Manichaeism, see also the groundbreaking analyses of P. BROWN, The Diffusion of Man- ichaeism in the Roman Empire, in JRS 59 (1969) 92-103.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 264264 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 265

lifetime (216-276), predate, if even by a short time, the arrival of the Manichaean movement in Egypt97. Thus these papyri of the Gospel of Thomas probably did not belong to Manichaeans. That leaves Christians at Oxyrhynchus as the likely owners of these papyri with the Gospel of Thomas. In their publication, Grenfell and Hunt wondered whether these sayings “circulated among a particular sect”98. The available evidence, however, does not allow us to distinguish separate Chris- tian groups99. I concur with Hurtado: “there is no particular reason to link these manuscripts with some distinctive circle of Christians. So far as we know, these artefacts were copied and used among the same Christian circles in which the other texts found in Oxyrhynchus and other ancient Christian sites functioned”100. Indeed, other papyrus finds testify to the diverse reading practices of Christians in this city, where besides the Gospel of Thomas, second- and third-century copies of the Gospels of Matthew, John, Mary and Peter101, among many other early Christian texts have been found. In addition to literary manuscripts, we have contemporaneous documentary evidence for Christians at Oxyrhynchus. As I have shown elsewhere, a small group of documents introduce us to a man called Sotas. The earliest known bishop from Oxyrhynchus in the third quarter of the third century102, Sotas presided over what appears to have been a fairly

97. See, for instance, G.G. STROUMSA, The Manichaean Challenge to Egyptian Chris- tianity, in B.A. PEARSON – J.E. GOEHRING (eds.), The Roots of Egyptian Christianity (Studies in Antiquity and Christianity), Philadelphia, PA, Fortress, 307-319, p. 310: “Although we know that Adda and Patteg, the Manichaean missionaries, reached Egypt much before the end of the third century, we cannot be quite specific about the dates and ways”. See also I. GARDNER – S.N.C. LIEU (eds.), Manichaean Texts from the Roman Empire, Cambridge, Cambridge University Press, 2004, esp. pp. 114-124. 98. GRENFELL – HUNT, P.Oxy. IV 654 (n. 4), p. 11. 99. See LUIJENDIJK, Greetings in the Lord (n. 7), p. 228: “Among the people whom we have investigated, little or nothing suggests divergent theological orientations. From the standpoint of papyrological investigation, theological inner-Christian divergences remain muted and indistinct”. 100. HURTADO, Greek Fragments (n. 5), p. 31. 101. Gospel of Matthew: P.Oxy. XXXIV 2683 + LXIV 4405; P.Oxy. LXIV 4403; P.Oxy. LXIV 4401; P.Oxy. I 2; P.Oxy. XXIV 2384 and P.Oxy. LXIV 4402. Gospel of John: P.Ryl.Gr. III 457 (provenance perhaps Oxyrhynchus); P.Oxy. X 1228; P.Oxy. II 208 + XV 1781; P.Oxy. LXV 4445, 4446, 4447, 4448; P.Oxy. VI 847; P.Oxy. XV 1780; P.Oxy. LXXI 4805; P.Oxy. XIII 1596. Gospel of Mary: P.Ryl.Gr. III 463 and P.Oxy. L 3525. Gospel of Peter: P.Oxy. XLI 2949. The Gospel of Mark is strikingly absent. 102. For the dating, see LUIJENDIJK, Greetings in the Lord (n. 7), p. 94 and p. 124. An unpublished Ethiopic text confirms my dating, informing that Sotas was chosen as new bishop of Oxyrhynchus under Alexandrian bishop Maximus (in office 264-282), as Alberto Camplani, one of the editors of this text, kindly informed me by email (June 19 and 20, 2010). On this manuscript, see A. CAMPLANI, Il patriarcato di Alessandria nella tarda antichità: L’identità del patriarcato alessandrino, tra storia e rappresentazione storiogra- fica, in Adamantius 12 (2006) 8-70.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 265265 44/07/11/07/11 11:5611:56 266 ANNEMARIE LUIJENDIJK

sizeable Christian community103. He provided Christian education that drew people from other communities, such as a certain Anos, visiting from Herakleopolis (P.Oxy. XXXVI 2785)104. In Sotas’s circle we also catch glimpses of the production of manuscripts105. Would Sotas read the Gospel of Thomas with students like Anos? Would he apply the Gospel of Thomas in his sermons, as Origen did? We do not know. As with most literary papyri, classical or Christian, we lack evidence that associates the papyri firmly with individuals known from documents106. Nevertheless, these documentary papyri provide an important backdrop for the reading of the literary fragments. We now return to our initial question: did Christians read the Gospel of Thomas liturgically or privately? Did they ever consider the Gospel of Thomas sacred scripture? The material evidence of the three Thomas papyri, read in conjunction with other papyri and Christian literary writings as the homilies of Origen, offered important insights about the reception of the Gospel of Thomas in the third century. The three Thoma sine Oxy- rhynchus papyri indicate a strong interest in that gospel at Oxyrhynchus in the mid-third century107. Looking closely at the manuscripts in their context, I established that the Gospel of Thomas in Oxyrhynchus was studied in private settings and also may have been recited in worship from a codex. Contemporary biblical scholar Origen, hailing from Egypt, mentioned the Gospel of Thomas in his homilies, albeit with mixed appreciation, once disapprovingly by name, and twice quoting a saying found also in the Gospel of Thomas.

103. From Sotas’s letters, we know by name ten men from this community: Sotas himself, Demetrianus (P.Oxy. XII 1492), Heron, Horion, Philadelphus, Pekusis, Naarous, Leo (PSI IX 1041), Heracles (PSI III 208), and, perhaps, Diphilus (P.Alex. 29). Two others arrived as visitors: a woman called Taion and a man called Leon (P.Oxy. XXXVI 2785). See also my cautious remark about the size of the Oxyrhynchite Christian community in LUIJENDIJK, Greetings in the Lord (n. 7), p. 115, n. 129. 104. See ibid., pp. 121-123. 105. Ibid., pp. 144-151. 106. On this topic, see W. CLARYSSE, Literary Papyri in Documentary “Archives”, in E. VAN ’T DACK – P. VAN DESSEL – W. VAN GUCHT (eds.), Egypt and the Hellenistic World. Proceedings of the International Colloquium Leuven, 24–26 May 1982 (Studia Hellenistica, 27), Leuven, Peeters, 1983, 43-61. For an exception, see A.M. LUIJENDIJK, A New Testament Papyrus and Its Documentary Context: An Early Christian Writing Exer- cise from the Archive of Leonides (P.Oxy. II 209/P10), in JBL 129 (2010) 575-596. 107. HURTADO, Greek Fragments (n. 5), pp. 28-29, observes: “the multiple copies of GThom clearly indicate a certain level of interest in this text among Christians in the early third century. If, however, we compare the number of copies of all literary texts in iden- tifiable Christian manuscripts dated prior to 300 CE, we can put the three copies of GThom into some perspective. Of course, any such comparison rests on the assumption (by no means incontrovertible) that the extant number of copies of a text from a given period is some general reflection of its comparative popularity at that time”.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 266266 44/07/11/07/11 11:5611:56 READING THE GOSPEL OF THOMAS IN THE THIRD CENTURY 267

Finally, my findings have larger implications for early Christian his- toriography. If we want to avoid writing about early Christianity from the hindsight-perspective of the Christian canon, a fourth-century his- torical construct, I have made clear that we should take the material evidence of these papyrus shreds seriously. These manuscripts in all their fragmentary messiness put the struggles of church fathers to enforce their preferred reading practices and theological agenda in stark relief. What I found was that in this time and place, the situation on the ground regarding Christian sacred scriptures is still open and fluid. For the project of rewriting the history of early Christianity, these glimpses into local religion suggest that contrary to elite ecclesiastical expressions, the reading practices of Christians were diverse and contrary to – or better, explanatory of – church leaders’ strong rhetoric108.

Department of Religion AnneMarie LUIJENDIJK Princeton University Princeton, NJ 08544 USA [email protected]

108. My thanks go to L.S. NASRALLAH for her helpful comments. I also presented different parts of this piece at the Center for the Study of Early Christianity at the Catholic University in Washington, D.C., at Princeton Theological Seminary in Princeton, N.J., and at the Mount Menoikeion Seminar in the Prodomos Monastery in Serres, Greece, and thank the audiences for their engaging questions and feedback.

994185_Clivaz_BETL242_11.indd4185_Clivaz_BETL242_11.indd 267267 44/07/11/07/11 11:5611:56