“Literate Media in Early Christian Groups: The Creation of a Christian Book Culture.” JECS 22, no. 1 (2014): 21–59.

Literate Media in Early Christ Groups: The Creation of a Christian Book Culture John S. Kloppenborg Journal of Early Christian Studies, Volume 22, Number 1, Spring 2014, pp. 21-59 (Article) Published by The Johns Hopkins University Press DOI: 10.1353/earl.2014.0004 For additional information about this article http://muse.jhu.edu/journals/earl/summary/v022/22.1.kloppenborg.html Access provided by University of Toronto Library (16 Mar 2014 15:38 GMT) Literate Media in Early Christ Groups: The Creation of a Christian Book Culture JOHN S. KLOPPENBORG The full emergence of Christian book culture in the fourth century was anticipated and prepared for by a series of developments in the second century: by presenting (anachronistically) the early heroes of the Jesus movement as skilled literate communicators; by the emergence of “textual communities” formed around the reading, study, and preservation of books; by valorizing literate culture through the embedding of allusions to books and book culture in Christian documents; and in the very construction of early Christian scriptural documents so as to render them more easily accessible to sub-elite readers. The author of Acts, writing probably early in the second century c.e.,1 represents the proponents of the earliest Jesus movement as ἄνθρωποι ἀγράμματοι καὶ ἰδιῶται, “illiterate and unskilled persons” (Acts 4.13). Luke’s use of ἀγράμματοι is undoubtedly apt: literacy levels in the ancient Mediterranean were generally low and there is little reason to suppose that Jesus’ first followers were very high on the literacy scales. His statement, This paper uses the standard system for designating New Testament papyri (Ƿ1, Ƿ2, Ƿ3. . . ; see Barbara Aland, et al., Novum testamentum graece, 28. Aufl. [Stuttgart: Deutsche Bibelstiftung, 2012]). Greek papyri are cited according to standard abbreviations. For abbreviations and publication details, see http://library.duke.edu /rubenstein/scriptorium/papyrus/texts/clist_papyri.html. 1. On the re-dating of Acts to the early second century, see Joseph B. Tyson, Marcion and Luke-Acts (Columbia, SC: University of South Carolina Press, 2006); Tyson, “The Implications of a Late Date for Acts,” Forum 1 third series 1 (2007): 41–63; Richard I. Pervo, “Dating Acts,” Forum 5:1 (2002): 53–72; Pervo, Dating Acts: Between the Evangelist and the Apologists (Santa Rosa, CA: Polebridge Press, 2006); Pervo, Acts, Hermeneia (Minneapolis, MN: Fortress Press, 2009), 5–7; Matthias Klinghardt, “The Marcionite Gospel and the Synoptic Problem: A New Suggestion,” NT 50 (2008): 1–27. Journal of Early Christian Studies 22:1, 21–59 © 2014 The Johns Hopkins University Press 22 JOURNAL OF EARLY CHRISTIAN STUDIES however, is not disparaging but ironic.2 His point is how the earliest followers of Jesus were able to craft and deliver a bold and compelling speech in the presence of elite and educated members of the Sanhedrin, despite the fact that they were illiterate amateurs. Luke’s explanation of their success and of the astonishment they provoked in their interrogators is that they were “filled with the Holy Spirit.” A very different picture emerges, both in literary representations— including Luke-Acts—and in the material record of the Jesus movement of the next three centuries. This picture suggests that literate media were prized and many of the heroes of the Jesus movement are remembered anachronistically as skilled literate communicators—not merely adept at speaking but skilled interpreters and performers of written texts. By the late second century, at least in some circles, Christian identity was connected with the possession of books, and various material evidence suggest that reading practices were developed and textual communities formed, in spite of what appears to have been the relatively low educational level of the majority of Christ believers. By the fourth century, a “media revolution”—to use Doron Mendels’s term—was in full swing.3 Christianity had become publicly visible in literate media and writers such as Eusebius exploited those media to enhance further the profile of Christians. Eusebius himself, a notorious book collector, mobilized his extensive library to create comprehensive literary accounts of religious and political history that placed at its center the emergence of Christianity under Augustus and its final triumph under Constantine. The fourth century also saw the creation of the Christian book par excellence, the large uncial biblical codices, Vaticanus and Sinaiticus, impossible prior to the fourth century due to the fact that book binding technology was hitherto incapable of binding in a single volume the more than 350 folios required for the entire Bible. As Mendels points out, Christians had achieved visibility through the written word—publicly accessible imperial edicts that aimed at the suppression of the movement 4 and especially through the edicts of toleration 2. See Thomas J. Kraus, “‘Uneducated,’ ‘Ignorant’ or Even ‘Illiterate’? Aspects and Background for an Understanding of ΑΓΡΑΜΜΑΤΟΙ (and ΙΔΙΩΤΑΙ) in Acts 4.13,” NTS 45 (1999): 434–49. 3. Doron Mendels, The Media Revolution of Early Christianity: An Essay on Eusebius’ Ecclesiastical History (Grand Rapids, MI: Eerdmans, 1999). 4. E.g., Stephen Mitchell, “Maximinus and the Christians in a.d. 312: A New Latin Inscription,” JRS 78 (1988): 105–24. Mitchell’s translation runs: “. . . let them take delight through the peace that has finally been allowed to them. And may those who, after being freed from those blind and wandering(?) by-ways, have returned to KLOPPENBORG / CHRISTIAN BOOK CULTURE 23 promulgated by Constantine. Eusebius was quick to record these edicts, both hostile and favorable, in his history, according them a special prominence. He quoted verbatim the edict of Gallienus to the bishops published in the wake of the Decian persecutions—an edict that expressly authorized the use of his rescript in order to prevent further persecution (HE 7.13). At 8.17 he again quotes an edict of Maximinus verbatim that expressly named the Christianoi (but no other group) as deviants, but conceded to Christians the right to exist and to build houses of assembly. Perhaps anticipating the modern aphorism that there is no such thing as bad publicity, Eusebius reproduced, again verbatim, Maximinus’s subsequent rescript, which ordered that those who persist in their accursed folly . . . be separated and driven far away from your city and neighbourhood, even as you requested; so that in accordance with your praiseworthy zeal in this respect, your city may be separated from all pollution and impiety, and, following its natural desire, may respond with due reference to the worship of the immortal gods (HE 9.7.12).5 In Book 10 Eusebius quoted at great length a series of five edicts and rescripts from Constantine: one granting toleration to the Christianoi (10.5.2–14); a second decree ordering the restoration of property to the Christians (10.5.15–17); a third ordering a synod (10.5.18–24); and two further rescripts granting funds to churches (10.6) and exempting clergy a right and goodly frame of mind, rejoice most of all, and, as though preserved from a sudden tempest or snatched from a grave illness, let them henceforward feel a more pleasant enjoyment of life. But as for those who have persisted in the abominable cult, let them be separated, just as you ask, far from your city and territory, and be removed, whereby, in accord with the praiseworthy zeal of your petition, your city, separated from the stain of every impiety, may respond, as it has been accustomed, to the sacred rites of the immortal gods with the worship which is owed to them. Moreover, so that you may know the degree to which your petition has been gratifying to us, behold, without any decree or any prayers on your part, with spontaneous accord according to our just and benevolent spirit, we grant permission to your devotion to request, in return for your religious resolution of that sort, whatsoever bounty you want. And may you do and request this now in the knowledge that you will obtain without any postponement something which, when granted to your city for all time, may as much bear witness to our own religious piety towards the immortal gods as it may show to your sons and grandsons that you have achieved rewards worthy of your traditions from our clemency. Fare well. Issued on 6 April in the second consulship of the emperors Constantine and Licinius (a.d. 312) at Sardis to the people of Colbassa.” 5. Eusèbe de Césarée, Histoire ecclésiastique, ed. Gustav Bardy, SC 31, 41, 55, 73, 4 vols. (Paris: Les Éditions du Cerf, 1952–58), 55:55; translation: Kirsopp Lake and J. E. L. Oulton, Eusebius: Ecclesiastical History, LCL, 2 vols. (London: Heinemann; Cambridge, MA: Harvard University Press, 1926–32), 2:349. 24 JOURNAL OF EARLY CHRISTIAN STUDIES from public liturgies (10.7). Mendels argues that Eusebius’s detailed reporting of imperial edicts is not simply to enliven his narrative, which culminated in Constantine’s edict of toleration, but because it so clearly demonstrate[s] the publicity victory of Christianity over all other religions in the marketplace. No religion but Christianity received such publicity and exposure from the imperial edicts, which were put up everywhere in the public sphere (and not just in one local temple or another). Even when toleration is mentioned . . . not one of the pagan religions is referred to by name. In other words, by quoting these edicts Eusebius wished to show that Christianity was the real winner because it received enormous publicity.6 A second engagement with literate media is seen in Eusebius’s use of codex technology. This allowed him to do what no one else could have done before him: to create a synoptic account of world history. In the second part of the Chronicon, Eusebius employed new codex technology in an innovative fashion: he divided each left and right page in a codex into multiple vertical columns, using these columns to align parallel chronologies and to allow, for the first time, careful synoptic comparison.7 As T. D. Barnes has plausibly suggested, Eusebius’s model was probably Origen’s Hexapla, which also made use of the codex form for a synoptic presentation of multiple versions of the Hebrew Bible. Origen’s massive Hexapla could not have been contained in a single or even two codices.8 But almost a century later Eusebius engaged in the ambitious project of publishing in codex form a collation of nineteen different state chronologies, exposing synchronisms and, just as importantly, differences and contradictions. The result was in the words of Williams and Grafton, “a comprehensive political, religious, and cultural history of the ancient world, one that served until the sixteenth century as the richest single source of information for anyone interested in the history of human culture.”9 A third and dramatic engagement with book culture: in contrast to the earlier Christian papyri that will be discussed below, the great fourth6. Mendels, Media Revolution, 237. 7. For an image, see the ninth-century manuscript, Oxford, Merton College, MS 315 online at: http://image.ox.ac.uk/show?collection=merton&manuscript=ms315. 8. T. D. Barnes, Constantine and Eusebius (Cambridge, MA: Harvard University Press, 1981), 120. Anthony Grafton and Megan H. Williams, Christianity and the Transformation of the Book: Origen, Eusebius, and the Library of Caesarea (Cambridge, MA: Harvard University Press, 2006), 104–5 point out that the Hexapla could never have been presented a single volume given its size: it would have filled 40 codices of 400 leaves (800 pages). 9. Grafton and Williams, Christianity and the Transformation of the Book, 137–40. KLOPPENBORG / CHRISTIAN BOOK CULTURE 25 century uncial codices, Vaticanus and Sinaiticus, have all the characteristics of the elite book: written scripta continua in a careful and consistent bookhand, with no ligatures or documentary characteristics, 42–48 lines per page, with ample margins, these codices exemplify what William Johnson has called an “egregiously elite product intended in its stark beauty and difficulty of access to instantiate what it is to be educated.”10 Skeat has argued that not only are the two manuscripts related, but that they are both from Caesarea and that the two are what remain of Eusebius’s compliance with Constantine’s request for fifty complete copies of the Bible.11 Whether or not Skeat’s argument is sustained, there is no doubt that a new niveau in Christian book production had been achieved, whether in Caesarea or Alexandria or elsewhere.12 Eusebius’s “bookishness” likely influenced him to retroject aspects of high literary culture into the activities of his predecessors. Kim HainesEitzen observes, for example, that Eusebius describes the “catechetical school” (τὸ τοῦ κατηχήσεως διδασκαλείον) in Alexandria at the time of Clement and Origen, supposing it to have a full curriculum of grammar, philosophical studies, and culminating in τὰ θεῖα παιδεύματα (HE 6.3.3).13 If this were an accurate representation of early third-century Alexandria one might also reasonably expect the existence of professional scriptoria for the copying of manuscripts. Yet as Haines-Eitzen notes, the clearest example of a NT codex produced in Alexandria, P.Bodmer 14–15 (Ƿ 75),14 displays none of the signs of having come from a professional scriptorium: 10. William A. Johnson, Readers and Reading Culture in the High Roman Empire: A Study of Elite Communities, Classical Culture and Society (New York: Oxford University Press, 2010), 21. 11. Theodore C. Skeat, “The Codex Sinaiticus, the Codex Vaticanus, and Constantine,” JTS n.s. 50 (1999): 583–625 (see Eusebius, Vita Constantini, 436–37). Bruce M. Metzger, The Text of the New Testament: Its Transmission, Corruption, and Restoration, 2nd ed. (New York: Oxford University Press, 1968), 7–8 doubts the Caesarean provenance of Sinaiticus and Vaticanus and suggests Alexandria, but allows nevertheless that they “are doubtless like those which Constantine ordered Eusebius to have copied.” 12. See the discussion in Grafton and Williams, Christianity and the Transformation of the Book, 178–232. 13. Eusebius, HE 6.3.3 (SC 41:87). Kim Haines-Eitzen, “Imagining the Alexandrian Library and a ‘Bookish’ Christianity,” in Reading New Testament Papyri in Context: Lire les papyrus du nouveau testament dans leus context, ed. Claire Clivaz and Jean Zumstein, Bibliotheca Ephemeridum theologicarum lovaniensium 242 (Leuven: Peeters, 2011), 214. 14. For the publication of the papyrus, see Victor Martin and Rodolphe Kasser, eds., Papyrus Bodmer XIV–XV: Evangiles de Luc et Jean (Cologny-Genève: Bibliotheca Bodmeriana, 1961). 26 JOURNAL OF EARLY CHRISTIAN STUDIES a uniform and calligraphic bookhand used for literary copies, corrections made by secondary scribes or an editor according to other exemplars, stichometrical markings used to calculate the pay of the scribe, various spelling errors due to dictation, and critical notes or markings. Moreover, the papyri that best preserve what appears to have been the earliest form of the New Testament text in Alexandria show virtually no recensional or editorial work at all—precisely what would need to be demonstrated for claims about a formal scriptorium and, even more, for claims about a bookish or scholarly Christianity in Alexandria.15 As will be noted more below, the materiality of early Christian manuscripts confirms the supposition that in the second and third centuries, while some Christian groups were in the process of developing “bookish” practices, there are strong signs that their manuscripts were prepared for non-professional readers. It would require imperial patronage, advances in book binding technology, and access to professional scriptoria to reach the level reflected in Eusebius’s Caesarea of the fourth century.16 ILLITERACY AND A LITERATE CULTURE Although the earliest partisans of the Jesus movement in Palestine and indeed the majority of the Mediterranean population even in Eusebius’s day were ἄνθρωποι ἀγράμματοι, the ancient Mediterranean was a literate culture.17 Estimates of literacy levels range from three to ten per15. Haines-Eitzen, “Alexandrian Library,” 216. In fact there are 116 corrections to Ƿ75 but, as James Royse notes, 111 of these were made by the original scribe himself and only five by later hands. See James R. Royse, Scribal Habits in Early Greek New Testament Papyri, New Testament Tools, Studies, and Documents 36 (Leiden: Brill, 2008), 625–56. In 15 instances there is the possibility of a second exemplar, but Royse considers it simpler to suppose that these were also correction to the scribe’s Vorlage (642). Three corrections were made by a second scribe, and two by a third scribe at some later date. 16. Evidence of trade in books in fourth-century Oxyrhynchus is found in P.Oxy. 63:4365, a letter addressed to a woman, asking her to lend the writer a copy of Esdras (probably 4 Ezra) in exchange for a copy of “the little Genesis” (either a miniature codex of Genesis, or the book of Jubilees). On this see Lincoln Blumell, Lettered Christians: Christians, Letters, and Late Antique Oxyrhynchus, New Testament Tools, Studies, and Documents 39 (Leiden and Boston: Brill, 2012), 169–70; Annemarie Luijendijk, “Greetings in the Lord”: Early Christians and the Oxyrhynchus Papyri, Harvard Theological Studies 60 (Cambridge, MA: Harvard University Press, 2008), 70–74. Blumell (Lettered Christians, 173–78) also collects instances of the trade in books among non-Christians at Oxyrhynchus. 17. Michael MacDonald (“Literacy in an Oral Environment,” in Writing and Ancient Near Eastern Society: Papers in Honour of Alan R. Millard, eds. P. Bienkowski, C. Mee, and E. Slater, Library of Hebrew Bible/Old Testament Studies [New York and KLOPPENBORG / CHRISTIAN BOOK CULTURE 27 cent.18 Despite low levels of literacy, however, the majority of the population was thoroughly familiar with literate modes of communication and data storage, as Alan Bowman has observed. Populations routinely encountered inscriptions, public notices, books and book sellers, libraries and other literate modes of communication, and there was a steady need to interact with bureaucracies on issues of tax declarations, census records, leases, loans, marriage and divorce documents, wills, and so forth.19 Roger Bagnall’s survey of Greek and Demotic documentary papyri underscores this fact: in the Ptolemaic period, petitions, receipts, and contracts—the kinds of documents generated by sub-elites in their financial and legal interactions—account for more than one-half of the surviving London: T & T Clark, 2005], 50) defines a literate society in this way: “I would define a ‘literate society’ as one in which reading and writing have become essential to its functioning, either throughout the society (as in the modern West) or in certain vital aspects, such as the bureaucracy, economic and commercial activities, or religious life. Thus, in this sense, a society can be literate, because it uses the written word in some of its vital functions, even when the vast majority of its members cannot read or write, as was the case, for instance, in early mediaeval Europe or Mycenaean Greece, where literacy was more or less confined to a clerical or scribal class.” 18. William V. Harris, Ancient Literacy (Cambridge, MA: Harvard University Press, 1989) estimates 10–15%. Lower estimates are given by Meir Bar-Ilan, “Illiteracy in the Land of Israel in the First Centuries c.e.,” in Essays in the Social Scientific Study of Judaism and Jewish Society: Volume II, ed. Simcha Fishbane and Stuart Schoenfeld (Hoboken, NJ: Ktav, 1992), 46–61; Catherine Hezser, Jewish Literacy in Roman Palestine, Texte und Studien zum antiken Judentum 81 (Tübingen: Mohr Siebeck, 2001). Hezser agrees with Bar-Ilan’s estimate of 3% as plausible. “It is quite reasonable to assume that the Jewish literacy rate was well below the 10–15 percent . . . which Harris has estimated for Roman society.” “If by ‘literacy’ we mean the ability to read a few words and sentences and to write one’s own signature only, then Jews probably came close to the Roman average rate” (49). 19. Alan K. Bowman, “Literacy in the Roman Empire: Mass and Mode,” in Literacy in the Roman World, ed. J. Humphrey, Journal of Roman Archaeology Supplementary Series 3 (Ann Arbor, MI: Journal of Roman Archaeology, 1991), 122: “The argument is, then, that a large proportion of the 80%+ illiterate population was thoroughly familiar with literate modes. Egypt, of course, may be exceptional and the argument that it was accustomed to a higher level of bureaucracy than other parts of the ancient world is duly put forward. But so may anything or everything else; regional differences need to be and are weighed carefully. But this is hardly compatible with the belief that literacy in Latin and Greek is in some sense a single phenomenon, sustained by a continuous cultural tradition, and that a line can be drawn between the literate and the illiterate population.” The Babatha archive provides a salient example of an illiterate person in routine contact with literary technologies. See Naphthali Lewis, The Documents from the Bar Kokhba Period in the Cave of Letters: Greek Papyri, Judean Desert Studies 2 (Jerusalem: Israel Exploration Society and The Hebrew University of Jerusalem, 1989). 28 JOURNAL OF EARLY CHRISTIAN STUDIES papyri, and this figure excludes the summaries of thousands of contracts preserved in village and nome archives.20 Although papyrological data is subject to the vicissitudes of archaeological discovery, there is no reason to believe that the profile that emerges from a few village archives in the Fayum or at Oxyrhynchus should not generalized to Ptolemaic Egypt as a whole. It was surely not the case that the inhabitants of a few Fayumic villages were the only ones to rely on the literate storage of their petitions, receipts, and contracts. Nor is there reason to think that Egypt in the Roman period suffered any loss of engagement with literate media nor indeed that other parts of the Mediterranean, where we do not have a papyrus record comparable to that found in Egypt, were quantitatively different in their reliance of data storage in the form of written texts. The level of illiteracy remained high, but few persons in the empire were oblivious to the reality and ubiquity of literate communication. Literate technologies were widespread among a population that was largely illiterate.21 It is not simply that large sectors of the population knew about written communication. Recent studies of literacy in the Roman world suggest that literacies, although admittedly restricted numerically, were not segregated into discrete sectors by language, function, and social register.22 On the contrary, because financial records, even if often compiled by slaves, ex-slaves, and free institores, have to be potentially auditable and comprehensible to landowners, this is not really scribal literacy. Likewise, because these documents linked the richest men in the community with their slaves and agents, this is not an example of commercial literacy or craftsman’s literacy. 20. Roger S. Bagnall, Everyday Writing in the Graeco-Roman East, Sather Classical Lectures 69 (Berkeley, CA and Los Angeles: University of California Press, 2011), 34–38. Bagnall concludes: “it looks very much as if Hellenistic Egypt was a society in which contracts were in extremely wide use, both between individuals and the government and among individuals; both Greek and Egyptian languages were commonly used for this purpose” (38). 21. Similarly, Harry Y. Gamble, Books and Readers in the Early Church: A History of Early Christian Texts (New Haven, CT: Yale University Press, 1995), 6–9. “It may seem paradoxical to say both that Christianity placed a high value on texts and that most Christians were unable to read, but in the ancient world this was no contradiction. In Greco-Roman society, the illiterate had access to literacy in a variety of public settings” (8). 22. See Brian V. Street, Literacy in Theory and Practice, Cambridge Studies in Oral and Literate Culture 9 (Cambridge, UK and New York: Cambridge University Press, 1984), who is critical of Goody’s treatment of literacy as “autonomous” (Jack Goody and Ian Watt, “The Consequence of Literacy,” Comparative Studies in Society and History 5 [1963]: 304–45). KLOPPENBORG / CHRISTIAN BOOK CULTURE 29 Roman landowners had good reason not to permit the development of segregated literacies. . . . Literacy approached the status of a generalized skill that it has in our societies. As a result those who learned to read in the army might make use of the skill in commerce and the literati could read— and might be astonished by—religious tracts emerging from unfamiliar sources. . . . The power of this generalized literacy was most widely felt beyond the narrow realm of administration.23 This “joined-up” nature of literacy was precisely what ancient writers depended upon for the impact of their work: the ancient writer did not produce a work that would be read directly by large numbers, but a work that, given the dispersed nature of Roman literacy, could be accessed by many. VALORIZING READING So how do we imagine the valorization of literate media and the development of a Christian book culture to have occurred, especially since the majority of Christians were presumably illiterate? This can be parsed into several discrete but mutually reinforcing factors. In the late second century, the book (or scroll) itself began to be treated as a symbol of Christian identity, a development that culminated in fourth-century tales in which the mere possession of a book was tantamount to a Christian confession. Second-century authors also displayed a tendency to represent the principal agents of the early Jesus movement as skilled literate communicators and to embed in their writings the values of a literate culture. During the same period, there is evidence of the creation of what Brian Stock called “textual communities” or William Johnson, “reading communities,” which instantiated the value of reading and text-oriented discussion, even amid predominantly illiterate audiences. And corresponding to the valorization of literate media, we can also observe in the construction of early Christian papyri efforts to produce texts that might be accessible to a sub-elite lector, created by scribes of rather middling scribal accomplishment. While the elite bookroll no doubt continued to reign in elite forms of sociability, epitomizing the social world of elite reading communities, these Christian scribes and lectors also created sub-elite reading communities, in which their books could be recited and discussed. 23. Greg Woolf, “Literacy or Literacies in Rome?” in Ancient Literacies: The Culture of Reading in Greece and Rome, ed. William A. Johnson and H. N. Parker, Studies in Book and Print Culture (Oxford and New York: Oxford University Press, 2009), 52, 53. 30 JOURNAL OF EARLY CHRISTIAN STUDIES The Book as Symbol of Identity Long before Eusebius, Christianity had begun to privilege literate culture as intrinsic to its identity. Tertullian conceded that Mercury had been the first to invent letters, but added that letters were essential both for commerce “and for our devotion to God.”24 Tertullian of course was literate and hence his literate devotion involved the direct consultation of texts. But his comment that letters are necessary to devotion should not necessarily be taken as simply autobiographical. The book, whether or not any particular Christian could read it, was in the process of becoming a symbol of Christian identity. In July of 180 c.e. twelve Christians were brought before Saturninus the proconsul of Carthage, accused of anti-Roman mores, and asked to return to their senses. When it became clear that the accused would not recant their error, Saturninus asked, “What do you have in your case (in capsa vestra)?” Speratus, one of the accused, replied, “books and letters of Paul, a righteous man” (libri et epistulae Pauli viri iusti).25 The proconsul’s question seems at first view irrelevant: once it was clear that Speratus and his friends were guilty, it hardly mattered what they had in their satchels. But the question and its answer from a literary perspective function to underscore the truth of Speratus’s declaration, Christianus sum.26 The proof of Speratus’s confession is demonstrated by his possession of a book. Seventy years later in mid third-century Caesarea Maritima, Marinus, a legionary about to be elevated to the rank of centurion was exposed by an enemy as a Christian and hence, so it was alleged, unable to sacrifice to the emperors. Being given some time to consider his situation, Marinus was taken to a church by its bishop, who pointed to the legionary’s sword and “at the same time brought and placed before him the book of the divine gospels” (τὴν τῶν θείων εὐαγγελίων γραφήν, Eusebius, HE 7.15.4 24. Tertullian, De corona 8.1 (PL 2.87B): Primus literas Mercurius enarraverit; necessarias confitebor et commerciis rerum et nostris erga Deum studiis; “Mercury may have been the first to teach letters; I will acknowledge that they are necessary both for the business transactions and for our devotion to God” (my translation). 25. Acts of the Scillitan Martyrs, 12. Text and translation: Herbert A. Musurillo, The Acts of the Christian Martyrs: Introduction, Texts and Translations, Oxford Early Christian Texts (Oxford: Clarendon Press, 1972), 80. 26. The connection between Christian identity and possession of Christian books is even clearer in the Acts of Euplus, which feature Euplus entering the council chamber with a copy of the pure gospels (τὰ ἄχραντα εὐαγγέλια ἐπιφερόμενος) and reading from them as proof of his identity as a Christian. See Musurillo, Acts of the Christian Martyrs, 310–19. KLOPPENBORG / CHRISTIAN BOOK CULTURE 31 [SC 41:190]).27 The bishop asked Marinus to choose between the two. The ending of the story is of course obvious. My point here is that whether or not the Scillitan martyrs or Marinus could read—and it is certainly unnecessary to suppose that they could—the book had come to symbolize and epitomize Christian identity and confession. A half a century later at the time of Diocletian, this equation was further strengthened in the practice of the confiscation and burning of Christian books; the book served as a surrogate for its putative author or authorizing source.28 Both the Roman persecutors and the Christian confessors understood that books and reading were at the core of Christian identity.29 The symbolic capital of the book was also highlighted in Christian art from a slightly later period: the traditio legis motif typically shows Christ flanked by Peter and Paul, handing either a scroll or a book to one of the disciples.30 This motif, especially common in sarcophagi from the midfourth century onward, quickly became a standard motif of Christian art. A particularly interesting example is a sarcophagus at the Louvre31 27. Eusebius, HE 7.15.4 (SC 41:190; trans. Lake and Oulton, LCL 2:173). 28. Daniel Sarefield, “The Symbolics of Book Burning: The Establishment of a Christian Ritual of Persecution,” in The Early Christian Book, ed. William E. Klingshirn and Linda Safran (Washington, DC: Catholic University of America Press, 2007), 159–75. Sarefield adduces the story related in Lucian’s Alexander 47, of Alexander’s hostility toward Epicurus and his public burning of a copy of the Kyriai Doxai as a surrogate for Epicurus himself. 29. See also the early fourth-century Gesta apud Zenophilum (CSEL 26:185–97), where prosecutors entered the houses of Christians to seize book (here called codices). Similarly, the Martyrdom of Agape, Irene and Chione (Musurillo, Acts of the Christian Martyrs, 279–93) from 303 c.e., where the governor declares Irene’s guilt on the basis of her possession of τοσαύτας διφθέρας καὶ βιβλία καὶ πινακίδας καὶ κωδικέλλους καὶ σελίδας γραφῶν τῶν ποτε γενομένων Χριστιανῶν, “so many parchments, books, tables, codices, and pages of writings of the former Christians” (Musurillo, 286; my translation, JSK). 30. E.g., the Junius Bassus sarcophagus in the Vatican Museum (ca. 359 c.e.) (Christ with a scroll); a fragment of a tomb relief, New York, Metropolitan Museum, 48.76.2, fourth quarter of the fourth century c.e. (Kurt Weitzmann, ed., Age of Spirituality: Late Antique and Early Christian Art, Third to Seventh Century [New York: Metropolitan Museum of Art, 1979], no. 502: Christ holding a scroll); the Sarcophagus of Stilicho in the Museo della civiltà romana a Roma (room 15) sculpted around 385 c.e.; a reliquary, end of the fourth century, Museo Archeologico di Firenze MBP:BA 71 (Christ with a scroll); a sarcophagus, Rome, fourth quarter of the fourth century c.e. (Jeffrey Spier, ed., Picturing the Bible: The Earliest Christian Art [New Haven, CT: Yale University Press, 2007], 242, cat. no. 64a; Christ with a scroll). 31. Sarcophagus (end of the fourth century c.e.), Louvre, Département des Antiquités grecques, étrusques et romaines, inv. 2296/MA 2958. Denon, Rez-de-Chaussée, Salle 27. 32 JOURNAL OF EARLY CHRISTIAN STUDIES from the late fourth century, depicting Christ with a codex (or tablet) in his hand, and the disciples with both rolls or codices. The transmission of Christian identity, doctrine, and teaching authority is symbolized in the handing over of a scroll or a book.32 Christian Heroes and Literate Culture Long before Eusebius’s depictions of earlier generations of Christians as participants in high literary culture we find evidence of a tendency to represent figures central to the Christ groups as engaged in literary culture. The second letter of pseudo-Paul to Timothy has “Paul” requesting that “Timothy” come and “bring the cloak that I left with Carpus at Troas, also the books, and especially the parchments” (καὶ τὰ βιβλία μάλιστα τὰς μεμβράνας) (2 Tim 4.13). This presented commentators who assumed that 2 Timothy was authentic with an immediate problem. Were the “parchments” in question are scrolls of the Torah or Prophets? And is it plausible to assume that Paul travelled with a library of books and parchments? T. C. Skeat, troubled by this implausibility, famously proposed that μάλιστα must be epexegetical (“books, I mean, notebooks”).33 2 Tim 4.13 has even been the basis of a claim to the authenticity of the letter. Thus, J. N. D. Kelly urged: Its [2 Tim 4.13] simple realism and naturalness are so striking that even critics who doubt the authenticity of the letter as a whole are often prepared to concede that it must come from the Apostle himself. It is extremely unlikely that an imitator in the ancient world would have thought of inventing banal details like these.34 32. The decoration of codices also underscored their symbolic value. John Lowden, “The Word Made Visible,” in The Early Christian Book, ed. William E. Klingshirn and Linda Safran (Washington, DC: Catholic University of America Press, 2007), 19 observes that the book in the form of a decorated codex was not simply meant to be read; it was meant to be seen. A very early example of this is the mid-fourth century, Codex II from Nag Hammadi, a codex of 176 pages in a single quire, was bound in leather and extensively decorated with incised lines, spirals, and ankh-like figures. See Lynda K. Ogden, “The Binding of Codex II,” in Nag Hammadi Codex II, 2–7: Together with XIII, 2*, Brit. Lib. Or.4926(1), and P.Oxy. 1, ed. Bentley Layton, NHS 20–21 (Leiden and New York: Brill, 1989), 1:19–25. According to Layton, the codex should be dated “by its association with other Nag Hammadi manuscripts, to the fourth century, and by the handwriting, to the first half of the same century” (at 4). NHC II is not typical of fourth and early fifth century codices, which often lack decoration entirely. 33. Theodore C. Skeat, “Especially the Parchments: A Note on 2 Timothy 4:13,” JTS n.s. 30 (1979): 173–77. 34. J. N. D. Kelly, A Commentary on the Pastoral Epistles; I Timothy, II Timothy, Titus, Black’s New Testament Commentaries (London: A. & C. Black, 1976), KLOPPENBORG / CHRISTIAN BOOK CULTURE 33 Pseudepigraphical compositions, however, were perfectly able to have such details. In Ps-Crates Letter 30, “Crates” returns a tunic (ἐξωμίς), typically worn by slaves and even some Cynics, to his wife Hipparchia, admonishing her that her efforts are better spent pursuing philosophy.35 This seemingly innocuous biographical detail in fact serves a higher rhetorical function to underscore the ideal Cynic disregard for material comfort and the supervening importance of philosophia. In 2 Timothy the mention of τὰ βιβλία μάλιστα τὰς μεμβράνας underscores the identity of Paul as a writer. The term μεμβράνα, which as a matter of fact appears for the first time in Greek literature here and is a loan word from Latin, normally refers to parchments for writing or notebooks, not finished books. It is thus beside the point to inquire whether the βιβλία were books of the Torah or whether the μεμβράνα were blank or filled. The point of the phrase in 2 Timothy is to advance the claim that Paul is a literate agent, who is able to use and read from books and to write his own notebooks. The significance of this should not be missed. Of course we know Paul as a letter writer who used a secretary. But in the undisputed letters, Paul refers to his acts of writing only in Gal 6.11 and 2 Cor 10.10. In the latter, he indicates that he is aware of unfavourable contrasts have been drawn between his letters, which seemed weighty, and an unimpressive personal presence. In the undisputed letters Paul does not construct himself primarily as a letter writer: he is an envoy of God and a builder who has assembled and maintained a network through his personal apostolic power. His letters are surrogates for his presence, which in 2 Cor 13.10 he threatens will be harsh if his does not find what he wants when he visits. 2 Tim 4.13, by contrast, prepares for a “Paul” who is known only through his letters, not through his personal presence, that is, a Paul of the late first or early second century. 2 Tim 4.13 allows the audience to know “Paul” through his notebooks and parchments. In the logic of pseudepigraphy, the parchments mentioned in 2 Tim 4.13 no doubt include the second letter 215; similarly, Ceslaus Spicq, Les epîtres pastorals, 2nd. ed., Etudes bibliques (Paris: Lecoffre, 1947), 392: “De tous les passages authentiques, II Tim iv, 13 est sûrement le moins apte à avoir été inventé,” citing J. S. Stevenson, Expository Times 34 (1923): 525; Walter Lock, The Pastoral Epistles (I & II Timothy and Titus), International Critical Commentaries (Edinburgh: T. & T. Clark, 1924), xxvi: “The Pauline authorship . . . in 1 and 2 Timothy is implied in the constant personal references either to St. Paul’s own life . . . or to his relations with Timothy.” 35. Text and Translation: Abraham J. Malherbe, ed. The Cynic Epistles: A Study Edition, SBL Sources for Biblical Study 12 (Missoula, MT: Scholars Press, 1977), 80–81. 34 JOURNAL OF EARLY CHRISTIAN STUDIES of Timothy itself.36 It should be noted that another pseudepigraphal work of the same vintage, 2 Pet 3.15–16, likewise treats Paul as a writer, and indeed as one so difficult that the “unschooled” (ἀμαθεῖς) misinterpret.37 Another early second-century writing, Acts, famously does not mention Paul’s letters, but does represent Paul as routinely engaging in debate “from the Scriptures” (Acts 17.2, διελέξατο αὐτοῖς ἀπὸ τῶν γραφῶν) and describes Apollos as “powerful in the Scriptures” (Acts 18.24, δυνατὸς ὢν ἐν ταῖς γραφαῖς) and as able to construct plausible arguments from the Scriptures (Acts 18.28). Acts leaves Paul in Rome, meeting with the leaders of the Jewish groups there, persuading them “from both from the law of Moses and from the prophets” (ἀπό τε τοῦ νόμου Μωϋσέως καὶ τῶν προφητῶν, Act 28.23). The construction of Pauline identity—and by implication, an ideal Christian identity—as fluent with literary technologies and able to quote not only the Hebrew Scriptures but Epimenides and Aratus’s Phenomona (Acts 17.28) reflects, as Vernon Robbins has shown, Luke’s social location rather than Paul’s.38 Robbins also notes that “every major character [in Luke-Acts] quotes (often lengthy passages) verbatim from the Jewish Scriptures very soon after being introduced to the implied reader”39 This tendency, to connect Christian identity with literacy and the use of literate media, is of a piece with the Pastoral writer and 2 Peter’s representations of Paul as defined through his literary activity. The book and the letter mark Christian identity. This is, of course, in contrast with the “historical Paul.”40 My point here is not that the “historical Paul” does not refer to the Scriptures; of course he does (except in 1 Thessalonians). The point, rather, is that in the representations of Paul, he was progressively converted into someone whose activities are intrinsically connected to high literate culture. 36. On pseudepigraphical technique, see John W. Marshall, “‘I Left You in Crete’: Narrative Deception and Social Hierarchy in the Letter to Titus,” JBL 127 (2008): 781–803. 37. 2 Pet 3.15–16: ὁ ἀγαπητὸς ἡμῶν ἀδελφὸς Παῦλος κατὰ τὴν δοθεῖσαν αὐτῷ σοφίαν ἔγραψεν ὑμῖν, ὡς καὶ ἐν πάσαις ἐπιστολαῖς λαλῶν ἐν αὐταῖς περὶ τούτων, ἐν αἷς ἐστιν δυσνόητά τινα, ἃ οἱ ἀμαθεῖς καὶ ἀστήρικτοι στρεβλοῦσιν ὡς καὶ τὰς λοιπὰς γραφὰς πρὸς τὴν ἰδίαν αὐτῶν ἀπώλειαν. 38. Vernon K. Robbins, “The Social Location of the Implied Author of Luke-Acts,” in The Social World of Luke-Acts: Models for Interpretation, ed. Jerome H. Neyrey (Peabody, MA: Hendrickson, 1991), 305–32. 39. Robbins, “Social Location,” 325. 40. Ronald F. Hock, “Paul’s Tentmaking and the Problem of His Social Class,” JBL 97 (1978): 555–64. KLOPPENBORG / CHRISTIAN BOOK CULTURE 35 Other depictions of the earliest circle of Jesus-followers as writers also belong to the late first century or early second century c.e. The fourth gospel, the product of the late first or early second century,41 parachutes its authority figure, the anonymous “beloved disciple,” into the circle of Jesus’ followers and in the second ending to the gospel expressly claims him as a writer: οὗτός ἐστιν ὁ μαθητὴς ὁ μαρτυρῶν περὶ τούτων καὶ ὁ γράψας ταῦτα (21.24). Likewise, in the incipit of the Gospel of Thomas, Thomas is claimed as a scribe of the hidden words of Jesus (P.Oxy. 4:654.1–3 = Gos. Thom. 1), and the opening scene of the Apocryphon of James (1.8–10) has James compose a “secret book” in Hebrew and send it to his correspondent and then describes the Twelve as sitting together, remembering Jesus’ public and secret teachings, and writing them in book (2.8–15). In the logic of pseudepigraphy, teachings that were originally delivered orally are now put into writing and delivered to a readership a century or more removed. David Brakke observes the critical interpretive shift: Now Christians find revelation in the written teachings of the Savior, especially in the parables (7.1–10; 7.35–8.10), for they are the media by which the learned few can gain intellectual understanding (νοεῖν); by contrast, the unlearned “many” (1.22) remain at the level of sense perception (αἰσθάνεσθαι), mediated by Jesus’ plain speech. Far from representing a “gnostic” protest against written scriptures in favor of a continued oral culture, Ap. Jas. testifies to the development within Christianity of an intellectual elite devoted to a literary culture of study, exegesis, and enlightenment.42 Correspondingly, the earliest tradents of the Jesus tradition are represented as writers and as interpreters of the written word. An even more dramatic instance of the tendency to promote the heroes of the Jesus movement as literate is seen in Luke’s representation of Jesus. 41. The dating of John up to now has been kept to the late first century largely because of the dating of Ƿ52 (John Rylands University MS 457), which contains John 18.31–33, 37–38, to the first quarter of the second century. Originally Ƿ52 had been dated on the basis of an early dating of P.Egerton 2. In the meantime, however, the discovery of P.Köln 6:255, a part of the Egerton manuscript, and its dating to the last quarter of the second century (see below, n.106), now puts into question the dating of Ƿ52. The earliest papyri of John are now all late second century: Ƿ52, Ƿ90 (= P.Oxy. 50:3523), and Ƿ66 (= P.Bodmer 2). Given this late dating, and since Justin knows nothing of John, it is now possible to imagine a later dating of John’s gospel. On Ƿ52 see Andreas Schmidt, “Zwei Anmerkungen zu P. Ryl. III 457,” APF 35 (1989): 11–12 and Brent Nongbri, “The Use and Abuse of Ƿ52: Papyrological Pitfalls in the Dating of the Fourth Gospel,” HTR 98 (2005): 23–48. 42. David Brakke, “Parables and Plain Speech in the Fourth Gospel and the Apocryphon of James,” JECS 7 (1999): 216. 36 JOURNAL OF EARLY CHRISTIAN STUDIES The scene in the Nazareth synagogue in 4.16–30 features Jesus unfolding a βιβλίον and locating a particular text (εὗρεν τὸν τόπον, 4.17). As I will note presently, this is no mean accomplishment, given the fact that ancient books and bookrolls did not have tables of contents, running headers, or chapter divisions, and, in Greek at least, were written scripta continua. In Luke’s imagination, Jesus is not only able to read; he is able to locate a particular text without the assistance of a trained lector and to perform that text in public. In addition, as Roger Bagnall has argued, the verb Luke uses—ἀναπτύσσω—suggests that Luke has in mind a book, not a bookroll or a scroll.43 That is, Luke not only represents Jesus as able to navigate his way through a book and to read from it, but also to use a media technology that was only recently introduced towards the end of the first century c.e.44 The contrast with what might reasonably be assumed about the “historical Jesus” is obvious.45 Scholars such as John Meier believe that Jesus had not only “craftsman literacy” but that he was also likely to have been schooled (at home?) and given a “reading knowledge” of Hebrew Scriptures.46 As Craffert and Botha note, however, this neglects approaches to literacy as a culturally- and socially-embedded practice and not an 43. Roger S. Bagnall, “Jesus Reads a Book,” JTS n.s. 51 (2000): 577–88. Bagnall observes that while βιβλίον can refer to a scroll, codex or even a letter, “I have not found a single case in which ‘unroll’ must be the meaning [of ἀναπτύσσω]. . . . The verb was used for a variety of actions of opening, including a common figurative use to mean ‘explain’” (584–85). Bagnall points out that the faulty interpretation of ἀναπτύσσω as “unroll” also informs the text-critical decision to prefer ἀναπτύξας (‫ א‬Dc Δ Θ Ψ 0123 λ φ 28 157 1010 it vg Orlat Eus1/2) to ἀνοίξας (A B L W Ξ 33 579 892 1241 sa bo eth Eus1/2). 44. See Colin H. Roberts and Theodore C. Skeat, The Birth of the Codex (London: Published for the British Academy by the Oxford University Press, 1983), 28–29, 63; Theodore C. Skeat, “The Origin of the Christian Codex,” Zeitschrift für Papyrologie und Epigraphik 102 (1994): 263–68. 45. See Peter F. Craffert and Peter J. Botha, “Why Jesus Could Walk on the Sea but He Could Not Read and Write: Reflections on Historicity and Interpretation in Historical Jesus Research,” Neotestamentica 39 (2005): 5–38. 46. John P. Meier, A Marginal Jew: Rethinking the Historical Jesus. Volume I: The Roots of the Problem and the Person, Anchor Bible Reference Library (New York: Doubleday, 1991), 273, 276. For a more cautious view, see Paul Foster, “Educating Jesus: The Search for a Plausible Context,” Journal for the Study of the Historical Jesus 4 (2006): 7–33, who concludes that while there is no clear evidence to decide whether Jesus had “some level of functional literacy . . . the balance of probabilities appears to favor the contention that Jesus at least possessed basic reading ability” (32–33). Foster rightly distinguishes between the ability to read and the ability to write, but does not distinguish between the ability to read simple texts and the ability to perform written texts in public. KLOPPENBORG / CHRISTIAN BOOK CULTURE 37 autonomous cultural value.47 Simply put, literacy as a value is dependent upon factors of social standing, occupation, and one’s place in the socioeconomic system. As a craftsman, it is possible that Jesus had some degree of numeracy and some basic ability to read. The ability to navigate and read lengthy texts, especially read in public, is another matter entirely: With regard to the historical Jesus we can clearly say that as a Galilean peasant the most plausible presentation of him should be that he was illiterate—particularly when seen in terms of what we understand literacy to mean. As a Galilean peasant he was at best able to recognize a few letters (meaning numbers) and construe a few names and/or inscriptional signs. What we call reading is “reading with comprehension,” an activity limited to but a few people in antiquity. Even those skilled with texts and documents did not read the way we do, as the rabbis graphically illustrate.48 Luke’s representation of Jesus as a literary performer appears to belong not to early first-century c.e. Galilee, but to the development of a literate Christ cult in the second century. Embedding Textual Practices In addition to the heroes of the early Jesus movement being represented anachronistically as readers, some of the earliest compositions of the Jesus movement embed literary practices as values. One of the earliest products of the Jesus movement, the Sayings Gospel Q, although as a written document was no doubt a scribal product,49 47. Street, Literacy in Theory and Practice, 121–25; Harvey J. Graff, The Labyrinths of Literacy: Reflections on Literacy Past and Present (London: Falmer, 1987). 48. Craffert and Botha, “Why Jesus Could Walk on the Sea,” 31. 49. That Matthew and Luke have consulted Q as text rather than as a set of oral information is clear from the level of verbatim agreement and, more importantly, the sequential agreements between Matthew and Luke. On this, see John S. Kloppenborg, “Variation in the Reproduction of the Double Tradition and an Oral Q?” ETL 83 (2007): 49–79. That it was the product of a scribal culture, probably a lowlevel scribal culture, has been argued by John S. Kloppenborg, “Literary Convention, Self-Evidence, and the Social History of the Q People,” in Early Christianity, Q and Jesus, ed. Kloppenborg and Leif E. Vaage, Semeia 55 (Atlanta, GA: Scholars Press, 1991), 77–102, esp. 83, who observes the scribal values embedded in Q: “the Sayings Gospel places a premium upon both clarity of perception, especially when it comes to matters of guidance (Q 6.40, 41–42), and good speech, the characteristic mark of good thinking (Q 6.45). Guidance and moral example are also the subjects of the sayings on judging (Q 6.37–38), scandal (Q 17.1–2) and forgiveness (Q 17.3b–4). . . . The saying preserved in Q 12.2 on the revelation of things hidden is not only a wisdom saying; it reflects a characteristic sapiential interest in what is hidden as an object of research and conceives the process of disclosure of the order of things as grounded in the relationship between God and the world.” See further, William E. 38 JOURNAL OF EARLY CHRISTIAN STUDIES shows only a slight engagement with written media, and mostly at the final stages of its redaction. As I have suggested elsewhere, however Q was actually used in the composition of Matthew and Luke, it was perhaps intended to function as an aide-mémoire, as a resource for rhetorical emulation rather than as a “source” for replication.50 Nevertheless, in the final stages of Q’s editing, Q became more invested in literate technologies. The final redaction of Q represents Jesus as one who can debate with the devil by quoting Scripture with the formula γέγραπται (4.4, 8, 10; cf. 7.27).51 Moreover, Q 16.17, also from the final redaction of Q,52 describes the impossibility of a yûd or a serif falling from the Law (ἢ ἰῶτα ἓν ἢ μία[ν] κεραία[ν] τοῦ νόμου [πεσεῖν]).53 This imagines the Torah not merely as a body of law but as letters on a page that can (or cannot) be damaged or lost. Ironically, although Q might have been conceived as a resource for re-oralization, Matthew and Luke employed it as a text to be copied. Matthew and Luke’s reproduction of Q is rather wooden, taking over more than fifty percent of Q’s words. While this is a great boon to those Arnal, Jesus and the Village Scribes: Galilean Conflicts and the Setting of Q (Minneapolis, MN: Fortress Press, 2001) and Giovanni B. Bazzana, “Village Scribes Behind Q: The Social and Political Profile of the Sayings Gospel in Light of Documentary Papyri,” in Auf Fels oder Sand Gebaut? Die Q-Forschung: Rückblicke—Einblicke— Ausblicke, ed. Christoph Heil, Bibliotheca Ephemeridum theologicarum lovaniensium (Leuven: Peeters, 2013), forthcoming. 50. John S. Kloppenborg, “Emulation of the Jesus Tradition in James,” in Reading James with New Eyes, ed. Robert L. Webb and John S. Kloppenborg, Library of New Testament Studies 342 (London and New York: T. & T. Clark International, 2007), 121–50. 51. On the editing of Q, see John S. Kloppenborg, The Formation of Q: Trajectories in Ancient Wisdom Collections, Studies in Antiquity and Christianity (Philadelphia, PA: Fortress Press, 1987; 2nd ed. Harrisburg, PA: Trinity Press International, 1999). Although the stratigraphic model for the composition of Q is not universally embraced, there is a nearly universal consensus that Q 4.1–13 belongs to the final stage of Q’s editing. See the tabulation of opinion in John S. Kloppenborg, “The Sayings Gospel Q: Literary and Stratigraphic Problems,” in Symbols and Strata: Essays on the Sayings Gospel Q, ed. Risto Uro, Suomen Eksegeettisen Seuran Julkaisuja, Publications of the Finnish Exegetical Society 65 (Helsinki: Finnish Exegetical Society Göttingen: Vandenhoeck & Ruprecht, 1996), 1–66. 52. John S. Kloppenborg, “Nomos and Ethos in Q,” in Gospel Origins and Christian Beginnings: In Honor of James M. Robinson, eds. James E. Goehring, Jack T. Sanders, and Charles W. Hedrick (Sonoma, CA: Polebridge Press, 1990), 35–48. 53. Greek reconstruction and translation: James M. Robinson, Paul Hoffmann, and John S. Kloppenborg, eds., The Critical Edition of Q: A Synopsis, Including the Gospels of Matthew and Luke, Mark and Thomas, with English, German and French Translations of Q and Thomas, Hermeneia Supplements (Leuven: Peeters; Minneapolis, MN: Fortress Press, 2000), 468. KLOPPENBORG / CHRISTIAN BOOK CULTURE 39 who reconstruct Q, this style of text-reproduction is in marked contrast to the way in which other source texts were typically reproduced by authors such as Josephus or Diodorus Siculus, who typically engaged in generous paraphrase and rhetorical emulation. Matthew and Luke’s reproduction of Q is closer in nature to the kinds of copying seen in the transmission of technical texts and manuals.54 The point in the context of this paper is that Q, irrespective of how it was intended to be used, quickly became part of the literary deposit of the Jesus movement. Part of the tendency to valorize literary culture can be seen in the redaction of various documents of the early Jesus movement. Mark’s use of γέγραπται (1.2; 7.6; 9.12, 13; 11.17; 14.21, 27) also engages literary culture and Mark 13.14, ὁ ἀναγινώσκων νοείτω, clearly has a reader rather than an auditor in view. Matthew’s editing of Mark and Q makes this tendency even more obvious. Matthew constructs the ideal reader of his book as someone who, if she does not actually read, understands that books—in particular the Septuagint—contain the key to understanding the significance of Jesus. Matthew almost doubles the number of fulfillment formulae present in Mark.55 The embedding of textual practices into the documents of the early Jesus movement corresponds to what Brian Stock has described as a critical stage in the formation of community identity: historicizing the community by giving it a past through intertextuality.56 54. Kloppenborg, “Variation.” 55. Interestingly, Matthew does not show a strong tendency to take over γέγραπται: all of his occurrences except perhaps Matt 2.5 are from his sources: 4.4Q, 6Q, 7Q, 10Q; 11.10Q; 21.13Mk; 26.24Mk, 31Mk. His favored term is τὸ ῥηθὲν ὑπὸ κυρίου διὰ τοῦ προφήτου λέγοντος (1.22; 2.15; 2.17; 2.23; 4.14; 8.17; 12.17; 13.35; 21.4; 22.31; 24.15; 27.9), which occurs only in Matthew. The γέγραπται formula finds its closest analogue in Dan 9.13 (‫)כאשׂר כתוב בתורת משׁה‬, 1QS (‫ ;)כיא כן כתוב‬CDC 11.18 (‫)כי כן כתוב‬ and 2Q25 1.3 (‫( )]כי [כן כתוב‬Qumran texts: Florentino García Martínez, The Dead Sea Scrolls Study Edition [Leiden, New York and Köln: Brill, 1997–98]). Aaron M. Gale (Redefining Ancient Borders: The Jewish Scribal Framework of Matthew’s Gospel [London and New York: T. & T. Clark International, 2005], 110–61) argues on the basis of evidences of allusion, gematria, and a range of scribal arguments that Matthaean scribes “expected the community to be literate in Greek as well as Hebrew” (158). Unless the entire Matthaean group was scribal, this seems unlikely. But that Matthew constructs an ideal reader who is literate is a different issue. 56. Brian Stock, Listening for the Text: On the Uses of the Past, Parallax (Baltimore, MD: Johns Hopkins University Press, 1990), 25–26. Stock’s analysis of the Waldensians led him to posit three stages of community formation: (1) oral contact (“the jongleur sings an oral text; the preachers read a text of the gospels”); (2) an educative process, leading to the study, memorization, and preaching of vernacular texts; and (3) the historicizing of the community, though intertextuality, by associat- 40 JOURNAL OF EARLY CHRISTIAN STUDIES Reading Communities The shift towards a valorization of literate technologies has social consequences or, perhaps, is a reflection of social structures. William Johnson has proposed the useful model of a “reading community”—a social mechanism that validates and valorizes particular practices concerning texts and their interpretation.57 Reading communities are sustained by at least two factors. First, they are created and sustained through institutions that promote the notion that knowledge is socio-culturally important. In the modern world schools and universities do this, but also government, industry, the judiciary, media, and other institutions that rely on certain levels and types of literacy in order to function. Facility in reading and writing becomes socio-culturally important because these institutions construct a world in which communication depends on literate technologies and in which literate accomplishment is rewarded. In the ancient world, the value of literate education was promoted not only by the gymnasium but also by such institutions as public assemblies and the courts, where knowledge of Homer, Hesiod, Virgil, and other classic texts marked the “cultured” person and where such knowledge counted as high moral achievement. The values of literate culture were reinforced by such institutions as libraries and booksellers. Second, reading communities are contrived by the teacher who creates a disposition that a certain kind of knowledge is important—that Homer should be interesting.58 Johnson recalls: When reading ancient epic in translation (Gilgamesh, Iliad, Odyssey, Aeneid), I have been deeply impressed at the high and general level of enthusiasm, indeed excitement, that the students bring to this reading of ing the community with earlier antecedents, in the case of the Waldensians, the Life of St. Alexis and the Life of Anthony. David Brakke has applied Stock’s model to the Apocryphon of James in “Parables and Plain Speech.” 57. Johnson’s “reading communities” seems to be an adaptation of the “textual communities” proposed earlier by Brian Stock (Listening for the Text, 23, 149–58). “The minimal requirement was just one literate interpreter, the interpres, who understood a set of texts and was able to pass his message on verbally to others. By a process of absorption and reflection the behavioral norms of the group’s other members were eventually altered. The manner in which the individuals behaved toward each other and the manner in which the group looked upon those it considered to be outsiders were derived from the attitudes formed during the period of initiation and education. The unlettered and semilettered members thereby conceptualized a link between textuality, as the script for the enactment of behavioral norms, and rationality, as the alleged reasonableness of those norms” (23). 58. William A. Johnson, “Toward a Sociology of Reading in Classical Antiquity,” AJP 121 (2000): 604. KLOPPENBORG / CHRISTIAN BOOK CULTURE 41 these texts. At least some of these texts are rather forbidding, after all, and not obviously to everyone’s taste. Moreover, not many of these same people, as 40-something stockbrokers or business executives would on their own find texts very engaging.59 Johnson argues that the value attached to reading such texts has “far less to do with cognition than with the construction of a particular reading community, one that validates itself through texts deemed important to a shared sense of culture and cultural attainment.”60 Reading in classical antiquity was a social act that functioned to reinforce elite culture. From an examination of such writers as Plato, Plutarch, and Pliny, Johnson collects instances of the practice of group reading. He summarizes its salient aspects: (1) the reading is a shared, group activity, where one person . . . brings out the meaning of the text for the rest; (2) the reading is of a difficult text; (3) the reading is, at least for some listeners, not a passive activity—the possibility of interruption is real; (4) implicitly, the goal of the reading is not simply to learn what [the author] had to say on this subject, but to promote a wider discussion, an extension of the dynamic interactions of the social group; (5) the reading becomes thereby both focus and springboard to a mutually understood set of group behaviors that serve to build the sense of an intellectual community.61 Thus, the practice of reading and recitation was not simply a matter of entertainment, but served as a basic component in the construction and reinforcement of elite intellectual culture and its shared values and commitments. The application of the notion of a reading community to the Christ groups at a lower rung of the social ladder is obvious: whatever the literacy levels among early Jesus followers and Christ groups, the depiction of its earliest purveyors as literate and the constant iteration of quotations of, and allusions to, the Scriptures reinforced the notion that books and the knowledge associated with books was of central importance. This is true whether one engaged in the practice of Matthew of trying to connect even minor aspects of Jesus’ activities to biblical texts, or Marcion’s 59. William A. Johnson, Readers and Reading Culture in the High Roman Empire: A Study of Elite Communities, Classical Culture and Society (New York: Oxford University Press, 2010), 12; “Sociology of Reading,” 603. 60. Johnson, Readers and Reading Culture, 12. 61. Johnson, “Sociology of Reading,” 618. 42 JOURNAL OF EARLY CHRISTIAN STUDIES privileging of certain texts—an expurgated version of Luke—and denying importance to the Hebrew Bible, or the Apocryphon of John’s inversionary reading of the Genesis story. The socio-cultural value of knowledge related to books was reinforced among the elite by the practice, richly illustrated in Pliny’s letters, of private and semi-public recitation of literary works, normally associated with meals.62 “In Pliny’s circle recitation was a critical mechanism for active engagement with literary pursuits, functioning so as to chart a middle course between the glory of widespread acknowledgement of one’s value as an author and the negative potential of rejection.”63 In the case of the gatherings of the early Jesus movement and Christ groups we know little before the middle of the second century about reading practices. Justin, representing practices in Rome, states And on the day called for the sun, there is a common gathering of all who live in cities or in the country, and the memoirs of the apostles (τὰ ἀπομνημονεύματα τῶν ἀποστόλων) or the writings of the prophets are read (ἀναγινώσκεται), as long as time allows. Next, after the reader has stopped (παυσαμένου τοῦ ἀναγινώσκοντος), the president admonishes and encourages with a speech to imitate these good things. Then we all rise together and pray and, as we before said, after we have finished praying, bread and wine and water are brought, and the president in like manner offers prayers and thanksgivings, according to his ability, and the people assent, saying Amen; and there is a distribution to each and a participation of the food over which thanks have been said, and for those who are not present some is sent through the ‘servants’.64 If we assume that Justin is not simply reporting on the idiosyncrasies of some Roman groups, what we have is a non-elite replication of the 62. Johnson, Readers and Reading Culture, 32–62. See Pliny, Ep. 3.1.9; 6.31.13; 9.36.49 9.40.3. 63. Johnson, Readers and Reading Culture, 53. 64. Justin, Apol. 67.3–5 (Edgar J. Goodspeed, ed., Die ältesten Apologeten. Text emit kurzen Einleitungen [Göttingen: Vandenhoeck und Ruprecht, 1914], 75; my translation). David Brakke (“Scriptural Practices in Early Christianity: Towards a New History of the New Testament Canon,” in Invention, Rewriting, Usurpation: Discursive Fights Over Religious Traditions in Antiquity, ed. Jörg Ulrich, AndersChristian Jacobsen, and David Brakke, Early Christianity in the Context of Antiquity 11 [Frankfurt: Peter Lang, 2011], 263–80) derives a set of “scriptural practices”— three sets of institutional context in which “Christian readers made judgments about texts, insured their reproduction, and reintroduced them to new generations” (271): (1) study and contemplation; (2) revelation and continued inspiration; (3) communal worship and edification. Brakke notes that Justin’s description has elements of the first and third practices (276). KLOPPENBORG / CHRISTIAN BOOK CULTURE 43 reading practices described by Pliny: reading of texts in connection with a meal, followed by some form of erudite interpretation or discussion. It is unnecessary to suppose that all who participated were literate. On the contrary, Justin specifically points to the role of the lector, ὁ ἀναγινώσκων, distinct from the president, who is responsible for reading. This maps well onto Pliny’s references to his trained lectors, probably slaves or freedmen,65 responsible for reading during Pliny’s dinners.66 Justin is not the only second-century example. As Brakke has argued, the Apocryphon of James embodies the three moments of Stock’s textual communities:67 “oral contact” by encountering written gospels in preaching and ritual; “education,” which in the case of Ap.Jas. requires effort and hard work to understand the significance of Jesus’ parables; and the “historicizing of the community” through such intertextual strategies as allusions to widely circulating gospels, portraying the disciples as involved in the composition of these gospels, fashioning the characters of Jesus and James as paradigms for the work of the teacher, and having Jesus interact with Peter and James in a manner that evokes the classroom dialogue.68 Whereas Justin’s group rooted itself intertextually in both the “memoirs of the apostles” and the “writings of the prophets,” the group represented by the Ap.Jas., evidently less oriented to the synagogue and the LXX, turned 65. Pliny, Ep. 8.1, names Encolpius (probably a slave) as his favorite lector, who suffered from dust in his throat and could not read; at 5.19.3 he names Zosimus, a freedman as a gifted comoedus. Raymond Starr has collected many examples of lectors as a standard part of an elite Roman household in “Reading Aloud: Lectores and Roman Reading,” CJ 86 (1990–91): 337–43. See also D. J. Allan, “Ἀναγιγνώσκω And Some Cognate Words,” CQ 30 (1980): 244–51. 66. See, however, the strange case of the “illiterate lector” (ἀναγώστης) recorded in P.Oxy. 33:2673 (Oxyrhynchus, 304 c.e.)—a lector’s declaration that he has handed over various items of the logistes for transport to Alexandria, which includes the codicil “I, Aurelius Ammonious swore an oath as aforesaid. I Aurelius Serenus, wrote this on his behalf because he does not know letters.” See the discussion of this in Malcolm Choat and Rachel Yuen-Collingridge, “A Church with No Books and a Reader Who Cannot Write: The Strange Case of P.Oxy. 33:2673,” Bulletin of the American Society of Papyrologists 46 (2009): 109–38; Peter J. Parsons, The City of the SharpNosed Fish: Greek Papyri Beneath the Egyptian Sand Reveal a Long-Lost World (London: Phoenix, 2007), 154; Gamble, Books and Readers, 3–4; Herbert C. Youtie, “ΑΓΡΑΜΜΑΤΟΣ: An Aspect of Greek Society in Egypt,” HSCP 75 (1971): 163; Ewa Wipszycka, “Un lecteur qui ne sait pas écrire ou un chrétien ou un qui ne veut pas se souiller? (P.Oxy. XXXIII 2673),” Zeitschrift für Papyrologie und Epigraphik 50 (1983): 117–21; Blumell, Lettered Christians, 184–87. 67. Above, n.53. 68. Brakke, “Parables and Plain Speech,” 217. 44 JOURNAL OF EARLY CHRISTIAN STUDIES not to all of the Jesus tradition, but especially to sayings of Jesus, in particular parables and “open” teachings.69 Finally, we can mention the Gospel of Thomas, in which Jesus’ speaking, iterated throughout with the formula ΠΕϪΕ ΙC, comes in the form of a historicized account of the result of literary activity of one of Jesus’ first disciples. And like the Apocryphon of James, Thomas stresses the educative moment, which consists in puzzling through Jesus’ words in a process described with the metaphor of “toil.” As with the Apocryphon of James, salvation is a matter of finding the interpretation of obscure words, and written words at that.70 It is unnecessary to assume that all early Christian groups developed in the same fashion, by forming themselves as reading communities. The accounts of Justin, the Apocryphon of James, and the Gospel of Thomas indicate that even when reading communities formed, they did not always select the same texts to valorize and did not historicize themselves in the same way. But those that did form help to account for the eventual emergence of Christianity as a book culture in the fourth century. Reading Made Easy? What were the mechanics of literary consumption in Christian circles? Johnson describes the bookroll as “an egregiously elite product intended in its stark beauty and difficulty of access to instantiate what it is to be educated.”71 The bookroll was both an aesthetic object72—as the Chris69. Brakke, “Parables and Plain Speech,” 218: “The circulation of multiple written gospels in the second century made possible a textually oriented mode of Christian socialization based on an educational model of salvation. Such teachers as Justin Martyr, Valentinus, Clement, and Origen placed the study of Christian and other sacred literature at the center of social and spiritual formation. Rather than being a primitive witness to the oral Jesus tradition or a ‘Gnostic’ protest against an emerging Christian written scripture, Ap. Jas. is an artifact of this culture of study within early Christianity” (203). 70. See John S. Kloppenborg, “Beyond Tinkering and Apologetics,” Annali di Storia dell’Esegesi 25:2 (2009): 23–35, esp. 31–33. 71. Johnson, Readers and Reading Culture, 21. 72. Johnson, “Sociology of Reading,” 613: “Unlike the utilitarian, documentary text, the bookroll was often used in a display setting: the reading accomplished by a lector, and in a social context such as after-dinner entertainment. . . . As a cultural signifier, the bookroll is analogous in many respects to statuary in a garden or to the luxurious plate on which dinner is served in an elite household. The literary roll exemplifies high culture not just in the demonstration that the owner is ‘literate’ and educated, but by means of the aesthetics the book roll also points up the refinement of the owner.” He earlier observes that the care taken in the production of bookrolls is exemplified in the fact that “the variation in column width from one end of a roll KLOPPENBORG / CHRISTIAN BOOK CULTURE 45 tian book would eventually become—and a text that was inherently and perhaps intentionally difficult to read. Hence, the bookroll served symbolically to mark off groups of elite “consumers” of texts, who had access to trained lectors, who could purchase calligraphic copies of texts, who had the education for erudite discussion, and who belonged to social networks which allowed for the circulation of texts and within which dinner invitations provided the occasions for intellectual exchange. The literary bookroll was dauntingly constructed. Not only was the text presented scripta continua with few paralinguistic markers and visual aids to assist the lector but there was also little in the way of punctuation, no indications of pauses in the text, no page numbers or running headers. These factors did not make the bookroll entirely inaccessible to non-elites but it does help to account for the necessity of having trained readers who were able not only to read, but to read with the appropriate emphasis, intonation, and diction appropriate to the work being recited.73 The lector had not only to master the ability to anticipate word boundaries and grammatical periods in a stream of continuous letters, but also to understand sufficiently the author’s style and usages so as to emulate the author’s meaning. Johnson notes that “phrasing emerges from the confluence of the author’s careful stylistic construction—the literary style—and the reader’s informed interpretation and rendering.”74 It seems unlikely that Christians in the second century had many trained readers able to negotiate the bookroll in all of its complexity. Indeed, the to the other almost never exceeds 2 millimeters—only two to three times the width of the pen’s nib” (612). 73. Johnson, “Sociology of Reading,” 610–12 notes that columns tended to be fifteen to twenty-five letters and the letters written with high legibility, both of which facilitated reading. 74. Johnson, Readers and Reading Culture, 25. Jerome complains that his opponent Jovinus wrote in such a way that it was difficult to determine where sentences begin and end and hence, what Jovinus meant: “However often I read him, even till my heart sinks within me, I am still in uncertainty of his meaning. Instead of clearness we have to make a choice between possible meanings. Everything starts from, everything depends upon, something else. It is impossible to make out any connection; and, excepting the proofs from scripture which he has not dared to exchange for his own lovely flowers of rhetoric, his words suit all matter equally well, because they suit no matter at all”; Quotiescumque eum legero, ubi me defecerit spiritus, ibi est distinctio. totum incipit, totum pendet ex altero: nescias quid cui cohaereat; et exceptis testimoniis scripturarum, quae illo venustissimo eloquentiae suae flore mutare non ausus est, reliquus sermo omni materiae convenit, quia nulli convenit (Jerome, Adv. Jovinum 1.3 [PL 23:212B]; trans. Philip Schaff, Nicene and Post Nicene Fathers, second series, 14 vols. [New York: Christian Literature Company, 1890–1900], 6:347). 46 JOURNAL OF EARLY CHRISTIAN STUDIES materiality of the earliest Christian manuscripts suggests that they were constructed in order to make public reading slightly easier, especially for the literate but not professionally-trained lector. Larry Hurtado recently observed that early Christian papyri exhibit a series of features that suggest that the codices were designed for public reading and for reading by persons other than the elite and their trained lectors:75 (1) wider margins, fewer lines per page, fewer letters per line, and slightly wider interlinear space;76 (2) a middle or high dot used to mark the end of a sense unit and the use of the paragraphos (⸏) in the margin to mark a new sense unit; (3) a diaeresis used to mark an initial vowel, where the preceding word ends with a vowel;77 (4) the use of a partial blank line at the end of a period,78 and ekthesis (a letter that protrudes into the left column) on the first line of the new sentence;79 and (5) the use of word spaces;80 None of these features, of course, appears in all early Christian manuscripts or even consistently in any manuscript. Nor are these features unique to early Christian manuscripts. E. G. Turner notes the use of the mid- or high dot in first- and second-century c.e. non-Christian manuscripts.81 Likewise, 75. Larry W. Hurtado, “Manuscripts and the Sociology of Early Christian Reading,” in The Early Text of the New Testament, ed. C. E. Hill and Michael J. Kruger (Oxford: Oxford University Press, 2011), 57–58 and Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids, MI: Eerdmans, 2006), 177–85, 210–29. I am indebted to Professor Hurtado for sending me a pre-publication version of this essay. 76. Hurtado relies on the analysis of Eric G. Turner, The Typology of the Early Codex, Haney Foundation Series Publication 18 (Philadelphia, PA: University of Pennsylvania Press, 1977), 86–87, who compared P.Chester Beatty 9–10 (Daniel, Esther, 34.4 x 12.8 cm with 49–54 lines/page and 17–23 letters/line) with P.Merton 3 (a copy of Homer) (32.5 x 13.7 cm with 52–54 lines/page, 32–38 letters/line) and the “Harris Homer” (Brit. Mus., Pap. 107: 29.5 x 13.0, 48–54 lines/page, 32–38 letters/line), and P.Oxy. 4:697 (Xenophon, Cyropaedia; 33.1 x 12.5, 60 lines/page, 40–45 letters/line) and PGM 1 iv (30.5 x 12.0, 50 lines/page, 31–39 letters/line). Likewise he compared Bodmer 14–15 (Luke-John) with 39–43 lines/page and 24–30 letters/line, with the Bodmer Menander Codex (47–54 lines/page, 25–34 letters/line). 77. E.g., in Ƿ52: οι ϊουδαι[οι η]μι[ . . . ] ουδενα ϊνα ολο[. . . ]. 78. As in Ƿ100 = P.Oxy. 66:4449; third/fourth century c.e. 79. As in P.Oxy. 65:4443 (second century c.e.). 80. Hurtado, Earliest Christian Artifacts, 171–89. 81. Eric G. Turner, Greek Manuscripts of the Ancient World (Princeton, NJ: Princeton University Press, 1971), 11 (hereafter GMAW). See also William A. Johnson, Bookrolls and Scribes in Oxyrhynchus, Studies in Book and Print Culture (Toronto: University of Toronto Press, 2003), passim. See, e.g., 21, 26, 31, 253, etc. KLOPPENBORG / CHRISTIAN BOOK CULTURE 47 the diaeresis is attested as early as the second century b.c.e.82 Ekthesis is used to mark new sense units in both literary (P.Cairo inv. 47454 [second century c.e.]) and documentary papyri (P.Mich. inv. 622 [42 c.e.]; P.Brit. Mus. 825 [155 c.e.]).83 The paragraphos is mentioned by Aristotle as a device to end a period (Rhet. 3.8.1409a.20) and is attested as early as the Derveni papyri (fourth/third century b.c.e.).84 Finally, the use of spaces is also seen in non-Christian papyri, although these normally mark the end of clauses or sentences rather than dividing words.85 Many of these features, however, appear in Greek manuscripts of biblical texts from Qumran and Nahal H≥ ever from the second and first centuries b.c.e., and it is a plausible assumption that copyists of early Christian texts were influenced by their practices.86 Sectional divisions are marked either with space between sections and a paragraphos in the left margin, or an empty space after the last word of the section and ekthesis and/or a paragraphos on the next line, beginning the next section.87 Even more significantly, while some of these manuscripts are written scripta continua, words spaces are found in some biblical texts, such as the earliest fragments of Deuteronomy in Greek (P.Ryl. 3:458)88 and the scroll of the 82. Turner, GMAW, 12. Johnson (Bookrolls, 21) notes that the “diaeresis is usually written on initial iota, usually not on initial upsilon.” He records many instances of the diaeresis in bookrolls. See 22, 23, 29, 31, 36, 252, 257, 264, 266, 270, 271, 274, 276, 278, 283, 285. 83. Turner, GMAW, 9–10 and Plates 44, 48, 59, 65, 69; Johnson, Bookrolls, 19, 274, 327 P.Mich. inv. 622 is now published as P.Mich. 2:121r. 84. William A. Johnson, “The Ancient Book,” in The Oxford Handbook of Papyrology, ed. Roger S. Bagnall (New York: Oxford University Press, 2009), 261. 85. Turner, GMAW, 10, citing P.Hamb. 646 (mid-third century b.c.e.); P.Brit. Mus. 131 (late first century c.e.); P.Oxy. 3:473 (138–60 c.e.). See also Frederic G. Kenyon, The Palaeography of Greek Papyri (Oxford: Oxford University Press, 1899), 26–27. 86. Thus Emmanuel Tov, Robert A. Kraft, and Peter J. Parsons, The Greek Minor Prophets Scroll from Nahal Hever: 8 Hev XII Gr., Discoveries in the Judaean Desert 8 (Oxford: Clarendon Press, 1990), 13; Hurtado, Earliest Christian Artifacts, 183. 87. See Emmanuel Tov, Robert A. Kraft, and Peter Parsons, 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), 139–42 and, e.g., 4QpapLXXLeva (late second/first century b.c.e., space between sections and a paragraphos in frag. 1 21, after Lev 26.13); 4QpapLXXLevb (first century b.c.e., spaces between sections and paragraphoi in frag. 27–31 6 after Lev 5.19 and in frag. 32); P.Fouad 266a–b (mid-first century b.c.e., use of both added space or empty lines between sections with paragraphoi); P.Oxy. 65:4443 (first/second century c.e., partially empty line, a paragraphos, and ekthesis after Esther 8.12); 8H≥ evXIIgr col. 19 (first century b.c.e., blank line, ekthesis + paragraphos), cols. 30–31 (paragraphos). According to Tov (184), the paragraphos is particularly common in Greek biblical manuscripts. 88. Ernest J. Revell, “The Oldest Evidence for the Hebrew Accent System,” Bulletin of the John Rylands Library 54 (1971): 214–22. 48 JOURNAL OF EARLY CHRISTIAN STUDIES Minor Prophets (8H≥ evXIIgr).89 While the use of word spaces is not consistent, the practice was likely due to the fact that Hebrew biblical manuscripts from the period use word spaces extensively. In any case, they made reading easier by indicating the boundaries of words in an otherwise unbroken stream of letters. Hurtado argues that the strong preference for the codex over the bookroll,90 the use of wider margins, fewer lines per page, shorter line-lengths, and the presence of readers’ aids represent deliberate choices to “turn away from the elitist format of high quality literary manuscripts.”91 That these choices were matters of deliberate identity formation and not pragmatic decisions related to the ease of reading is difficult to show, but at least the effect of the adoption of the codex was to mark off Christian literary production from that of high literary culture, for whom the bookroll remained the ideal.92 Hurtado makes much of the number of lines per page as a significant difference from pagan literary works. According to E. G. Turner, second- and third-century codices typically had more than forty-five lines per page.93 89. Tov, Kraft, and Parsons, Scribal Practices, 131–33. For word spaces in Greek, see Tov, Kraft, and Parsons, Greek Minor Prophets Scroll, plates X (col. 16), XI (col. 17), XVII (col. 30–31), XIX (col. B1–2). 90. Hurtado (“Sociology of Early Christian Reading,” 56) observes that in the second and third centuries c.e. the striking disproportion in the distribution of codices and bookrolls is striking. While bookrolls dominate in both centuries, identifiably Christian works (2% in the second century and 12% in the third) represent respectively 27% and 38% of the extant codices. About 76% of Christian works take the codex form—striking because the preference for the bookroll for non-Christian literature would not change until the fourth century, rising from 4.9% in the second, to 21% in the third, to 56.3% in the fourth (Hurtado, Earliest Christian Artifacts, 43–93). 91. Hurtado, “Sociology of Early Christian Reading,” 59 (emphasis original); Hurtado, Earliest Christian Artifacts, 69. 92. Johnson, “The Ancient Book,” 267: up to the end of the second century, “the bookroll constituted the enduring image of what a classical literary text was meant to be.” It is tempting to suggest that the Gesta apud Zenophilum, which refers to the seizure of codices and the Martyrdom of Agape, Irene and Chione (above n.29), which refers to διφθέρας καὶ βιβλία καὶ πινακίδας και κωδικέλλους καὶ σελίδας γραφῶν τῶν ποτε γενομένων Χριστιανῶν, but does not mention τόμοι, the normal term for bookrolls, reflects a knowledge of the preference of Christians for codices. The Acts of the Scillitan Martyrs (above) refers only to libri et epistulae, terms that are indifferent to the exact format of writing. 93. Turner, The Early Codex, 85–87, 96–97. He notes for example, P.Köln inv. 3328 (Lollianus): 59+ lines; P.Berol. inv. 13236 (Thucydides): 60 lines; P.Ant. 3:162 (Homer): 51 lines; PSI 6:720 (Aristophanes): 56 lines; P. Mert. 3 (Homer): 52–54 lines; P.Oxy. 4:697 (Xenophon): 60 lines; P.Oxy. 32:2647 (Hesiod): 58–61 lines; PGM 1 iv: 31–39 lines; Bodmer Menander codex: 47–54 lines. KLOPPENBORG / CHRISTIAN BOOK CULTURE 49 By contrast, P.Bodmer 2 (Ƿ66) has between 15 and 26 lines on a (16.2 x 14.2 cm) page, compared with 24–25 lines on a similarly-sized codex of Philo.94 Metzger calculated that P. Rylands 3:457 (Ƿ52) would have had only eighteen lines per page and that P.Chester Beatty 2 (Ƿ46) had 26–32 lines.95 Many other early Christian papyri are of the same order. A selection of second and third-century manuscripts illustrates the point:96 Papyrus Date (c.e.) P.Oxy. 64:4404 (Ƿ104) Lines per page 2nd cent. ca. 31 P.Oxy. 50:3523 (Ƿ ) late 2nd cent. ca. 23–24 P.Ryl. 3:457 (Ƿ52) late 2nd cent. 18 90 46 P.Chester Beatty 2 (Ƿ ) ca. 200 26–32 P.Bodmer 2 (Ƿ66) ca. 200 ca. 20 P.Oxy. 34:2683 + 64:4405 (Ƿ77) 2nd/3rd cent. ca. 20 P.Oxy. 64:4403 (Ƿ103) 2nd/3rd cent. ca. 19–20 P. Rylands 1:5 (Ƿ32) early 3rd cent. 26–27 P.Oxy. 1:1 (Gos. Thomas) early 3rd cent. ca. 3597 P.Oxy. 12:1596 (Ƿ28) 3rd cent. 25–26 P.Oxy. 64:4401 (Ƿ101) 3rd cent. ca. 32–33 106 P.Oxy. 65:4445 (Ƿ ) 3rd cent. ca. 36 P.Oxy. 65:4446 (Ƿ107) 3rd cent. ca. 33 94. Roger A. Pack, The Greek and Latin Literary Texts from Greco-Roman Egypt, 2nd ed. (Ann Arbor, MI: University of Michigan Press, 1965), no. 1344 = PSI 2:1207 + P.Oxy. 9:1173; 11:1356; 18:2158 + P.Haun. 1:8. Bruce M. Metzger (Manuscripts of the Greek Bible: An Introduction to Greek Palaeography [Oxford: Clarendon Press; New York: Oxford University Press, 1981], 66) reports 17 lines/page for Ƿ 66 but only for οθ* (= 79). 95. Metzger, Manuscripts of the Greek Bible, 62, 64. 96. This table is compiled from the data in http://chrles.multiply.com/photos?& =&album=&page_start=40 (accessed Jan. 13, 2013 but now unavailable) and, where estimates are missing, from my calculations of page lengths where the recto and verso contain text from continuous portions of the gospels, and thus allow for an accurate estimate of page size. The statistics for Oxyrhynchus Papyri vol. 64 are from Peter M. Head, “Some Recently Published NT Papyri from Oxyrhynchus: An Overview and Preliminary Assessment,” Tyndale Bulletin 51 (2000): 5. 97. Joseph A. Fitzmyer, “The Oxyrhynchus Logoi of Jesus and the Coptic Gospel According to Thomas,” in Essays on the Semitic Background of the New Testament, SBL Sources for Biblical Study 5 (Missoula, MT: Scholars Press, 1974), 355 n.2 calculated the column length at 38 lines per page, but this seems too high. My reconstruction of the missing lines, using Fitzmyer’s own retro-translation, indicates 35 lines per page, which is comparable in column height to P.Oxy 65:4445, and 66:4497 and to the manuscripts listed in the next note. 50 JOURNAL OF EARLY CHRISTIAN STUDIES P.Oxy. 65:4447 (Ƿ108) 3rd cent. ca. 23 P.Oxy. 65:4448 (Ƿ109) 3rd cent. ca. 26 P.Oxy. 66:4495 (Ƿ111) 3rd cent. ca. 21–22 P.Oxy. 66:4497 (Ƿ113) 3rd cent. ca. 35 P.Oxy. 66:4498 (Ƿ114) 3rd cent. ca. 2798 It is clear that these papyri—all of them from codices—have significantly fewer lines per page than the non-Christian codices of the second and third centuries discussed by Turner, and fewer than the fourth-century uncial codices, Sinaiticus and Vaticanus, at 42–48 lines per page. These compare, however, rather more favorably to second and third-century bookrolls in column length (which after all represent by far the largest proportion of surviving books), although they still fall on the lower end of the range. In his survey of hundreds of bookrolls from Egypt, Johnson reports that in the second and third centuries, literary manuscripts range from 24–42 lines per column. Even more significantly, the line-leading99 of literary papyri ranges from 4.1mm/line to 7.5mm/line, with a median in the 5.5±1.0 range.100 Papyrus Date c.e. Leading (mm/line) Lines per page P.Oxy. 2:223 (Homer) 3rd cent. 7.0–7.55 24–25 P.Oxy. 2:228 (Plato) 2nd cent. 5.1 32 P.Oxy. 2:230 (Demosthenes) 2nd cent. 6.8–7.1 35–36 P.Oxy. 3:844 (Isocrates) 1st/2nd cent. 5.45–5.7 39–41 P.Oxy. 7:1017 (Plato) 2nd/3rd cent. 7.0–7.45 33–35 P.Oxy. 10:1250 (Achilles Tatius) 3rd/4th cent. 4.6 41–42 P.Oxy. 17:2100 (Thucydides) 2nd cent. 5.0 37–38 98. There are, of course, early Christian manuscripts with more lines per page: for example, P.Bodmer 14–15 (Ƿ75; third century c.e.: 38–45 lines/page); P.Oxy. 4:654 (Gos. Thom.: third century c.e.; a bookroll with 42+ lines/column); P.Oxy 69:4707 (Hermas; third century c.e.: 55 lines/page); P.Chester Beatty 1 (Ƿ45; third century c.e.: 39 lines/page); P.Chester Beatty 9–10 (Esther, Ezekiel, Daniel: third century c.e.; 45–57 lines/page). 99. “Leading” is used in two senses: in typesetting it refers to the interlinear space; it is also used to refer to the distance between one baseline and the next. I use it here in the latter sense. 100. Johnson, Bookrolls, 83–84. See also Johnson, “The Ancient Book,” 259, where he reports the “typical” bookroll dimensions in the Roman period as: height: 25–33 cm; column height: 12–26 cm; column width (prose): 4.5–9.0 cm; letters per line: 13–24 (at the extremes, 10–30); lines per column: 25–50 (at the extremes, 18–64). KLOPPENBORG / CHRISTIAN BOOK CULTURE P.Oxy. 17:2101 (Xenophon) 3rd cent. 4.55 36 P.Oxy. 17:2102 (Plato) late 2nd cent. 5.35–5.5 29 P.Oxy. 22:2335 (Euripides) 2nd cent. 4.1–4.2 37–39 P.Oxy. 49:3436 (Dinarchus) 2nd/3rd cent. 5.6 35 P.Oxy. 49:3447 (Strabo) early 2nd cent. 6.0–6.1 34 P.Oxy. 60:4040 (Aeschines) 2nd/3rd cent. 4.9–5.3 33–37 51 Second- and third-century Christian manuscripts for which line-leading can be calculated display a similar range (3.7–8.5).101 Papyrus Date c.e. Leading (mm/line) Lines per page P.Oxy. 64:4404 (Ƿ104) 2nd cent. 4.44 ca. 31 P.Oxy. 50:3523 (Ƿ90) late 2nd cent. 5.0–5.2 ca. 23–24 P.Ryl. 3:457 (Ƿ52) late 2nd cent. 8.5 18 P.Egerton 2 late 2nd cent. 6.1 unknown P.Oxy. 34:2683 + 64:4405 (Ƿ ) 2nd/3rd cent. 5.0 ca. 20 P.Bodmer 2 (Ƿ66) ca. 200 5.5 ca. 20 77 1 P.Oxy. 1:2 (Ƿ ) 3rd cent. 5.7 ca. 30 P.Oxy. 50:3527 (Hermas) early 3rd cent. 6.2 25 Paris, Bibl. Nat. Gr. 1120 (Ƿ4) 3rd cent. 3.7 36 P.Oxy. 69:4707 (Hermas) 3rd cent. 5.3 30 P.Oxy. 71:4803 (Ƿ119) 3rd cent. 5.9 40 P.Oxy. 71:4805 (Ƿ121) 3rd cent. 5.0 37 P.Oxy. 66:4449 (Ƿ100) 3rd/4th cent. 6.3–6.6 35 It can be seen immediately that the line-leading of early Christian papyri falls generally within the range of that of contemporary bookrolls, with P.Rylands 3:457 (Ƿ52) representing an extreme case of spacing (8.5mm), but not wildly disproportionate with a second-century text of Homer which has a comparable number of lines per column (P.Oxy. 2:223: 7.0–7.55mm). Although several early Christian papyri fall at the lower end of the range of lines/page and a few display quite generous line-leading, both of which features might have made reading easier, the more significant aid to reading came in two forms: the presence of word spaces, and the use of paralinguistic marks as punctuation. 101. I have calculated leading from photographs that include ruler marks, either by measuring the column height and dividing by the number of lines, or in the case of fragmentary papyri, by dividing the visible column height by the number of visible lines. 52 JOURNAL OF EARLY CHRISTIAN STUDIES Word spaces are extremely rare in Greek literary texts, which are normally laid out scripta continua,102 but evidenced in some Greek manuscripts of the Jewish Bible (above). It seems plausible that Christian scribes adopted this practice from their Judaean counterparts. Punctuation marks of various kinds are attested in Greek literary manuscripts, but quite common in some Christian manuscripts. I note here instances of both word spaces and punctuation in a selection of early, mostly second-century, papyri: P.Oxy. 64:4404 (Ƿ104; mid-second century c.e. containing Matt 21.34–37, 43, 45) is now the oldest NT papyrus extant. Although Head reports that apart from a few rough breathing marks, “no other lectional or punctuation feature is found,”103 it should be observed that there are small spaces between αυ[τ]ου_προς (→ l. 3), λα]βειν_τους (→ l. 4), and απεκτειναν_ο(ν) (→ l. 6).104 P.Ryl. 3:457 (Ƿ52; late second century c.e., with John 18.31–33, 37– 38) not only shows generous line-leading but the use of diaereses at → 1 and 2, and word spaces, twice to mark the beginning of two clauses (→ 2: αποκτειναι ουδενα_ϊνα ο λογος and ↓ 2: κοσμον_ϊνα μαρτυρησω), and once at → 3 (ειπες_σημαίνων). P.Egerton 2 (late second century c.e.,105 a fragment of an unknown gospel) displays many paralinguistic marks: word spaces at frag. 1 ↓ 4 (bis), 6, 8, 9, 15; frag. 1 → 5, 10, 12, 13, 14; frag. 2 → 1, 2, 3, 4, 5, 8 (bis), 9, 15; and frag. 2 ↓ 8. A diaeresis appears at frag. 2 ↓ 7 (ϊο[ρδ]ανου). Most importantly, the scribe used paragraphoi 102. Johnson, “The Ancient Book,” 261. 103. Head, “Some Recently Published NT Papyri,” 10. 104. I use an understrike (_) to mark a word space, → for recto and ↓ for verso. 105. The dating of P.Egerton 2 from the early second century to the late second is partly a consequence of the discovery of P.Köln 6:255, now identified as part of the same papyrus. H. I. Bell and Theodore C. Skeat, Fragments of an Unknown Gospel and Other Early Christian Papyri (London: British Museum, 1935) had originally declared that P.Egerton 2 was “unquestionably the earliest specifically Christian manuscript yet discovered in Egypt” (at 1) and “conservatively” dated it to the mid-second century. The appearance of an apostrophe (καὶ ἀνένεγ’κον) in P.Köln 6:255 → 21 suggests, however, a date late in the second century or early in the third. Michael Gronewald, “Unbekanntes Evangelium oder Evangelienharmonie (Fragment aus dem ‘Evangelium Egerton’),” in Kölner Papyri (P. Köln) 6, ed. Michael Gronewald, Sonderreihe Papyrologica Coloniensa 7 (Köln: Reinisch-Westfälischen Akademie der Wissenschaften und Universität Köln, 1987), 136–37: “Nachzutragen ist, daß sich in dem Kölner Fragment nun auch Apostroph zwischen Konsonanten (ἀνένεγ’κον) wie in P.Bodmer II findet, was nach E.G. Turner, GMAW, 13 eher ins dritte Jahrhundert weist. Doch auch bei einer eventuellen Datierung um 200 würde P.Egerton 2 immer noch zu den frühesten christlichen Papyri zählen.” KLOPPENBORG / CHRISTIAN BOOK CULTURE 53 at frag. 1 ↓ 4, 14, indicating a new period, and mid-dots (·) also to mark off new grammatical periods at frag. 1 → 8, 10, 12, 14, 17, 19; frag. 1 ↓ 3, 6, 12, 15, 16; frag. 2 → 10, 11, 12; frag. 2 ↓ 11, 12, 13 and frag. 3 → 5. It should be noted, additionally, that the middle dots are not the additions of a reader, as in some manuscripts, but the work of the original scribe. Finally, as Roberts observes, there is a tendency of the scribe to enlarge the first letter following a pause (frag. 1 → 1, 9; frag. 2 ↓ 11, 13), a practice that is associated with documentary papyri rather than literary manuscripts.106 P.Oxy. 60:4009 (Gospel of Peter?;107 second century c.e.) has a partial blank line (→ 10), apparently separating two periods, and ekthesis signaling the beginning of a new sense unit (↓ 10). P.Oxy. 34:2683 + 64:4405 (Ƿ77; late second/third century c.e., containing Matthew 23.30–34, 35–39) uses a paragraphos twice, to signal pauses at the beginning of new clauses at vv. 32 and 34 and between vv. 36 and 37 the scribe leaves almost a whole blank line.108 P.Oxy. 64:4403 (Ƿ103; second/third century c.e., containing Matt 13.55–56; 14.3–5) has three mid dots (↓ 4; → 2, 4), the latter two separating grammatical periods, and the former either a mistake or offering a non-conventional break between αἱ ἀδελφαὶ αὐτοῦ and the question, οὐχὶ πᾶσαι πρὸς ἡμᾶς εἰσιν? P. Chester Beatty 2 + P.Mich. 222 (Ƿ46; ca. 200 c.e., a large manuscript with Pauline letters and Hebrews) shows generous interlinear spacing, and the beginning of word spaces—not between every word but, for example, in Hebrews 1.1, between πολυμερως and καιπολυτροπως, and between πάλαι and οΘΣελαλησας and before τοισπατρασιν. 106. Colin H. Roberts, Manuscript, Society, and Belief in Early Christian Egypt, The Schweich Lectures for 1977 (London and New York: Oxford University Press, 1979), 16. 107. Dieter Lührmann, Fragmente apokryph gewordener Evangelien in griechischer und lateinischer Sprache, Marburger theologische Studien 59 (Marburg: N. G. Elwert Verlag, 2000), 72–95 has identified P.Oxy. LX 2009 as part of the Gospel of Peter even though there are only slight overlaps between this fragment and the Akhmîm manuscript. This identification has been challenged by Paul Foster, “Are There any Early Fragments of the So-Called Gospel of Peter?” NTS 52 (2006): 1–28. For my purposes, it is unnecessary to enter into this debate, since my concern is only with the paleography of P.Oxy. 60:4009. 108. See Head, “Some Recently Published NT Papyri,” 7. 54 JOURNAL OF EARLY CHRISTIAN STUDIES P.Bodmer 2 (Ƿ66; ca. 200 c.e., a nearly complete copy of John), written in a very legible hand, uses a system of blank line-endings followed by ekthesis on the following line to mark sectional divisions,109 and mid-dots to mark grammatical periods (e.g., p. 1 ↓ 1, 2, 3, 6, 7 etc.). A diaeresis often appears on an initial iota following a vowel. P.Oxy. 4:654 (Gospel of Thomas sayings 1–7; mid-third century c.e.)110 is part of a bookroll written on the verso (↓) of a land survey. The scribe used a diaeresis once (↓ 21) and placed paragraphoi after each of the six sayings. Additionally the scribe used a line-filler (diplê obelismenê, or forked paragraphos [⸐]) at the end of each of the sayings and prior to the next λέγει Ἰη(σοῦ)ς.111 Annemarie Luijendijk notes on the one hand the re-use of a papyrus and the fact that P.Oxy. 4:654 is a roll rather than a codex as indications that it was intended for private, rather than liturgical use,112 but also notes Roberts’s comment a propos of P.Ryl. 1:1 (a copy of Deuteronomy) that “not all texts written on improvised material need have been private.”113 She suggests that the layout of the text, with multiple lectional aids, might permit the supposition of use in a liturgical setting or for “reading out loud in a different context, for instance, in an educational setting.”114 109. See the discussion in Victor Martin, Papyrus Bodmer II: 1: Evangile de Jean chap. 1–14. Supplément: Evangile de Jean, chap. 14–21 (Cologny-Genève: Bibliotheca Bodmeriana, 1956–62), 1:18–21. 110. Harold W. Attridge, “The Gospel of Thomas: The Greek Fragments,” in Nag Hammadi Codex II,2–7 Together with XIII,2*, Brit. Lib. Or.4926(1), and P.Oxy. 1, 654, 655, ed. Bentley Layton, NHS 20 (Leiden: Brill, 1989), 97 dates P.Oxy. 4:654 to the mid-third century; Annemarie Luijendijk, “Reading the Gospel of Thomas in the Third Century,” in Reading New Testament Papyri in Context: Lire les papyrus du nouveau testament dans leus context, ed. Claire Clivaz and Jean Zumstein, Bibliotheca Ephemeridum theologicarum lovaniensium 242 (Leuven, Paris, and Walpole, MA: Peeters, 2011), 245, following B. P. Grenfell and A. S. Hunt, eds., The Oxyrhynchus Papyri. Part IV, Egypt Exploration Society, Graeco-Roman Memoirs 4 (London: Egypt Exploration Society, 1904), 1, who date it to the second half of the second century. 111. The filler is erroneously placed after “Jesus says” in ↓ 27. 112. Luijendijk, “Reading the Gospel of Thomas,” 251, citing Roberts, Manuscript, Society, and Belief, 9, who takes re-used papyri to be a sign of private use. Similarly, Roger S. Bagnall, Early Christian Books in Egypt (Princeton, NJ: Princeton University Press, 2009), 76. 113. Roberts, Manuscript, Society, and Belief, 9. 114. Luijendijk, “Reading the Gospel of Thomas,” 253–54. Apropos of P.Oxy. 1:1 Larry Hurtado (“The Greek Fragments of the Gospel of Thomas as Artifacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654 and Papyrus Oxyrhynchus 655,” in Das Thomasevangelium: Entstehung, Rezeption, KLOPPENBORG / CHRISTIAN BOOK CULTURE 55 To these papyri one might compare Papyrus Oxyrhynchus 75:5048, an early second-century copy of Xenophon’s Memorabilia (2.6.12–21), written in a clear bookhand, 21 x 24 cm, with 30 lines/column and a lineleading of 5.1mm/line. The scribe supplied a few middle dots (→ col. ii 1, 4; iii 19, 20, 26) and paragraphoi at ii 5, 16; iii 26; iv 12 to mark a change in speaker. As a collection of anecdotes about Sokrates, the content of the text is generally comparable to the gospel texts noted above, as are the column height and the line-leading. One can easily suppose that this roll was used for public reading in a “textual community”; yet the elegant bookhand and the relatively few lectional markings likely presuppose a lector with greater skill than that needed for the Christian papyri surveyed above. The same, incidentally, might be said of P.Bodmer 14–15 (Ƿ 75; early third century c.e.; 38–45 lines/page), a copy of portions of Luke and John. It shows considerably more lines per page than other early manuscripts, with fewer obvious word spaces. In the first five lines there are only two word spaces (l. 4 ουδε_εν; l. 5: ζωηην_και), middle dots at ου κατελαβεν · εγενετο, and high dots at ουδε εν ˙ and ονομα αυτου ιωαννης ˙ ουτος ηλθεν. A diaeresis can be seen at φωτος ϊνα παντες. One wonders whether the smaller interlinear space and more lines per page are related to the smaller number of paralinguistic signs. That is, does this suggest that Ƿ 75 is the product of a more accomplished scribe, and for a lector who does not require the kinds of readers’ aids noted above in other early papyri? It must be stressed that none of these readers’ aids is unique to early Christian documents, that none is deployed consistently throughout any of the papyri that have been examined, and that there was no consensus on Theologie, ed. Jörg Frey and Jens Schröter, Beihefte zur Zeitschrift fur die neutestamentliche Wissenschaft 157 [Berlin and New York: Walter de Gruyter, 2008], 22–24) has argued that the column height of the papyrus, which following Fitzmyer (above n.97) he puts at 37–38 lines, suggests that it was for personal reading rather than public reading. He also observes that the hand suggests a scribe of “modest literary ability” and concludes that the absence of scribal devices associated with public reading indicate that it was a copy for private use. As indicated above (above n.97), the column height should be set at 35 rather than 37–38 lines, putting P.Oxy. 1:1 well within the range of other papyri that Hurtado believes to be used for public reading. The line length (15–17 letters) is in fact significantly shorter than most of the papyri studied above, and the scribal hand is not significantly different from other early Christian papyri. Luijendijk (“Reading the Gospel of Thomas,” 255–57) quotes Grenfell and Hunt’s assessment of the papyrus as “a leaf of a handsomely-written book which may well have been a valuable trade copy,” which leads her to conclude that P.Oxy. 1:1 may have been prepared for public use. It is certain that neither the column height nor the scribal hand is significantly different from other papyri that we might assume were read publicly. 56 JOURNAL OF EARLY CHRISTIAN STUDIES which punctuation marks to use. Thus concessions to the non-professional lector vary widely from one papyrus to another. The scribes of Ƿ 75 and P.Oxy 75:5048 evidently could assume (or presume) a somewhat more accomplished lector—one who could negotiate tightly packed lines and fewer deviations from a scripta continua presentation—, while the scribes of P.Egerton 2, P.Chester Beatty 2 and most of other Christian papyri discussed above accommodated readers who required greater interlinear space, paragraphoi, middle dots and even word spaces.115 None of this should be construed to imply that reading was now a simple matter or that there was growing or widespread literacy in second-century Christ groups. These scribal practices suggest that the intended readers in the second-century Christ groups were not drawn from likes of Pliny’s professionally trained slave-lectors. The scribal hands in these papyri suggest something similar. While there are a few early Christian instances of calligraphic hands with bilinear lettering,116 most of the Christian texts surveyed above lack the characteristics of bookhands, and show instead the documentary nature of their hands:117 115. Apropos of P.Oxy. 5:850 (a fourth-century c.e. miniature codex), Michael J. Kruger, The Gospel of the Savior: An Analysis of P.Oxy. 840 and Its Place in the Gospel Traditions of Early Christianity, Texts and Editions for New Testament Study 1 (Leiden and Köln: Brill, 2005), 49–50 observes that middle dots were used not only to mark the end of sentences, but pauses within grammatical periods. He counts 22 middle dots in 45 lines of text representing 12–14 complete sentences. This represents an average of two dots per sentence. “The abundance of these points demonstrates that they do not just mark the beginning of new sentences, but also help the reader sift through the small grammatical points, such as when to pause at the appropriate points.” 116. E.g., P.Oxy. 50:3527 (Hermas; early third century c.e.). For the meaning of “bilinear” see the next note. 117. Guglielmo Cavallo and Herwig Maehler, Hellenistic Bookhands (Berlin: Walter de Gruyter, 2008), 7: “. . . for the sake of simplicity we have defined as ‘bookhand’ those scripts that keep the letters separate and bilinear, i.e., confined between notional upper and lower parallel lines, generally with the exception of phi and psi, slowly written in the attempt to emulate the regularity seen in most stone inscriptions.” By contrast, Ann Hanson describes the documentary style as “individual letters seldom received full articulation, and the scribe’s nubbed pen remained in contact with the surface of the papyrus, producing a chain of letters joined together in ligature” (“Ancient Illiteracy,” in Literacy in the Roman World, ed. J. Humphrey, Journal of Roman Archaeology Supplementary Series 3 [Ann Arbor, MI: Journal of Roman Archaeology, 1991], 173). In practice, there is not a sharp division between literary bookhands and documentary hands. As Cribiore points out, “between [literary and documentary hands] there is an almost infinite range of different levels, and sometimes it is difficult to decide whether a hand belongs to one or the other category” (Raffaella Cribiore, Writing, Teachers, and Students in Graeco-Roman Egypt, American Studies in Papyrology 36 [Atlanta, GA: Scholars Press, 1996], 97). KLOPPENBORG / CHRISTIAN BOOK CULTURE 57 lettering with cursive aspects, the presence of ligatures, and abbreviations (nomina sacra). Indeed, many of the features that I have called “readers’ aids”—word spaces, an enlarged first letter of a new unit, and ekthesis— are elements common in documentary papyri but rare in literary texts.118 Roberts concludes that these papyri are “the work of men not trained in calligraphy, and not so accustomed to writing books” and refers to the hands are “reformed documentary.”119 In addition to the various readers’ aids evidenced in early Christian papyri, there is a wealth of palaeographic features pointing to what Johnson calls “workaday” hands:120 rounded alphas, looping mus, frequent ligatures, and hands that are anything but bilinear.121 These manuscripts, it should be noted, are very different from the classical bookhands seen in the fourth-century codices, Vaticanus and Sinaiticus, in which ligatures are rare and the text is consistently scripta continua. The presence of documentary forms in early Christian papyri may find an explanation in the suggestion that these manuscripts were produced not by the expert copyists responsible for literary texts intended for the elite consumption. Haines-Eitzen of course notes that some scribes were 118. Johnson, “The Ancient Book,” 262: “common in documents but rare in bookrolls is the use of indentation (eisthesis) and reverse indentation (ekthesis); in literary papyri these are normally used only to indicate verse groups or changes in meter. In general, then, lectional aids were few, and little by way of help or intervention interrupted the flow of letters. Thorough training was necessary for one to be able to read this scriptio continua readily and comfortably.” 119. Roberts, Manuscript, Society, and Belief, 14. Cavallo points out that this modified documentary hand emerged in the Antonine period typified by the hands of PSI 5:446 (133–37 c.e.), which shows ligatured alphas, epsilons, and gammas, and P.Oxy. 3:473 (13 c.e.), where the alpha, mu, and upsilon are made in single looping actions; letters are grouped into words (hence, additional micro-spaces between words), spaces appear after periods, and there are some ligatured epsilons (Guglielmo Cavallo, “Greek and Latin Writing in the Papyri,” in The Oxford Handbook of Papyrology, ed. Roger S. Bagnall [New York: Oxford University Press, 2009], 129). This type of hand eventually turned into what Cavallo calls the “Alexandrian majuscule,” the first instances of which are found in P.Egerton 2 and P.Bodmer 2 (Ƿ66). 120. Johnson, “The Ancient Book,” 266. 121. While the earliest NT papyrus to date, P.Oxy. 64:4404 (Ƿ104), has essentially bilinear, with separate letters with only rounded alphas, P.Oxy. 50:3523 has ligatured epsilons and a looping alpha connected to the following letter. P.Egerton 2 has almost all alphas looping and joined to the next letter, mus are looping, and most epsilons are ligatured; P.Ryl. 3:457 is generally a bilinear hand with distinct letters, but many of the epsilons are joined to the following letter; P.Oxy. 34:2683 + 64:4405 (Ƿ77) shows a strong preference for ligatures, especially on the alpha, epsilon, gamma, and tau and at Matt 23.37 the letters αγειταν in ἐπισυνάγει τὰ νοσσία form a single string of connected letters. Ligatures are also a characteristic of P.Chester Beatty 2, in which the alphas are generally not ligatured but many of the epsilons are. 58 JOURNAL OF EARLY CHRISTIAN STUDIES multifunctional and were able to traverse both the bilinear style preferred by booksellers and the cursive styles appropriate to correspondence and recordkeeping.122 But she also notes that, although in the early Roman period one can find both literary and documentary styles in both classical and Christian papyri—perhaps attributable to the multifunctionality of (some) scribes—, the difference . . . is a question of degree: stated negatively, while there are no Christian papyri that are free from a dual influence, there are classical papyri that have no similarities whatsoever to documentary practice.123 Stated another way, the absence in early Christian papyri of a pure literary or bilinear style suggests that, while Christian scribes might also have been multifunctional, they did not have a specific training or extensive experience in copying literary texts. . . . Since experience, by its very definition, is based upon practice, we might well suppose that the earliest copyists of Christian literature were trained professional scribes, whose multifunctionality may well have been suited best for a private context.124 This would appear to locate the scribes of the Christian papyri under consideration somewhat lower on the scribal hierarchy. This seems congruent with the fact that their productions appear to be aimed at lectors who were also less able to negotiate the elite bookroll, but were, with some readers’ aids, still able to recite from a codex. It is not impossible that in some instances the scribe and the lector was in fact the same person. CONCLUSION What Eusebius could take for granted—libraries, the existence of codex technology capable of binding large volumes, including entire Bibles, professional scriptoria in which Christian documents could be copied, and readers able to negotiate the complexities of professionally copied texts— were the results of the political, social, and economic changes brought by Constantine’s embrace of Christianity. Although Eusebius could imagine his predecessors as operating in schools, with professional scriptoria, there 122. Kim Haines-Eitzen, Guardians of Letters: Literacy, Power, and the Transmitters of Early Christian Literature (Oxford: Oxford University Press, 2000), 63. 123. Haines-Eitzen, Guardians of Letters, 66. Similarly, Roberts, Manuscript, Society, and Belief, 15: “Works of secular literature are also written in such [semidocumentary] hands, but there is not the same preponderance of them.” 124. Haines-Eitzen, Guardians of Letters, 68. KLOPPENBORG / CHRISTIAN BOOK CULTURE 59 is little evidence that this was the case. The literary artifacts of the second and third centuries owe more to the work of lower level documentary scribes than they do to the professionals responsible for Vaticanus and other fine codices. In spite of widespread illiteracy and modest resources, and no doubt because of constant exposure to literate media, some second-century Christ groups took steps in the direction of a book culture. Both literary and material evidence suggests that literacy and literate culture were valorized: the principal agents of the Jesus movement came to be represented as skilled literate communicators; the book, perhaps even a codex, became a special symbol of a Christian confession; what might be called “textual communities” formed around the reading, study and preservation of books; and allusions to books and book culture were recursively embedded in Christian documents, so as to historicize those groups by grounding their beliefs and practices in other more ancient writings. Corresponding to the valorization of literate media, we can also see in the construction of early Christian papyri the effort to produce texts that might be accessible to a sub-elite lector, created by scribes of rather middling scribal accomplishment. While the elite bookroll no doubt continued to reign in elite forms of sociability, epitomizing the social world of elite reading communities, these Christian scribes and lectors also created sub-elite textual communities, in which their books could be recited and discussed.125 John S. Kloppenborg is Professor and Chair of the Department for the Study of Religion, University of Toronto, Toronto, Canada 125. I would like to thank Larry Hurtado and the two anonymous readers for comments and criticisms that have assisted greatly in the revision of this paper.