The Future of New Testament Textual Scholarship: From H. C. Hoskier to the Editio Critica Maior and Beyond (Wissenschaftliche Untersuchungen Zum Neuen Testament) 9783161566622, 9783161566639, 3161566629

Citation preview

Wissenschaftliche Untersuchungen zum Neuen Testament Herausgeber / Editor Jörg Frey (Zürich) Mitherausgeber / Associate Editors Markus Bockmuehl (Oxford) · James A. Kelhoffer (Uppsala) Tobias Nicklas (Regensburg) · Janet Spittler (Charlottesville, VA) J. Ross Wagner (Durham, NC)

417

The Future of New Testament Textual Scholarship From H. C. Hoskier to the Editio Critica Maior and Beyond Edited by

Garrick V. Allen

Mohr Siebeck

Garrick V. Allen, born 1988; 2015 PhD; Lecturer in New Testament Studies at Dublin City University, Republic of Ireland. orcid.org/0000-0003-4428-1775

ISBN 978-3-16-156662-2 / eISBN 978-3-16-156663-9 DOI 10.1628/978-3-16-156663-9 ISSN 0512-1604 / eISSN 2568-7476 (Wissenschaftliche Untersuchungen zum Neuen Testament) The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliographie; detailed bibliographic data are available at http://dnb.dnb.de. © 2019 Mohr Siebeck Tübingen, Germany. www.mohrsiebeck.com This book may not be reproduced, in whole or in part, in any form (beyond that permitted by copyright law) without the publisher’s written permission. This applies particularly to reproductions, translations and storage and processing in electronic systems. The book was printed on non-aging paper by Gulde Druck in Tübingen, and bound by Großbuchbinderei Spinner in Ottersweier. Printed in Germany.

Table of Contents Preface ......................................................................................................... IX

Intellectual History of Textual Scholarship Garrick V. Allen The Patient Collator and the Philology of the Beyond: H. C. Hoskier and the New Testament ........................................................... 3 Juan Hernández Jr. Hoskier’s Contribution to the Apocalypse’s Textual History: Collations, Polyglots, Groupings................................................................... 39 Martin Karrer Herman Charles Hoskier and the Textual Criticism of Revelation.................. 51 Jan Krans Hoskier in the Spiritual World ...................................................................... 69 Jennifer Wright Knust On Textual Nostalgia: Herman C. Hoskier’s Collation of Evangelium 604 (London, British Library Egerton 2610; GA 700) Revisited........................... 79 Peter J. Gurry ‘A Book Worth Publishing’: The Making of Westcott and Hort’s Greek New Testament (1881) ..................................................................... 103

The Status Quaestionis and Future of Textual Scholarship Stanley E. Porter The Domains of Textual Criticism and the Future of Textual Scholarship .... 131

VI

Table of Contents

Gregory Peter Fewster Finding Your Place: Developing Cross-reference Systems in Late Antique Biblical Codices ................................................................ 155 Christina M. Kreinecker Papyrology, Papyrological Commentary, and the Future of New Testament Textual Scholarship ....................................................... 181 Jacob W. Peterson Patterns of Correction as Paratext: A New Approach with Papyrus 46 as a Test Case ............................................................................................. 201 Dirk Jongkind Redactional Elements in the Text of Codex B ............................................. 231 H. A. G. Houghton The Garland of Howth (Vetus Latina 28): A Neglected Old Latin Witness in Matthew .................................................................................... 247 Curt Niccum Hoskier and His (Per)Version of the Ethiopic .............................................. 265 Thomas J. Kraus Ostraca and Talismans: The Story of Two Former Text-Critical Categories and What to do with Them Today .............................................. 283 An-Ting Yi The Critical Apparatus of Stephanus’ Greek New Testament of 1550: Early Printed Editions and Textual Scholarship ........................................... 305 Tommy Wasserman Methods of Evaluating Textual Relationships: From Bengel to the CBGM and Beyond........................................................................... 333 J. K. Elliott Thoroughgoing Eclectic Textual Criticism: Manuscripts and Variants of Revelation.............................................................................................. 363 Jill Unkel Speaking in Tongues: Collecting the Chester Beatty Biblical Manuscripts ... 379

Table of Contents

VII

Editing the New Testament in a Digital Age D. C. Parker The Future of the Critical Edition ............................................................... 395 Catherine Smith Old Wine, New Wineskins: Digital Tools for Editing the New Testament.... 407 Klaus Wachtel The Development of the Coherence-Based Genealogical Method (CBGM), its Place in Textual Scholarship, and Digital Editing ....... 435 Annette Hüffmeier Apparatus Construction: Philological Methodology and Technical Realization........................................................................... 447 Contributors ............................................................................................... 461 Bibliography .............................................................................................. 463 Ancient Sources Index ................................................................................ 503 Modern Author Index ................................................................................. 513 Subject Index ............................................................................................. 521

Preface The title of this volume is intentionally audacious – how can any individual, or even the collective groupthink that his book represents, claim to pronounce or to even glimpse the future of a discipline, especially a discipline as multifaceted as textual scholarship? Even the great Bruce Manning Metzger was ambivalent about addressing such a topic in another context.1 The truth is that we cannot speak definitively about the future of our disciplines, for none of us, we of unclean lips, are soothsayers or prophets (as far as I can tell). But imagining the future by way of analysis of the past, which is the working method of most of the contributions in this volume, provides perspective on the present, which offers a roundabout way of self-critical reflection on our present projects and future aspirations. We imagine the future by analysing the past in an effort to better understand how we came to be in our present. This perspective is the underlying contention of this book and the larger project of which it is an outgrowth, the HoskBib project funded by the Irish Research Council’s New Foundations scheme, with support from Dublin City University’s School of Theology, Philosophy, and Music and the Faculty of the Humanities and Social Sciences. Herman Charles Elias Hoskier, as I argue in my own article, is an ideal figure to use as a baseline platform for this discussion. The goal of this project was never to better understand Hoskier, his life, work, and context – although this is a by-product as most contributors touch on Hoskier as a point of departure for their own studies – but to ask how this semi-obscure figure, an outcast in his time, working in a period of intense disciplinary instability helps us to image what we do not yet know about our field and its critical questions. Hoskier’s work is valuable for this critical self-reflection because it does not lend itself to hagiography; the mixture of his positive and negative contributions to the discipline makes him ideal for seeing the variegated value of our own work. He also provides some solace insofar as poorly designed studies or idiosyncratic projects can sometimes make positive contributions to the discipline beside themselves. This point, at least for me, is comforting. This volume is not interested in resurrecting Hoskier as a paragon of text-critical virtue, and even less an attempt to justify his critical take on the Textus Receptus, his invective polemics, or his allencompassing polyglot theory. But it is interested to probe how technological, 1 Bruce M. Metzger, “The Future of New Testament Textual Studies,” in The Bible as Book: The Transmission of the Greek Text, ed. S. McKendrick and O. O’Sullivan (London: British Library, 2003), 201–208.

X

Preface

ideological, and political changes in textual scholarship from a century ago altered the discourse in the field, and to see if the vast technological changes in media and materials that are part and parcel of textual scholarship today are having similar effects. To this end, the volume explores “textual scholarship” a rather new, but helpful collocation, as a broader phenomenon. Textual scholarship, at least in this iteration, refers not only to the practice of textual criticism and textual analysis of chirographic traditions, but also the integrally linked facets of manuscript literacy, like paratexts and various forms of material philology. Many of the studies in this book deal with explicitly textual issues, but the reality that these texts are transmitted on manuscripts or printed artefacts whose arrangement, format, and larger bibliographic contexts inform analysis, is never far from the broader conversation. Textual scholarship is more than textual criticism, it is a tradition of scholarship that focuses on the most primary of sources in all their textual and material variety. Therefore, contributions explore a variety of topics – from papyrology to paratexts, from Westcott and Hort to Chester Beatty, from text types to the CBGM, from Stephanus to the Garland of Howth – in an effort to understand how our past has brought us to the present, so that we can make some (hopefully informed) guesses about the future of the discipline. What will the next generational task of textual scholarship on the New Testament be? I have some ideas, but perhaps other answers lie among the various critical emphases, and even contradictions and arguments, that appear in the articles that follow this preface. The book is divided into three main sections: (1) Intellectual History of Textual Scholarship; (2) The Status Quaestiones and Future of Textual Scholarship; and (3) Editing the New Testament in a Digital Age. The first of these sections focuses primarily on the context of textual scholarship at the turn of the twentieth century, exploring in rich detail the motivating factors and intellectual pressures that contributed to the production of some lasting critical tools. Much of this section focuses especially on Hoskier and his connections and rivalries with other more prominent figures, but attention is also given to others like Westcott and Hort. This section argues that those who practice textual scholarship are complicated and indeed human and that the intellectual context of our own endeavours is often more complex than we ourselves are aware. Most of our connections and debts to the past are unknown, unacknowledged, and obscured by lacunae is institutional, social, and personal memory. The second section deals more directly with possible avenues of future research. These studies are always informed by study of the past of course, but provide some innovative pathways for engaging the tradition, be that specific textual, material, or intellectual traditions. These studies demonstrate a method, an idea, or phenomenon that requires further investigation. (Calling all PhD Students and Postdocs!) The final section of the book deals directly with a constellation of major editorial projects – the Editio Critica Maior (ECM) – and the challenges it produces.

Preface

XI

These projects, currently being carried out in Münster, Wuppertal, and Birmingham, are leading the way not only in re-evaluating textual relationships, but also the materiality of the tradition, and the ways that both text and material are handled in a digital medium. A major development that the next generation of textual scholars will continue to grapple with and refine is the change to digital modes of editing and analysis. These studies bring this facet of scholarship to the forefront, demystifying the editorial process for the uninitiated and calling out for further engagement from a diverse array of scholarly disciplines. The future of textual scholarship is collaborative. To conclude, I should, as is customary, note that this book is the product of a conference held at Dublin City University’s All Hallows Campus on 28–30 August 2017. In addition to the financial support from the bodies listed above, I would like to thank my colleagues, especially Bradford Anderson, Jonathan Kearney, and Ethna Regan for their support in organising this event. Thanks also go to the presenters and delegates for making the meeting a true learning experience. Dublin 9

Garrick V. Allen

Intellectual History of Textual Scholarship

The Patient Collator and the Philology of the Beyond: H. C. Hoskier and the New Testament Garrick V. Allen The shared goal of this volume is to explore the developing fault lines in textcritical and editorial praxis that pervade the discipline of New Testament studies. These changes in the field are as multifaceted as the reasons for their manifestation, wrought by fundamental transformations in media, changes in theological attitudes toward the wording of the New Testament, re-evaluations of the significance of the history of the tradition, and many other factors. The work of Herman Charles Elias Hoskier (1864–1938), who published under the name H. C. Hoskier but was called Charles by friends, was selected as the lens through which to analyse changing trends in research because Hoskier himself was a transitional figure who was active precisely a century ago. This article unpacks this decision by examining the eccentricities of Hoskier’s life, work, and his contribution to textual scholarship on the New Testament. Hoskier provides a model, sometimes a cautionary one, for grappling with substantial disciplinary instability and for personal dedication to a sometimes thankless vocation, both of which are recurring themes in his body of work. I conclude the discussion with a complete and extended annotated bibliography of Hoskier’s publications, because many (if not all) of his books and articles are out of print and difficult to locate, and because the foibles of his individual outputs are explained, at least in part, when his broader body of work is taken as a whole. The discussion that follows is my justification for selecting Hoskier as an interlocutor for this volume, even though his milieu was populated by many other more distinguished individuals and more adept analytical minds. Hoskier keeps a volume like this from becoming nakedly hagiographic. Hoskier’s transitional status is firstly represented in the contours of his biography. 1 He lived and produced his scholarship in a period defined by change and upheaval. Born in Blackheath, Kent to a prominent merchant banker, Hoskier is much less renowned than his father (also called Herman Hoskier), who made his name shipping cotton past the Union blockade at the outset of 1 For a fuller biographical treatment, cf. Garrick V. Allen, “‘There is No Glory and No Money in the Work’: H. C. Hoskier and New Testament Textual Criticism,” TC: A Journal of Biblical Textual Criticism 23 (2018): 1–19.

4

Garrick V. Allen

the American civil war in 1859, as the head of the largest bank in the world (the Union Bank of London in 1881), and as the financial director of Arthur Guinness Son & Co. Brewery from 1886. We are still able to enjoy a pint of Guinness today, in part, because of the labours of Hoskier’s father. Hoskier benefitted from his father’s prominence, earning a place at Eton College (1878–1881) – although perhaps “earning” is not the right word here in this culture of privilege – and a healthy inheritance, equivalent to nearly $28 million in current terms. 2 Hoskier took his father’s connections and financial backing to Gilded Age Manhattan in the mid-1880s, settling in the East Egg enclave of South Orange, New Jersey and marrying Amelia Wood in a heralded ceremony graced by many a titan of finance in 1888. Following what was by all accounts a successful career in brokerage and finance for the firms Hoskier, Wood & Co. and the L. von Hoffmann & Co., he retired in 1903 to the lucrative career of textual criticism. He briefly returned to finance to co-chair J. P. Morgan’s short-lived Foreign Finance Corporation – a precursor to the World Bank – following the First World War. Although he published his first book A Full Account and Collation of the Greek Cursive Evangelium 604 in 1890, he lamented the lack of relaxation available to him in his career as a financier, even though he was also a noted man of leisure, collecting numismatics, incunabula, manuscripts, objets d’art, and horses. 3 Hoskier valued intellectual pursuits over and above financial gain and security, rejecting the dominant ethos of übercapitalist latenineteenth century New York. A career change at the age of thirty-nine marks the first major transition in Hoskier’s life. He never held an academic post, but appears to have lived off the wealth that he had amassed as inheritor and financier. The level of Hoskier’s idealist commitment to his personal convictions is also on display in his participation in what I suspect is the defining event in his life: the First World War. After producing a number of publications from 1910–1914, Hoskier’s bibliography has a five-year gap that reflects his volunteer service in an American detachment of the French Ambulance corp. He saw combat, was twice wounded on the Western front (injuries that slowed his rigorous scholarly work), awarded the Croix de Guerre, and made member of the Legion of Honour. His volunteer service was motivated by his family background on one hand – he had deep connections to France (his uncle Emile £211,027 7s. 6d. according to John Orbell, “Hoskier, Herman,” in Oxford Dictionary of National Biography (2004) (https://doi.org/10.1093/ref:odnb/49026) [accessed 19 February 2018]. 3 Hoskier’s financial issues famously led to the selling of some of his collections. Cf. A Catalogue of a Portion of the Valuable Library of H. C. Hoskier, Esq. of South Orange, New Jersey, U.S.A. (London: Sotheby, 1908) and Auctions-catalog einer höchst bedeutenden Sammlung Griechischer und Römischer Münzen, Collection H. C. Hoskier (Munich: Hirsch, 1907). 2

The Patient Collator

5

Hoskier was a prominent French banker, for example) and boasted Serbian heritage – and by his son’s own zeal for the justice of the Allies’ cause on the other. 4 Ronald Wood Hoskier, Hoskier’s son and a student at Harvard, was the first American fighter pilot to perish in the war. He served in the Escadrille de La Fayette and was shot down over San Quentin on 23 April 1917. 5 Both Hoskier and his family continued to advocate for the victims of the war long after its conclusion. 6 Following the war, Hoskier resumed his text-critical work. Closely aligned with Henry A. Sanders of the University of Michigan, as well as other prominent scholars like J. Rendel Harris with whom he left a voluminous correspondence,7 Hoskier donated much of his library, manuscripts, and coins to the University of Michigan library, benevolence that earned him an honorary Master of Arts in 1925 and an appointment as Honorary Curator of the University’s Museum of Archaeology (1929). 8 He only accepted the latter position after receiving assurances that it required no actual work. He moved from New Jersey 4 Cf. the introduction in Literary Fragments of Ronald Wood Hoskier 1896–1917 (Boston: McKenzie, n.d.), 5–8, which I suspect was edited by H. C. Hoskier, although the editor is anonymous and it bears no date. The copy that I have consulted was donated by H. C. Hoskier himself to the University of Michigan Library. 5 Cf. “American Flier Killed in Combat,” New York Times, 25 April 1917. 6 Amelia Wood Hoskier, for example, wrote a letter to the editor of The New York Times that was published on 8 February 1926, advocating for French Refugees. 7 Hoskier carried on a long-lasting and rather intimate correspondence with Harris touching on a range of issues, including Harris’ survival of a German torpedoing off Corsica, Hoskier’s experience at the Western front, text-critical concerns, and serious interest in spiritualism, although it is not clear that Harris reciprocated this interest (Hoskier refers to Harris’ “discreet silence” on the matter, which to me indicates that Harris was more interested in the text-critical aspects of their conversation, Birmingham Library DA21/1/2/1/25/7, 17 May 1922). Among Hoskier’s correspondence with Harris exists a short work by Hoskier, unpublished, reflecting on his appreciation of Patience Worth, a supposed spirit of a seventeenth century woman in a long-term communication with a Mrs Curran of St. Louis (Appreciation of “P.W.” by an Outsider, signed H.C.H.), along with a copy of a book composed by Patience through the medium Mrs Curran (DA21/1/2/1/25/3, 19 December 1921). The earliest letter to Harris (7 May 1917) is characteristic of Hoskier idiosyncrasies, discussing Harris’ “escape of the Boche torpedoes,” the death of his son in a dogfight, sarcastic thoughts on the news that C. R. Gregory was a German lieutenant on the Western front, and hope that pro-German academics in Great Britain would “see the light” (DA21/1/2/1/25/1). Harris also uses Hoskier as a sounding board for expensive purchases of manuscripts, which Hoskier at times offers to fund on his behalf (DA21/1/2/1/25/4, 20 January 1922). Hoskier also informs Harris that he has met Patience in person, since the family that received the spirit communication named an adopted child after the spirit (DA21/1/2/1/25/6, 22 April 1922). Their letters are not always entirely friendly in tone (e.g. DA21/1/2/1/25/15, 25 March 1928). 8 The British Museum was also a benefactor of Hoskier’s donations. Cf. H. R. Hall, “Other Donations to the Egyptian and Assyrian Department,” The British Museum Quarterly 5/2 (1930): 48–9.

6

Garrick V. Allen

to Jersey in the Channel Islands in 1927, where he travelled frequently to France. In June 1938, three months before his death on 8 September, Hoskier was awarded an honorary ThD from the Universiteit van Amsterdam. Amendments to his will after his wife’s death in 1929 (amended 2 August 1935) indicate his poor financial state. He requests not to be transported to his family plot in South Orange, but to be buried in a simple box on Jersey. He also notes that his son Walter in fact owes him £359 and that this amount should be deducted from his inheritance should there be any. Not to dissuade potential PhD students, but Hoskier is proof that one rarely gets rich on textual scholarship. 9 The many significant transitions of Hoskier’s life are deeply connected to his body of academic and philosophical work. The primary aim of Hoskier’s activities was, even to the end of his life, to reclaim the value of the Textus Receptus that Westcott and Hort had dethroned decisively, with much assistance from distinguished predecessors, in Anglophone scholarship in 1881. Hort’s introduction in particular satisfied Hoskier’s need for a foil, even though he adopted many of the presuppositions that stand behind Westcott and Hort’s method, including the idea that textual criticism properly done can fully and securely recover the “original” or “true” text of the New Testament, that scribal proclivities were always aimed toward the mechanical reproduction of texts, especially sacred ones, and that the goal of editorial work was the identification and removal of accreted errors. 10 Hoskier is often identified, and therefore dismissed, with John William Burgon (1813–1888), the indefatigable champion of all things traditional, who, in addition to being a thoroughgoing polemicist and righteous supporter of the Textus Receptus and the authenticity of the long ending of Mark, fought to keep women out of Oxford and considered changing student housing policies since some residences employed women who had previously been incarcerated. The introduction to Hoskier’s Greek Cursive Evangelium 604 (1890) reinforces his connection to Burgon, since it contains an anecdote that opens at midnight with Burgon on the staircase of his Chichester home, recounting his assertion that “as certainly as the sun will rise to-morrow morning, so surely will the traditional text be vindicated.”11 Burgon and his acolytes perceived the paradigm9 In a letter to J. Rendel Harris, Hoskier notes that “I have completely ruined myself in Jersey & would accept a few old piece of furniture from the manor” (DA/21/1/2/1/25/22, 14 March 1929). 10 Cf. B. F. Westcott and F. J. A. Hort, eds., The New Testament in the Original Greek: Introduction and Appendix (New York: Harper & Brothers, 1882), 1–3, 6–7, 24–30. Although Hort in particular served as Hoskier’s nemesis, he could also have benefitted from the example of Westcott and Hort’s partnership, which they describe thusly: “No individual mind can ever act with perfect uniformity, or free itself completely from its own idiosyncrasies: the danger of unconscious caprice is inseparable from personal judgement” (here 17). 11 H. C. Hoskier, A Full Account and Collation of the Greek Cursive Evangelium 604 (with two facsimiles) (London: David Nutt, 1890), v. Hoskier is also referred to as a “scion

The Patient Collator

7

shifting edition of Westcott and Hort and its influence on the Revised Version as outright assaults on orthodoxy, tradition, and divine inspiration, often responding as combatants in a holy war. 12 Traces of this influence are felt in Hoskier’s pre-war writings, but Hoskier never considered himself one of Burgon’s followers, although they did share some critical goals and suppositions. The transitional nature of Hoskier’s project is on full and clear display in the ways that Hoskier breaks from Burgon, especially in his changing rhetorical strategies for vindicating the Textus Receptus. Hoskier’s academic work can be divided into pre and post-war epochs; the former defined by polemical attempts to vindicate the Textus Receptus through rigorous textual data and invective prose, and the latter characterised by methodological devotion to digesting the totality of the evidence, although the polemical edge of his rhetoric never entirely dissipated. The war changed the tenor and tenacity of Hoskier’s project; it is no accident that his work that has endured was produced after the war. The obvious pinnacle of Hoskier’s pre-war rhetoric is found in the two-volume Codex B and its Allies: A Study and an Indictment (1914),13 which constitutes an attempt to undermine Codex Vaticanus as a witness to Hort’s neutral text, as well as Hort’s methodological principles. 14 The critique fails in its of the Burgon school” in a review of Henry A. Sanders’ New Testament Manuscripts in the Freer Collection, in The Biblical World 42 (1913): 59–69 (here 59) by a certain A. D. Nonetheless, the relationship between Hoskier and Burgon is not so clear-cut, and although Hoskier continued to hold affinities for Burgon’s quest to justify the Textus Receptus, there is no evidence that he did so out of allegiance to Burgon, even though he did have personal knowledge of Burgon’s library, noted in a letter to J. Rendel Harris (DA 21/1/2/1/25/2, 6 October 1920). Cf. Daniel B. Wallace, “Historical Revisionism and the Majority Text Theory: The Cases of F. H. A. Scrivener and Herman C. Hoskier,” NTS 41 (1995): 280–85: “neither Scrivener nor Hoskier followed in Burgon’s steps” (here 281). See also the introduction to Annette Hüffmeier’s article and also both Tommy Wasserman and Jennifer Knust’s articles in this volume. 12 Cf. The Oxford Debates on the Textual Criticism of the New Testament (London: George Bell, 1897) and John William Burgon, The Revision Revised: Three Articles Reprinted from the Quarterly Review (London: John Murray: 1883). 13 Codex B and its Allies: A Study and an Indictment, 2 vols. (London: Bernard Quaritch, 1914). 14 Westcott and Hort’s method for evaluating witnesses led to an extremely high valuation of readings that belonged to a document that they felt was usually correct, especially if it was ancient, aesthetically beautiful, and in uncial script. Cf. The New Testament in the Original Greek, 10–11, 30–9, 60–2, 232 especially the section on “Internal Evidence of Documents” and “Internal Evidence of Groups,” where the valuation of the overarching textual character of a particular witness or group of witnesses, evaluated partially by genealogical reconstruction can at times override intrinsic and transcription probability. A large portion of Hort’s introduction (93–179) is devoted to establishing the hierarchical relationships between his main (and ancient) textual families, the neutral, Alexandrian (α), Western (β), and Syrian (δ), of which the neutral text is clearly the group that takes priority, due to the

8

Garrick V. Allen

virulence and lack of structure. The first pages, enveloped in legal language of indictment and accusation, aim to prick “the bubble of codex B,” and lay “hundreds of separate accounts” (apparently on a reading by reading basis) against Westcott and Hort. 15 The conflict is personal, as Hoskier’s confrontation of Alexander Souter – on the first page of the preface! – demonstrates. He notes that despite a negative review of his Concerning the Genesis of the Versions of the N.T. (1910), Souter ended up expressing gratitude for my collations…but added some very strong advice to hold my tongue as regarded commenting on the evidence so painfully accumulated…I refuse to be bound by such advice. I demand a fair hearing on a subject very near my heart, and with which by close attention for many years I have tried to make myself sufficiently acquainted to be able and qualified to discuss it with those few who have pursed a parallel course of study. 16

Hoskier never got his “fair hearing,” due in large part to the ad hominem and almost panicked nature of his discourse. If Edgar J. Goodspeed described Hoskier’s earlier two volumes on the versions as “a mass of individual textual notes, with an occasional paragraph of bold generalization,”17 then the same can easily be made of Codex B and its Allies. And the work is indeed just so: a series of collations designed to undermine the text of Codex B as a legitimate witness of the “true text” and support some other of Hoskier’s idiosyncratic pet theories, like the deep antiquity of the versions and the idea that Mark was initially composed in both Latin and Greek simultaneously. In contrast, the modus operandi of Hoskier’s work changed fundamentally following the war in a way that still animates text-critical projects like the Editio Critica Maior that emphasise comprehensiveness. From 1919 until his death in 1938, Hoskier retained an interest in editing Greek and Latin manuscripts that he perceived to preserve especially important texts, like The Text of Codex Usserianus 2., or r2 (“Garland of Howth”) (1919) and The Complete Commentary of Oecumenius on the Apocalypse (1928), or other traditions that he found interesting, like De Contemptu Mundi: A Bitter Satirical Poem of 3000 Lines upon the Morals of the XIIth Century by Bernard of Morval Monk of Cluny (1929). But a larger overarching project, centred on the New Testament Apocalypse took pride of place in his trajectory, and it remains the most

periphrastic and interpolatory nature of some Western readings and the conflate nature of a number of Syrian readings, not to mention the fact that, according to Hort, the Alexandrian grammatical schools would have kept “a more than usual watchfulness over the transcription of the writings of the apostles” (p. 127). Codex B is Hort’s preeminent witness to the preSyrian neutral text (pp. 150–51, 170–72, 210, 220–60). 15 Codex B, 1.i. 16 Codex B, 1.i. 17 Edgar. J. Goodspeed, “Review: Hoskier’s Study of the New Testament Versions,” AJT 16 (1912): 652–54 (here 653).

The Patient Collator

9

important aspect of his body of work. Although he published a number of studies on Revelation, including a five-article series in the Bulletin of the John Rylands Library (1922–1924), his two-volume magnum opus Concerning the Text of the Apocalypse (1929) represents a herculean individual effort to compile a master collation of every known witness of the book of Revelation in an edition organized around the 1550 Stephanus text. Polemical interjections remain essential to the fabric of Hoskier’s discourse in the post-war period, but methodological principles of patience and data aggregation replace pure textual rhetoric. Hoskier never achieved Lachmannian “scientific” proof of the “originality” of the Textus Receptus for the Apocalypse – his underlying and sometimes stated goal – but he did provide a valuable resource for textual scholarship that accurately supplements hand editions of the New Testament and provides access to now-lost artefacts and their texts (e.g. GA 241, Hoskier 47). Concerning the Text did not rescue the Textus Receptus, but undermined it further by clearly demonstrating the fundamental uncertainty of many places in the tradition and the peculiarities of many individual witnesses. The methodological purity of Hoskier’s post-war programme continues to inform textual criticism on the New Testament, which is now grappling with basic changes in media, digital infrastructure, and the requirements of funding bodies, even if his influence remains primarily subconscious. The production of collaborative digital workspaces and electronic transcriptions is now making it possible for editors and scholars to once again build comprehensive sets of data for New Testament works beyond Revelation. 18 Hoskier too, utilised technological innovations to produce his lasting contribution. Using his once vast personal financial resources, he purchased photographs of manuscripts from far-flung libraries and personal collections, plying the improvements in the cost of photographic technology in the early twentieth century to his advantage, while at the same time complaining about the prices that libraries charged for reproductions. Continuing changes in modern text-critical praxis are enabled by technological changes in the field, like the burgeoning archives of quality digital images, published transcripts, and digital editions. Although his project ultimately failed in its stated goals, Hoskier’s working method and reliance on modern technology anticipated more sophisticated modes of research, many of which are visible in the work of the contributors to this volume.

18 The Editio Critica Maior of Revelation is currently in production at the Institut für Septuaginta- und biblische Textforschung at the Kirchliche Hochschule Wuppertal/Bethel under the supervision of Martin Karrer. Cf. Marcus Sigismund, “Die neue Edition der Johannespokalypse: Stand der Arbeit,” in Studien zum Text der Apokalypse II, ANTF 50, ed. M. Sigismund and D. Müller (Berlin: De Gruyter, 2017), 3–17, and Darius Müller, “Zur elektronischen Transkription von Apokalypsehandschriften: Bericht zum Arbeitsstand,” 19– 30 in the same volume.

10

Garrick V. Allen

The final piece of evidence for the transitional nature of Hoskier’s complex life is the series of philosophical treatises that he authored in the late 1920s and 1930s. As the flustered polemics of Codex B and its Allies demonstrates, prewar Hoskier is motivated by an explicit desire to uphold what he perceived as orthodox Christian adherence to a traditional text form of the New Testament. However, following the war marked by his own service and the tragic death of his son, the apologetic strain in his academic writings ebbs and a moral call for humanity to realise its own essential deific essence gains prominence in these writings. His philosophy is important because it is never entirely divorced from his academic work, as Jan Krans’ article in this volume incisively demonstrates. Hoskier appeals to mediums and spirit guides to enforce his textual decisions and even includes readings created by a spirit at a nineteenth century séance. He also comments in a copy of a letter sent to J. Rendel Harris that he is “in slight touch himself with the other side.” 19 These moves are undergirded by his philosophic ideals, laid out in his panentheistic treatises that were part of a much larger re-enchantment of the world known as theosophy, a movement that garnered a significant amount of popular interest after the war in Europe, even though its origins can be traced to the early nineteenth century. Two of these works were written under the appropriately esoteric pseudonym Signpost. 20 Although these writings, at times, are prescient in their suggestion that the world was hurtling toward another major conflict, the prose is often as incohesive and its message is incoherent.21 In his self-proclaimed creed in the form of a prayer, What is Nirvana?, Hoskier builds his case using familiar Christian language. For example: And so, Great Father – see, I dare to call Thee Father – taught by Him of lowly Nazareth, – Thine Angel-Messenger, – Gabriel, God-man – I bow my head I bow my knees, I bow my

Copy of a letter to Mrs Curran 1 December 1921 (DA/21/1/2/1/25/3). When D. C. Parker, An Introduction to the New Testament Manuscripts and their Texts (Cambridge: Cambridge University Press, 2008), 231 asks “was he [Hoskier] serious” about including readings from spirit communications in the apparatus of Concerning the Text, the answer must undoubtedly be affirmative. 20 H. C. Hoskier, In Tune with the Universe (London: Rider & Co., 1932) and Hoskier, The Back of Beyond (London: C. W. Daniel, 1934), although Hoskier wrote the preface for In Tune with the Universe in his own name (in which he declares that “Signpost has lifted a corner of the veil by his differentiation between the vibrational world and the non-vibrational”), and although H. C. Hoskier is included on the title page of The Back of Beyond. 21 See for example, Hoskier’s comments on Japan in 1930: “If I turn my eyes to Nippon, I am but bewildered. Is the backwater of to-day but the maelstrom of to-morrow? Who knows?” (‘The Bronze Hoses’: A Comment on the Prose-Poem of Amy Lowell [Portland, ME: Mosher Press, 1930], 14). This appraisal is commensurate with his pessimistic appraisal of nation states at this time (pp. 13–8). In In Tune, 120–21 Hoskier also calls for a body like the United Nations. Cf. also Back of Beyond, 62–70. 19

The Patient Collator

11

dust-clad spirit and acknowledge Thee: All-Good, All-Wise, All-Just, All-True, All-Pure, to be…………….Me…………..And I…………Thee. 22

Hoskier contends that all that exists is really one living eternal organism and that human conflict of all forms is inimical to the all-encompassing “All-Life.” The goal of human activity, according to his theosophonic philosophy, is to transcend the “vibrational world of effects” to what he calls the “back of beyond” or the “world of Causes lying behind it,”23 a place accessible by looking inside oneself in an effort to locate one’s own “godhood” where we find “our coequality with that Essence.” 24 Ultimately, for Hoskier, once we realise that we are “essentially deific,” we are free to escape to the world of ultimate causes; 25 since all are one and one are all, death has no significance and organized religion is a false path to “Wholeness.” 26 This is borne out also by his burial wishes recorded in his will: I declare that I die in the certainty of the continuity of life both molecular and spiritual; attached to no particular school of thought nor to any particular religion, but in love with all man’s striving towards the recognition of his birthright as part of an indivisible All-life, which in reality constitutes the Whole, so that he cannot ‘die’ in any sense whatever.

Many have scoffed at Hoskier’s naïveté for believing in spirits and theosophy and have thereby written-off his contribution to the discipline. But it is important to remember that his interest in the occult was shared by many, especially following the rebirth of theosophy in Europe after the trauma of the war. These aspects of Hoskier’s work are not disqualifying, but instead further illuminate his context and influences. The persistence of resolve found in some corners of Hoskier’s scholarship are identifiable also in his life. There is no division for him between academic work and the working out of the complexities of life and the ultimate fate of humanity. In this sense, Hoskier views textual scholarship as essential to understanding the world and as a basic foundation for life. It is more than academic tedium and even more than an essential preliminary task necessary for interpretation; it is not just an essential discipline in the humanities, but an essential discipline for humanity, whether or not we agree with his philosophical proclivities. The detail-oriented rigor and persistence required for textual criticism were characteristics that Hoskier perceived as essential to living. Textual scholarship mirrors life, and textual rhetoric is the language of science. Despite his idiosyncrasies and foibles, all textual scholars can see parts of themselves in Hoskier’s principled integrity and earnestness of conscience, H. C. Hoskier, What is Nirvana? (Portland, ME: Mosher Press, 1930), 12. E.g. In Tune, 1. 24 “You don’t look up, but you look within.” “No longer Three in One, but All in One, and we are not only of it or a part of it, but It Itself” (In Tune, 5, 7); cf. Back of Beyond, 21. 25 Cf. Back of Beyond, 41–5 on how to accomplish this task, according to Hoskier. 26 Back of Beyond, 28–9, 53–7, 70–82. 22 23

12

Garrick V. Allen

even if we ultimately exclude spirit communications from our apparatuses and decide to keep our philological efforts to this side of the Back of Beyond. The many transitions in Hoskier’s life and work coalesce to mark him as a transitional figure in textual scholarship on the New Testament. He stands between the Lachmannian sensibilities of the nineteenth century and the radical insecurity of textual traditions and editorial decision that defined some quarters of editorial work in the twentieth and twenty-first centuries.27 A basic tension of his medial position is his insistence in many publications that one must first gather all data before making any interpretations or judgments, even though he has an obvious agenda and makes interpretive judgements throughout. The contradiction reinforces the concept that, even if one claims otherwise, all textual scholarship is interpretive, all collation is rhetoric, and every manuscript has a voice as a legitimate witness to the tradition. Hoskier could not have admitted this, but his method and body of work speaks plainly. This is precisely why using him as a frame to re-imagine the discipline of New Testament textual scholarship is both legitimate and right. He is not the most famous or decorated scholar of his generation; he never held an academic post, earned a university degree, relieved a funding body of its reserves, had a PhD student, won any awards, or sold many books – many of the copies of his books that I have examined at different European libraries were in fact donated by him personally to these institutions. But he created the space through his methodological emphases for important projects and trends in the field that are finally being realised today. Hoskier provides space to imagine what we do not yet know – to think about how the choices made by scholars and editors today will change the discipline in the future.

27 Cf. Sebastiano Timpanaro, The Genesis of Lachmann’s Method, trans. G. W. Most (London: The University of Chicago Press, 2005), esp. 84–9, 119–38, sections that enlighten Hoskier’s intellectual context, his simultaneous radical departures from it, and his unknowing accession to his own time.

The Patient Collator

13

Annotated Bibliography28 A Full Account and Collation of the Greek Cursive Evangelium 604 (with two facsimiles). London: David Nutt, 1890. 29 Hoskier’s first book is both misleading and instructive for understanding his larger body of work. The preface opens with an anecdote of a conversation with John William Burgon about the vindication of the Textus Receptus and the book is fawningly dedicated to Burgon. 30 However, as I mentioned above, the relationship between Hoskier’s work and Burgon is complicated, and although Burgon is the better rhetor, Hoskier is the more creative scholar. The book is instructive insofar as it introduces Hoskier’s dominant mode of discourse (collation and textual notes) and an early insight into his larger, but developing project (the scientific vindication of the Textus Receptus, against Hort [cf. pp. cxv–cxvi], through study of the text of neglected New Testament manuscripts). The main argument that the evidence in the book is designed to support is that the text of the fourth and fifth century uncials is corrupt, a point 28 Hoskier is also credited with another book that is not included in this bibliography because it is comprised of tables of calculations for identifying the prices of securities in arbitrage sales between the New York and London stock exchanges. H. C. Hoskier, Table of Arbitrage Parities between New York and London (London: Richard Clay and Sons, 1892). This volume has nothing to do with textual scholarship, but it does demonstrate Hoskier’s interest in careful data aggregation, and this type of work appears to be the equivalent of textual criticism for the financial sector, although I imagine that it is much more lucrative. Another, unpublished article, entitled “Λόγια or the ‘Oracles of God’ and χορηγία or the Supply of the Spirit by Direct Intervention between God and Man,” was sent by Hoskier to J. Rendel Harris on 22 June 1922 (DA21/1/2/1/25/9–10), seeking advice for a publisher. One was not found for this article that, among other things, leans heavily on the comparison of prayer and the new-fangled wireless. 29 Cf. the positive review of Appendix C of this work in E. Nestle, “Some Points in the History of the Textus Receptus of the New Testament,” JTS 9 (1910): 564–68; the criticism in Isaac H. Hall, “The Title-Page of the Elzevir Greek New Testament of 1624,” JBL 10 (1891): 147–50; and A. Plummer’s review in The Classical Review 4/10 (1890): 478, who refers to the work as “a labour of love.” Wilhelm Bousset, Textkritische Studien zum Neuen Testament, TU 9/4 (Leipzig: Hinrichs’sche, 1894), 118–19 argues that Hoskier refuses to comment on the significance of his data, but Hoskier rebuts that it would be too hasty in Concerning the Genesis of the Versions, ix–x. In Concerning the Text, 1.xxxviii he retorts again: “This is foolish. We have had too many cheap and hasty deductions from insignificant or insufficient data.” Bousset, “Textkritik II,” Theologische Rundschau 17 (1914): 187–206 (esp. 199–200) also critiques Hoskier. He begins his review thusly: “Ein Textkritiker, der ganz einsam und abseits von den gewöhnlichen Wegen seine Bahnen zieht, ist Hoskier.” 30 See also p. vi: Burgon’s “Magnum Opus, had he lived to edit it, would have for ever vindicated his reputation, his views, his methods, nay, the very manner of expressing himself, if by a too decided front he had made himself enemies and curtailed the extent of his hearing for a time.”

14

Garrick V. Allen

reinforced by the continued unearthing of texts that that disagree with them in substantive ways, especially when these new texts are independent in their textual affiliations. Hoskier goes about this by examining the text of GA 700 and comparing it to readings in earlier uncial witnesses, an appraisal that illustrates to his satisfaction that “the compilers of such [uncial] codices were, to an enormous extent, their own critics, leading them to altogether independent treatment of the Sacred Text” (p. xiv). The age of the manuscript does not guarantee the quality or age of the text (p. xv), and the text of GA 700 is of higher quality than more ancient exemplars. Hoskier comments first on the palaeographic, codicological, and scribal profile of the manuscript, as well as offering corrections to antecedent collations, as he is often wont to do (pp. i–xxviii). This is followed by a list of singular readings, counting 270 in this Gospel manuscript, and a list of readings poorly attested elsewhere without comment (pp. xxix–cxv). The introduction gives way then to the collation of the manuscript in toto against Stephanus’ 1550 edition (pp. 1–43). The book concludes with ten appendices that describe other manuscripts, correct some of Scrivener’s collations, collate various printed editions, contain library reports, and other text-critical concerns. The book also provides some delightful notes on the importance of collating and careful study of the documents, 31 alongside some invective statements that anticipate the full-blown polemics of Codex B and its Allies (e.g. p. xvi). Appendix J – a note on 1 Tim 3:15 – is a reprint of an article Hoskier published in Clergyman’s Magazine in February 1887. 32 See also Jennifer Knust’s article in this volume. The Golden Latin Gospels: JP in the library of J. Pierpont Morgan (formerly known as the “Hamilton Gospels” and sometimes as King Henry the VIIIth’s Gospels) now edited for the first time, with critical introduction and notes, and accompanied by four full-page facsimiles. New York: Private Printing, 1910.33 This volume is a sumptuously produced edition and discussion of a seventh or eighth century purple bicolumnar Vulgate manuscript that the famed financier 31 E.g. p. vi: “Though seemingly dry and laborious work (and of a truth it is the latter to a large extent) some of the most wonderful truths, some of the most interesting problems present themselves to his mind as letter by letter, line by line, and page by page the patient collator toils along slowly at his task.” See also p. xxi: “Die grösste Frucht unserer Arbeit ist oft die Arbeit selbst.” 32 Cf. also Wilbur N. Pickering, The Identity of the New Testament Text III (Eugene: Wipf & Stock, 2012), 71. 33 Cf. Hoskier’s correspondence with Rudyard Kipling in November 1910 on the provenance of the manuscript’s scribe in T. Pinney, ed., The Letters of Rudyard Kipling, vol. 3 1900–10 (London: Palgrave Macmillan, 1996), 464–66. Cf. reviews by Edgar J. Goodspeed,

The Patient Collator

15

J. Pierpont Morgan purchased from Hoskier’s own dealer and sometimes publisher Bernard Quaritch of London. Only 200 hundred copies were printed in a private printing arranged by Morgan (printed by Frederic Fairchild Sherman), including some colour images of the manuscript printed on the finest Italian paper with interlocking sea horse watermarks that bear the text “FFS Italy.” The volume is an artistic work regardless of its contents; the beauty of the printed edition corresponds to the aesthetics of the manuscripts it aims to represent. In his review, Sanders refers to it as “a book-lovers’ prize” (p. 218). The introduction to the volume, however, is as arduous to read as it was onerous to construct. It attempts to localise the production context of the manuscript, which Hoskier locates in the UK or Ireland, even though JP – the siglum he invents for the manuscript – is in “a class by itself as regards English and Irish MSS” (p. xv). The arguments of the volume are twofold: (1) to acknowledge the high value of JP’s text (the stemma on p. xcviii emphasises the significant place this witness has in Hoskier’s reconstruction of the tradition); and (2) to demonstrate that the Greek uncials were influenced by readings particular to JP and its tradition (e.g. pp. liv–lxvii). 34 Hoskier’s overriding polyglot theory takes shape here. The means of making these arguments is through data in the form of extensive collation, which Hoskier makes for each Gospel in the manuscript, even though much of the data is repeated in his lengthy introduction (116 pages, followed by 71 pages of “Preliminary Remarks”). The collations are made against the Sixto-Clementine Vulgate of 1592 and it includes readings from dozens of other witnesses (see pp. 75–8; collations pp. 80–344). An interesting feature of the introduction is that Hoskier is sometimes specific about the mechanics of how the polyglot theory plays out in process of copying. For example, he imagines the working conditions of the scribe of Codex Sinaiticus thusly: “Project yourself in theory into the cell or cabinet de travail of the scribe of ‫ א‬about A. D. 400. You find him surrounded by his library at his desk. You think to find him close to the Apostolic autographs. But “The Golden Latin Gospels,” The Biblical World 38 (1911): 67–70, which, again, is quite negative, noting, “in all this one feels that Mr. Hoskier, in his natural enthusiasm for a notable and beautiful manuscript, has been carried too far” (p. 70). In particular Goodspeed criticises Hoskier’s identification of forty different scribal hands in such an expensively crafted manuscript; he does, however, admit Hoskier’s “extraordinary learning and diligence,” despite his “discursive and casual, rather than orderly” working method (pp. 69–70). Cf. also Henry A. Sanders’ review in The American Journal of Philology 32 (1911): 218–20 and Arthur H. Weston’s review in Classical Philology 8 (1913): 378–82, who is pedantically critical of Hoskier’s paragraphing and the linguistic peculiarities of his prose. 34 This basic goal of the project is explicated in the subtitle to the collation, which includes the phrase “etiam in multis locis explicatur de testimonia codd.Graec.” Understanding the shape of the Greek text is Hoskier’s overriding goal in examining Latin manuscripts like JP.

16

Garrick V. Allen

the retrospect of 350 years to him seems just as great as that of 1850 years today to us…But approach closer and watch him at his task. He sits with a handsome volume open on his left. As far as we can see, it is bicolumnar, but his immaculate sheets of thin white vellum overlie parts of it, and possibly cover a third and yet a fourth column, containing Syriac and Coptic in parallel columns. At any rate, what he appears now to be transcribing from is GraecoLatin in separate columns, the Greek in the left-hand column. He is at John ii:14, and as his eye goes to the Greek column, he reads βοας και προβατα, the προβατα in the line below; προβατα then is the last thing in his mind. As his eye passes over the Latin he sees oves, the last thing on his retina. What more natural than for him to invert and write προβατα και βοας” (pp. lxiv–lxv). This volume is in many ways the fountainhead of the larger project of identifying polyglot interference in the uncials, a project patronised in this instance by the prominent Morgan and his manuscript. This volume is a direct outworking of Hoskier’s relationships cultivated on Wall Street. Such patrons are rare and beautiful butterflies. Concerning the Genesis of the Versions of the N.T. Remarks Suggested by the Study of JP and the Allied Questions as Regards the Gospels. 2 vols. London: Bernard Quaritch, 1910.35 This book stands in the medial position between Hoskier’s edition of JP (see previous entry) and his transcription of Codex Usserianus 2 (cf. esp. pp. 109– 340). Using his edition of the Golden Latin Gospels as a starting point for his discussion, Hoskier makes a number of critical points that crop up in other works, including his negative appraisal of the text of B and other uncials (e.g. pp. 387–88), criticism of Hort (e.g. pp. viii–ix, 57–60, 97), his polyglot theory, the importance of minuscule witnesses (pp. 61–3), and polemic rebuttals of perceived opponents. 36 This volume is essentially an aggregate of multiple 35 Cf. Henry A. Sanders, “Hoskier’s Genesis of the Versions,” American Journal of Philology 33 (1912): 30–42, who accepts the basic polyglot principle of Hoskier’s theory, but rejects his assertion for two early concurrent forms of Mark; and an anonymous, mostly positive, review in The Academy and Literature 82 (1911): 107. 36 Many, but not all of these attacks are religiously charged. For example, Bousset’s critique of Hoskier’s 1890 book for his inaction to drawing conclusions is “foolish” (p. x), because Hoskier is building a cumulative case. On Burkitt: “This is truly unscientific of Professor Burkitt, and he must know a great deal better than that” (p. 61); on von Soden’s volumes on Cyprian: “I may be very stupid, but I have failed to glean anything new from them, and I do not see in what direction his labours tend” (p. 78); on Albert Edmunds: “Mr. Edmunds is apparently blissfully ignorant, when he write himself down ‘as a Christian believer though attached to no sect or Church whatever,’ that he is in Marcion’s class, and is returning to the vain gods of the second century” (p. 107c); on Tischendorf and WestcottHort’s use of text types: “we have used the foregoing example, and have illustrated it as

The Patient Collator

17

studies that all coalesce around the interrelations of the versions and the Greek text – especially Latin traditions – and their deep antiquity, construed at times as nearly concurrent compositions alongside the Greek. If one was forced to identify a thesis, it would be that “there is abundant evidence that the mistakes in ‫ א‬and D, with other like survivals in other Greek and Latin MSS., are due to the use of a polycolumnar polyglot in copying…whichever way we turn we are met with polyglots” (pp. 15–6; cf. p. 75). Hoskier’s reconstruction of the materiality of the tradition forms his view of its textual history. This book also provides the first evidence of his long-term work on the Apocalypse. He notes that as of 1910, he had already collated over one hundred manuscripts and that the text of many of the later minuscules go “far back of ‫( ”א‬p. 17), because they bear the influence of trilingual (Syriac-Graeco-Latin) or even perhaps quadrilingual (Syriac-Graeco-Coptic-Latin) manuscripts that predate ‫( א‬p. 23). His work on the Apocalypse is deeply connected to his other pet theories. He is correct, however to critique a range of suppositions in textual criticism, many of which have also been critiqued in recent discourse, like the dissolution of geographically bound textual families (p. 24)37 and the inflexible application of rules like the preference of the shorter reading (pp. 375– 76). The book is valuable for understanding Hoskier’s programme not only in terms of content, but mode of argumentation, which is, once again, eminently textual in orientation. The main body of the work is a collation that illustrates the relationship between r2 and other Latin texts, attempting to identify the witness that best preserves the archetype of the tradition and to argue for the close relationship between the Latin and Syriac, which explains his turn to the Diatessaron following lengthy discussion of the Latin (pp. 341–69). Volume 2 is over 400 pages of appendices, comprised primarily of collations of various manuscripts (e.g. the books of Dimma and Moling, among other Latin manuscripts) and comments on some recent text-critical publications and the medical discourse on the blood and water that flowed from Jesus’ side at his execution (John 19:34). This is a prime example of the rhetoric of text and data that Hoskier employs in service to those without access to the manuscripts, but also in service to his arguments and polyglot theory. The versions, especially when they agree with the Textus Receptus, are more valuable witnesses to the text than the early Greek uncials and papyri. This point comes through clearly in Hoskier’s note on W referring to readings that he identifies as “IInd or IIIrd century glosses” (p. 2.379): “that the Church knew what she was doing when she disallowed the reproduction of such unscriptural addenda, and her wisdom is profusely as space will allow, in order to show in how senseless a way Tischendorf and Westcott-Hort clung to ‘type’ as a fetish, though opposed to good scholarship and common sense and the consentient voice of the document” (p. 395). 37 This was applied, not surprisingly, especially to Hort’s Western text (e.g. p. 55). Cf. also Concerning the Date of the Bohairic Version, 124.

18

Garrick V. Allen

justified in every respect as we recover the pseudohagiographa of the early centuries.” The idea that the church heavy-handedly influenced the transmission of particular text forms reappears in a number of Hoskier’s writings. “The Elzevir New Testaments of 1624 and 1633.” JTS 12 (1911): 454–57.38 In this article, Hoskier responds to three articles in the July 1910 fascicle of JTS, one by Eb. Nestle and two others by F. C. Burkitt.39 His first quarrel is with Nestle, who impugns his ability to accurately collate, referring to Hoskier’s transcriptions of readings from the 1624 and 1633 Elzevir editions (esp. Heb 9:12 and Rom 6:4). 40 Hoskier reacts to Nestle’s rhetorical question of “was Hoskier struck with blindness?” by pointing out typographic changes to different print runs of the Elzevir editions. 41 He defends his own transcriptions, but acknowledges that other printings of the 1933 edition follow the text that Nestle believes mistaken. Hoskier takes offence at the questioning of his accuracy, in which he rightfully takes great pride: 42 “would it not have been more generous of Dr Nestle to have asked me to verify my references before he pilloried me, and threw doubt on my accuracy?” (p. 455). Hoskier goes on to comment on a recent article by Burkitt, arguing that Burkitt is incorrect to argue against an Irish provenance for the Latin manuscript

Cf. the response by F. C. Burkitt, “Additional Note,” JTS 12 (1911): 457–59, who sarcastically comments that he is “extremely interested to hear that it [i.e. Codex Claromontanus, Vat. Lat. 7223] was written in Ireland by an Irishman, and I am sure that readers of the Journal will be grateful if Mr Hoskier will publish the colophon or note which establishes this important fact” (pp. 457–58). Referring to Hoskier’s reputation, Burkitt notes: “I would not quibble at a word, but Mr Hoskier has such a well-deserved reputation for minute accuracy in textual matters, and he is so severe on the lapses of other people, that his statement might very well be understood to imply a higher degree of similarity between Z and the Vulgate portion of Cod. Claromontanus that I imagine to exist” (p. 458). 39 Eb. Nestle, “Some Points in the History of the Textus Receptus of the New Testament,” JTS 11 (1910): 564–68; F. C. Burkitt, “Euangelium Gatianum,” JTS 11 (1910): 607–11; Burkitt, “A Gothic-Latin Fragment from Antinoe,” JTS 11 (1910): 611–13. 40 Cf. Hoskier, A Full Account and Collation of the Greek Cursive Evangelium 604, appendix C. 41 Nestle does, however, refer to Hoskier as “one of the most conscientious workers in the field” and comments on his admiration (“there can be no greater admirer of his patience than I”), impressions that perhaps accounts for his indignant tone in Hoskier’s rebuttal of Scrivener’s comments on the differences between the 1624 and 1633 Elzevir editions in Plain Introduction to the Criticism of the New Testament, 3rd ed. (London: George Bell, 1883), 441–43 (Nestle, “Some Points,” 565). 42 Cf. Kirsopp Lake, The Text of the New Testament, 6 th ed. (London: Rivingtons, 1933), 76, who notes that Hoskier is “an almost supernaturally accurate collator.” Hoskier and Lake appear to have had somewhat of a professional rivalry, although it was probably one-sided. Cf. a letter from Hoskier to J. Rendel Harris 4 February 1929 (DA21/1/2/1/25/19). 38

The Patient Collator

19

h. This manuscript, according to Hoskier “is Irish, was written in Ireland by an Irishman, and has Irish decoration” (p. 456). He concludes with a comment on the Gothic version, arguing that its Greek base is a hitherto unknown form that has a very old “Graeco-Latin-Syriac stem,” a classic example of Hoskier’s position on the origin of the versions, based in this instance, on Burkitt’s review of an edition of an actual Latin-Gothic bilingual manuscript prepared by Paul Glaue.43 “The Authorized Version of 1611.” Bibliothecha Sacra 68 (1911): 693–704. Writing on the three-hundredth anniversary of the publication of the King James Version, Hoskier takes the opportunity to make two main points. First, he again criticises the “unfortunate and overzealous” 1881 revision based on Westcott and Hort’s text; and, second, he argues “against any revision at the present time. I feel that this cannot be successfully handled to-day” (p. 693). He reasons along these lines because, although much new Greek and versional material had been published since 1881, there are many missing pieces of the puzzle, and because a suitable methodological basis for a revision has not been established. Even if an agreeable method of “removing errors” in the Greek text can be identified, the issue of translation creates another set of critical issues, and Hoskier points to a lengthy list of perceived errors made by the 1881 revisers as evidence for the problematic nature of revising the 1611 version. At the heart of his objection to a new revision is the idea that “our Bible of 1611 is so precious – obtained through fire and sword, blood and much tribulation – that we cannot safeguard it enough” (p. 696), although he does identify some alterations that are necessary, although not necessarily substantive, like the substitution of “flock” for “fold” in John 10:16. “New Edition of the Codex Veronensis (b).” The American Journal of Philology 32 (1911): 220–21. This piece reviews E. S. Buchanan’s edition of the Latin text of the purple manuscript b (codex Veronensis). 44 He commends the author for correcting the faulty collation of Bianchini, 45 especially in the light it now sheds on the reading in Luke 23:34. Hoskier commends the volume to Americans in particular, 43 Paul Glaue and Karl Helm, Das gotisch-lateinische Bibelfragment der Universiätsbibliothek der Gießen (Gießen: Töppelmann, 1910). 44 E. S. Buchanan, The Four Gospels from Codex Veronensis (b), being the First Complete Edition of the Evangeliarum Purpureum in the Cathedral Library at Verona, with two facsimiles (Oxford: Clarendon, 1911). 45 Giuseppe Bianchini, Evangeliarium quadruplex latinae versionis antiquae seu veteris italicae (Rome: de Rubeis, 1749).

20

Garrick V. Allen

connecting its text to Irenaeus (it “lay on his desk”) and his polyglot theory: “we are taking more interest in textual criticism these days and what may not American scholars accomplish if, properly equipped, they lend their aid in unravelling the interesting questions which arise in connection with the great Graeco-Syriac-Latin base of all these manuscripts” (p. 221). Concerning the Date of the Bohairic Version: Covering a Detailed Examination of the Text of the Apocalypse and a Review of the Some of the Writings of the Egyptian Monks. London: Bernard Quaritch, 1911. 46 This book represents both an explicit preview of Hoskier’s argumentation in Concerning the Text of the Apocalypse (see below), but also another articulation of his polyglot theory. He blends these concerns in his attempt to “exhibit the coptic element in ‫ א‬in the Apocalypse as an answer to those who think the bohairic version is later than the time of ‫( ”א‬p. 1). In order for the text of Sinaiticus to have been influenced by the Coptic tradition, the Coptic translations must antedate Sinaiticus. Hoskier aims to show that both of these propositions are correct and the weapon he wields in this debate is, once again, extensive collation. Sinaiticus was copied from a Coptic-Graeco polyglot (p. 3), and even corrected from a diglot (p. 7). The legion examples that Hoskier adduces to prove this contention are only convincing in aggregate, even though many of the individual examples are easily explained by other routes beyond polyglot linguistic interference, like mechanical errors in copying, inner-Greek variation, or other less tendentious reasons. For example, few would hold that the variant ισχυσεν/ισχυσαν in Rev 12:8 is truly influenced by a parallel Coptic reading (pp. 34–5). The second part of the book examines the quotations of some Egyptian monastic writers that Hoskier dates to the fourth century (pp. 117–92). The quotations of these writers show that they were using polyglot manuscripts before the production of ‫( א‬e.g. pp. 159–60), an argument that supports his two main arguments in the book. The comments of Goodspeed in his dual review of this volume and Concerning the Genesis of the Versions (1910) clearly summarize the critical response to these volumes: “That it all demands for its explanation Mr. Hoskier’s quaint theory few will believe, while the extreme positions in which that theory involves its advocates and the anterior improbability of the existence or use of such polyglots in antiquity complete its discomfiture. Mr. Hoskier’s view Cf. the review of this volume and Concerning the Genesis of the Versions (1910) in Goodspeed, “Hoskier’s Study of the New Testament Versions,” 652–54; cf. also a brief statement in S. Gaselee, “Christian Egypt,” Archaeological Report (Egypt Exploration Fund) (1911–1912): 54–79 (here 57–8), noting Souter’s critique of Hoskier’s polyglot theory. 46

The Patient Collator

21

might have been more clearly and compactly and less dogmatically presented. In particular his criticisms of Dr. Hort’s textual methods and conclusion suggest that he, like many of Hort’s critics, has not fully understood them.” 47 The book also illustrates the progress that Hoskier had made on his Apocalypse collations. A list precedes the volume of the first 123 manuscripts that he has collated, along with a preliminary list of textual groups (eighteen of them) and a short list of “important single manuscripts.” “Evan. 157 (Rome. Vat. Urb. 2).” JTS 14 (1912–1913): 78–116; 242–93; 359– 84 (3 parts). This lengthy three-part study explores the text of the carefully prepared twelfth century Vat. Urb. gr. 2 (GA 157), a text that is of importance to Hoskier because of its close relationship with the Textus Receptus, especially in Matthew and Mark. These studies are mainly comprised of textual notes and the relationships between this manuscript’s interesting Greek readings and the versions, along with collations made against the 1550 Stephanus text. Part II provides some analysis, as well as collations of Luke. The main argument of the series is that the manuscript’s text is influenced by linguistic interference from antecedent bilingual copies, perhaps Graeco-Latin, Graeco-Latin-Coptic, and perhaps also some influence from Syriac and other versional traditions (pp. 243–52). Part III is a collation of John without further comment. “The Lost Commentary of Oecumenius on the Apocalypse.” The American Journal of Philology 34 (1913): 300–14. This article relays in detail Hoskier’s re-discovery of the full Oecumenius commentary in GA 2053, although it was first noticed by Franz Diekamp in 1901.48 Hoskier identifies this manuscript and its tradition as particularly important for a number of reasons: (1) it offers direct insight, according to him, into the text used by Oecumenius in late antiquity (ca. 600 CE); (2) the text is older than Andrew of Caesarea; and (3) the Oecumenius commentary quotes the lemmatic text of Revelation extensively, as well other New Testament texts, like the disputed words from the cross in Luke 23:34. Hoskier also collates parts of the text with other prominent witnesses, tying the text to a range of versional traditions in support of his polyglot theory. He concludes with a condemnation of

Goodspeed, “Hoskier’s Study of the New Testament Versions,” 653. Franz Diekamp, “Mitteilungen über den neu aufgefundenen Kommentar des Oekumenius zur Apokalypse,” Sitzungsbericht der Kgl. Preuß. Akademie der Wissenschaft 13– 14 (1901): 1046–1056; Diekamp, “Neue über die Handschriften des Oekumenius-Kommentares zur Apokalypse,” Biblica 10/1 (1929): 81–4. 47 48

22

Garrick V. Allen

von Soden’s edition for failing to include readings from important minuscule manuscripts. “The New Codex ‘W.’” Expositor 5/5 (May 1913): 467–80. This article responds to the edition of the Freer Gospel codex published by Henry A. Sanders. 49 It functions in part as a review of this work and it adds a number of further observations on W. These include the idea that W was copied from a papyrus book, not unlike P.Oxy. 2, and that its material layout is wellremoved from the early bi- and tri-lingual traditions that Hoskier perceives to stand at the early stages of the tradition, along with textual commentary arranged to support his early polyglot theory and the close relationship between W and Coptic traditions (pp. 475–76). Hoskier also perceives “retranslation” from Greek to Latin to Greek in some of W’s readings (p. 477). The article concludes with the notation “(to be continued.),” but I can find no obvious second part of the article. “Von Soden’s Text of the New Testament.” JTS 15/3 (1914): 307–26. This article constitutes Hoskier’s review of Hermann von Soden’s entire project.50 He does not approve: “instead of writing a eulogy on his work I regret to have to condemn it strongly” (p. 307). Hoskier’s objections are many. (1) The apparatus of the edition is riven with errors and based in part on out-ofdate or otherwise problematic collations. Related to this is the fact that von Soden introduced his own system of abbreviation and nomenclature, a feature of text-critical praxis that Hoskier bemoans in a number of his works (pp. 307– 308, 322–23). (2) According to Hoskier, von Soden’s eclectic approach leads him to “invent scripture,” which leads to further confusion in the apparatus (pp. 308–13). This is worse than Hort’s slavish devotion to B and leads to “grotesque” readings or “grave errors” (pp. 312, 316–22), including the omission of readings from ‫ א‬not found in Tischendorf’s apparatus. (3) Hoskier can divine no system in von Soden’s textual decisions (pp. 313–16), since he follows numerous witnesses in constructing his text, even when the evidence is scant. In essence, he objects to von Soden’s reliance on internal criteria. He concludes: “But to state these matters is only to make a partial impression on my readers

49 Henry A. Sanders, The New Testament Manuscripts in the Freer Collection. The Washington Manuscript of the Four Gospels (New York: Macmillan, 1912); Sanders, Facsimile of the Washington Manuscript of the Gospels in the Freer Collection, 2 vols. (New York: Macmillan, 1912). 50 Hermann Freiherr von Soden, Die Schriften des Neuen Testament in ihrer ältesten erreichbaren Textgestalt, auf Grund ihrer Textgeschichte, 2 vols. (Göttingen: Vandenhoeck & Ruprecht, 1911–1913).

The Patient Collator

23

of the grievous state of things in this latest book on a most intricate subject. Es ist zum Weinen” (p. 326). The timing of Hoskier’s very negative review, published in April 1914, could not have been worse, a point not lost on some of his contemporaries. 51 Its publication was initially slated for the January edition, the same month of von Soden’s death in a U-Bahn accident in Berlin on 15 January. In Codex B and its Allies, 461 Hoskier regrets the tone of his review in light of the circumstances, although he continues to be perniciously critical of von Soden’s work, even in his 1937 supplement to his JTS article (fasc. 2, pp. 18–9). Codex B and its Allies: A Study and an Indictment. 2 vols. London: Bernard Quaritch, 1914. This substantial two-volume work is the pinnacle of Hoskier’s polemical attack on Hort, the editors of 1881 Revised Version (who perpetrated the “heresy of our time” [p. 422]), Codex Vaticanus, and other detractors. Like his other works, this book is comprised mainly of hundreds of pages of collations designed to undermine the value of B and related points, creating a full frontal assault on the principles of Westcott and Hort and other critics, especially as Hoskier becomes increasingly exasperated at the end of the volume, often couched in the language of apologetics and idolatry. 52 Hoskier chafes at the Cf. W. Sanday, “Baron Hermann von Soden,” JTS 15 (1914): 305–306: “I cannot close this brief and inadequate tribute to a scholar of great eminence without a word, which must be also one of sympathy, for the author of a criticism which appears later in this number. Mr. Hoskier is well known as one of the most independent and most incisive of our writers, with a passion for precise detail. It has fallen to his lot to review the great book, and he has done so in a sense that is adverse, and even hostile. I know that he regrets the unhappy coincidence which brings out his criticism at this particular juncture. It is one thing to throw down a gage of battle before a champion who is in possession of the lists and in the fullest vigour to defend his own cause, and another thing to issue a like challenge over a newly closed grave. All who are connected with The Journal of Theological Studies would have wished to avoid such a coincidence; but the article was already paged for the January number of the Journal, and on the eve of being printed off at the time of Baron von Soden’s death, and the publication of it could only be deferred for the moment” (here p. 306). 52 Some particularly titillating examples include the following. On Matt 22:10: “I think it is criticism gone absolutely wild and mad to accept νυµφων here, and it is unpardonable of Hort to put νυµφων in his text without any alternative in the margin and equally wrong of Soden” (p. 66). On followers of Hort: “How many more instances of this kind must I adduce before the worshippers of B and the obsequious slaves of Hort will allow that I am right?” (p. 84). Responding to a criticism of Alexander Souter: “Don’t condemn me in this cavalier fashion then, if you please, but look into these matters a little more carefully” (p. 313). On Burkitt: “He has said, rather unnecessarily, of me that I do not know the difference between a dilettante and a scholar. However that may be, I think I can detect the difference between an unbeliever and a believer!” (p. 357). Hoskier goes on to accuse Burkitt of apostasy; if you don’t have anything nice to say, I guess it should be said in a footnote. Although 51

24

Garrick V. Allen

idea that B preserves a neutral text, charging “an indictment against the MS B and against Westcott and Hort, subdivided into hundreds of separate counts” (p. i). The book is therefore a somewhat roundabout attempt to “sing the Deathsong of B as a neutral text” (p. iii). The larger argument that stands in the background of this book is, again, the idea that the texts of the uncials are in fact corrupted and that they fell into disuse because of this fact. Therefore, the Textus Receptus, the text that the church relied upon until Westcott and Hort, is the true text. The Textus Receptus is ancient, and was bungled into the text found in the uncials by polyglot interference. B’s text in particular is the result of the conscious editing of scribes who presume to know better than the inspired authors they were tasked with transmitting. These scribes “mutilated” the text. Another explanation for this phenomenon that Hoskier posits is the concurrent composition of Mark in both Greek and Latin. The Latin text was, according to Hoskier, later translated into Greek, creating two early Greek forms of Mark (pp. 126–207). This theory strains credulity on a number of fronts. Despite the obvious issues with such a strong stance, the book also develops Hoskier’s methodological profile, emphasising tedious attention to the detail of the tradition writ large, the patient digestion of data, and hard fought conclusions. One of his main charges that he levels at his opponents, tendentious though it is, is that their consensus conclusions are based only on partial data. The conventional wisdom, for Hoskier, is based on partial and inadequate information. This belief explains the depth of his textual rhetoric and devotion to collation, as well as his extended critique of Hort’s method (pp. 1–13) and criticism of the shorter reading canon (e.g. p. 54). The book also demonstrates that Hoskier’s critical goal is a religious desire to find the “true,” unadulterated text of the New Testament, and that those who identify this text wrongly are heretics, leading the young and impressionable astray. Hort’s text is not only wrong, but heretical (pp. 468–73). The second volume of the book attempts to further discredit the text of the uncials by driving a wedge between ‫ א‬and B in the Gospels. The point again is to undermine the value of key witnesses for Westcott and Hort. Once again, deficiencies in ‫ א‬are chalked up to copying from polyglots (e.g. pp. 146–47), relying on reams of data devoid of much analysis. The Text of Codex Usserianus 2., or r2 (“Garland of Howth”) with Critical Notes to Supplement and Correct the Collation of the Late T. K. Abbott. Old Latin Biblical Texts. London: Bernard Quaritch, 1919. Hoskier began preparation of this volume in 1914 and its publication was delayed by the Great War (p. iii). He devotes an entire book to this particular he is quite high on C. H. Turner: “I suppose that it will readily be conceded that C. H. Turner is without question the most brilliant writer on Textual Criticism today” (p. vii).

The Patient Collator

25

manuscript because it is a witness to an old Latin text. The transcription that comprises the main portion of the piece – and which is a feat of printing with its many special characters that mirror the palaeography of the manuscript – is necessary because previous publications, especially T. K. Abbott’s, are shot through with errors that number in the thousands if orthographical peculiarities are considered (p. iii; cf. the thirty-seven pages of corrections to Abbott’s collation in the appendix).53 The transcription follows not only the textual peculiarities, but also attempts to present the form, line structure, and palaeographic profile of the manuscript. It is an attempt at a manual facsimile. See Hugh Houghton’s article in this volume that explores this manuscript and its text in detail. “Manuscripts of the Apocalypse – Recent Investigations, part I.” Bulletin of the John Rylands Library 7/1 (1922): 118–37. This series of five articles functions as a precursor to Concerning the Text of the Apocalypse. They deal with minuscule manuscripts from the tenth century onward, since, according to Hoskier, many later manuscripts “throw much light on the transmission of the text” (p. 118). This first article also articulates his motivation for devoting so much time to the Apocalypse, and it illuminates his own knowledge of the ground-breaking nature of his method: “Never before has a comprehensive examination such as this been undertaken for any book of the New Testament. I selected the Apocalypse simply because it was possible for an individual to handle the matter within his lifetime” (p. 118). His three critical observations in this article are (1) that the text of Erasmus’ choice of manuscript corresponds to the majority textual form; (2) that he has not identified any Complutensian text witnesses that bear the singularities of Stunica’s edition; and (3) that his own numeration system is superior to both Gregory’s and von Soden’s. He analyses the text of manuscript 200 (GA 2329), arguing that the text’s age rivals that of the great uncials. His high valuation of this witness corresponds to his lengthy treatment of it in Concerning the Text, 1.637–52 and its status as a “consistently cited witness” in NA28. 54

T. K. Abbott, Evangeliorum versio Antehieronymiana ex codice Ussheriano (Dublin: Hodges and Figgis, 1884). 54 Cf. also Darius Müller, “Die Apokalypse-Handschriften GA 2329 und 2351: Textkritische und textgeschichtliche Beobachtungen zu zwei ‘ständigen Zeugen für die Apokcalypse’ in Nestle-Aland28,” in Studien zum Text der Apokalypse II, ANTF 50, ed. M. Sigismund and D. Müller (Berlin: De Gruyter, 2017), 369–410. Hoskier mentions this article in a letter to J. Rendel Harris on 17 May 1922 (DA21/1/2/1/25/7) and that he plans to write a second in the series. 53

26

Garrick V. Allen

“Manuscripts of the Apocalypse – Recent Investigations, part II.” Bulletin of the John Rylands Library 7/2 (1923): 256–68. This article deals with manuscripts 201 (GA 2351), a partially preserved manuscript with a previously unknown commentary – the so-called Scholia in Apocalypsin.55 In addition to rapping Adolf von Harnack on the knuckles for producing a poor edition and listing out the edition’s errors, 56 Hoskier catalogues the significant readings of this manuscript and its shared agreements with other text clusters. He continues with a brief analysis of the commentary and, after listing out the many errors of Harnack’s edition in this portion of the text, concludes with Harnack that Origen was the likely author, although some scholia can be traced to other sources. He concludes with a short note on 202 (GA 2352), which he identifies as having a Complutensian text. “Manuscripts of the Apocalypse – Recent Investigations, part III.” Bulletin of the John Rylands Library 7/3 (1923): 507–25. This article examines manuscript 143 (GA 2050), dated to 1107. This piece reads very much like an entry in volume 1 of Concerning the Text, containing brief material descriptions, palaeographic comments, and information on its familial relations. This particular witness, for Hoskier, “stands quite apart from any traditional family groups” (p. 508), an observation that is supported by recent Text und Textwert data that shows 2050 agreeing with LA 2/ in 51% of readings. 57 Hoskier uses this textual data to argue that the manuscript is the progeny of an ancient Greco-Coptic manuscript produced and used in North Africa (p. 508). Because he perceives the peculiarities of the text to be ancient, Hoskier carefully catalogues its significant readings, drawing deep connections

This manuscript has attracted much interest recently. Cf. Garrick V. Allen, “The Reception of Scripture and Exegetical Resources in the Scholia in Apocalypsin (GA 2351),” in Commentaries, Catenae and Biblical Traditions, ed. H. A. G. Houghton (Piscataway: Gorgias, 2016), 141–63; P. Tzamalikos, An Ancient Commentary on the Book of Revelation: A Critical Edition of the Scholia in Apocalypsin (Cambridge: Cambridge University Press, 2013). 56 C. Diobouniotis and A. Harnack, Der Scholien-Kommentar des Origenes zu Apokalypse Johannis, TU 3/8 (Leipzig: Hinrichs’sche, 1911). Hoskier (p. 256) states that “this publication is not only faulty and inaccurate, but the pride of the scholar has caused Harnack to print his suppositious emendations in the text of the work and the real readings of the MS. are relegated to the footnotes, an inverted and pernicious manner of editing a document, so far unique, to which the present writer seriously objects.” Hoskier is methodologically opposed to conjectural emendation as a principle. 57 M. Lembke et al., eds., Text und Textwert der griechischen Handschriften des Neuen Testaments. VI. Die Apokalypse. Teststellenkollation und Auswertungen (Berlin: De Gruyter, 2017), 568–69. 55

The Patient Collator

27

to the Coptic tradition. The text of this manuscript, for Hoskier, touches “the faint spots” of the tradition and gets behind the text of Codex Sinaiticus. “Manuscripts of the Apocalypse – Recent Investigations, part IV.” Bulletin of the John Rylands Library 8/1 (1924): 236–75. The fourth article in the series examines manuscript 130 (GA 1854) in depth, mustering textual analysis in support of the deep antiquity of the versions and the idea that the Greek text was influenced at an early stage by the proliferation of polyglottal exemplars: “its original polyglot base is veneered with a much later strain of eclectic polyglot readings and re-readings” (p. 236). In other words, polyglots on top of polyglots; piles of polyglot polyglots. This analysis is supported further by a thirteen page “P.S.” notation that also advances this same critical aim. This theory, for Hoskier, answers a legitimate critical observation: that these medieval manuscripts seem to preserve, in part, very ancient texts, texts that are often mixed and which sometimes correspond to versional traditions, a phenomenon also known as block mixture. Ancient polyglots at the very base of the beginning of transmission is the only all-encompassing response that Hoskier can divine (pp. 240–75). He clarifies this relationship: “I do not mean to say that all the sporadic agreement with the Versions, – first with one and then with another, – is all due to their reflect action on the Greek. But I do mean to say that most of it is” (p. 270, emphasis original). “Manuscripts of the Apocalypse – Recent Investigations, part V.” Bulletin of the John Rylands Library 8/2 (1924): 412–43. In the final article of the series, Hoskier examines the readings of a number of manuscripts, focusing again on their close relationship to the versions (e.g. GA 94 141 208 2081). For example, GA 2081 (Hoskier 179) is of value because it stands close to many readings in Crawford’s Syriac edition, even though it was “carelessly copied in the early stages of its reproduction” (p. 413). Hoskier also, in a moment of candour, recognizes that his polyglot theory may be difficult to swallow: “I hope, however, [the reader] may gradually come to see that I am not perversely afflicted with any wish to over-emphasise [my polyglot theory]. I simply cannot get away from it” (p. 420). The article concludes with a lengthy poetic composition meditating on divine names in the Bible which reappears at the end of the introduction in the first volume of Concerning the Text of the Apocalypse, a pious articulation of his motivation to undertake detailed textual work. It is followed by the comment that “this is why I am concerned to recover, as far as it is possible, the exact wording of this sublime message to the Churches and to Humanity-at-large” (p. 443).

28

Garrick V. Allen

Immortality. Boston: The Stratford Company, 1925. This volume is the first of Hoskier’s philosophical works, previewing a number of ideas that he develops in the more substantial volumes in the 1930s. Ideas like the “back of beyond” are introduced here – the metaphysical location of human origins where individuality ceases – along with the essential deific nature of all humanity and organized religion’s inability to properly reflect the metaphysics of the universe. The goal of life is not to cheat death or experience eternal life on earth, but to “take wings and leave earth” (p. 1) in a metaphorical sense, although the mechanics of this operation are not entirely clear. In order to fundamentally reconceive notions of the goal of existence and models of the afterlife, Hoskier appeals to comparative religious studies and argues at length for the veracity of various forms of spirit communication, arguing that eastern religions stand closer to the origins of an initial revelation between some “extramundane Creator and Ruler” (pp. 2–3) than do Christianity or Judaism. He examines side-by-side, for example, texts from the Bahgavad-Gita, portions of the New Testament, 58 and spirit communication through a certain Mdme. de Watteville and other mediums (pp. 5–13). 59 The book progresses by exploring the thought of important – at least by Hoskier’s measure – ancient literati from Hesiod to Plato, Moses to Sophocles, divining their proximity to the origins of the first “Revelation.” These traditions in all their breadth function as witnesses to ancient events. In a way, Hoskier exercises text-critical operations on tradition-historical resources, attempting to remove errors to arrive at a religion unadulterated by tradition, change, and human error. Accessing this event will help humanity to realise the nature of their immortality, which has been muddled by the passage of time. The rest of Hoskier’s philosophical tracts develop these ideas further, sometimes in unexpected ways. The ideas herein are, in addition to being controversial (to put it mildly), sometime troubling, like his adoration of the French theosophist Joseph Alexander Saint-Yves d’Alveydre (and other French spiritualists), 60 whose Eurocentric and supersessionist rhetoric dominates in Hoskier’s extended interaction with his work. 61 The book is 58 For Hoskier, the New Testament is riven with “occultist” happenings (pp. 115–28), mustering passages that connect Jesus’ perspective to his own theosophonic viewpoint. He even offers his text-critical work and the unearthing of new readings as further evidence of the occult nature of the New Testament (e.g. pp. 127–28). 59 Hoskier also mentions “Mrs de W.” briefly in a letter to J. Rendel Harris, 14 June 1922 (DA21/1/2/1/25/11). 60 This book also contains Hoskier’s only explicit comments on what he calls Spiritism or Spiritualism and spirit communication, perspectives that also impinge on his text-critical work in Concerning the Text (cf. pp. 136–246). For Hoskier, “incredulity in spirit phenomena is simply and solely due to ignorance” (p. 150). 61 Cf. Saint-Yves d’Alveydre, Les Mission des Juifs (Paris: Calmann Lévy, 1884); cf. also the spiritualist writings of Jean-Baptiste Eugène Nus, e.g. Les Grands mystères: vie

The Patient Collator

29

fundamentally an esoteric history of the world from the first Egyptian dynasty (the time of the theocratic and esoteric “Kingdom of Ram”) to modern Spiritualism and theosophy, built on surmise, conjecture, and coincidence, but also experience with ancient manuscripts, archaeology, and the progressive evolutionary science of the early twentieth century. An interesting connection between the argument of this work and Hoskier’s academic publications is the idea that the scientists and the wise are the true guardians of truth and justice preserved from the earliest forms of “true Religion” (pp. 98–9). Hoskier’s harping on the “scientific” nature of textual criticism – even critiquing Hort in Codex B and its Allies by exclaiming “where is the science?” – is not a mere methodological critique, but a theological one. This volume explains some of the invective of his pre-war work. Hoskier’s “living God” is “WISDOM AND SCIENCE” (p. 102), and Jesus is the historical figure who embodied ancient “true Religion” in its fullest sense (pp. 103–105). The book demonstrates that Hoskier’s philosophical and religious views are derived in part from nineteenth century reports of conversations between spirits and French mediums, and it is clear that he has had first hand experiences in such settings. The Complete Commentary of Oecumenius on the Apocalypse: Now Printed for the First Time from Manuscripts at Messina, Rome, Salonika, and Athos. Ann Arbor: University of Michigan Press, 1928 [repr. Wipf & Stock, 2008]. 62 This edition, which follows the text of GA 2053 as far as possible, is built on the evaluation of nine manuscripts. 63 Hoskier correctly places Oecumenius before Andrew historically, and points out Andrew’s dependence on Oecumenius. He prefaces the edition with a substantial discussion of Oecumenius’ language, interpretive proclivities, intertexts, modes of quotation, the familial characteristics of his lemmatic text, theology, and strategies of interpretation. universelle, vie individuelle, vie sociale (Paris: Noirot, 1866), which Hoskier references on pp. 105–14. 62 Reviewed by Ernst Benz in Gnomon 6 (1930): 341–45; Alexander Souter in The Classical Review 43/6 (1929): 240; K. Staab in Byzantinische Zeitschrift (1931): 374–79: Aimé Puech in Revue de Philologie, de Littérature et d’Histoire Anciennes (1929): 425–26; P. Ubaldi in Rivista di Filologia e di Istruzione Classica 59 (1931): 419–20; Henri-Charles Puech in Revue de l’Histoire des Religions 106 (1932): 465–66; J. Behm in Orientalsiche Literaturzeitung 37 (1934): 170; A. C. in Aegyptus 11/4 (1931): 509. Hoskier first mentions his plan to publish the Oecumenius commentary in a letter to J. Rendel Harris of 17 May 1922 (DA21/1/2/1/25/7) and comments on the “unkind” review in JTS 31 (1929): 54–8 (which also reviews Concerning the Text) in a letter of 18 December 1929 (DA21/1/2/1/25/23). 63 Cf. the recent edition of Marc de Groote, ed., Oecumenii Commentarius in Apocalypsin (Leuven: Peeters, 1999) and an English translation by John N. Suggit, Oecumenius: Commentary on the Apocalypse, The Fathers of the Church 112 (Washington, D.C.: The Catholic University of America Press, 2006).

30

Garrick V. Allen

Like much of Hoskier’s editorial work, the preface is structured in an intuitive way (at least intuitive to Hoskier), where each of these topics are interwoven according to Hoskier’s own concerns which are not always explicit. (Although Puech’s review notes that “son Introduction est fort instructive,” despite the multiple accenting mistakes in the text.) The preface gives special emphasis to Oecumenius’ mystical interpretation and his lengthy quotation of Methodius (e.g. pp. 12–3), as well as the customary description of the manuscripts utilised in the construction of the edition and their textual relationships (pp. 16–25). The edition is now superseded by Marc de Groote’s 1999 edition, but Hoskier’s text is aesthetically pleasing and his apparatus accurate. Concerning the Text of the Apocalypse: Collation of All Existing Available Greek Documents with the Standard Text of Stephen’s Third Edition Together with the Testimony of Versions, Commentaries and Fathers. 2 vols. London: Bernard Quaritch. 1929.64 The two stout volumes of Concerning the Text of the Apocalypse are the greatest of Hoskier’s contribution to textual scholarship. Despite the presence of idiosyncrasies like the polyglot theory (e.g. p. 1.444), the presence of readings divined from spirit communication (p. 1.xxxviii), 65 exclamations about poor collations, 66 and strident critiques of Hort (e.g. pp. 1.xlvii–lxvi), this work represents the first ever attempt to catalogue every reading from every known manuscript for a particular New Testament work. The task took over thirty years to complete and is the pinnacle of Hoskier’s methodological emphasis on comprehensiveness, patient tedium, and data aggregation (e.g. p. 1.128). The first volume (751 pages in all) contains a substantial introduction and a description of every manuscript in his edition. The prolegomena is consistently Hoskierian, lacking a clear structure and jumping from topic to topic without Reviewed very positively by León Vaganay in Revue des Scienes Religieuses 13/2 (1933): 295–96: “En un temps où les critiques sont si nombreux, il peut paraître étrange qu’un livre de valeur passe inaperçu. C’est cependant ce qui est arrivé, du moins dans les milieux de langue française, pour le dernier ouvrage de M. Hoskier sur le texte de l’Apocalypse. Sauf erreur, aucune revue n’en a parlé. Or c’est là, à notre avis, une œuvre monumentale digne de retenir l’attention.” And “enfin, et surtout, il nous donne là un bel exemple de labeur ingrat, accompli quand même avec le soin le plus scrupuleux, notamment en ce qui concerne la tradition des manuscrits grecs. Il mérite de ce fait la reconnaissance de tous les savants qui auront à s’occuper de l’Apocalypse. Son œuvre restera longtemps l’instrument de travail indispensable.” Vaganay even argues that Hoskier’s approach is superior to that of Westcott-Hort and von Soden. 65 Other spiritist perspectives are found through the volumes. E.g. p. 1.358 on the significance of the various precious stones in Revelation 21 and a lengthy paragraph in Latin on the numerology of Rev 13:18 (p. 2.365). 66 E.g. “It is painful to realize how few men are qualified to collate accurately, even when striving to do their best” (p. 7). 64

The Patient Collator

31

discussion on their relatedness. This section provides a number of insights into Hoskier’s motivation and method for such a herculean project. He is influenced not simply by textual questions, but theological ones. He interprets Rev 8:9 in light of the Great War and argues that the sinking of one-third of the world’s shipping during the conflict is a fulfilment of the text – the Apocalypse is important to understand because the eschaton is drawing nigh (p. 1.ix). 67 The goal of the project, however, is threefold: (1) to understand the full history of the text; (2) to prove the error of the “elder documents,” likely referring to the uncials; and (3) to identify Greek readings that antedate the fourth century, readings that are now lost among the mass of minuscules (p. 1.x). Implicit also in this introduction is a call to adopt his method as a model, a mode of operation that will answer the remaining questions of the field, and a desire to critique the revisers of the Authorized Version (pp. 1.xii–xiv). The Textus Receptus, apart from its obvious errors, will be vindicated by way of textual criticism as a venerable and ancient form of the text, protected by God’s providence (p. 1.137, 317). Hoskier also lays out his conclusions relating to Revelation’s textual families, asserting relationships by juxtaposition and sometimes collation. Hoskier’s attempt at constructing text families, however, is less than successful, a point that his early champions even noted. 68 His families were undermined by Josef Schmid’s seminal study, 69 whose work has been further undercut by the recent Text und Textwert data. 70 Following the introduction, the catalogue of manuscripts commences, allowing only six pages for a discussion of the uncials. The average manuscript description comments upon a number of issues, including the location, owner, call number(s), date and means of collation, family characteristics, palaeographic observations, mention of writing support, scribal habits, comments on the bibliographic context of Revelation, and engagement with previous scholarship on the specific manuscript. Each of these features are discussed in order to properly date the document in an effort to understand the context of its text, which is the underlying goal of these profiles. The group to which the manuscript belongs, according to Hoskier, is noted at the head of the entry, with no further comment, although he often shows via collation how the manuscript 67 On this same page, he also notes the negative effect of his war wounds on his right hand and eyesight. Cf. also his letter to J. Rendel Harris of 22 April 1922, where he notes that his eyes are “giving out,” although he has still enough sight to able to critique Harnack and Gregory (DA21/1/2/1/25/6). 68 Cf. Henry A. Sanders, “The Beatty Papyrus of Revelation and Hoskier’s Edition,” JBL 53/4 (1934): 371–80, even though Sanders’ article takes Hoskier’s groupings as the basis for his analysis of the relationship between P47 and the rest of the tradition. 69 Josef Schmid, Studien zur Geschichte des griechischen Apokalypse-Textes. 2. Teil. Die alten Stämme (Munich: Karl Zink, 1955). 70 Cf. Lembke et al., Text und Textwert (2017). Cf. Hoskier’s listing of families in Concerning the Text, 1.7–12.

32

Garrick V. Allen

preserves the special readings of that sub-group. For example, the group of Apoc. 20 (GA 175) is identified as 4-20-48-64-74 and eight readings are produced in support of this association (p. 1.38). For the most part these profiles are uneven, and the connection between groupings and the material in a given profile are often unexpressed, even though the primary aim of the catalogue is to eliminate witnesses and families in an attempt to identify the “original” text (p. 1.108). He identifies ten major groups (including a group inclusive of all the uncials), along with a number of “composite documents” and important single manuscripts (cf. pp. 2.23–4). In usual Hoskierian fashion, there are also a number of entertaining, and even charming, polemic moments. 71 The second volume comprises a re-edition of Stephanus’ 1550 text that includes a comprehensive apparatus of every Greek reading in every manuscript catalogued in volume one, along with versional witness (Syriac, Sahidic, Bohairic, Armenian, Ethiopic, Arabic, Latin) and church fathers (Dionysius, Hippolytus, Methodius, Oecumenius, Victorinus, Tertullian, Cassiodorus, Primasius, Cyprian, Tyconius, Apringius, pseudo-Ambrose). These latter data are not as trustworthy as the Greek readings derived directly from manuscripts. This volume represents a continuation of the arguments of the first volume, attempting to show the corruption in the uncial manuscripts and to demonstrate that the Textus Receptus is indeed a quality representative of the “original” text. The mode of argumentation differs, however, insofar as it is purely textual. And Hoskier even acknowledges the rhetorical edge of his edition: “for there is an argument on every page” (p. 2.7). For each verse, the text of Stephanus heads the page in bold type, followed by a comprehensive list of every variation from this text in every manuscript. The volume comprises 649 pages and is the most complete collection of data on the Apocalypse aggregated to this date, even though 71 further witnesses to Revelation have been discovered since its publication. De Contemptu Mundi: A Bitter Satirical Poem of 3000 Lines upon the Morals of the XIIth Century by Bernard of Morval Monk of Cluny (fl. 1150) Re-edited, with Introduction and Copious Variants from all the Know MSS. London: Bernard Quaritch, 1929. 72

71 E.g. after a detailed list of errors in Reiche’s collations: “Surely we can cease here. Of all the unwelcome tasks this is the worst, though we have but shown a small part of Reiche’s shortcomings. It is pitiful, pitiful. We can all forgive a man for failing to record some readings which have escaped his eyesight, but deliberately to misrepresent and misquote throughout is not permissible. Reiche’s dust cannot rise up and apologize, but we can learn a lesson not to mar these studies with such wicked pitfalls. God knows enough exist naturally” (p. 151). 72 Reviewed by Samuel H. Cross in Speculum 5 (1930): 451–52; H. E. Butler in The Modern Language Review 25 (1930): 359–60, in which he refers to Hoskier’s disorderly

The Patient Collator

33

This handsome folio edition of the twelfth-century satirical Latin poem of Bernard de Cluny demonstrates the breadth of Hoskier’s interests and technical skill. It also shows that his primary form of rhetoric, even in issues beyond the New Testament, is textual scholarship. This is the only critical eclectic edition that Hoskier ever constructed beside his edition of the Oecumenius commentary,73 even though he edited numerous individual documents and reissued Stephanus’ 1550 text along with a comprehensive list of readings in the Apocalypse in Concerning the Text. The preface demonstrates his perception of Bernard’s genius, as well as his disdain for previous editions and translations of his work, especially when they self-censor portions of the poem. “I feel convinced,” he says, “that Bernard’s poem has only to be put in its entirety in the hands of the intelligent Public of to-day for this reading public to rise up and call him blessèd” (p. xi). The organization of the edition is Hoskierian insofar as its introduction lacks a clear structure and the greatest depth of all textual minutiae are plumbed (cf. his catalogue and description of the eighteen manuscripts on pp. xxii–xxxiv). In his review, Butler’s main critique is the structure of the introduction, which he calls “diffuse, ill-arranged and full of repetitions.” 74 The goal of the edition is not utility, but comprehensiveness, even though the main text is relatively unencumbered. His editorial work still serves as the editio princeps for this medieval work. The Bronze Horses: A Comment on the Prose-poem of Amy Lowell. Portland, ME: Mosher Press, 1930. This short book, a mere eighteen pages, uses the controlling metaphor of the bronze horses from Amy Lowell’s poem as a way to describe the cyclical nature of time. It represents a continuation of Hoskier’s philosophical trajectory, but it is ultimately pessimistic in its prescient appraisal of the belligerence of nation-states that might lead to another great conflict. It contains some linguistic charms, but also many tortured sentences. Although published in 1930, the date on the last page is 1926. It is not hard to imagine why it took so long to locate a publisher. What is Nirvana? Portland, ME: Mosher Press, 1930. 75

introduction and cumbrous apparatus as a “labour of love.” Hoskier describes finishing the manuscript in a letter to J. Rendel Harris of 20 February 1929 (DA21/1/2/1/25/20). 73 Although he follows the text of GA 2053 for the Oecumenius commentary whenever possible. 74 Butler in Modern Language Review, 360. 75 Mentioned in a letter to J. Rendel Harris, 10 January 1930 (DA21/1/2/1/25/25).

34

Garrick V. Allen

This humble volume comprises a slim thirteen pages. The preface orients the discussion around the afterlife in religious traditions, focusing on the idea of nirvana that Hoskier adopts from Hinduism. He describes the main body of the work as a “Credo” of his own perspective on the afterlife, which is “quite in accord with boundless love, and which will be found also quite in harmony with a scientific attitude towards universal motion and universal chemistry, and which answers every question of every seeker-after-truth, and also provides a satisfactory solution to an apparently hitherto insoluble problem” (p. 7). The main thrust of the piece is that all people all essentially deific and that this essential nature is attainable after the “unclothing” of death, but I remain sceptical that Hoskier has unlocked all the answers to the afterlife, which is a goal almost as ambitious as reconstructing the “original” text of the New Testament. In Tune with the Universe. London: Rider & Co., 1932 [Pseudonym “Signpost”]. This is one of Hoskier’s most substantive philosophical treatises, building off his perspectives on the afterlife in What is Nirvana? Interestingly, Hoskier writes an obsequious forward to Signpost’s work in his own name. Esoteric jargon defines the work, whose primary argument is that all people are part of the “All Life,” a reality that disposes of individualism in all its forms: “realize that you are not your brother’s keeper, but that you are your brother” (p. iv). He calls for an internal disarmament that defines much of human interaction (e.g. p. 29), provocatively using the language of early 1930s European political discourse. The treatise gives the overall impression that Hoskier views modern nation-states and liberal democracies as the negative outworking of human individualism (cf. his solution on pp. 120–21). 76 Nonetheless, the goal of human existence, for Hoskier, is to transcend the vibrational world to the “world of causes lying behind it” (p. 1) by ceasing to look to transcendent deities, by looking inside oneself to their own inner “godhood”: “you don’t look up, but you look within” (p. 5). Within this treatise he appeals to Christian traditions – or at least couches his language in conventional Christian pious terminology – and other non-conventional avenues, like various societies for psychic research (p. 13). This book is at times progressive, but also reflects predominant and problematic racial and class stereotypes of the period (e.g. pp. 31–8, 129). Finally, the book offers access to Hoskier’s perceptions of his major life events, including his marriage (p. 70), service in the Great War (pp. 94–5), and his numismatics collecting (pp. 131–32). “Concerning ‫ את‬and its Very Special Use in the Old Testament.” Pages 96–117 in Amiticiæ Corolla: A Volume of Essays Presented to James Rendel Harris, 76

Cf. also Hoskier, Back of Beyond, 47, 62–4, 68–70.

The Patient Collator

35

C.Litt. on the Occasion of His Eightieth Birthday. Edited by H. G. Wood. London: University of London Press, 1933. This article best represents a fusion of Hoskier’s academic work and philosophical proclivities, as he explores the numinous and esoteric value of the “designatory particle” in Hebrew. He finds previous grammatical and linguistic explanations of this word “arid” (p. 96) and notes that: “Well, in the first place, it is significant that it is composed of the first and last letters of the Hebrew alphabet, and therefore undoubtedly corresponds to the Alpha and Omega of all things, the wholeness of the thing characterized or designated, the oneness, the entireness of a thing, or, as Fabre d’Olivet 77 says, its ‘ipseity’ in toto, through and through” (p. 96). He advocates for highly invested exegetical signification in the use of ‫את‬. For example, Gen 1:1 should be translated, “At the first in principle created He, the Elohim, the whole ipseity of the Heavens with that of the Earth” (emphasis original). Similar translations are applied to examples of ‫ את‬in the Psalter and Isaiah. The word, then, functions not as a marker of the direct object, but as a circumlocution for fullness, principality, or basic inherent forms, and he appeals to folk etymologies to support this distinction. Ultimately, this theory is an attempt to locate the concept of the “Back of Beyond,” a major part of his theosophonic philosophy, in the Old Testament, however obliquely. I have been unable to locate any citations of this article. 78 The Back of Beyond. London: The Daniel Company, 1934 [Pseudonym “Signpost”]. This book covers much of the same ground as In Tune with the Universe, but includes a fuller discussion of the characteristics of the world of non-vibrationdom, the place where the world of effects is transcended – the “ocean of primordial being” (p. 11). This is something that need not wait for death to be 77 Fabre d’Olivet (1767–1825) was a French author and dabbling Hebraist whose hermeneutical approaches and philosophy influenced many occultists, Hoskier included. Hoskier also cites d’Olivet’s Histoire philosophique du genre humain, ou L’homme considéré sous ses rapports religieux et politiques dans l’état social, à toutes les époques et chez les différents peuples de la terre, précédée d’une dissertation introductive sur les motifs et l’objet de cet ouvrage, 2 vols. (Paris: Hubert, 1824) in In Tune with the Universe, viii and engages deeply with his work in Immortality, a book that also briefly mentions the mystical symbolism of particular Hebrew graphemes (pp. 259–61). I suspect that d’Olivet’s La Langue hébraïque restituée et les veritable sens des mots hébreux rétabil et prouvé par leur analyse radicale (Paris: Eberhart, 1815) forcefully influenced this particular article. 78 This article is similar to an extract on the “House of Jad” ( ‫ )יד‬that Hoskier sent to J. Rendel Harris on 17 May 1922 (DA21/1/2/1/25/7), but which was never published as far as I can tell. This type of overwrought lexicographic exegesis appears in muted ways in a number of Hoskier’s works.

36

Garrick V. Allen

accomplished, but which is accessible by “birth into a new dimensional activity” (p. 17). Esotericism, once again, reigns in this book, but the goal is revelation: “the following contribution is made with the objective of loosening some of the scale which obscure [Man’s] vision, and of enabling him to participate here and now in the advancement of his own interests, to the point where competition shall yield to a glorious co-operation, and when man shall rejoice in the drama in which he is an enforced actor, and play his part to the satisfaction of himself and of his audience, instead of wondering what it is all about” (p. 8). He even appeals to Jesus’ relationship to God (cf. John 10:30) as an analogy (p. 21), and points to the destabilizing nature of recent papyrological discoveries of the New Testament, arguing that recent manuscript discoveries fundamentally question established ecclesiological structures (pp. 36–7). This perspective explains his interest in the Chester Beatty Papyri that dominates his final publications. 79 Textual scholarship remains essential for his philosophical discourse, even if only by analogy. Organized religion of all stripes, however, is a false path (pp. 53–7, 77–82). Hoskier seeks freedom from the material world in the ethereal plane of the “Back of Beyond” (p. 97). The front matter of the book also suggests that Hoskier was in 1934 preparing a concordance to the Sanskrit epic Mahabharata. There is no evidence that this was ever published, but a manuscript that may be a partial draft of this work is now at the Bibliothèque municipale de Lyon. See Jan Krans’ contribution to this volume. “A Study of the Chester-Beatty Codex of the Pauline Epistles.” JTS 38 (1937): 148–63. Like much of his textual analysis, Hoskier’s interest in P46 , which carries on also into his final book, is tied to his polyglot theory. “We need only examine the text,” he opines, “in order to rest assured that we are in the presence of something which is contemporaneous with, or which may have preceded the compilation of, the Sahidic version; thus, the circumstantial evidence is definite,80 for that is generally attributed to a period circa A.D. 190” (p. 149). P46 is especially important for Hoskier because it provides the opportunity to reach a form of the Greek text that antedates the Sahidic, a version that he thinks unduly influenced later forms of the Greek tradition through polyglottal linguistic interference. Even though P46 helps to do this, it is already marked by linguistic influence from Coptic traditions, Latin, and Syriac (pp. 149–50). Hoskier waxes eloquent on being in the presence of an ancient witness, but the 79 In a preceding study, Henry A. Sanders, “The Egyptian Text of the Four Gospels and Acts,” HTR 27/2 (1933): 77–98 defends some of Hoskier’s points in Codex B and its Allies, but criticises his rejection of the papyri as valuable textual witnesses (pp. 77–8). 80 Is circumstantial evidence ever definite?

The Patient Collator

37

substance of the article is in his discussions of linguistic oddities and the ethics of scribes as copyists. There is little new insight on the text of P46. His technical ability to handle diverse materials and linguistic traditions is obvious, but not necessarily his analytical aptitude. I suspect that Hoskier is interested to comment on this manuscript because it is both new and very ancient, and his prose and structure of the article reflect his antiquarian interest, along with his usual idiosyncrasies, which are at this stage equal parts charming and alarming, like his continued and unrelenting assault on renderings in the Revised Version (p. 160). Appendix to an Article on the Chester-Beatty Papyrus of the Pauline Epistles known as P46 in the Journal of Theological Studies no. 150 Setting Forth Here in Supplementary Detail the Shorter Text of that Important Document. Oxford: Oxford University Press, 1937. The end of Hoskier’s previous article mentions a supplementary publication that he had prepared, which focused on further text-critical data pertaining to P46. Readers of the journal were encouraged to purchase this additional information since the Journal refused to publish it. This additional data comprises in this volume, which is two very short fascicles that further reflect on his previous article. With his polyglot theory still intact, Hoskier examines the omissions in P46 (vis-à-vis the Textus Receptus), arguing that many of the peculiarities of the text are the result of a “critical” approach of the scribe, by which he seems to infer intentional alterations: “revising hands were continuously busy trying to improve Paul’s epistolary methods” (p. 2). Like his earlier critiques of the text of Vaticanus, Hoskier intends to show that the early manuscripts are more corrupt than the Textus Receptus due to intentional editorial activity, although some of the “omissions” of P46 likely reflect “original” readings. The volumes includes the following data: (1) a supplementary list of omissions; (2) reflections on the degenerative nature of textual transmission (the written vs. the oral word); and (3) readings that P46 shares with the Ethiopic version (see Curt Niccum’s article in this volume). Together, the two fascicles comprise twenty-seven pages. A Commentary on the Various Readings in the Text of the Epistle to the Hebrews in the Chester-Beatty Papyrus P46 (circa 200 A.D.). London: Bernard Quaritch, 1938. The goal of Hoskier’s final book is to affirm the recently published P46 of the Chester Beatty collection81 as a most valuable witness in reconstructing the text 81 F. G. Kenyon, The Chester Beatty Biblical Papyri: Descriptions of Twelve Manuscripts on Papyrus of the Greek Bible (London: Emery Walker, 1933–1937).

38

Garrick V. Allen

of the Pauline Epistles (pp. 65–6), focusing his attention on Hebrews. But, like much of his previous work, this overarching argument is not always easy to divine among the mass of collations and scattered analysis. The book also contains a number of unexpected digressions – like the appeal to Egyptian hieroglyphics, the Book of the Dead, Sanskrit, compound verbs in the Septuagint, and the first person singular pronominal forms in Hebrew to explain a variant at Heb 9:3 (pp. 7–28, 73) – which contributes to a somewhat chaotic structure. Hoskier continued to digest new discoveries into his larger programme of critique of Hort’s text and of polyglot proclivities. The main body of the work is a textual commentary on particular readings – not always following the serial order of Hebrews – that emphasises primarily connections to the versions, especially the Bohairic (e.g. p. 63). For Hoskier, this form of the text traces its ancestry back to a different archetype than the “stereotyped” text in use before P46’s discovery – it goes back to “a very different original draft” (p. 33). This book is also interesting because Hoskier, in places, reflects on his preferred method of textual criticism. For example, he emphasises patient tedium and labour: “I commend this to any who still think there may be some short cuts to be had in textual criticism. Let this disabuse them. Indeed, there is no road at all, from where we stand, but a tangled brush. No coach and four may gallop up to it; no high-powered modern car may purr up alongside it, but the Shrine, which houses the remains of our titledeeds – the Δαβίρ, the Oracle – is plainly visible” (p. 67). The book closes (pp. 74–6) with an esoteric aside, unrelated from his preceding analysis, which attempts to connect Hoskier’s academic work to his philosophical ideals, even citing his own In Tune with the Universe. Hoskier is as much a philologist of the New Testament as he is a philologist of the beyond. 82

82 Cf. his analysis of Patience Worth’s spirit communications, noting also “to us, who study the comparative value of messages from beyond…” (DA21/1/2/1/25/5).

Hoskier’s Contribution to the Apocalypse’s Textual History: Collations, Polyglots, Groupings Juan Hernández Jr. It is easy, perhaps too easy, to dismiss the work of H. C. Hoskier as that of a madman, as that of someone so bound by idiosyncratic opinions and peculiar theories that one seeks justifications – even legitimate ones – to overlook his work. After all, what, if any, of his unusual ideas have survived to this day? The feeling only increases the further one wades into page after page of soul-deadening data – worsened, no doubt, by the tapestry of eccentricities holding it all together. Pulling at the conceptual threads of Hoskier’s thinking activates the law of diminishing returns. The cost of understanding him is too high, especially if one already knows he is wrong. That this appears to have been the case is borne out by the early reviews of his text-critical work. We are hard-pressed to find any lengthy, detailed treatment of Hoskier’s theoretical musings. The engagement, available in a handful of journals, appears to have been brief and sporadic. Piecemeal. Appreciative of his industry. Dismissive of his ideas. Extended treatment was unnecessary.1 And yet, Hoskier’s thinking, however beset by imperfections, appears to have been altogether prescient. We operate within his field of vision, even as his shade looms over our current projects.

1 See for example: Ernest de Witt Burton, The Biblical World 40 (1912): 72; Walter Drum, “Ecclesiastical Library Table: Text,” The Ecclesiastical Review 47 (1912): 747; E. J. Goodspeed, “Review of H. C. Hoskier, Concerning the Genesis of the Versions of the NT; Concerning the Date of the Bohairic Version: Covering a Detailed Examination of the Text of the Apocalypse and a Review of Some of the Writings of the Egyptian Monks,” AJT 16 (1912): 652–54; A. Plummer, “Review of H. C. Hoskier, A Full Collation of the Greek Cursive Codex Evangelium 604 (with two facsimiles) [Egerton 2610 in the British Museum]. Together with Ten Appendices (London: Nutt, 1890),” Classical Review 4 (1890): 478, col. 1–2.

40

Juan Hernández Jr.

Collations and Polyglot Theories Hoskier’s Collations At one level, the judgments of Hoskier’s peers have stood the test of time; the concerns were warranted. At the same time, the very eccentricities that pervade Hoskier’s text-critical works assure them a nearly endless shelf life. The unmatched industry and nearly impossible standards of exacting detail are part-andparcel of his status as an eccentric and – ultimately – what keeps his work from obsolescence. Hoskier’s theories of textual transmission, of course, were dead on arrival. His collations – particularly those of the Apocalypse’s Greek manuscripts – were immune to such a fate. Precision assured their longevity. Nowhere is this clearer than in Concerning the Text of the Apocalypse.2 The work transcends its theoretical trappings, discredited as they are. Criticisms of von Soden The industry Hoskier would bring to the examination of manuscripts was on display early. His dissatisfaction with the status quo illustrates it. A number of statements signal the direction of his thinking well in advance of the publication of his magnum opus. The most unforgettable, perhaps, are the criticisms levelled at von Soden’s edition of the New Testament.3 Hoskier stood indignant at the publication of an apparatus that was “positively honeycombed with errors.”4 Von Soden is indicted on several counts, including the invention of scripture, the faulty use of quotations, and the work’s complicated system and grave errors.5 Only his collations of the codices at Sinai, Jerusalem, and Athos escape condemnation.6 The problem, as Hoskier saw it, was that others had collected the material for von Soden.7 Digesting the material through the eyes of his assistants – when documents were unexamined or wrongly collated – only eroded the reliability of the apparatus. It is a mistake Hoskier was determined to avoid.8 2 H. C. Hoskier, Concerning the Text of the Apocalypse: Collations of All Existing Available Greek Documents with the Standard Text of Stephen’s Third Edition together with the Testimony of the Versions, 2 vols. (London: Bernard Quaritch, 1929). 3 Hoskier, “Von Soden’s Text of the New Testament,” JTS 15 (1914): 307–26. 4 Hoskier, “Von Soden’s Text,” 307. 5 Hoskier, “Von Soden’s Text,” 307–26; cf. Hoskier, “Manuscripts of the Apocalypse – Recent Investigations, part IV,” Bulletin of John Rylands Library 8/1 (1924): 240–41. 6 Hoskier, “Von Soden’s Text,” 307. 7 Hoskier, “Manuscripts of the Apocalypse,” 240–41. 8 Not all of Hoskier’s criticisms are well founded. See Maurice A. Robinson, “Von Soden’s Folly” (Unpublished Ph.D. Seminar Paper, 1977), 48–60; James R. Royse, “Von Soden’s Accuracy,” JTS 30 (1979): 166–71.

Hoskier’s Contribution

41

The most serious charge – that von Soden had “invented Scripture” – is also the most revealing. Hoskier rails against a form of eclecticism that conflates two or more variants to offer readings unattested in the manuscript tradition. Von Soden had thus increased the number of textual variants rather than reduced them. Hort at least had the decency to stick with his beloved Codex Vaticanus. Von Soden, however, “prints the documentary evidence from his head.”9 Considering All the Evidence Hoskier’s commitment to the evidence of readings runs throughout his text-critical works. Even Gwynn’s study of the Syriac10 – to which Hoskier is indebted – faces criticism for harmonizing Syriac readings to the Greek without manuscript support.11 The move was unnecessary. Hoskier would find Greek support even for the unattested readings,12 a scenario that repeats itself elsewhere. That every manuscript ought to be examined – irrespective of age, script, material, or language – was equally important to Hoskier. Shortcuts were prohibited. There was so much mixture and so much variation among the readings of even the oldest codices that additional evidence was necessary to separate unworthy readings from worthy ones. Hoskier’s comprehensive approach to the accumulation of data puts him in no rush to reconstruct the text of the Apocalypse (or that of any other book). Pursuing his collations and the study of individual manuscripts was a critical first step – no matter how time-consuming or taxing – that ought to precede any reconstruction. Textual reconstruction was for later generations. As Hoskier put it, Let us publish our researches and our deductions if we will, but leave [the] authoritative revision of the text for the future. There is plenty of time. The whole question is more or less academic, and the study of minutiae first is the only key which will unlock the remaining mysteries…Exhaustive methods are the only ones worth using, and accurate transcriptions or photographic copies the only ways of presenting the primary evidence of important documents.13

The sentiment keeps pace with current thinking, with the exception, of course, of delaying textual reconstruction. In that case, the future is now. For Hoskier, however, it was sufficient to provide future revisers with the material “they themselves [had] despised to accumulate and digest.”14

Hoskier, “Von Soden’s Text,” 310. See Hoskier, Concerning the Text of the Apocalypse, 2.9; cf., J. Gwynn, The Apocalypse of St. John: In a Syriac Version Hitherto Unknown (Dublin: Hodges, Figgis, and Co., 1897). 11 Hoskier, Concerning the Text, 2.9. 12 Hoskier, Concerning the Text, 2.9. 13 Hoskier, Concerning the Genesis of the Versions of the New Testament (Gospels) (London: Bernard Quartich, 1910), 1.x–xi. 14 Hoskier, Concerning the Text, 1.xiii. 9

10

42

Juan Hernández Jr.

Hoskier’s Polyglot Theory The pursuit of exacting detail is not the entire story. Hoskier’s meticulous cataloguing of minutiae serves a larger purpose. Hoskier was not simply interested in laying the groundwork for a new text that lay somewhere in the future. He sought to illuminate the transmission history of the New Testament itself – a history far more complex than his peers or predecessors could imagine. And it was the data, according to Hoskier, that tells the story. At the centre of it all lies Hoskier’s polyglot theory. According to Hoskier, early polyglots circulated well before the rise of the great fourth and fifth century uncials. Moreover, many of the “textual errors” of the great uncials betray the influence of these bilingual, trilingual, and even quadrilingual manuscripts. The theory surfaces more fully in Concerning the Genesis of the Versions of the New Testament and then in Concerning the Date of the Bohairic Version.15 The latter responds to the claims of Guidi and Leipoldt16 – both of whom date the Bohairic version late – and forms the basis for Hoskier’s textual groupings for the Apocalypse’s manuscripts. These are displayed more fully in Concerning the Text of the Apocalypse. A series of cascading analogies support Hoskier’s polyglot theory. Hoskier observes that Codex Bezae, a Graeco-Latin manuscript, preserves Greek readings that appear to fall under the influence of the Latin. These Latin readings, in turn, seem to be as old as, if not older than, the Greek. Hoskier then finds that the Latin, Greek, and Syriac versions together appear to shed light on the origin of the old Latin at the very threshold of the second century. The Coptic also strikes him as betraying Syriac and Latin influences. All told, Hoskier advances the idea that copying a bilingual manuscript, like Codex Bezae, could not have developed from a single manuscript, bilingual or otherwise. Rather, several manuscripts were being copied in close proximity. The juxtaposition of such manuscripts in several languages was akin to the multilingual inscription on the cross in Greek, Aramaic, and Latin. It was thus far more likely that Codex Bezae arose in this kind of environment than in another. It was not enough, however, to simply argue that there were other bilinguals – whether Syriac-Greek, Syriac-Latin, or Coptic-Latin – alongside the GrecoLatin and Greco-Coptic codices. As noted, Hoskier thought it more probable that there were trilingual manuscripts in Syriac, Greek, and Latin, and possibly even a great quadrilingual manuscript (!) in Syriac, Coptic, Greek, and Latin. These 15 Hoskier, Concerning the Date of the Bohairic Version: Covering a Detailed Examination of the Text of the Apocalypse and a Review of Some of the Writings of the Egyptian Monks (London: Bernard Quaritch, 1911). 16 I. Guidi, Il testo copto del testament di Abramo (Rome, 1900); J. Leipoldt, Sinuthii Archimandritae Vita et Opera Omnia, vol. 1 (Leipzig, 1906); cf. Hoskier, Concerning the Date, 117; Edgar J. Goodspeed, “Review: Hoskier’s Study of the New Testament Versions,” AJT 16 (1912): 652–54.

Hoskier’s Contribution

43

manuscripts, of course, would have been older than the Codices Sinaiticus or Bezae and set the stage for an alternative framework for understanding the New Testament’s early transmission history. Scholars could therefore no longer grant the uncials undue weight as independent witnesses to textual error. The use of multi-columned polyglots had to be taken into account as well. Hoskier’s theory also serves to bring readers closer to the original. That is to say, to the original language (or languages) of the New Testament. Hoskier thought it probable that books like Matthew, the Apocalypse, Mark, and even Luke were originally written in Aramaic.17 In fact, according to Hoskier, the endless Syriac forms in one of the Latin versions show that it is a polyglot and bolster the case for an Aramaic original or originals.18 The idea of Aramaic originals, of course, was not distinctive to Hoskier.19 The notion that polyglots were evidence of it, however, appears to have been his innovation.20 Problems with Hoskier’s Polyglot Theory Hoskier’s polyglot theory is, of course, fraught with difficulties. It was not plausible in Hoskier’s day and it remains so in ours. The reasons are clear. First, none of these “early polyglots” has survived – not the bilinguals, let alone the trilinguals or even the legendary quadrilingual. Further, the fact that skins large enough for polyglots were found – or that manuscripts with three or four columns have been discovered – does not alter the fact that no pre-fourth century, multi-columned polyglots are extant. Hoskier knows this of course, which is why he argues on internal grounds rather than external ones. We also note that Hoskier spends little-to-no time thinking through the hurdles of polyglot construction. What socio-economic or cultural conditions have to be in place for there to have been “a flurry” of polyglots or for these to have been circulating in the third, second, or even first centuries? What about the later diglots and triglots that are extant? What circumstances facilitated their production? At the very least, the analogies could have proved useful, although in all likelihood these would have upended his theory. It is the internal evidence, however, that fails so spectacularly. The study of the versions was still in its infancy and highly unreliable in Hoskier’s day. This remained the case a quarter of a century later when Josef Schmid would publish his Studien zur Geschichte des griechischen Apokalypse-Textes.21 In fact, this is Hoskier, Genesis of the Versions, 1.21–2. Hoskier, Genesis of the Versions, 1.21–2. 19 Hoskier was just one among many scholars who held this position in the early twentieth century. 20 It is unclear how committed Hoskier was to the theory of an Aramaic original, as it does not appear to resurface in later works. 21 J. Schmid, Studien zur Geschichte des griechischen Apokalypse-Textes. 2 Teil: Die Alten Stämme (Munich: Zink, 1955), now available in English, Josef Schmid, Studies in the 17 18

44

Juan Hernández Jr.

precisely why Schmid chose to focus on the Greek tradition alone.22 Only experts could use the versions without risk. Further, no reliable critical editions or monographic studies were available to examine the question of translation style or Greek Vorlagen.23 If this was the case in Schmid’s day, how much more so in Hoskier’s? A Reified Polyglot Theory But perhaps the greatest problem with Hoskier’s polyglot theory is that it does not function as a theory. Hoskier’s theory becomes an established fact in his hands. The stability of this “fact” is what allows him to account for textual variation in seemingly endless ways. These always manage to support his theory rather than refute it. The readings of Codex Sinaiticus, for example, may stem either from carelessness or from a deliberate clarification, but in both instances, the scribe may have drawn from the polyglot exemplar. Its columns – whether displaying the Syriac, Latin, or Coptic – provide the scribe and Hoskier with a rich resource for speculation. Whether the changes are accidental or deliberate, the polyglot theory remains a common denominator. Hoskier’s polyglot theory, perhaps predictably, elides into a theory of polyglot minds. The scribe’s thinking becomes another source for speculation. Scribal thought processes, for example, may reflect an interesting “polyglot mind.”24 Greek choices are made to bring out the best Coptic. Textual errors – even the same textual error – can arise from a glance at the Syriac or from the scribe’s Coptic mental processes.25 Random agreements with the Coptic merely show that the scribe made occasional rather than systematic use of the versions. The stability of the theory is a sine qua non. The explanatory options, however, vary and multiply.

Textual Groupings for the Apocalypse Hoskier’s Groupings With the polyglot theory firmly in place (and collations completed), Hoskier is able to separate the manuscript tradition into a variety of groups. There are seven

History of the Greek Text of the Apocalypse: The Ancient Stems, ed. and trans. J. Hernández Jr., G. V. Allen, and D. Müller (Atlanta: SBL, 2018). 22 Schmid, Studien, 2.x–xi. 23 Schmid, Studien, 2.x–xi. 24 Hoskier, Concerning the Date, 39. 25 Hoskier, Concerning the Date, 46, 49, 82.

Hoskier’s Contribution

45

such major groupings for the Apocalypse, four of which appear to have subgroups. The polyglot connections run throughout. The first is not a “family group” per se but three sets of uncials followed by cursives.26 The Codices Sinaiticus, Alexandrinus, Ephraemi, 025 and 046 (mistakenly dubbed B by Hoskier), plus a group of eighty cursives make up the first set. The second and third sets include 051 and 052 respectively, each of which is accompanied by a handful of cursives. The inclusion of these manuscripts into the first grouping appears to be dictated by their status as uncials, a relatively useless criterion today. The Erasmian Family27 is the second grouping and features three subgroups.28 One of these is its most ancient stem, whose archetype is surmised to have been a Graeco-Syriac bilingual.29 The third is the Complutensian, also with three subgroups, one of which has a Coptic background.30 The fourth is the (mistakenly labelled) B family, again, 046 plus a series of cursives.31 The fifth is Arethas.32 The sixth is the Graeco-Latin with two subgroups.33 And the seventh is the Egyptian, which exhibits a “syriacizing” character, also with a couple of subgroups.34 The seven groups initially surface in Concerning the Date of the Bohairic Version35 and are expanded in Concerning the Text of the Apocalypse.36 Hoskier’s manuscript groupings never caught on however; the idiosyncrasies are too many. The families are disbanded once the polyglot theory is discounted. The arrangements are altered, if not dissolved, without the threads of Hoskier’s multipurpose theory or his use of the versions. None of the groups remain intact in Schmid’s subsequent landmark study. In fact, Schmid dismisses Hoskier’s theories as “phantastisch.”37

Hoskier, Concerning the Text, 2.23. Hoskier associates the Erasmian family closely with the Textus Receptus, a connection of singular importance for his reconstruction of the New Testament’s transmission history. 28 Hoskier, Concerning the Text, 2.23. 29 That subgroup consists of 2067-743 + 2419-2051-2055-2064. See Hoskier, Concerning the Date, 3; Hoskier “Manuscripts of the Apocalypse – Recent Investigations, part V,” Bulletin of John Rylands Library 8/2 (1924): 420; and Hoskier, Concerning the Text of the Apocalypse, 1.xxxiii. 30 That subgroup consists of 42-367-468-757sup (21:9-fin.)-(1626). See Hoskier, Concerning the Text, 1.xxxiii, 2.23. 31 Hoskier, Concerning the Text, 2.23. 32 Hoskier, Concerning the Text, 2.23. 33 Hoskier, Concerning the Text, 2.23. 34 Hoskier, “Recent Investigations, part V,” 420; Hoskier, Concerning the Text, 2.23. 35 Hoskier, Concerning the Date of the Bohairic Version, 2–3. 36 Hoskier, Concerning the Text, 2.23. 37 Schmid, Studien, 2.29 n.2. 26 27

46

Juan Hernández Jr.

Altered Groupings The Erasmian Family, for example, finds twenty-one38 of its twenty-nine manuscripts39 relocated to the Andreas Text. And one40 is placed within the Koine tradition. And that most ancient stem, whose archetype was purportedly a GrecoSyriac bilingual, is found to consist almost entirely of manuscripts linked to the Andreas tradition. Hoskier’s Erasmian Family thus no longer exists as an entity in Schmid’s work, except (for the most part) as members of the Andreas subgroups. The Egyptian Family, similarly, finds eight41 of its ten manuscripts relocated among the Andreas subgroups. And the so-called “syriacizing” character of its Egyptian manuscripts or their Coptic traces42 proves inconsequential for their textual groupings or transmission history. The Arethas Family, on the other hand, fares a little better. Five manuscripts43 are initially identified by Hoskier as belonging to this group, with eight44 forming a subgroup. Schmid expands the original five into seven45 and reduces the eight of the subgroup into six,46 with two manuscripts lost to the O Family.47 The number of manuscripts in Schmid’s original subgroup, however, remains at eight since two manuscripts are added to the Arethas Text.48 The Complutensian Family perhaps weathers the transition from Hoskier to Schmid best. There are thirty-three such manuscripts listed under the Complutensian Family in Hoskier’s work,49 three shy of Schmid’s thirty-six.50 Two of Hoskier’s Complutensian manuscripts are dropped by Schmid, with one placed 38 The twenty-one manuscripts are 205 2026 2028 2029 2031 2033 2043 2056 2057 2044 2045 2054 2059 2065 2068 2069 2081 2083 2091 (2595) 2186. See Schmid, Studien, 2.26. 39 Hoskier’s full Erasmian group: 1 209 2026 2028 2029 2031 2033 (2038) 205 205 abs 2043 2056 2057 2044 2045 2049 2054 2059 2065 2068 2069 2081 2083 1894 2091 (2595) 2186 1668 1903. See Hoskier, Concerning the Text, 2.23. 40 2078, see Schmid, Studien, 2.27. 41 The eight are 2014 2015 2034 2036 2042 2043 2082 2037. See Schmid, Studien, 2.26. 42 2014 and 2030 respectively. 43 These include 91 175 242 1934 617. See Hoskier, Concerning the Text, 2.23. 44 The subgroup consists of 314 2016 664 (2070 2305) 2075 2077 (1094). See Hoskier, Concerning the Text, 2.23. 45 Schmid’s expanded group is 91 175 242 256 617 1934 2017. See Schmid, Studien, 2.27. 46 The six are 314 664 2075 2077 (1094). See Schmid, Studien, 2.27. 47 2070 2305. So Schmid, Studien, 2.29; cf., Hoskier, Concerning the Text, 2.23. 48 39A 2419. See Schmid, Studien, 2.27. 49 These are 60 35 2014 432 2023 2035 1957 2041 824 757 2061 986 1072 1075 1894 1328 1503 2352 1551 1733 1617 1771 1745 1746 1740 1637 1652 1774 2196 1864 1903 1865 1248. See Hoskier, Concerning the Text, 2.23. 50 Schmid’s list consists of 60 35 2014 432 2023 2035 1957 2041 824 757 2061 986 1072 1075 1328 1503 2352 1551 1733 1617 1771 1745 1746 1740 1637 1652 1774 2196 1864 1903 1865 1248. See Schmid, Studien, 2.28.

Hoskier’s Contribution

47

in an Andreas subgroup51 and the other eliminated altogether.52 Schmid nonetheless adds five additional Complutensian manuscripts,53 thereby returning to a total of thirty-six. As for the subgroup with a supposed Coptic background, most of its manuscripts appear to belong to the Koine Text. And as for Hoskier’s Greco-Latin group, its nine members54 are expanded into ten by Schmid and are divided almost evenly between two distinct families: Family 10455 and Family 336.56 These two families consist of mixed-texts that form their own subgroup. There are actually four such mixed texts in the Apocalypse according to Schmid. These include: 1) the Arethas Text; 2) the Complutensian Text; 3) Families 104/336; and 4) the O Family. Each of these is a mixture of the Andreas and Koine textual traditions. Thus, the four families reviewed so far – the Erasmian, the Complutensian, the Arethas, and the Greco-Latin – either turn out to reflect the Andreas tradition, the Koine tradition, or are a mixture of both. The same applies to the B family: it is mostly Koine. And the uncials that were lumped together by their status as such are now divided into two groups: the first consists of Alexandrinus and Ephraemi Rescriptus, which is largely representative of the “neutral” text;57 and the second consists of Sinaiticus, which together with P47, forms a very old text.58 The remaining uncials are repatriated to the Koine, Andreas, or the mixed-text traditions. Thus, unmoored from their polyglot connections and connected only by their affinities with the Greek tradition, an entirely new framework of textual relationships emerges in Schmid’s study. And Hoskier’s knotty groupings disintegrate. “Mystical Texts” One further observation is noteworthy. The Apocalypse is extant in a number of manuscripts that preserve no other New Testament books. This widely recognized phenomenon is distinctive to the Apocalypse’s transmission history. Hoskier thought such manuscripts merited special attention because of their independence from “ecclesiastical standardization.” Exactly what Hoskier meant by “standardization” is not altogether clear, but the presumption that a work’s appearance in a given manuscript frees it from the vagaries of a complex textual history is problematic. In fact, of the eighteen independent manuscripts singled 2014, see Schmid, Studien, 2.26. 1894, see Schmid, Studien, 2.28. 53 These are 1384 1732 2926 2431 2434. See Schmid, Studien, 2.28. 54 The nine members are 104 336 1918 459 628 582 680 922 620. See Hoskier, Concerning the Text, 2.23. 55 Family 104: 104 459 680 922 2493, see Schmid, Studien, 2.28. 56 Family 336: 336 582 620 628 1918, see Schmid, Studien, 2.28. 57 See Schmid, Studien, 2.146–47. 58 See Schmid, Studien, 2.146–47. 51 52

48

Juan Hernández Jr.

out by Hoskier,59 nearly all of them fit within known textual families. Thirteen, for example, belong to the Andreas subgroups60 and three to the Koine.61 Another two belong to the O Family62 and the Arethas text respectively.63 Two more belong to the Complutensian Family64 and another two are “stand alone” manuscripts. Nothing is made of the latter by Schmid.65 The preservation of the Apocalypse in largely non-biblical manuscripts does not preclude its connection to well-established textual families.

Conclusion Despite questions over Hoskier’s polyglot theories and textual groupings, Concerning the Text of the Apocalypse remains an indispensable resource for the study of the book – an assessment that applies almost exclusively to the second volume. Its data exceeds that of Tischendorf66 and von Soden,67 and laid the foundation for Schmid’s landmark study. The breadth, depth, scope, and detail of its collations are unlikely to be surpassed anytime soon. Scholars must build upon the second volume – and tolerate its idiosyncratic labels, groupings, and numbering system.68 The first volume can be set aside without harm. The second cannot.

These are 2018 2019 2020 2025 2038 2058 2048 2050 2055 2059 2074 2077 2080 2329 1732 2196 1678 254. See Hoskier, Concerning the Text, 1.x. 60 All of these manuscripts belong to the Andreas subgroups: 1678 2020 2059 2080 2074 and 2019 2038 2055 1732 2038 2047 2050 254, see Schmid, Studien, 2.26. 61 2025 2058 2048, see Schmid, Studien, 2.27. 62 2018, see Schmid, Studien, 2.29. 63 2077, see Schmid, Studien, 2.27. 64 1732 2196, see Schmid, Studien, 2.28. 65 2050 2329. For 2050 see Schmid, Studien, 2.25, 34, 42. For 2329 see Schmid, Studien, 2.9, 25, 31, 34, 42. 66 Constantin Tischendorf, Novum Testamentum Graece, 8 th ed., 2 vols. (Leipzig: Giesecke and Devrient, 1869–1872). 67 Hermann von Soden, Die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt hergestellt auf Grund ihrer Textgeschichte, 2nd ed. (Göttingen: Vandenhoeck & Ruprecht, 1911). 68 The idiosyncrasies that fill the pages of Concerning the Text of the Apocalypse reflect a broader set of peculiarities in Hoskier’s life and work. Anyone who explores Hoskier’s other writings discovers that he is an invariably complex figure. At times, he reads like an Orthodox Christian with a high view of Scripture and a narrow view of truth. At others, he appears to be a syncretist, who draws from multiple religious traditions in a quest for universal truths. The most striking irony, however, is Hoskier’s devotion to Spiritism. It is a bit unsettling to find that the very individual who derided von Soden for “inventing scripture” engages in Spiritism in order to establish it! See section titled “Spirit testimony” in Hoskier, Concerning the Text, 1.xxxviii. 59

Hoskier’s Contribution

49

Idiosyncrasies aside, Hoskier’s scholarship also remains a model of productivity, precision, clarity, and comprehensiveness. The passage of time appears to have vindicated Hoskier in surprising ways as well. We might disagree with Hoskier’s handling of Westcott and Hort, for example,69 but we would side with him today on the insufficiency of text types for defining textual relationships.70 We may reject Hoskier’s use of the versions for his brand of reconstruction, but today we are on a path toward integrating them more fully in our textual work. And who would take issue with Hoskier’s demand for “accurate transcriptions,” “exhaustive methods,” and “photographic copies”? It is today’s standard. The digital humanities, for example, furnish exactly what Hoskier called for – new and improved images for paleographical and codicological observations, as well as a bank of (electronic) collations with exacting accuracy.71 The number of digitization projects available to anyone with a keyboard exemplifies Hoskier’s highest aspirations.72 “Exhaustive methods are the only ones worth using, and accurate transcriptions or photographic copies the only ways of presenting the primary evidence of important documents.”73 The individual and collaborative projects currently examining all of the versions also validate Hoskier’s commitment to all the data. Every ancient version of consequence for understanding the New Testament’s transmission history is being investigated. The Syriac, the Latin, the Coptic, the Ethiopic, the Armenian, the Georgian, and even the Gothic74 have each received renewed attention, many See for example H. C. Hoskier, Codex B and Its Allies: A Study and an Indictment, 2 vols. (London: Bernard Quaritch, 1914). 70 For the state of the question, see Eldon Jay Epp, “Textual Clusters: Their Past and Future in New Testament Textual Criticism,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 519–77; see also Georg Gäbel, “‘Western Text,’ ‘D-Cluster,’ ‘Bezan Trajectory,’ Or What Else?,” in ECM Acts 3, 83–136; and Klaus Wachtel, “On the Relationship of the ‘Western Text and the Byzantine Tradition of Acts – A Plea Against the TextType Concept,” in ECM Acts 3, 137–48. 71 Cf. the various ECM projects in Münster, Birmingham, and Wuppertal: http://ntvmr.uni-muenster.de/; http://ntvmr.uni-muenster.de/web/apokalypse-edition/open/ blogs/final-transcription-of-apoc-104. Note also: http://egora.uni-muenster.de/intf/pro jekte/ecm.shtml; http://www.birmingham.ac.uk/schools/ptr/depart-ments/theologyandreli gion/research/projects/gospel-john.aspx [accessed 10 June 2018]. 72 E.g. The Center for the Study of New Testament Manuscripts: http://www.csntm.org [accessed 10 June 2018]. 73 Hoskier, Genesis of the Versions, 1.x–xi. 74 Peter J. Williams, “The Syriac Versions of the New Testament,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 143–66; Philip Burton, “The Latin Version of the New Testament,” 167–200 (in the same volume); Christian Askeland, “The Coptic Versions of the New Testament,” 201–29 (in the same volume); Askeland, “The Sahidic Apocalypse in Early Islamic Egypt,” in Studien zum Text der Apokalypse, ANTF 47, ed. M. Sigismund, M. Karrer, and U. Schmid (Berlin: De Gruyter, 2015), 271–87; Rochus Zuurmond, revised 69

50

Juan Hernández Jr.

under the auspices of the multigenerational Editio Critica Maior project.75 The limitations and possibilities of the genre for textual reconstruction are thus clearer than ever before and we are closer to settling longstanding questions over citation habits, translation styles, textual transmission, and Vorlagen recovery. Even polyglots – actual material polyglots – garner greater attention today.76 And Schmid’s landmark study – which forever wiped away Hoskier’s fraught textual groupings and overhauled our understanding of the Apocalypse’s textual history – has had its own groupings revised and updated.77 A reminder that every reconstruction is a by-product of the human imagination and subject to its limitations. There are no platonic ideals in textual criticism. Only projects. And in Hoskier’s case, whatever the oddities of his textual work, his instincts about a comprehensive approach to the manuscript data were accurate and his industry above reproach. As such, we operate within his field of vision, and the agenda we execute is his.

by Curt Niccum, “The Ethiopic Version of the New Testament,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 231–52; Martin Heide, “Zur Vorlage und Bedeutung der äthiopischen Bibelübersetzung,” in Studien zum Text der Apokalypse, ANTF 47, ed. M. Sigismund, M. Karrer, and U. Schmid (Berlin: De Gruyter, 2015), 289–313; S. Peter Cowe, “The Armenian Version of the New Testament,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 253–92; Jeff W. Childers, “The Georgian Version of the New Testament,” 293–327 (in the same volume); Carla Falluomini, “The Gothic Version of the New Testament,” 329–50 (in the same volume). 75 The most recent publication of ECM Acts, for example, even includes articles on select versions consulted for the edition. See Siegfried G. Richter und Katharina D. Schröder, “Hinweise zur Verzeichnung der koptischen Versionen,” ECM Acts 3, 72–80; Siegfried G. Richter, “Der Wegweiser zum Mittelägyptischen in der ECM,” ECM Acts 3, 221–27; Andreas Juckel, “Der ‘harklensische Apparat’ der Acta Apostolorum,” ECM Acts 3, 228–45. 76 K. Treu, “Griechisch-Koptisch Bilinguen des Neuen Testaments,” in Koptologische Studien in der DDR, Wissenschaftliche Zeitschrift der Martin-Luther-Universität, ed. J. Irmscher (Halle: Matin-Luther-Universität, 1965), 95–123; P. Nagel, “Koptischen Bibelhandschriften des Alten Testaments aus frühislamischer Zeit,” in Die Koptische Kirche in den ersten drei islamischen Jahrhunderten: Beiträge zum gleichnamigen Leucoren-Kolloquium 2002, Hallesche Beiträge zur Orientwissenschaft 36, ed. W. Beltz and H.-M. Schenke (Halle: Martin-Luther-Unversität, 2003), 131–55; Bruce M. Metzger, “Bilingualism and Polylingualism in Antiquity; with a Checklist of New Testament Manuscripts Written in More Than One Language,” in The New Testament Age: Essays in Honor of Bo Reicke, ed. W. C. Weinrich (Macon: Mercer University Press, 1984), 2.327–34; Marcus Sigismund, “Die griechisch-lateinischen Apk-Bilinguen (und andere mehrsprachige Manuskripte der Apk),” in Studien zum Text der Apokalypse, ANTF 47, ed. M. Sigismund, M. Karrer, and U. Schmid (Berlin: De Gruyter, 2015), 315–64. 77 Markus Lembke et al., Text und Textwert der griechischen Handschriften des Neuen Testaments VI: Die Apokalypse; Teststellenkollation und Auswertungen, ANTF 49 (Berlin: De Gruyter, 2017).

Herman Charles Hoskier and the Textual Criticism of Revelation Martin Karrer 1. The Work and Life of H. C. Hoskier Herman Charles Hoskier’s (1864–1938) work on Revelation is both fascinating and bewildering. He was financially independent for most of his life,1 funding his work at his own expense. This commitment deserves the highest respect. He also regarded text-critical research into the Apocalypse to be indispensable and wrote: “textual study must always be the forerunner of any interpretation. The solution of any textual difficulties must precede any final and authoritative explanation of the text.”2 This critical goal is likewise admirable. Yet, at the same time, he ran from controversy to controversy. He refused to follow the main stream in textual criticism and provoked other scholars, never holding a chair or academic post. He pursued his text-critical targets like a matador waving a red flag in front of a bull.3 And he read the prophecies of Revelation in a way that was curiously imbued with the atmosphere of the fin de siècle. The times were, as he thought, far removed from the beginnings of Christianity, gloomy, and characterised by an “agony of confusion.” The “ultra-Modernists” laboured, as he said, in vain. The Seer of the Apocalypse, however, cries out;4 his loud voice must be heard immediately in the here and now.5

Cf. the sketch of Hoskier’s biography in Garrick Allen’s contribution to the present volume and Garrick V. Allen, “‘There is No Glory and No Money in the Work’: H. C. Hoskier and New Testament Textual Criticism,” TC: A Journal of Biblical Textual Criticism 23 (2018): 1–19. 2 H. C. Hoskier, Concerning the Text of the Apocalypse. Collations of All Existing Available Greek Documents with the Standard Text of Stephen’s Third Edition, 2 vols. (London: Bernard Quaritch, 1929), 1.x. 3 Cf. Juan Hernández Jr.’s article in the present volume. 4 H. C. Hoskier, “Manuscripts of the Apocalypse – Recent Investigations, part V,” Bulletin of The John Rylands Library 8/2 (1924): 412–43 (here 442). 5 He supported this opinion in the Prolegomena of his opus magnum Concerning the Text of the Apocalypse by way of the Great War: the destruction of one third of the ships of the world between 1914 and 1918 corresponded to Rev 8:9 in his view, since the Apocalypse 1

52

Martin Karrer

From the older generation of scholars, Hoskier regarded only Frederick H. A. Scrivener (d. 1891), a conservative, as a friend.6 Otherwise, he delighted in polemics. He diminished the worth of the Codex Vaticanus against Westcott and Hort and called the theories concerning the “Syrian” (in today’s terminology, Byzantine) text of the New Testament an “old bosh.”7 He attacked Hermann von Soden’s edition,8 which was, at the time, the most up-to-date treatment of the text of Revelation. It was, according to Hoskier, unreliable and “worthy of the strongest condemnation.”9 It is no wonder that his habits and text-critical theses elicited some negative reactions from his contemporaries. This perspective is captured in a caricature produced in his life time celebrating prominent residents of South Orange, New Jersey.10 In this image, Hoskier sits at his desk, proud and elegant. Regal furniture along with a painting bespeaks his affluence. But his head is larger than his body; he is literally “egg-headed,” overly intellectual. Codices lay before him and bookrolls behind. He studies them thoroughly and accurately. Nevertheless, he is alone, isolated. He stares to a point far off in the distance. Said in another way, his access to the manuscripts, his accuracy and his care may be marvellous, but his self-confidence compels him to distance himself from his peers. He lives in his own world. Hoskier’s research on the text of Revelation took place in the midst of all these tensions. In what follows, I sketch his work on this text, proceeding from his early studies, on to his deliberations concerning the textual history of Revelation, and finally to his masterpiece, the great collation of the Apocalypse. We will see that some of his contributions remain relevant and continue to stimulate critical engagement in current research.

2. The Early Investigation into the Textus Receptus Hoskier’s early research focused on the Textus Receptus and the Gospels. In 1890 (when he was 26 years old), he published the transcription of a Gospel records that τὸ τρίτον τῶν πλοίων διεφθάρησαν (“a third of ships were destroyed”; Concerning the Text, 1.x). 6 Hoskier, Concerning the Text, 1. xi: “My old friend Dr. Scrivener…” 7 H. C. Hoskier, Codex B and Its Allies. A Study and an Indictment, 2 vols. (London: Bernard Quaritch, 1914), 1.270. 8 Hermann von Soden, Die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt hergestellt auf Grund ihrer Textgeschichte (Göttingen: Vandenhoeck & Ruprecht), 1902–1913. 9 H. C. Hoskier, “The Lost Commentary of Oecumenius on the Apocalypse,” The American Journal of Philology 34 (1913): 300–14 (here 314). 10 The caricature was posted by Peter Gurry on The Evangelical Text Criticism Blog (24 March 2017), http://tinyurl.com/yb8mgprz [accessed 6 June 2017].

Herman Charles Hoskier

53

manuscript along with a pioneering investigation into the origin of the printed text of the New Testament.11 He was thrilled by every detail and worked very accur-ately. This initial study established his renown, and his accuracy was verified by succeeding transcriptions. His fame as a detail-minded collator spread to the continent. Even Eberhard Nestle, the founder of the Novum Testamentum Graece, praised Hoskier’s meticulousness in 1910.12 Let us look back in order better to understand the achievement of his first study. The Textus Receptus developed from the edition of the New Testament by Erasmus (published in 1516 and updated in following editions up to 1535).13 Erasmus had only a few manuscripts at his disposal. Concerning Revelation, he used just one manuscript, the minuscule GA 2814, which he procured via Johannes Reuchlin (therefore, famously titled Codex Reuchlini; later on Gregory Apk 1 / Hoskier 1). In this particular manuscript, the text of Revelation is embedded in the Byzantine commentary written by Andrew of Caesarea, and Erasmus had to fill up lacunae (esp. Rev 22:16–21) and make some minor improvements to the text due to the state of the manuscript.14 Erasmus’s narrow-based edition was controversial.15 The Complutensian Polyglot which was published soon thereafter (distributed from 1522 onwards) allowed textual comparisons. Yet, Erasmus did not devote much attention to the Complutensis. He did not consider its text of Revelation and other parts of the New Testament better than that of his own edition; the polyglot, too, used only one or two manuscripts of the Apocalypse (these manuscripts were young and H. C. Hoskier, A Full Account and Collation of the Greek Cursive Codex Evangelium 604 (London: David Nutt, 1890), Appendix B (a comparison of editions from Erasmus up to Elzevir’s 1624 printing) and Appendix C, “Collation of Elzevir 1624 with Elzevir 1633.” 12 E. Nestle, “Some Points in the History of the Textus Receptus of the New Testament,” JTS 11 (1910): 564–68 (here 565). 13 Critical edition: A. J. Brown, ed., Novum Testamentum ab Erasmo recognitum, Opera Omnia Desiderii Erasmi Roterodami 6. Revelation: part IV (Leiden: Brill, 2013). 14 Cf. M. Heide, Der einzig wahre Bibeltext? Erasmus von Rotterdam und die Frage nach dem Urtext, 5 th ed. (Nürnberg: VTR, 2006), esp. 86–111; J. Krans, “Erasmus and the Text of Revelation 22:19: A Critique of Thomas Holland’s Crowned With Glory,” TC: A Journal of Biblical Textual Criticism 16 (2011): 1–19; J. M. Ross, “The Ending of the Apocalypse,” in Studies in New Testament Language and Text, NTSup 44, ed. J. K. Elliott (Leiden: Brill, 1976), 338–44. Erasmus acted jointly with a team (Gerbel, Oecolampad) in the printing house of Froben in Basel; see M. Karrer, “Der Codex Reuchlins (Minuskel 2814 GA), Erasmus und die Textgeschichte der Apokalypse,” (forthcoming). 15 Cf. U. Dill, “Kontroversen: Erasmus verteidigt seine Ausgabe,” in Das bessere Bild Christi. Das Neue Testament in der Ausgabe des Erasmus von Rotterdam, ed. U. Dill and P. Shierl (Basel: Schwabe, 2016), 167–79; R. Coogan, Erasmus, Lee and the Correction of the Vulgate: The Shaking of the Foundations (Genf: Droz, 1992) and A. Coroleu, “On the Reception of Erasmus’s Latin Version of the New Testament in Sixteenth-Century Spain,” The Bible Translator 67 (2016): 56–68. 11

54

Martin Karrer

are lost today). Thus, Erasmus made do with small corrections in his succeeding editions.16 An indeed, his edition outmatched the Complutensis. After Erasmus’s death, various printers (from Stephanus in 1550 up to Elzevir 1624 and 1633) made some changes to Erasmus’ text, following new manuscripts and new ideas. The most prominent example is a conjecture by Theodore Beza. He suggested that, at Rev 16:5, the text read ὁ ἐσόµενος (“he who will be”) instead of ὁ ὅσιος (“the holy one”). This reading alluded to Exod 3:14 and was taken up by the King James Version: “O Lord, which art, and wast, and shalt be.”17 Other corrections followed, but were minor in scope. In the foreword to their 1633 edition, Bonaventura and Abraham Elzevir dared the statement: Textum ergo habes nunc ab omnibus receptum.18 That statement became eponymous for the Textus Receptus. The Textus Receptus dominated up to the nineteenth century. The textual differences between Erasmus and Elzevir did not interest scholars in that period. Thus, Hoskier breaks through with a new perspective. He explored the genesis and early development of the Textus Receptus by comparing the editions of Erasmus, Stephanus, and so on up to Elzevir’s 1624 and 1633 printings. He noted all the differences, including punctuation, accents and breathings for the whole New Testament. In Revelation, he found very few differences between editions. The result is relevant for today and future research: the Textus Receptus of Revelation is not to be traced primarily to the text of the 1633 Elzevir edition, but rather directly back to the sixteenth century, to Erasmus and Stephanus. Even the conjecture of Beza in Rev 16:5 did not prevail in the long run against Erasmus and Stephanus’ texts: Hoskier demonstrated that Elzevir accepted Beza’s change only in the 1633 edition (not the 1624 edition).19 This was in fact a concession to the King James Version, and was corrected already in the time of the Textus Receptus. The same applies to a second change in the 1633 edition, namely the omission of καί at Rev 22:3.20 In Scrivener’s edition (1887), which, for the last time, printed the Textus Receptus as the main text, neither of these variant readings are noted in the apparatus, and the main text is based on Cf. M. Karrer, “Das Neue Testament des Erasmus und Luthers,” Theologische Zeitschrift 73 (2017): 299–324. For details concerning Revelation, see Brown, Novum Testamentum 6/4, 11–3. 17 Thus, thanks to Beza, we find a unique description of the name of God in Revelation deriving from Exod 3:14; there is no parallel in either in Hellenistic Jewish writings or in the New Testament connecting the potential future tense formulation within God’s name evoked by the Exodus passage (‫)ֶא ְהיֶה‬. J. Krans, Beyond What is Written: Erasmus and Beza as Conjectural Critics of the New Testament (Leiden: Brill, 2006) proffers more information on early conjectures. 18 Η ΚΑΙΝΗ ΔΙΑΘΗΚΗ. Novum Testamentum…, ed. Bonaventura and Elzevir (Leiden, 1633), fol. *2v. 19 Hoskier, Full Account, Appendix C, 15. 20 Noted in Hoskier, Full Account, Appendix C, 15. 16

Herman Charles Hoskier

55

Stephanus’ 1550 edition instead of Elzevir’s 1633 edition.21 Hoskier proved decisively that Scrivener’s preference for Stephanus was correct (an important element of their friendship). Hoskier’s meticulous research remains relevant up to the present day, since it facilitated two further critical investigations of the early editions of the New Testament. First, in 1910 Eberhard Nestle drew attention to the fact that one of the exemplars of Elzevir’s 1633 edition did not agree entirely with the information presented by Hoskier. A fruitful exchange between the two scholars22 resulted in the following finding: the printing process was so slow in the fifteenth and sixteenth century and, even in 1633, corrections were made during the drawn out process of the actual printing of the edition. For that reason, differences in early examples of the same print run are indicative of the fact that the exemplars originated in different stages of the print run. In a sense, the correspondence between Hoskier and Nestle anticipated the modern discipline of analytical bibliography of early printed books, which eventually led to the conclusion that every book of the early printing period is a unique witness, even if part of the same edition. Second, Hoskier’s observations helped some eighty years later to another discovery: Henk Jan de Jonge used details of Hoskier’s collation of the New Testament to solve the riddle of who prepared the text for the Elzevir print. He demonstrated that Jeremias Hölzlin, the Greek scholar from Leiden, was the editor responsible for Elzevir’s 1633 edition.23 The current hand edition of the Greek New Testament, NA28, does not present the history of the Textus Receptus and its conjectures in the apparatus. Hoskier’s study fills this gap. It therefore remains an important interlocutor for specialists in addition to the electronic databases which list the New Testament conjectures of the old editions in a more comprehensive way.24 Hoskier’s meticulous cataloguing of differences in orthography (e.g. edition 1624 ἕλκος / 1633 ἕλκκος at 16:2), punctuation (e.g. colon / middle dot after λέγοντες, disputed at 13:4 and 15:3), final consonants (e.g. ἐστι or ἐστιν at 21:16), breathings and accentuation (e.g. 18:10, where 1624 reads κρῖσίς whereas 1633 attests κρίσις, followed by modern editions) draws attention to a 21 F. H. A. Scrivener, Novum Testamentum Textûs Stephanici… (London: George Bell, 1887), 583 on Rev 16:5. 22 See Nestle, “Some Points,” 565–68, and the response by H. C. Hoskier, “The Elzevir New Testaments of 1624 and 1633,” JTS 12 (1911): 454–57. 23 H. J. de Jonge, “Jeremias Hoelzlin: Editor of the ‘Textus Receptus’ Printed by the Elzeviers Leiden 1633,” in Miscellanea Neotestamentica. Studia ad Novum Testamentum Praesertim Pertinentia a Sociis Sodalicii Batavi c.n. Studiosorum Novi Testamenti Conventus Anno MCMLXXVI Quintum Lustrum Feliciter Complentis Suscepta, NTSup 47, ed. T. Baarda et al. (Leiden: Brill, 1978), 105–28. 24 See the Amsterdam database of conjectures (created by Jan Krans) linked to the New Testament Transcripts database of the Institut für Neutestamentliche Textforschung Münster.

56

Martin Karrer

further problem in every edition. Modern editions normalize the orthography and text-organizing signs. The text-structuring devices (e.g. punctuation) and typographical matters (ranging from orthography to movable-ν) of the manuscripts are lost to large degree. It is only in recent years that all these details and their import have again garnered similar attention as devoted to them by Hoskier. Nevertheless, the Editio Critica Maior (ECM) of the New Testament will outpace Hoskier in the end, since it maintains the orthographica of manuscripts in the electronic transcriptions that underlie the edition. The ECM of Revelation will record the text-structuring characteristics of selected manuscripts and list the accents of manuscripts at places where semantic differences arise depending on accentuation. It is even planned to include select readings from some historic editions (e.g. Erasmus, Stephanus, Beza) in the apparatus. Hoskier’s habits of collation anticipated future research trends.

3. The Search for an Organization of Textual History The number of known manuscripts of the New Testament has increased century to century, and yet, additional criteria were needed to deal with various ambiguities and difficulties of the New Testament text. Erasmus instigated the pertinent deliberations at the outset of humanist engagement with the text. In his annotations, he preferred the witness of Church Fathers and held the annotations of Lorenzo Valla, the most famous humanist of the fifteenth century,25 in high regard. Vice versa, he downgraded the Latin version because he was convinced of deteriorations in the transmission of the Vulgate and wanted to improve the Latin text.26 Later generations upgraded the importance of the ancient translations of the Greek New Testament. The Latin traditions (Vetus Latina and Vulgate), especially, increased in text-critical value; at the time of Hoskier, Bousset regarded the Vulgate to be highly relevant for the text of Revelation,27 against Erasmus’ perspective. What is more, the nineteenth century brought great advances in the knowledge of the other ancient translations like Syriac, Coptic, Ethiopic, and

Cf. A. Perosa, ed., Lorenzo Valla. Collatio Novi Testamenti (Firenze: Sansoni 1970). H. J. de Jonge, “Novum Testamentum a nobis versum. The Essence of Erasmus’ Edition of the New Testament,” JTS 35 (1984): 394–413 (395–407); de Jonge, “Erasmus’ Translation of the New Testament, Aim and Method,” The Bible Translator 67 (2016): 29–41. 27 W. Bousset, Die Offenbarung Johannis, KEK 16, 6 th ed. (Göttingen: Vandenhoeck & Ruprecht, 1906), esp. 149 and 155 on the AC Vulgate group (repr. 1966). Hoskier, Concerning the Text, 1.xxxviii criticises Bousset’s text-critical studies from 1894 (Textkritische Studien zum Neuen Testament, TU 11.4, Leipzig: Hinrichs’sche, 1894) but does not deal with Bousset’s commentary. 25 26

Herman Charles Hoskier

57

Armenian. This led to disputes as to how to establish relationships between these divergent witnesses. Every textual critic from the sixteenth up to the early twentieth century cited Church Fathers and versions as textual witnesses in addition to the Greek manuscripts. Nevertheless, the majority of scholars were cautious in their appraisal of these additional witnesses, owing to the difficult questions concerning the retranslations of foreign languages into Greek and the knowledge and date of the Church fathers. R. H. Charles, for instance, did not always cite Hippolytus.28 Here, Hoskier swam against the tide in his evaluations of Revelation’s text. He paid special attention to Hippolytus, who, according to the early Church tradition, wrote his Commentarius in Joannis Evangelium et Apocalypsin shortly after 200 CE.29 Furthermore, he tied the old text form of Hippolytus to that of the young Textus Receptus.30 Consequently, the value of the Textus Receptus increased in Hoskier’s perspective. Looking back at Hoskier’s early work, one catches sight of a special love for the Textus Receptus. Hoskier inherited this love from his text-critical ancestors and Scrivener. But this love biased him for a textual theory that went beyond Scrivener and the textual criticism of his generation. He committed mistakes, like not using the text of Hippolytus in its best form31 and overestimating the possibility for conclusions from the only existing fragments. As a result, the link from Hippolytus to the Textual Receptus could not hold. Josef Schmid, who in 193032 (still during Hoskier’s lifetime) started the next great exploration on the text of the Apocalypse, sharply criticised Hoskier’s thesis and cast considerable doubt on the value of Hoskier’s use of patristic citations.33 Hoskier’s thesis concerning the versions was even bolder. He suggested that the ancient versions may be traced to the earliest period of transmission of the New Testament writings (first to third centuries CE), and postulated the existence of ancient multilingual (tri- or quadrilingual) editions for many of the New

28 Esp. R. H. Charles, A Critical and Exegetical Commentary on the Revelation of St. John, ICC, 2 vols. (New York: Scribner’s Sons, 1920). Hoskier, Concerning the Text, 1.xxxvii criticises Charles’ rejection of a reference to Hippolytus on one example passage. 29 See P. Prigent and R. Stehly, eds., “Les Fragments du De Apocalypsi d’Hippolyte,” Theologische Zeitschrift 29 (1973): 313–33. Interpretations of Revelation are also found in M. Richard, ed., Hippolytus: Commentarius in Danielem, GCS NF 7 (Berlin: De Gruyter, 2000). Nevertheless, Hippolytus’ quotations of the text of Revelation are fragmentary and incomplete. 30 Cf. Hoskier, Concerning the Text, 1.xlviii. 31 He passed over the edition of Hippolytus by H. Achelis and N. Bonwetsch, eds., Hippolytus Werke, vol. 1, Exegetische und homiletische Schriften (Leipzig: Hinrichs’sche, 1897). 32 This date is mentioned by J. Schmid, Studien zur Geschichte des griechischen Apokalypse-Textes, vol. 2, Die Alten Stämme (Munich: Karl Zink, 1955), xi. 33 Schmid, Studien, 8 n.1.

58

Martin Karrer

Testament writings.34 For Revelation in particular, he undertook to prove this theory in a 1911 study.35 His thesis proved untenable immediately, based on his view of the age of the Coptic and Syriac manuscripts, which were of special interest to Hoskier. Edgar J. Goodspeed dismissed Hoskier’s theory in an important review from 1912,36 but Hoskier himself never abandoned this idea.37 Despite the lack of acceptance of his views on the versions, we should not overlook an essential side effect: Hoskier’s thesis compelled him to document the ancient versions of Revelation in all their breadth. He ventured on that difficult project, learned the languages as far as allowed by the current state of lexicography and grammar, and collated the accessible texts. That major undertaking helped generations of scholars and should earn our respect despite his project’s many errors.38 Josef Schmid, who critiqued Hoskier one generation later, was well aware of the problems of associated with working with the old versions and avoided them altogether.39 Today, scholars from various disciplines share this burden. The Vetus Latina of Revelation was edited not earlier than two generations after Hoskier, accomplished superbly by Roger Gryson in 2000–2003 (with sharp critiques against Hoskier).40 The Syriac version was in recent years newly investigated and is now edited by Martin Heide in preparation for the ECM; the Sahidic Coptic text has also been recently edited by Christian Askeland, and variants from the Ethiopic are currently being prepared for the use in the ECM by Curt Niccum.41 While Hoskier’s use of the versions is definitively outdated by these 34 Cf. H. C. Hoskier, Concerning the Genesis of the Versions of the N.T., 2 vols. (London: Bernard Quaritch, 1910) and Juan Hernández’ contribution to the present volume. 35 H. C. Hoskier, Concerning the Date of the Bohairic Version: Covering a Detailed Examination of the Text of the Apocalypse and a Review of Some of the Writings of the Egyptian Monks (London: Bernard Quaritch, 1911). 36 Edgar. J. Goodspeed, “Review: Hoskier’s Study of the New Testament Versions,” AJT 16 (1912): 652–54. 37 Cf. Hoskier, Concerning the Text, 1.xlviii: “the scholars, who reviewed my previous volume on the date of the Bohairic Version of the Apocalypse, refused…to believe in the existence of Graeco-Syriac documents in the first, second and third centuries, unless I could offer tangible proof of a fragment of papyrus or parchment bearing a few lines of such a bilingual text. The faith that is in me, thereagainst, is based on this study of origins.” 38 See Curt Niccum’s article in this volume. 39 Schmid, Studien, x–xi and 10–2. 40 R. Gryson, ed., Apocalypsis Johannis, VL 26/2 (Freiburg: Herder, 2003), 93 comments on Hoskier’s work (cf. n.69 below). 41 These three editions are made electronically for the ECM Revelation. Cf. M. Heide, “Die syrische Apokalypse oder Offenbarung an Johannes. Kritische Edition der harklensischen Textzeugen,” in Studien zum Text der Apokalypse II, ANTF 50, ed. M. Sigismund and D. Müller (Berlin: De Gruyter, 2017), 81–187; Heide, “Die syrische Johannes-Apokalypse. Zum gegenwärtigen Stand der Forschung,” in Die Johannesoffenbarung: ihr Text und ihre Auslegung, ABG 38, ed. M. Karrer and M. Labahn (Leipzig: Evangelische

Herman Charles Hoskier

59

editions and the forthcoming publication of the ECM of Revelation, he nonetheless deserves credit for providing an overview, for the first time, of ancient translations at the variation-units of the Apocalypse. His unusual views concerning the textual history of the New Testament isolated Hoskier.42 In that situation, he published his aforementioned investigation into Codex Vaticanus (1914) and attempted to diminish the importance of this most famous New Testament codex (B). 43 He overestimated his abilities, as a contemporary reviewer charged.44 The Vaticanus controversy is of less importance for our subject, since the codex does not contain the Apocalypse (either it was missing already in the scriptorium or the portion with the Apocalypse was lost later). Indirectly, the debate stands in line with Hoskier’s assessment of Erasmus. Let us briefly return to the sixteenth century, a time when B was available in the Vatican Library, although the relevance and age of the codex were unknown. Neither Erasmus nor the editors of the Complutensian Polyglot made a journey to consult it. After the publication of his edition, Erasmus asked for variants in the codex when looking for arguments in the dispute over the Comma Johanneum. Transcriptions of illustrative passages and variants were sent to him. This list of 365 readings is lost, unfortunately. Nevertheless, we know Erasmus’ reaction: he observed that variants of B agreed with the text of the Vulgate against his proposals for a correction of the Latin text. However, he did not attempt his own thesis concerning the Vulgate, but concluded in 1534 that the Vulgate influenced the text of Vaticanus.45 This conclusion allowed Erasmus to defend his own edition, the

Verlagsanstalt, 2012), 71–82; Heide, “Zur Vorlage und Bedeutung der äthiopischen Bibelübersetzung,” in Studien zum Text der Apokalypse, ANTF 47, ed. M. Sigismund, M. Karrer, and U. Schmid (Berlin: De Gruyter, 2015), 289–314; C. Askeland, “An Eclectic Edition of the Sahidic Apocalypse of John,” in Studien zum Text der Apokalypse II, ANTF 50, ed. M. Sigismund and D. Müller (Berlin: De Gruyter, 2017), 33–79; Askeland, “The Sahidic Apocalypse in Early Islamic Egypt,” in Studien zum Text der Apokalypse, ANTF 47, ed. M. Sigismund, M. Karrer, and U. Schmid (Berlin: De Gruyter, 2015), 271–88; C. Niccum, “Apokalypse Now: The Ethiopic Version of Revelation Fifty Years after Hofmann,” in Studien zum Text der Apokalypse II, ANTF 50, ed. M. Sigismund and D. Müller (Berlin: De Gruyter, 2017), 211–30. 42 In 1914 Wilhelm Bousset put it well when he wrote: Hoskier is “a textual critic, who pursues his course all alone, away from usual roads.” (Bousset, “Textkritik II,” Theologische Rundschau 17 [1914]: 187–206 (here 199). 43 Hoskier, Codex B. 44 “This…task has proved beyond his power”: (H. St. John Thackeray, “Book Note: Codex B and its Allies,” Journal of Hellenic Studies 35 (1915): 275–76, (here 275). 45 P. S. Allen et al., eds., Opvs epistolarvm Des. Erasmi Roterodami Erasmus, vol. 2 (Oxford: Clarendon, 1910), nr. 2906, lines 39–45. Cf. J. K. Elliott, “‘Novum Testamentum editum est’: The Five-Hundredth Anniversary of Erasmus’s New Testament,” The Bible Translator 67 (2016): 9–28, esp. 19 and 22 n.36.

60

Martin Karrer

forerunner of the Textus Receptus. He erred in his underappreciation of B long before Hoskier. But did it make sense for Hoskier to diminish B anew centuries later? In the nineteenth century, in the midst of the demise of the Textus Receptus, B became a chief witness of the critical editions against the Erasmian and Byzantine textual tradition. In reaction to this development, Hoskier downgraded B and voted for a continuing preference of the Textus Receptus. His approach was too narrow, but would go on to win unexpected followers in the twentieth century, attracting attention from various Majority Text theorists, since Hoskier appeared to advocate for the Byzantine Majority text and Textus Receptus.46 Hoskier’s criticism of B, however, did not mean that he was uncritical of Erasmus. He presented for his own position when he was confronted with the rediscovery of the Codex Reuchlini, Erasmus’ base text for his edition of Revelation (GA 2814). The codex was found in the library of Maihingen (Fürsten von Oettingen) in 1850.47 Franz Delitzsch evaluated the problematic conjectures in Erasmus’ edition and condemned his editorial laxness,48 a criticism that Hoskier applauded: “Delitzsch’ [sic]…investigations of the Erasmian texts are worthy of all praise. He is hard enough on Erasmus.”49 Nevertheless, Hoskier did not doubt that Erasmus utilised an important and widely disseminated text form. He classified related manuscripts and called the text, which was typical for Andrew’s commentary tradition, the “Erasmian family.”50 Thus, the Erasmian text became part of Andrew’s textual form, the “Andreastext” as it was called later on. To be sure, Hoskier overvalued this specific textual form – the “Andreastext” is more differentiated, holding many subgroups, than he thought.51 Nonetheless, his grouping was ground-breaking. Additionally, Hoskier identified the textual form of the manuscripts that were used in Alcalá near Madrid for the Complutensian Polyglot. He categorised other manuscripts of this group and showed that the “Complutensian text” is later than Note the comments in D. B. Wallace, “Historical Revisionism and the Majority Text: The Cases of F. H. A. Scrivener and Herman C. Hoskier,” NTS 41 (1995): 280–85; and Wallace, “The Majority Text Theory: History, Methods, and Critique,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 711–44, esp. 715. 47 T. F. Karrer, “Antiquarisches,” Zeitschrift für die gesammte lutherische Theologie und Kirche 11 (1850): 121–24, esp. 122. 48 Franz Delitzsch, Handschriftliche Funde 1. Die Erasmischen Entstellungen des Textes der Apokalypse. Nachgewiesen aus dem verloren geglaubten Codex Reuchlin (Leipzig: Dörffling & Franke, 1861). 49 Hoskier, Concerning the Text, 1.7. 50 Hoskier, Concerning the Text, 1.8; cf. Hoskier, “Manuscripts of the Apocalypse – Recent Investigations, part I,” Bulletin of the John Rylands Library 7 (1922): 118–37, esp. 118. 51 State of research in M. Lembke et al., eds. Text und Textwert der griechischen Handschriften des Neuen Testaments. VI. Die Apokalypse: Teststellenkollation und Auswertungen, ANTF 49 (Berlin: De Gruyter, 2017). 46

Herman Charles Hoskier

61

the text found in Erasmus’ main manuscript (GA 2814).52 This particular finding was confirmed by later research: the text of the Complutensian Polyglot is of middle Byzantine origin.53 In the end, Hoskier’s appraisal of the versions and Hippolytus was erroneous, as was his verdict against Codex Vaticanus. His theory concerning the early transmission of the Apocalypse and other New Testament writings was untenable. But he laid the foundations for the differentiations of textual groups from the Byzantine period onwards, and he cut his losses in abstaining from his own reconstruction of the oldest text of Revelation.54 One may add another aspect in ecumenical perspective. GA 2814 (Codex Reuchlini) was purchased in Constantinople for the Council of Basel,55 a council that assayed the possibility of a union between the Eastern and Western churches. Erasmus and many generations after him, including Hoskier, neglected this historical context. Perhaps the interest in this codex and the “Erasmian” text will be renewed from a historical perspective, once the scholars rediscover the ecumenical aspects of the New Testament’s textual history.

4. The Collation of the Greek Witnesses to the Text of the Apocalypse Hoskier’s magnum opus is the collation of the Greek witnesses to the text of Revelation published in 1929.56 Transcriptions of manuscripts, examination of their textual worth, and text grouping had begun long before. Scrivener, for instance, transcribed thirteen minuscules of Revelation,57 and Tregelles had

52 Hoskier, “Recent Investigations, part I,” 119; Hoskier, Concerning the Text, 1.xv– xxxiii and 1.8. 53 Cf. M. Lembke, “Der Apokalypsetext der Complutensischen Polyglotte und sein Verhältnis zur handschriftlichen Überlieferung,” in Studien zum Text der Apokalypse, ANTF 47, ed. M. Sigismund, M. Karrer, and U. Schmid (Berlin: De Gruyter, 2015), 33–134; and U. B. Schmid, “Editing the Apocalypse in the 21st Century,” in Book of Seven Seals: The Peculiarity of Revelation, its Manuscripts, Attestation, and Transmission, WUNT 363, ed. T. Kraus and M. Sommer (Tübingen: Mohr Siebeck, 2016), 231–40, esp. 233–34. 54 Schmid, Studien, 8 puts it more sharply: “Eine eigene Rekonstruktion des Urtextes hat Hoskier nicht unternommen. Dies ist aber kein Verlust für die Wissenschaft.” 55 All the manuscripts from Basel used by Erasmus were procured for the Council of Basel in the fifteenth century. Cf. M. Wallraff, S. S. Menchi, and K. von Greyerz, eds., Basel 1516: Erasmus’ Edition of the New Testament (Tübingen: Mohr Siebeck, 2016) and U. Dill and P. Schierl, eds., Das bessere Bild Christi, Das Neue Testament in der Ausgabe des Erasmus von Rotterdam (Basel: Schwabe, 2016). 56 Vol. 2 of Hoskier, Concerning the Text. 57 Schmid, Studien, 4 n.1.

62

Martin Karrer

transcribed the Codex Reuchlini (GA 2814),58 among other examples. But not a single scholar before Hoskier undertook a comprehensive examination of all the manuscripts of one single New Testament book or sought to compile all their variant readings. Hoskier set foot in new territory. He was aware of his innovation and proud to dedicate thirty years of his life to it. He collected photographs or copies of the manuscripts, successively completed the collation, and reworked his manuscript descriptions several times.59 In all of these years, he was motivated by the desire to demonstrate that it was feasible to manage a full collation of a biblical book. He was convinced that other scholars would follow him and accomplish comparable collations of further New Testament books.60 Although this was not to be, due to the abundance of New Testament manuscripts, the goal had the desired effect. After Hoskier, all manuscripts for each biblical book were examined at Teststellen for the Text und Textwert volumes, which prepared the selection of manuscripts for the ECM volumes. These selected manuscripts are then fully transcribed. Hoskier’s comprehensiveness extends also beyond the Text und Textwert volumes. It is a dream at least of the ECM Revelation to extend the full collation to all manuscripts, made possible by the electronic medium of editing that can be updated as new manuscripts become available. Hoskier’s longstanding commitment to his project imposed a substantial personal financial burden. Von Soden and Lietzmann, who promoted textual scholarship in Germany, received financial support through Elise König. Hoskier, by contrast, was on his own financially. His wealth melted in the second half of his life, and this financial strain endangered the printing of his two volume investigation. The work was ready in 1927, as Hoskier writes in the preface, but the publication in the academic context of the University of Michigan was delayed

S. P. Tregelles, “A Few Notes on Codex Reuchlini of the Apocalypse, together with a collation of it’s [sic] text with the common editions,” in Franz Delitzsch, Handschriftliche Funde 2. Neue Studien über den Codex Reuchlins und neue Textgeschichtliche Aufschlüsse über die Apokalypse aus den Bibliotheken in München, Wien und Rom (Leipzig: Dörffling & Franke, 1862), 1–16. 59 The extensive work (Hoskier, Concerning the Text) betrays its long growth at multiple places. On occasion, Hoskier himself indicated dates of entries; for instance, he examined MS 12 (GA 181) in 1901, but supplemented it ten years later in 1911 (Concerning the Text, 1.24). 60 In his own words: “Towards the close of our labours, it becomes apparent that the task we suggest to others of the collation of the existing rich material of the other books of the N.T. is not at all superhuman, and ought to have been undertaken long ago” (Hoskier, Concerning the Text, 1.xii). 58

Herman Charles Hoskier

63

despite Hoskier’s proactive approach.61 Publication only took place two years later, when Hoskier’s friends stepped in to help.62 The lover of frank speech that he was, Hoskier articulated his endeavours for getting manuscripts in a special “Note” at the beginning of volume 1: I do not feel that I am under any obligations to the public or private libraries, the authorities of which have allowed their MSS. to be photographed at my expense, as this is an insurance on their part against destruction of such records by fire, and I have had to provide such libraries as a rule with two and sometimes three copies of each MS., which copies are probably kept apart from the MSS. themselves. But I am, of course, under considerable obligations to the Librarians of these institutions.63

I would not mention this statement had some of the manuscripts he collated not been lost subsequently. I may note here two manuscripts from Dresden: GA 2039 (= Hoskier 90 / Sächsische Landesbibl. A 95) from twelfth century and GA 241 (= 47 Hoskier / Sächsische Landesbibl. A 172) from the eleventh century. GA 2039 was burnt in the Second World War, but is documented by Hoskier, and my quest for Hoskier’s photos of GA 241 in Dresden yielded a surprising result: there are no photographs, but the manuscript is not burnt as was promulgated in the last generation. It was handed over to the Soviet military administration on 26 August 1947 as a part of the collection Matthaei, and is located today in Moscow.64 Access to this manuscript that was once thought lost is now possible,65 thanks in part to the hints of Hoskier.66 Other manuscripts of Revelation remain inaccessible, or are preserved only in incomplete photographs. In these fortunately rare cases, Hoskier’s collation currently constitutes the only extensive evidence. The ECM Revelation takes this

61 Several reports indicate that Hoskier supported the University of Michigan collection in the 1920s and early 1930s: Michigan Alumnus 36 (1930): 264; University of Michigan, The President’s Report for 1932–1933, Suppl. p. 247; University of Michigan Official Publication 41 (1939): 243, and suppl. 397. 62 See the reference at the beginning of the volume: Hoskier, Concerning the Text, 1.iii. 63 Hoskier, Concerning the Text, 1.vii. 64 The entry in Lembke et al., Text und Textwert, 5 must be complemented. I thank the Saxon State and University Library Dresden for informing me by email, 12 June 2017. 65 I am very grateful to Mikhail Seleznev who identified the manuscript in Moscow (personal correspondence in 2017) and the Institut für Neutestamentliche Textforschung Münster for providing digital photographs. 66 Hoskier thought that GA 241 not only was a witness to the Byzantine Koine text, but that it also contained valuable agreements with Hippolytus and ancient versions (Concerning the Text, 1.xxvi, 133–37). Schmid corrected Hoskier in a series of articles in 1936 (Schmid, “Untersuchungen zur Geschichte des griechischen Apokalypsetextes. Der K-Text,” Biblica 17 [1936]: 11–44, 167–201, 273–93, 429–60, esp. 438) and again in 1955 (Studien, 27). In Schmid’s view, both manuscripts of Dresden fully belong to the Koine tradition; GA 241, which Hoskier regarded so highly, has less value for the text of Revelation than Hoskier thought.

64

Martin Karrer

into consideration, recording readings for GA 242 1678 and 2039 using Hoskier’s collation.67 It was difficult for a long time to scrutinise the quality of Hoskier’s work because his photographs were lost.68 Josef Schmid, the great scholar of the textual history of Revelation of the mid-1950s,69 was the first who rigorously re-examined many of the manuscripts of Revelation. Although he criticised Hoskier’s theses regarding the early transmission of Revelation, Schmid spoke in Hoskier’s favour concerning the collation: Hoskier “collated the majority of manuscripts, which had been then known, independently and with great accuracy. Here lies the value of his work.”70 Schmid held his collations in high esteem and, at the same time, was critical when he perceived numerous inaccuracies in other areas.71 Other scholars augmented the critiques, looking at Hoskier’s errors in handling the patristic evidence and ancient versions. I quote the editor of the Vetus Latina of Revelation, the aforementioned Roger Gryson. Hoskier’s collation is, in his view, neither “an edition, nor a history of the text, but a simple collection of variants, coupled with endeavour to characterise and classify the Greek manuscripts; to our knowledge, its reliability has never been put to systematic scrutiny…If the author [Hoskier] treated the Greek witnesses with the same flippancy as the Latin ones, not to mention the oriental versions, there is truly a reason to be worried.”72 The contrast between Schmid and Gryson’s overall evaluation of Hoskier can now be solved. The Institut für Neutestamentliche Textforschung (INTF) in Münster owns microfilms of almost all manuscripts of the New Testament which have been put to use by the ECM Revelation project, outstriping the analytical Lembke et al., Text und Textwert, 7*. The search for Hoskier’s private copy of the photographs has started. How nice it would be if the current search for hard copies in Hoskier’s own estate (carried out by Garrick Allen) proved successful! 69 We cannot discuss Josef Schmid further here. His important work on the Apocalypse has recently been published in English as Juan Hernández Jr., Garrick V. Allen, and Darius Müller, trans. and eds., Studies in the History of the Greek Text of the Apocalypse: The Ancient Stems (Atlanta: SBL, 2018). 70 Schmid, Studien, 8. This corresponds to Hoskier’s comment in his Prolegomena (Concerning the Text, 1.x.): “Let it be clearly understood at the outset that my investigation of the transmission of the text of the Apocalypse has been and is being made without the slightest prejudice or bias, or preconceived ideas.” 71 Schmid, Studien, 8 n.2 asserts strongly: “Die paläographische Beschreibung der Hss ist ganz unzulänglich.” 72 Gryson, Apocalypsis, 93 (the Engligh translation is my own): “une édition, ni une histoire du texte, mais un simple recueil de variantes, assorti d’un effort pour caractériser et classer les manuscrits grecs; a notre connaissance, la fiabilité n’en a jamais été systématiquement eprouvée…Si l’auteur a traité les témoins grecs avec la même légèreté que les latins, sans parler des versions orientales, il y a vraiment de quoi s’inquiéter.” 67 68

Herman Charles Hoskier

65

potential of earlier times.73 The Wuppertal team of the ECM (directed by the author of this contribution) gathered two datasets for the test passages of Revelation’s Text und Textwert volume. One dataset was based purely on the INTF photographs, and the second also considered Hoskier’s collation.74 The comparison of data corresponded to Schmid’s observations, confirming the high regard for Hoskier’s collation of Greek manuscript, which to this day has not been seriously challenged.75 To put it plainly: in his collation of manuscripts, Hoskier was not led by his own interests, or his theories of textual history and prophecy. His collations are of high quality. On the other hand, many witnesses unknown to Hoskier were discovered in the last century, as the number of the registered manuscripts has grown from 252 to 310. J. Keith Elliott76 and David Aune77 have taken the trouble to coordinate Hoskier’s idiosyncratic numbering with the data of the current editions, and individual changes were executed in the Kurzgefasste Liste through the preparatory work on the ECM.78 The selection of manuscripts for the ECM transcends Hoskier’s important work. These observations have significant consequences. On the one hand, the previous studies of Hoskier retain their value. Hoskier, for example, documented the solecisms in the manuscripts well beyond his predecessors, and investigations on the peculiarities of scribal habits in the main manuscripts of Revelation are correct to use his collations.79 Moreover, Hoskier presents a repository for variants in all the manuscripts known to him, a body of data that will not be overtaken until the complete

73 Only a few manuscripts seen by Hoskier are now lost or missing. I mentioned the most important ones and add that already Hoskier himself lamented manuscript losses during First World War (Hoskier, Concerning the Text, 1.ix). 74 Cf. Lembke et al., Text und Textwert, 18*–25* (98*–103*). 75 See, e.g., D. C. Parker, An Introduction to the New Testament Manuscripts and Their Texts (Cambridge: Cambridge University Press, 2008), 230. 76 J. K. Elliott, “Manuscripts of the Book of Revelation Collated by H.C. Hoskier,” JTS 40 (1989): 100–11. 77 D. E. Aune, Revelation 1–5, WBC 52A (Nashville: Thomas Nelson, 1997), cxxxix– cxlviii. 78 See U. Schmid, “Die Apokalypse, überliefert mit anderen neutestamentlichen Schriften – eapr-Handscriften,” in Studien zum Text der Apokalypse, ANTF 47, ed. M. Sigismund, M. Karrer, and U. Schmid (Berlin: De Gruyter, 2015), 421–41, esp. 432–36; J. K. Elliott, “Recent Work on the Greek Manuscripts of Revelation and the Consequences for the Kurzgefasste Liste,” JTS 66 (2015): 574–84; and Lembke et al., Text und Textwert, 9*–12*. 79 Cf. James R. Royse, Scribal Habits in Early Greek New Testament Papyri (Leiden: Brill, 2008), 47 and J. Hernández Jr., Scribal Habits and Theological Influences in the Apocalypse, WUNT 2/218 (Tübingen: Mohr Siebeck, 2006), 201. Naturally, the newer studies always make use of, in addition to Hoskier, the best possible access to the documents, facsimiles, and now also microfilms, digital images, etc.

66

Martin Karrer

transcription of all manuscripts known today has been achieved by the ECM.80 Let us not forget that the ECM is not meant to ensure absolute completeness – Hoskier’s ideal – but rather to offer a representative selection, whereby its users can get a proper overview of the material. To put this more succinctly, the apparatus of the ECM will cite around 110 manuscripts, as is usual in the ECM for all New Testament books.81 This number is more than one third of all the manuscripts of Revelation. At the same time, a number of Hoskier’s manuscripts will be omitted in favour of witnesses that came to light after 1929.82 In effect, only about 80 manuscripts that Hoskier collated will be fully represented in the ECM of Revelation. Whoever wishes to consult the textual variants of the more than 160 other manuscripts that Hoskier collated will still have to refer to his volumes. It is hoped, of course, that this necessity will cease one day, when the electronic edition of the Apocalypse gradually incorporates all the manuscripts. But there is still a long way to go. In contrast to his lasting contributions, we have to mention Hoskier’s limitations. First, the new imaging technology enables researchers to recognise readings that Hoskier missed or transcribed incorrectly. For instance, multispectral images of Codex Ephraemi Rescriptus (C)83 show that our manuscript does not read βλεπεις at Rev 1:11 (as Tischendorf perceived), but βλεψεις. Also, at 3:7, it does not read κλιει (as Tischendorf and Lyon suggested), but the first hand rather wrote κλιω, which was then corrected to κλιων.84 Hoskier’s collation requires a thoroughgoing update based on modern technological capabilities. Second, there are serious imprecisions in Hoskier’s treatment of patristic evidence, extending beyond his treatment of Hippolytus. For instance, he discovered the importance of the Oecumenius text, an old textual form, and criticised von Soden for his failure to recognise it.85 But he himself examined only a meagre number of manuscripts of Oecumenius (the oldest Byzantine commentary on the Apocalypse). Hoskier’s edition of this commentary (1928), which supplemented Incidentally, Hoskier’s collation has proven valuable even in producing transcriptions for the ECM, especially, as something of a “secondary confirmation” at places where the currently available images are unclear or where the manuscript’s state of preservation has deteriorated since the time of Hoskier. 81 These manuscripts are selected according to the groups, which emerged in the whole breath of the transmission. The exact number cannot be given yet, since, in theory, transcriptions can be added until the end of the editorial process (ca. 2023). 82 Hoskier did not have access to important materials like P47, which was first edited in 1934, or P115. Cf. P. Malik, P.Beatty III (P47): The Codex, Its Scribe, and Its Text, NTTSD 52 (Leiden: Brill, 2017) and D. C. Parker, “A New Oxyrhynchus Papyrus of Revelation: P115 (P.Oxy. 4499),” NTS 46 (2000): 159–74. 83 These multispectral images were acquired by the Graduiertenkolleg Dokument – Text – Edition of the University of Wuppertal, in which the Kirchliche Hochschule Wuppertal (the main location of the ECM-Apocalypse project) participates. 84 I am grateful to Darius Müller (Wuppertal) for this reference. 85 Hoskier, “Oecumenius,” 314. 80

Herman Charles Hoskier

67

his collations,86 was from the very beginning outmoded. Marc de Groote produced an adequate edition finally outdating Hoskier in 1999.87 Schmid edited the second ancient commentary, written by Andrew of Caesarea in 1955–1956,88 and J. Hernández has recently brought some text historical consequences of Schmid’s treatment of the “Andreastext” to the fore.89 The text forms are clearer now, and the sequential citations (formulated by the interpreters inside their commentaries) of both of these major Byzantine commentaries can be collated only on the basis of these newer editions. Third, Hoskier’s references to the ancient versions must always be read in a way that respects the limited knowledge of his time and the context of his (obsolete) efforts to prove the existence of an early multilingual edition of the Apocalypse. For example, Sahidic witnesses betray some affinities with the Greek text in the papyri. This fact is easily explained, since both originated in Egypt, but Hoskier sought to identify parallels even at places where simple scribal errors seem more probable.90 To put it briefly, Hoskier’s collation is to be held in high esteem, but to be used with critical caution.

5. Conclusion Hoskier esteemed clear words and puzzling quotations. He was convinced that he was the best text-critical scholar of his time, setting the standard for accurate collations and promoting the reflection on textual history of the Gospels and Revelation. At the same time, he embellished the title page of his investigation into Codex Vaticanus with the first aphorism of Hippocrates in Greek; I paraphrase:

H. C. Hoskier, The Complete Commentary of Oecumenius on the Apocalypse. Now Printed for the First Time from Manuscripts at Messina, Rome, Salonika and Athos (Ann Arbor: University of Michigan Press, 1928). 87 M. de Groote, ed., Oecumenii Commentarius in Apocalypsin (Leuven: Peters, 1999); de Grotte, ed., Index Oecumenianus (Gent: Olms-Weidemann, 2001). 88 J. Schmid, Studien zur Geschichte des griechischen Apokalypse-Textes I. Der Apokalypse-Kommentar des Andreas von Kaisareia. Text (München: Karl Zink, 1955); Schmid, Studien zur Geschichte des griechischen Apokalypse-Textes I. Der Apokalypse-Kommentar des Andreas von Kaisereia. Einleitung (München: Karl Zink, 1956). 89 J. Hernández, Jr., “The Creation of a Fourth-Century Witness to the Andreas Text Type: A Misreading in the Apocalypse’s Textual History,” NTS 60 (2013): 1–15 and other studies. 90 Nevertheless, each instance must be discussed separately. For example, Hoskier (Concerning the Text, 2.545) relates the reading αγγε of the first hand of Sinaiticus (instead of ἄγγελον) to the Coptic eke and interprets it in the sense of αλλον (cf. ἄλλον ἄγγελον by corrector 2). On this, see Hernández, Scribal Habits, 73 n.75. 86

68

Martin Karrer

Life is short, art is long (ὁ βίος βραχύς, ἡ δὲ τέχνη µακρή)…and decision difficult. Not only must one be prepared and willing to do what is right (τὰ δέοντα) himself, but one must be prepared as well to support also the ill of the subject (νοσέων) and other external circumstances (καὶ τοὺς παρεόντας, καὶ τὰ ἔξωθεν).91

It seems to me that Hoskier related this quotation to his own labour. The first part of the aphorism expresses that his life was too short (βραχύς) to bring his art (τέχνη µακρή) to completion, and that the text-critical decisions he faced proved too difficult to produce a critical text of the Apocalypse. In that situation, he did what was necessary (τὰ δέοντα). He threw a theory concerning the origin of the Apocalypse and the New Testament into the ring, and he confidently collated the manuscripts of the Apocalypse, being the first scholar to do so exhaustively for any New Testament book. His text historical reflections are passé. His collation remains. The second part of the aphorism sounds sceptical. Hoskier construes the contemporary text-critical situation as “ill,” brought into a bad state, and he is sure, that an improvement depends on favourable circumstances needing acknowledgement and affirmation by others. He alludes to his conflicts with other scholars, annoyed by the opposition that he himself not seldom provoked. Let us turn this aphorism into a wish, following Hoskier’s analogy: May Hoskier’s readers acknowledge his merits, and may the currently available sources lead to new textual discoveries through good scholarly cooperation and favourable external circumstances.92

Hoskier, Codex B, vol. 1 title page, my own free translation. I thank Peter Malik for a first English translation, Niklas Voltmann for some further research, and Garrick Allen for English corrections. 91 92

Hoskier in the Spiritual World Jan Krans My contribution to this volume concerns a remarkable passage in the Prolegomena to the first volume of Hoskier’s Concerning the Text of the Apocalypse.1 Study of Hoskier’s worldview and sources turns out to be necessary in order to understand the shape of his text-critical work. The continuing importance of his edition of Revelation is widely acknowledged, and rightly so, but, as I will show, knowledge of this edition is essential to its proper use, a point that can be made for all historical editions of the New Testament. The passage itself runs as follows: I have laid under tribute [i.e. made serviceable to me], for what it may be worth, one of the most reliable and successful of our investigators [Hoskier refers to a person, soon to be mentioned] of psychic (or intercosmic) phenomena by dragging into my net at [Rev] xxi. 4 the only example of which I have cognisance of spirit-communication from an entity who was a bit of a textual critic, or at any rate one who was acquainted with various readings. I refer to a communication of the whole of this verse from the air, obtained by Baron Goldenstubbe on the 28 Oct. 1856, in the presence of a reliable witness (Count d’Ourches). See p. 81 and plate vi. No. 34, where the writer [i.e. the entity] had placed ὁ θεός in brackets. This is quite an interesting adjunct to our studies. It was not a case of automatic writing. In all these experiments the writing was done by the communicator without human hands or instruments. See Goldenstubbe, ‘La réalité des Esprits,’ Paris 1857. Later in the verse the τα in τα πρωτα is not clear and may be σα. It looks like the writing of a modern Greek. This is a place where the forbears of ℵ and syrS went so egregiously wrong, and where 143 now gives us a new variant of ταυτα. 2

This passage is, to say the least, intriguing. There is a verse of Revelation, written directly without human intervention, not even by means of a medium in trance, and Hoskier takes it seriously. Very seriously, despite his “for what it may be worth,” for he calls Goldenstubbe – that should actually be de Guldenstubbé – “reliable and successful” in obtaining such written communication “from the air,” and considers the result to be “quite an interesting adjunct to our studies.” 1 My thanks go to Garrick Allen for organising the Hoskier meeting, as well as to him, Juan Hernández and Jeff Cate for providing me with valuable materials and suggestions. 2 H. C. Hoskier, Concerning the Text of the Apocalypse. Collations of All Existing Available Greek Documents with the Standard Text of Stephen’s Third Edition Together with the Testimony of Versions, Commentaries and Fathers. A Complete Conspectus of All Authorities, 2 vols. (London: Bernard Quaritch, 1929), 1.xxxviii. Text in square brackets are my own insertions.

70

Jan Krans

In the following, I will try to explain what Hoskier did, and what the origin of his information was. Apparently Hoskier believes in this instance in spiritual communication. This is confirmed by the way he records the information in the second volume: Cf. Goldenstubbe: ‘La realité des Esprits’ (Paris, A. Franck 1857) ubi (pl. vi No. 34) – in hoc versu integer exscripto – verba haec “ὁ θεός” inter uncinos includebantur a scriptore (qui ex aethere scribebat sine calamo vel stylo, sine graphide nec atramento humano).3 … where – in this verse, written out in full – these words ὁ θεός were included between brackets by the writer [read: the spirit] (who wrote from the air [or ether] without reed or pen, without human pencil or ink [translation is my own].

His uncertainty about σα is also recorded.4

Hoskier’s Spiritual World In order to properly understand the background of Hoskier’s remarks quoted above, a brief sketch of his spiritual world is needed. During the preparation for the Dublin meeting in August 2017, Juan Hernandéz and I collected digital copies of Hoskier’s lesser known religious books. There are at least five of these.5 Immortality, a real book, though strange, gives Hoskier’s views on religion, spirits, history, and science. It was followed by two small booklets, What is Nivana? and The Bronze Horses, of a more meditative nature. Then In Tune with the Universe, more than 200 deep and strange pages on vibrations and non-vibrations, and finally The Back of Beyond, hard to describe, but let us say that it is about the pilgrimage to wholeness. If I were to give a short characterisation of Hoskier’s beliefs, I would call it Neo-Gnosticism, New Age, with a good dose of spiritualism – belief in spirit communication – and a fascination in science. Sparks, eternity, energy, universe, evolution, reincarnation, are recurrent along with other esoterica. The inevitable India connection is there as well. The public library of the Channel island Jersey still preserves Hoskier’s own eight-volume copy of the Mahabharata, and his typescript index of it went to the Jesuit library there, and can now be found in the Bibliothèque municipale de Lyon, together with a nice Hoskier, Concerning the Text, 2.575. Hoskier, Concerning the Text, 2.577. 5 H.C. Hoskier, Immortality (Boston: Stratford, 1925); Hoskier, What is Nirvana? (Portland, ME: Mosher, 1930); Hoskier, The Bronze Horses. A Comment on the Prose-Poem of Amy Lowell (Portland, ME: Mosher, 1930); Hoskier, In Tune with the Universe (London: Rider, 1931); Hoskier, The Back of Beyond (London: The Daniel Company, 1934). A list in Back of Beyond happily mentions Hoskier’s writings without any distinction between scholarly and religious ones. See the annotated bibliography of Garrick Allen’s contribution in this volume. 3 4

Hoskier in the Spiritual World

71

part of his scholarly library, including some hand copies of his own books, with marginal notes. The reality of spirit communication is beyond any doubt to Hoskier. In fact, “incredulity in spirit phenomena is simply and solely due to ignorance.” 6 That quote is from his book Immortality, in which he also gives seventeen proofs of spiritualism, and Baron de Guldenstubbé appears again in the fourteenth. 7 So here it becomes obvious how seriously Hoskier took this phenomenon, and how important it was to him. The date as well is significant: 1925, four years before his two volumes on Revelation were finally published.

Louis de Guldenstubbé Not much is known about de Guldenstubbé. He was Estonian, and lived in Paris in the 1850s, where he organized séances, and then moved on to even higher goals. The result is the book Hoskier used: The reality of the spirits and the marvelous phenomenon of their direct writing,8 under the heading of “positive and experimental pneumatology.” The latter has a meaning quite different from the one common in theological schools of thought, of course. The word “experimental” is typical for the scientific age, but de Guldenstubbé also insists that the phenomenon of writings appearing from the beyond is as old as writing itself. He points to the Bible for this – no, not to the Pericope de Adultera, but the stone tablets received by Moses (twice!) and the writing on the wall in Daniel. 9 He may even have gotten his inspiration from these texts. He experimented with pieces of paper left alone in a box, upon which, after a while, miraculously – or actually not more miraculously than if it were through chemistry or electricity, Hoskier would say – written text and drawings Hoskier, Immortality, 150. Hoskier, Immortality, 142–43. Hoskier describes the phenomenon as follows: “‘Directwriting,’ by which is meant the receipt on paper, without intervention of human hands, or the visible intervention of the operating agency, of written messages from the dead, such as are recorded in the book of the Russian Baron de Goldenstubbe, Paris 1857, or in the book by Robert Dale Owen: “The Debatable Land,” London 1871, see pp. 294–301, who with Kate Pox saw the operator at work. This kind of communication is difficult, but not as rare as it seems to the uninitiated, nor is it necessary to supply the operator with a pencil. At times they fabricate for themselves the necessary writing material, and the locality most favorable for this experiment is said to be the propinquity of ancient tombs in the fluidic ambiance of old cathedrals and churches.” 8 Louis de Guldenstubbé, La réalité des esprits et le phénomène merveilleux de leur écriture directe demontrées par… (Paris: Franck, 1857) (online: http://gallica.bnf.fr/ark:/ 12148/bpt6k8513383, accessed 1 March 2018). 9 Cf. de Guldenstubbé, Réalité, 73: “confirmé…par le témoignage de la Bible…” 6 7

72

Jan Krans

appeared. At first he provided a pencil as well, but it turned out that such a tool was not necessary. 10 He took care to have his eye-witnesses provide the paper, to make sure that no-one could accuse him of tampering with it. He found that churches and cemeteries yielded the best results, since spirits would linger around their tombs or be attracted to holy places. Yet his own Paris lodgings at Rue de Versailles were good enough as well. De Guldenstubbé describes direct writing as something that settles beyond any doubt the existence of spirits and of life after death, contrary to “automatic script” or all other medium-related phenomena. It is the ultimate victory of the spiritual over materialism and skepticism. It should be noted that “spiritualism” had both a long tradition and a tremendous upsurge first in the United States around 1850, rapidly spreading to Europe and connecting itself with such currents as locally existed there.11 To de Guldenstubbé, the main message from the beyond is actually surprisingly simple: the spirits convey that there is life after death, that they care about the living (those who are not yet dead), and that they can communicate with them. In my view the entire phenomenon of “direct writing” or “pneumatography” is spurious, even though I do not know how it was done in practice. With magicians such as Dynamo or Darren Brown around these days who demonstrate even more marvelous feats, one can easily image that some sort of fraud or self-deceit, or a combination of both, must have taken place. Fortunately, de Guldenstubbé’s book contains 67 plates of samples of direct writing out of more than 2000 that were produced in the course of some years. This means that the phenomenon can also be evaluated on the basis of its alleged results. Most texts are in Roman characters, and in majuscule script, but some are in minuscules. De Guldenstubbé collected numerous signatures, and mostly not of the less famous of history. It started with mere initials (Plates 6; 17; 19; 22), but signatures followed: Virgil (30); St. John (31); St. Paul (32); Melchizedek (33); Cicero (46); Plato (47); Spencer (48); Isocrates (49); Abelard (54); Hippocrates (64), and probably a few have been missed in this list. By the way, prior to producing the signatures of Cicero and Plato, the spirit is said to have first played the piano for a quarter of an hour. He/she/it should perhaps have stuck to music.

de Guldenstubbé, Réalité, 68. See John Patrick Deveney, “Spiritualism,” in Dictionary of Gnosis and Western Esotericism, ed. W. J. Hanegraaff (Leiden: Brill, 2006), 1074b–82a. 10 11

Hoskier in the Spiritual World

73

The Greek Some of de Guldenstubbé’s samples are in Greek script and language, and those are of course crucial to this contribution. Plate 27 provides the very first instance of Greek that appeared in de Guldenstubbé’s home, 4 October 1856, with three witnesses present: a professor, a count, and a baron.

Fig. 1. Automatic Writing from de Guldenstubbé (Plate 27)

A mere glance at this text makes comment almost superfluous. The clumsiness in writing is striking, including outright errors. The “entity” who produced these characters did not know Greek. The verse is 1 Cor 15:55: “Death, where is your sting? Hades, where is your victory?” Its content illustrates precisely the point all these spirit writings were intended to have: death is no threat. 12 From a text-critical perspective the spirit follows the majority reading here: ποῦ σου, θάνατε, τὸ κέντρον; ποῦ σου, ᾅδη, τὸ νῖκος, against the modern critical text (MCT) that has ποῦ σου, θάνατε, τὸ νῖκος; ποῦ σου, θάνατε, τὸ κέντρον. The latter does not have ᾅδη, but θάνατε twice, and has κέντρον and νῖκος in the inverse order. De Guldenstubbé refers to the last line in the image as the signature, but I have no idea what it says. A few weeks later, writings de Guldenstubbé called “signed by Saint John” and “signed by Saint Paul” appeared. Here the latter is shown (Plate 32). The writing is very similar to the earlier one, and just as clumsy and faulty. It could be said that actually two texts are combined here: first part of 1 Cor 15:9 is given (ἐγὼ γάρ εἰµι ὁ ἐλάχιστος τῶν ἀποστόλων); second the “signature” Παῦλος ἀπόστολος Ἰησοῦ Χριστοῦ. Text-critically interesting, perhaps, is the 12 Cf. de Guldenstubbé, Réalité, xii: “…ce fameux verset fut écrit directement en grec et signé par un Esprit inconnu.”

74

Jan Krans

word order Ἰησοῦ Χριστοῦ. The MCT order is always Χριστοῦ Ἰησοῦ, but the inverse order is attested for the beginnings of 2 Corinthians, Ephesians, Colossians, and 1 and 2 Timothy.

Fig. 2. Automatic Writing from de Guldenstubbé (Plate 32)

But what did Hoskier see in the case of Revelation? In de Guldenstubbés own words, it concerns “a strange figure, and Greek writing drawn in the presence of the count of Ourches, 28 October 1856, in the author’s lodgings. This writing brings to mind verse 4 of chapter 21 of Revelation.” 13

Fig. 3. Automatic Writing from de Guldenstubbé (Plate 34)

This is one out of two examples in which some sort of Greek minuscule script is used – the other one being some lesser-known Greek poetry14 – but as you can

de Guldenstubbé, Réalité, 81. Plate 37. De Guldenstubbé does not tell us what text it is, but it turns out to be a fragment of Solon, and as it seems directly derived from a book, for instance Poetae minori 13 14

Hoskier in the Spiritual World

75

see, it is just as clumsy and erroneous as the other examples. The spirit is even dotting the iotas, which makes me wonder how on earth Hoskier could say that “[i]t looks like the writing of a modern Greek.” Again, content-wise, we have a verse dealing with consolation and the end of death and suffering. In what is probably the most “friendly” transcription that could be made, the text is as follows: και εξαλειπει (ὁ θεὸς) πᾶν δακρυον απὸ τῶν οφθαλµῶν αὐτων, καὶ ὁ θανατος οὐκ ἐσται ἐτι οὐτε πενθος οὔτε κραυγὴ οὔτε πονος οὐκ εσται ἔτι ὁτι σα πρῶτα ἀπηλθον. Again, text-critically, it is not the critical text that we see: MCT omits ὁ θεός, has ἐκ instead of ἀπό (𝔐), and has ὅτι in square brackets. The two points Hoskier singles out are indeed clear: the bracketing of ὁ θεός in the first line and σα instead of τα before πρῶτα in the last line. The latter can be dealt with most easily: Hoskier, even if he took the writing seriously, should not have recorded it. It is a nonsense reading by a scribe (of whatever nature) whose scribal habits make one expect these types of errors. But what about the brackets around ὁ θεός, the very detail that prompted Hoskier’s remark on “an entity who was a bit of a textual critic”? Here I did what every modern textual critic would do (at least I would hope so). I realised that the use of brackets is something not seen in manuscripts, but in editions from the sixteenth century onwards to indicate parentheses, and from the eighteenth onwards to mark text-critical alternatives. I also noticed that the spirit uses the text current until the turn to the critical text in the course of the nineteenth century. So my question became straight-forward: are there contemporary editions that have brackets here? It turns out there are. All editions by Georg Christian Knapp, from 1797 onwards, have this feature, for example the fourth London edition.15

Fig. 4. Detail of Knapp’s Text of Rev 21:4

Graeci III (Leipzig: Kuhnen, 1823), 137 lines 59–62. The handling of the Greek text betrays errors and problems quite similar to the ones observed at Rev 21:4. 15 Georg Christian Knapp, Η καινη διαθηκη. Novum Testamentum Graece. Recognovit atque insignioris lectionum varietatis et argumentorum notationes subiunxit…Editio quarta (London: Priestley, 1824), 637.

76

Jan Krans

The spirit obviously did not bother much about the difference between ordinary and square brackets. One may wonder why Hoskier did not think of such a connection with a printed text. Perhaps the lure of the spiritual world was simply too strong for him. And what would he have concluded had he noticed the similarity with Knapp’s edition?

Concluding remarks Hoskier’s appeal to de Guldenstubbé’s samples can be safely buried among the curious aberrations of history. Even someone not yet sceptical about the phenomenon of spirit communication would lose all faith in it just for the sheer banality of the messages received from the beyond. In the words of an anonymous contemporary reviewer: “if the dead were to return to converse with us, it would certainly be in order to bring us something better than samples of their bad writing.”16 Hoskier’s appeal to spirit communication is an exception in his text-critical work, but it is part of his larger religious outlook, as we have seen. In fact only very rarely do his spiritualist penchants spill over into his scholarly work, though perhaps he himself would have protested at such a distinction.17 There is another instance of spirits meddling with Hoskier’s scholarly affairs, though the case would still merit further research. In the 1920s the idea was hotly debated that Mark’s Gospel as we have it in Greek might be the translation of a Latin original. 18 Hoskier himself had even more complicated ideas on the issue which need not occupy us now. 19 The University of Birmingham happens to preserve a collection of letters by Hoskier to James Rendel Harris, and the catalogue contains the following note on the one dated 25 March 1928: “Hoskier asks Mrs Curran to help him with his New Testament research: ‘Patience Worth, asked the direct question by me, says Mark’s Gospel was not first written in Latin.’” 20 Revue critique des livres nouveaux 26 (1858), 132–34, here 134: “Si les morts revenaient s’entretenir avec nous, ce serait assurément pour nous apporter quelque chose de mieux que des spécimens de leur mauvaise écriture.” 17 Hoskier does prefer an eccentric spiritual explanation of the number 666 in Rev 13:18, which is even more surprising since he so admirably collects the many gematria possibilities listed in a good number of manuscripts (Apocalypse, 2.365). 18 See especially Paul-Louis Couchoud, “L’Évangile de Marc a-t-il été écrit en Latin?” RHR 94 (1926): 161–92. 19 See H. C. Hoskier, Concerning the Genesis of the Versions of the N.T. Remarks Suggested by the Study of JP and the Allied Questions as Regards the Gospels (London: Bernard Quaritch, 1910); see also Codex B and Its Allies (1914), 1.126–39. 20 DA21/12/1/25. The letters contain more on Patience Worth, and also other promising elements. 16

Hoskier in the Spiritual World

77

In this case, Hoskier consults a famous medium to have a spirit settle an important question of New Testament textual research. Luckily for Markan scholarship Mrs Curan gave the correct answer. The other direction, Hoskier’s text-critical work spilling over in his spirituality, occurs as well. He, for example, prefers a singular reading in Rev 21:11, but we read of this only in his booklet Immortality. 21 Convictions such as Hoskier’s were not rare in the intellectual climate of his day, as odd as they may seem at first glance. Although in the case of Rev 21:4, the entire artifice shows, once again, how elements of editions may become something else entirely. It would be absurd to judge Hoskier’s scholarly work in general based on this instance of bad judgment, or, worse, on the perceived lack of Christian quality of his beliefs. Such ad hominem arguments are best left to those who do not know better. And, yes, there is a scholarly desideratum of true critical historical study of editions of the Greek New Testament. Such a thing would make for a wonderful research project, with thrilling stories to be unearthed just as the one presented here. Extensive reuse of information should be mapped, as well as gradual innovation. The long-lasting repercussions of errors once made should be laid bare, as well as the underlying mechanisms. Historical editions such as Wettstein’s (1751–1752) contain information on manuscripts that are now completely or partially lost. A more realistic perspective on our field should ensue: not everything is to be discarded that was not produced in the digital era, not all is progress. Questions tend to return because they have been forgotten, and dependency on earlier editions and traditions is the rule rather than the exception. Editions are like time capsules: they express the scholarly climate and technical possibilities of their time, which deserve to be studied, in order not only to better understand these editions and their times, but also to better locate ourselves and our present day in the history of scholarship.

21 On p. 128, Hoskier wants to read not ἔχουσαν τὴν δόξαν τοῦ θεοῦ, καὶ ὁ φωστὴρ αὐτῆς (καί is not in MCT), but instead of καὶ ὁ φωστὴρ αὐτῆς he opts for ὡς φωστὴρ αὐγῆς. He refers to a Greek manuscript (and to the Syriac), which turns out to be his no. 111 (GA 1611).

On Textual Nostalgia: Herman C. Hoskier’s Collation of Evangelium 604 (London, British Library Egerton 2610; GA 700) Revisited Jennifer Wright Knust The presence of the original is the prerequisite to the concept of authenticity.1

Intellectual engagement, as David Scott has described it, takes place within a “problem-space” around which “a horizon of identifiable stakes (conceptual as well as ideological-political) hangs.”2 Within this “problem-space,” the problems that get posed as problems, the particular questions that seem worth asking about the problems, and the kinds of answers that seem worth having “alter historically because problems are not timeless and do not have everlasting shapes.”3 Old questions lose their salience and old answers “appear lifeless, quaint, not so much wrong as irrelevant.” The reception of H. C. Hoskier’s achievements confirm Scott’s analysis: his collations endure but his “tapestry of eccentricities,” his propensity for turning a pet theory into a fact, and his interest in the “science” of spiritualism seem utterly baffling.4 Yet Scott would push the analysis further. As a postcolonial historian, Scott’s explicit interests stretch beyond intellectual history to include an accounting of what particular forms of argumentation inadvertently accomplish, and why. Historians – and perhaps textual critics can be included here as well – reconstruct the past in order to delineate future possibilities and necessities, and they do so for reasons that often lie beyond the scope 1 Walter Benjamin, “L’ouvre d’art à l’époque de sa reproduction mécanisée,” Zeitschrift für Sozialforschung 5/1 (1936): 41: “Le hic et nunc de l’original forme le contenu de la notion de l’authenticité, et sur cette dernière repose la représentation d’une tradition qui a transmis jusqu’à nos jours cet objet comme étant resté identique à lui-même”; Benjamin, Das Kunstwerk im Zeitalter seiner technischen Reproduzierbarkeit (Frankfurt: Suhrkamp, 1974), 12: “Das Hier und Jetzt des Originals macht den Begriff seiner Echtheit aus.” Benjamin, “The Work of Art in the Age of Mechanical Reproduction,” in Illuminations, ed. H. Arendt (New York: Schocken Books, 2007), 220. The French version, the first published version of Benjamin’s essay, is quite a bit longer (and less politically pointed) than the German version translated by Zohn in 1968; see Howard Caygill, Walter Benjamin: The Colour of Experience (London and New York: Routledge, 1998), 98. 2 David Scott, Conscripts of Modernity: The Tragedy of Colonial Enlightenment (Durham: Duke University Press 2004), 4. 3 Scott, Conscripts, 4. 4 See Jan Krans’ article in this volume.

80

Jennifer Wright Knust

of the arguments they explicitly present.5 Scott therefore encourages critics to ponder their own times, to ask why they compose the narratives they compose, and to wonder if what they compose will achieve the world that they want. Herman C. Hoskier’s close study of Egerton 2610 offers a fitting introduction to the problem spaces New Testament textual criticism has inhabited as well as to the early stages of his own remarkable career. A provocateur and an energetic amateur contributor to the intellectual currents of his day, Hoskier approached Egerton 2610 with great enthusiasm; this manuscript, he hoped, would emerge as a crucial witness to the “original text” of the New Testament. He designed his detailed collation as a contribution to ongoing efforts to identify this by then highly valued text, but in a way that would move the discipline closer to (rather than further from) the Textus Receptus. This effort failed. Nevertheless, his approach illuminates the “horizon of identifiable stakes” animating textual criticism, including one seemingly unshakeable principle: since the goal is the extraction of the “original text” (now Ausgangstext) from extant manuscript witnesses, individual manuscripts are valuable chiefly for the evidence they offer about a reconstructed lost text; no manuscript can ever be that text. Thus, once a witness is evaluated for its possible contribution to the recuperated “better text,” the manuscript as an object or artefact becomes irrelevant. In Hoskier’s work, and in textual criticism more broadly, this now familiar process of extraction and abstraction from manuscript to text encapsulates the disembodied stance “modern science” has often demanded. Indeed, abstraction marks a practice as science: by seeking to escape the particularities of distinctive cultural forms as well as overt human investments, modern science pursues universal truths by separating subject from object and reasoned assessment from mundane human life. The production of a text for the church but not by the church thereby enabled critics like Hoskier to convey the New Testament text into this new, scientific age. Injecting science into an emerging Protestant textual criticism governed by reason and not theology, the church’s sacred text could be restored and then returned to the faithful. Hoskier’s position as an outsider/insider to this project, however, inadvertently called attention to the motivations informing the gambit: as an outsider, he dared to admit that his scientific pursuit had a theological goal in mind. In his methodology, however, he adapted to the role of insider: displaying little interest in the circumstances and life of the manuscript he scrutinized, he sought only that manuscript’s text. The text alone was the question, a presupposition that has been symptomatic of a genealogical science that sets out to identify origins, not abiding human connections, and texts, not those who have preserved these texts.6 Textual criticism employs analytical 5 Scott, Conscripts, 46. Cf. Joan W. Scott, “Fantasy Echo: History and the Construction of Identity,” Critical Inquiry 27/2 (2001): 284–304. 6 According to the British Library catalog, the manuscript travelled from a private collection in the Byzantine East to the Albani library in Rome, to a bookshop in Germany, and

On Textual Nostalgia

81

distance, empirical study, and methodological refinement to identify relationships between texts (not manuscripts) and readings (not people). Yet, as both Scott and Hoskier can remind us, people and their problems are always involved.

Burgon’s Ghost In a preface to his Full Account and Collation of the Greek Cursive Manuscript Evangelium 604, Hoskier dedicated the publication to the memory of Dean John Burgon “in recognition of his true friendship, fellowship and helpful intercourse,” as well as to “his unflagging efforts to prepare reliable foundations upon which might securely rest the true science of textual criticism of the New Testament.”7 This dedication is followed up with a compelling account of a visit to Burgon’s study; at midnight, as the younger scholar and his mentor mounted the stairs, the Dean offered a prophecy that, Hoskier laments, has yet to come true: As surely as it is dark now, and as certainly as the sun will rise to-morrow morning, so surely will the traditional text be vindicated and the views I have striven to express be accepted. I may not live to see it. Most likely I shall not. But it will come.8

The publication of the collation of Evangelium 604 (GA 700) was intended to bring Burgon’s prediction about the vindication of the traditional text closer to fruition. In 1882, Burgon had published a letter in the Guardian affirming the great value of this codex to the discipline; not only does this manuscript agree with the traditional text at many places in the Gospels, at the time of its “discovery,” it also offered the first material confirmation in a Gospel text of a phrase in the Lord’s Prayer cited by Gregory of Nyssa, “Thy holy Spirit come upon us and cleanse us” (Luke 11:2).9 Extolling the value of this manuscript for the classification of the Gospels as well as for its corroboration of the traditional text, Burgon offered another prediction: “studious men will obtain real help in the classification of Evangelia by paying diligent attention to many a circumstance like the foregoing,” a reference to the para-textual details of this “charming little copy.” Then, indirectly conjuring another partisan defender of the Textus then, finally to the British Museum: https://www.bl.uk/catalogues/illuminated manuscripts/ record.asp?MSID=8103&CollID=28&NStart=2610 [accessed 13 June 2018]. 7 H. C. Hoskier, A Full Account and Collation of the Greek Cursive Codex Evangelium 604 (with two facsimiles) [Egerton 2610 in the British Museum], together with ten appendices (London: David Nutt, 1890), iii. 8 Hoskier, Full Account, v. 9 John Burgon, Letter to the Guardian, 26 July 1882, inserted as a frontispiece to the microfilm copy of GA 700 (British Museum Egerton 2610); available at the New Testament Virtual Manuscript Room: http://ntvmr.uni-muenster.de [accessed 8 June 2018]. The traditional text is vindicated at: Matt 5:22; 6:13; Mark 6:20; 7:19; 10:21; Luke 9:55, 56; 10:41; 22:43, 44; 23:34, 38; and John 1:18, 5:3–4.

82

Jennifer Wright Knust

Receptus, George Salmon, Regius Professor of Divinity at Trinity College Dublin, he professed to know a “learned professor at Dublin who, to some extent, will be of my opinion.”10 Burgon’s predictions, however, were only partially realised. Egerton 2610/GA 700/Evangelium 604 certainly has been received as a worthy example of a Byzantine tetraevangelion as it stood in some quarters in the eleventh century, but, despite Hoskier’s efforts to the contrary, it cannot be described as playing an important role in textual criticism as currently practiced.11 To the contrary, at this point it would be difficult to find a reputable New Testament textual critic interested in the value of this manuscript for its contribution to our understanding of the “original,” “initial,” or Ausgangstext, though it does preserve some important readings and therefore remains valuable as a matter of historical interest. As Alfred Plummer stated in his 1890 review of Hoskier’s book: The wish of Mr. Hoskier is that the truth should prevail. If it is found in the opposite direction from that which he is looking for, he will not lament that the material, which he has so conscientiously collected and so clearly presented for the use of scholars, has been instrumental in the discovery or confirmation of it. 12

Plummer was right; if anything, Hoskier’s careful work contributed, in the end, to the further demise of the Textus Receptus and a further embrace of what Plummer, channelling Burgon and Hoskier, called “the heresies of Westcott and Hort.”13 In 1890, it was still possible to at least hope to discover a manuscript or manuscripts that could somehow overturn the ascendancy of the so-called “Neutral” text, an effort that Hoskier also pursued on other grounds.14 This battle for the survival of the Textus Receptus can now seem quaint – a relic of an earlier age that, by 1890, was decisively coming to an end – and the preface to Hoskier’s Full Account does read like a kind of last gasp. Still, Burgon and Hoskier’s predictions about “types,” and the possible contribution of GA 700 to identifying them, did enjoy a longer life. Grouped as one of the witnesses to the controversial “Caesarean text,” 700 went on to play a starring role in debates about text types Burgon, Letter. I would like to thank Jan Krans for his assistance in identifying this “learned professor” as George Salmon. 11 Kurt and Barbara Aland identify the manuscript as “category III”; i.e., “manuscripts of a distinctive character with an independent text, usually important for establishing the original text but particularly important for the history of the text” (see their Text of the New Testament: An Introduction to the Theory and Practice of Modern Textual Criticism, trans. E. F. Rhodes [Grand Rapids: Eerdmans, 1987], 131, 155 [quotation from 155]). 12 Alfred Plummer, “Review of A Full Account and Collation of the Greek Cursive Codex Evangelium 604 (With Two Facsimiles) [Egerton 2610 in the British Museum. Together with Ten Appendices by Herman C. Hoskier,” The Classical Review 4/10 (1890): 478. 13 Plummer, “Review,” 478. 14 H. C. Hoskier, Codex B and Its Allies: A Study and an Indictment, 2 vols. (London: Bernard Quaritch, 1914). 10

On Textual Nostalgia

83

and their origins, with Kirsopp Lake and likeminded colleagues placing its text in this broader group and B. H. Streeter working to establish this text’s connection to Caesarea.15 Thus, 700 has been important, even if it did fail to rescue the “traditional text,” and its unique readings (above all its confirmation of Gregory of Nyssa’s text of the Lukan Lord’s Prayer) rightly inspire further reflections on the persistence of late antique church traditions even in much later texts. Hoskier’s characteristically sharp introductory comments to Full Account also raise other questions about the broader commitments of textual critics to the texts they study and the methods they pursue: by beginning his book with a dedicatory declaration of devotion to Burgon and a kind of ghost story involving a prediction delivered at midnight on the stairs of a cold, darkened house, Hoskier acknowledged that his interests extended well beyond scientific curiosity and text-critical advancement, though he did intend to use science (as he defined it) to achieve his ends. Animated by theological convictions as well as personal ties, Hoskier’s collation, he forthrightly admitted, was undertaken as a concrete expression of his desire to discover the text he hoped to find as well as to defend a mentor he admired, “a misjudged man by many, as hard a worker as any, as generous and true a heart as any brother could desire.” I would like to honor Hoskier’s example by accepting his challenge to admit that I (we) have broader commitments at stake when editing texts, examining manuscripts, mining dusty shelves, thumbing through well-worn books, or, in the digital age, scrolling through images of them. For the sake of my argument, I will label this practice “textual nostalgia” and describe the world created by this nostalgia as a Protestant text-critical habitus. Of course, it is not surprising that late nineteenth century scholars participated in the broader intellectual, political, and theological currents of their day; this is to be expected. But the uncannily disconnected connection I detect between Hoskier’s work and ours may help to illuminate contemporary problem spaces, to identify what is still at stake, and to seek a way forward. 15 Kirsopp Lake suggested that Family 13, Θ, 565, 700 and 28 were part of a group of related manuscripts (Kirsopp Lake and R. P. Blake, “The Text of the Gospels and the Koridethi Codex,” HTR 16 [1923]: 267–86 and Lake, Codex 1 of the Gospels and its Allies, TS 7 [Cambridge: Cambridge University Press, 1902]). B. H. Streeter argued that Caesarea was the origin of this type of text (The Four Gospels: A Study of Origins [New York: MacMillan. 1925], 572–81 and Streeter, “Codices 157, 1071 and the Caesarean Text,” in Quantulacumque: Studies presented to Kirsopp Lake by Pupils, Colleagues, and Friends, ed. R. P. Casey, S. Lake, and A. K. Lake [London: Christophers. 1937], 149–50). Also see Frederik Wisse, The Profile Method for Classifying and Evaluating Manuscript Evidence, SD 44 (Grand Rapids: Eerdmans, 1982), 22–3. As Eldon Epp points out, however, the theory of a Caesarean text type can no longer be sustained (“Textual Clusters: Their Past and Future in New Testament Textual Criticism,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes [Leiden: Brill, 2014], 89–92, 104).

84

Jennifer Wright Knust

Nostalgia Nostalgia has been interpreted as a structuring condition of modern life, a form of “homesickness” (Heimweh) that, in Lynne Huffer’s terms, “both creates and obliterates a lost object.”16 As such, it plays a role in a broader symptom of modernity: the melancholic sense that the original has been lost and yet remains deeply necessary. As Walter Benjamin observed in 1936: “The authenticity of a thing is the essence of all that is transmissible from its beginning, ranging from its substantive duration to its testimony to the history which it has experienced.”17 This authenticity, however, is threatened, if not destroyed, Benjamin argued, by fascism, mechanization, and reproducibility, as “the original” (in Benjamin’s essay, auratic, pre-modern art) recedes behind multiplicity, producing critics who analyse instead of participants who experience.18 It is not that modernity has lost authenticity; rather the sense that authenticity has been lost is fundamental to the production of the idea “we are modern.” Nineteenth-century textual nostalgia fits well within Benjamin’s observations. According to political scientist Kimberly Smith, the term Heimweh (from which the English term “nostalgia” was drawn), was invented in 1688 to describe the despair of Swiss mercenaries fighting in foreign wars and was identified as a “sad mood originating from the desire to return to one’s native land” with symptoms that included “persistent thinking of home, melancholia, insomnia, anorexia, weakness, anxiety, smothering sensations, and fever.”19 “Nostalgia” proper entered English as a translation for Heimweh in 1756 when J. G. Keysler’s Travels through Germany, Hungary, Bohemia, Switzerland, and Lorrain was published: the fine mountain air of Switzerland, Keysler reported, leads the Swiss to “feel a kind of anxiety and uneasy longing after the fresh air to which they were accustomed from their infancy, without being able to account for such

16 Lynne Huffer, Maternal Pasts, Feminist Futures: Nostalgia, Ethics, and the Question of Difference (Stanford: Stanford University Press, 1998), 3. Huffer is addressing a logic that “comes out of a particular Western, deconstructive tradition of thought,” and not homesickness/Heimweh per se. 17 Benjamin “L’ouvre d’art,” 43: “L’authenticité d’une chose intègre tout ce qu’elle comporte de transmissible de par son origine, sa durée matérielle comme son témoignage historique.” Das Kunstwerk, 221: “Die Echtheit einer Sache ist der Inbegriff alles von Ursprung her an ihr Tradierbaren, von ihrer materiellen Dauer bis zu ihrer geschichtlichen Zeugenschaft.” 18 Cf. Rainer Rochlitz, The Disenchantment of Art: The Philosophy of Walter Benjamin, trans. J. M. Todd (New York: Guilford Press, 1992), 155–57 and Caygill, Benjamin, 102– 103 19 Kimberly K. Smith, “Mere Nostalgia: Notes on a Progressive Paratheory,” Rhetoric and Public Affairs 3/4 (2000): 509–10.

On Textual Nostalgia

85

disquietude.”20 Taken up in medical, psychoanalytical, and philosophical literature to describe a more generalized longing to escape the present by means of the past, by the nineteenth century, Heimweh-Nostalgia referred broadly to the retrospective re-imagination of a place that never fully existed. The history of these two terms is not accidental, Smith argues; rather, their emergence was a “rhetorical artifact of the politics of industrialization,” a politics that imagined Europe marching forever forward by means of rational administration and ever-improving technology but mourned what was felt to be lost in the process.21 This type of nostalgia is also utopian, Andreas Huyssen avers, in the sense that it re-claims an imaginary past that was never there to begin with: the decay of what went before leaves its traces in ruins, fragments, and artefacts which in turn invite an awareness of the contingency of every human project, including those undertaken in the name of the triumphant present. 22 Gazing at remainders, the material traces of a lost past invite both re-animation of the lost object in the present and a palpable awareness that the past is irrevocably absent. Provoked by a promise of a return, modern nostalgia therefore contains within it the knowledge that one can never actually go home again. Textual nostalgia, as I am formulating it, refers to the twin sense that (a) thanks to the irreversibility of time, critics will almost certainly be prevented from recovering the desired past textual object in its plenitude, and (b) that the project of recovery must necessarily be undertaken anyway, in fulfilment of the progressive march forward characteristic of modern scientific advancement. Such a process of recuperation within progress is, nineteenth-century textual critics openly acknowledge, likely to hurt – it is supposed to hurt, as all progress must – but this pain, they also argued, is a small price to pay for a deeply desired, scientifically reclaimed authentic text. 20 J. G. Keysler, Travels through Germany, Hungary, Bohemia, Switzerland, and Lorrain. Giving a True and Just Description of the Present State of those Countries; their Natural, Literary, and Political History; Manners, Laws, Commerce, Manufactures, Painting, Sculpture, Architecture, Coins, Antiquities, Curiosities of Art and Nature, Etc., 2nd ed., 4 vols. (London: G. Keith, A. Linde, S. Crowder, P. Davey and B. Law, T. Field, and E. Dilly, 1756), 1.141. Keysler explains that “nostalgia” signifies “a pain from being denied a return to one’s own country.” 21 Smith, “Mere Nostalgia,” 506–507; also see Linda Austin, “The Nostalgic Moment and the Sense of History,” Postmedieval: A Journal of Medieval Cultural Studies 2/2 (2011): 130. 22 “What is allegedly present and transparent whenever authenticity is claimed is present only as an absence. It is the imagined present of a past that can now be grasped in decay…Especially in the post-Enlightenment discoveries of origins and national identities, the present of modernity appeared – more often than not – as a ruin of authenticity and a better and simpler past…the authentic ruin as a product of modernity itself rather than as a royal road toward some uncontaminated origin” (Andreas Huyssen, “Authentic Ruins: Products of Modernity,” in Ruins of Modernity, ed. J. Hell and A. Schönle [Durham: Duke University Press, 2010], 21–2; also see Huyssen, “Nostalgia for Ruins,” Grey Room 23 [2006]: 6–21).

86

Jennifer Wright Knust

Thus defined, textual nostalgia can readily be detected in Hoskier’s attempt to restore the Textus Receptus to its rightful status. As he states: Our scientific contention briefly then is this: that every passage must be weighed most carefully in whatever Manuscript occurring apart from as well as in connection with the characteristics of that MS, or of other so-called sister MSS, which agree with it in general or in that particular place. 23

This kind of disciplined weighing will likely lead, Hoskier hypothesized, to the victory of the church’s traditional text: ‫א‬, B, D, L, Δ, etc. are as “lonely” as 28, 33, 69, 131, and 700 once their singular readings are examined, placing them equally in the company of the “scribes or rather the compliers” who acted as if they were “critics” engaged in an “altogether independent treatment of the Sacred Text.”24 Rejecting the genealogical method as practiced by Westcott and Hort, Hoskier argued that every manuscript must be subjected to “the most searching criticism”; no single manuscript, including a minuscule like 700, can too quickly be taken as representative of an entire family and then either dismissed or elevated to a status of closest to or farthest away from “the original text.” ‫ א‬and B might also be wrong, Hoskier insisted, and thus the “Revisers” (that is, those responsible for the Revised Version English translation) “fell into a trap laid for them by Lachmann, Tischendorf, and Westcott and Hort” when they allowed themselves to be persuaded to “narrow down” the evidence to a few witnesses and to “blindly oppose” others.25 The solution is to treat each manuscript individually, weighing every witness intrinsically and extrinsically, “up and down, and round and round” while seeking to identify “what is foisted onto the Sacred Text and what is the genuine survival of ancient readings.”26 This is where the Revisers and the textual critics who inspired them failed, he implies: he is willing to admit that “the bulk of the cursives may be wrong even when numerically holding the balance,”27 but the Revisers have displayed an unscientific favouritism for a few peculiar witnesses without keeping the majority reading in view. Science may require that certain traditional readings be abandoned, Hoskier admits, but he at least is willing to suffer such wounds for the sake of the Sacred Text.28 What about the Revisers? They seem determined to cling to ‫ א‬or B at all costs and, even when they select a reading that conflicts with ‫ א‬or B, their reasoning is often flawed; favouring the shorter reading, they overlook the majority reading, and are led astray as a result.29 Hoskier, Full Account, xvii. Hoskier, Full Account, xiv. 25 Hoskier, Full Account, xvii. 26 Hoskier, Full Account, xv. 27 Hoskier, Full Account, xvi, emphasis added. 28 Hoskier, Full Account, xvii. 29 Hoskier, Full Account, xvii. 23 24

On Textual Nostalgia

87

The point here is not to defend Hoskier’s perspective per se but rather to notice the logical structure of this rhetoric:30 the goal is to restore the genuine Sacred Text; the method must be “scientific” (all available data must be considered, analysis must proceed rationally and systematically, and technological advancements should be employed whenever possible); the approach must be “unbiased”; and the results of this critical work must be accepted, even when the result is felt to be against the critic’s own self-interest or theological convictions. Westcott and Hort offered a similar rationale for their own methods, despite Hoskier’s differences with them. Progress in the discipline of textual criticism, they argued, consists not in the growing perfection of an ideal in the future, but in approximation towards complete ascertainment or definite facts of the past, that is, toward recovering an exact copy of what was actually written on parchment or papyrus by the author of the book or his amanuensis.31

Such work is made difficult by the act of transmission itself, the long history of the development of the New Testament canon, unwitting scribal incorporation of various harmonized and interpolated readings, the loss of so many early copies, as well as a number of other factors.32 Consequently, criticism demands a disciplined attempt to strive against individual bias,33 including conformity “to rationally framed or rather discovered rules” which “impose salutary restraints” on “arbitrary or impulsive caprice.”34 An approximated authorial “original” text can then be realised by the identification of “neutral” readings uncorrupted by Western, Alexandrian, and (especially) Syrian influence. Unlike Hoskier, who in Full Account advocated for scrutiny of each manuscript individually, Westcott and Hort’s placement of manuscript witnesses within textual families they labelled Western, Alexandrian, and Syrian situated them more firmly within the genealogical discourse that was a hallmark of Victorian science:35 working to identify the character of individual texts, they placed 30 Cf. Daniel Wallace, “Historical Revisionism and the Majority Text Theory: The Cases of F. H. A. Scrivener and Herman C. Hoskier,” NTS 41 (1995): 280–85. 31 B. F. Westcott and F. J. H. Hort. The New Testament in the Original Greek, 2 vols. (London: Macmillan, 1881–1882), 2.3 32 Westcott and Hort, New Testament, 2.4–11. 33 A problem they sought to address, at least in part, by working independently and then comparing their results prior to printing their edition: “This combination of completely independent operations permits us to place far more confidence in the results than either of us could have presumed to cherish had they rested on his own sole responsibility. No individual mind can ever act with perfect uniformity, or free itself completely from its own idiosyncrasies: the danger of unconscious caprice is inseparable from personal judgment” (Westcott and Hort, New Testament, 2.17). 34 Westcott and Hort, New Testament, 2.65. 35 On the centrality of racialized genealogies to of nineteenth-century scientism, see Maurice Olender, The Languages of Paradise: Race, Religion, and Philosophy in the Nineteenth

88

Jennifer Wright Knust

them within hierarchalizable family trees that, in theory at least, could identify what was “corrupt” and what was, relatively speaking, “neutral.” As contemporaries of Darwin and heirs to the philological work of, for example, Johann Gottfried Herder,36 by exploring the question of origins, families, and descent, New Testament scholars certified their participation in a broader scientific discourse determined to trace languages, nations, races, human cultures, and species to their points of origin.37 In textual criticism, this procedure also had the advantages of reducing the vast number of surviving manuscripts into manageable groupings and potentially linking these texts to specific locales, thereby offering a window not only into texts but also into the contexts in which they arose.38 Indeed, though Hoskier accused Westcott and Hort of applying their genealogical method unsystematically, he also adopted the method in his subsequent work.39 His later studies did not fare any better at vindicating the traditional text,

Century, trans. A. Goldhammer (Cambridge, MA: Harvard University Press, 1992); Molly Rogers, Delia’s Tears: Race, Science and Photography in Nineteenth-Century America (New Haven: Yale University Press, 2010), esp. 117–30, 195–211); Veronika Lipphardt, “Isolates and Crosses in Human Population Genetics; or, A Contextualization of German Race Science,” Current Anthropology 53/5 (2010): S69–S82; Will Abberley, “Race and Species Essentialism in Nineteenth-Century Philology,” Critical Quarterly 36/1 (2012): 45–60; and Douglas A. Lorimer, “Science and Secularization of Victorian Images of Race,” in Victorian Science in Context, ed. B. Lightman (Chicago: University of Chicago Press, 2012), 212–35. For a helpful overview of the history of text types as an editorial strategy, as well as a discussion of their rejection in current contexts, see Tommy Wasserman and Peter J. Gurry, A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method (Atlanta: SBL, 2017), 7–10 (with bibliography). On Wescott and Hort as avid participants in the genealogical science of their age, see Yii-Jan Lin, The Erotic Life of Manuscripts: New Testament Textual Criticism and the Biological Sciences (Oxford: Oxford University Press, 2016), 71–7. 36 Johann Gottfried Herder, “Essay on the Origin of Language” (1772), in On the Origin of Language: Two Essays by Johann Gottfried Herder and Jean-Jacques Rousseau, trans. J. H. Morann and A. Gode (New York: Frederick Ungar, 1966), 87–176. In 1860, Hort wrote to Westcott asking, “Have you read Darwin? How I should like a talk with you about it! In spite of difficulties, I am inclined to think it unanswerable. In any case it is a treat to read such a book” (F. J. A. Hort, Life and Letters of Fenton John Anthony Hort, ed. A. Hort [New York: MacMillan, 1896], 1.414). Their discussion of the implications of Darwin’s work continued for many years (see, e.g., 1.430–32) 37 See David C. Parker in this volume. Also see Rachel Bryant Davies, Troy, Carthage, and the Victorians: The Drama of Classical Ruins in the Nineteenth Century Imagination (Cambridge: Cambridge University Press, 2018) for further discussion of the entanglements of science with classical and biblical studies in the nineteenth century. 38 Epp, “Textual Clusters,” 523–41. 39 Full Account and the investigation of individual manuscripts failed to protect the traditional text, but perhaps a refined genealogical method could, or so Hoskier seems to have hoped. See the excellent and extensive overview of his work by Garrick V. Allen in this volume.

On Textual Nostalgia

89

but Hoskier’s meticulous collations did contribute to the development of text type family theories articulated by others.40 Hoskier, Westcott, and Hort differed in the lessons they drew from their application of genealogical methods, but they were united in their elevation of scientific detachment as a necessary precondition for all scholarly inquiry. Science, they agreed, requires a scrupulous bracketing of confessional commitments. As Westcott and Hort conceded, however, the work of editing the New Testament text imposes constraints on editors that other texts do not. They therefore sought a compromise capable of trading one set of hurt feelings (those experienced by the church) for another (those experienced by scientists):41 None can feel more strongly than ourselves that it might at first sight appear the duty of faithful critics to remove completely from the text any words or passages which they believe not to have originally formed part of the work in which they occur. But there are circumstances connected with the text of the New Testament which have withheld us from adopting this obvious mode of proceeding.42

In other words, critics are professionally obligated to delete fraudulent passages once identified, but, when confronted by the tender sensibilities of the Christian faithful (the “circumstances connected with the text of the New Testament”), critics may have no choice but to print the text against their own, scientifically obtained results. Westcott and Hort therefore retained some of the verses and passages they would have preferred to delete, albeit in brackets, accompanied by notes, or placed in an appendix. Though he did not print his own edition, Hoskier made similar concessions: he was prepared to accept the pain of removing passages he might prefer to preserve, he reported, if only he could be satisfied that the case for their omission had been fairly and properly adjudicated. He objected to Westcott and Hort’s bracketing decisions, but not to the custom per se.43

Hoskier’s attempt was widely regarded as a failure, even in his own lifetime. According to Bruno Latour, this performance of “standing outside” is a key ideological proposition among those claiming to be “modern” (We Have Never Been Modern, trans. C. Porter [Cambridge, MA: Harvard University Press, 1993], 100–29). 42 Westcott and Hort, New Testament, 2.294. 43 As he noted in his even more polemical Codex B and Its Allies, the decision to bracket, for example, πρὸς αὐτόν at Luke 9:62 “on the sole authority of B,” without accounting for the omission of the same phrase in 604 (GA 700) exposes their “arbitrary judgement”; he continued, “These remarks should suffice as to definite examples of the unscientific use of the margin as well as of the text whether bracketed or not” (Hoskier, Codex B, 1.1, 3). The now familiar custom of bracketing formerly beloved texts offers a striking material expression of the textual nostalgia I am describing here. Such bracketing is a fitting material symbol of what it has meant to produce a “modern text,” a visible proof of a willingness to embrace the secular distancing modern criticism demands as well as an overt writing in of the cost of this procedure for those tempted to resist its premises and its conclusions. For further discussion, see Jennifer Knust and Tommy Wasserman, To Cast the First Stone: The 40 41

90

Jennifer Wright Knust

Thus the validity of the work of scientific textual criticism was certified, at least in part, by an acknowledgment that a return of the textual object to its original state – to its “home” – requires editors to embrace scientific advancement as the first priority and to willingly suffer the consequences of what this science has wrought. Such a first step/second step approach to knowledge partakes in the “structure of belatedness” characteristic of modernist empiricism.44 From this perspective, (1) the authentic text has already been lost to the vagaries of time, (2) recuperation will only partially succeed, and (3) what is restored is a better text but not necessarily the text that either the critic or the critic’s audience might want. Only then, after the scientific work is complete, can interpretation commence. In the case of the New Testament, the hypothetical restoration of the author’s own text had an extra benefit: by recuperating the words of the authors, textual criticism had the potential to connect readers not only to the true intentions of those who composed these books but also to the divine author who inspired them.45 Like Swiss mercenaries, the recipients of this improved text may be required to descend from the familiar, fresh mountain air of their primitive villages – that is, from the security of their traditional texts – to dwell instead in the industrialized city where progress reigns. This “modern city,” however, offers compensatory access to authors and indirectly to God himself.46

Sine interpretatione as a Professional Protestant Habitus In the 1830s, Karl Lachmann insisted that critical editing must in every case be performed without recourse to interpretation (recensere sine interpretatione); Transmission of a Gospel Story (Princeton: Princeton University Press, 2018), 29–32 (esp. Table 1.1). 44 I borrow the term “structure of belatedness” from Courtney Weiss Smith, Empiricist Devotions: Science, Religion, and Poetry in Early Eighteenth-Century England (Charlottesville: University of Virginia Press, 2016), 7–10. As Dale Martin points out, this twostep procedure was also fundamental to eighteenth- and nineteenth-century Protestant theology, which argued that historical investigation must proceed theological application (Biblical Truths [New Haven: Yale University Press, 2017], 9–23). 45 On the anachronism of the modern assumption that textual meaning resides in the intention of the author, see, for example, Karen L. King, “‘What is an Author?’ Ancient Author-Function in the Apocryphon of John and the Apocalypse of John,” in Scribal Practices and Social Structures among Jesus Adherents: Essays in Honor of John S. Kloppenborg, ed. W. E. Arnal et al. (Leuven: Peeters, 2016), 15–42 (with bibliography). 46 Courduan’s popular teaching text Handmaid to Theology: An Essay in Philosophical Prolegomena (Grand Rapids: Baker, 1981) (repr. Eugene: Wipf & Stock, 2009) summarizes the procedure and its aims: “The Bible was written by men under the inspiration of the Holy Spirit; yet it reflects the humanity of its authors in wide-ranging ways…Scriptural interpretation…must begin with a careful study of the written text; and theology must begin with exegesis, utilizing the historical-grammatical method of hermeneutics” (23–4).

On Textual Nostalgia

91

theological conviction has no place in editing, even when Christian sacred texts were in view.47 Hoskier’s remarkable acknowledgement of the significant theological stakes for him personally was therefore unusual, a further indication that he was an outsider to the text-critical endeavour as undertaken in professional academic contexts. Indeed, his enthusiasm for vindicating the traditional text, seemingly at any cost, marked him as an amateur.48 Westcott and Hort were more reticent than Lachmann and less forthright than Hoskier: they were well aware that some would be offended by a decision to edit away familiar passages. Thus, they adopted the emerging practice of graphically setting aside certain traditional verses with brackets or appendices but printing them anyway. Strikingly, Lachmann himself made no such concession to church tradition: for him, brackets indicated scientific textual uncertainty, not a hesitation to interrupt the received text for the sake of the faithful; indeed, he presented his edition as a provocation, designed to irritate scholars into a more critical approach.49 Arguing that editorial work must avoid the “Papism” of over-adherence to tradition, in his edition he simply deleted passages and verses that he identified as interpolations into the fourth-century text (in his view, the earliest recoverable textual object).50 Westcott and Hort sought a compromise, one which preserved the distinction between scientific and ecclesial commitments, even if Burgon and Hoskier found their edition to be supremely irritating. Nevertheless, Westcott and Hort also embraced a Lachmannian stance: agreeing that critically edited texts are produced for the church but not from within the church, their reasoned detachment ostensibly enabled them to step outside of Roger Lüdeke, “Materialität und Varianz,” in Regeln der Bedeutung: Zur Theorie der Bedeutung literarischer Texte, ed. F. Jannidis et al. (Berlin: De Gruyter, 2003), 454–59; Helmut Müller-Sievers, “Reading without Interpreting: German Textual Criticism and the Case of Georg Büchner,” Modern Philology 103/4 (2006): 499–500, and Lin, Erotic Life, 50–4. 48 Carolyn Dinshaw, How Soon is Now? Medieval Texts, Amateur Readers, and the Queerness of Time (Durham: Duke University Press, 2012), 23: “In fact, not ‘scientific’ detachment but constant attachment to the object of attention characterizes amateurism.” The closing plea of A. Souter’s review of Concerning the Versions of the N.T. is telling: “We cannot afford to do without [Hoskier’s] valuable co-operation in New Testament textual criticism, but would suggest that he confine his energies to the collection and accurate presentation of material, and leave theorizing to others, at least in the meantime” (“The Genesis of the Versions of the Gospels,” Review of Concerning the Versions of the N.T. Remarks suggested by the study of P and allied questions as regards the Gospels by H. C. Hoskier, JTS 13 [1911]: 122). 49 Karl Lachmann, “Rechenschaft über seine Ausgabe des Neuen Testaments,” Theologische Studien und Kritiken 3 (1830): 845: “Mein schönstes Ziel aber ist erreicht, wenn, was ich gethan habe, ein Unfang wird, der die Nachfolger fördert und zur Vollendung in gleichem Sinne reizt.” 50 Lachmann praises Bentley for his resistance to “Papism” and the text-critical progress he was able to achieve as a result (“Rechenschaft über seine Ausgabe,” 819). 47

92

Jennifer Wright Knust

their roles as Anglican clergymen when applying the rules of science.51 As such, they sought to inhabit a secular rather than ecclesiastical problem-space where human reason, technological advancement, and the collection and analysis of relevant data labours to return a more authentic text to readers, who are then free to interpret as they wish. In the process, textual criticism, so important to the production of Protestant/Roman Catholic Christian difference, was pulled into intra-Protestant controversies, as the new textual science demanded resistance not only to Catholic sacramentalism (Lachmann’s dismissive “Papism”) but also a disavowal of Protestant confessional investments in their own “traditional text.”52 Operating out of the logic of sola scriptura, which implies that textual preferences cannot and should not be imposed by a clerical body or by tradition,53 the newly scientific textual criticism laboured to print a text free from superstitious over-investment in belief, even if that belief was Protestant. As such, the eclectic texts, manuscript studies, collation projects, and elaborations of textual families produced by Hoskier, Westcott, Hort, and their contemporaries operated within a Protestant cultural habitus that they both inhabited and helped to build.54 The performance of critical distance, claimed by Hoskier and more carefully maintained by Westcott and Hort, was therefore not only nostalgic, it was also a way of claiming modern science for the Protestant cause. Yet, as clergymen and churchgoers, the borderline separating secular science from confessional 51 Both Westcott and Hort were Christians trained in theology and ordained in the Anglican communion. 52 As Ron Hendel points out, “The clashing discourses of biblical philology and biblical inerrancy” were born out of a matrix of theological controversy “engendered by the Protestant heresy and intensified by arguments over variants in the MT [Masoretic Text], LXX [Septuagint], and SP [Samaritan Pentateuch].” Textual variants in the Hebrew played a starring role in debates about authority, heresy, and salvation, as editions like the Complutensian Polyglot demonstrate in material form (Steps to a New Edition of the Hebrew Bible [Atlanta: SBL, 2016], 271). On the entanglements of science with the rise of a Protestant “secularism,” see J. Z. Smith, Drudgery Divine: On the Comparison of Early Christianities and the Religions of Late Antiquity (Chicago: University of Chicago Press, 1990). 53 As Daniel J. Treier points out, Protestants “believe that the basic message of Scripture is clear and can therefore be used to enlighten other biblical texts, outside of any institutional mandate” (The Cambridge Companion to Evangelical Theology, ed. T. Larsen and D. J. Treier [Cambridge: Cambridge University Press, 2007], 35). Also see Knust and Wasserman, To Cast the First Stone, 307–308. 54 Pierre Bourdieu, Homo Academicus, trans. P. Collier (Stanford: Stanford University Press, 1988), 21: “Scientific discourse demands a scientific reading, capable of reproducing the operations of which it is itself the product.” As in other cultural realms, scholarly work depends upon a shared habitus that requires those working within it to operate within the limits of “everything we can only know as insiders, and everything we cannot or do not wish to know as long as we do remain insiders” (1). Also see Bourdieu, The Field of Cultural Production: Essays on Art and Literature, ed. and trans. R. Johnson (New York: Columbia University Press, 1993), 106–11, 122, 161–75.

On Textual Nostalgia

93

commitment was dangerously porous, making performances of critical distance all the more important. The differing receptions of Hoskier and Burgon both by their contemporaries and by later critics illustrate this point. At his death in 1938, J. Rendel Harris defended Hoskier as much more than a “disciple or a copyist of Dean Burgon”; no, Harris claimed, “in the sphere of textual study Hoskier was a freelance and a great controversialist,” but his “genial spirit outlived controversies and misunderstandings and left him at the end in the foremost ranks of Biblical science.” 55 Daniel Wallace made a similar argument in 1995, in the context of a refutation of the “Majority Text” position: Hoskier, he argued, cannot easily be cited in detail by those who defend this text because, unlike Burgon, he was a “bona fide textual critic.”56 Importantly, Harris did not mention Hoskier’s interest in the scientific consultation of spirits though, as Jan Krans has shown, he certainly knew all about it.57 Instead, Harris worked to characterise him as a “scientist,” which necessarily meant underplaying the theological and spiritualist convictions he was quite eager to claim during his own lifetime.58 By contrast, Burgon is remembered for placing his devotion to plenary inspiration of scripture ahead of scientific assessment of the evidence at hand, marking him as a throwback to the earlier era that Protestant scientific criticism necessarily rejects. Nineteenth century insiders to this text-critical habitus knew better than to challenge the preeminent place of science in the editorial enterprise. Those who refused to accept this version of textual nostalgia were in danger of being analogized not only to the “papists” but also, in Dipesh Chakrabarty’s terms, to the European “peasant” or the colonial “native” who remains trapped within a precritical, pre-political, pre-modern past: “unreason and superstition” stand for their backwardness, their stubborn refusal of progress, and their unwillingness to embrace the painful discipline of accepting the truths of science.59 Indeed, Wallace’s description of more contemporary advocates of the Majority Text dismisses his colleagues in precisely these terms:

55 J. Rendel Harris, “Obituary for Mr. H. C. Hoskier,” The Times, Monday 19 September 1937, 8. Also see the overview of Hoskier’s writings by Allen in this volume. 56 Wallace “Majority Text,” 281: “Besides Burgon, MT [Majority Text] apologists also appeal to the writings of F. H. A. Scrivener and Herman C. Hoskier for a scholarly defence of their position. The attraction to these two, no doubt, is that they are both somewhat more recent than Burgon and are recognized as bona fide textual critics. What is conspicuous by its absence, however, is any detailed citing of either Scrivener or Hoskier by MT champions. The reason for this is not too hard to find: neither Scrivener nor Hoskier followed in Burgon’s steps.” 57 See Krans in this volume. 58 See, for example, his work What is Nirvana? (1930) which, Allen reports, Hoskier mentioned in a letter to Harris in that same year (Allen this volume). 59 Dipesh Chakrabarty, Provincializing Europe: Postcolonial Thought and Historical Difference (Princeton: Princeton University Press, 2000), 238.

94

Jennifer Wright Knust

(1) They are superstitious: “The use of made of Scrivener and Hoskier by modern day traditional text advocates reveals a disturbing twofold pattern: on the one hand, their perception of results determines allegiance. Questions of method rarely surface. All that matters is that the traditional text is affirmed.” (2) They are poor scholars: “On the other hand, their perception of results is not based on an examination of a given scholar’s writings. Typically, little more is known about his views that that he is theological conservative, makes positive references to the TR, and criticizes Hort’s favoured MSS.” (3) And they refuse progress: “Because of such shibboleths, MT proponents have been repeatedly misled into soliciting unwitting support from the dead voices of the past. Such not only is intellectually dishonest, but also raises questions as to what drives this need for champions.”60

Partisans in favour of the majority text would surely argue otherwise – and Maurice Robinson explicitly challenges Wallace.61 But the accusation that one form of Protestant textual criticism privileges theological investment over scientific rigor while the other willingly embraces it, despite any cost to the church, continues to police a supposed dividing line between professional academic work and an amateurish, “primitive” application of a (pseudo-) science designed to rescue the church’s text rather than the text for the church. Paradoxically, Wallace, Robinson, Burgon, Hoskier, Westcott, and Hort have each sought to confirm the conviction that the New Testament reveals the divine will but by means of science rather than faith.62 Indeed, to borrow Bruno Latour’s terms, by “conjoining inventions of scientific facts and citizens” this form of secularism opened the door for “a wholly individual and wholly spiritual religion” that enabled moderns to be “both secular and pious at the same time.”63 The argument was therefore not about whether the New Testament text should be regarded as the inspired Word of God but rather how this text should be defended and by whom. To be clear, I am not defending the legitimacy of the Majority Text position. I do not think nor would I ever argue that the Byzantine Majority Text can be defended as the “initial text” on any ground that makes sense within the discipline of scientific textual criticism as currently practiced.64 Rather, I am

Wallace, “Majority Text,” 285. Maurice A. Robinson and William G. Pierpont, The New Testament in the Original Greek: Byzantine Textform (Southborough, MA: Hilton Book Publishing, 2005), 540. 62 Cf. https://danielbwallace.com/about/ [accessed 8 June 2018]. 63 Latour, Never Modern, 33; cf. Latour, Pandora’s Hope: Essays on the Reality of Science Studies (Cambridge, MA: Harvard University Press, 1999), esp. 14, 24, 113–14, 134. 64 Though the Byzantine (“Syrian”) text has turned out to be much more important to the current reconstruction of the Ausgangstext than Westcott and Hort anticipated. See, most recently, Holger Strutwolf, Georg Gäbel, Annette Hüffmeier, Gerd Mink, and Klaus Wachtel, Introduction to the Novum Testamentum Graecum: Editio Critica Maior, vol. 3, Acts of the Apostles (Stuttgart: Deutsche Bibelgesellschaft, 2017), Part 1.1, 30*–31*. This decision is based, in part, on the important re-evaluation of the evidence by Klaus Wachtel; 60 61

On Textual Nostalgia

95

suggesting instead that the dispute Hoskier stages between himself, his mentor Burgon, and Westcott, Hort, Tischendorf, and Lachmann participated in and worked to produce a habitus that had already determined, in advance, what the argument could be about. What the argument could not be about had also been decided. For example, no matter how vehemently Hoskier defended the Textus Receptus on the basis of the minuscules he collated, it would not have occurred to him to argue that actual Byzantine texts could be “authentic” on their own merits. Similarly, Westcott’s and Hort’s Alexandrian, Western, Syrian, and Neutral scheme of characteristic readings – drawn, as Yii-Jan Lin has shown, from nineteenth century genealogical science – also disallowed the importance of any particular manuscript or tradition to their reparative project, despite their preference for B: a corrupted text demands precise genealogical analysis, they argued, and no single textual “family” can stand in for the “original text.”65 The argument was therefore about textual witnesses and their appropriate scientific evaluation, not manuscripts and their value as embodied remnants of past or present Christian practices. In Pierre Bourdieu’s terms, a habitus necessarily involves meticulous surveillance of the boundaries of thinkable and unthinkable thoughts, as well as articulate-able and non-articulate-able arguments. In the context of modernity, an a priori over commitment either to the plenary inspiration of scripture (i.e., “superstition”) or to church tradition (i.e., an undesirable, outmoded past, the very past responsible for corrupting the text to begin with) are outside the bounds of acceptable thought, let alone practice. Burgon lingers as a text-critical spectre because of his refusal to play by these rules or perhaps, more provocatively, because he serves as a warning to others of what can sometimes happen when critics allow the stakes of their disciplines rise too closely to the surface. In 1890, Hoskier was caught in the cross hairs of this dilemma: as a fellow scientist, he had an obligation to pursue a reasoned method capable of restoring the original text, but the traditional text he was after, he candidly admitted, was important because it is Sacred and traditional, and not only for its scientific, historical, and cultural value. He therefore sought to combine the operative principles of a scientific textual criticism that disavows dogmatic theological convictions when applying scientific methodologies even as he undertook what was, ultimately, a theological project designed to recuperate a lost original Sacred Text that, he was convinced, coincided with the traditional text and not with B. Operating within the discipline bound him to the rules of this discipline and, in the end, what he named as his quest, if not the data he collected, proved futile.

see, for example, Der byzantinische Text der katholischen Briefe, ANTF 24 (Berlin: De Gruyter, 1995). 65 Lin, Erotic Life of Manuscripts.

96

Jennifer Wright Knust

Textual Nostalgia Revisited In her reflections on the modern nostalgic condition, Svetlana Boym identifies what she views as two forms of nostalgia: restorative and reflective. Restorative nostalgia, she suggests, “attempts a transhistorical reconstruction of the lost home.”66 This form of nostalgia protects the absolute truth, seeks total reconstruction of the monuments of the past, and tends toward nationalism, a longing for purity, and a desire to squelch anxieties about the incongruities of past and present.67 Reflective nostalgia offers a more productive counterpart, expressive of a shared yearning for a different time that understands common estrangement as well as the inherent contradictions of modernity.68 Reflective nostalgia seeks connection with others who, it is remembered, also suffer loss, but restorative nostalgia looks away, demanding a narrow vision that recognizes compatriots exclusively and only insofar as they participate in a similar quest for the same imaginary lost object. Described in these terms, nineteenth century textual nostalgia appears to be more restorative than reflective: engaged in a nostalgic, recuperative drive for a lost object, the “original autograph,”69 textual critics defended the necessity of their work by articulating what they interpreted as a long history of corruption. Scouring surviving manuscripts now littered with dross, these critics looked to eliminate additions and accretions introduced, they concluded, by ignorant actors inhabiting social worlds that valued texts in all the wrong ways and for the wrong reasons.70 From this perspective, manuscripts like Egerton 2610 (i.e., GA 700/Evangelium 604) were important not for their status as (formerly) sacred Svetlana Boym, The Future of Nostalgia (New York: Basic Books), xvii. Boym, Future, 49: “The past for the restorative nostalgic is a value for the present; the past is not a duration but a perfect snapshot. Moreover, the past is not supposed to reveal any signs of decay; it has to be freshly painted on its ‘original image’ and remain eternally young.” 68 Boym, Future, 49–55, 115–16, 354. 69 Cf. Eldon J. Epp, “The Multivalence of the Term ‘Original Text’ in the New Testament,” HTR 92/3 (1999): 245–81; Michael Holmes, “From ‘Original Text’ to ‘Initial Text’: The Traditional Goal of New Testament Textual Criticism in Contemporary Discussion,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 637–88; cf. Peter Shillingsburg, “Text as Matter, Concept and Action,” Studies in Bibliography 44 (1991): 31–82. 70 The ecclesiastical setting of the so-called Syrian text, for example, was perceived to be a liability that needed to be overcome; this text may be, Westcott and Hort explained, “entirely blameless on either literary or religious grounds” but it showed “no marks of either critical or spiritual insight” and was “more fitted for cursory perusal or recitation than for repeated or diligent study (Westcott and Hort, New Testament, 2.135). It is perhaps not too difficult to imagine the readers they had in mind when they identified a desire for a functional liturgical text with corruption and a desire for repeated, diligent study with those more able to achieve critical and spiritual insight. 66 67

On Textual Nostalgia

97

objects, but for their roles as possible witnesses to a now lost immaterial text.71 The debate among textual critics, including Hoskier, was therefore not over manuscripts and their lengthy, distinctive histories but over textual witnesses and their value as data points on a journey toward a hypostatic ideal that only detached scientists could be trusted to recover.72 Within the constraints of Heimweh-Nostalgia and the sense that authenticity has been lost, such a quest succeeded at offering nineteenth-century European and North American Christians a scientifically verifiable New Testament text upon which Protestant faith could more securely rest, at least for a time. This scholarship also provided a durable map of what scientific New Testament textual criticism has been: a secular, scientific discipline rooted in Christian humanism and the Protestant Reformation that seeks to improve both texts and lives for the betterment of society and, indirectly, the Christian faith. Still, the ghosts of Lachmann, Burgon, Hoskier, Westcott, and Hort haunt the discipline in less salutary ways. By demanding resolute performances of detachment from their true object (the recuperation of God’s Word ) and claiming that science is an effective prophylactic against both interpretation and the intrusion of human desire, this criticism has bestowed a more ambivalent legacy to current textual practice, one that continues to measure texts, traditions, and those who edit them on a scale that runs from primitive to civilized, embodied to disembodied, and amateur to professional, at least implicitly. Moreover, by transforming manuscripts into abstracted, immaterial witnesses that point toward a vanished transcendent ideal, textual criticism has also been known to disparage the dignity of those who value different texts differently. Restorative textual nostalgia tends to flatten or ignore provenance (where and how a manuscript travels throughout its life) in favour of a close investigation of provenience (where and how the manuscript was found and/or made) capable of denigrating those who once copied, preserved, and held these manuscripts (i.e., the “evidence”). As long as the fundamental importance of Egerton 6410 (Evangelium 604/GA 700) – as well as ‫ א‬and B – resides primarily in what that object can indicate about a reclaimed lost text behind every distinctive text, extraction and abstraction will prevail, denuding manuscripts of

Hoskier and Burgon actually shared this view (Hoskier, Full Account, xviii). Nineteenth and early twentieth-century debates about the provenance of Codex Bezae offer a good example of this phenomenon (e.g., Frederick H. Scrivener, Bezae Codex Cantabrigiensis: Being an Exact Copy, in Ordinary Type, of the Celebrated Uncial GraecoLatin Manuscript of the Four Gospels and Acts of the Apostles. Edited with a Critical Introduction, Annotations, and Facsimiles [London: Bell and Daldy, 1864]; J. Rendel Harris, Codex Bezae: A Study of the So-Called Western Text of the New Testament [Cambridge: Cambridge University Press, 1891]; F. E. Brightman, “On the Italian Origin of Codex Bezae, II. The Marginal Notes of Lections,” JTS 1 [1900]: 446–54; J. Rendel Harris, The Annotators of the Codex Bezae [with Some Notes on Sortes Sanctorum] [London: C. J. Clay and Sons, 1901]). 71 72

98

Jennifer Wright Knust

their ongoing contexts, histories, and material lives.73 Such an approach, I fear, obscures not only the material histories of manuscripts but also the material settings, desires, and institutional frameworks that have made textual criticism a habitus and problem-space available at all.74 Times have changed, and so has textual criticism. Fulfilling Hoskier’s wishes, though in a direction that moves away from rather than toward the ends he sought, the landscape of relevant texts has significantly grown. Byzantine texts are now taken ever more seriously and, thanks to the Coherence Based Genealogical Method (CBGM), the pursuit of text types and textual families has been put to an end; a much more complex understanding of genealogical relationships between and among texts has now emerged.75 Among those who employ the CBGM, it is no longer possible to evaluate texts on the basis of an implicitly Orientalizing scale that analogizes text types to nations, geographies, or cultures and dismisses them accordingly.76 As Burgon predicted, paratextual material has become ever more important, though perhaps not in the ways he hoped, as corrections, canon tables, section numbers, scholia, marginal commentary, chapter lists, and illuminations take centre stage in the work of interpreting, contextualizing, and understanding the activities of the scribes, scholars, and churchmen who produced and preserved these manuscripts.77 Thankfully, the Editio Critica 73 I would like to thank Liv Ingeborg Lied for helping me understand this more clearly. Cf. Donna Haraway, “Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective,” Feminist Studies 14/3 (1988): 575–99. 74 In Donovan Schaefer’s words, “Texts are produced by bodies that are enmeshed in their partial worlds and trying to negotiate those worlds in their own distinctive ways” (“The Promise of Affect: The Politics of the Event in Ahmed’s The Promise of Happiness and Berlant’s Cruel Optimism,” Theory & Event 16/2 [2013], discussing Lauren Berlant, Cruel Optimism [Durham: Duke University Press, 2011]). Though none of these thinkers address New Testament textual criticism per se, I have been persuaded to this point of view by critical theorists like, for example, Sarah Ahmed (e.g., The Promise of Happiness [Durham: Duke University Press, 2010]; Jaspir K. Puar (e.g., The Right to Maim: Debility, Capacity, Disability [Durham: Duke University Press, 2017]; and Donovan Schaefer, [e.g., Religious Affects: Animality, Evolution, and Power [Durham: Duke University Press, 2015]). 75 See discussion in Wasserman and Gurry, New Approach, 7–11. 76 The comments of Strutwolf et al., are refreshing: “Since the Textus Receptus was overcome by the scholarly textual criticism of the 19th century, there is tenacious negative bias against the Byzantine majority text…the decision against the majority text was often made easily, without seriously considering the quality of the variants in question” (“Introduction,” 30*). Still, all genealogical schemes tend to attract human explanations that easily fall into fantasies of difference, similarity, and belonging capable of supporting any number of human cultural projects, whatever the intentions of the scientists pursuing this work; see Nadia Abu El-Haj, The Genealogical Science: The Search for Jewish Origins and the Politics of Epistemology (Chicago: University of Chicago Press, 2012). 77 Tommy Wasserman and I, for example, looked to manuscript paratexts to re-interpret the reception of the pericope adulterae, thereby challenging the current New Testament consensus that the story was either marginalized or ignored. The “Paratexts of the Bible” project

On Textual Nostalgia

99

Maior (ECM) offers much more than a hypothetical Ausgangstext: an ingenious textual apparatus and the advancements of the digital age have made it possible for historians to work backwards, away from this eclectic edited text and toward the patristic writers and papyri, lectionaries and translations, early editions and conjectural emendations – to name just some of the available evidence – that have contributed to the kaleidoscopic history of this text.78 Importantly, the INTF and the IGNTP are invitational in spirit and practice: eminently collaborative and democratic, these institutes welcome both professional and amateur scholars within an expanding network of interested, invested parties.79 Scientific textual editing has advanced not only as a discipline but also as a humanistic endeavour. Still, Hoskier’s example does invite a momentary pause. As the nineteenth century controversies that energized him illustrate, technological, and methodological advancements do not occur in a vacuum. Neither carefully executed manuscripts nor ingenious New Testament texts will succeed at producing a universally valid, eternal text. Instead, they materialize the desires, hopes, and dreams of those who produce them and, as such, they keep the New Testament alive. They also preserve in material form the horizon of identifiable stakes operative in that edition’s own now. And it is the now that is actually our home: we live here, in this world, not in the past and not yet in the future.80 Neither manuscript collation nor technological advancement will save us, despite the promises of interconnectivity, hypertextuality, and progress without bounds attributed in our own digital age to computer algorithms and cyber-speed. Current fantasies of endless progress through proliferating knowledges democratically sorted by benevolent experts will achieve something approximating human connection and enhanced understanding only if human actors insist that they do. Yet learned performances of detachment from context, history, desire, and embodiment have not lent themselves well to reflective versions of nostalgia. Neglecting the provenance of Egerton 2610 and focusing exclusively on its text, for example, Hoskier missed the opportunity to place this manuscript within other histories of loss and therefore to connect his own experiences with the lives of those he tacitly dismissed. Hoskier mourned the defeat of the Textus Receptus, the passing of his mentor Dean Burgon, the retreat of his own sense of place within a rapidly changing modern world, and also his son, who was killed in (http://paratexbib.eu/index.html) offers an exciting and more comprehensive example of this new interest. 78 See Annette Hüffmeier and Klaus Wachtel in this volume. 79 For example, variants where editors continued to debate their own conclusions employ a “split guiding line” to indicate the alternative to the a reading (the text printed in the “primary line” as the text of the edition (Strutwolf et al., 24*) and the CBGM has been made freely available so that editors can test the method as well as the conclusions and even apply the method for their own purposes. 80 Cf. Boym Future, 350: “One is not nostalgic for the way the past was but for the way it could have been.”

100

Jennifer Wright Knust

World War I,81 but these were far from the first instances of displacement, misfortune, and loss encountered by those responsible for the survival of Egerton 2610. Once a prized possession of a well-to-do Byzantine family, this Tetraevangelion contains a Greek note in a later hand about the birth of a daughter.82 Added in 1229, this inscription indicates that the family of one “Theodora” intended to record their private history on the blank folia of this devotional book, a common practice at the time, though clearly they were forced to abandon this goal for one reason or another.83 The manuscript next appears in Rome among the library of books collected by the once famous Albanis.84 Illustrious contributors to the Counter Reformation and the Italian Renaissance, the Albani family included a pope, several cardinals, and a Senator among their number. They were also renowned antiquarians; the considerable collection of antiquities amassed by Cardinal Alessandro Albani formed the founding deposit of the remarkable Museo Capitolino (opened 1734), the first public museum of its kind.85 Later Albanis were less prosperous and entrepreneurial, however, and successive waves of misfortune troubled both the family and their books: reduced circumstances in the eighteenth century led to the sale of some of their treasures; some of their books were looted during the Napoleonic occupation; and the last Albani heir died in

See Garrick V. Allen in this volume. London, British Library Egerton 2610 f. 295v: εγενηθη (εγεννηθη) η αυθεντοπουλλα (αυθεντοπουλα) ηµον (ηµων) ει (η) κυρα Θεοδορα (Θεοδωρα) η Κοντοφρενος µινι (µηνι) Αυγούστου ηµερα γ΄…(?) γ΄ ετους χψλη΄. (Transcription available at Pinakes: http://pinakes.irht.cnrs.fr/notices/cote/39437/.) 83 As Stratis Papaioannou observes, this practice underscored an “underlying desire to inscribe private history into the imagined permanence of books.” He concludes, “Ultimately, it is here, mostly in the margins of books, that the joyous and traumatic memories of a Byzantine person attempted, perhaps a bit feebly, to become memory and history, to enter the Byzantine historia.” (“Byzantine historia,” in Thinking, Recording, and Writing History in the Ancient World, ed. K. A. Raaflaub [London: Wiley Blackwell, 2014)], 297–313, here 307–8]). 84 The manuscript is marked by Albani family book stamps. 85 On the library, see Cecil H. Clough, “The Albani Library and Pope Clement XI,” Librarium: Zeitschrift der Schweizerischen Bibliophilen Gesellschaft/Revue de la Sociéte Suisse des Bibliophiles 12/1 (1969): 13–21. For an account of the transfer of several thousand Albani books to the Catholic University of American, see B. M. Peebles, “The Bibliotheca Albana Urbinas as represented in the library of the Catholic University of America,” in Didascaliae: Studies in Honor of Anselm M. Albareda, ed. S. Prete (New York: B. M. Rosenthal, 1961), 327–51. On the place of Alessandro Alabani’s collection in the founding of the Museo Capitolino, see Carol Paul, “Benedict XIV’s Enlightened Patronage of the Capitoline Museum,” in Benedict XIV and the Enlightenment: Art, Science, and Spirituality, ed. R. Messbarger, C. M. S. Johns, and P. Gavitt (Toronto: University of Toronto Press, 2016), 341–66. As Anthony Cutler points out, however, this enthusiasm for collecting was not innocent (see his “From Loot to Scholarship: Changing Modes in the Italian Response to Byzantine Artifacts, ca. 1200-1750,” Dumbarton Oaks Papers 49 [1995]: 237–67, here 264). 81 82

On Textual Nostalgia

101

1852, after which the remaining books were sold off.86 By the time this manuscript reached the British Museum, it had been orphaned, traded, and sold many times over. Hence its folia, bindings, stamps, and inscriptions carry traces of a longer, richer story than its text alone can indicate. Perhaps Hoskier’s neglect of this readily available information – did he simply miss seeing the Greek note and the Albani stamps? – can prod current critics to claim our own nostalgia, textual and otherwise, not so that we can turn away from the lives that came before and those who value texts differently, but so that we can notice shared loss and shared life and care.

86 Clough, “The Albani Library,” 16–7. According to Clough, a number of volumes were lost when the collection Theodor Mommsen had purchased for the Prussian Government sunk in a wreck off the Strait of Gibraltar in 1862.

‘A Book Worth Publishing’: The Making of Westcott and Hort’s Greek New Testament (1881) Peter J. Gurry In many respects, the nineteenth century marks a high point in research on the text of the New Testament. In terms of manuscript discoveries, collation work, and new editions of the text, no period can claim so much effort expended by so many with such significant and lasting effects.1 This is the age of Karl Lachmann (1793–1851), S. P. Tregelles (1813–1875), Constantin von Tischendorf (1815– 1874), F. H. A. Scrivener (1813–1891), B. F. Westcott (1825–1901), F. J. A. Hort (1828–1892), and Eberhard Nestle (1851–1913). Each of these scholars produced editions with importance well beyond their own lifetimes. Besides editions, the nineteenth century was a time of increased manuscript discovery and study. Tischendorf published important facsimiles like those of Codex Sinaiticus and Codex Vaticanus, and scholars like Tregelles and Scrivener brought additional manuscripts to wider attention through meticulous collations. Little wonder that Eldon Epp says that “the period from Lachmann to Westcott-Hort, 1831– 1881, undoubtedly constitutes the single most significant fifty-year period in the history of NT textual criticism.”2 In addition to this tremendous activity – in fact, as a result of it – this is also the period in which the long-simmering frustration with the Textus Receptus reaches its tipping point.3 By the end of the century, the first and only officiallysanctioned revision of the Authorized or King James Version was published to a widespread and often strong reaction.4 With the public anxious to see the result 1 A good overview can be found in Dirk Jongkind, “The Text and Lexicography of the New Testament in the Eighteenth and Nineteenth Centuries,” in A History of Biblical Interpretation: Volume 3 The Enlightenment through the Nineteenth Century, ed. D. F. Watson and A. J. Hauser (Grand Rapids: Eerdmans, 2017), 274–99. 2 Eldon J. Epp, “Textual Criticism,” in The New Testament and Its Modern Interpreters, ed. E. J. Epp and G. W. MacRae (Atlanta: Scholars Press, 1989), 80. 3 Perhaps overdoing his military metaphor, Epp likens Lachmann’s edition to D-Day and Westcott and Hort’s to V-Day (“Textual Criticism,” 81). 4 The New Testament was published on 17 May 1881 and the Old Testament on 19 May 1885. Bruce M. Metzger, The Bible in Translation: Ancient and English Versions (Grand Rapids: Baker, 2001), 102.

104

Peter J. Gurry

of a decade of work, the New Testament sold two million copies in Britain in just a few days.5 Beyond its updated English, the Revised Version, as it was known in Britain, was momentous because it was based on a new eclectic Greek text, one that differed in 5,614 instances with the Greek text thought to be behind the King James Version.6 This text was different from, but influenced by, the text of another monumental New Testament edition, one that was published just five days before it. That edition was produced by the Cambridge pair Westcott and Hort and titled The New Testament in the Original Greek, published in two volumes, the first being the text and a short introduction and the second being a much larger introduction and an appendix.7 It is this edition that is widely credited with finally ending the long, uncontested reign of the Textus Receptus among New Testament scholars,8 and eventually, by way of proliferating English translations, of the wider Bible-reading public as well.9 Of their edition, Stephen Neill says “the effect was comparable to that which is attained when an ancient Greek statue that has lain for hundreds of years at the bottom of the sea is raised and cleaned and seen again in its pristine beauty.”10 Geddes MacGregor, The Making of the Bible (Philadelphia: J. B. Lippincott, 1959), 211. This is my own count, based on differences recorded in F. H. A. Scrivener, The Parallel New Testament: Greek and English (Cambridge: Cambridge University Press, 1882). F. C. Cook, however, reports that Scrivener’s own notes have the number as 5,788 (The Revised Version of the First Three Gospels [London: John Murray, 1882], 222). My thanks to Alan Cadwallader for this reference from his forthcoming book The Politics of the Revised Version (New York: Bloomsbury, 2018). 7 Westcott and Hort were, of course, members of the revision committee as well. Their role is discussed especially vividly in A. H. Cadwallader, “The Politics of Translation of The Revised Version: Evidence from the Newly Discovered Notebooks of Brooke Foss Westcott,” JTS 58/2 (2007): 415–39. 8 Robert F. Hull, The Story of the New Testament Text: Movers, Materials, Motives, Methods, and Models (Atlanta: SBL, 2010), 105; C. L. Church, “Westcott, B(rooke) F(oss) (1825–1901), and F(enton) J(ohn) A(nthony) Hort (1828–1892),” in Dictionary of Major Biblical Interpreters (Downers Grove: IVP, 2007), 1042. Metzger and Ehrman title the longest section of their chapter on the “Modern Period” of New Testament textual criticism as “The Overthrow of the Textus Receptus.” Bruce M. Metzger and Bart D. Ehrman, The Text of the New Testament: Its Transmission, Corruption, and Restoration, 4 th ed. (New York: Oxford University Press, 2005), 170–94; cf. the division of Modern Criticism into pre- and post-Westcott and Hort in Kirsopp Lake, The Text of the New Testament, 6 th ed. (London: Rivingtons, 1953). 9 Interestingly, while the King James Version is no longer the best-selling translation, results from the 2012 General Social Survey found that it remains by far the most read translation in the U.S. See Philip Goff, Arthur E. Farnsley II, and Peter J. Thuesen, “The Bible in American Life Today,” in The Bible in American Life, ed. P. Goff, A. E. Farnsley II, and P. J. Thuesen (Oxford: Oxford University Press, 2017), 9–10. Even still, the Revised Version broke the dam, as it were, and ushered in the flood of translations since its time. 10 Stephen Neill, The Interpretation of the New Testament: 1861–1961 (London: Oxford University Press, 1964), 75. 5 6

‘A Book Worth Publishing’

105

Granting Neill’s penchant for exaggeration in that archaeological image, few will argue with the basic point that Westcott and Hort’s edition was monumental in its importance. Much has been written on their method and its results, but what has not as often been discussed is its long gestation in the twenty-eight years over which they laboured together. In this chapter, I want to consider the personal side of their work. It is fairly well-known that the two did much of this work by correspondence, working individually and then consulting each other as they finished their sections. What is less known is that much of that correspondence survives unpublished along with the early installments of their edition in the various libraries of the University of Cambridge. While these have been used by a few scholars,11 no one has ever fulfilled the wish of Westcott’s son that someone should give separate attention to his father’s text-critical work through “a careful digestion of the mass of correspondence on the subject.”12 While a complete study of these unique sources must await further work,13 I hope to present here a first pass and tell “the ‘human story’ behind this vast project [that] has never really been considered before.”14 Since we all work on the shoulders of giants and since Westcott and Hort’s work still looms so largely over the entire discipline, a consideration of their past editorial work makes a valuable point of departure for thinking about the future of textual scholarship. The story can be told in three parts, looking at the impetus and impediments of their work, followed by a section considering some of the implications for our own work today.

1. Impetus Westcott and Hort were both trained at premier English prep schools (Westcott at King Edward VI and Hort at Rugby) and both attended Cambridge where

11 The letters about their edition are held at the Cambridge University Library as Add. MS 6597 which is a selection and transcription made in 1917 by G. A. S. Schnieder, librarian of Gonville and Caius College. The original letters have not survived it seems. The installments referenced below are held at Trinity College Library (a copy of the Gospels is also held at St. John’s College). 12 Arthur Westcott, Life and Letters of Brooke Foss Westcott, 2 vols. (London: Macmillan, 1903), 1.398; hereafter LLW. These letters are also probably referenced with the same hope in Arthur F. Hort, Life and Letters of Fenton John Anthony Hort, 2 vols. (London: Macmillan, 1896), 2.240; hereafter cited as LLH. 13 I hope to publish the complete letters along with a collation of the installments in book form. For this study I have also checked the massive Macmillan Archives at the British Library on which see Philip V. Blake-Hill, “The Macmillan Archive,” The British Museum Quarterly 36 (1972): 74–80. 14 Graham A. Patrick, F. J. A. Hort: Eminent Victorian (Decatur: Almond, 1987), 84.

106

Peter J. Gurry

Westcott, being the older of the two, was one of Hort’s tutors.15 They would form a fast friendship that would last the rest of their lives. During much of their work they were separated with Hort doing parish ministry in the small town of St. Ippolyts about thirty miles from Cambridge and Westcott caring for boys as a schoolmaster at Harrow School in London. By 1872 they would be reunited in Cambridge where they lived just doors down from one another on St. Peter’s Terrace. Between the two of them, they held Cambridge’s three most prestigious chairs of divinity: Westcott being the Regius Professor of Divinity and Hort the Hulsean and later the Lady Margaret’s. Their time together in Cambridge was again not to last, ending in 1890 when Westcott was appointed Bishop of Durham.16 Except for the shared years in Cambridge, they kept up a lively correspondence, writing about all manner of topics, not least of which was their text. 1.1 Early Interest The first inclinations toward the work of editing the New Testament are found in Westcott’s diary entry on 8 February 1846 shortly after his twenty-first birthday. “If I am enabled,” he wrote, “what a glorious employment for one’s leisure hours it would be to prepare a new edition of the New Testament. If it please God, may I be allowed to do this, and enabled to do it in a proper spirit.”17 Hort’s first reference is even more explicit and comes in a letter to the Rev. John Ellerton written in 1851.18 There he writes about his experience working slowly through his Greek text: I had no idea till the last few weeks of the importance of texts, having read so little Greek Testament, and dragged on with the villainous Textus Receptus. Westcott recommended me to get Bagster’s Critical, which has Scholz’s text, and is most convenient in small quarto, with parallel Greek and English, and a wide margin on purpose for notes. This pleased me much; so many little alterations on good MS. authority made things clear not in a vulgar, notional way, but by giving a deeper and fuller meaning. But after all Scholz is very capricious and sparing in introducing good readings; and Tischendorf I find a great acquisition, above all, because he

15 Westcott also tutored J. B. Lightfoot. A curious fact is that the faces of these three, along with those of Dean Stanley, F. D. Maurice, and Dr. Arnold are placed on the faces of the apostles in the clerestory windows of Great St. Mary’s church in the center of Cambridge. 16 A short overview of their lives can be found in Bruce M. Metzger, “The Westcott and Hort Greek New Testament – Yesterday and Today,” The Cambridge Review 103 (1981): 71–2. 17 LLW, 1.42. The entry for April 6 of the same year says, “My Greek Testament comes at last – which I trust may be my companion for many, many years to come. May I not fail to ‘remember,’ and in all things to set it in my greatest treasure, my surest comfort; and so may all my friends.” 18 There is a reference in 1849 to the first volume of Henry Alford’s Greek New Testament about which Hort quips that “it seems good, and not superstitious” (LLH, 1.115). But Arthur writes that the letter of 1851 is “perhaps, the first signs of interest in the text of the Greek Testament” (LLH, 1.174).

‘A Book Worth Publishing’

107

gives the various readings at the bottom of his page, and his Prolegomena are invaluable. Think of that vile Textus Receptus leaning entirely on late MSS.; it is a blessing there are such early ones.19

By the beginning of 1853,20 he and Westcott had begun to plan a new edition of the Greek New Testament, one that would be useful in schools and by clergy. In the Spring, Hort met with Westcott in Birmingham to solidify their plans and on 19 April he wrote again to Ellerton about his plans with Westcott: One result of our talk I may as well tell you. He and I are going to edit a Greek text of the N. T. some two or three years hence, if possible. Lachmann and Tischendorf will supply rich materials, but not nearly enough; and we hope to do a good deal with the Oriental versions. Our object is to supply clergymen generally, schools, etc., with a portable Gk. Test., which shall not be disfigured with Byzantine corruptions. But we may find the work too irk some.21

1.2 A Publisher and a Plan By April of the same year, their plan had been accepted by the budding booksellers-turned-publishers, Alexander and Daniel Macmillan, whose congeniality and work ethic won them the loyalty of many of Cambridge’s best minds at the time.22 On the very same day he wrote to Ellerton, Hort wrote to Westcott to inform him of good news from Macmillan. I promised to let you know Macmillan’s answer about the Greek Testament. It has been slow in coming, but is quite favorable. He fully approves of the plan, but leaves all to our own discretion, saying that they will be delighted to “do their part, that is to say, to take all risk and publish, and push.” Besides describing our plan, I told him of my own schemes of examining and restoring (if possible) the texts of the several versions (beside the Peshito, in which we both 19 LLH, 1.211. The very next paragraph in this letter mentions the beginning with Westcott and several others of “a society for the investigation of ghosts and all supernatural appearances and effects, being all disposed to believe that such things really exist, and ought to be discriminated from hoaxes and mere subjective delusions.” This marks an unlikely connection with Herman Hoskier who is known to have an interest in the paranormal. Unlike Hoskier, however, Hort is never known to have cited such encounters in his text-critical work! See the infamous discussion of a “spirit testimony” in H. C. Hoskier, Concerning the Text of the Apocalypse: Collations of All Existing Available Greek Documents with the Standard Text of Stephen’s Third Edition, Together with the Testimony of Versions, Commentaries and Fathers. A Complete Conspectus of All Authorities, 2 vols. (London: Bernard Quaritch, 1929), 1.xxxviii; and see the essays by Jan Krans, Juan Hernández Jr., and Garrick Allen in this volume. 20 LLH, 1.239–40. 21 LLH, 1.250. 22 See Charles Morgan, The House of Macmillan (1843–1943) (London: Macmillan, 1944), 27–49 (esp. 30). Charles L. Graves writes that Hort’s “Snark-like habits were not compatible with a wide range of acquaintances, but [his] friendship with the brothers, based on their common love of books, was intimate and life-long” (Life and Letters of Alexander Macmillan [London: Macmillan, 1910], 33). Daniel died in 1856 and so most of their correspondence is with Alexander.

108

Peter J. Gurry

hoped to effect something), so soon as I should have learned the languages sufficiently: and likewise my intention to go through such of the earlier Fathers as have been examined but cursorily: and likewise to see whether anything can be gleaned from the Latin or cursive Greek MSS at Cambridge. I told him at the same time that these were schemes which might very probably not be realised; and indeed several times urged that there must be no definite agreement, as we might wish to desist for various reasons; I referred in particular to the change in the Greek Professorship which (if Thompson were elected) might cause a good text to be published by authority, and to swamp our edition. He says that no text is published by the Pitt Press under the Greek Professor’s authority; but that the Pitt Press merely reprints Lloyd, and Scholefield published a text of his own. He adds that, if I carry out my scheme, our book will be worth publishing, whatever any one else may do. 23

Already in these letters we see Hort’s plan to produce an edition that would be useful to students and would leave behind “Byzantine corruptions.” The goal of a text simple enough to use in schools is an important one that would play a recurring role in their later discussions, especially as it touched on the format and structure of the edition. Among other features, it explains why there was no apparatus and why the famed introduction and appendix were kept to a separate volume. At various points, Westcott pleaded with Hort to keep these additional materials brief but he won only a few of these particular battles.24 While giving little ground on the introduction, Hort was as keen as Westcott to preserve the “compactness” of the text volume.25 Also visible in these letters is Hort’s characteristic tendency to take on gargantuan projects. Here he is planning to examine and restore “the texts of several versions,” of which he has yet to even learn the language! In a letter to Westcott later in the same year he would propose an even bigger project which, he says, “I hope will not frighten you by its bigness.”26 The plan was for a complete listing of the citations of the New Testament by the church fathers and an essential feature was to be its basis on “secure correct texts” such that it would not “add materially to the labour if we were to pay attention to the text of the fathers Cambridge University Library, Add. MS 6597, letter 4, 19 April 1853. Writing to Hort on 10 June 1869, Westcott says, “I am anxious that the Introduction and Appendix should be as short as possible, especially the Introduction. If you explain our method it will be enough. I do not think that anything would be gained by any defence of it. In the main the book will speak best for itself” (Add. MS 6597, letter 132; cf. letters 72, 80, 83, 131, 133–35). 25 Referring to the idea of publishing the text volume in parts, Hort wrote to Westcott on 20 January 1865 saying, “Separate publication would surely give a false idea of the book as an elaborate performance for the learned rather than the simple original New Testament for common use. Compactness would of course be lost” (Add. MS 6597, letter 94). On at least two occasions, Hort found it necessary to defend this single publication with Macmillan who was, at the time, in discussions with Philip Schaff in the United States about a diglot edition. See letters of 3 May 1871 and 24 June 1877 in British Library, Macmillan Archive, vol. CCCIX. 26 Add. MS 6597, letter 5, 19 and 25 September 1853. 23 24

‘A Book Worth Publishing’

109

themselves, and not merely their quotations: we should thus be amassing materials for a series of sound editions of the Greek fathers.” To his credit, I suppose, Hort at least realised that the project could take “10, 20, or more years.” In a role he would reprise again over the next twenty-eight years of editing, Westcott had to gently talk Hort off the ledge: What shall I say of your great scheme, for I must get that off my mind before I can speak of anything else? Such a plan would indeed involve a complete edition of the Greek Fathers as a beginning…Are you prepared for this? However deeply I must sympathize with the work, yet I feel sure that it may not be entered on at a short notice, and I hardly know how far my whole life is not already pledged to another work.27

After this, there is no more record of this project and Hort resigns himself to securing the citations that he and Westcott agreed were indispensable to establishing the text.28 This was, of course, not the only time Hort hatched a massive project he could never hope to complete.29 At other points he would plan to join Westcott and Lightfoot in a series of commentaries covering the New Testament, a translation and expansion of Georg Winer’s Greek grammar, a role in a new translation of the works of Plato, and an expansive three-volume history of the church from Abraham up through the birth of Christ.30 Of the latter, E. G. Rupp has not unfairly said that it was “so hopeless that only Hort could have seriously envisaged it.”31 Aside from their Greek New Testament and its introduction, Hort published no major books in his own lifetime,32 a fact that almost all who study him attribute to a crippling perfectionism “which prevented him from publishing unless all the facts were available.”33 Add. MS 6597, letter 6, 12, 13, and 14 October 1853. See Add. MS 6597, letters 7–8. 29 Cf. Patrick, Eminent Victorian, 22. 30 On the planned commentary series, see LLH, 1.372; 1.417–18; 2.230; for the Winer translations, LLH, 1.373; 2.134–35; Add. MS 6597, letter 22, 3 May 1858; and for Plato, LLH, 1.349. The former two never materialized in any form and the commentaries he did publish were posthumous and incomplete. Hort’s inability to finish a project also affected his lectures as it was said that the term would nearly end before he had covered the preliminaries of the subject and that he might only cover ten verses of an epistle according to David L. Edwards, Leaders of the Church of England: 1828–1944 (Oxford: Oxford University Press, 1971), 189; Neill, Interpretation, 91. 31 E. G. Rupp, Hort and the Cambridge Tradition: An Inaugural Lecture (Cambridge: Cambridge University Press, 1970), 9. 32 Unless one counts his Two Dissertations as an exception as does his son (LLH, 2.175– 76). His five other books were published posthumously. These along with his articles are listed in Appendix III of LLH, 2.492–95. 33 Patrick, Eminent Victorian, 10. Cf. Edwards, Leaders, 185. It should not be thought, however, that Hort was myopic in his interests. His skill in botany is well-known and comes through in a letter written to his publisher during one of many trips taken to the Alps for health reasons in which he admits that “except some Greek Testament (and perhaps Plato), 27 28

110

Peter J. Gurry

At one point in 1882, after Hort wrote to his friend and recently elected Archbishop A. C. Benson in dismay about the state of the church as he saw it, Benson wrote back rebuking Hort, asking rhetorically, “but what if those who have insight prophesy in closets – when they ought to be speaking from the housetops?”34 This hesitancy of Hort’s stands in stark contrast to Westcott, who published voluminously throughout his career,35 seeing many of his books into multiple editions in his lifetime. In these distinctions, we get an important indication of the personality traits that each would bring to their editing of the New Testament.

2. Impediments With the benefit of hindsight, the most striking aspect of their earliest letters is the optimism with which they approach their edition. In fact, it would take them twenty-eight years, but here they are in the 1850s speaking in terms of one or two years! Given this, it will be worthwhile to consider why the edition took so much longer than originally anticipated. 2.1 Optimism The first explanation may simply be the confidence that often comes with youth. In 1853, for example, Hort would tell a friend that we “hope we may perhaps have it out in little more than a year.”36 A few years later and the scope began to dawn on Hort, but not so much that it dulled his optimism. Here he is writing to Westcott in January 1857: I have been meaning…to write to you about the present aspect of our New Testament scheme. The further one advances, the more one sees of the immense amount of work that must be done before the text can be anything like finally settled; but it seems to me that for us to put off publication till that time would be to put if off till our old age, perhaps for ever. On the other hand we have now a large and varied mass of material, quite sufficient to enable us to reduce the number of doubtful readings to an amount extremely small in comparison with the corrections which can be made with certainty. The one great desideratum has been the Curetonian Syriac, and now that is being given to us approximately in Tischendorf’s new edition. Is it not therefore advisable for us to begin at once to go carefully through the New Test. with a view to printing as soon as the revision is done? Idle as I have been in the matter for some time past, I I am going to give myself up in the Alps to botany and geology” (LLH, 1.425). Likewise, Chadwick points out that his hesitancy to print was countered by a “fluence in letters and in speech” (Owen Chadwick, The Victorian Church, Part II: 1860–1901 [London: SCM, 1972], 49). 34 Cited in Edwards, Leaders, 184. 35 See the list of thirty-one books in Graham A. Patrick, The Miners’ Bishop: Brooke Foss Westcott, 2nd ed. (Werrington: Epworth, 2004), 257–58. 36 LLH, 1.264.

‘A Book Worth Publishing’

111

feel more and more sure that we ought to do the work, and that Dr Tregelles’ edition does not in the least supersede the necessity for ours; and further that ours ought to be published as soon as possible. We must make up our minds to many imperfections: I trust we may live to remove some of them at a future date. Do pray let me have your ideas on the subject as soon as you conveniently can. 37

Three days later, Westcott responded in kind: My work has been desultory, but I have a fairly clear idea of the general principles by which I believe that we should proceed. Perhaps I am hardly so sanguine as you are as to the possibility of reducing almost infinitely the doubtful readings. But we may compare some few points and see whether we agree in conclusions, starting as we do from somewhat different points. I am particularly anxious to replace the Oxford Mill by a creditable school text as soon as possible.38

2.2 Attention to Detail In fact, their optimism was not entirely unwarranted at this point. Already by summer of the same year, they had specimens in hand from C. J. Clay, the printer for the Cambridge University Press. Naturally, Hort was opinionated about all level of minutiae, not least the printing. After commenting on various typographic options in a letter to Westcott in May 1857, he wrote, “let me know exactly what you think about this matter and the whole appearance of the page.”39 This matter of the printing would be a recurring concern for Hort. In September 1864, he was still dissatisfied, writing to Macmillan about the latest specimen that “I cannot admire it” and similarly to Westcott the same week that, “as regards printing, are we to acquiesce in the old specimens, or seek for new? I can put up with the old but am hardly satisfied.”40 By the end of the month, he would acquiesce, telling Westcott, “I have just found the old specimens of type for the text (you have some, have you not?), and am content to keep them, if you are. I doubt whether we should easily improve as to beauty, clearness, or compression.”41 Add. MS 6597, letter 13, 13 January 1857. Add. MS 6597, letter 14, 16 January 1857. 39 Add. MS 6597, letter 16, 20 May 1857. From the mention of Matt 9:10 the specimen was at least of that chapter, but judging by the discussion of Matthew being finished soon in May 1865 (LLH, 1.35), it seems that all of Matthew had been printed by 1857. 40 Letter to Alexander Macmillan, 2 September 1864; British Library, Macmillan Archive, vol. CCCIX; letter to Westcott is Add. MS 6597, letter 73, 3/6 September 1864. 41 Add. MS 6597, letter 74, 23 September 1864. The type settled on was a form of the ubiquitous Porson for most of the text and Decker Uncial for Old Testament references. The latter is especially peculiar and seems to have been used for the first time in the early installments of the edition (J. H. Bowman, Greek Printing Types in Britain in the Nineteenth Century: A Catalogue [Oxford: Oxford Bibliographical Society, 1992], 1–2, 63). Later, in 1895, Macmillan would re-typeset the whole edition in their recently-commissioned Macmillan Greek. The bold and quirky new type flopped and was soon abandoned (see J. H. Bowman, “Macmillan Greek,” Journal of the Printing Historical Society 19/20 [1990]: 103– 24, esp. 111, 115). But its use for only the second time in Westcott and Hort’s edition (the 37 38

112

Peter J. Gurry

It will hardly surprise, then, that the two men would discuss other matters of formatting down to the level of punctuation, capitalization, and even the table of contents and how it would affect the running heads. As a testament to their attention to detail on these matters, we should mention a letter from Westcott to Hort written ten years after publication, about a single comma!42 At one point, Hort even found it necessary to send Macmillan hand-drawn sketches of his vision for the outside cover.43 The letters bear full witness to Arthur Hort’s testimony that “very great pains were taken over all details of orthography, typography, and punctuation. Considerable alterations were more than once made in the stereotype plates, and there was doubtless rejoicing at the University Press when the ideal of such fastidious workmen was at length realised.”44 Hort’s interest in matters of orthography would, perhaps most of all, test the resolve of his co-editor as they revised each other’s work. At one point, still in the early stages, Westcott wrote to Hort that generally indeed I feel very, very great repugnance to the whole work of revision. I do not see my way to a positive result nearly so clearly as I once did. Perhaps I think that the result of labour is wholly unequal to the cost, and indeed too often worthless…I cannot express to you the positive dislike – I want a stronger term – with which I look on all details of spelling and breathing and form. How you will despise me, but I make the frank confession nevertheless, and am quite prepared to abide by it.45

This time it was Hort’s turn to talk Westcott off the cliff: I quite enter into your feeling about revision. It comes over me now and then…But believing it to be absolutely impossible to draw a line between important and unimportant readings, I should hesitate to say that the entire labour is disproportionate to the worth of fixing the entire text to the utmost extent now practicable. It would, I think, be utterly unpardonable for us to give up our task, and, if so, every reason conspires to urge us to finish it as quickly as possible, if only to get the burden off our necks. Every right-minded person, I suppose, has a relative contempt for orthographic details. Their dignity comes from their being essential to complete

first being Plato) is a further testament to the importance the publisher ascribed to their edition even fourteen years after initial publication. A digital scan of this rare edition can be seen at the Hathi Trust Digital Library at https://hdl.handle.net/2027/hvd.32044024148884 [accessed 23 March 2018]. 42 LLW, 1.153. The offending comma was at the end of Luke 24:45 and it was changed to a period. 43 Letter to Alexander Macmillan, 2 December 1870; British Library, Macmillan Archive, vol. CCCIX. 44 LLH, 2.243. In light of this detail, Robert Hull may be right that the non-capitalization of “neutral” text in the introduction is “a matter of principle, not simply of typography,” in that they viewed their neutral text as distinct from their other texts in that it is not a recension in any strict sense. See Hull, Story, 103–104 n.30. 45 Add. MS 6597, letter 58, 7 May 1862. Similar expressions about the work on the Revised Version are found in LLW, 1.397.

‘A Book Worth Publishing’

113

treatment. And I confess, when once at work upon them, I find a certain tepid interest as in any research depending on evidence and involving laws.46

However tepid Hort’s interest in orthography was at times, his letters reveal a man far more enthusiastic about this than Westcott.47 It was just this kind of attention to detail, however, that won Hort his highest praise among his contemporaries. To this day one is hard-pressed to find any edition of the Greek New Testament that shows as much concern for orthography as theirs, a concern on full display in thirty-two pages of “Notes on Orthography” at the end of the Appendix.48 The price paid for all this attention to detail, of course, was further delay. 2.3 Revision Aside from matters of formatting and orthography, the bulk of the work seems to have been given to revising each other’s textual decisions. The two scholars made a habit of working independently and then resolving their differences. Occasionally, they would meet in person to do this, but, except for their years together in Cambridge, most of this work was done by letter. It is in these letters that we find a rich pool of examples about how they went about their work. The work ebbed and flowed over the years, constrained not only by their other obligations at Harrow School and as a vicar, but also because of Hort’s ill-health, an unfortunate price he paid for many late nights spent working into the dawn as a student. Additionally, the pastoral work, for which Hort was constitutionally illsuited, took its toll and resulted in a breakdown.49 Nevertheless, the work was never set aside for long. In April 1858, Hort notes in a letter to Westcott that they have made it through 20 and a half chapters of Matthew even as revision remains “a slow process.”50 To give a sense of just how slow the rate could be, they finished Matthew and were only three chapters into Mark by January of the next year.51 By early 1865, they had finished a draft of the Gospels and had the proofs for them in hand. The 46 LLH, 1.445. At one point, Hort did speak of “a great heap of troublesome and unremunerative work to do for the orthography and other such externals of the text” (LLH, 2.34). 47 In the context of the RV, however, Westcott too could defend the study of “trivialities,” even the spelling of Colossae. See B. F. Westcott, Lessons from Work (London: Macmillan, 1901), 167. 48 One might contrast this with the brief comments on format, type, punctuation, and paragraphing in the NA26, 43*. Today, one finds Hort’s orthographic concern most alive in another Cambridge Greek New Testament. See P. J. Williams, “The Greek New Testament, Produced at Tyndale House,” Early Christianity 8/2 (2017): 277–81; Dirk Jongkind et al., eds., The Greek New Testament Produced at Tyndale House, Cambridge (Wheaton: Crossway, 2017), 508–12. 49 Edwards, Leaders, 187, 189. 50 LLH, 1.398. 51 LLH, 1.402.

114

Peter J. Gurry

procedure from that point on was to use one proofreader for each editor. Hort sent his to his friend A. A. Vansittart who himself was toiling away at an edition of Hebrews and shared his love of detail, not least in matters of spelling.52 Westcott seems to have found his proofreader in Rev. Hilton Bothamley although none of their correspondence is included in the Cambridge letter collection.53 All of this meant, according to Hort, that “everything will pass under four pairs of eyes; but this plan involves loss of time as a matter of course.”54 By 1870, an installment of all four Gospels was printed and circulated privately by the publishers,55 brought to press, Hort tells us, because of the work of the revision committee.56 This printing would include a “temporary introduction” written by Hort which was markedly different from the final introduction included with volume one.57 A second printing would be made in July 1871 followed by installments of Acts in February 1873, the Catholic Letters in December of the same year, the Pauline epistles in February 1875, and Revelation in December of 1876.58 These tended to track with the schedule of the revision committee as attested by letters from Hort to Macmillan requesting their printing in time for the committee’s next meetings.59 Along with the letters, these installments give us a unique window on the editorial decisions made in this period. The existence of the 1870 edition of the Gospels is of special interest. It goes unmentioned in their final edition and is not noted in their respective Life and Letters and I have not seen any later scholars

52 Add. MS 6597, letters 106–109, April–May 1865. The edition of Hebrews was never published to my knowledge, but Vansittart’s sensitivities may be sampled in his brief note “On Two Triple Readings in the New Testament,” Journal of Philology 3/6 (1871): 357–59. 53 Though he is mentioned by Hort in letter 78, 8 October 1864. 54 LLH, 2.35. Later W. F. Moulton would also provide corrections to the 1870 printing according to his William F. Moulton: A Memoir (London: Isbister, 1899), 176. 55 See letter of Hort to Macmillan, 26 December 1870 in the Macmillan Archives and cf. LLH, 2.148. 56 LLH, 2.137. For a careful look at supplementary role this played alongside other editions used by the committee, see Cadwallader, Politics of the Revised Version, chap. 4. 57 Hort to Macmillan on 28 June 1870: “I have just sent you…of the temporary introduction, which may go to press as soon as you like. I send also my clean copy of the gospels, which I shall be glad to have back when you have done with it. I forgot to say that the first leaf of the following sheet containing the story of the Woman taken in Adultery will have to go with the last of the gospels” (British Library, Macmillan Archive, vol. CCCIX). For the role of this introduction in the revision committee’s work, see Cadwallader, “Politics of Translation,” 430. 58 As noted in B. F. Westcott and F. J. A. Hort, The New Testament in the Original Greek: Introduction, Appendix, 2nd ed. (London: Macmillan, 1896), 18. Cf. LLH, 2.148; LLW, 1.430. 59 See letters 17 January and 21 November 1873 in British Library, Macmillan Archive, vol. CCCIX. In the former letter, forty-five copies of Acts are requested.

‘A Book Worth Publishing’

115

discuss it.60 It is, however, described briefly by Eduard Reuss in 187261 and comes up in the authors’ unpublished correspondence.62 The only copy I am aware of is held in reserve at Trinity College, Cambridge.63 Its most obvious difference with the 1871 installment and the final 1881 edition is to be found at the end of John’s Gospel where the pagination has been affected.64 The issues are three: how to relate chapter 21 to the preceding chapters, how to relate 21:24– 25 to each other and the formatting of the passage on the woman caught in adultery (7:53–8:11). Starting with the last first, the Pericope Adulterae is not to be found at all in the 1870 installment. This, however, may be a printing error as it is mentioned in the introduction as being found “alone at the end of the Gospels with double brackets”65 just as it would be in the 1871 installment and still in the final 1881 edition. Likewise, Reuss mentions this placement explicitly in his summary of the edition in 1872. Perhaps it was a misprint due to a failure in communication. However, it is clear from the letters that the final placement was their plan from very early on. In a letter to Vansittart in May of 1865, Hort explained their reasoning: I firmly adhere to the Pericope so treated, though conscious that it may cause scandal. Let me repeat more clearly than before. This is one of many passages which belong in a sense to the New Testament, and which we feel we cannot expel from it, and yet which do not belong to the originals of its component books. The other such passages or clauses we leave (in at least one case, Mt 27.49b we insert) in their proper places for two reasons: those passages could not stand independently from their very nature, and the contexts are little or not at all injured by the interpolation, which of course is plainly marked. Here both conditions are reversed: the Pericope can very well stand by itself, and St John’s narrative is miserably interrupted by its insertion. To put it in the appendix would be to expel it from the New Testament: we can therefore only place it as an omitted chapter of the ΕΥΑΓΓΕΛΙΟΝ . It will, I trust, like the other passages stand within ⟦ ⟧.66

The exception being A. H. Cadwallader. Eduard Reuss, Bibliotheca Novi Testamenti Graeci cuius editiones ab initio typographiae ad nostram aetatem impressas quotquot reperiri potuerunt (Braunschweig: Schwetschke, 1872), 246–47. 62 Notably in Add. MS 6597, letters 134–36, July and August 1869. 63 The introduction is signed “Christmas 1870” and there is a sticker from the “University Press, Cambridge” hand-dated 17 November 1870 and initialed as what appears to be “W.C.” The library class number for this particular volume is 272.c.87.27. 64 Like the ending of John, Westcott and Hort discussed the ending of Romans (Add. MS 6597, letters 139–40, February 1870), but the formatting was not changed there between the 1875 installment and the final edition in 1881. Cf. F. J. A. Hort, “On the End of the Epistle to the Romans,” Journal of Philology 3/5 (1871): 51–80. 65 F. J. A. Hort and B. F. Westcott, The Gospels (Cambridge: Cambridge University Press, 1870), xxvii. 66 Add. MS 6597, letter 107, 4 May 1865. 60 61

116

Peter J. Gurry

In this case, there is continuity from 1865 all the way to the final edition in 1881. In the two other examples from John, however, the editors would change their opinion. In the 1870 installment, John 20:30–31 is marked off as a separate section with blank lines before and after and by capitalizing the first three words (πολλὰ µὲν οὖν). This is followed by a double horizontal rule to mark the start of chapter 21 with the first two words (µετὰ ταῦτα) again capitalized. John 21:25 is then similarly marked off with another double horizontal rule and capitalized initial words. By the 1871 installment, however, all this formatting changes so that 20:30–31 and 21:25 are both just normal paragraphs (no spaces, no capitalization) and both horizontal rules are removed completely. As it happens, Westcott and Hort discussed this very change in July and August of 1870, providing us with a good illustration of their method that Hort would later describe as “two editors of different habits of mind, working independently and to a great extent on different plans, and then giving and receiving free and full criticism wherever their first conclusion had not agreed together.”67 Westcott writes first: Lately I have been led to recur to my old belief that John 21.24, 25 must go together. The verses form the note, as I think, of the Ephesian elders to the Appendix. Why did we reject that arrangement? v. 24 is not a natural close to the narrative, and v. 25 is hardly complete by itself. Together they form a natural whole: contrast the οἴδαµεν with οἷδεν 19.35. Are you inflexible? or the printers?68

To which Hort replied: On John 21.24f. we had no discussion. You accepted my proposal, the only question having been about ⟦ ⟧, which I barely raised. At the time I thought our arrangement certain; but I confess I have at times hesitated. Certainly I cannot agree in all you now say. 24 seems to me specially and certainly connected with 23; 25 a separate comment on the whole chapter. What you say of οἴδαµεν is true; but then the contrast with οἶµαι is surely stronger still. Briefly 24 does seem as if it might easily be an addition of the elders, but 25 does not seem to go with it; and yet one can hardly think here of an addition to an addition. But may not 25 have been part of the Gospel in its earliest form, following closely on 20.31? The three classes of facts then run well together: (1) those of the other Gospels, (2) those of this Gospel, (3) those not recorded. I do not see why 21.1–23, with its supplementary note by the elders should not have been inserted before the closing verse. I should not scruple to alter the plates in any way we think fit. But this explanation hardly requires an alteration, though we might have preferred a large space before 24, and perhaps ΟΥΤΟΣ .69

Westcott again: I wish we could agree about John 21. Your view does not in substance contradict mine. If St John wrote originally 21.25 continuously with 20.30f, and later wrote 21.1–23, and this addition was inserted either by himself or (it may well be) by the elders before 21.25, and they

Westcott and Hort, Introduction, Appendix, 17–8. Add. MS 6597, letter 134, 6 July 1869. 69 Add. MS 6597, letter 135, 8 and 9 July 1869. 67 68

‘A Book Worth Publishing’

117

further inserted 21.24 as a note of their own, I do not see that any difficulty remains unexplained. How best to signify in printing the nature of 21.24, is not so easy to say.70

In the end, they would settle with the formatting noted above with none of these paragraphs marked out so drastically as in their 1870 printing. Remarkably, Hort remained optimistic, noting in the first paragraph of his “temporary introduction” that he “hoped that in a few months the whole New Testament may be published.”71 Discussions like this one would last another eleven years. 2.4 The Publisher’s Push Naturally, this eleven-year wait from first privately circulated printing to final edition did not please their publisher. As editing dragged on, Macmillan began to grow increasingly anxious for its completion. He had promised Hort from the very beginning that the publisher would “do their part, that is to say, to take all risk and publish, and push.”72 They had taken enough risk and now it was time to push. The Gospels portion of the edition was advertised as forthcoming in prominent journals already in 1871 and again in 1877.73 Given the installments and the promise in their introduction, Macmillan can hardly be faulted for this. But Hort seems not to have been alerted and both advertisements prompted quick and worried letters, insisting that the text not be published in parts. After the first, Hort wrote to Macmillan: Surely you cannot have forgotten that it was agreed, in consequence of American difficulties not to publish any Part I separately, but merely to send out privately a very limited No. of copies of the Gospels, together with the most temporary introduction which had been written when the separate introduction was contemplated?74

When the second advertisement came out in 1877, Hort again heard about it second-hand, this time while in London at a meeting of the Revised Version committee. He wasted no time, writing from Westminster’s Jerusalem Chamber with Add. MS 6597, letter 136, 5 August 1869. Hort and Westcott, The Gospels, v. 72 See the letter cited above in the introductory section. 73 Letters to Macmillan, 3 May 1871 and 24 June 1877 in British Library, Macmillan Archive, vol. CCCIX. The journals mentioned are the Journal of Philology and the Quarterly, respectively. I have yet to locate these advertisements. 74 Letter to Macmillan, 3 May 1871 in British Library, Macmillan Archive, vol. CCCIX. The “American difficulties” is a reference to Macmillan’s discussions with Philip Schaff to publish the edition in America. Hort only learned of the potential agreement after C. J. Clay informed him that he had sent a sheet of John to Schaff. Hort was disturbed because “we have decided to change the arrangement of that page considerably. In so key…and so important a passage we did not wish to place our rejected arrangement in Schaff’s hands” (Hort, letter to Macmillan, 13 December 1869 in British Library, Macmillan Archive, vol. CCCIX). A few days later, Hort heard that the agreement with Schaff had fallen through, telling Macmillan that it is “a matter of relief” to him. 70 71

118

Peter J. Gurry

the same basic objection: “We have never contemplated dividing it (as published) with volumes…I feel more than ever sure that our text will be quite misunderstood if it comes out without the introduction which is now in preparation.”75 Whether or not Macmillan strategically placed these notices to spur Hort to action is impossible to say. But, either way, by the middle of 1878, the issue came to a head and Macmillan did not mince his words. He wrote to Hort, threatening to send the required paper to Clay at the press “with order to print an edition of 1250 copies” to be finished in time for the August start of school.76 “You have had your way for several years now,” he went on, “to our serious loss, and deference to your judgment has been already carried beyond due limits.” He also noted that he had received word that the Pitt Press was soon to put out its own Greek New Testament by J. J. S. Perowne and Macmillan was worried that Westcott and Hort’s might find in its text a copycat of their instalments.77 Writing at the same time to Westcott about the delay, Macmillan was even more blunt: I wish you could persuade our friend Hort to let the text, that you and he have elaborated so thoroughly, come out. Does he think it needful that the last hair in all our beards should be actually blanched before recognising the fact that we are all getting up in years? I can see no sense in which his delay is right. He seeks a perfection that would lead to no existence before it reached no possible fault. He is getting to be a critical Buddhist. We practical English would think of him as trying to catch his shadow.78

Hort seems to have been adequately harried by all this as testified by a response of fourteen-pages, the longest letter of his in the Macmillan Archives. While Macmillan was quite unimpressed by Hort’s protests to hold off a little longer (“I read all your arguments with amazement. They can only be meant to convince yourself, and is that needful?”), it seems to have been enough to ease Macmillan from his perch and he relented yet again. In the end, it would be three more years before final publication and even then, the results, according to D. McKitterick, were “far from perfect even after so many years, and the reprints of both volumes were heavily corrected.”79 75 Letter to Macmillan, 24 June 1877 in British Library, Macmillan Archive, vol. CCCIX. In this letter we also learn the deferential reason for the order of names (Westcott then Hort): “There is another mistake in the advertisement. My name (for the first time, as far as I know) occurs before Westcott’s, though he is both my senior and also Regius Professor.” 76 Graves, Life and Letters, 345. 77 The fear was not without warrant as the text was already being used and cited by various other authors as, for example, in F. H. A. Scrivener, Six Lectures on the Text of the New Testament and the Ancient Manuscripts Which Contain It, Chiefly Addressed to Those Who Do Not Read Greek (London: George Bell and Sons, 1875). For others, see Cadwallader, Politics of the Revised Version, chap. 4. 78 Graves, Life and Letters, 344. 79 David McKitterick, A History of Cambridge University Press, Volume 3: New Worlds for Learning 1873–1972 (Cambridge: Cambridge University Press, 2004), 66. The introduction to the smaller manual edition published in 1885 notes that it follows “the second and

‘A Book Worth Publishing’

119

2.5 Publication What finally seems to have pushed Hort to press was the impending publication of the Revised Version. Hort was intent on having their text out before it. Exactly why is not clear from the letters, but we may surmise that he wanted the publication dates to reflect the true direction of influence. If so, his hope was all too fulfilled as many wrongly concluded that the Westcott and Hort text was the underlying text for the RV when, in fact, it had been used by the committee which had to agree to a two-thirds majority on all changes. The confusion, as well as the suggestion that the committee was bound to their text, annoyed Hort. He wrote to Westcott, “I too have of late been inclining to think that the absurd identification of texts should be repudiated, and perhaps by ourselves if no one else will do it, though it would come better from others.”80 This did eventually happen to Hort’s apparent satisfaction81 with the publication of a defense by two members of the revision committee.82 They pointed out, among other things, that the Greek text behind the Revised Version stood with Westcott and Hort’s against the Textus Receptus and against Lachmann, Tischendorf, or Tregelles in only sixty-four places.83 Given over 5,000 changes to the Textus Receptus, their text’s particular influence on the resultant text is obviously minimal however much their personal influence may have been otherwise.84 As it was, the first volume of Westcott and Hort’s edition was published on 12 May 1881 just ten days after they signed the contract85 and five days before corrected impression of the larger edition of the text, issued in December 1881” (p. 541). Letters to Hort contain a few dozen suggestions for corrigenda (Add. MS 6597, letters 187, 199, 202). 80 Add. MS 6597, letter 184, 4 February 1882. 81 Add. MS 6597, letter 186, 24 May 1882. Of the actual influence of their text, Patrick says it was “relatively modest: the Company accepted their text before others in only 64 places” (Patrick, Miners’ Bishop, 28). 82 C. J. Ellicott and E. Palmer, The Revisers and the Greek Text of the New Testament by Two Members of the New Testament Committee (London: Macmillan, 1882) which was a response to three anonymous articles later published as John W. Burgon, The Revision Revised: Three Articles Reprinted from the Quarterly Review (London: John Murray, 1883). E. Palmer himself edited the Oxford edition of the Greek text behind the RV in Η Καινη Διαθηκη: The Greek Text with the Readings Adopted by the Revisers of the Authorised Version (Oxford: Clarendon, 1881). 83 Ellicott and Palmer, Revisers and the Greek Text, 41. For perspective, Cadwallader reports that the Westcott-Hort text had sixty-six unique readings just in Matthew and none were adopted by the RV (Politics of the Revised Version, chap. 4). 84 There has been much interest in Westcott and Hort’s possible influence in swaying the committee against F. H. A. Scrivener who tended to defend the traditional readings of the Textus Receptus. See Cadwallader, “Politics of Translation,” 424–26 and Cadwallader, Politics of the Revised Version, chap. 4. 85 The contract is kept in the Macmillan Archives of the British Library and stipulates that after publisher’s expenses, “the profits remaining…are to be divided into three parts,

120

Peter J. Gurry

the Revised Version appeared. 86 The gap with the RV was enough to satisfy Hort who found it “a great relief that it appeared before the Revised Version,”87 a view that Westcott apparently shared.88 There would be no peace for Hort, however, until the introduction and appendix volume was out. Westcott had finally surrendered it completely to him by the end of 1880: I find it quite vain to attempt to fit my work to yours. Our conceptions of the Appendix are wholly different. I think that it will be best – indeed the only thing – to give this, as the Introduction, wholly to you. You must follow your judgment…You will, I am sure, understand me. Perhaps you will think we wrong; but I am quite clear that I could be of no use in working on this scale.89

In June of 1881 it was nearing completion, and now there arose the issue of how to credit the work. Westcott wrote to Hort that “I have been thinking very often how we could make it clear that the Introduction volume is your work. I have done what I can privately, but I feel very strongly that the fact should be made known publicly.”90 There follows some discussion of how to word the acknowledgment of shared labour while crediting the actual writing to Hort. Even after it appeared on 4 September 1881, a new problem arose with regard to the profits from it, which were not initially evenly distributed. Immediately upon discovery, Hort wrote to Westcott: “it had never occurred to me that any question could arise as to a difference between the two volumes of our text in the matter of division of profits, and now that the question is raised, I fail to recognize the justice of the suggested difference.”91 The reason being that Volume II is a joint work, as simple matter of fact. That you entrusted me with the final redaction is an altogether subsidiary circumstance: what belongs to a part of the last two or three years must vanish in the sum of the twenty-eight. You last of all men should wish to make equality the true criterion of justice…it would be too grievous if there were a breach of solidarity now.92

Even Mrs. Hort became involved, lending her assurances to Westcott that the matter should be resolved with an even split. To this Westcott acquiesced, if only to the solution and not the principle, saying, “I am glad still that I gave Macmillan one part each to be paid to the said Revs. F. J. A. Hort and Revd. Canon Westcott and the other to belong to Messrs. Macmillan and Co.” 86 For the dates, see LLH, 2.234–35. The publication date of the first volume and of the Revised Version are conflated by Philip Schaff in his introduction to the American edition and reversed by MacGregor, Bible in the Making, 210. 87 LLH, 2.282–83. 88 Hort writing to Westcott: “I feel most strongly with you the importance of preceding the Revised Version” (Add. MS 6597, letter 163, 12 February 1880). 89 Add. MS 6597, letter 168, 25 November 1880. 90 Add. MS 6597, letter 170, 23 June 1881. Cf. Patrick, Miners’ Bishop, 22. 91 Add. MS 6597, letter 191, 12 October 1882. 92 Add. MS 6597, letter 191.

‘A Book Worth Publishing’

121

direction which he could have silently followed, but of course I will do what you like, and Mrs. Hort decides.”93 2.6 Reception The response to Westcott and Hort’s edition was immediate and significant. Although the Revised Version vastly overshadowed it,94 the attention naturally spilled over to the edition of two key members of the translation committee. There was special interest in the Introduction as it set the principles which were only briefly described in the first volume. Within four months, Westcott reported to Hort that Macmillan only had 300 copies left of volume two and was at the ready with a second printing.95 Although I have found no record of how large the first print run was, we may take some measure based on Macmillan’s threat (cited earlier) to print 1,250 copies of the text volume. In that case, we may surmise that the second volume had sold nearly 1,000 copies in just four months. But even with both volumes published, work remained. Within a month of the second volume’s appearance, Macmillan had approved of a “school reprint,” wishing “no time to be lost about it.”96 Hort did not waste any time, spending “many hours, magnifying glass in hand, in search of broken letters and other minute blemishes”97 to produce the manual edition which would fulfill the original design of the edition. For this, the marginal readings would be reduced in number and moved to the foot of the page to accommodate a smaller page size. This edition was published in 1885. Again, sales figures were not found in my searches,98 but a copy of the student edition from 1906 lists reprints almost every year or two in the first thirty years. On my own shelf is a reprinted edition with dictionary from 1969 that is listed as the forty-seventh printing which amounts to one printing just under every two years. If we roughly estimate 1,000 copies at each printing, this totals over 500 copies per year. Even more than its popularity, the longevity is what impresses. Editions printed as recently as 1974 are easily found in American library catalogues online. A 2007 re-issue by Hendrickson Publishers means that even now,

93 Add. MS 6597, letter 193, 12 October 1882. Both editors were zealous at various points to preserve dual credit. Letters from each are extant in which they insist on “WH” as the proper citation of their work in others’ editions. See LLW, 1.403 and Add. MS 6597, letter 194, 13 and 17 October 1882. 94 McKitterick, History, 3.66. 95 Add. MS 6597, letter 180, 4 January 1882. 96 Add. MS 6597, letter 178, 4 November 1881. 97 LLH, 2.235. 98 Though I remain hopeful that they exist somewhere in the massive Macmillan Archives.

122

Peter J. Gurry

nearly 140 years later, it remains in print.99 In fact, despite the claim to the contrary,100 it is not clear that it has ever been out of print. If there is a more obvious testament to the edition’s importance, I have not found it.

3. Implications Many more specific details that could be drawn from the correspondence and early installments remain to be explored, but the above suffices to give a picture of their work from which to reflect on text-critical lessons that remain for us in the twenty-first century. We highlight these under three headings. 3.1 Manuscript Work The first is that textual criticism is a discipline that requires just that: discipline. Nowhere is this more needed than in the study of manuscripts themselves. Since text-critical work is best done with what Hort called “a knowledge of documents,” hours spent with a codex or piece of papyrus will often repay the effort far more than the equivalent time spent with an apparatus alone. As Hort would write in the introduction to the 1870 and 1871 installments of the Gospels, the evidence supplied by printed editions has been to a certain extent augmented, as well as frequently verified and reexamined. Indeed the needful experience could hardly be otherwise acquired: the exigencies of our task demanded a personal acquaintance with the outward phenomena of MSS, with the continuous texts of individual MSS and versions, and with the varying conditions under which the New Testament is quoted or referred to by the Fathers; for no information at second hand can secure the conveyance of a correct and vivid impression of the true facts by bare lists of authorities.101

This emphasis on avoiding “second hand work” is quite a different picture of Westcott and Hort’s method than the one sometimes gained from modern authors. Bart Ehrman, for example, says that they “collated no MSS themselves, but instead applied themselves to the study of collations and apparatuses made by other scholars (see, for example, Introduction, 144). As a result, their knowledge of the documents was secondhand and partial.”102 The Alands make 99 B. F. Westcott and F. J. A. Hort, The Greek New Testament: With Dictionary (Peabody: Hendrickson, 2007). The edition was reformatted and introduced by Eldon J. Epp. No doubt Hort would have had strong feelings about the reformatting! 100 J. Ed Komoszewski, M. James Sawyer, and Daniel B. Wallace, Reinventing Jesus: How Contemporary Skeptics Miss the Real Jesus and Mislead Popular Culture (Grand Rapids: Kregel, 2006), 284–85. 101 B. F. Westcott and F. J. A. Hort, The Gospels (Cambridge: Cambridge University Press, 1871), vi. 102 Bart D. Ehrman, Studies in the Textual Criticism of the New Testament, NTTSD 42 (Leiden: Brill, 2006), 16 n.24; repr. from NovT 29/1 (1987).

‘A Book Worth Publishing’

123

a similar claim, saying that “the fact should be noted (on which there is general agreement) that neither Westcott nor Hort ever actually collated a single manuscript but worked completely from published material, i.e., critical editions (viz., Tischendorf).”103 Without getting into their dissatisfactions with Tischendorf,104 this is simply not true.105 While they certainly made ample use of published facsimiles, the collations of others, and the editions of Tischendorf and Tregelles, their correspondence reveals their many visits to see manuscripts at the British Museum, in Cambridge, Oxford, and on the continent during holidays – all for the purpose of study and collation. A catalogue of these and other insights stemming from their knowledge of manuscripts would expand beyond the space we have here,106 but suffice it to say that their firsthand knowledge of manuscripts was hardly deficient. To claim that their work “unfortunately…marks a step away from firsthand engagement with the ancient manuscripts”107 can hardly be sustained. Colwell and Tune are much closer to the mark in saying that “Hort’s knowledge of manuscripts of the New Testament was encyclopedic.”108 In this, they provide a model to be followed especially in the wonderous age of digital images; there remains no substitute for firsthand encounters with manuscripts. 3.2 Appropriate Confidence A second observation we may draw is that not all text-critical confidence should be confused with arrogance. The Victorian era is not infrequently (and not always unduly) charged with its share of overconfidence in human ability, and Westcott and Hort were not entirely immune. But there is one form of this vice from which 103 Kurt Aland and Barbara Aland, The Text of the New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism, trans. E. F. Rhodes, 2nd ed. (Grand Rapids: Eerdmans, 1989), 18. The claim is further expanded to say they never even “personally examined a single ancient manuscript” in David L. Dungan, A History of the Synoptic Problem: The Canon, the Text, the Composition, and the Interpretation of the Gospels (New York: Doubleday, 1999), 295; emphasis added. My thanks to Peter Head for this latter reference. 104 See, e.g., F. J. A. Hort, “Reviews,” The Journal of Classical and Sacred Philology 4 (1859): 373–84. 105 The same correction to the record is made in Cadwallader, Politics of the Revised Version, chap. 4. 106 At the least, we should mention Hort’s key role in the sensational discovery of Codex Amiatinus’ origins as described in LLH, 2.254–58 and Christopher de Hamel, Meetings with Remarkable Manuscripts (London: Allen Lane, 2016), 62. Cf. Hort’s correction of F. H. A. Scrivener on the contents of Codex Porphyr (Add. MS 6597, letter 159, 16 February 1877). 107 Stanley E. Porter and Andrew W. Pitts, Fundamentals of New Testament Textual Criticism (Grand Rapids: Eerdmans, 2015), 140. 108 E. C. Colwell and Ernest W. Tune, “Method in Classifying and Evaluating Variant Readings,” in Studies in Methodology in Textual Criticism of the New Testament, NTTS 9 (Leiden: Brill, 1969), 102.

124

Peter J. Gurry

Westcott and Hort need to be vindicated. It is the claim that they were arrogant in their text-critical aims. Eldon Epp, for example, speaks of their “audacious move…beyond the kind of modest proposal of Lachmann – to establish the text of the fourth century – to the unqualified claim to have established the text of the NT ‘in the original Greek.’” For Epp, it would have been much better for them to aim only at the text of the second century and, since they did not, he finds their “final daring thrust” for the original text to be “overkill.”109 He finds the problem right on the first page with what he calls “the arrogant title of their edition.”110 Robert Hull follows a similar line of critique, calling them “audacious enough to entitle their edition The Original Text of the New Testament.”111 Aside from the problem that Hull has the wrong title (it is The New Testament in the Original Greek), this criticism raises the question of what they did mean by their title. If permitted, I would like to suggest that the emphasis of the title is not so much on the New Testament “in the original text” as it is on the New Testament “in its original language.” After all, the New Testament as they knew it was found in multiple ancient languages (Latin, Syriac, etc.), but their edition was the New Testament in its original language of Greek.112 Although I have yet to find a letter in which they discuss what they meant by the title, we may have what we are looking for right on the spine of the book itself where the title is shortened to “The New Testament in Greek.” Now, this could easily be dismissed as the publisher’s doing were it not for Hort’s letter already mentioned. Along with his cover drawing, he specifically asks Macmillan that the spine read “THE NEW TESTAMENT IN GREEK ” with “ CAMBRIDGE ” at the bottom.113 This is, of course, exactly how we find it on all editions and it offers more than a passing hint about where the emphasis lies in the full title. Whatever we make of the title, the introduction certainly does speak of the work in exalted terms as “an attempt to present exactly the original words of the New Testament, so far as can now be determined from surviving documents” Epp, “Textual Criticism,” 81–2. Eldon J. Epp, “Critical Editions and the Development of Text-critical Methods, Part 2: From Lachmann (1831) to the Present,” in The New Cambridge History of the Bible: From 1750 to the Present, ed. J. Riches, vol. 4 (Cambridge: Cambridge University Press, 2015), 27. 111 Hull, Story, 97. 112 In this, I suggest that Westcott and Hort are closer to their predecessors in their use of “original” to refer to language than they are to us who naturally assume it refers to text. On this point, note the difference between the use of “original” in Matthew Henry (1662–1714) versus in Bart Ehrman as pointed out in P. J. Williams, “Ehrman’s Equivocation and the Inerrancy of the Original Text,” in The Enduring Authority of the Christian Scriptures, ed. D. A. Carson (Grand Rapids: Eerdmans, 2016), 394–96. 113 Letter to Macmillan, 2 December 1870 in British Library, Macmillan Archive, vol. CCCIX. 109 110

‘A Book Worth Publishing’

125

and that it is “another attempt to determine the original words of the Apostles and writers of the New Testament.”114 But even here we must not miss the important qualifiers (“so far as” and “attempt to”). More importantly, to call the edition overconfident is to ignore the hundreds of places where alternative text is marked in the margin as having some probability of being original not to mention the sixty-three “suspected readings” where the original was thought to be lost or nearly so.115 What may be most instructive is how their contemporaries viewed their confidence. Giving quite the opposite opinion to that of Epp is Edward Hutton who found it necessary to praise their uncertainty in 1911: Our final text must therefore often be difficult of determination, and here Drs Westcott and Hort have shown their wisdom in giving a much larger number of alternative readings than any other critic, and thus better representing the present state of New Testament criticism. In other words, while the principles of criticism are satisfactory enough, the paucity of authorities makes it unsafe to be too confident in all cases. Hesitation is the truest wisdom, and in the New Testament best represents the present state of the case. Infallibility is the mark of the ignoramus, or of the charlatan.116

In this, we think Westcott and Hort illustrate how it is possible, after no small effort, to remain properly confident without arrogance. We may agree with Hutton that “hesitation is the truest wisdom” and still think that it is also a mark of wisdom to know when not to hesitate. In Westcott and Hort’s case, the legitimacy of their aim should be judged more by the effort with which they pursued it than by the degree to which we think they achieved it. With the vast increase in material witnesses since 1881, we can be all the more confident in our pursuit of the original text even if it remains out of reach in some cases.117 3.3 Collaboration A final lesson is the value of collaboration, one that New Testament textual critics should be commended for pursuing with renewed success in the last several decades.118 We have already noted Hort’s insistence that Westcott receive due credit

Westcott and Hort, Introduction, Appendix, 1, 16. Westcott and Hort, Introduction, Appendix, 291–93. 116 Edward A. Hutton, An Atlas of Textual Criticism Being an Attempt to Show the Mutual Relationship of the Authorities for the Text of the New Testament Up to about 1000 A.D. (Cambridge: Cambridge University Press, 1911), 9. 117 I have discussed how such a quest for the original text can be maintained despite the useful shift toward the goal of the “initial text” in Peter J. Gurry, A Critical Examination of the Coherence-Based Genealogical Method in New Testament Textual Criticism, NTTSD 55 (Leiden: Brill, 2017), 90–101. 118 I think here especially of the collaboration between the IGNTP and INTF projects for an Editio Critica Maior, on which see Klaus Wachtel and David C. Parker, “The Joint IGNTP/INTF Editio Critica Maior of the Gospel of John: Its Goals and Their Significance 114 115

126

Peter J. Gurry

for the Introduction and Appendix which he openly acknowledged penning. But this has not stopped others from trying to slice up the overall credit for the great achievement of the edition. F. F. Bruce, for example, refers to Westcott as “the junior partner in the textual partnership,”119 whereas the Times’s obituary for Westcott says he “must be given the merit of having by his earnest cheerfulness kept up the courage of his shy and nervous colleague.”120 Graham Patrick has been right to push back against this latter characterization, noting the letters cited above in which Westcott despairs of editing and it is Hort who presses Westcott to keep on.121 The real problem in trying to divvy credit is that each scholar brought unique qualities that compensated for the other’s weaknesses.122 Had it not been for Hort, the edition would be far less ambitious and exacting; had it not been for Westcott, it would never be finished! Hort himself recognized their differences, telling Macmillan that “I work more slowly than he does, and examine more authorities, and find it answer [sic]; for I have obtained much very valuable evidence unused before.”123 It was just this attention to detail that would be appreciated so much by another famous editor six years after Hort’s death. Writing to Westcott at the completion of his own important edition, Eberhard Nestle praised the book saying, “I testify once more with the greatest pleasure, I never handled a book made up with so much care and thoughtfulness in the smallest details as your edition.”124 One can almost see a wry smile crossing Westcott’s face as he read this and thought back to his hatred of those “smallest details.” But when combined with Westcott’s ability to progress and move on, Hort’s tendency to look “through the microscope, not only when it was necessary, but even when he needed a telescope” became “of utmost value” to their text-critical work.125 In this, the editors illustrate well the value of complementary skills and temperaments when brought to bear on projects, like editing an editio critica maior of the New Testament, that now seem far too much for any individual to finish. From this we can all learn to pursue similar such fruitful collaborative work.

for New Testament Scholarship” (paper presented at the SNTS meeting, Halle, 2005), http://epapers.bham.ac.uk/754/1/2005_SNTS_WachtelParker.pdf. 119 F. F. Bruce, “Westcott’s Significance,” in The Epistles of St. John: The Greek Text, with Notes and Addenda (Grand Rapids: Eerdmans, 1966), x. 120 LLW, 1.398. 121 Patrick, Eminent Victorian, 82; Patrick, Miners’ Bishop, 21–2. 122 As Patrick rightly says, the unpublished correspondence throws light on the fact that “although this was a joint work, two very different personalities were involved” (Eminent Victorian, 84). 123 LLH, 1.392. 124 Letter to B. F. Westcott from Eberhard Nestle, March 1898 held in the Westcott archives at Westcott House, Cambridge. 125 The descriptions are from Chadwick, Victorian Church, Part II, 49.

‘A Book Worth Publishing’

127

4. Conclusion Praise for the accomplishments of Westcott and Hort is not hard to find (nor is criticism if one enjoys internet commentary). James Turner says they did nothing less than restore “English textual philology to the glory days of the eighteenth century.”126 Even when their work is said to be finally superseded, it is noteworthy how often it is still the benchmark, still what needs to be superseded. NA26, for example, was the first to break with Nestle’s original method for establishing the text and yet it points back, not to one of his editions as we might expect, but to Westcott and Hort whose edition was then nearly a century old.127 At the time, there was criticism of the current state of textual studies for not having advanced beyond Westcott and Hort and Aland was responding to it.128 To that criticism, Gordon Fee memorably closed an important study of his from that time by saying that, “if all of this means that we still appear to be crossing the Atlantic in an 1881 ship, it may be that they built them better in those days. But more likely the point of wonder is not that we still follow [Westcott and Hort], but that they, without our discoveries and advances, revealed such remarkable judgments.”129 Whatever one thinks of the seaworthiness of their boat today, there can be no doubt from either their effort or the longevity of its fruit that their publisher was correct to say that it was, indeed, a book worth publishing.130

126 James Turner, Philology: The Forgotten Origins of the Modern Humanities (Princeton: Princeton University Press, 2014), 358. 127 Kurt Aland et al., eds., Novum Testamentum Graece, 26 th ed. (Stuttgart: Deutsche Bibelgesellschaft, 1979), 43*. Cf. E. C. Colwell, “Hort Redivivus: A Plea and a Program,” in Studies in Methodology in Textual Criticism of the New Testament, NTTS 9 (Leiden: Brill, 1969), 148–71. 128 See Eldon J. Epp, “The Twentieth Century Interlude in New Testament Textual Criticism,” JBL 93/3 (1974): 386–414 and the response from Kurt Aland, “The TwentiethCentury Interlude in New Testament Textual Criticism,” in Text and Interpretation: Studies in the New Testament Presented to Matthew Black, ed. E. Best and R. McL. Wilson (Cambridge: Cambridge University Press, 1979), 1–14. 129 Gordon D. Fee, “P75, P66, and Origen: The Myth of an Early Textual Recension in Alexandria,” in Studies in the Theory and Method of New Testament Textual Criticism, ed. E. J. Epp and G. D. Fee, Studies and Documents 45 (Grand Rapids: Eerdmans, 1993), 273; originally published in New Dimensions in New Testament Study, ed. R. N. Longenecker and M. C. Tenney (Grand Rapids: Zondervan, 1974), 19–45. 130 Gratitude is due to Jan Krans and Mike Holmes for reading the chapter and noting several corrections and to Alan Cadwallader for permitting me early access to the relevant chapter of his forthcoming book.

The Status Quaestionis and Future of Textual Scholarship

The Domains of Textual Criticism and the Future of Textual Scholarship Stanley E. Porter A revolution in textual scholarship began in October of 1966 in Baltimore, Maryland. A major conference was held at Johns Hopkins University, entitled “The Languages of Criticism and the Sciences of Man.”1 The conference, ironically sponsored by the Ford Foundation (a foundation whose fortune was firmly grounded in modernism), was intended to introduce French structuralism to North America, and so featured papers (mostly originally delivered in French) by a number of well-known French structuralists in explication of what was called “les sciences humaines.”2 These papers, retrospectively at least, look very predictable in their advocacy of the structuralist agenda, such as the nearly metaphysical status of structure, unification of signifier and signified, and the centrality of linguistic analysis, especially phonology, as explanatory of most phenomena, including literary ones and many others.3 However, there was one presented paper that, even at the time, stood out from the others and marked the beginning of a transition in thought, so that when the volume of essays was published in 1971, a mere five years after the original conference, the editors recognized that something of significance had happened. What had happened was that structuralism had entered the era of poststructuralism.4 The paper that stood out

1 See Richard Macksey and Eugenio Donato, eds., The Structuralist Controversy: The Languages of Criticism and the Sciences of Man (Baltimore: Johns Hopkins University Press, 1972 [1971]). For a brief summary of structuralism, see Stanley E. Porter and Jason C. Robinson, Hermeneutics: An Introduction to Interpretive Theory (Grand Rapids: Eerdmans, 2011), 155–67. The importance of this movement and its impact upon a wide variety of critical disciplines cannot be overestimated. 2 Peter Caws, “‘Discussion’ of Nicolas Ruwet, ‘Linguistics and Poetics,’” in The Structuralist Controversy: The Languages of Criticism and the Sciences of Man, ed. R. Macksey and E. Donato (Baltimore: Johns Hopkins University Press, 1972), 314. 3 See throughout the volume, but esp. Nicolas Ruwet, “Linguistics and Poetics,” in The Structuralist Controversy: The Languages of Criticism and the Sciences of Man, ed. R. Macksey and E. Donato (Baltimore: Johns Hopkins University Press, 1972), 296–313. 4 See Richard Macksey and Eugenio Donato, “The Space Between – 1971,” in The Structuralist Controversy: The Languages of Criticism and the Sciences of Man, ed. R. Macksey

132

Stanley E. Porter

was by Jacques Derrida, entitled “Structure, Sign, and Play in the Discourse of the Human Sciences,”5 in which he questioned the notion of structure, attempted to sever the relationship of sign and signified, endorsed notions of play and freedom in sign systems, and deconstructed the structural platform of one of its major figures, Claude Lévi-Strauss. One can see in his essay – which had been anticipated in earlier work (some of it known to the conference participants)6 – the major ideas that would be developed further in Derrida’s work and in many areas of poststructuralism. I use this as an example because Derrida’s conference paper, delivered at a conference not unlike the one in which this paper was first delivered, marked not just a moment in structuralist or literary or linguistic thought, but a major shift in the notion of text and with it textual scholarship. Prior to the late 1960s or early 1970s, there had been a relatively consistent view of the concept of text, although admittedly it was usually an assumed and pre-critical understanding.7 A text was a product of an intentional act, in written form (unless being referred to metaphorically), and bounded, finite, and even static in its structure and parameters and the limits of its interpretation. This notion of text was prominent across most disciplines. The philological text was defined by its being a grammatical text to be studied by those interested in matters of language, especially the best examples written by the greatest authors of the past. The literary text was closely related to the philological text, with emphasis originally placed upon the author and then later the text itself as the locus of meaning, finding its most enduring representation in the so-called New Criticism. The linguistic text, like the philological text, was characterised as subject to linguistic analysis in the Saussurian and then structural fashion as an encoded instance of a linguistic sign system. The text-critical text was arguably a type of philological text, and this is how it has traditionally been treated and categorized in New Testament studies. The text-critical text places high emphasis upon an actual, physical text, as textual criticism consists in comparing not just texts, but specific wordings of texts for their similarities and differences.8 Texts are not just bounded entities but they are recognizably finite and even flawed, and as a result suitable for comparison at and E. Donato (Baltimore: Johns Hopkins University Press, 1972), ix–xiii, where they recognize that in the course of five short years much had changed. 5 Jacques Derrida, “Structure, Sign, and Play in the Discourse of the Human Sciences,” in The Structuralist Controversy: The Languages of Criticism and the Sciences of Man, ed. R. Macksey and E. Donato (Baltimore: Johns Hopkins University Press, 1972), 247–65. 6 Richard Macksey, “Concluding Remarks,” in The Structuralist Controversy: The Languages of Criticism and the Sciences of Man, ed. R. Macksey and E. Donato (Baltimore: Johns Hopkins University Press, 1972), 320. 7 For problems with the notion of text, see Jorge J. E. Gracia, A Theory of Textuality: The Logic and Epistemology (Albany: State University of New York Press, 1995), xx. 8 Cf. John Mowitt, Text: The Genealogy of an Antidisciplinary Object (Durham: Duke University Press, 1992), 223–24.

The Domains of Textual Criticism

133

the most minute literal levels for their convergences and divergences. Textual criticism, although it implies a perhaps unachievable text (the Urtext), operates on a horizontal level in its seeking an elusive original text, but the comparison is between one text and another. At most, a given text is a representation, even if flawed, of the original text. This text-critical perspective played a huge role in the notion of textual studies from its rise in the sixteenth century all the way into the twentieth century, especially in the study of ancient texts, including the Bible, where textual criticism is characterised as “lower criticism” that provides the firm and solid foundation for more interpretive “higher criticism.”9 This traditional notion of a text-critical text continues to occupy such a place in textual criticism in New Testament exegesis handbooks, where often the first stage in exegetical procedure is textual criticism.10 The traditional viewpoint summarized the situation up to around the 1970s, when the notion of text underwent a major and radical poststructuralist reconceptualization. This reconceptualization was part of a general intellectual foment that transpired within many if not most intellectual disciplines, especially those that had structuralist underpinnings, in which many of the foundational assumptions and orientations of such disciplines – such as the notion of stable texts, something that was deeply engrained in most of these disciplines mentioned above up to this point – underwent a radical revaluation. The basic transformation that occurred involved at least the following differences: between the traditional notion of text as product and the notion of text as process and the emphasis upon its productivity;11 between text and textuality as the process by which various texts and other phenomena are interwoven into a tapestry of intertextuality, a move from meaning as simplex and univocal to its being (to use the language of Bakhtin) dialogical and heteroglossic;12 and between stability in interpretation and fluidity and even infiniteness or unboundedness in meaning.13 The result is what has been called the semiological text, that is, the text as a socially and culturally embedded and unstable semiological system, with shifting 9 For a brief history of textual criticism, see Stanley E. Porter, How We Got the New Testament: Text, Transmission, Translation (Grand Rapids: Baker, 2013), 12–7. Cf. also Fredson Bowers, Textual and Literary Criticism (Cambridge: Cambridge University Press, 1959); and G. Thomas Tanselle, A Rationale of Textual Criticism (Philadelphia: University of Pennsylvania Press, 1989). 10 Craig L. Blomberg and Jennifer Foutz Markley, A Handbook of New Testament Exegesis (Grand Rapids: Baker, 2010), 1–29. 11 Mowitt, Text, 6; cf. Julian Wolfreys, ed., The Continuum Encyclopedia of Modern Criticism and Theory (New York: Continuum, 2002), 870 (in the glossary). 12 Although Mikhail Bakhtin preceded this reconceptualization, his writings were important in the poststructural revolution, especially as they were reinterpreted by Julia Kristeva (see below). On Bakhtin and dialogism, see Michael Holquist, Dialogism: Bakhtin and his World (London: Routledge, 1990). 13 Mowitt, Text, 15.

134

Stanley E. Porter

relations to society, culture, author, reader, and even language itself.14 The lines were drawn in a number of different ways among various conceptual fields. These include opposition between modernists who believed in stable and fixed texts, even if they were subject to interpretation, and postmodernists who were unconvinced of such stability; structuralists who posited almost ontological categories of fundamental oppositions as a type of metaphysical framework for human existence and poststructuralists who believed in the instability of the sign and there being nothing other than fluid notions of language and meaning; ontologists who believed in reality other than the text and pragmatists or de-ontic thinkers who were more concerned with the notion of the creativity of language than its meaning; those who placed emphasis upon the individual, especially the individual as interpreter, and those who were concerned with rethinking the nature of the subject in light of communities of practice and belief and behavior;15 and those who placed emphasis upon cognitive processing and those who emphasized knowledge as a social and communal construct to be negotiated. The four leading proponents in this major shift in the notion of text were, as mentioned above, Derrida, but also Julia Kristeva, Roland Barthes, and Michel Foucault.16 The contribution of these four major thinkers of the second half of the twentieth century cannot be underestimated. I can only summarize here what others have expatiated upon at far greater (and necessary) length and with much greater depth. Derrida, besides dissolving the bounds between author and text, was enamored with the play of language and the instability of signification;17 Kristeva was concerned with moving from language as a stable code to the text

14 Cf. Mowitt, Text, esp. 6–7, 10, but whose entire volume is dedicated to exploring the “semiological text.” The language of semiology pervades poststructuralism. 15 See Paul Smith, Discerning the Subject (Minneapolis: University of Minnesota Press, 1988). 16 See Mowitt, Text, who devotes a chapter each to Derrida (83–103), Kristeva (104–16), and Barthes (117–38). He does not devote a chapter to Foucault, but does discuss him at some length (e.g. 30–7, but also elsewhere). Mowitt notes the significance of the French journal Tel Quel in this intellectual movement (Kristeva was an editor and Derrida and Barthes had varying levels of involvement). For Tel Quel, see Patrick ffrench [sic] and Roland-François Lack, The Tel Quel Reader (London: Routledge, 1998), with essays by all but Derrida of those mentioned above. See James Williams, Understanding Poststructuralism (Stocksfield: Acumen, 2005), among many volumes dedicated to this diverse movement. 17 For important sources of Derrida’s perspective, see Jacques Derrida, “Semiology and Grammatology: Interview with Julia Kristeva,” in Positions, trans. A. Bass (Chicago: University of Chicago Press, 1981), 15–36, which refers to or alludes to his major notions; Derrida, Writing and Difference, trans. A. Bass (Chicago: University of Chicago Press, 1978), which also contains his “Structure, Sign, and Play in the Discourse of the Human Sciences” (see above); and Derrida, Of Grammatology, trans. G. C. Spivak (Baltimore: Johns Hopkins University Press, 1976).

The Domains of Textual Criticism

135

as process,18 the notion of text becoming a productive force, that is, textuality,19 and with intertextuality;20 Barthes brought the notion of the text as structural artefact under attack and turned to other notions to provide justification, such as pleasure;21 and Foucault was preoccupied with social and cultural anti-conventionalism regarding structures of knowledge, power, and society.22 The poststructuralist movement (or whatever one wishes to call what transpired in the rejection of convention in its various forms in the 1970s or so) led not only to social and cultural transformation (some rightly saw it as a revolution), but to a radical reconceptualization of text. Text became a fluid, malleable, composite, variably interpretable, impossible to circumscribe and limit, and even deconstructive and deconstructing notion. All of the fields that I have mentioned above – and many others as well that had been influenced by structuralism – were themselves forced either to outright reject or to reconceptualize the notion of text. Philological and literary texts moved the center of interest from the text to the reader, by means of reception history, with increasing interest in reader-oriented strategies and destructive tendencies of texts and readers.23 With the recognition For important sources of Kristeva’s perspective, see Julia Kristeva, Σηµειωτική: Recherches pour und sémanalyse (Paris: Seuil, 1969), esp. the essays on “Le texte et sa science” and “La productivité dite texte”; Kristeva, Revolution in Poetic Language, trans. M. Waller (New York: Columbia University Press, 1984). Two useful collections of Kristeva’s works are Toril Moi, ed., The Kristeva Reader (New York: Columbia University Press, 1986) (pp. 75–88 are a translation of the first essay mentioned above); and Kelly Oliver, ed., The Portable Kristeva, rev. ed. (New York: Columbia University Press, 2002). 19 Plays on and extensions of the word “text” into such things as “textuality,” “texture,” “intertextuality,” and the like have served to figuratively extend the notion of what a text is, does, and contains to a variety of fields. 20 The literature on intertextuality is immense and growing, especially the sanitized kind that refers to direct lines of influence (such as the Old Testament in the New Testament) in New Testament studies. For an introduction to the issues, see Graham Allen, Intertextuality, 2nd ed. (London: Routledge, 2011). 21 For important sources of Barthes’s perspective, see Roland Barthes, Essais critiques (Paris: Seuil, 1964), with essays “Ecrivains et écrivants” and “L’activité structuraliste”; Barthes, Image – Music – Text, trans. S. Heath (New York: Hill and Wang, 1977), containing the essays “The Death of the Author” and “The Grain of the Voice”; Barthes, The Pleasure of the Text, trans. R. Miller (New York: Hill and Wang, 1975); and Barthes, The Rustle of Language. trans. R. Howard (New York: Hill and Wang, 1986). 22 For important sources of Foucault’s perspective, see Michel Foucault, Language, Counter-Memory, Practice: Selected Essays and Interviews, ed. D. F. Bouchard, trans. D. F. Bouchard and S. Simon (Ithaca: Cornell University Press, 1977), with “What Is an Author?”; and Paul Rabinow, ed., The Foucault Reader (New York: Pantheon, 1984). 23 See Robert C. Holub, Reception Theory: A Critical Introduction (London: Routledge, 1984); two helpful collections of essays in reader-response, Jane Tompkins, ed., ReaderResponse Criticism: From Formalism to Post-Structuralism (Baltimore: Johns Hopkins University Press, 1980); and Susan R. Suleiman and Inge Crosman, eds., The Reader in the Text: Essays on Audience and Interpretation (Princeton: Princeton University Press, 1980); and, 18

136

Stanley E. Porter

of socially and culturally embedded readers came the notion of cultural studies as the appropriate response to the understanding of texts, and along with it the development of what might appear as fragmented interests in such things as colonial and post-colonial studies, New Historicism, feminist and gender studies, and various other ideological criticisms – all reflecting in various ways the resulting intertextuality of the productive text.24 Linguistic texts soon came to be discussed in relationship to discourse. If text is the instantiation, discourse is the socially and culturally embedded environment in which such a text functions or is used.25 In this paper, I would like to examine three areas where I see the poststructuralist reconceptualization of text affecting textual scholarship in its past, present, and potential future. These three areas are: the goal and purpose of textual criticism, the conceptualization of textual variant, and the place of so-called paraand metatextual data.

The Goal and Purpose of Textual Criticism The influence of the poststructural view of text on textual criticism has come to be known in text-critical circles as the new or material philology.26 In 1990, an

among a number of works on deconstruction, the Yale deconstructionists in Harold Bloom et al., Deconstruction and Criticism (New York: Continuum, 1979). 24 For an overview of such literary criticism, see Raman Selden et al., A Reader’s Guide to Contemporary Literary Theory, 4th ed. (London: Prentice Hall, 1997); Lois Tyson, Critical Theory Today: A User-Friendly Guide, 2nd ed. (London: Routledge, 2006); and Mads Rosendahl Thomsen et al., eds., Literature: An Introduction to Theory and Analysis (London: Bloomsbury, 2017). See also Vincent B. Leitch, Cultural Criticism, Literary Theory, Poststructuralism (New York: Columbia University Press, 1992), who chronicles some of the reactions to these developments. 25 See Sara Mills, Discourse (London: Routledge, 1997), 8–27 for varying definitions of discourse (although Mills emphasizes the role of Foucault in the development of the notion of discourse). Discourse analysis has become a dominant approach to texts, especially in linguistics. 26 An excellent introduction to this subject is provided by Matthew J. Driscoll, “The Words on the Page: Thoughts on Philology, Old and New,” in Creating the Medieval Saga: Versions, Variability and Editorial Interpretations of Old Norse Saga Literature, ed. J. Quinn and E. Lethbridge (Odense: University Press of Southern Denmark, 2010), 87–104, which I draw on here in my description. He sees this “material philology” as under the influence of poststructuralism, “which, among other things, de-emphasised the importance of the author, focusing instead on the inevitably collaborative nature of literary production, dissemination and reception and the cultural, historical and ideological forces at work in these processes” (93). For an excellent history and description of the issues in textual criticism, see D. C. Greetham, Textual Scholarship: An Introduction (New York: Garland, 1994), 295–346, who also sees the new philology as “symptomatic of a general shift in critical

The Domains of Textual Criticism

137

issue of the journal Speculum: A Journal of Medieval Studies, edited by Stephen Nichols then of Johns Hopkins University, brought the movement to the fore.27 Matthew Driscoll summarizes the foundational principles of this “material philology” as these: (1) “Literary works do not exist independently of their material embodiments,” thus rendering important “the relationships between the text and such features as form and layout, illumination, rubrics and other paratextual features, and, not least, the surrounding texts”; (2) “These physical objects come into being through a series of processes in which a (potentially large) number of people are involved; and they come into being at particular times, in particular places and for particular purposes, all of which are socially, economically, and intellectually determined; these factors influence the form the text takes and are thus also part of its meaning”; and (3) “These physical objects continue to exist through time, and are disseminated and consumed in ways which are also socially, economically and intellectually determined, and of which they bear traces.”28 I note the several similarities with how I have described the change in perspective on text brought about by poststructuralism, although I also note that there is a New Historicist element to the definition (perhaps due to the wider influence of what has come to be called cultural criticism).29 New Testament studies has itself undergone a transformation in its view of text in regard to textual criticism, which has come to be called “narrative textual criticism” or perhaps better “sociohistorical textual criticism.” Some of the major proponents of this viewpoint are William Petersen,30 Jacobus Petzer,31 David

theory from a reliance on an author’s imputed meaning to the free play of meaning associated with post-structuralism” (341). 27 Stephen G. Nichols, “Introduction: Philology in a Manuscript Culture,” Speculum 65 (1990): 1–10; cf. also Nichols, “Why Material Philology?” Zeitschrift für deutsche Philologie 116 (1997): 1–21. Driscoll notes other prior influences upon the movement, “Words on the Page,” 90–3. 28 Driscoll, “Words on the Page,” 90–1. 29 On New Historicism, see Paul Hamilton, Historicism, 2nd ed. (London: Routledge, 2003), 130–50. 30 William L. Petersen, “What Text Can New Testament Textual Criticism Ultimately Reach?” in New Testament Textual Criticism, Exegesis and Church History: A Discussion of Methods, ed. B. Aland and J. Delobel (Kampen: Kok Pharos, 1994), 136–52, esp. 136– 40. 31 Jacobus H. Petzer, “The History of the New Testament – Its Reconstruction, Significance and Use in New Testament Textual Criticism,” in New Testament Textual Criticism, Exegesis and Church History: A Discussion of Methods, ed. B. Aland and J. Delobel (Kampen: Kok Pharos, 1994), 11–36, esp. 36.

138

Stanley E. Porter

Parker,32 Eldon Jay Epp,33 Michael Holmes,34 and Gerd Mink as a representative of those who use the “coherence-based genealogical method” (CBGM).35 There are variations in the arguments of the various proponents, but all of them seem to revolve around the problematic nature of the original text and whether emphasizing the search for such a text should be abandoned in favour of examining the contexts in which texts appear and changes are made, that is, the sociohistorical dimension. Petersen first problematized the notion by asking whether the original text means one that we actually have (such as a fourth-century codex) or one that we reconstruct and does not actually exist in any given manuscript. Petzer contends that, no matter whether one uses the traditional (text types) or what he calls Münster (tendencies) approach to textual criticism, one is still a distance from the original text. Parker questions both the strength with which assertions about the original text have been made and the achievability of such a quest, especially when discovery of an original text among many texts makes such a task highly problematic, and especially when no such text ever existed. Parker is instead interested in the “living text,” that is, the tradition, development, or continuing life of the text. Epp similarly raises questions about what is meant by the “original text,” especially when there are earlier or preexistent forms of texts or sources, an autographic text form, a canonical text form, and an interpretive text form. Holmes questions the notion of an autograph when there are possibilities of multiple editions (Mark; Acts) or multiple copies (Romans; Ephesians). Finally, D. C. Parker, The Living Text of the Gospels (Cambridge: Cambridge University Press, 1997), 1–7; cf. Parker, “What Is the Text of the New Testament?” in The Reliability of the New Testament: Bart D. Ehrman and Daniel B. Wallace in Dialogue, ed. R. B. Stewart (Minneapolis: Fortress, 2011), 95–104, esp. 97–8. 33 Eldon Jay Epp, “The Multivalence of the Term ‘Original Text’ in New Testament Textual Criticism,” HTR 92 (1999): 245–81, esp. 276–78; repr. in Epp, Perspectives on New Testament Textual Criticism: Collected Essays, 1962–2004, NTSup 116 (Leiden: Brill, 2005), 551–93, esp. 586–88. 34 Michael W. Holmes, “From ‘Original Text’ to ‘Initial Text’: The Traditional Goal of New Testament Textual Criticism in Contemporary Discussion” and “Reasoned Eclecticism in New Testament Textual Criticism” (orig. 1995), in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis Second Edition, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 637–88 and 771–802; Holmes, “Text and Transmission in the Second Century,” in The Reliability of the New Testament: Bart D. Ehrman and Daniel B. Wallace in Dialogue, ed. R. B. Stewart (Minneapolis: Fortress, 2011), 61–79, esp. 74–8; and Holmes, “What Text Is Being Edited? The Editing of the New Testament,” in Editing the Bible: Assessing the Task Past and Present, ed. J. S. Kloppenborg and J. H. Newman (Atlanta: SBL, 2012), 91–122, esp. 105–12. 35 Gerd Mink, “Contamination, Coherence, and Coincidence in Textual Transmission,” in The Textual History of the Greek New Testament: Changing Views in Contemporary Research, ed. K. Wachtel and M. W. Holmes (Atlanta: SBL, 2011), 141–226 (others in the volume have differing perspectives, even those from the CBGM project). There are other proponents as well. See Holmes, “From ‘Original Text,’” 638–41. I agree with Holmes that Bart D. Ehrman does not belong in this list, despite Epp’s placing him here. 32

The Domains of Textual Criticism

139

Mink, who is one of the major developers of the CBGM, notes that all variants in all manuscripts can be taken into consideration with their coherence relationships. From this a stemma is created and, along with others, a larger set of stemmatic relationships established among manuscripts that lead back to an “initial text.” This initial text represents the “reconstructed form of text from which the manuscript transmission started.”36 There is no doubt that a major shift, if not in substance, at least in emphasis, has occurred in textual criticism in New Testament studies, with the introduction of the sociohistorical or new philological approach. There is no doubt also some merit to the material philological approach insofar as it raises questions about and garners appreciation for such issues as the history of textual development, the importance of sociohistorical context for text-critical change, and the more modest goals of what can be achieved in textual criticism regarding establishing an original text based on authorial intention. However, there are also some questions to be raised regarding the relation between the two. This text-critical approach, even though it has been highly influenced by the poststructuralist agenda (virtually none of the biblical scholars that I have read seems to acknowledge such influence in any substantive way), appears to witness little recognition of this line of descent and connection. Have biblical textual critics genuinely understood the deconstructive agenda, do they understand the implications of the severing of the sign and signified, do they really wish to welcome the death of the author, do they really wish to see texts as open-ended, infinite, and offering no conclusive interpretation but rather the “free play of meaning,”37 do they wish to see texts as processes rather than products, and do they wish to see them as intertextual collages of other texts that have left their traces? There is no way to answer this, as there is little indication that these issues have been raised for New Testament textual critics. Thus, on the one hand, there is a sense in which New Testament textual criticism has not gone far enough in its embrace of the

36 Mink, “Contamination,” 145. The CBGM claims to have eliminated text types. However, they essentially retain the Byzantine text type (indicated by Byz) and also speak of a “Western cluster of variants.” See Tommy Wasserman and Peter J. Gurry, A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method (Atlanta: SBL, 2017), esp. 7–10. The position of the Editio Critica Maior of Acts as a result of the CBGM on these issues is clearly made in Klaus Wachtel, “Notes on the Text of the Acts of the Apostles,” in ECM Acts 1.1, 28*–33*, esp. 30*–2*; Georg Gäbel, “‘Western Text,’ ‘D-Text Cluster,’ ‘Bezan Trajectory,’ Or What Else? – A Preliminary Study,” in ECM Acts 3, 83–136; and Klaus Wachtel, “On the Relationship of the ‘Western Text’ and the Byzantine Tradition of Acts – A Plea Against the Text-Type Concept,” in ECM Acts 3, 137–48. Despite their protests to the contrary, I find this an odd (and unconvincing) argument against text types, when two of the three recognized text types are essentially retained. I believe that this is just one of many major questions that calls the entire project into question, as I have indicated in Porter, How We Got, 32–3. 37 Greetham, Textual Scholarship, 341.

140

Stanley E. Porter

poststructuralist text-critical agenda. On the other hand, there is a sense in which such proposals have too easily gone too far in abandoning the notion of an original text, simply because of its difficulty, when there are other possible explanations that have been proposed within the field of textual criticism when problems with an authorially-intended text are raised. I deal with several issues that emerge from this recent discussion. First, there are a number of examples of problematic authorial texts that are proposed in textual criticism that supposedly argue against the effort to find an authorial text. These include such examples as the two stories of King Lear by Shakespeare and the evolution of Wordsworth’s Prelude, cited by Parker (along with Mozart’s Don Giovani). But there are other well-known texts as well, such as Shakespeare’s Hamlet, Herman Melville’s Typee, Theodore Dreiser’s Sister Carrie, Stephen Crane’s Red Badge of Courage, Pascal’s Pensées, Boccaccio’s Decameron, and perhaps most appropriately for the context of this conference James Joyce’s Ulysses, among many others.38 All of these examples may be problematic for a narrow view of authorial text, usually related to a narrow view of authorial intention, but each of them as a problem has been addressed by textual critics, including by those who argue for some form of authorially intended text. These difficulties do not necessarily argue against an authorial text, only that finding one may be difficult and not as definitive for these instances as for some other texts. Second, the terms of the discussion are arguably more complex than many biblical textual critics have recognized. In other words, the opposition is not simply between an authorial text as a narrowly construed singular entity and the sociohistorical text, as if there is not some recognition of the complexity and even ability to bridge the gap by each side. This accommodation occurs in various forms. For example, some textual critics attempt to find a usable compromise by making a distinction between the work, the text, and the artefact. The work is the idealized or abstracted “original” text that is sought in textual criticism, the text is the wording of the particular work, found in two dimensions: the substance of the text and the accidental features such as punctuation, spelling, etc. The artefact is an example of an entity that contains this text, such as a personal copy of a text.39 According to this scheme, the originalists or intentionalists focus upon the substance of the text and the non-intentionalists on the accidents. This distinction has not proved entirely satisfactory to either side in the debate, with some arguing that the so-called incidentals are as important for establishing the text and its meaning as any other part,40 and others arguing that the wording itself is problematic. Other textual critics have argued that regardless of one’s view of See Greetham, Textual Scholarship for discussion of most of these examples. Driscoll, “Words on the Page,” 93–5. 40 Greetham, Textual Scholarship, 343. The traditional differences in punctuation between the UBS and NA editions present an interesting case in point. 38 39

The Domains of Textual Criticism

141

intentionality in textual criticism, either form implies an original on the basis that one’s reconstruction of the textual history and the priority of variants points to some variants being earlier than others and accounted for on the basis of transmissional change. Once one works back far enough, one must arrive at an original text, or at least an idealized hypothesis of one. For example, those that problematize Shakespeare by arguing that a text of a play cannot be original because the part for a particular player is out of proportion to the others – and hence probably originated with that player who embellished his (and the actors were all male then) role – is creating an argument for an earlier and hence arguably more authorially accurate version without the embellished role. Similarly, if we recognize that certain parts of a biblical book – such as the long ending of Mark – did not appear or were not incorporated into Mark’s Gospel until the early second century, then we are arguing that a version that is without this ending is closer to the authorial text. Third, in light of the poststructuralist re-evaluation, there has been a reconceptualization of what is meant by authorial text in relation to authorial intention. For all the merits of deconstruction, with its instructive realization that texts often do not follow or even violate their own internal logic, criticisms about the selfnegating assertions of deconstruction have rendered it less than useful as a textcritical (or, for that matter, interpretive) strategy.41 But there have been many other attempts to come to terms with the notion of intention that do not end up abandoning an authorial text. Some of these include: the notion of a rolling or internal intention, whereby text-critical evaluation is made not against a textually external or fixed intention but the reconstructed authorial intention as it is inscribed in the text through the writing process; or an intention that is not necessarily the final intention of the author (although it might be, as with Wordsworth’s Prelude of 1850) but may be a first intention (such as the first edition of a work) or an intention somewhere along the way in the authorial development of the text, such that even though there may not be a single authorial text there is one deemed for various reasons to be the authorial text for interpretation or reconstruction on the grounds of good reasons; the concept of a published or received text, by which one endorses the text that was promulgated or promoted by the author or that was received by the wider reading public as the text of a particular work, regardless of other editions that may have been incidentally produced (such as Crane’s Red Badge of Courage);42 and, finally (but not conclusively), the concept of there being more than one text under consideration, each with equal authorial authority, on the basis of their not being editions of each other but, for the sake of textual reconstruction and interpretation, being two 41 See John M. Ellis, Against Deconstruction (Princeton: Princeton University Press, 1989), among numerous others. 42 Cf. James Thorpe, Principles of Textual Criticism (San Marino: Huntington Library, 1972), 50–79, esp. 50.

142

Stanley E. Porter

different texts, even if they have similar or even overlapping content (such as the two plays by Shakespeare of King Lear).43 Fourth, and finally at least for me on this point, non-authorial New Testament textual criticism has not fully embraced the implications of such an approach to an open-ended and non-finite text. The CBGM, as well as a number of other textcritical projects, promises constant revisions of the text of the New Testament as the variants are evaluated and as other manuscripts are discovered and added to the calculation. In other words, such projects promise many different potential texts of the New Testament. Therefore, it seems that in keeping with this approach to textual criticism, the text of the New Testament should no longer be considered a static text to be published in printed and bound form, but should be made available through the variety of digital technologies that are conducive to such a textual approach. This has been suggested,44 but has yet to be fully realised, especially because of matters of copyright (and dare I also say money). The text of the New Testament, as it is constantly revisited and revised, should be openly accessible to all who wish to use it. Further, such a text, in keeping with its sociohistorical status, should be open to intertextual engagement by others, even those who are not formally engaged in these projects. As Driscoll so poignantly and presciently says, “as far as scholarly editions are concerned, the failure of the electronic edition ever really to take off is due to a large extent, I have come to believe, to the inability of textual scholars to see, and embrace, the real potential of digital media, as doing so would inevitably involve relinquishing the more-or-less total control textual scholars have tended to want to maintain over the way in which ‘their’ texts are presented.”45 We have certainly seen this resistance in New Testament textual studies, where those who have produced what they claim are accurate representations of ancient texts also copyright and restrict access to such texts. Driscoll continues: “The majority of the electronic texts produced in the last decade and a half have thus been static and read-only, essentially trying to reproduce the printed text on the screen.”46 New Testament textual criticism has made manuscript photographs available online, but it continues to restrict open access to its texts, and thus brings its practice into direct conflict with its theory regarding the text. They have produced a static text, rather than the fluid text that represents the various sociohistorical phases of textual development. A way forward might be, as Driscoll suggests, “rather than mere electronic versions of printed texts what we ought possibly to be thinking of are interactive text archives, where the user determines to a much greater extent the 43 See Greetham, Textual Scholarship, 341–44, for some of these theories. These theories are not unlike some of the other discussions regarding intention in other disciplines. 44 See D. C. Parker, “Through a Screen Darkly: Digital Texts and the New Testament,” JSNT 25/4 (2003): 395–411. 45 Driscoll, “Words on the Page,” 103–104. 46 Driscoll, “Words on the Page,” 104.

The Domains of Textual Criticism

143

nature and scope of the content and how that content is presented.”47 He issues this plea to the “material philologists”: they “have all argued that textual instability…is so fundamental a feature of chirographically transmitted texts that rather than trying to bring order to this chaos we should celebrate it. Here, finally, we have a means of doing so.”48 If New Testament textual critics wish to pursue such a material philology, the next step must be open access to the Greek text of the New Testament.

The Conceptualization of Textual Variant If textual scholarship is about texts, whether we conceive of these in relation to the old or the new philology, then textual variants are about differences in these texts. The notion of a textual variant is one of the most important concepts in textual scholarship, because textual variants provide one of the major reasons for the existence of textual scholarship. As Greetham says of them, they are the area “most familiar to textual criticism.”49 Whether one believes that the authoriallyintended text can be reconstructed or not, textual scholarship is concerned with those places where some kind of textual change – whether this is a corruption or not, whether intentional or by accident, whether determined or mechanical50 – has been introduced and therefore must be dealt with. These variants are evaluated in various ways, according to the theory of textual editing that one endorses, but before one can apply such a theory one must identify and examine the textual variants. Because the notion of a textual variant is so fundamental to textual criticism, and the term is widely used in work on the subject, one might legitimately expect that the term is robustly defined and hence widely understood within the discipline.51 I find it surprising how the meaning of the term appears to be assumed by those in the discipline to the point that the term is rarely defined. For example, in Greetham’s thorough and relatively recent book, there is no definition of scribal variants, but instead the recognition that a variant has something to do with the “form of the words themselves.”52 This definition is better than most, where any definition whatsoever is often lacking.

Driscoll, “Words on the Page,” 104. Driscoll, “Words on the Page,” 104. 49 Greetham, Textual Scholarship, 279. 50 Greetham, Textual Scholarship, 279. 51 Some of what follows is dependent upon Stanley E. Porter and Andrew W. Pitts, Fundamentals of New Testament Textual Criticism (Grand Rapids: Eerdmans, 2015), 80–6. I thank my co-author for his work in developing this material in the book. 52 Greetham, Textual Scholarship, 279. 47 48

144

Stanley E. Porter

The traditional definition of a textual variant seems to be any variation in wording between two manuscripts regardless of length or type of variation. As a result, a variant might be simply a single letter that indicates a variant spelling up to and including relatively lengthy passages, such as the longer ending of Mark (16:9–20) or the unit that contains the story of the woman caught in adultery (John 7:53–8:11), and anything in between. We can probably agree that we need far greater precision in defining what constitutes a variant than this. And indeed there has been work that has helped to define a textual variant more precisely. However, as Eldon Epp states, “the clarification, definition, and delimitation of the term ‘textual variant’ are vastly more complex and difficult matters than at first would appear.”53 There are a number of reasons for this. “The common or surface assumption is that any textual reading that differs in any way from any other reading in the same unit of text is a ‘textual variant,’ but this simplistic definition will not suffice,”54 for the reasons already noted above. We clearly need a more stringent criterion for definition of a variant. The second reason is that a textual variant must mean a “‘significant’ or ‘meaningful textual variant,’ but immediately this raises the further question of the meaning of ‘significant’ or ‘meaningful.’”55 There has been some progress on both fronts, although not as much as one might imagine. The first part of the question addresses the matter of what constitutes a variant. Epp makes the distinction between a reading and a variant, with a variant constituting a significant reading, to be distinguished from other types of readings that have no value, such as nonsense readings.56 This is probably not a useful distinction, because it requires that we have a definitive and inclusive definition of significance. What may appear at one time to be of no account may prove on later examination to be of account. What may at first appear to be nonsense may under further scrutiny in fact make more sense than it once did. This does not appear to be the way forward in defining variant. A second important distinction is what to call such a variant. Ernest Colwell and Ernest Tune proposed the term “unit of variation” or “variation-unit.”57 They define a “variationunit” as “length of the text wherein our manuscripts present at least two variant

53 Eldon Jay Epp, “Toward the Clarification of the Term ‘Textual Variant,’” in Studies in the Theory and Method of New Testament Textual Criticism, E. J. Epp and G. D. Fee (Grand Rapids: Eerdmans, 1993), 47–61, here 48; repr. in Epp, Perspectives on New Testament Textual Criticism, 101–24. 54 Epp, “Toward the Clarification,” 48. 55 Epp, “Toward the Clarification,” 48. 56 Epp, “Toward the Clarification,” 48. 57 Ernest C. Colwell and Ernest W. Tune, “Method in Classifying and Evaluating Variant Readings,” in Studies in Methodology in Textual Criticism of the New Testament (Leiden: Brill, 1969), 96–105, here 97 (originally published in an earlier form in JBL 83 [1964]: 253– 61).

The Domains of Textual Criticism

145

forms,”58 with the extent including “those elements of expression in the Greek text which regularly exist together.”59 Epp and others have subjected this definition to scrutiny on a number of fronts. Epp finally arrives at the following definition: “that segment of text, constituting a normal and proper grammatical combination, where our manuscripts present at least two ‘variants.’”60 However, Epp recognizes that this definition itself is insufficient, because of the failure to define what constitutes a “normal and proper grammatical combination.” On the one hand, the so-called unit may consist of more than one variant that is unrelated to the others or simply multiple variants, especially in circumstances where such variants are tabulated for analysis. On the other hand, what constitutes such a “proper grammatical combination” is subject to question, depending upon the co-text. As a result, Epp defines such a unit as “the shortest or smallest possible grammatical unit…that is, the shortest grammatically related segment of text that still will encompass all the variants from across the manuscript tradition that present themselves at that point.”61 This definition is not an improvement. The third point that Epp makes is that the term “variation-unit” has “become standard.”62 This may be true in New Testament textual criticism, but it is far from standard in other types of textual criticism. I admit that I have not surveyed every book on textual criticism, but I checked a number of textual criticism books in classical studies and literary studies and did not find a single one that used the term “variation-unit.”63 This does not mean that the terminology or, more importantly, the concept is not important and useful, but it is far from the standard. Even if we concede that the notion of “variation-unit” is helpful, it has not been sufficiently well-defined to be of use, as examples by Epp and Fee abundantly illustrate – or at least their attempts to explain them illustrate. For both, there is no consistent syntagmatic (horizontal) or paradigmatic (vertical) criteria that provide for greater precision in defining a variation-unit or, more importantly, the parameters of such a unit and the relationship that variants within such a unit might have with other units and their larger function within the given Colwell and Tune, “Method,” 97. Colwell and Tune, “Method,” 99. 60 Epp, “Toward the Clarification,” 61 (statement in italics in original). Fee has essentially endorsed the proposal of Colwell and Tune, with some of his own suggestions regarding addition/subtraction, substitution, and word order, none of which actually seems to help the situation. See Gordon D. Fee, “On the Types, Classification, and Presentation of Textual Variation,” in Studies in the Theory and Method of New Testament Textual Criticism, E. J. Epp and G. D. Fee (Grand Rapids: Eerdmans, 1993), 62–79. 61 Epp, “Toward the Clarification,” 61. 62 Epp, “Toward the Clarification,” 49. 63 I surveyed a number of works published after Colwell and Tune’s article (1964) and found none that use it. These include works by Renehan, Thorpe, Tanselle, West, and Greetham. There may be others that use such terminology, but these are fairly “standard” works and their failure to use the terminology belies the assertion. 58 59

146

Stanley E. Porter

text, not simply as a variant but as an element within the larger text and its context. It is at this point that we may introduce some linguistic criteria to help us understand the variation-unit more precisely. I am acutely aware of the critique of structural linguistics leveled by the poststructuralists, not only in one paper in particular in 1966 but on many other occasions. Such criticism objected to the metaphysical pretentions of the structuralists, the emphasis upon phonology as a representation of reality (with the sound system in its binary character acting as a surrogate of ontology), the concentration upon minimalist units of analysis, the close linkage between signifier and signified, the grammatically based definition of meaning with emphasis upon semantics (over pragmatics), and the contextless analysis that often accompanied such linguistic description, possibly among other such criticisms. Many if not most of these criticisms are probably defensible.64 However, whereas this may have characterised much of structuralist linguistics of that time, there have been other forms of linguistics that have developed since that time that have addressed many – though admittedly not all – of the criticisms that have been leveled against structuralist linguistics. Structuralist linguistics has taken many different forms, but many of the models have continued to emphasize minimalist units with emphasis upon a-contextual semantics, and made absolutistic claims regarding, if not reality, at least the human mind, language, and even cognition. Such claims need not be made. Some linguistic models have been developed that are less formalist and more functional in nature. Even though they may be code-based, they encompass units larger than the clause and are highly sensitive to the importance of the environment of usage, whether that is the immediate linguistic environment or a larger contextual and even cultural usage in which language functions. Such a model is found in Systemic Functional Linguistics (SFL). Here is not the place to offer a defense of SFL within the poststructuralist critique, but I believe that SFL presents linguistic criteria that can address some of the shortcomings of textual criticism by proposing a more rigorous definition of the variation-unit. Systemic Functional Linguistics has two features that are important for the definition and interpretation of variation-units: the first is the notion of rank structure, and the other is the notion of co-textual (or contextual) function.65 I will provide an illustration of how rank structure can help to make the notion of a variation-unit more rigorous and then provide an example of how such a definition can be extended beyond the traditional parameters of its use to provide 64 See Thomas G. Pavel, The Spell of Language: Poststructuralism and Speculation, trans. L. Jordan and T. G. Pavel (Chicago: University of Chicago Press, 1989), for a strong critique of the linguistic understanding of the poststructuralists. 65 These features are identified and defined in most works on SFL. The standard introduction is Michael A. K. Halliday, Halliday’s Introduction to Functional Grammar, 4 th ed., rev. by C. M. I. M. Matthiessen (London: Routledge, 2014).

The Domains of Textual Criticism

147

insight into the co-textual function of a given variation-unit. The rank scale in SFL proceeds in a scale of increasing size and complexity from the morpheme to the word to the word group (phrase) to the clause to the clause complex (and conceivably beyond to the paragraph and paragraph complex). The morpheme is the smallest unit of structure, followed by the others until the entire text is reached. Each higher elemental rank consists of elements from the next lower level, so that words make up word groups and, conversely, words are composed of morphemes. The identification of this rank structure of Greek enables us to identify variation-units on the basis of their grammatical structure as the fundamental identifying distinction. So the variants that make up a variation-unit are categorized according to the rank at which they are located, with the possibility that multiple variants can be contained within a variation-unit as part of hierarchical grammatical structure. For example, a word group consists of words that are organized in such a way as to consist of a head term (usually a noun but possibly other elements used substantivally) and its various modifiers (adjectives, prepositional phrases, and even participles, etc.). A variation-unit at the word group level might consist of a word group that has variants of various types within that unit, such as a different ordering of the elements or substitution of different words for the major elements (the categories of Fee are pertinent here, but not sufficient).66 With this system, one may identify variants within variationunits, and even variation-units within larger variation-units. The purpose is to define the boundaries of such a unit on the basis of the rank structure of Greek, without mixing categories, and to establish the relationship among these units, also on the basis of Greek lexicogrammar. An example that has been used in several books but that has proved problematic of explanation is the variation-unit in John 7:1. There are five variations on the readings in this variation-unit: (1) (2) (3) (4) (5)

µετὰ ταῦτα περιεπάτει ὁ Ἰησοῦς µετὰ ταῦτα περιεπάτει Ἰησοῦς περιεπάτει ὁ Ἰησοῦς µετὰ ταῦτα περιεπάτει µετ᾽ αὐτῶν ὁ Ἰησοῦς περιεπάτει ὁ Ἰησοῦς

Fee has trouble explaining this variation-unit, because he identifies it as consisting of two sets of manuscript agreements, those such as (1) and (2) with µετὰ ταῦτα / ὁ Ἰησοῦς, and the issue of the article with Ἰησοῦς. There is of course also the issue of the placement and wording of µετ᾽ αὐτῶν. If this is considered at least three variation-units, then there is great potential for confusion, because these are grammatically analyzable on several different ranks: word, word group, and clause.67 66 67

See Fee, “On the Types,” 63. Fee, “On the Types,” 64–5 and 69.

148

Stanley E. Porter

Rather than seeing these three different types of variants at different ranks, I suggest (as Andrew Pitts and I have elsewhere shown) that this is a variationunit at the clause rank and this establishes the boundaries of the unit. At the clause level, there is both ordering and substitution occurring: with the first variation concerned with ordering of the word groups with the prepositional phrase either as the first (1 and 2), second (4), or third (3) clausal element, and the second variation being its presence or absence (5). Within this clause level variationunit, there are two word group variants, the first the variant being with (1, 3, 4, and 5) or without (2) the article and the other the variant head term of the prepositional phrase (4, as opposed to 1, 2, and 3).68 As an example of the contextual implications of textual variation, I wish to take the example of the use of the finite verb form in the primary clause in Rom 5:1.69 The variation-unit consists of a morphological variant of an /o/ vowel length, determining whether the reading is the first person plural indicative verb ἔχοµεν or the first person plural subjunctive verb ἔχωµεν. I have argued elsewhere that the best reading here, on the basis of external evidence, is with the subjunctive. I do not intend to revisit that discussion here. However, one of the arguments against such a finding, even though the textual evidence is overwhelmingly in its favour and it was the favoured reading from the advent of textual criticism until about the middle of last century, was the supposed fact that utilisation of a subjunctive form introduces exhortation when Paul’s argument requires the certainty of the indicative. I wish instead to show how the variationunit, even if it is defined morphologically, functions in an increasing, expanding, and ever-widening co-text and context, until the use of the subjunctive is seen to be formative for the argument of the entire letter to the Romans. Before I do that, however, I must point out that much of the misunderstanding regarding this variation-unit begins with a misunderstanding of the semantics of the indicative and subjunctive mood-forms. This confusion is the result of both linguistic and theological misunderstandings. The linguistic ones are the mistaken belief that the indicative mood-form indicates the certainty of what is and the subjunctive mood-form, especially in the first person plural (its so-called hortatory use), indicates a lack of certainty and merely speculation or wish or hypothesis. One can see that if one wishes to establish theological truths, these are better established with indicative rather than subjunctive mood-forms. However, these are not the semantics of these mood-forms. This is not the place to review important developments in Greek linguistics over the last thirty or so years – but they have been many and they have been important, to the point that they have See Porter and Pitts, Fundamentals of New Testament Textual Criticism, 85–6. For my most recent treatment of this variation-unit, see Stanley E. Porter, The Letter to the Romans: A Linguistic and Literary Commentary, NTM 37 (Sheffield: Sheffield Phoenix, 2015), 112–15, esp. 114–15, with reference to secondary literature. In this commentary, I also place Rom 5:1 in its larger context. 68 69

The Domains of Textual Criticism

149

left many New Testament scholars with an inadequate understanding of the Greek language – except to say that the indicative mood-form indicates the user’s assertion regarding a process and the subjunctive mood-form indicates the user’s projection of a process, regardless of their relation to “reality.” With this in mind, I think that we can see how co-textual and contextual understanding of the use of the subjunctive ἔχωµεν can help us to understand not just this particular textual variant in Rom 5:1, but how it has far greater implications. The variation-unit in Rom 5:1 is a morphological variant, that is, the smallest rank at which this variant functions is at the morphological level, the alternation of an omicron and omega that indicates the difference between an indicative and a subjunctive form. However, one can also see that this morphological variant also affects the meaning of the word, the verb itself, and this affects the verbal word group (consisting of a head term, the verb, with no modifiers). This word group is the predicator of the clause εἰρήνην ἔχωµεν πρὸς τὸν θεὸν διὰ τοῦ κυρίου ἡµῶν Ἰησοῦ Χριστοῦ. This clause may be rendered, for the sake of this paper, as “let us enjoy peace with God through our Lord Jesus Christ.” This clause is the primary clause within the clause complex that consists of the initial participle secondary clause, “having therefore been justified by faith.” Compressing an argument to its smallest, this clause complex marks the transition in the argument of the letter to the Romans. The body of the letter, which is inclusive of Rom 1:16–11:36, is divided into several parts. The first, Rom 1:18–4:25, describes the sinful human situation and its forensic solution, with human sinfulness that leads to condemnation resolved not by the law of God but by God’s righteousness. The second part of the argument, Rom 5:1–21, sees reconciliation as the climax of Paul’s argument, in which personal relation is established between God and humanity. The fulfillment of the first part of the argument allows for appropriation of the second, in which Paul exhorts the reader to embrace reconciliation as something that follows in one’s relation to God on the basis of having been justified. This perspective has similarities to the use of internal textcritical criteria, but with some exceptions. There is more attention to the co-textual function than simply noting authorial tendencies and closer attention to the discourse function than simply the notion of authorial style. However, the important factor is that the variation-unit is analysed as part of a hierarchy of textcritical levels, so that it is fully integrated into the larger text.

The Place of so-called Para- and Metatextual Data The effects of poststructuralism on the nature of text have led to a number of important further ramifications in textual criticism that have implications for textual scholarship. One of these has come about through further definition of the notion of text. If one might characterise the longstanding notion of text as static and as product, and the poststructuralist notion of text as process and as

150

Stanley E. Porter

productive, then the result of this dialectic (if one dares to use such a framework) is a moderating view of text that rejects the extremes of both yet appreciates what has been learned from the residual advances. In a work that to a large extent reflects this moderating view, the philosopher Jorge Gracia approaches the subject as a discussion of a “theory of textuality.” I note that he assumes that textuality now demands further definition, but he also recognizes that in order to do so he must begin with a notion of text. However, his notion of text necessarily takes into account the poststructuralist notion of textuality. As a result, Gracia offers the following definition of texts: they are “semantically significant artifacts in relation to which authors, audiences, and contexts play interesting and idiosyncratic roles.”70 This definition then forms the basis for the rest of his volume. Each of the parts of the definition merits detailed consideration. The major factors that I note are the following: the importance of meaning as indicated by texts being “semantically significant” even if there is not a specific theory of meaning that is agreed upon, the status as “artifact,” and their having “interesting” even if “idiosyncratic” relationships, which covers a host of non-predictable or not necessarily stable relationships, with “authors” as their producers (Gracia clearly does not go so far as to see the author as having vanished), “audiences” as their users, and “contexts” as their necessary environment of function and use (Gracia’s emphasis is upon the cultural function of texts in conjunction with the above factors, as determinative of meaning, even if meaning is not finally singularly determine). This definition, even if it would not be completely satisfactory to those of either the old or new schools of thought on texts (it introduces too much imprecision and fluidity for the former and too much certainty and fixedness for the latter), attempts to capture the reality and definable existence of the text, while also recognizing that texts are not simply static products, but without introducing a number of undefinable fluid boundaries that make interpretation and analysis more problematic than they already were when they were seen as static entities. Despite a number of innovations, many of them the result of the digital age, there is still a perception within textual criticism that textual criticism is simply concerned with textual variants. The result is that individual texts are seen as repositories of readings to be taken in isolation. This view has been in place from some of the earliest formulations of textual criticism up to the present, including the CBGM. The traditional hand edition of the Greek New Testament, still widely used in New Testament scholarship, encapsulates such an approach by its critical apparatus. There is a sanitized eclectic text placed in the upper portion of the page and a collection of variants, arranged according to sigla representing the repositories of such readings, at the bottom of the page. The various manuscripts, despite any individuality that they may have, are treated as sources of the variants, which are then grouped according to similarity or difference. There are of 70

Gracia, Theory of Textuality, xv.

The Domains of Textual Criticism

151

course reasons for such an approach, such as economy, access, and even usability. However, this also represents a view of meaning of texts that has been questioned by the poststructuralist revolt. Without endorsing the deconstructive extremes, one can note that the notion of meaning reflected in most textual criticism pushes toward a narrow view of textual meaning as definitive and even singular. A more expansive view of meaning – one that recognizes the difficulties in determining meaning, the notion that a text is a process or at least represents a process – is minimized. The emphasis is simply upon wordings, as if they exist apart from actual instances of text. One of the results of poststructuralism is the recognition of intertextuality that expands the borders of texts. These expansive frontiers were always there in the form of para- and metatextual data that comprise each textual artefact. These para- and metatextual data have increasingly been explored, but their full significance has not yet been fully realised. Such data include a wide array of items.71 I would classify these among what is no doubt an even larger list: the ink, the lettering not just in its broad style but in the particular ways letters are formed, the material peculiarities of this artefact (e.g. is writing done around a strand of papyrus or a hole in a parchment), spacing, letter sizing, placement upon the sheet (margins), lines per page or column, number and types of columns, scroll or codex and what kinds and sizes and contents, the length of material on a sheet and other codicological features, corrections, any markings indicating paragraphing or completing of lines, etc., supra- and sublinear lines, the use of nu-lines and sometimes smaller lettering, nomina sacra, other forms of figuration such as staurograms, abbreviations and other kinds of shortened or reduced or otherwise transcribed forms, phonetic notation of any type, accentuation of any type, punctuation of any type, or any related features, headings, notes, marginalia of any type, or prosodic and discourse features, whether they were part of the original scribal hand or added later and for whatever reason, and no doubt numerous other features that I simply have not identified – the list is not inclusive but meant to be suggestive of the scope and potential for such analysis. There has been research into most if not all of these features, some more intensely than others, but there is much that is not yet known about many if not most of them – such as the contexts that produced them, the economics that impelled them, and the beliefs that inspired them. Hurtado’s book on Christian artefacts is one of the most recent and comprehensive. His book shows, however, that we still have very limited knowledge of the function of a variety of these See Larry W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), who is one of the few who takes these things seriously in a single volume. However, I note that despite the title of the book, he does not, so far as I can tell, ever really define what an artefact is. This is an important topic in relation to defining text. See Gracia, Theory of Textuality, 44–52. 71

152

Stanley E. Porter

para- and metatextual data. An example would be the nomina sacra. Besides various theories of their origins, there is still a major question about their meaning or function, besides the issue of their development.72 We know even less about their contextual function, whether within the textual artefact itself or within the larger scribal, communal, or theological context. In the concluding chapter of his volume, Hurtado refers to what he calls “readers’ aids.”73 These include such things as diacritical marks, punctuation, and the like. These are very important elements, but Hurtado’s formulation of the discussion makes them appear as if they are something other than the text. This simply is not so. The use of a diaresis is integral to the text and making clear that we have some type of vowel distinction, depending upon whether organic or inorganic diaresis is utilised. However, the one large question that seems always to be avoided or not yet resolved is how such features are part of what it means for understanding and interpreting the meaning of a text. I have written elsewhere on some of these features of early Christian manuscripts. I determined from examining some of the characteristics of such manuscripts that early Christianity was a culturally literate community (or perhaps communities), a literarily creative community, a theologically reflective and interpretive community, and a consciously theological community.74 Even these categories only provide broad areas of further contemplation, but without necessarily integrating all of the considerations into providing a new interpretation of a particular text-critical artefact, one that takes all of these data into consideration along with their contextual function. As an example, let’s consider for a moment the nomina sacra (again!). Whether they were derived from secular or religious practice (such as found within Judaism), one of the major issues is whether they are abbreviations or not. In other words, are they simply some type of space-saving device? (This is beside the question of whether the particular nomen sacrum is a true abbreviation or a suspension, a distinction that is itself interesting and worthy of further discussion.) They probably were some type of space-saving device, but the more important question is why they were used. Were they simply used to save space? There is some argument to be made for this, as they do save space, and the words that were transferred into nomina sacra did increase, thus potentially increasing the space saved. However, most scholars do not think that this is a sufficient explanation, because they are not used in all manuscripts and not even used regularly in

72 See Hurtado, Earliest Christian Artifacts, 95–134, for a recent but far from conclusive discussion. 73 Hurtado, Earliest Christian Artifacts, 177–85. 74 Stanley E. Porter, “What Do We Know and How Do We Know It? Reconstructing Early Christianity from Its Manuscripts,” in Christian Origins and Greco-Roman Culture: Social and Literary Contexts for the New Testament, ed. S. E. Porter and A. W. Pitts (Leiden: Brill, 2013), 41–70.

The Domains of Textual Criticism

153

particular manuscripts. There is the further issue of the particular kinds of words that became nomina sacra, when there was already a Greek abbreviation system in place. Thus, there seems to be something more to their use than simply the saving of manuscript space. What was it? They seem to be related to use of particular theological nomenclature within early Christian manuscripts. It makes sense that, for some reason – possibly devotion, respect, focus, or highlighting, or some other reason – early Christians used these nomina sacra. This still leaves open a host of questions: what compelled them to develop these nomina sacra to begin with, why do they not use them whenever such words appear and what motivates the use and non-use of them, did all Christian communities use them similarly and if so, why and if not, why not? And these occur among numerous other questions to which we do not know the answers. However, these nomina sacra do offer an open window into the meanings of the texts in which they are – and are not – used. Such a fact is completely overlooked by our modern eclectic texts and their textual apparatus, when such a factor constitutes, I believe, a major interpretive key to the meaning of the texts in which they do – and do not – appear. We may not know what this precise meaning is, but there is meaning and interpretation nevertheless, even intertextual connection, that begs for further explanation.

Conclusion In this paper, I have tried to offer some context for some of the domains of textual criticism, in light not only of its past but also of its present and future. There is much to do and to think about regarding textual criticism. We are in an interesting and potentially productive time within textual criticism when there is potential for reconceptualization of some of the fundamentals of the discipline. I believe that it is important to know something of the intellectual environment that has helped to create the situation that we are in, but more important is to understand not just how we got here but what it means to be here. I have identified three areas where I believe that there is potential for much further exploration. These include the goal and purpose of textual criticism especially as it relates to the authorially-intended text, the concepts of textual variant including variation-unit and linguistic criteria for analysis, and the para- and metatextual data as part of the complex of poststructuralist meaning. Each of these areas is at an important crossroads for further conceptualization in order to move textual scholarship forward.

Finding Your Place: Developing Cross-reference Systems in Late Antique Biblical Codices Gregory Peter Fewster In the early third century CE, Origen, a Christian scholar from Alexandria, migrated to the city of Caesarea Maritima.1 A teacher of Greek grammatical scholarship and a practitioner of Alexandrian techniques of textual criticism, Origen’s move to Caesarea roughly coincided with his adaptations of classical scholarship to Jewish and Christian texts.2 The third and fourth centuries were a time of immense upheaval in the modes and practices of book production, as Anthony Grafton and Michelle Williams have documented. And individuals such as Origen and his student Eusebius were important contributors to these material shifts in knowledge production and access. Origen’s massive undertaking, the Hexapla, tabulated alternative versions of the Hebrew Bible in six parallel columns to enable its textual criticism. And Eusebius likewise arranged textual data into tables to visually display world history in his Chronicon and to conveniently access parallel material among the four canonical Gospels. Tabular arrangement leveraged codex technology for the purposes of a variety of textual criticisms, becoming a hallmark of Caesarean scholarship.3 Grafton and Williams’ work is one example of a small but growing set of scholarship that is taking seriously, rather than just giving passing notice to, the materiality of Christian editorial and scribal practice in late antiquity. And along 1 I wish to thank Jennifer Knust and Tommy Wasserman for sharing material from their once-forthcoming book To Cast the First Stone: The Transmission of a Gospel Story (Princeton: Princeton University Press, 2018), along with the other conference attendees for their helpful and encouraging remarks. I am also grateful to Jeremiah Coogan for sharing forthcoming publications on the Eusebian apparatus. 2 See discussion in Bernhard Neuschäfer, Origenes als Philologe, 2 vols. (Basel: F. Reinhardt, 1987); Timothy David Barnes, Constantine and Eusebius (Cambridge, MA: Harvard University Press, 1981), 83–95; Ronald E. Heine, Origen: Scholarship in the Service of the Church (Oxford: Oxford University Press, 2010), 22–5. Peter Martens, in fact, considers the work of Origen’s Hexapla to have been initiated in Alexandria. Peter W. Martens, Origen and Scripture: The Contours of the Exegetical Life (Oxford: Oxford University Press, 2012), 45–6. 3 Anthony Grafton and Megan Williams, Christianity and the Transformation of the Book: Origen, Eusebius, and the Library of Caesarea (Cambridge, MA: Harvard University Press, 2006).

156

Gregory Peter Fewster

with this growing interest, book historians and textual critics are devoting more attention to the study of paratextual features in manuscripts and print books.4 This kind of work has productive potential for the increasingly diversifying field of textual scholarship on the New Testament. As research on the text of the New Testament undergoes shifts in its theory, method, and goals,5 we are also seeing heightened interest in the early manuscripts that contain New Testament texts as artefacts, especially under the auspices of the term “scribal habits.”6 These are certainly not mutually exclusive endeavours; detailed study of manuscripts and scribal practice informs text-critical method. However, the text of the New Testament persistently exhibits a gravitational pull as a desirable object of study, over and against other features of the manuscripts that contain it. As much as these recent studies uncover interesting features of individual early Christian manuscripts, there is still much to learn about the social and material history of early Christian book production. This includes issues of codicology and palaeography, but also extends the scope of relevant data to the various paratexts that facilitated the reading of New Testament (as well as other texts) in antiquity. In addition to the types of bookish transformations described by Grafton and Williams at the hands of Origen and Eusebius, biblical manuscripts from late antiquity attest to an even wider diversity of materials that reflect the kinds of innovative modification of book technologies endemic to Caesarea. These materials emerge with striking rapidity and variability, especially when we consider the particularities of individual manuscripts. This essay aims to contribute broadly to the growing body of scholarship on the development of paratexts by early Christian scribes and scholars and in the material history of the book by providing an account of the emergence and 4 This conversation has been prompted to a great degree by Gérard Genette, Paratexts: Thresholds of Interpretation, trans. J. E. Lewin (Cambridge: Cambridge University Press, 1997). Additional contributions relevant to this essay can be found throughout the notes. 5 I refer primarily to the incorporation of the Editio Critica Maior into the NA28 (2012), and the publication of the ECM Acts volume, which involves the use of the Coherence-Based Genealogical Method and a shift in editorial goals from the “original text” to the Ausgangstext. Other recent editions, making use of different editorial principles include Michael Holmes, The Greek New Testament: SBL Edition (Atlanta: SBL, 2010) and Dirk Jongkind et al., The Greek New Testament (Cambridge: Cambridge University Press, 2017). 6 Larry Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006) represents an important influence to this kind of scholarship, though note the earlier David Parker, Codex Bezae: An Early Christian Manuscript and its Text (Cambridge: Cambridge University Press, 1992). See also chapters 1 and 2 of Tommy Wasserman, The Epistle of Jude: Its Text and Transmission (Stockholm: Almqvist & Wiksell, 2006); and more recently, W. Andrew Smith, A Study of the Gospels in Codex Alexandrinus: Codicology, Palaeography, and Scribal Hands, NTTSD 48 (Leiden: Brill, 2014); Alan Mugridge, Copying Early Christian Texts, WUNT 362 (Tübingen: Mohr Siebeck, 2016); and Peter Malik, P.Beatty III (P47): The Codex, Its Scribe, and Its Text, NTTSD 52 (Leiden: Brill, 2017).

Finding Your Place

157

transmission of two roughly-contemporaneous paratextual cross-reference systems: Eusebius’ Gospel-harmony system known as the Eusebian apparatus, and a system of cross-referential chapter divisions, or kephalaia, which make up a substantial portion of the Euthalian apparatus. As Eusebius and Euthalius make clear in their respective explanations of their editorial activity, the production of cross-referencing systems comes with significant debt to the work of predecessors. This essay takes these references as an invitation to pursue the influence of these specific debts on Eusebius and Euthalius and to comparatively contextualize their work more broadly within paratextual systems of classical and late antiquity. Such comparison is suggestive both genealogically and analogically, and is driven by a focus on three elements of cross-referencing systems in particular that comprise the Eusebian and Euthalian systems: tabular organization, marginal sigla, and physical relationship of paratexts to the main text. Schematizing paratextual systems this way accommodates the present concern for Eusebian and Euthalian material in late antiquity, while setting parameters for future studies. It opens the door for further comparison between late antique paratextual systems and their various components, as well as scholarship on their transmission and development into the medieval period.

Eusebius and Euthalius among Late Antique Biblical Paratexts Late antiquity saw an explosion in the production of beautiful, deluxe biblical codices, which included the incorporation of numerous paratextual materials.7 Many of these incorporated some sort of marginal notation system that corresponded with information, such as short chapter descriptions (kephalaia), tabulated at the head of the text. Modern day indices or tables of contents are not-sodistant cousins to these technologies. The correspondence between prefixed, tabulated information and marginal sigla would allow the reader to flip back and forth between text and paratext – one of the benefits of codex technology – enabling alternative modes of reading and study that supplement linear scanning. Cross-referencing paratexts were not consistent in their form or function, however. Kephalaia could be tabulated at the head of book or separated and placed in the upper margins, or both. And marginal numerals could correspond with tabulated kephalaia or simply indicate a chapter division. Such diversion is striking, but also difficult to account for given that many of these paratextual systems including Gospel kephalaia and chapter divisions marked by marginal numerals in Acts or the Pauline epistles appear on the biblical page without any 7 This is not to say that earlier books did not include paratextual elements. P46 (P.Beatty II + P.Mich. inv. 6238), for example, contains page numbers and letter-titles/superscriptions. However, the paratextual materials of these later parchment codices are much more substantial and complex.

158

Gregory Peter Fewster

explanation by their ancient scribes or editors. However, two additional paratextual apparatuses were furnished with explanations of their function and origin: the Eusebian Gospel canons and the Euthalian apparatus of Acts and the Epistles.8 These explanatory materials occur, respectively, in a brief prefatory letter written by Eusebius to Carpianus and in Euthalius’ Prologue to his edition of the Pauline epistles. Both of these explanatory materials provide unparalleled windows into late antique Christian editorial practice, serving as important data in an account of the emergence of paratextual cross-reference systems in late antique biblical codices.9 However, making use of these materials is not without challenges, since 8 There is some debate concerning the identity of the composer of the so-called Euthalian apparatus, including which materials derive from this initial compositional moment and which were added later. Eric Scherbenske includes an up-to-date discussion on the identity of the initial editor in Canonizing Paul: Ancient Editorial Practice and the Corpus Paulinum (Oxford: Oxford University Press, 2013), 118–20; see also Albert Ehrhard, “Der Codex H ad Epistulas Pauli und ‘Euthalios Diaconos’: Eine palaeographisch-patrologische Untersuchung,” Zentralblatt für Bibliothekswesen 8 (1891): 385–411; Günther Zuntz, “Euthalius = Euzoius?” VC 7/1 (1953): 16–22. As Charles Willard’s research shows, a large number of paratextual materials come to be included in the Euthaliana (A Critical Study of the Euthalian Apparatus [Berlin: De Gruyter, 2009]). Most scholars today tend to follow the indications of the Prologue of the Pauline epistles, which seems to articulate a distinct edition of the corpus Paulinum characterised by its Prologue along with Divine Testimonies, lections/lection table, and kephalaia. See also the important arguments of J. Armitage Robinson, who distinguishes initial and expanded layers of the Divine Testimonies and argues that the Martyrium Pauli is a secondary addition to the Prologue and dependent upon it, and Nils Dahl, who argues that the hypotheses included in many Euthalian-type manuscripts derive from an initially independent edition of the corpus Paulinum. See J. Armitage Robinson, Euthaliana: Studies of Euthalius Codex H of the Pauline Epistles and the Armenian Version, with an Appendix Containing a Collation of the Eton MS of the Pseudo-Athanasian Synopsis (Cambridge: Cambridge University Press, 1895), 17–9, 28–30; Nils A. Dahl, “The ‘Euthalian Apparatus’ and Affiliated Argumenta,” in Studies in Ephesians: Introductory Questions, Text& Edition-Critical Issues, Interpretation of Texts and Themes, ed. D. Hellholm, V. Blomkvist, and T. Fornberg (Tübingen: Mohr Siebeck, 2000), 244–45, 253–57. In this essay, the name Euthalius will be used without comment for the sake of convenience and does not take a stake in the debate. 9 Some scholars have already found these materials useful in elaborating on Eusebius and Euthalius’ editorial goals. On Eusebius, see Carl Nordenfalk, Die spätantiken Kanontafeln: Kunstgeschichtliche Studien über die eusebianische Evangelien-konkordanz in den vier ersten Jahrhunderten ihrer Geschichte. Textband (Göteborg: O. Isacsons, 1938), 46–9; Nordenfalk, “The Eusebian Canon-Tables: Some Textual Problems,” JTS 35/1 (1984): 97–8; Thomas O’Loughlin, “Harmonizing the Truth: Eusebius and the Problem of the Four Gospels,” Traditio 65 (2010): 1–29; Martin Wallraff, Kodex und Kanon: Das Buch im frühen Christentum (Berlin: De Gruyter, 2013), 29–37; Satoshi Toda, “The Eusebian Canons: Their Implications and Potential,” in Early Readers, Scholars and Editors of the New Testament: Papers from the Eighth Birmingham Colloquium on the Textual Criticism of the New Testament, ed. H. A. G. Houghton (Piscataway: Gorgias, 2013), 27–43; Matthew R. Crawford,

Finding Your Place

159

these prefatory materials are not preserved in the earliest manuscripts that contain elements of the systems. For example, Codex Sinaiticus (‫א‬01), the earliest manuscript to contain Eusebian numeral-sets in its margins, does not appear to have included the Epistula ad Carpianum or the canon tables. Crucial portions of both Codex Alexandrinus (A02) and Codex Coislinianus (HP015) are lost or damaged, and thus lack the prefatory material that may have corresponded with the Eusebian numerals in the former manuscript and Euthalian materials in the latter. In the effort to make sense of each paratextual apparatus as a whole, scholars must rely on critical editions of the prefatory and other materials, reconstructed from later manuscripts.10 This procedure is not unwarranted. However, for both the Eusebian and Euthalian prefatory material, the critical texts are very old and their limited manuscript base has prompted some concern.11 And with respect to the interests of this essay, a priori reliance on a reconstructed text in critical editions has the potential to idealize and fix each cross-referencing apparatus and thus overdetermine the analysis of their emergence as distinctive systems. As Jerome McGann explains, even insofar as the critical edition “reconstitute[s] for the reader, in a single text, the entire history of the work as it has emerged into the present…it equally embodies an illusion about its own historicity (or lack thereof).”12 This “illusion” as McGann calls it, is all the more pernicious with respect to paratextual materials, since their non-linguistic variations from manuscript to “Ammonius of Alexandria, Eusebius of Caesarea and the Origins of Gospel Scholarship,” NTS 61/1 (2015): 1–29; Francis Watson, The Fourfold Gospel: A Theological Reading of the New Testament Portraits of Jesus (Grand Rapids: Baker, 2016), 109–23; Jeremiah Coogan, “Mapping the Fourfold Gospel: Textual Geography in the Eusebian Apparatus,” JECS 25/3 (2017): 337–57; Coogan, “Rewriting the Temple Incident with Eusebius of Caesarea,” JBL (forthcoming); Knust and Wasserman, To Cast the First Stone. On Euthalius, see Robinson, Euthaliana, 11–27; Willard, Critical Study, 1–56; Scherbenske, Canonizing Paul, 120–57. 10 For editions of Eusebius’s Epistula ad Carpianum, see Eberhard Nestle, “Die Eusebianische Evangeliensynopse,” Neue Kirchliche Zeitschrift 19 (1908): 40–51, 93–114, 219– 32 with minor edition in the NA28, 85*–94*. See also critical discussion and translation in Harold H. Oliver, “The Epistle of Eusebius to Carpianus: Textual Tradition and Translation,” NovT 3 (1959): 138–45. For editions, of Euthalius’ Prologues and Euthalian material more broadly, see Laurentius Alexander Zacagnius, Collectanea Monumentorum Veterum Ecclesiae Graecae ac Latinae quae Hactenus in Vaticana Bibliotheca Delituerunt. Tomus Primus (Rome: Typis Sacra Congreg. de Propag. Fide, 1698), 515–709; J. P. Migne’s PG 85.627–790; Hermann Freiherr von Soden, Die Schriften des Neuen Testaments, vol. 1/1 (Berlin: Duncker, 1902), 637–82; Vemund Blomkvist, Euthalian Traditions: Text, Translation and Commentary (Berlin: De Gruyter, 2012), 45–117 (99–117), reproducing von Soden’s text. 11 See, for example, Nordenfalk, “The Eusebian Canon-Tables”; Blomkvist, Euthalian Traditions, 5. 12 Jerome J. McGann, A Critique of Modern Textual Criticism (Chicago: University of Chicago Press, 1983), 93–4.

160

Gregory Peter Fewster

manuscript (i.e., position and arrangement) significantly alter the paratexts’ function.13 The following analysis continues to benefit from the work done by critics E. Nestle and L. A. Zacagnius to reconstruct respectively the prefatory material of Eusebius and Euthalius, but it also prioritizes the documentary evidence of extant manuscripts. Familiarity with the reconstructed texts of the Epistula ad Carpianum and Euthalius’ Prologues is indispensable. Eusebius, Euthalius, and the other unnamed producers of biblical paratexts share editorial affinities with one another and betray dependencies upon foregoing scribes and scholars. It is the task of comparison to bring these genealogies to light. However, this exercise also benefits from the highly analogical component of comparison, as emphasized by scholars such as Jonathan Z. Smith and John S. Kloppenborg.14 According to their formulations, interested comparison – and in this case I am interested in three particular features – anchors the comparative exercise and prevents hypotheses of unwarranted dependence. Comparison that supposes analogical relationships between manuscripts’ paratextual elements and elements of formatting allows primarily for an appreciation of distinctive uses of such features and their combination in the Eusebian and Euthalian systems, as well as genealogical explanations when additional evidence permits. As he explains in his Epistula ad Carpianum, Eusebius was interested in showing parallels between the canonical Gospels in a way that did not disrupt the actual Gospel text, as Ammonius an earlier scholar had done. So after assigning consecutive numerals to each pericope in each of the four Gospels, Eusebius composed ten tables that each represented a possible iteration of a Gospel parallel (e.g., Matt-Mark-Luke or Matt-Mark, etc.). He then noted the parallels by listing corresponding pericope numerals in the appropriate table, while including the table number underneath the corresponding paragraph number in the margin of the Gospel book. The reader of the Gospel text could thus find parallel accounts in other Gospels by following the marginal numeral-sets back to the appropriate canon, which would direct the reader to the parallel pericope in another Gospel.15 Eusebius’ Gospel canon system is clearly distinct from the kephalaia-type paratexts described above, with respect to the tabulated content. However, the crossreferencing facilitated by marginal numeration operates according to shared 13

This point relies upon Genette’s notion of the paratext as a “threshold of interpreta-

tion.” Jonathan Z. Smith, Drudgery Divine: On the Comparison of Early Christianities and the Religions of Late Antiquity (Chicago: University of Chicago Press, 1990), 36–53; John S. Kloppenborg, “Disciplined Exaggeration,” NovT 59/4 (2017): 390–414. 15 Nordenfalk suggests that Eusebius may have had other functions in mind that he did not articulate in the Epistula. Namely, the construction of 10 tables, when there conceivably could have been more, imbues the tables with a somewhat mystical status, carried on in the elaborate decorations of the canons in later manuscripts. Carl Nordenfalk, “Canon Tables on Papyrus,” Dumbarton Oaks Papers 36 (1982): 29–30. 14

Finding Your Place

161

principles. And the canons’ association with Eusebius places their origin slightly earlier than the biblical codices mentioned above. The Euthalian apparatus resembles the kephalaia-system much more closely than the Eusebian canons do. As he explains in a section of the Prologue to the Pauline epistles, Euthalius adapted a list of chapter headings from someone he calls a “Christ-loving father” – one of a few debts Euthalius admits, including one to Eusebius for information concerning Paul’s martyrdom.16 And these chapter headings Euthalius claims to have prefixed to each letter along with other paratextual material. The Prologue describes this as follows: Now let this be said about them [i.e., components of politaeia] with reference to our epitome. But in what follows, we will prefix to each letter a short exposition of the chapters, worked out by one of the wisest of our fathers (a Christ-lover). Not only will we do this, but we have summed up precisely the precise division of the lections and approved list of the Divine Testimonies. This we will publish just after this Prologue.17

The Prologue’s editorial notice lacks extensive description of the paratextual materials and their navigation, highlighting primarily the elements included in the edition, the debt to the wise father, and the relative location of some of the elements. However, these comments begin to make more sense when read next to manuscripts of the Euthalian edition and other late antique biblical manuscripts containing analogous paratextual elements. Although it lacks this explanatory Prologue, Codex Coislinianus prefixes kephalaia in tabulated form to the epistles with which they correspond. And each kephalaion is assigned a numeral, which is duplicated in the margin of the Epistles at the appropriate location.18 This pattern is visible, with some variation, in later Euthalian manuscripts. Readers could thus gain a sense of the contents of the given Pauline Epistle before even reading it, as well as find topical paragraphs with relative ease thanks to the numeration system. Not only do the kephalaia summarize the contents of the Epistles, they also reflect a programmatic mode of exposition. As Vemund Blomkvist has argued, the language of the kephalaia displays a systematic interest in exhortative language, which corresponds with the language in the pre-text (i.e., the verse with which a given kephalaion 16 According to the Prologue to the Catholic Epistles, Euthalius composed his own kephalaia and Divine Testimonies for the Catholic Epistles. See Robinson, Euthaliana, 20–6; Willard, Critical Study, 47–56, for discussion of the relationship between the summaries. Several scholars consider Pamphilus to be the likely composer of the Pauline kephalaia. See F. C. Conybeare, “On the Codex Pamphili and the Date of Euthalius,” Journal of Philology 23/46 (1895): 241–59; Conybeare, “The Date of Euthalius,” ZNW 5/1 (1904): 39–52; Dahl, “Euthalian Apparatus,” 240. 17 See the Greek text in Zacagnius, Collectanea, 528–29; Blomkvist, Euthalian Traditions, 107. 18 Since Codex Coislianius is so fragmentary, many of the marginal numerals do not have corresponding kephalaia tables extant, and vice versa. And, as I will discuss below, not all marginal numerals are present even when anticipated by the kephalaia table.

162

Gregory Peter Fewster

most closely corresponds).19 And Eric Scherbenske adds that this adopted system in the kephalaia fits with Euthalius’ editorial interests, insofar as he recorded them in the Prologue: preliminary instruction centred on a Pauline ideology of politaeia (citizenship).20 However closely the interests of the Prologue can be coordinated with a reading of the kephalaia, Codex Coislinianus and other Euthalian manuscripts contain features not explicitly expressed in that Prologue.21 These divergences are pertinent to the question of the emergence of cross-reference systems in late antique biblical codices. Euthalius states in the Prologue only that he prefixed expositions of chapters (προτάξοµεν τὴν τῶν κεφαλαίων ἔκθεσιν)22 to each epistle. In contrast to Eusebius’ description of his apparatus, Euthalius does not indicate that he numerated his kephalaia, and he does not suggest that the kephalaia are part of the cross-reference system or explain how a reader might navigate such a system. Yet in Coislinianus, kephalaia tables are materially present as part of a cross-referencing system facilitated by marginal numerals. There is thus a gap here between what Euthalius describes in his editorial comments as fundamental features of his edition and their manifestation in the manuscript record. According to their prefatory discourses, Eusebius and Euthalius differed slightly according to their editorial goals. Eusebius was interested in constructing relations between the four canonical Gospels and Euthalius seemed interested in directing the reading of the Pauline corpus for the purpose of social formation and theological instruction. However, both of them furthered their editorial projects through the adaptation and production of paratextual materials, ancillary systems that would mediate material engagement through these texts. There are ambiguities associated with Eusebius and Euthalius’ descriptions of their editorial projects and how we might imagine this to have worked out materially on the page. But the manuscript record shows that both of their paratextual apparatuses would quickly prompt engagement with a codex according to principles of cross-referencing, relying on annotated margins of the main text and tabulated information that was ancillary to the main text. That is to say, from a documentary standpoint, by the fifth and sixth centuries, the Eusebian apparatus and the 19 Blomkvist, Euthalian Traditions, 121–42. This analysis builds on a previous and more focused analysis of paraenetic language in the kephalaia. See David Hellholm and Vemund Blomkvist, “Parainesis as an Ancient Genre-Designation: The Case of the ‘Euthalian Apparatus’ and the ‘Affiliated Argumenta,’” in Early Christian Paraenesis in Context, BZNW 125, ed. J. Starr and T. Engberg-Pedersen (Berlin: De Gruyter, 2004), 465–519. 20 See Scherbenske, Canonizing Paul, 133–36, 141–46. 21 The most significant portion of “non-Euthalian” material that comes to be included in Euthalian editions is the Hypotheses/Argumenta. Nils Dahl argues that these originated in a distinct edition, later combined with the Euthalian edition. Dahl, “Euthalian Apparatus,” 244–45, 253–57. See the full discussion of what Willard calls “minor pieces” in Critical Study, 59–92. 22 Zacagnius, Collectanea, 528; Blomkvist, Euthalian Traditions, 107.

Finding Your Place

163

Euthalian apparatus existed as fully-established cross-referencing systems.23 And according to their prefatory material, the components of the cross-referencing systems derived from earlier materials, but were adapted to contemporary editorial projects. Tabulation Tabulation of prefatory material certainly predates the work of Eusebius and Euthalius. Tables of contents or indices, for example, are attested for classical letter collections, including those of Cicero (particularly, Ad familiares) and Pliny the Younger.24 It is difficult to know when the tables of contents for Cicero’s Ad familiares were compiled, whether with an initial compiling of letters or subsequently in their transmission history.25 However, a partial manuscript of the late fifth or early sixth century, likely of the ten-book collection of Pliny’s Epistles, contains a table of contents for book 3 (Morgan Library, Ms. M.462, 48r–49r), which Roy Gibson argues may derive from Pliny himself.26 23 In addition to the Greek Codex Coislinianus, Gothic and Syriac manuscripts provide late antique witnesses. See Günther Zuntz, The Ancestry of the Harklean New Testament (London: British Academy, 1945); James W. Marchand, “The Gothic Evidence for the ‘Euthalian Matter,’” HTR 49/3 (1956): 159–67; Sebastian P. Brock, “The Syriac Euthalian Material and the Philoxenian Version of the NT,” ZNW 70 (1979): 120–30. Early attestation for Eusebian material exists in Greek, Latin, and Ge’ez, though much of it is also fragmentary. See Nordenfalk, Die spätantiken kanontafeln, 1.165–220 (Latin), 221–55 (Syriac); Nordenfalk, “Canon Tables on Papyrus.” The Latin tradition is particularly curious since it seems to have been taken over from the Greek at two separate intervals. See H. A. G. Houghton, “Chapter Divisions, Capitula Lists, and the Old Latin Versions of John,” Revue Bénédictine 121/2 (2011): 320–21. 24 Ad familiares, Biblioteca Medicea Laurenziana, Plutei 49.9 (Codex Mediceus 49.9), 18v, 32r, 49r, 18r, 100v–101r, 134r, 150v–151r, 174v, 209r–210v, 260v–261r. 25 The publication of Cicero’s letters (both when and in what shape) is a particularly fraught issue, which makes it difficult to determine when the tables of contents were incorporated into the manuscript tradition. See especially Mary Beard, “Ciceronian Correspondences: Making a Book Out of Letters,” in Classics in Progress: Essays on Ancient Greece and Rome, ed. T. P. Wiseman (Oxford: Oxford University Press, 2002), 103–44; Roy Gibson, “On the Nature of Ancient Letter Collections,” JRS 102 (2012): 56–78; R. W. McCutcheon, “A Revisionist History of Cicero’s Letters,” Mouseion: Journal of the Classics Association of Canada, III, 13/1 (2016): 35–63. If Shackelton Bailey is correct that Tiro compiled the collection, then there might be reason to suppose that he was involved in the production of tables of contents. D. R. Shackelton Bailey, “Introduction,” in Epistulae ad Familiares, vol. 1 (Cambridge: Cambridge University Press, 1977), 24–5. However, even positing Tiro as the compiler of this particular collection does not make a claim for the shape of others (both the ordering of books and the ordering of letters in each book), which has implications for the paratexts. 26 Gibson’s argument relies on the peculiar function of the indices, which provide more complete identification of Pliny’s correspondents through double-names, otherwise usually referred to by a single name in the letter headings. It is difficult to imagine who else could

164

Gregory Peter Fewster

A more immediate antecedent to the type of tabulation seen in the systems of both Eusebius and Euthalius likely comes from the use of columns begun by Origen and Ammonius in Alexandria. Origen is well known for his Hexapla, six columns of competing versions of the Hebrew Bible, whose parallels could be useful for critical, apologetic, or even exegetical purposes.27 As Grafton and Williams note, with its six columns, the Hexapla pushed third century codex technology to its limits.28 But it was precisely this codex technology that would enable parallel material to visually “make sense.” For parallel columns in a bookroll simply facilitated linear reading, not comparison. Very little direct documentary evidence remains for the Hexapla itself outside of two palimpsest manuscripts. The earliest of these manuscripts is from the Cairo Genizah (Cambridge University Library, T-S 12.128; Rahlfs 2005), containing portions of Psalm 22 in the translations of Aquila, Symmachus, and the LXX.29 A considerably larger tenth-century manuscript (Bibliotheca Ambrosiana O 39; Rahlfs 1098) also contains portions of the psalms and is better called a Tetrapla insofar as it lacked the initial column containing the Hebrew text.30 Additional extant manuscripts have access to this sort of information, though Gibson is open to a roughly-contemporary editor having such knowledge. See Roy Gibson, “Starting with the Index in Pliny,” in The Roman Paratext: Frame, Texts, Readers, ed. L. Jansen (Cambridge: Cambridge University Press, 2014), 41–5 (plates: 42–4). See E. A. Lowe and E. K. Rand, A Sixth-Century Fragment of the Letters of Pliny the Younger: A Study of Six Leaves of an Uncial Manuscript Preserved in the Pierpont Morgan Library, New York (Washington D.C.: Carnegie Institution of Washington, 1922), 13 on the date of the manuscript. 27 See discussion of the various potential purposes for the Hexapla in Joachim Schaper, “The Origin and Purpose of the Fifth Column of the Hexapla,” in Origen’s Hexapla and Fragments: Papers Presented at the Rich Seminar on the Hexapla, Oxford Centre for Hebrew and Jewish Studies, 25 th–3 rd August 1994, TSAJ 58, ed. A. Salvesen (Tübingen: Mohr Siebeck, 1998), 11–4; T. M. Law, “Origen’s Parallel Bible: Textual Criticism, Apologetics, or Exegesis?” JTS 59/1 (2008): 1–21; Martens, Origen and Scripture, 47–63. 28 Grafton and Williams, Christianity and the Transformation of the Book, 103. 29 See C. Taylor, Hebrew-Greek Cairo Genizah Palimpsests from the Taylor-Schechter Collection: Including a Fragment of the Twenty-Second Psalm according to Origen’s Hexapla (Cambridge: Cambridge University Press, 1900). The manuscript likely had all six columns. See R. G. Jenkins, “The First Column of the Hexapla: The Evidence of the Milan Codex (Rahlfs 1098) and the Cairo Genizah Fragment (Rahlfs 2005),” in Origen’s Hexapla and Fragments: Papers Presented at the Rich Seminar on the Hexapla, Oxford Centre for Hebrew and Jewish Studies, 25 th–3 rd August 1994, TSAJ 58, ed. A. Salvesen (Tübingen: Mohr Siebeck, 1998), 88–102. Rahlfs assigns this manuscript to the seventh century, whereas Taylor characterises the “sloping uncials” of the manuscript as post-600. A. Rahlfs, Septuaginta, Vol. X. Psalmi cum Odis (Göttingen: Vandenhoeck & Ruprecht, 1931), 13; Taylor, Hebrew-Greek Cairo Genizah Palimpsests, 13. 30 See Iohannis Mercati, Psalterii Hexapli Reliquiae. Pars Prima. Codex Rescriptus Bybliothecae Ambrosianae 0 39 SVP. Phototypice Expressus et Transcriptus (Rome: Bybliotheca Vaticana, 1983). Rahlfs assigns 1098 its tenth-century date. Rahlfs, Psalmi cum Odis, 12.

Finding Your Place

165

appear to transmit “Hexaplaric” texts, deriving from one of Origen’s columns. However, the extant palimpsests, especially the Cairo Genizah fragment, clearly show the tabular arrangement of competing translations of the Hebrew Bible. Emerging from the same scribal milieu as Origen’s Hexapla was the more modest contribution of Ammonius of Alexandria, who produced a Gospel harmony, or as Matthew Crawford prefers to call it, a Diatessaron Gospel.31 No manuscripts of Ammonius’ Diatessaron Gospel remain. As such, a primary source for Ammonius’ editorial work actually comes from Eusebius’ Epistula ad Carpianum, in which he acknowledges the work of Ammonius to produce the Diatessaron Gospel, emphasizing its problems in order to justify the production of a new system of apprehending Gospel parallels.32 Eusebius describes Ammonius’ Diatessaron Gospel as a synopsis – portions of Mark, Luke, and John were placed in parallel columns next to corresponding passages of a continuous text of Matthew. As Crawford notes, there is some ambiguity concerning the identity of Ammonius and thus his relationship to Origen and Eusebius. The Ammonius of Eusebius’ Epistula may have been Origen’s teacher or a rough contemporary of Origen, each option suggesting a different direction of influence between the Diatessaron Gospel and the Hexapla.33 Regardless of the precise identity of Ammonius and the direction of influence, both Ammonius’ Gospel harmony and Origen’s Hexapla leveraged the visual power of parallel columns to show relative discrepancies between texts that had, on a more abstract scale, definite affinities. Eusebius’ scholarship reflects debts to both Origen and Ammonius. The Epistula ad Carpianum indicates direct influence of Ammonius on Eusebius’ canon tables.34 But scholars such as T. D. Barnes and Grafton and Williams, consider Eusebius to have drawn on the tabular strategies of information management of the Hexapla in the production of his Chronicon, a text that relied heavily on parallel columns to visually display the temporal synchrony of world-historical

Crawford, “Origins of Gospel Scholarship,” 11. See the text in NA28, 84* and an English translation in Oliver, “The Epistle of Eusebius to Carpianus,” 144. Jerome confirms this dependency and places Ammonius in Alexandria with Origen: “Ammonius, vir disertus et valde eruditus in philosophia, eodem tempore Alexandriae clarus habitus est: qui inter multa ingenii sui et praeclara monumenta etiam de consonantia Moysi et Jesu elegans opus composuit, et evangelicos canones excogitavit, quos postea secutus est Eusebius Caesariensis” (Vir. ill. 55, PL 23.667). 33 See Mark Edwards, “Ammonius, Teacher of Origen,” The Journal of Ecclesiastical History 44/2 (1993): 169–81; Crawford, “Origins of Gospel Scholarship,” 3–6. 34 Earlier treatments of the Eusebian tables attribute the paragraph numeration to Ammonius, but Crawford corrects this position by attributing both the paragraph numeration and Canon numeration to Eusebius. Crawford, “Origins of Gospel Scholarship,” 22. 31 32

166

Gregory Peter Fewster

events.35 These ambiguities of influence between Origen and Ammonius and between Origen, Ammonius, and Eusebius do not allow for a strong genealogical hypothesis, but do reflect realistically the complexity of scholarly dependence. Strategies of tabulation were being passed along and developed between Christian scholars, and the Library at Caesarea seems to have furnished both the space and the materials for Eusebius to develop a penchant for tabulation, seen in Eusebius’ Chonicon as well as canon tables of the Gospels and the Psalms.36 Eusebius’ Chronicon reflects a fairly simple adaptation of Alexandrian tabulation strategies.37 However, Eusebius’ Psalm and Gospel tables reflect a more complicated adaptation, inverting the tabulation strategies of Ammonius and Origen. Whereas Ammonius visually exposed parallels through disrupting the text of three Gospels, Eusebius maintained the integrity of each Gospel text through a more abstract mode of tabulation. Parallel columns were maintained, but they tabulated relations marked by representative numerals, and not the text-deemedparallel itself. Tabulation in the mode of Eusebius’ Gospel canons thus involved the creation of a discrete textual unit. The production of new and discrete textual units is precisely what we see in paratexts like the indices of Pliny’s letters or the kephalaia of Euthalius’ edition of the corpus Paulinum. The Euthalian Prologue does not explicitly mention the tabulation of kephalaia, but the formatting of the kephalaia is hinted at through Euthalius’ depiction of dependence. Just as Eusebius accounts for his Gospel divisions – the raw material for his canon tables – by the work of Ammonius, Euthalius states that he drew upon the kephalaia of another for his edition. Both take over content from predecessors for their new paratext but modify the formatting. Euthalius betrays some reliance on Eusebius as well, explicitly marking his dependence on what he calls “the chronological tables of Eusebius of

35 T. D. Barnes sees the Hexapla as the clear inspiration for Eusebius’ tabular format, though Barnes also understands the Chronological canons to be a “genuine innovation” with respect to the representation of chronological time. Barnes, Constantine and Eusebius, 120. Drawing on Barnes, Grafton and Williams state that “Origen’s Bible, laid out in columns, seems the obvious prototype for Eusebius’s efforts to lay out time in the same way.” Grafton and Williams, Christianity and the Transformation of the Book, 142. 36 Eusebius’ Psalm tables are a much lesser-known and less-attested paratext. See discussion in Martin Wallraff, “The Canon Tables of the Psalms: An Unknown Work of Eusebius of Caesarea,” Dumbarton Oaks Papers 67 (2013): 1–14. 37 There is little direct evidence for Eusebius’ Greek Chronicon. However, the tabular layout has been preserved in Latin versions that likely derive from Jerome. See Alfred Schoene, Die Weltchronik des Eusebius in ihrer Bearbeitung durch Heironymus (Berlin: Weidmann, 1900); Rudolf Helm, Eusebius Werke, siebenter Band. Die Chronik des Hieronymus (Leipzig: Hinrichs’sche, 1926), 2.ix–xxviii; Rudolf Helm, Hieronymus’ Zusätze in Eusebius’ Chronik und ihr Wert für die Literaturgeschichte (Liepzig: Dieterich, 1929); Alden A. Mosshammer, The Chronicle of Eusebius and Greek Chronographic Tradition (Lewisburg: Bucknell University Press, 1979), 29–38.

Finding Your Place

167

Pamphilus” (τῶν χρονικῶν κανόνων εὐσεβίου τοῦ παµφίλου)38 for the section on the chronology of Paul’s preaching in the Prologue.39 Although this reference provides no positive evidence for Euthalius’ dependence on Eusebius on matters of formatting, it does reflect interaction with Eusebian scholarship and awareness of the tabular format of the Chronicon. In contrast to the marginal kephalaia of Acts in Codex Sinaiticus, the Gospels of Codex Alexandrinus, or even the marginal Euthalian kephalaia of some later minuscules (e.g., GA 1828), the language of “prefixing” that Euthalius used in his Prologue implies that the headings were gathered together in one place, before each letter (συντόµως ἐπιστολὴν ἐν τοῖς ἑξῆς προτάξοµεν τὴν τῶν κεφαλαίων ἔκθεσιν).40 This is precisely what we see in Codex Coislinianus as well as other kephalaia, such as in the Gospels in Codex Alexandrinus. >> κεφάλαια τῆϲ πρὸϲ τιµό >> θεον ὰ ἐπιϲτολῆϲ παῦλου· α

β

γ

δ

περὶ τῆϲ ἐιϲ ἀγαπην θυ ὀδηγίαϲ τὴν ἀπροϲ δε ῆ· νοµικῆϲ ἀνάγκηϲ· περὶ τῆϲ ἑἀυτοῦ ἐκλογῆϲ· εἰϲ ἐυαγγελιϲ τὴν ἐκδιώκτου· κατὰ χά ριν θυ · παραγγελία περὶ πίϲτεωϲ καὶ ἐυ ϲυνειδήτου διακονίαϲ· ἧϲ ἄν ευ κίν δυνοϲ· περὶ εὐχῆϲ, ὅτι ὑπὲρ πάντων· ὅτι παντα χοῦ ἀκάκωϲ ἀταράχωϲ· ϲεµνῶϲ·

>> Chapters of “to Timothy >> the first” a letter from Paul 1

2

3

4

Concerning, in order to love God, guidance, which does not need the force of the law. Concerning the election of him as an evangelist from being a persecutor, according to the grace of God. Instruction about faithful and with good conscience service, without which there is danger. Concerning prayer, that it is for all people, everywhere, without evil, quiet and solemn.

Fig. 1. Transcription and Translation of 1 Timothy Kephalaia (Paris, BnF Suppl. gr. 1074, 6v)

Located prior to the main text, these discrete textual units (i.e., tabulated kephalaia) supplement the reading of their corresponding text by abstracting key features of a paragraph into a brief summative statement. Such abstraction is not 38 Zacagnius, Collectanea, 529; Willard, Critical Study, 152; Blomkvist, Euthalian Traditions, 108. 39 Euthalius also demonstrates awareness and use of Eusebius later in the Prologue, as he defends his account of Paul’s stay in Rome on the basis of the authority of “Eusebius the chronographer” (εὐσεβίῳ τῷ χρονογράφῳ; Zacagnius, Collectanea, 534; Willard, Critical Study, 155; Blomkvist, Euthalian Traditions, 111.). 40 Zacagnius, Collectanea, 528; Blomkvist, Euthalian Traditions, 107.

168

Gregory Peter Fewster

random. Rather, it reflected the interests of the editor, being construed in principled and systematic ways through the repetition of key words or phrases. The practice of abstraction that produced the Euthalaian kephalaia sets them apart from the other prefatory paratextual materials included in the Euthalian Prologue. In this Prologue, Euthalius outlined Paul’s life and martyrdom, and Paul’s supposed intentions with respect to each letter. Narrative prefaces such as these could contextualize the letters of Paul. Euthalius alludes to this potential by bookending his epitome to Paul’s letters with descriptions of his express interest in publishing the letters in an annotated collection, i.e., politaeia (citizenship).41 Kephalaia could do the same, but more concisely. Whereas kephalaia – both the Euthalian kephalaia and those found in Gospel manuscripts – are discursive abstractions of the material of the main text, Eusebius reduced that abstraction to its minimal component, a numeral. But such reduction makes the tabulated material – Eusebius’ canons – nonsensical without the aid of sigla placed strategically in the margins of the main text, a feature that could also extend and specify the function of kephalaia.42 Marginal Sigla Origen’s innovative use of columns in a codex was not his only contribution to the type of textual scholarship that would characterise the Library at Caesarea. He also brought with him from Alexandria some of the text-critical techniques associated with the great Homeric scholars Zenodotus, Aristophanes, and Aristarchus. The Library at Alexandria fostered a criticism of Homeric and other texts that desired to preserve the integrity of the received manuscripts while making critical interventions into readings presumed to be corrupt. Critical techniques associated with Aristarchus in particular combined textual commentary, frequently in a supplementary handbook, with critical sigla placed in the margins of the edition of the Iliad, Odyssey, as well as Herodotus and possibly Plato.43 For the Alexandrian textual critics, marginal sigla, such as obeloi (÷ or –), asteriskoi (※), or the diplé (>), performed a variety of functions. They could indicate lines deemed spurious, lines repeated elsewhere, or lines whose language or style

41 Euthalius concludes the Prologue by writing, οὕτως ἡ πᾶσα βίβλος περιέχει παντεῖον εἶδος πολιτειῶν κατὰ προσαύξησιν (“Thus, the book encompasses the entire shape of citizenship, progressing towards it”; Zacagnius, Collectanea, 528; Blomkvist, Euthalian Traditions, 107.). 42 Curiously though, as Nordenfalk notes, in some codices (specifically the Latin Codex Brixianus) brief descriptions of the content of pericopae were added within the canon table. Such modifications extend the function of the canon apparatus, such that it describes pericopae like kephalaia do. See Nordenfalk, “Canon Tables on Papyrus,” 37–8 and plate 10. 43 Concerning the possibility that Aristarchus worked on the Platonic corpus, see Francesca Schironi, “Plato at Alexandria: Aristophanes, Aristarchus, and the ‘Philological Tradition’ of a Philosopher,” The Classical Quarterly 55/2 (2005): 423–34.

Finding Your Place

169

warranted comment, and thus acted as an essential mediator between the criticised text and corresponding textual commentary.44 It is challenging to provide definitive statements concerning the function and appearance of these signs, given that the majority of this evidence comes from few manuscript fragments or medieval scholia.45 However, a siglum’s ability to point the reader to a supplementary commentary remains a crucial component of this text-critical practice. Scholars are split on whether or not Origen included Alexandrian critical sigla in the Hexapla itself, insofar as the parallel columns of text would synoptically reveal textual variation between the versions, similar in that respect to Ammonius’ Gospel harmony. And marginal sigla are lacking in the few extant witnesses to the multi-column Hexapla (those noted above).46 There is also no evidence that Origen created a textual commentary, like those of the Alexandrians. In fact, this may have been unnecessary precisely because of the Hexapla’s synoptic format.47 However, there is literary evidence to support the claim that Origen did indeed employ Alexandrian critical sigla, including comments in his epistle to Julius Africanus (Ep. Afr., PG 11.60–61) and in his commentary on Matthew (Comm. Matt. 15.14; PG 13.1293).48 In both cases, Origen states that he marked with obeloi instances where the Greek text included an element lacking in the Hebrew, whereas asteriskoi would indicate that the material recorded on this manuscript was present in the Hebrew but was lacking in the Septuagint. It is crucial to note that in Comm. Matt. 15.14, Origen references other editions ambiguously (i.e., See discussion in Rudolf Pfeiffer, History of Classical Scholarship: From the Beginnings to the End of the Hellenistic Age (Oxford: Clarendon, 1968), 171–233; L. D. Reynolds and N. G. Wilson, Scribes and Scholars: A Guide to the Transmission of Greek and Latin Literature, 4 th ed. (Oxford: Oxford University Press, 2013), 9–11; Francesca Schironi, “The Ambiguity of Signs: Critical Σηµεια from Zenodotus to Origen,” in Homer and the Bible in the Eyes of Ancient Interpreters, ed. M. R. Niehoff (Leiden: Brill, 2012), 88–100. Schironi outlines several other sigla and their range of functions. 45 Schironi, “Ambiguity,” 87. 46 See discussion in Schaper, “First Column”; Natalio Fernández Marcos, The Septuagint in Context: Introduction to the Greek Version of the Bible (Leiden: Brill, 2000), 213–15; Grafton and Williams, Christianity and the Transformation of the Book, 88; Francesca Schironi, “P.Grenf. 1.5, Origen, and the Scriptorium of Caesarea,” BASP 52 (2015): 194–96. It is also worth noting that Eusebius’ description of the Hexapla makes no reference to critical sigla (Eusebius, Hist. eccl. 6.16; G. Bardy, Eusèbe de Césarée. Histoire Ecclésiastique, vol. 2, SC 41 [Paris: Éditions du Cerf, 1955], 109–11). 47 Schironi states that Origen provided no commentary. Schironi, “Ambiguity,” 104–105; Schironi, “P.Grenf. 1.5, Origen, and the Scriptorium of Caesarea,” 215. 48 Jerome describes Origen’s system similarly in Praef. in Pentateuch (PL 28.179) and Praef. in Paral. (PL 28.1393), in which he describes the obeloi and asteriskoi and Origen’s criticism in relation to the text of Theodotian (possibly a source for expansions from the Hebrew). 44

170

Gregory Peter Fewster

those used to illuminate textual variation); he does not make clear that these other editions made up the Hexapla. But it is clear, according to this description, that Origen does not record these markings on a Hexapla codex. Instead, obeloi and asteriskoi appear in several manuscripts (most notably P.Grenf. 1.5), which represent some sort Hexaplaric revision – an LXX with additions marked with obeloi and asteriskoi.49 Francesca Schironi considers this confluence of data as evidence that Origen was involved in producing this Hexaplaric revision.50 However, both T. D. Barnes and Peter Gentry defer that task to Eusebius and Pamphilus, whose editorial work (they claim) was responsible for the Hexaplaric text popular in Palestine and Syria.51 At the very least, Eusebius does indicate awareness presumably of Origen’s use of Alexandrian sigla in his comments on Psalms 66 and 72, where he states that passages were obelized because they were µὴ κείµενον δὲ ἐν τῷ Ἑβραῖκῷ (PG 23.673, 849).52 Origen’s use of critical sigla, stemming from Alexandrian critical scholarship, provides some genealogical parameters for the sigla employed in both Eusebian and Euthalian paratexts. However, there are limitations to a strong genealogical explanation. If Schironi’s hypothesis that it was Origen’s Hexaplaric expansion of the LXX that contained obeloi and asteriskoi is correct, then these sigla were fully-functional in isolation from any explanatory materials. In fact, she cites this as one of two marked advantages over the Alexandrian practice, the other being the reduced ambiguity of the meaning of the sigla.53 However, explicating This is argued forcefully by Schironi, pointing especially to Codex Colberto-Saravianus (G), Codex Marchalianus (Q), and most importantly P.Grenf. 1.5, as evidence for an expanded LXX with critical sigla. See Schironi, “Ambiguity,” 106–107; Schironi, “P.Grenf. 1.5, Origen, and the Scriptorium of Caesarea,” 181–223; followed also by Knust and Wasserman, To Cast the First Stone. It is important to note that the text of these manuscripts and the critical sigla are not identical (P.Grenf. 1.5 most closely resembles Origen’s practice), but this likely reflects further editorial alteration over the course of transmission. See also Peter Gentry’s work on the Hexaplaric signs in the Syriac Job tradition. Peter J. Gentry, The Asterisked Materials in the Greek Job, SCS 38 (Atlanta: Scholars, 1995). 50 Schironi, “P.Grenf. 1.5, Origen, and the Scriptorium of Caesarea,” 212–15. 51 See Barnes, Constantine and Eusebius, 95–8; Peter J. Gentry, “The Aristarchan Signs in the Textual Tradition of LXX Ecclesiastes,” in In the Footsteps of Sherlock Holmes: Studies in the Biblical Text in Honour of Anneli Aejmelaeus, ed. K. De Troyer, T. M. Law, and M. Liljeström (Leuven: Peeters, 2014), 463–78. Barnes’ argument relies considerably on colophons found in 2 Esdras and Esther in Codex Sinaiticus, which claim that their respective texts derive from a copy of Pamphilus, itself derivative of the Hexapla. However, there is reason to doubt the accuracy of these claims. David Parker is inclined to dispute their historical accuracy, cf. Parker, Codex Sinaiticus: The Story of the World’s Oldest Bible (London: British Library, 2010), 83. Indeed, it is worth noting that colophons claiming Caesarean provenance appear in the Euthalian tradition as well (Codex H and GA 88) and cannot both be true. See Harold S. Murphy, “On the Text of Codices H and 93,” JBL 78/3 (1959): 228– 37. 52 These, too, are noted in Barnes, Constantine and Eusebius, 98. 53 Schironi, “Ambiguity,” 107–108. 49

Finding Your Place

171

Origen’s own dependence on and transformation of Alexandrian critical practice adds analogical purchase to understanding the use of sigla in Eusebian and Euthalian systems, by showcasing a range of functions that marginal sigla could perform.54 Indeed, this range is evident already in late antique biblical codices, whose marginal notation operates on a binary between what I call referential and non-referential sigla. Non-referential sigla – that is, sigla that do not direct the reader outside the main text – are not uncommon in late antique biblical codices. Obeloi and asteriskoi – those used according to Origen’s scheme and represented in P.Grenf. 1.5 – are examples of non-referential sigla that perform text-critical functions. Codex Vaticanus (B03) and Codex Sinaiticus55 also contain marginal notation (in this case numerals) to divide into subsections the Pauline epistles and Acts, respectively.56 Reading aides, such as paragraphoi, could also be included in this category. Referential sigla, like those used in Alexandrian text-critical practice, are more common in late antique biblical codices. However, kephalaia systems and the Eusebian apparatus tend to use numerals as opposed to the Alexandrian sigla (obeloi, asteriskoi, etc.), a modification that allows sigla to indicate unit sequencing (a function of some non-referential numerals) alongside their cross-referential function. The ability for a siglum to facilitate cross-referencing is an important feature, deployed in a variety of ways and with varying degrees of complexity. Although Acts in Codex Sinaiticus contain marginal numerals, these numerals do not seem to correspond with the kephalaia in the upper margin.57 The resulting ambiguity showcases the significance of sigla for cross-referencing. By contrast, Codex Alexandrinus makes use of a double reference system in the 54 At some point, Alexandrian sigla made their way also into the criticism/transmission of the New Testament. And these have no obvious connection with Origen. Knust and Wasserman, To Cast the First Stone observe that Codex Basiliensis (E07) includes an asterisk at the Pericope Adulterae, which would also perform the liturgical function of prompting the reader to skip over the Pericope. Philip Payne also argues that a scribe of Codex Vaticanus occasionally used a “distigme obelos” to mark interpolated text. Philip B. Payne, “Vaticanus Distigme-Obelos Symbols Marking Added Text, Including 1 Corinthians 14.34–5,” NTS 63/4 (2017): 604–25. 55 Milne and Skeat incorrectly refer to the numerals in Acts in Codex Sinaiticus as “Euthalian.” H. J. M. Milne and T. C. Skeat, Scribes and Correctors of the Codex Sinaiticus (London: British Museum, 1938), 38. As Jongkind notes, any correspondence between the two systems likely derives from inherent sense divisions, rather than a genealogical relationship between the text-division systems themselves. Dirk Jongkind, Scribal Habits of Codex Sinaiticus (Piscataway: Gorgias, 2007), 122 n.135. 56 See discussions of these systems in Greg Goswell, “Ancient Patterns of Reading: The Subdivisions of the Acts of the Apostles in Codex Sinaiticus,” JGRChJ 7 (2010): 68–97; Goswell, “An Early Commentary on the Pauline Corpus: The Capitulation of Codex Vaticanus,” JGRChJ 8 (2011): 51–82. 57 See Jongkind, Scribal Habits, 122.

172

Gregory Peter Fewster

Gospels. A wedge-shaped paragraphos (in Matthew and Mark) and a crossshaped siglum (in Luke and John) in the side margins reference numerated kephalaia titles in the upper margin, which are also tabulated and prefixed to each Gospel. Numerals that begin each kephalaion enable cross-referencing between upper-marginal and prefixed iterations.58 The Eusebian system is also a complex one, whereby pericope numerals and canon numerals in the side margin facilitate cross-referencing with the canon tables and their contents. However, in many cases, such as in Codex Alexandrinus, a wedge-shaped siglum appears below the canon numeral to specify where on the page the particular pericope begins.59 Finally, Coislinianus includes a very elegant system, at least as far as the crossreferencing function goes. Side margins include numerals, which correspond to numerated kephalaia titles, tabulated and prefixed to each of Paul’s epistles. Both the Eusebian and Euthalian apparatus are imperfectly numerated in their manuscript witnesses. Codex Coislinianus is severely defective, noticeable even in its fragmentary state. For instance, marginal numeration frequently does not appear where we would expect. This is the case when our knowledge of the location of a chapter division comes from later manuscripts. For example, we would expect a ιη at Heb 12:12 (BnF Suppl. gr. 1074, 5r), or an ε at 1 Tim 2:11 (BnF Suppl. 1074, 6v).60 Alternatively, this expectation is created by extant kephalaia tables in Coislinianus, whereby extant pages lack the corresponding numeral. For example, we would expect to read ι at Gal 5:2 (BnF Suppl. gr. 1074, 1r),61 or a β at 1 Tim 1:12 (BnF Suppl. gr. 1074, 7v, with corresponding kephalaia at BnF Suppl. gr. 1074, 6v). Another kind of error, misplacement, occurs in Coislinianus, though off only by a single verse. The γ expected at 2 Tim 2:3 is located at 2:4 (Turin, B.I.v., 2r) and the γ expected at Tit 2:1 is located at 2:2 (BnF Coislin 202, 12r).62 58 See discussion of these features in Greg Goswell, “Early Readers of the Gospels: The Kephalaia and Titloi of Codex Alexandrinus,” JGRChJ 6 (2009): 134–74. 59 Andrew Smith notes that this combination of features is not always consistent; there are a few instances where the paragraphos is absent. Smith, Study of the Gospels, 143. On the other hand, section beginnings are frequently marked with two parallel dots in Codex Sinaiticus, as Jongkind has observed. Jongkind, Scribal Habits, 112. The use of paragraphos with numerals is also present in the marginal numerals of Acts in Codex Sinaiticus. See Milne and Skeat, Scribes and Correctors, 38. 60 For a reconstructed text of the kephalaia and their expected location, see Blomkvist, Euthalian Traditions, 55, 59. 61 The kephalaia table is located of fol. 7r on the Athos fragments. See transcriptions in L. Duchesne, “Fragments des Epîtres de S. Paul,” Archives des Missions Scientifiques et Littéraires 3/3 (1876): 427; M. H. Omont, Notice sur un très ancien Manuscrit Grec en Onciales des Épitres de Saint Paul, Conserve à la Bibliothèque Nationale (H ad Epistulas Pauli) (Paris: Imprimerie Nationale, 1889), 22; Kirsopp Lake, Facsimiles of the Athos Fragments of Codex H of the Pauline Epistles (Oxford: Clarendon, 1905), plate 13. 62 For a reconstructed text of the kephalaia and their expected location, see Blomkvist, Euthalian Traditions, 56–7.

Finding Your Place

173

Similar problems exist in the early witnesses to the Eusebian apparatus. Both Dirk Jongkind and Andrew Smith have identified transmission errors of the Eusebian apparatus in Codex Sinaiticus and Codex Alexandrinus, respectively. Like Coislinianus, Sinaiticus lacks marginal numeral-sets in 16 locations, according to Jongkind’s count.63 Errors of misplacement also occur in Sinaiticus, some of which are brought on by the fact that it has multiple columns or because the scribe opted to leave space between Eusebian sections that would otherwise be virtually on top of one another.64 In Alexandrinus, Smith describes what he calls “cascading errors,” where an error of omitting or duplicating a canon numeral, a section numeral, or a numeral-set causes errors to multiply until rectified by the scribe.65 With the inclusion of multiple paratextual systems in a single codex, individual features also compete for space in the margins. By placing kephalaia in the top margin, this problem is solved, as is the case for the competing chapter numerals and kephalaia in Acts of Codex Sinaiticus. But the kephalaia in the Gospels of Codex Alexandrinus utilise marginal sigla, which puts them in direct competition with the marginal Eusebian sigla. For example, kephalaion µ of Mark, represented by a wedge-shaped paragraphos, is tucked in between two ekthetic capitals, moved to the right apparently to leave space for two Eusebian numeral-sets that partially overlap (8v).66 A far more severe instance of competition occurs in Coislinianus at Heb 10:5, where notation for another cross-referencing system of the Euthalian apparatus – the Divine Testimonies, which indicates a citation (in this case to the Psalms) – was preferred to the kephalaia numeral, which was omitted.67 Though early witnesses to the apparatus of Eusebius and that of Euthalius contain more or less significant errors, their mistakes fit the trend of the transmission of other contemporaneous systems. The variation and competition of marginal sigla in the Euthalian and Eusebian systems raise significant questions about copying practices of marginal sigla, especially insofar as they facilitate cross-referencing. Competition between marginal material, in particular, speaks to the challenges faced by late antique editors and scribes who wished to combine formerly-distinct paratextual systems into a Readers may consult Jongkind, Scribal Habits, 281 (appendix III). See Jongkind, Scribal Habits, 112–19. 65 Smith, Study of the Gospels, 143–56. Similar errors also appear in Codex Sinaiticus. Jongkind describes quite a variety of errors of sequence, misplacement (sometimes in the wrong column!), or omission, usually rectified quickly. In at least one case, similar to Alexandrinus, the error appears to have been present in the exemplar. For details on these errors, see Jongkind, Scribal Habits, 113–15. 66 This overlap may be due to the scribe writing out the Eusebian section numerals prior to the canon numerals, as Smith suggests. Smith, Study of the Gospels, 153–55. 67 Note that Omont does not record this divine testimony in his edition. Omont, Épitres de Saint Paul, 38. For discussion on the Divine Testimonies, see Robinson, Euthaliana, 17– 20; Willard, Critical Study, 29–46. 63 64

174

Gregory Peter Fewster

single bound codex. This practice of combination is well-attested. Examples include the two kephalaia systems and the Eusebian material in the Gospels of Codex Alexandrinus,68 or the combination of Euthalian kephalaia and hypotheses/argumenta in medieval manuscripts of the Pauline epistles.69 With respect to divisions in the Gospels, these late antique scribes benefitted from variations in earlier editors’ interpretation of paragraph and pericope divisions. Variation and competition of marginal sigla also reveal the significance of those numerals and reading practices associated with them. Errors of omission, sequence, and mismatch (in the case of the Eusebian apparatus) render the crossreferencing utility of these systems nonsensical, though it does not foreclose other uses for the paratextual materials. A mismatch in the combination of Eusebian section and canon numeral eliminates the primary function of the apparatus – to help readers navigate Gospel parallels through schematic mapping.70 Omission creates similar issues. The missing numeral-set at Mark 6:47 in Codex Sinaiticus (221r, noted by Jongkind) would be bewildering to the reader guided there after coming across the parallel marked at Matt 14:24 (208r). Omissions of marginal numerals are also difficult to explain for Codex Coislinianus, not least because of its fragmentary state. Combinations of kephalaia tables and the marginal numerals appear to have been inscribed by two different hands. For example, one scribe rendered kephalaia tables or marginal numerals in Galatians, Hebrews, and Titus with no decoration, whereas another scribe included decorations to kephalaia tables or numerals of Colossians and 1 Timothy. The appearance of the paratexts of Titus suggests that the scribe of a given kephalaia table also inscribed the numerals of the same epistle. Numerals were thus unlikely to be a scribal afterthought and thus may reflect errors in the exemplar. Such errors eliminate cross-referencing opportunities. But at least in the case of the kephalaia, the value of the tabulated material as a summary of the contents is maintained. Errors of misplacement are less grievous and may indicate that precise placement was not required for effective cross-referencing. In most of the examples of misplaced marginal sigla, the difference is not great. By virtue of the function of both Eusebian and Euthalian systems, readers could easily infer the text being referenced, insofar as parallel Gospel material would partially direct the reader and the main text of the Pauline epistles would be summarized by the kephalaia. Physical Relationship The question of the physical relationship of paratextual features to the main text takes us beyond the isolated phenomena of marginal sigla and tabulation to issues of modes of navigation. How do paratextual features combine so as to function See Smith, Study of the Gospels, 156–61. That these materials reflect to distinct editions is argued by Dahl, “The ‘Euthalian Apparatus’ and Affiliated Argumenta,” 244–45, 253–57. 70 Jeremiah Coogan underscores this function effectively (“Mapping,” 337–57). 68 69

Finding Your Place

175

as thresholds for engagement with a book-object? Both tabulation and marginal sigla have the opportunity to work on their own. Origen’s Hexapla, Ammonius’ Diatessaron Gospel, Eusebius’ Chronicon, and possibly the kephalaia of the “Christ-loving father” referenced in the Euthalian Prologue employed strategies of tabulation to great effect. Additionally, Origen’s Hexaplaric sigla and some of the marginal numerals present in Codex Vaticanus and Sinaiticus (among other contemporaneous codices) perform distinctive and recognizable functions as non-referential sigla. However, the combination of these two features together mediates very particular relationships with the main text, a combination that clearly distinguishes the Eusebian and Euthalian apparatuses from other contemporaneous paratextual systems. Eusebius’ description of his apparatus in the Epistula ad Carpianum flags the possibility of cross-referencing, whereby marginal numeral-sets associated with the pericope of one Gospel draws the reader to the canons, and from there to a parallel pericope in another Gospel. Martin Wallraff seizes upon the physical implications of Eusebius’ innovation, suggesting that it makes full use of the possibilities afforded by the codex format and even requires a physical apparatus (perhaps a finger) as a placeholder in order for the reader to flip from one page to another.71 The physicality of reading, according to Wallraff’s account, underwent a notable shift with the Eusebian apparatus, particularly if we imagine how this would have worked with a book-roll. Navigating Gospel parallels would require endless unrolling and rerolling. Wallraff’s argument, however, becomes even more pointed when considered alongside a cross-referencing technology already discussed: Alexandrian critical philology. When Aristarchus was undertaking his critical work on Homer, his available book-object was a roll. And his critical sigla directed readers to a separate commentary in order to understand what error the siglum represented.72 In this case, sigla in the Aristarchan and Eusebian mode both direct the reader outside the text-object being read. Reading as consultation is possible with the roll.73

Wallraff, Kodex und Kanon, 30–3. See discussion in Pfeiffer, History of Classical Scholarship, 218. Additionally, Shironi provides a visual mock-up of what these two texts, edition with sigla and supplementary commentary may have looked like. Schironi, “Ambiguity,” 92–5. It is important to note, however, that Alexandrian scholars did not exclusively place text-critical commentary in a separate book, as Reynolds and Wilson, Scribes and Scholars, 247 note. However, there is evidence for an Aristarchan commentary on Herodotus in a fragment from the third century CE (P. Amh. 2.12). See Bernard P. Grenfell and Arthur S. Hunt, The Amherst Papyri: Being an Account of the Greek Papyri in the Collection of Lord Amherst of Hackney. Part II Classical Fragments and Documents of the Ptolemaic, Roman and Byzantine Periods (London: H. Frowde, 1901), 3–4. 73 Here I depart from Wallraff, Kodex und Kanon, 32 who suggests that the consultation was a new mode of reading enabled by the combination of codex and canon (“Das neue 71 72

176

Gregory Peter Fewster

What distinguishes these two modes of reading then, besides the book-technology itself, is the fact that the Eusebian system requires multiple steps. The cross-referencing function of the Eusebian apparatus works effectively as far as its description in the Epistula ad Carpianum goes, and indeed supports the notion of important shifts in modes of engaging book-objects. But the manuscript record challenges this somewhat. One of the noticeable features of the early manuscripts containing Eusebian material is the presence only of marginal numeralsets.74 Neither canon tables nor the Epistula ad Carpianum are prefixed to Greek manuscripts until the sixth century in Codex Purpureus Rossanensis (Σ042), though versional evidence witnesses to its appearance in the fifth.75 Examples include the Latin Codex Harliensis (BL Harley MS 1775) or the Ge‘ez Garima I.76 The Gothic Codex Argenteus and the Latin Codex Brixianus represent a strange variation on the standard prefixing of the canons and Epistula, where truncated canons are inscribed in lower margins of the codex.77 It may be significant that those manuscripts that prefix the important prefatory material of the Eusebian apparatus are Gospel codices. It is difficult to know what to make of this curious state of the evidence. Eusebius’ description of the canon system in the Epistula clearly indicates the essential relationship between both tables and numeral-sets, not to mention that the Epistula itself is needed to make sense of the otherwise opaque paratextual features. The fact that Codex Sinaiticus in particular lacks the prefatory materials is curious, given its connection with the library at Caesarea in general and Eusebius himself in particular, as is alleged by some scholars.78 It is hard to imagine that Medium generiert ein neues Leseverhalten: nicht nur kontinuierliches, sondern auch konsultierendes Lesen”). 74 Jongkind thinks that the Eusebian canon tables were probably never present. Jongkind, Scribal Habits, 110. 75 Another roughly-contemporaneous witness to the prefatory material is MMA X.455, fragments only of the Epistula ad Carpianum and canons, at the Metropolitan Museum of Art. The Met lists the date of these fragments somewhere between the fourth and seventh century (“Papyrus Fragments of the Canon Tables,” http://www.metmuseum.org/art/collection/search/474440, accessed 5 March 2018). The fragments’ association with Epiphanius seems to place it the late sixth century. See H. G. Evelyn White, “Greek Ostraca and Papyri,” in The Monastery of Epiphanius at Thebes, Part II, ed. W. E. Crum (New York: Publications of the Metropolitan Museum of Art Egyptian Expedition, 1926), 302–305; Nordenfalk, “Canon Tables on Papyrus,” 30–1. 76 On the Latin tradition, see H. A. G. Houghton, The Latin New Testament: A Guide to Its Early History, Texts, and Manuscripts (Oxford: Oxford University Press, 2016), 200– 202. On the Eusebian material in the Garima Gospels, see essays in Judith S. McKenzie and Francis Watson, eds., The Garima Gospels: Early Illuminated Gospel Books from Ethiopia (Oxford: Manar al-Athar, 2016). 77 See plates in Nordenfalk, Die spätantiken kanontafeln, taf. 150, a and b. 78 See for example Milne and Skeat, Scribes and Correctors, 66–9; T. C. Skeat, “The Codex Sinaiticus and the Codex Vaticanus and Constantine [1999],” in The Collected

Finding Your Place

177

Eusebius would authorize the production of a biblical codex that imperfectly reproduced his apparatus, perhaps mitigating against this connection between Sinaiticus and Eusebius in the first place.79 In any case, that the codex is likely such a close contemporary of Eusebius continues to render the absence of the prefatory material conspicuous. Once again, Alexandrian textual criticism provides, through analogy, one possible solution. The absence of Eusebius’ prefatory material in codices containing their corresponding numeral-sets could be explained by the existence of that material in a pamphlet that was exterior to and thus supplementary to the Gospel codex itself, analogous to the separate commentary of the Alexandrian textual critics. In such a scenario, Sinaiticus, Alexandrinus, and some other early biblical codices lack the canons and Epistula precisely because their exemplar did as well. Fragments of the Epistula ad Carpianum or canons, such as MMA X.455, may attest to this prefatory material as a supplementary pamphlet, but they could just as easily be fragments of the initial pages of a Gospel codex. Analogical comparison has its limits in the absence of sufficient material evidence. On the other hand, the existence of supplementary paratextual book-objects in the case of the Eusebian apparatus provides interesting context for Euthalius’ description of the kephalaia in his Prologue to the Epistles of Paul. As noted above, concerning the kephalaia adapted from the “Christ-loving father,” Euthalius specifically indicates that he prefixed (προτάξοµεν) kephalaia to each epistle.80 It is difficult to take this reference any other way than that these kephalaia were materially prefixed, since Euthalius pictures each epistle as receiving a set of summaries. And Coislinianus attests to this material prefixing in partially-extant kephalaia tables for Galatians (Mt. Athos, 7r), Hebrews (BnF Coislin 202, 5r–v), 1 Timothy (BnF suppl. gr. 1074, 6v; BnF Coislin 202, 10r– v), and Titus (Bnf Coislin 202, 11r), as does Codex Alexandrinus with respect to Gospel kephalaia (5v [Mark], 19r–v [Luke], 42r [John]). The benefit of prefixing, in contrast to the Alexandrian practice of producing a supplementary commentary, is the assured presence of paratextual material. As Wallraff hypothesized with respect to Gospel codices containing the Eusebian apparatus, readers of a Euthalian edition of Paul would similarly encounter it as Biblical Writings of T. C. Skeat, NTSup 113 ed. J. K. Elliott (Leiden: Brill, 2004), 193–237. Even Christfried Böttrich, who doubts that the codex resulted from Constantine’s edict, considers it to have been a master copy at the Caesarean Christian scriptorium. Böttrich, “Codex Sinaiticus and the Use of Manuscripts in the Early Church,” ExpTim 128/10 (2017): 470– 75. 79 Nordenfalk in fact argues that Eusebius would have maintained a master copy of a Gospel book with canons at the library, serving as an exemplar for the production of additional Gospel books by Caesarean scribes and the archetype for the canons in all manuscripts. He also suggests that the fifty books ordered by Constantine may have contained the canon system. Nordenfalk, Die spätantiken kanontafeln, 1.49–50. 80 Zacagnius, Collectanea, 528; Blomkvist, Euthalian Traditions, 107.

178

Gregory Peter Fewster

a book-object to be consulted, using a finger or another object to vacillate between prefixed kephalaia table and the page signalled by its corresponding marginal numeral. In actual practice, however, this could be laborious. For marginal numerals were not always located with precision or recorded with accuracy. But in spite of the opportunity for error in two cross-referencing systems that made use of tabulated prefatory material and marginal numerical sigla, the Euthalian apparatus and Euthalian apparatus, became almost ubiquitous in late antique and medieval biblical codices.

Conclusion The fourth century marks an important period in the history of the book, during which, scholarly innovators like Eusebius and Euthalius constructed new ways of representing information, and consequently, of materially navigating bookobjects. However, both cross-referencing paratextual systems benefitted from practices of textual scholarship and modes of book production that preceded them. Modern textual scholars routinely make note of some of these genealogical debts, but this essay has pushed that discussion further, through attention to how the Eusebian and Euthalian systems materially combined distinctive traditions of tabulation and referential sigla. The greatest challenge to producing an account of the emergence of paratextual cross-referencing systems is the state of the manuscript evidence. Although textual scholars and book historians are well-acquainted with the challenges of textual reconstruction, necessitated by textual variation and fragmentary witnesses, the temptation to idealize critical reconstructions of these paratextual systems, and consequently their purported originators, is great. That scholars have not subjected the paratexts to the same degree of critical scrutiny as the texts they garnish contributes to the reliance on the reconstructed texts; their idealizing becomes a methodological necessity.81 This essay took, as a methodological dictum, a preference for early manuscript evidence over and above critical editions of the Eusebian canons and the Euthalian apparatus. And while the reconstructed texts of the Epistula ad Carpianum and Euthalius’ Prologue to the corpus Paulinum find considerable support in their earliest witnesses, the variants and absent associated paratextual materials raise questions concerning how those texts materially manifested in their earliest phases of production and how effectively they accomplished their creator’s expressed goals. In order to answer these lingering questions, and to prompt new ones, improved knowledge of the manuscript evidence (both Greek and versional) is a necessity. 81 This is in the process of being rectified with the Paratexts of the Bible project. See an introduction to this project in Martin Wallraff and Patrick Andrist, “Paratexts of the Bible,” Early Christianity 6 (2015): 237–43.

Finding Your Place

179

The prospect of improved knowledge of biblical paratexts opens productive avenues for future scholarship on the social and material history of New Testament texts. The relationship between the transmission of New Testament texts and the transmission of associated paratextual systems remains an open question. Harold Murphy’s comparison of the text of Codex Coislinianus and GA 88 showed significant textual divergence, in spite of sharing the Euthalian apparatus, including a colophon attesting to correction against a Pamphilian codex of the corpus Paulinum.82 The extent to which this type of divergence exists with respect to other Euthalian codices is not presently clear, nor indeed how revisions of the text of the corpus Paulinum or the Gospels extended to their paratexts. But the vagaries of these relationships hold potential value for how we account for the textual history of the corpus Paulinum or the fourfold Gospel and the functional relationship between text and paratext. In addition, the Eusebian and Euthalian materials (and other biblical paratexts for that matter) tell us some things about the production, reception, and material engagement by readers of Gospel books or Pauline codices in late antique Caesarea under Origen, Eusebius, Pamphilus, and Euthalius. But the manifestation of these paratexts in other late antique codices from Egypt and Ethiopia to northern Italy and further on to the Greek east has the potential to shed important light on the production, reception, and material engagement by readers in those particular locales. So too with these manuscripts’ medieval descendants. If textual scholarship on the New Testament desires to produce comprehensible histories of these texts, their paratextual components cannot be ignored or treated as a secondary concern. But even if we understand on a theoretical level that the relationship between text and paratext is significant, this essay has shown that idealized representations of certain paratextual systems break down upon the analysis of individual manuscripts, while telling us more about how these particular manuscripts were produced and could have been used. Insofar as manuscripts of the New Testament number in the thousands, there is much work to be done.

82 See Murphy, “On the text of Codices H and 93,” 228–37. Note that his use of the designation Codex 93 uses an older system, and he indeed studied what is now known as GA 88.

Papyrology, Papyrological Commentary, and the Future of New Testament Textual Scholarship Christina M. Kreinecker This article looks at the future of textual scholarship on the Bible from a papyrological viewpoint, focussing on documentary papyri. It starts at Herman C. Hoskier’s times with the beginnings of papyrology in the nineteenth century (1), discusses major developments in the field from the present (2), and looks ahead into the future (3), identifying challenges and opportunities for biblical studies and biblical scholars. The article is a reflection on how documentary papyri and papyrology as a discipline contribute to New Testament studies and their future.

1. The Beginnings of Papyrology and the Discovery of its Significance for Biblical Studies 1.1. Excavating the Past: The Discovery of Papyri Although the middle of the eighteenth century had already seen the discovery of about 800 ancient papyrus rolls at Herculaneum in Italy, it was not until a hundred years later that the first collectors of papyri emerged. The initial interest, insofar as it went, was in literary texts, “complete texts, archives, or whole rolls.”1 When Egyptian farmers expanded their cultivated land in the second half of the nineteenth century, major finds were made in the Fayyûm with its ancient capital Arsinoe, but also in Heracleopolis and Hermopolis. The last two decades of that century were marked by an enormous hunt for papyri with little respect to languages, genres, and actual places of discovery. Sensational literary findings like Aristotle’s Constitution of Athens and apocryphal writings like the Gospel of Peter, the Apocalypse of Peter, and the Book of Enoch, resulted in an increased

1 Eric G. Turner, Greek Papyri: An Introduction (Oxford: Oxford University Press, 1980), 21. Among the earliest findings of literary texts are two rolls of Hyperides, a poem of Alcman, and manuscripts containing Homer. Among the earliest documentary collections are Ulrich Wilcken’s two volumes of Ptolemaic documents (UPZ, 1927 onwards).

182

Christina M. Kreinecker

interest in texts preserved on papyri. Scholars, including notably Carl Wessely, Ulrich Wilcken, Frederic Kenyon, and Ernest Wallis Budge, rose to the new academic challenges presented by palaeography, preservation, and other issues. New series for publishing the recently discovered material were also established.2 Major papyrus collections of today had their beginnings during Hoskier’s life. The world’s largest collection in Oxford, for example, was supported by the newly established Egypt Exploration Society (1882), a non-profit organisation based on a crowed-funding principle with private and public donors. The society’s most famous excavators, Bernard P. Grenfell and Arthur S. Hunt, discovered thousands of Greek and Latin papyri from Graeco-Roman times in the Fayyûm and, in particular, the ancient city of Oxyrhynchus. After their systematic search for sites of Graeco-Roman villages during the late nineteenth century, their efforts were crowned with success at rubbish dumps near Oxyrhynchus. They published the first volume of the newly established series P.Oxy. in 1898; this still maintains its now classic layout of literary texts first and documentary texts second.3 Other places and scholars joined the game and the excitement of the new discoveries resulted in a multitude of publications. Having a papyrus from Egypt became a status symbol for the wealthy and important. Financial support for excavations were therefore given, for example, by the Austrian Archduke Rainer whose papyrus purchases became the foundation of today’s second largest papyrus collection in the world, which is part of the Austrian National Library in Vienna. Altogether, the period of the late nineteenth and early twentieth centuries up to the First World War saw many exciting initial findings. Texts of which little or nothing was known were found, and various scholarly disciplines took up the challenge.4 New methodologies and scholarly tools were developed to evaluate the discovered material. In 1899 Frederic G. Kenyon published his Palaeography of Greek Papyri. In 1901 Ulrich Wilcken started the Archiv für Papyrusforschung, the first journal dedicated to papyrological issues. Wilcken and Ludwig Mitteis’ Grundzüge und Chrestomathie der Papyruskunde (1912) gave a first outline of the value of papyri for the scientific world, in particular from a historical and juridical perspective. From 1913 onwards Friedrich Preisigke published a collection of corrections, alternative readings, and new insights on

1892 brought the Berliner Griechische Urkunden (BGU) and 1893 the Catalogue of Greek Papyri in the British Museum. In 1895 the Corpus Papyrorum Raineri (CPR) in Vienna followed, and in 1896 Les Papyrus de Genève (P.Gen.). B. P. Grenfell and A. S. Hunt started the Oxford Papyri series (P.Oxy.) in 1898. 3 Turner, Greek Papyri, 26–7. 4 Turner, Greek Papyri, 17–53; Hélène Cuvigny, “The Finds of Papyri: The Archaeology of Papyrology,” in The Oxford Handbook of Papyrology, trans. A. Bülow-Jacobsen, ed. R. S. Bagnall (Oxford: Oxford University Press, 2009), 59–78. 2

Papyrology

183

already published material as Berichtigungsliste.5 In 1915 he also began to publish various Greek dictionaries based on the papyrological evidence.6 In the same year he established the Sammelbuch (SB), a collection of individually published papyri in various journals, thereby assuring easy access to these dispersed publications. At the very latest, the year 1919 marks the formation of papyrology as a new discipline with Schubart’s publication of his introductory work Einführung in die Papyruskunde. 1.2. Documentary Papyri Papyri are traditionally classified into two categories: literary and documentary papyri, texts with characteristics of both are called semi-literary papyri. Both categories are relevant text sources for New Testament studies.7 It is obvious that literary papyri, particularly those containing biblical writings, apocrypha – some of them only known by title, some of them unknown to the modern world until their discovery on papyrus – and theological treatises are of utmost value for ancient history, classics, theology and, most of all, biblical studies and ecclesiastical history. However, documentary papyri, upon which this article focuses, are also of high relevance to biblical studies and early Christianity. This is not only true for documentary papyri with a Christian background but also for documents without any direct Christian relationships. 1.2.1. Documents Pointing to a Christian Background The most obvious benefit for New Testament studies and early Christianity comes from those documentary papyri that show a Christian background. These texts may contain information about the daily activities of Christians, but also about the development of the early church, its structure and its administration. Studying these texts allows observation on the development of the church through the centuries. 5 The latest volume is F. A. J. Hoogendijk and A. Jördens, eds., Berichtigungsliste der Griechischen Papyrusurkunden aus Ägypten, vol. 13, collected by J. M. S. Cowey und F. A. J. Hoogendijk (Leiden: Brill, 2017). 6 Cf. Friedrich Preisigke, Wörterbuch der griechischen Papyrusurkunden, mit Einschluß der griechischen Inschriften, Aufschriften, Ostraka, Mumienschilder usw. aus Ägypten, 2 vols. (Berlin: Selbstverlag der Erben, 1925–1927); Preisigke, Besondere Wörterliste, vol. 3 of Wörterbuch der griechischen Papyrusurkunden, mit Einschluß der griechischen Inschriften, Aufschriften, Ostraka, Mumienschilder usw. aus Ägypten, ed. E. Kießling (Berlin: Selbstverlag der Erben, 1931); Emil Kießling, Wörterbuch der griechischen Papyrusurkunden, mit Einschluß der griechischen Inschriften, Aufschriften, Ostraka, Mumienschilder usw. aus Ägypten, vol. 4 (Berlin: Selbstverlag der Erben, 1944) (repr. Wiesbaden: Harrassowitz, 1993). 7 For a general overview see Christina M. Kreinecker, “Papyrologie (NT),” in Das wissenschaftliches Bibellexikon im Internet (www.bibelwissenschaft.de/stichwort/52056/).

184

Christina M. Kreinecker

Deciding whether a papyrus text is of Christian provenance is challenging. Above all it cannot be stressed enough that there are no Christian documentary papyri, or, as Roger Bagnall has phrased it when talking about letters, “There are no ‘Christian letters’…only letters written by Christians, to Christians, or both. It is not helpful to analytic clarity for scholars to ‘baptize’ letters.”8 Nevertheless, can one determine when a text was produced by a Christian? Certainty in this regard is more easily obtained in later rather than in earlier centuries when names of bishops or correspondence between churches and ecclesiastical communities are explicit or traceable. In order to decide whether a person believing in Jesus Christ has written a text certain features have been identified to make such an authorship “more likely.” Lincoln Blumell has brought together a useful list of decisive markers,9 among which are the following: (1) Rare are those instances where the letter writer identifies himself or herself as a Christian, only Heras in P.Oxy. XLIII 3149 from the fifth century CE seems to fall in that category. It is of little surprise that in an everyday correspondence it does not seem necessary to mention the fact that one is a Christian when writing to someone familiar who would know the sender personally anyway. (2) Texts in which Christians are mentioned can be found from the third century CE onwards. However, the fact that someone is a Christian is usually not the point of interest, it is rather clarifying the identity of a person. One example is SB XVI 12497 (first half of the third century CE), a list of names of people who are on public service. Among other names this list contains a certain Antonius Dioskorus who is specified as being a Christian. Other early examples are the business letter SB VI 9557 (264 or 282 CE) which mentions Christian clerics or the summons of the Christian Petosorapis, P.Oxy. XLII 3035 (256 CE). (3) From the fourth century CE onwards correspondence among clerics, in particular bishops, of various sorts can be found, e.g. accounts, letters to monks asking for prayers, healing or other support. (4) Biblical names and quotations may indicate a Christian background. Most of the time, however, they could also point to a Jewish (unless from New Testament writings) or Manichean background. (5) The same hesitation when it comes to biblical names is also true for monotheistic expressions, certain terms and symbols typically but not exclusively used in Christian circles. If, for example, nomina sacra are present, the likelihood of a Christian background rises, though even these markers cannot give absolute proof. Likewise not every decorative cross present on a papyrus hints at a Christian background. 8 Roger S. Bagnall, “Review of Malcolm Choat, Belief and Cult in Fourth-Century Papyri,” BASP 43 (2006): 205–209, here 207. 9 Lincoln H. Blumell, Lettered Christians: Christians, Letters, and Late Antique Oxyrhynchus, NTTSD 39 (Leiden: Brill, 2012), 36–88.

Papyrology

185

Therefore, declaring a letter to be of Christian authorship is usually the result of more than one identification marker being present in the text. However, even after establishing a document’s probable Christian background, the question still remains as to the significance of this identification: are these texts different from others in terms of genre, language, style, general concerns etc.? This is not the case. On the contrary, documentary papyri written by Christians have the same language, the same context and deal with the same matters as many thousands of papyrus texts deriving from a non-Christian background at the same time and from the same area: business letters, personal letters, petitions etc. However, this does not diminish the value of these texts for historical, religio-historical, social, geographical and many other studies and observations. It draws the attention to the larger spectrum of documentary papyri in general and non-Christian documentary papyri in particular. 1.2.2. Documents Pointing to a Non-Christian Background Among the first to point out the potential of documentary papyri for New Testament studies on a large scale was a contemporary of Herman Hoskier (1864– 1938), the German theologian and classicist Adolf Deißmann (1866–1937). His life and contribution to both the scientific community and politics is the focus of Albrecht Gerber’s monograph Deissmann the Philologist.10 Gerber covers the period of Deißmann’s life from when he was discovering and analysing papyri more theologically during his work as a Greek lexicographer, through to his philological investigations of the Greek Bible (Septuagint and New Testament). He focuses on Deißmann’s study tours to the East (including Greece, the Levant, and Egypt) as well as his contributions to the Ephesus excavations. Deißmann’s political and ecumenical activities earned him two nominations for the Noble Peace Prize in 1929 and 1930. He became the Rektor of the Berlin University (1930/31) and was awarded six honorary doctorates,11 four from Britain, one from the United States and one from Sweden: Aberdeen (1906), St. Andrews (1911), Manchester (1912), Oxford (1929), Wooster College, Ohio (1929), and Uppsala (1932). Deißmann’s “influential trilogy (Bibelstudien, Neue Bibelstudien and Licht vom Osten) changed the course of biblical studies everywhere and, perhaps more importantly, established his contention that the linguistic history of postclassical Greek needed to be rewritten, and to be set firmly in its socio-politically correct historical context.”12 His books Bibelstudien (1895) and Neue Bibelstudien (1897) were published in English as one volume entitled Bible Studies (1901, 2nd ed. 1909). Summarising the “revolutionised” perspective of “the specialists” of

Albrecht Gerber, Deissmann the Philologist, BZNW 171 (Berlin: De Gruyter, 2010). Cf. Gerber, Deissmann the Philologist, 228, 236, 319, 333, 338, 351 and 369. 12 Gerber, Deissmann the Philologist, 123. 10 11

186

Christina M. Kreinecker

his times, he stresses that “the starting-point for the philological investigation of the New Testament must be the language of the non-literary papyri, ostraca, and inscriptions.”13 Deißmann showed that the language of the New Testament is not peculiar or a unique language but, on the contrary, part of the spectrum of Koiné Greek at the time. Slightly polemically he insists: There was a time when the Greek of the New Testament was looked upon as the genuinely classical; it was supposed that the Holy Spirit, using the Apostles merely as a pen, could not but clothe His thoughts in the most worthy garb. That time is past…Paul did not speak the language of the Homeric poems or of the tragedians and Demosthenes, any more than Luther that of the Nibelungen-Lied.14

Deißmann proves that the Greek language of biblical writings is in fact Koiné Greek. Referring to the language of the New Testament as “Biblical Greek” as if it was a separate language is therefore not only misleading, but historically and linguistically inadequate. In his monograph Light from the Ancient East, Deißmann gives three dimensions in which documentary papyri are of significance for biblical studies: linguistically, literary-historically and socio-historically. On a linguistic level,15 Deißmann takes his observations from his volume Bible Studies a step further by giving examples of how the syntax, semantics, and style of the Koiné Greek used in documentary papyri are similar and therefore comparable to New Testament writings. On a literary-historical level,16 documentary papyri offer contemporary reference to genres and authorial intentions. Due to thousands of preserved papyrus letters, it is first and foremost the letter genre about which papyri hold significant information concerning, for example, letter conventions, style, and formulas. On a socio-historical and also religio-historical level,17 papyri provide contemporary insights in the actual ancient world and its economy, law, agriculture, daily deeds, and all sorts of individual topics relevant for the Bible, for example, slavery, family, women, associations etc. This information is of particular value as it gives evidence from actual happenings, which can deviate from the theory or more idealised reality presented in literary texts or legal decrees. However, some caution is required due to the fragmentary picture that documentary evidence provides of the past. It cannot be known for sure whether a

13 Adolf Deissmann, Light from the Ancient East: The New Testament Illustrated by Recently Discovered Texts of the Graeco-Roman World, new and completely rev. ed. with eighty-five illustrations from the latest German edition, trans. L. R. M. Strachan (London: Hodder and Stoughton, 1927) (repr. Peabody: Hendrickson, 1995), 63. 14 Adolf Deissmann, Bible Studies: Contributions Chiefly from Papyri and Inscriptions to the History of the Language, the Literature, and the Religion of Hellenistic Judaism and Primitive Christianity, trans. A. Grieve, 2nd ed. (Edinburgh: T&T Clark, 1909), 63. 15 Deissmann, Light from the Ancient East, 62–145. 16 Deissmann, Light from the Ancient East, 146–251. 17 Deissmann, Light from the Ancient East, 252–392.

Papyrology

187

given text preserves complete details on a given subject that it addresses. Most of the time the answer or the reply to communication is not preserved, personal details are unknown to the outsider, and information is lacking in the eye of the modern reader. Lack of evidence, however, does not necessarily indicate the absence of certain topics, words or phrases in ancient everyday life; it is often a combination of luck and coincidence that dictated which texts survived and which did not.

2. The Application of Papyrology for New Testament Studies Today: Current Examples Deißmann’s interest in documentary papyri was also shared by others. From 1914 to 1929 James Hope Moulton and George Milligan published the lexicon The Vocabulary of the Greek Testament Illustrated from the Papyri and Other Non-Literary Sources, which has been revived only lately by Greg H. R. Horsley and John A. L. Lee.18 However, apart from these ground-breaking insights into the significance of documentary papyri for biblical studies, it was not long before academic investigations almost ceased entirely.19 Individual scholars, such as John Lee White with his works on epistolary formulas, and Hans Dieter Betz’s work on magical papyri, kept the awareness of documentary papyri alive in the biblical scholarly community.20 In the 1980s Greg H. R. Horsley established the series New Documents Illustrating Early Christianity (New Docs), a collection of published documentary papyri of particular interest for biblical studies. Stephen R. Llewelyn and James R. Harrison are the present editors of the series.21 18 James H. Moulton and George Milligan, The Vocabulary of the Greek Testament Illustrated from the Papyri and Other Non-Literary Sources (London: Hodder and Stoughton, 1929). Cf. Gregory H. R. Horsley and John A. L. Lee, “A Lexicon of the New Testament with Documentary Parallels: Some Interim Entries, 1,” Filología Neotestamentaria 10 (1997): 55–84, and John A. L. Lee and Gregory H. R. Horsley, “A Lexicon of the New Testament with Documentary Parallels: Some Interim Entries, 2,” Filología Neotestamentaria 11 (1998): 57–84. 19 Cf. Gerber, Deissmann the Philologist, 365. 20 Cf. Hans Dieter Betz, “Zum Problem der Auferstehung Jesu im Lichte der griechischen magischen Papyri,” in Hellenismus und Urchristentum: Gesammelte Aufsätze, vol. 1 (Tübingen: Mohr Siebeck, 1990), 230–61. Cf. John L. White, The Form and Function of the Body of the Greek Letter: A Study of the Letter-Body in the Non-Literary Papyri and in Paul the Apostle, SBLDS 2 (Missoula: SBL, 1972). For more details and further bibliography cf. Peter Arzt-Grabner, Philemon, Papyrologische Kommentare zum Neuen Testament 1 (Göttingen: Vandenhoeck & Ruprecht, 2003), 39–43. 21 The latest volume is Stephen R. Llewelyn and James R. Harrison, Greek and Other Inscriptions and Papyri Published 1988–92, New Documents Illustrating Early Christianity, vol. 10 (Grand Rapids: Eerdmans, 2012). The journal Early Christianity, established in 2010, pursues a similar objective with its rubric “New Discoveries” and its particular reports

188

Christina M. Kreinecker

A systematic analysis of documentary papyri had been overdue for many reasons, one of them being that the number of published papyri since Deißmann has increased tenfold at least.22 It was a team of New Testament scholars in Salzburg who accepted the challenge at the end of the 1980s and established a new series, the Papyrological Commentaries on the New Testament.23 2.1. Papyrological Commentaries on the New Testament The Papyrological Commentaries on the New Testament take seriously Deißmann’s observations on the value of documentary material for biblical studies and apply them systematically to New Testament writings. After several years of intense study and financial support by the Austrian Science Funds (FWF), the series was launched in 2005 with its first publication, the commentary on Philemon.24 Each volume of the Papyrological Commentaries consists of at least two parts: a topical analysis and a verse-by-verse commentary. The investigations are driven by the main question: how might contemporaries have understood the New Testament text at the time when it came into being? This main question is based on Deißmann’s observation that the language of the New Testament falls within the wider spectrum of Koiné Greek. The main question of the series triggers others: What connotations and possible meanings did words and phrases have at the time? What and where do documentary texts speak of matters mentioned in or relevant to the New Testament, like slavery, pseudepigraphy, life situation of women, life in associations, etc.? The method necessary to answer such questions is rigorous and was laid out in the first volume of the series in 2005. This methodology forms the groundwork for all subsequent volumes already published and those still to follow.25 Before any comparison between papyrological evidence and the New Testament text is made the “comparison value” of the documentary material is considered. Documents that are closer to the biblical texts in terms of chronology, geography, genre, thematic parallels, etc. are of higher value. To give one example, a personal letter from 55 CE talking about a male slave child is of higher value when focusing on the letter of “Neuigkeiten aus der Papyrologie” by Peter Arzt-Grabner, cf. so far Early Christianity 3 (2012): 111–15, and Early Christianity 6 (2015): 561–69. 22 Peter Arzt-Grabner, “Neues zu Paulus aus den Papyri des römischen Alltags,” Early Christianity 1 (2010): 131–57, here 132. 23 Cf. Arzt-Grabner, Philemon, 42–3. 24 Arzt-Grabner, Philemon. Further published volumes are the commentary on 1 and 2 Corinthians and 2 Thessalonians: Peter Arzt-Grabner et al., 1. Korinther, Papyrologische Kommentare zum Neuen Testament 2 (Göttingen: Vandenhoeck & Ruprecht, 2006); Christina M. Kreinecker, 2. Thessaloniker, Mit einem Beitrag von Günther Schwab, Papyrologische Kommentare zum Neuen Testament 3 (Göttingen: Vandenhoeck & Ruprecht, 2010); Peter Arzt-Grabner, 2. Korinther, unter Mitarbeit von Ruth E. Kritzer, Papyrologische Kommentare zum Neuen Testament 4 (Göttingen: Vandenhoeck & Ruprecht, 2014). 25 Cf. Arzt-Grabner, Philemon, 43–56; Kreinecker, 2. Thessaloniker, 19–31.

Papyrology

189

to Philemon due to its closeness in time and content, than, for example, an account for grain delivery from 250 BCE. Texts with obvious or suspected Christian background are not taken into consideration as the series does not seek to reflect on the earliest reception within a Christian context. Likewise, due especially to their often literary status, other sources such as texts of the Second Temple Period or the Septuagint itself, doubtlessly relevant for the understanding of New Testament writings and language of their time, are not considered. Neither are other documentary texts besides those preserved on papyri, tablets or ostraca. For example, inscriptions are not considered not least because they represent a different part of society and a different level of public notice than represented by documentary papyri. Though these resources are unquestionably of utmost interest, the series adopts a pragmatic approach that focuses on documentary papyri exclusively.26 It is almost impossible to predict the outcome of papyrological investigations beforehand. Most of the time it is impossible to foresee what kind of information documentary papyri will provide that is relevant to the New Testament. On several occasions the texts have brought unexpected insights to light, for example, concerning slavery. Evidence from ancient slave sales contracts presents the slave Onesimus in a new light. Traditionally, Onesimus has been acknowledged as a fugitive, a runaway slave escaping his master. Slave contracts, however, provide an alternative interpretation. Onesimus may have been a truant, a slave who lingered and stayed away from home for longer than his errand could plausibly have taken. Whatever actually happened between the slave and his master remains unknown, yet the evidence from ancient slave contracts indicates that Paul decided to write about Onesimus in terms of a truant, not a runaway. Accepting this possibility requires a shift in understanding the dynamics of Paul’s requests to Philemon on his slave’s behalf.27 Another major shift from traditional exegetical interpretations concerns the genre of ancient letters. Studying letter conventions and typical formulas in documentary letters – in particular those of personal interest – leads to the conclusion that the beginning of a letter body sometimes holds either thanks or a rebuke. At this place of a letter, thanks are often mentioned for good news the sender 26 The inscriptional material is ample and extensive and needs to be considered in its own right. The new collaborative project “Epigraphy and New Testament” of Markus Öhler, Thomas Corsten (both Vienna) and Joseph Verheyden (Leuven) is dedicated to this task. Cf. Thomas Corsten, Markus Öhler and Joseph Verheyden, Epigraphik und Neues Testament, WUNT 365 (Tübingen: Mohr Siebeck, 2016). 27 Cf. Peter Arzt-Grabner, “Onesimus erro: Zur Vorgeschichte des Philemonbriefes,” ZNW 95 (2004): 131–43. Arzt-Grabner, “‘Neither a Truant nor a Fugitive’: Some Remarks on the Sale of Slaves in Roman Egypt and Other Provinces,” in Proceedings of the TwentyFifth International Congress of Papyrology, Ann Arbor, July 29–August 4, 2007, ed. T. Gagos and A. Hyatt (Ann Arbor: Scholarly Publishing Office The University of Michigan Library, 2010), 21–32; Arzt-Grabner, Philemon, 101–108.

190

Christina M. Kreinecker

previously received about the addressee’s good health or recovery from illness.28 A rebuke on the other hand regularly expresses the sender’s irritation that the addressed person has not written a letter. The Greek word regularly used to express rebukes is θαυµάζω.29 One example is P.Mich. VIII 479,4–8 (early second century CE): θαυµά|ζω πῶς ἀναπλεύσας οὐκ ἀντέγραψάς | µοι περὶ τῆς σωτηρίας σου, ἀλλὰ ἕως σή|ερ[ο]ν ἀγωνιῶ διότι νωθρευόµε|νος ἀπʼ ἐµοῦ ἐξῆλθες (“I am astonished that after you sailed upriver you did not write to me about your wellbeing, but until today I have been anxious because you were indisposed when you left me”).30 Other reasons for rebukes are that the addressee has not come for a visit,31 did not do as the sender had asked,32 spent too much money (for example, on irrigation33 or on oil34), or some other obvious misunderstanding,35 among others. An example with a temporal proximity to the Pauline letters is BGU III 850, 1–7 (76–84 CE).36 Χαιρήµων Ἀπολλωνίωι | τῶι φιλτάτωι χαί[ρ]ειν. | θαυµάζωι (l. θαυµάζω) ἐπὶ τῆι [ἀσ]υ̣ντ̣αξ̣ ίᾳ σου, καίτοι ἐµοῦ σ̣ε̣ πολλὰ | ἐρωτήσαντος. παρακληθεὶς | [ο]ὖν ἄδελφε, ἄνε̣λ̣θε πρὸς | [ἡµ]έραν µίαν

28 For examples see Peter Arzt-Grabner, “Paul’s Letter Thanksgiving,” in Paul and the Ancient Letter Form, ed. S. E. Porter and S. A. Adams (Leiden: Brill, 2010), 129–58. ArztGrabner, Philemon, 135–40. 29 Cf. Christina M. Kreinecker, “Emotions in Documentary Papyri: Joy and Sorrow in Everyday Life,” in Yearbook 2011: Emotions from Ben Sira to Paul, ed. R. Egger-Wenzel and J. Corley (Berlin: De Gruyter, 2012), 451–72, 455–56; Kreinecker, 2. Thessaloniker, 146–47. 30 Cf. H. C. Youtie and J. G. Winter in P.Mich. VIII p. 68. For further examples cf. PSI V 502,12 (7 July 257 BCE); BGU IV 1041,12 (second century CE); P.Mich. VIII 500,3 (second century CE); P.Ryl. II 235,6 (second century CE); P.Mich. XV 751,4 (late second century CE); P.Mich. III 209,6 (late second or early third CE); PSI XV 1553,9 (first half third century CE); P.Oxy. XIV 1770,8 with BL I 467 (late third century CE); SB XXII 15603,5 (late third century CE); P.Mert. I 28,4 (late third century CE); SB III 6222,4 (end of third century CE); P.Oxy. I 123,5 (third or fourth century CE); LIX 3997,3–4 (third or fourth century CE). 31 See P.Oxy. LXVII 4627,3 (late third century CE). 32 Cf. P.Lille I 26,4 with BL VI 59 (third century BCE), P.Mert. II 80,3 (second century CE); P.Oxy. I 113,20 (second century CE) and also P.Prag. I 109, Recto 3 (249–269 CE); P.Freib. IV 69,2 (second or third century CE); P.Mil.Vogl. IV 256,3 (second or third century CE). 33 P.Oxy. XLII 3063,11 (second century CE). 34 P.Oxy. XXXVI 2783,6 (third century CE). 35 Cf. e.g. P.Yale I 30,9–10 (ca. 265 BCE), a misunderstanding about broken goods. 36 The text includes the corrections from Bror Olsson, Papyrusbriefe aus der frühesten Römerzeit (Uppsala: Almquist & Wiksells, 1925), nr. 47.

Papyrology

191

Chairemon to the beloved Apollonios, greetings. I am astonished by your disorganisation, despite the fact that I have asked you many times. You are, therefore, requested, brother: come for a day …

In Gal 1:6–7 Paul expresses his astonishment and irritation exactly in the style of ancient letter conventions: θαυµάζω ὅτι οὕτως ταχέως µετατίθεσθε ἀπὸ τοῦ καλέσαντος ὑµᾶς ἐν χάριτι [Χριστοῦ] εἰς ἕτερον εὐαγγέλιον, 7 ὃ οὐκ ἔστιν ἄλλο, εἰ µή τινές εἰσιν οἱ ταράσσοντες ὑµᾶς καὶ θέλοντες µεταστρέψαι τὸ εὐαγγέλιον τοῦ Χριστοῦ. I am astonished that you are so quickly deserting the one who called you in the grace of Christ and are turning to a different gospel – 7 not that there is another gospel, but there are some who are confusing you and want to pervert the gospel of Christ.

Papyrological evidence shows that Paul’s use of the word θαυµάζω is not just an arbitrary choice but the convention for expressing astonishment in a letter. Together with further observations on letter writing conventions, phrases and styles, this draws a picture of Paul as a skilled letter writer with his own individual style exercised within the boundaries of ancient letter writing. This conclusion can particularly be drawn by comparison with the personal style of individuals, which can be studied where several letters written by the same author are preserved.37 These brief examples may suffice to illustrate the value of documentary papyri for biblical studies: they can add additional information, clarification and, occasionally, alternative perspectives to long-held convictions in the field. Studying documentary papyri is, however, not a substitute for traditional methods of biblical research, but a supplement. As I mentioned above the nature of documentary material has to be taken into account: it is fragmentary and the picture that it provides today may be altered or extended by new findings and publications appearing on a regular basis. All conclusions drawn from papyri are therefore tentative rather than ultimate; they are conclusions in progress and reflect the current state of the discipline and its sources. 2.2. Documentary Papyri and Women The historical information that documentary papyri hold on a large scale offers other benefits for New Testament studies, as already pointed out by Deißmann. Women’s studies have been one major field. Documentary papyri preserve more information about the actual life of women than any literary source. Most importantly, they preserve direct evidence of women, that is letters written by

37 Cf. Günther Schwab, “Stileigenheiten als Identitätsmerkmale in Papyrusbriefen,” in 2. Thessaloniker, Christina M. Kreinecker, Mit einem Beitrag von Günther Schwab, Papyrologische Kommentare zum Neuen Testament 3 (Göttingen: Vandenhoeck & Ruprecht, 2010), 39–60.

192

Christina M. Kreinecker

women, even if they are few in number.38 Studying women from a papyrological viewpoint has brought various unexpected insights into, for example, their social status in general, their involvement in legal matters,39 and the important economic role they played. One particular genre that provides a vast amount of information regarding women’s everyday lives are census declarations from Roman Egypt. More than 300 declarations are preserved on papyrus from 12–259 CE. The fact that from 19/20 CE onwards such declarations had to be made every 14 years,40 makes them an excellent resource for studying the development and change in human population over a longer period.41 From these declarations Roger S. Bagnall and Bruce Frier42 have made deductions about the social realities for women in the first centuries of the Roman rule in Egypt, the period which coincides with the composition of the New Testament. They show, for example, that while average life expectancy did not differ between the genders in a significant way,43 65% of all women who were married at least once in their life (95% of all women),44 were married for the first time between the age of 12 and 20. Conversely, men were on average 7.5 years older than women upon their first marriage.45 Due to 38 Roger S. Bagnall and Raffaella Cribiore, Women’s Letters from Ancient Egypt, 300 BC–AD 800 (Ann Arbor: University of Michigan Press, 2006), 19: “Private letters form a relatively stable part of the total documentation from the late Hellenistic period until late antiquity. Women’s letters as a percentage of the letters vary somewhat more, but as a percentage of total documentary papyri there is a remarkable stability from the first century BC to the fourth century AD, with the percentage right at 1 percent of the total, plus or minus a tenth of a percent.” 39 The texts show that despite regulations against it, women occasionally functioned as co-sureties and acted in legal matters without a guardian even when one was legally required, cf. Claudia Kreuzsaler, “Der Rechtsalltag von Frauen im Spiegel der Wiener Papyri,” in Emanzipation am Nil: Frauenleben und Frauenrecht in den Papyri, ed. H. Froschauer and H. Harrauer (Wien: Phoibos Verlag, 2005), 1–17, and Edgar Kutzner, Untersuchungen zur Stellung der Frau im römischen Oxyrhynchos (Frankfurt: Peter Lang, 1989). 40 Cf. P.Oxy. II 254 with BL X 137–138 and XI 143 (ca. 13–26 CE). The first census in Egypt was conducted under the reign of Augustus in 11/10 BCE. Until 19/20 CE the census was conducted in a seven-year cycle. 41 Cf. Fritz Mitthof, “Junge Heirat, viele Geburten, frühe Verwitwung oder Scheidung: Demographische Aspekte des Frauenlebens im römischen Ägypten,” in Emanzipation am Nil: Frauenleben und Frauenrecht in den Papyri, ed. H. Froschauer and H. Harrauer (Wien: Phoibos Verlag, 2005), 65–73, here 67. 42 Roger S. Bagnall and Bruce W. Frier, The Demography of Roman Egypt (Cambridge: Cambridge University Press, 1994). 43 Bagnall and Frier, The Demography of Roman Egypt, 75–110. 44 Cf. Bagnall and Frier, The Demography of Roman Egypt, 110–34. 45 Cf. Bagnall and Frier, The Demography of Roman Egypt, 118–21. One of the main reasons for this is that men were expected to be able to sustain a family and the economic basis for that had to be build up. The age difference has – among others – the effect that women regularly became widows in their thirties or forties. Most of these widows would

Papyrology

193

high infant mortality rates, at least 6 live births were required in order to bring two children to adulthood and thereby keep the reproduction rate stable. The observation of a slight increase in the population of Egypt in the early centuries CE suggests that women were having rather more children, i.e. about 7–8 live births per woman.46 From these basic facts one can deduce that a major part of a woman’s everyday life was concerned with being pregnant and nursing. Various papyrus letters talking about pregnant women and women who lost their babies, the deaths of women during child birth, and also the prayers for good news of healthy deliveries attest to these demographical observations. Documentary papyri also attest that women were far more economically active than is apparent from other (literary) sources.47 Even if their economic role does not equal that of men, they still played important roles both inside and outside the home, for example, as weavers, managers of the household, day workers, slaves, agricultural workers, or as workers in other trades and services.48 A significant number of women owned property, for which they paid taxes and which they personally managed. Various forms of documents illustrate this: declarations of property, sales contracts, purchase contracts, contracts of lease or rent, loan contracts, and others. From declarations of property, it is clear that women owned, in part or in full, a wide variety of assets: animals such as sheep, goats, and camels, money, grain, slaves, houses, land (on average 2.5 hectares), luxury items, jewellery, ships, mills, irrigation plants, and other items.49 The main reasons for women to have property in the first place are either inheritance or dowry.50 Papyrological evidence suggests that in Roman Egypt between onefifth and one-third of all landowners were women, and about 16–25% of all land was owned by women.51 Women were perfectly able to maintain and manage their own business, for which P.Oxy. VI 932 (late second century CE) is one example. The woman,

then try to live with either their siblings or their grown-up children. Those who became single mothers and ran their own household were a minority, cf. Mitthof, “Junge Heirat,” 72. 46 Mitthof, “Junge Heirat,” 72. 47 Jane Rowlandson, Women and Society in Greek and Roman Egypt: A Sourcebook (Cambridge: Cambridge University Press, 1998), 218: “The documentary evidence of the papyri tells us far more about the economic lives of women than we can deduce from any literary evidence.” 48 Cf. Csaba A. La’da, “Die wirtschaftliche Stellung der Frau im hellenistischen, römischen und byzantinischen Ägypten (332 v. Chr.–642 n. Chr.),” in Emanzipation am Nil: Frauenleben und Frauenrecht in den Papyri, ed. H. Froschauer and H. Harrauer (Wien: Phoibos Verlag, 2005), 35–49. 49 Cf. Bagnall and Cribiore, Women’s Letters, 70–1. 50 Cf. Kreuzsaler, “Der Rechtsalltag von Frauen im Spiegel der Wiener Papyri,” 3. 51 La’da, “Die wirtschaftliche Stellung der Frau,” 46.

194

Christina M. Kreinecker

Thais, writes to one of her slaves and workmen, a certain Tigrios, who may have been the steward of her estate: Θαῒς Τιγρίῳ τῶι ἰδίῳ χαίρειν. ἔγραψα Ἀπολιναρίῳ ἵνα γένηται ἐν τῇ Πέτνη ἵνα µ[ε]τρήσηι. ἐρῖ (l. ἐρεῖ) σο δὲ Ἀπολινάριος πῶς τὰ θέµατα καὶ τὰ δηµόσια· τὸ ὄνο5 µα ὃ ἂν αὐτός σοι εἴπῃ. ἂν ἔρχῃ ἄφες ἀρτάβας ἓξ ἰς (l. εἰς) τοὺς σάκκους σφραγίσας λαχανοσπέρµου ἵνα πρόχιροι (l. πρόχειροι) ὦσι, καὶ ἐὰν δύνῃ ἀναβῆναι ἵνα ἐπιγνοῖς τὸν ὄνον. ἀσπάζεταί σε Σαραποδώρα κ(αὶ) Σ Σαβῖνος. τὰ 10 χοιρίδια χωρὶς µοῦ µὴ πώλι (l. πώλει). ἔρρωσο.

Thais to her Tigrios, greeting. I wrote to Apolinarios to come to Petne for the measuring. Apolinarios will tell you how the deposits and the taxes (stand). He will tell you the name, whatever it is. I Ιf you come, take out six measures of vegetable seed and seal t them in sacks, so that they may be ready. And if you can, go up and find out about the donkey. Sarapodora and Sabinos salute you. Do not sell the piglets without me. Good bye.

These socio-historical insights from documentary papyri about women and their opportunities in life have an impact on biblical studies and might support or change certain readings and interpretations of texts. For example, women mentioned as co-workers around Paul may be viewed differently, but also women in Galilee “supporting Jesus and his disciples” (Luke 8:3) may have been more than simply women praying for the good outcome of Jesus’ mission or women of limited means. Documentary papyri are a treasure trove in this regard, and may provide further proof of or contradict evidence to existing interpretations of the Bible and in some cases they may even lead to new ones. The engagement of New Testament studies with documentary papyri is worth pursuing and the Papyrological Commentaries as well as other projects like the New Docs or the New Moulton-Milligan Project are ongoing. The outcome, however, is pending.

3. Papyrology and New Testament Studies: The Future Instead of raising the question of what results biblical studies could expect from future discoveries or analysis of still unpublished documentary material, I will rather focus on the scholarly communities in which biblical studies is located. When it comes to the future of textual scholarship on the Bible with a papyrological interest in mind, two necessities are apparent and both are connected to previous scholarship: the first is the need to bring down disciplinary boundaries and the second is the necessity of academic friendship, a long-valued custom within the papyrological discipline.

Papyrology

195

3.1. Overcoming Obstacles and Thinking Outside the Box Both Herman C. Hoskier and Adolf Deißmann breathed new life into their disciplines. Hoskier looked at biblical manuscripts in a different way than was considered standard. Later on in his life he took one step further and applied his knowledge from manuscripts to other areas such as spiritualism.52 Even if biblical scholarship has developed in a different direction, Hoskier is still respected today for the sheer amount of material that he gathered and took into consideration for his studies. Deißmann on the other hand did not accept the barriers between philology and theology which were taken for granted in his time. Overcoming the academic pigeonholing of his discipline, he brought major shifts to the linguistic understanding of the New Testament and set the ground for an entirely new way of evaluating biblical texts through documentary papyri. He redefined biblical studies as a discipline and like Hoskier – though in a different way – took his academic insights further, as can be seen in his political and ecumenical activities particularly on the eve of the First World War (Völkerverständigung), which earned him two nominations for the Noble Peace Prize in 1929 and 1930.53 In their time, both Hoskier and Deißmann stood on the edge of what was considered standard scholarship and methodology. Both worked outside the box of the traditional understandings of their disciplines. Much can be learned from their examples and gained for the future of textual scholarship when it comes to overcoming the obstacles and facing the challenges that both the present and imminent future hold. In terms of discipline, multidisciplinary and interdisciplinary collaboration will be even more vital than it already is. The vast knowledge relevant for textual scholarship, to give one example, cannot be held by one individual anymore – if ever it could. To find a person who controls Greek, Latin, Coptic, Syriac, Armenian, Gothic, Java, Python, and philosophy is a rare discovery. However, as some current research projects and collaborations have shown, for example COMPAUL at the University of Birmingham, bringing together specialist knowledge from various subjects, languages, and disciplines brings new insights and results to the scientific community. In terms of COMPAUL, the new evaluation of Patristic quotations has been made possible through a combination of joint knowledge from disciplines and fields like New Testament textual criticism, patristics, classical philology, linguistics, theology, computer sciences, and digital humanities.54 Fundamental 52 Cf. H. C. Hoskier, What is Nirvana? (Portland, ME: The Mosher Press, 1930). Cf. also Hoskier’s publications under his pseudonym “Signpost”: In Tune with the Universe (London: Rider & Co, 1932), and The Back of Beyond (London: The Daniel Company, 1934). 53 Cf. Gerber, Deissmann the Philologist, 209–341. 54 Cf. Christina M. Kreinecker, “The Earliest Commentaries on Paul as Sources for the Biblical Text: A New Research Project at the Institute for Textual Scholarship and Electronic Editing at the University of Birmingham (COMPAUL),” Early Christianity 3 (2012): 411–

196

Christina M. Kreinecker

research has become more complex than ever, and it will remain so in the future, requiring the work of inter and multidisciplinary teams of people collaborating beyond institutional and political borders as well. This collaboration will bring shifts within the disciplines, and challenge collaborators’ understanding of how data and material is examined. Some of these changes concerning New Testament textual scholarship are overdue, for example, when it comes to classifying texts and manuscripts. The bipolar distinction between literary and documentary texts does not cover semi-literary ones and even with three categories the boundaries remain fluid. In New Testament textual criticism the distinction between papyri and majuscules is somewhat diffuse. In addition, the evidence from the early versions and non-continuous biblical manuscripts requires further consideration as it pertains to reconstructing texts: catenae manuscripts, the early versions, so-called amulets, and pieces of fabric holding biblical quotations are still waiting to be situated within the field in a responsible way.55 There is a more fundamental challenge to textual scholarship on the Bible which will affect the future: the academic landscape is becoming more and more monoglot. Despite their outspoken commitment to multilingualism, various international academic associations seem to struggle with this, both in papyrology and biblical studies. Despite the fact that the triennial International Congress of Papyrology maintains the tradition of providing platforms for papers in the four major languages of the discipline (German, French, Italian, and English), it is a fact that fewer people join French and Italian sessions and applications for papers in languages other than English have dropped significantly in recent years. The same tendency can be observed in biblical studies. For example, during the Society for New Testament Studies (SNTS) annual General Meetings, where lectures and papers in English, German, and French are offered, attendance figures have dropped for non-English sessions. Moreover, the quantity of quoted

15; H. A. G. Houghton, “The Use of the Latin Fathers for New Testament Textual Criticism,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes, 2nd ed. (Leiden: Brill, 2014), 375– 405; Houghton, “Recent Developments in New Testament Textual Criticism,” Early Christianity 2 (2011): 245–68. 55 Cf. D. C. Parker, An Introduction to the New Testament Manuscripts and Their Texts (Cambridge: Cambridge University Press, 2008), 35–46; Parker, Textual Scholarship and the Making of the New Testament (Oxford: Oxford University Press, 2012), 38–40; ArztGrabner, “Neuigkeiten aus der Papyrologie”; Arzt-Grabner, “Neuigkeiten aus der Papyrologie (2).” For more details see e.g. Brice C. Jones, New Testament Texts on Greek Amulets from Late Antiquity, LNTS 554 (London: Bloomsbury, 2016); Peter M. Head, “Additional Greek Witnesses to the New Testament (Ostraca, Amulets, Inscriptions, and Other Sources),” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes, 2nd ed. (Leiden: Brill, 2014), 429–60.

Papyrology

197

German scholarly contributions in English publications, in comparison to the number of English citations in German publications within the field, is lower. And finally, there will be challenges from outside the academic field altogether, some of which are already looming on the horizon. On a political scale “Brexit” has to be mentioned, since it seems inevitably to bring changes to the possibilities for collaboration of institutions and scholars within Europe and might impose on the personal good will of scholars. Politics impinges on the building and maintaining of important scholarly networks. These are just some of the challenges, challenges that are being met by the continued development of projects and ideas that address these fundamental concerns. This is also the lesson that can be learned from Hoskier and Deißmann’s examples for the future of textual scholarship: obstacles are no reason for withdrawal or self-pity. On the contrary, bringing down boundaries is an academic task for all time. 3.2. The Need for Academic Friendship Papyrology as a discipline comes with a long and outspoken tradition of friendship, the amicitia papyrologorum. In 1933 at the Third Papyrological Congress in Munich the German scholar Leopold Wenger mentioned the “friendship of papyrologists” in his closing speech. At the time, his main concerns connected to that term were internationalism and the absence of odium, the grudge against other scholars and their work, as he allegedly witnessed in other disciplines. Over the years a more active approach to friendship has been brought forward among papyrologists: a personal dimension and friendship both professional and personal between individual scholars, though at the same time with the realistic view that personal relationships cannot be maintained with several hundreds of people in a growing field. As president of the Association Internationale de Papyrologues (AIP) in 2010, Roger Bagnall reflected on friendship in his opening speech of the 26th International Congress of Papyrology in Geneva. He optimistically stated: “We recognize that…we can do more together than separately.”56 In a globalised world where national boundaries have shifted in terms of importance and a “shortage of academic positions” seems ubiquitous, institutions have cooperated “more readily today than before, to the benefit of all.”57 According to Bagnall, papyrology as a discipline has always distinguished itself in having “some welcome willingness to welcome scholars from outside an institution to bring their expertise to bear on publishing collections, too.”58 It is clear, however, that none Roger S. Bagnall, “The Amicitia Papyrologorum in a Globalized World of Learning,” in Actes du 26e Congrès international de papyrologie: Genève, 16–21 Août 2010, ed. P. Schubert (Genève: Droz, 2012), 1–5. 57 Bagnall, “Amicitia Papyrologorum,” 2. 58 Bagnall, “Amicitia Papyrologorum,” 2. 56

198

Christina M. Kreinecker

“of this happens of itself…Underlying all of it is a willingness to share resources, to be open to others and welcome them, and the courage to believe that openness and generosity are good things for all involved rather than a threat to those who have.”59 Active friendship is vital in a world where “the continuing usefulness of our learned disciplines is not taken for granted by too many people outside the ranks of their practitioners.”60 Applying these observations to textual scholarship on the Bible, one of the major challenges for the present and future is communication about one’s findings and their meaning inside and outside the scientific community. Some examples from New Testament textual criticism may suffice to illustrate this point. The fact that even New Testament scholars today sometimes mistake the text of editions such as the Nestle-Aland or the Greek New Testament as the text needs attention.61 Apparently even more explanatory work from the editors but also more “active reception” by those who are using these editions will be necessary in the future. The same holds true for the CBGM, which does not stand for “Complicated Beyond Grasping Method,” but “Coherence Based Genealogical Method,” an effective working tool for dealing with the fact that the transmission of the biblical text is contaminated.62 A recent publication by Tommy Wasserman and Peter J. Gurry is one essential step to make the method more easily accessible for people generally interested in the field.63 Additionally, the significance and impact of new findings have to be explained – even within the scientific community – be it individual fragments of hitherto unknown Christian writings or entire manuscripts that have been thought lost but rediscovered like Fortunatianus’ commentary on the Gospels in 2012.64 In times when outreach is on every academic institution’s mission statement, this has to be done without overstatements resulting in sensational media speculations about early dating and without understatements, resulting in keeping discoveries under wraps and away from any publicity, let alone the scholarly community. These are only a few examples where textual scholarship on the Bible could benefit from still more active communication. Others could be added. A clear commitment to active academic friendship involves “resource-sharing,”65 for example visibly sharing online, be it an open access policy as standard, Bagnall, “Amicitia Papyrologorum,” 2. Bagnall, “Amicitia Papyrologorum,” 3. 61 Cf. Parker, Textual Scholarship, 13 and 146. 62 Cf. e.g. Gerd Mink, “Contamination, Coherence, and Coincidence in Textual Transmission,” in The Textual History of the Greek New Testament: Changing Views in Contemporary Research, ed. K. Wachtel and M. W. Holmes (Atlanta: SBL, 2011), 141–216. 63 Tommy Wasserman and Peter J. Gurry, A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method (Atlanta: SBL, 2017). 64 Lukas Dorfbauer, ed., Fortunatianus Aquileiensis. Commentarii in evangelia, CSEL 103 (Berlin: De Gruyter, 2017), 1–2. 65 Bagnall, “Amicitia Papyrologorum,” 3. 59 60

Papyrology

199

digitalisation of manuscript images, or free full public catalogues of collections and manuscripts. The advantages of this approach are obvious, and can already be seen in projects relying on collaboration and sharing both knowledge and data. The most successful example from papyrology are The Papyrological Navigator and Editor,66 which guarantee some independence from grants for the survival of online instruments, as they can be “maintained by groups of users spread around the globe, with quality control provided by an editorial board checking work that is submitted online” by a “constantly shifting and widely distributed body of volunteer contributors.”67 In terms of New Testament studies, the New Testament Virtual Manuscript Room (NTVMR), the Workspace for Collaborative Editing, and electronic tools developed at the University of Birmingham, such as the Version Editor are examples for current best practice, which allow a glimpse of the exciting possibilities in the future.68 Academic friendship is not a monopoly of papyrological scholars, but is has been an outspoken disciplinary concern. It is also vital and of utmost importance for the future of textual scholarship on the Bible. When friendship is applied at an institutional level, this means maintaining and expanding already existing good practices, e.g. strong advocacy for open access, open resources, digital humanities, and those platforms that allow inter- and multidisciplinary collaboration and connection. 3.3. An Afterthought Were Hoskier and Deißmann friends? So far I cannot detect any evidence that Hoskier and Deißmann personally knew one another. However, they did share a mutual acquaintance, the papyrologist and Quaker theologian J. Rendel Harris, who was interested in many fields and amenable to more than one context of research.69 In the end, the future of any academic discipline lies with the individual scholars who are willing and able to further it beyond the well-known and well-established realms. Hoskier and Deißmann are both examples of this reality in their own ways.

See http://papyri.info/ [accessed 4 March 2018]. Bagnall, “Amicitia Papyrologorum,” 4. 68 For the NTVMR see http://ntvmr.uni-muenster.de/de; for the Workspace for Collaborative Editing see https://www.birmingham.ac.uk/research/activity/itsee/projects/workspace.aspx; cf. H. A. G. Houghton and C. J. Smith, “Digital Editing and the Greek New Testament,” in Ancient Worlds in Digital Culture, ed. C. Clivaz, P. Dilley, and D. Hamidović (Leiden: Brill, 2016), 110–27. 69 The Catalogue of the papers of James Rendel Harris at the Woodbrooke College in Selly Oak, Birmingham (UK), illustrates that Harris was in regular correspondence with both scholars, c.f. e.g. the reference numbers JRH/1/3/27, JRH/1/4/5, JRH /1/4/7, JRH/1/4/6/7 and JRH/1/4/9 (numbers 45, 54–68, 111–117 and 160) and JRH/1/3/27 (https://www.woodbrooke.org.uk/wp-content/uploads/2017/02/jrh1.pdf, accessed 4 March 2018). 66 67

Patterns of Correction as Paratext: A New Approach with Papyrus 46 as a Test Case Jacob W. Peterson It was in reaction to Zuntz’s treatment of the corrections in P46 that Tasker claimed, “he must surely be lacking in imagination who does not feel some thrill at such an attempt to see reflected in the mistakes and corrections of a document written over seventeen hundred years ago images of the ‘life-situation’ which produced them.”1 In his study on Codex Sinaiticus, Parker similarly wrote, “the detailed study of the correctors and their corrections is worthwhile, because of the insight it offers into the ways in which readers studied and used a biblical manuscript in antiquity.”2 In general, corrections have been treated as isolated events revealing the concerns or theology of the reader or community at that particular point in the text. This chapter develops a method for approaching the corrections of an entire manuscript as a cohesive whole that can be analysed for insight into its later users. To do so, corrections have been reframed as a type of paratext and analysed using the insights from that field of study. The chapter begins with an introduction to paratexts and the key terms involved. This is followed by an explanation of the method for investigating patterns of correction within manuscripts. Finally, the method is applied to the oldest manuscript of the Pauline Epistles, P46, as a test case. The result of applying this new method to P46 reveals the manuscript to have likely been used and read in a Jewish-Christian church setting. The aim of this new method is to aid textual scholarship in moving beyond the collection of raw text-critical data on corrections to find patterns and meaning within the mass. This move requires viewing corrections not as separate points of data, but as clues potentially pointing to a larger cultural framework within which a manuscript was used and altered. These cultural frameworks have, with few exceptions, remained a mystery to scholars. Where applicable, this new method provides a tool by which text critics can further evaluate the tendencies within particular manuscripts and offer explanations for the readings and R. V. G. Tasker, “The Text of the ‘Corpus Paulinum,’” NTS 1 (1954–1955): 190. D. C. Parker, Codex Sinaiticus: the Story of the World’s Oldest Bible (Peabody: Hendrickson, 2010), 79. 1 2

202

Jacob W. Peterson

corrections they contain. Not only are studies of the patterns in individual manuscripts valuable in and of themselves, they collectively contribute to a better understanding of the social settings and influences affecting the transmission of the New Testament text. This offers an interdisciplinary future for textual scholarship wherein corrections become another potential window into the history of reception, exegesis, and the church.

1. Paratexts: A Brief Introduction 1.1 Genette The recasting of the elements of a work existing alongside the main text as “paratexts” has a relatively short history in the grand scheme of literary analysis. The project that would ultimately lead to the study of paratexts as a phenomenon began in 1979 with the publication of Introduction à l’architexte3 in which Gérard Genette shifted the focus of his literary studies from the text itself onto those “transtextual” features that bring the main text “into relation (manifest or hidden) with other texts.”4 Although the term “paratextuality” appears in Introduction à l’architexte, the first development of paratexts as a new lens for literary analysis comes in Genette’s Palimpsestes: la littérature au second degré.5 Therein, paratexts are defined as the generally less explicit and more distant relationship that binds the text properly speaking, taken within the totality of the literary work, to what can be called its paratext: a title, a subtitle, intertitles; prefaces, postfaces, notices, forewords, etc.; marginal, infrapaginal, terminal notes; epigraphs; illustrations; blurbs, book covers, dust jackets, and many other kinds of secondary signals, whether allographic or autographic.6

The third volume in Genette’s transtextuality trilogy, Seuils, is dedicated to elucidating the study of paratexts.7 There we find that the paratext is what enables a text to become a book and to be offered as such to its readers and, more generally, to the public. More than a boundary or a sealed border, the paratext is,

Gérard Genette, Introduction à l’architexte (Paris: Seuil, 1979). Gérard Genette, The Architext: An Introduction, trans. J. E. Lewin (Berkeley: University of California Press, 1992), 81, cf. 81–2 for discussion of “transtextuality.” 5 Genette, Palimpsestes: la littérature au second degré (Paris: Seuil, 1982). “Paratextuality” appears on p. 82 in The Architext, but is defined in the way that hypertextuality is later developed in Palimpsestes as the study of pastiche and parody. Genette introduces four types of transtextuality in Introduction à l’architexte and then develops this into five in Palimpsestes where the meaning of “paratext” that this essay focuses on is introduced. 6 Gérard Genette, Palimpsests: Literature in the Second Degree, trans. C. Newman and C. Doubinsky (Lincoln: University of Nebraska Press, 1997), 3. 7 Genette, Seuils (Paris: Seuil, 1987). 3 4

Patterns of Correction as Paratext

203

rather, a threshold, or…a ‘vestibule’ that offers the world at large the possibility of either stepping inside or turning back.8

The bulk of Seuils contains Genette’s attempt at inventorying, analysing, and categorizing the various types of paratextual objects that made their way into the Republic of Letters, but that had been largely ignored in terms of their literary function.9 Before beginning that task, he offered some important definitions as well as a means of defining the “status of a paratextual message” in terms of its “spatial, temporal, substantial, pragmatic, and functional characteristics” that answer the where, when, how, from whom, to whom, and to do what questions.10 It is worth a quick look at these key concepts as they are used again later in this chapter.11 The spatial characteristic concerns the location of the paratext in relation to the main text and has two divisions: the peritext and the epitext. Peritexts are those paratextual elements such as titles, prefaces, dedications, and chapter headings, located within the book itself. Epitexts are those elements that exist outside the book, at least initially, such as authorial correspondences, personal notes, and media interviews. The temporal aspect of paratexts deals with the timing of the appearance of the paratext in relation to the first appearance of the text.12 There are prior paratexts, for instance publication announcements, that appear before the initial publication. The most common temporal paratexts are the original paratexts, such as the preface, that appear at the same time as the original publication. Then there are paratexts that appear subsequent to the original publication, which fall into two classes: the later paratext and the delayed paratext. No strict definitions are offered for their difference, only the understanding that a preface written for a second edition issued shortly after the first is necessarily different than one written some decades later.13 This distinction also calls for differentiating whether the paratext was introduced in the author’s lifetime or after, and for these Genette uses the terms “anthumous” and “posthumous,” respectively. It should also be remembered that paratexts are not permanent. They can be changed or removed in later editions of printed works or erased from handwritten works. The substantial characteristic of a paratext refers to the form which the paratext takes. Typically, a paratext is also a text. Other, non-textual substances are 8 Gérard Genette, Paratexts: Thresholds of Interpretation, trans. J. E. Lewin (Cambridge: Cambridge University Press, 1997), 1–2. 9 Cf. Genette, Paratexts, 13–4. 10 Genette, Paratexts, 4. 11 The following few paragraphs are summaries of the more detailed accounts given by Genette in Paratexts, 4–13. 12 I have intentionally left “first appearance of the text” vague, as did Genette, to avoid discussions of “original text” that plague textual criticism, especially of the New Testament, as well as literary studies; cf. Genette, Paratexts, 5, esp. n.9. 13 Genette, Paratexts, 5–6.

204

Jacob W. Peterson

illustrations that might accompany a work, the typeface and formatting of the text itself, and so-called “factual” paratexts. This latter item requires explanation. There are facts that surround works that undoubtedly influence the reader’s reception of them. For instance, knowing the genre and historical time period in which a piece was created shapes the way it is understood. Other similar factors are the author’s age, sex, personal history, and education level. Many of these evade readers and the importance of each factor varies, but Genette is correct to note that “in principle, every context serves as a paratext.”14 The pragmatic aspect of a paratext deals with the sender and their degrees of authority, the addressee, and the illocutionary force of the message. Concerning the sender, there are three types of paratext: authorial, publisher’s, and allographic. The first two are clear enough, but the third needs clarification. Genette asserts that “by definition, something is not a paratext unless the author or one of his associates accepts responsibility for it, although the degree of responsibility may vary.”15 Accordingly, an “allographic” paratext cannot refer to just anything written by someone other than the author or publisher; it must have been written with their acceptance. This naturally lead to issues over responsibility for a paratext’s message. For this, there are two categories: official and unofficial. An official paratext is one produced by the author or publisher, for which they bear responsibility. An unofficial paratext generally comes from the authorial epitext or an allographic writer, for which the author can evade responsibility by claiming misrepresentation, unintended publication, privacy, or any manner of separation from ultimate responsibility. There are two main types of addressees to highlight. The first is reached by the public paratext, which is all those elements that are open for anyone to read and interact with. The private paratext, as the public paratext’s antithesis, relates to those elements reserved for a smaller audience, even to the author alone in a private diary. Finally, the illocutionary force of a paratext concerns what the paratext communicates. A paratext, among other things, can communicate information, an interpretation, and can even be performative. The final aspect of a paratext is its functional characteristic. Whereas a paratext typically is described by one term from a prescribed list at the exclusion of others for its spatial, temporal, substantial, and pragmatic aspects, it may have many functional characteristics. Furthermore, while the gamut of paratexts in a work might be delayed allographic, each particular paratext will likely have its own specific function (i.e. chapter numbers will have a purpose distinct from notes or a preface). The diversity of function precludes easy categorization and each paratext should be evaluated individually within the work, although fundamental themes are employed regularly by authors.

14 15

Genette, Paratexts, 8. Genette, Paratexts, 9; cf. 3.

Patterns of Correction as Paratext

205

1.2 Paratexts and New Testament Studies Genette’s work opened a new realm for scholarly research and formed the conceptual basis for many subsequent studies,16 but it is the study of paratexts in manuscripts, and New Testament manuscripts in particular, that is central to the present study.17 Genette’s study was concerned with print culture, which, unsurprisingly, has a different production process and aim than hand written copies of books in a pre-printing press age. This difference has been nicely stated by Matti Peikola: In that [print] context, the various paratextual elements mediating a book to its reader emerge essentially as devices conveying ideas of the author and those of the producer (publisher, printer, translator, editor etc.) as to how the book and its texts are to be consumed. In manuscript production, however, which in fifteenth-century England continued to be largely based on individual acts of commissioning rather than on speculative trade, the paratextual packaging of a book may also, importantly, reflect the needs of a specific consumer – the individual (or collective) patron, owner, or reader. For this reason, it can be argued, the paratexts of manuscript books can potentially provide more direct information concerning individual reading practices than can be inferred from their printed equivalents.18

In sum, the study of manuscript paratexts shifts the focus from the author to the receiving community. As will be discussed in the next section, the “author” (as scribe) of the paratexts will, in many cases, be a part of the receiving community. Paratexts have received a good amount of attention in biblical studies in recent years with a number of important studies published. The paratexts so far explored in New Testament manuscripts are typically those most recognizable as paratexts. There have been works on the titles of books,19 particular features in the

16 See, for instance, Helen Smith and Louise Wilson, eds., Renaissance Paratexts (Cambridge: Cambridge University Press, 2011) and the repeated reference to Genette in Laura Jansen’s introduction to her edited volume, The Roman Paratext: Frame, Texts, Readers (Cambridge: Cambridge University Press, 2014). 17 For studies on non-biblical manuscript paratexts, see, for instance: Menico Caroli, Il titolo iniziale nel rotolo librario greco-egizio: con un catalogo delle testimonianze iconografiche greche e di area vesuviana (Bari: Levante, 2007); Giovanni Ciotti and Hang Lin, eds., Tracing Manuscripts in Time and Space through Paratexts, Studies in Manuscript Cultures 7 (Berlin: De Gruyter, 2016); Francesca Schironi, ΤΟ ΜΕΓΑ ΒΙΒΛΙΟΝ: Book-ends, End-titles, and Coronides in Papyri with Hexametric Poetry, American Studies in Papyrology 48 (Durham: American Society of Papyrologists, 2010). 18 Matti Peikola, “Manuscript Paratexts in the Making: British Library MS Harley 6333 as a Liturgical Compilation,” in Discovering the Riches of the Word: Religious Reading in Late Medieval and Early Modern Europe, ed. S. Corbellini, M. Hoogvliet, and B. Ramakers (Leiden: Brill, 2015), 44–5. 19 E.g. Simon J. Gathercole, “The Titles of the Gospels in the Earliest New Testament Manuscripts,” ZNW 104/1 (2013): 33–76.

206

Jacob W. Peterson

textual tradition of a single book,20 on features of individual manuscripts,21 and several studies on the various reference schema, summaries, divisions, arrangements and collections of the various New Testament corpora.22 Additionally, there is an ongoing research project dedicated to cataloguing the paratexts of the Greek New Testament manuscript tradition.23 One consistently recurring feature of manuscripts that has thus far not been addressed as a paratext is corrections.24 In the following section, I propose a new method for studying corrections in manuscripts as paratexts.

2. A New Approach The study of corrections in manuscripts has historically aimed at cataloguing them, determining which hand was responsible, and analysing the effect of the corrections. This new approach explores the frequency and location of corrections in manuscripts to see if patterns emerge.25 Just as scholars are able to study the habits of individual scribes to create a profile of their interests and the types of copying errors they were likely to introduce,26 the argument advanced here is 20 E.g. Garrick V. Allen, “Paratexts and the Reception History of the Apocalypse,” JTS (forthcoming 2019). 21 E.g. Greg Goswell, “Ancient Patterns of Reading: The Subdivision of the Acts of the Apostles in Codex Sinaiticus,” JGRChJ 7 (2010): 68–97; Peikola, “Manuscript Paratexts in the Making,” 44–67. 22 E.g. Matthew R. Crawford, “Ammonius of Alexandria, Eusebius of Caesarea and the Origins of Gospels Scholarship,” NTS 61/1 (2014): 1–29; T. J. Lang and Matthew R. Crawford, “The Origins of Pauline Theology: Paratexts and Priscillian of Avila’s Canons on the Letters of the Apostle Paul,” NTS 63/1 (2017): 125–45; Eric W. Scherbenske, Canonizing Paul: Ancient Editorial Practice and the Corpus Paulinum (Oxford: Oxford University Press, 2013); Louis Charles Willard, A Critical Study of the Euthalian Apparatus, ANTF 41 (Berlin: De Gruyter, 2009). 23 The project is “Paratexts of the Bible” (paratexbib.eu, accessed 4 March 2018). 24 To my knowledge, the lone exception in which corrections play a role in paratextual analysis is Emanuel Tov, “Paratextual Elements in the Masoretic Manuscripts of the Bible Compared with the Qumran Evidence,” in Antikes Judentum und Frühes Christentum. Festschrift für Hartmut Stegemann zum 65. Geburtstag, BZNW 97, ed. B. Kollmann, W. Reinbold, and A. Steudel (Berlin: De Gruyter, 1999), 73–83. Even here, though, corrections play a supporting role in the overall analysis and are not studied as paratexts proper. 25 “Location” means more than the physical location of the correction (e.g. fol. 64→, interlinear between L4 and L5). The location is the content and contextual location (e.g. a correction at Gen 1:4 would have a content location of the creation of light and a contextual location of the creation narrative). 26 See for instance Edgar B. Ebojo, “A Scribe and His Manuscript: an Investigation into the Scribal Habits of Papyrus 46 (P. Chester Beatty II––P. Mich. Inv. 6238)” (PhD Thesis, University of Birmingham, 2014); Juan Hernández Jr., Scribal Habits and Theological Influences in the Apocalypse: the Singular Readings of Sinaiticus, Alexandrinus, and

Patterns of Correction as Paratext

207

that corrections, particularly when viewed collectively, may be indicative of the communities that inserted them. For a modern analogy, it is possible to look at a well-worn book to see where a reader spent most of his time. Soiled edges at specific places, dog-eared pages, marginal notes, underlines, and highlights are all indicators of a reader’s interest in select passages as well as broad areas of the text.27 In the same way, the presence of corrections in a manuscript, and especially their grouping around certain types of passages, reveals a marked interest in and use of those passages from which a profile of its users can be built. The method outlined below attempts to systematize a means of building this profile in suitable manuscripts. 2.1 Corrections as Paratexts Analysing corrections as paratexts creates a number of difficulties. Perhaps foremost among these is Genette’s requirement that “the author or one of his associates accepts responsibility for it.”28 The initial step away from such a requirement was discussed above when noting the shift in focus when studying manuscript paratexts from the author to the receiving community for which the manuscript was created. The problem remains, however, that corrections to a text are distinct from the text as initially prepared. Accordingly, a few further clarifications are needed in order to properly situate corrections within broader examinations of paratexts. Corrections in literary manuscripts, except in rare cases, certainly fail to have authorial acceptance in the Genettian sense.29 This is due, in part, to the fact that the very notion of “authorial acceptance” is itself problematic when dealing with manuscripts of ancient literary corpora. Paratexts in New Testament manuscripts, for example, are the work of a scribe who might be working more than a thousand years after the work’s initial composition. A scribe will invariably fail at perfectly transmitting his exemplar for a variety of reasons. When this happens, clear cut notions of an “author” begin to diminish. The original author still has claim to the text (e.g. the apostle Paul to Romans), but the subsequent scribe

Ephraemi, WUNT 2/218 (Tübingen: Mohr Siebeck, 2006); James R. Royse, Scribal Habits in Early Greek New Testament Papyri, NTTSD 36 (Leiden: Brill, 2008). 27 In the reverse direction, we should not be surprised to find a scholar of 1 Corinthians have their Nestle-Aland text show signs of heavy use there, but also show wear in 2 Corinthians and other parts of Paul, with lighter use in the Gospels and at the places in their LXX/MT texts from which Paul quoted. 28 Genette, Paratexts, 9. Genette does allow for one break in this definition at p. 5 n.8. 29 These exceptions are, of course, autographic manuscripts that have been corrected by the author. The qualification of “literary” manuscripts is made here to distinguish from documentary manuscripts, which are generally unlikely to have been copied and are therefore more likely autographic.

208

Jacob W. Peterson

has, perhaps unwittingly, created something new in that particular manuscript.30 Further still, later correctors adjust the original text of the author and the text transmitted by the scribe. In this way, each subsequent corrector acts as an author, even if they do not view themselves as such, since inherent in the copying process is the introduction of changes due to error and by consciously or unconsciously shaping the text to their own tradition. This idea is not new, for Luciano Canfora has previously described scribes as “the true author of the texts that managed to survive” and compares their changes to translators who fill in voids stating, “so the copyist supplements, believing to improve it, a text in which he is perfectly identified: precisely because he is a copyist.”31 The underlying assumption in studies on scribal habits is that scribes had an effect on a particular manuscript to a sufficient extent that the end product reveals something about that scribe. The recognition of even later users (i.e. correctors) as another type of author seems a natural extension of the principle. Just as we can learn about scribes by their habits, the effects of these later users also reveal something about them and their interests. The noticeable traces from scribes and later users are what Vito Lorusso referred to when writing that, In fact, as borderlands of the text, paratexts are possibly the main sources from which one can retrieve information about the temporal and spatial context in which manuscripts were 30 At the extreme, one only needs to think of the Western text of Acts as an example of a non-Lukan, and thus non-authorial, scribal creation. Schmid’s well-founded critique of Ehrman and Kannaday’s creation of a “scribes as authors” category focuses on intentionality as a marker of authorship; cf. “Scribes and Variants – Sociology and Typology,” in Textual Variation: Theological and Social Tendencies?, ed. H. A. G. Houghton and D. C. Parker (Piscataway: Gorgias, 2008), esp. 3–9. “Authorship” for the present study encompasses, irrespective of intent, the creation of something new via changes to a base text that affect later readership. Thus, scribes act as “authors” in this wider sense by introducing variants and using paratexts in the form of page layouts, title, inserting readers aids, etc. that fundamentally shape how a text is read. Certainly, some of these features will be inherited, but not all since no text is perfectly copied and new errors are introduced with each successive stage. The rest of Schmid’s chapter on editorial variants and readers’ notes, items requiring intent, works well with the point I make in the following paragraphs about correctors as another type of author. Relevant to the focus on P46 in this chapter, see the discussion by Holmes on how early readers of the Pauline corpus in turn affected its transmitted text in “The Text of P46: Evidence of the Earliest ‘Commentary’ on Romans?” in New Testament Manuscripts: Their Texts and Their World, TENT 2, ed. T. J. Kraus and T. Nicklas (Leiden: Brill, 2006), 189–203. 31 Translations are my own; from the original, “Il vero artefice dei testi che sono riusciti a sopravvivere” and “così il copista integra, credendo di perfezionarlo, un testo in cui si è perfettamente identificato: appunto perché copista.” Luciano Canfora, Il copista come autore (Palermo: Sellerio, 2002), 15–6 cf. wider discussion, esp. 15–24. Similarly, Caroli writes about P.Würzb. inv. 1, “In definitiva, il titolo del papiro, che non trova riscontri nella tradizione indiretta, rappresenta uno di quei casi in cui, per dirla con Canfora, il copista è nello stesso tempo ‘autore’ di una parte, tutt’altro che marginale, dell’opera letteraria, qual è il titolo” (Il titolo iniziale nel rotolo librario greco-egizio, 146).

Patterns of Correction as Paratext

209

produced and used. In particular, paratexts help reconstruct the history of an object that might have passed through several hands on what may have been a long journey before reaching the library in which it is preserved today.32

Correctors, then, should be understood as fulfilling an authorial role because their corrections and marginal notes create something new, albeit derivative. Kenney writes that a reader needing to fix the errors in his copy “had virtually to make his own edition of his book.”33 No one will equate this work with that of the original author, but receiving communities have shaped their texts with the end result being a text that perhaps never existed before and never again. Importantly, these later interactions with the text may offer us a window into the world in which they happened. It may be useful to pause briefly before setting out a method for studying this corrective activity in order to define the paratextual status of corrections according to their spatial, temporal, substantial, pragmatic, and functional characteristics. A few of these can be described globally for corrections in general. The spatial element is, perhaps, the most straightforward, with corrections always occupying the peritext of the manuscript. Corrections are inserted into the margins, between lines, written over the text, or represented by various critical sigla for deletions (e.g. expunging dots, slashes, cross outs, etc.). As for the substantial characteristic, the only concession to be made against all corrections being textual is for certain types of deletions. Erasures or physically cutting out a portion of text might be classified as material in substance, but this perhaps nit-picks too much since the effect is still a textual one. Some factual substances might be known about a corrector. For instance, a colophon might indicate the corrector’s name and other details and, in some cases, a date for the corrections can be assigned on palaeographical grounds. These can potentially aid our understanding of why correctors made the changes they did based on the wider cultural milieu. Concerning the functional aspect of corrections, Genette’s initial comment about paratexts in general proves insightful. He writes that the paratext “is at the service of a better reception for the text and a more pertinent reading of it (more pertinent, of course, in the eyes of the author and his allies).”34 Any correction, whether to fix an obvious error or to introduce a “better” interpretation serves to provide an improved reading experience or clearer understanding of the text as

Vito Lorusso, “Locating Greek Manuscripts through Paratexts: Examples from the Library of Cardinal Bessarion and other Manuscript Collections,” in Tracing Manuscripts in Time and Space through Paratexts, ed. G. Ciotti and H. Lin (Berlin: De Gruyter, 2016), 224. 33 E. J. Kenney, “Books and Readers in the Roman World,” in The Cambridge History of Classical Literature: Volume 2: Latin Literature, ed. E. J. Kenney and W. V. Clausen (Cambridge: Cambridge University Press, 1982), 18. 34 Genette, Paratexts, 2. 32

210

Jacob W. Peterson

the corrector sees it. The remaining characteristics, temporal and pragmatic, need to be described on the basis of the corrector involved.35 To this end, there are three distinct types of correctors to be discussed: the original scribe, the diorthotes, and later users. The original scribe’s corrections breakdown into two further subcategories. The first type are in scribendo corrections, which are those created in the process of writing where a mistake is immediately corrected. The second type of corrections are those that come about during revision or double-checking the finished product. Both types of corrections have the same pragmatic status as authorial, official, public paratexts. Despite an actual temporal difference between the two types of corrections, both count as original paratexts since they appear at the same time as the final form of the text.36 The corrections by the diorthotes are official, public paratexts. Their corrections might best be related to the publisher’s paratext since they amount to a final polish of the text’s presentation before sending out the finished manuscript. The diorthotes’ corrections are also original paratexts since they accompany the initial publication of the text and, except for presumably rare cases, are anthumous. The later correctors can be assessed individually or as a collective. The main difference in paratextual status between the correctors will be temporal. Both because this difference is largely inconsequential and the focus of this study is overall patterns of correction, it is prudent here to assess them collectively. As a collective these corrections have a delayed, and more often than not, posthumous temporal status. Things get more confused as we consider their pragmatic status. Later corrections have two pragmatic paratextual statuses depending on how their authorial status is judged. They are at once unofficial allographic paratexts, since they lack the original scribe’s or author’s consent and authorial, official paratexts if they are extended co-authorial status based on their creation of a new text. Reaching a decision on which status is the dominant one will impact whether the corrections are studied primarily as separable events (unofficial allographic) or as part and parcel of the creation of a new text (official authorial).37 The former views corrections as singular, largely disconnected events whereby the original text (the scribe’s text) is fixed or modified. The latter views corrections as part of a larger, connected enterprise of shaping large sections of the text for a particular community, thereby creating an essentially new text. This study elevates the second perspective.

35 The illocutionary force is not stated for each type of corrector since that must be assigned on a correction-by-correction basis. 36 The only time their real temporal difference would matter is in cases where the scribe consulted a different exemplar to check against his copy, although this would not affect their temporal classification. 37 Neither view should be considered the “correct” view; rather, each has a certain utility depending on the study.

Patterns of Correction as Paratext

211

2.2 Method for Studying Patterns of Correction as Paratexts The purpose of this chapter is the development of a method for viewing the corrective activity in manuscripts as potentially indicative of the community that used them. As the title indicates, it is not individual corrections that are of ultimate importance but, rather, patterns of correction. The basis for the idea is that individual corrections are not random, but represent points where the text was being read and used. The theory, then, is that under certain circumstances patterns of correction can be indicative of the interests of the manuscript’s correctors, readers, or community.38 The following subsections outline the proposed method by first introducing the types of manuscripts for which the method is best suited, then outlining the data needed before beginning any analysis, then developing a process for mining the data, and finally describing a means of investigating the corrections for patterns. 2.2.1 Locating Suitable Manuscripts The first task is to find a manuscript that is suitable for study since the method cannot be applied to every manuscript. The most obvious criteria are that the text must be corrected and must be extant for a large section of text.39 The next criterion of a suitable manuscript is signs of sporadic corrections. At this point we have reached a tripartite catch-22 with the next two steps (covered in Sections 2.2.2–3). Determining if the corrections are likely sporadic requires a decent understanding of the corrective activity as well as an understanding of where the corrections are within the manuscript. Fortunately, an incomplete understanding of these two points is sufficient for the moment and a more complete understanding can be delayed until the next stage.40 Perhaps the simplest method for doing this is quickly scanning through images or transcriptions of a manuscript and noting whether there is a steady flow of corrections or gaps between them. Caution should be exercised when consulting a list of corrections since, in general, these lists are abstracted away from the

38 As a parallel to scribal habits, the tendency to correct certain portions of a manuscript and not others might be called “correcting habits.” 39 A manuscript containing only a few fragments is not large enough to determine what a community was and was not most interested in based on corrections and the lack thereof. The more extant the manuscript the better, with confidence in the results being proportional to the amount of text the manuscript contains. 40 This first step could be skipped, but it is intended to be a quick method for finding promising manuscripts. Step two is quite laborious, so step one is aimed at preventing wasting time on less well-suited manuscripts. A manuscript that does not pass this stage should not necessarily be eliminated from consideration forever, however. It is possible that the corrections by the scribe and diorthotes distort initial observations about the manuscript’s corrections.

212

Jacob W. Peterson

physical layout of a manuscript, thereby making the process prone to incorrect impressions. A key clarification is that both the total number of corrections and the rate of correction are insufficient and misleading guides for manuscript selection. As a thought experiment, assume a 100-page manuscript. If it has 200 corrections, then the correction rate is two corrections per page. It is easy to mistake that rate as indicating there are two corrections on every page. However, it is impossible to know by the rate alone whether there are two corrections on every page or twenty-five corrections on eight pages. The former has no discernible spacing between corrections whereas the latter does. Alternatively, if the manuscript has fifty corrections, then the rate is one correction for every two pages. Yet, this lower rate does not mean the corrections are necessarily any more spread out. They could be every other page as the rate allows or they could cluster on a few pages. Thus, neither the total number of corrections or the rate of correction are indicators of a good manuscript for this method. While the method can work with any number of corrections as long as they are sporadic, this is not to say that a manuscript with few corrections is just as good as one with copious corrections. More corrections, as with a more complete manuscript, allows for more secure results, but it is not an a priori indicator of a more suitable manuscript. In sum, the first task is to find a suitably large manuscript with corrections and then to gain a good initial understanding of the corrective activity so as to determine whether they are evenly spread or more sporadic in their occurrences. If a manuscript passes this stage, it can proceed to the more thorough vetting of the following stages. 2.2.2 Gathering the Data Once a manuscript has been determined to be a likely candidate for studying patterns of corrections, the next step is to create a comprehensive list of the corrections. Thankfully, such a list will have already been produced for many of the most important manuscripts, which will only need to be confirmed and updated with new findings. In manuscripts with corrections by more than one hand the corrections need to be assigned as far as is possible. Assigning hands is likely the step that introduces the most amount of variability in the final results. There are other paratexts beside the corrections that can be collected at this stage to assist later analyses. Marginal notes are often indicative of a more prolonged interaction with a particular piece of text and, therefore, should be noted. Similarly, the inclusion of commentary at particular points indicates the potential importance of that text. Reading aids, particularly when sporadically included, are especially pertinent. These can range from the simple reading marks that indicate sense-line divisions to more formalized marginal apparatus (e.g. Eusebian Canons).

Patterns of Correction as Paratext

213

2.2.3 Mining the Data The next step is to plot the paratextual data to facilitate investigating patterns in the corrections. Using the data generated in the previous stage, the corrections and other paratextual data should be organized by hand and manuscript page. Organizing the data by recto and verso of each folio allows the corrective activity to be visualized and analysed more readily. The hands are recorded so that the corrections of the scribe and diorthotes can be disregarded, since they are most likely not indicative of the community that later used the manuscript. Their task was to produce a complete manuscript and their corrective activity should be void of the types of patterns sought in this study.41 This process allows for the work of the later correctors to be isolated. It will then be immediately recognizable if the corrections are sporadic or evenly distributed within the manuscript. Additionally, noting recto and verso locations for corrections can provide explanations for corrections that do not obviously fit a pattern. A correction may be viewed as less random and problematic for the pattern when it is recognized as being on the page opposite of other corrections when the manuscript was opened.42 2.2.4 Investigating the Patterns If a manuscript has passed through the previous stages and shows signs of correction spread, the corrections can then be investigated for thematic patterns. Whereas a singular correction is merely the work of a hand in a particular area, a pattern is the aggregation of corrections around a common theme. The basic process is to read the passages surrounding each correction to find if there are common topics or if each deals with a different aspect of a larger theme. The scope should not be restricted to just the passage containing a correction, but should be expanded to include the content between corrections that appear close together. The premise behind this is that it is unlikely that the first correction in a cluster represents the first word read and the last correction the last. Accordingly, the surrounding content on those particular pages should be included in the thematic study. In a manuscript that was corrected by multiple hands, any overlap in their corrective activity strengthens the notion that the topic covered was important 41 While a scribe (and the diorthotes) could have been tasked with producing a manuscript for a particular community and shaped its readings accordingly, this would hardly be recognizable in the corrections. That type of tampering with the text would be uncovered by studying the textual variants, as in E. J. Epp, The Theological Tendency of Codex Bezae Cantabrigiensis in Acts (Cambridge: Cambridge University Press, 1966). On the relationship between scribes and users in early Christian communities, see Schmid, “Scribes and Variants,” 10–3. 42 The assumption here is that if a corrector was working on the verso of a folio, he might notice and correct an error on the recto of the following folio.

214

Jacob W. Peterson

for the receiving community. Additionally, it is conceivable that a manuscript would have a sufficient number of corrections by each hand that individual patterns might be distinguishable alongside a collective pattern. The method thus proposed is far from an exact science and can easily lend itself towards identifying patterns that are not there. As far as possible, expectations and wishes must be tempered and conclusions expressed only as confidently as the evidence permits. External verification for the findings should be sought when possible. If the patterns of correction suggest a certain theme, it may be possible to corroborate that theme’s importance at the time(s) of correction by looking into the wider cultural and specific local milieus.

3. Application of New Method: Patterns of Correction in P46 With the above considerations in mind, the method of studying patterns of corrections can now be applied to the earliest manuscript of the Pauline Epistles, P46, as a test case. This section will follow the order prescribed: first offering an initial assessment of the suitability of P46 for the study, then gathering the relevant paratextual data, next mining that data, and finally investigating the findings for any thematic patterns. 3.1 An Initial Look at the Suitability of P46 Ebojo remarked in his doctoral thesis that “because of the interplay of many hands, however temporally distant from one another, P46 becomes an even more interesting manuscript that unveils quite graphically the sociology of ancient book production.”43 Yet, the study of corrections in P46 has historically been concerned with cataloguing them, determining which hand was responsible, and analysing their effect on the text. Ebojo advanced the discussion by considering how the corrections were made. It has been widely recognized that the corrections in the manuscript decrease sharply as it goes on, with the bulk of the corrections being to the text of Hebrews. While the correction rate was determined by Ebojo to be “a little over one correction every page,” he rightly noted this point as misleading since 65% of the corrections occur in the first half of the manuscript and, by his count, only 102 of the 172 pages (59%) contain corrections.44 These statistics indicate that the corrections are not evenly distributed but are spread in P46, giving it the appearance of a good candidate for investigating patterns of correction. Explanations for the decreasing and scattered corrections have not been seriously attempted. This gap in the literature led Ebojo to remark, “what is noticeable, however, in both Zuntz’s and Royse’s observations is their 43 44

Ebojo, “A Scribe and His Manuscript,” 322. Ebojo, “A Scribe and His Manuscript,” 292–93.

Patterns of Correction as Paratext

215

absence of discussion as to the probable reasons for such selective correction events, and their implications to the principle of ‘text guarding’ role of corrections.”45 Indeed, Ebojo’s own conclusion was that “‘other hand’ corrections in P46 were done randomly and unsystematically, i.e., spot-checking.”46 The following application of the method for studying patterns of correction aims at showing randomness to be an insufficient explanation. 3.2 Gathering the Data in P46 There are two relevant types of paratexts in P46 that have to be accounted for in this section: the corrections and reading marks. 3.2.1 The Corrections In a recent study on the corrections in P46, I have proposed several new corrections and eliminated a few previously identified corrections. The study resulted in an updated total number of corrections of 202.47 The breakdown of the corrections is given in Table 1.48 The overall distribution of the corrections remains largely the same as in previous studies, with Hebrews receiving disproportionate attention from correctors (43.6%). It should be remembered, however, that the percentages displayed are relative since the books vary considerably in length and completeness in P46. Even still, Hebrews has more corrections than the two Corinthian letters combined, both of which are fully preserved and roughly commensurate in length. This raw data and the updated record of the corrections can now be used to look for patterns of corrections in the manuscript.

Ebojo, “A Scribe and His Manuscript,” 295 n.20. Ebojo, “A Scribe and His Manuscript,” 295, cf. 307, 322. Ebojo groups the second to fifth hands as “other hands” with the third through fifth sub-classified as “non-contemporaries” or “later hands.” Further down this same page, the suggestion of “incompetence in doing a ‘good’ job” is made. This notion will be picked up in my own discussion in §3.3.1 on which hands to include in the investigation. 47 This table originally accepted for publication in Jacob W. Peterson, “An Updated Correction List for Chester Beatty BP II—P.Mich. Inv. 6238 (Gregory–Aland Papyrus 46 [𝔓46]),” BASP 56 (2019). 48 The scheme devised by Ebojo was used for hand identifications; cf. “A Scribe and His Manuscript,” 301–19, 638–55. For the table, ‘M1’ is the first hand to work on the manuscript (i.e. the original scribe), ‘M2’ is the second hand (i.e. the first corrector), ‘M3’ is the third hand, and so forth. ‘ED’ refers to those corrections that were “extremely difficult” to assign to a hand. 45 46

216

Jacob W. Peterson

Table 1. Distribution of Corrections by Hand and Book Romans Hebrews 1 Corinthians 2 Corinthians Ephesians Galatians Philippians Colossians 1 Thessalonians Total %

M1 12 22 28 17 8 9 6 3 0 105 52.0%

M2 0 56 9 1 0 0 0 0 0 66 32.7%

M3 2 3 3 0 0 0 0 0 0 8 4.0%

M4 1 1 1 3 0 0 0 1 0 7 3.5%

M5 1 2 1 1 0 0 0 0 0 5 2.5%

ED 1 4 2 2 0 0 2 0 0 11 5.4%

Total 17 88 44 24 8 9 8 4 0 202

% 8.4% 43.6% 21.8% 11.9% 4.0% 4.5% 4.0% 2.0% 0.0%

3.2.2 Other Paratexts in P46 One of the immediately recognizable features of many leaves in P46 are small tick marks and dots placed throughout the text. The marks appear heavily in Romans and Hebrews, at the end of 1 Corinthians, the opening page of 2 Corinthians, and a couple of pages in Philippians. Sanders is the first to make mention of them and concludes that, although they are similar to the hand that wrote the page numbers, the ink indicates they are from a hand different from the main scribe or the corrector. His conclusion from this is that they were likely inserted “by some later reader for his convenience, when reading in the church.”49 Kenyon subsequently indicates that these were “added by another hand, perhaps that which has inserted the page-numeration.”50 Zuntz later correctly distinguishes between two sets of reading marks. The first, distinguished by dark ink, were by the ex-officio corrector in his scheme who also added the page numbers and stichoi. The second were by an unidentified “other hand” that used pale ink.51 Most recently, Ebojo has attributed the two sets of reading marks to the fourth and fifth hands (M4 and

Henry A. Sanders, A Third-Century Papyrus Codex of the Epistles of Paul, University of Michigan Studies Humanistic Series 38 (Ann Arbor: University of Michigan Press, 1935), 17. 50 Frederic G. Kenyon, ed., The Chester Beatty Biblical Papyri, Fasciculus III, Supplement: Pauline Epistles, Text (London: Emery Walker, 1936), xiv. 51 Günther Zuntz, The Text of the Epistles: A Disquisition upon the Corpus Paulinum (London: Oxford University Press, 1953), 253. Zuntz’s discussion is less than clear, with n.3 being the key for piecing the clues together. The footnote is attached to a discussion of the ex-officio corrector that lacks mention of that hand inserting reading marks. The discussion of the pale ink marks as outliers only makes sense in comparison to other darker reading marks that are not mentioned, though possibly indicated in n.5. Zuntz’s “other hand” is Ebojo’s M4 based on their identification of the hand as the same that made the correction at 1 Cor 16:7. 49

217

Patterns of Correction as Paratext

M5).52 The range of opinions here, particularly as they relate to the role of the second hand that also added the page numbers and stichoi, is potentially problematic. If the second hand was a diorthotes and added reading marks, then these should be excluded from consideration per the method developed in section 2.2.3. 3.3 Mining the Data in P46 Now that all of the relevant paratextual features in P46 have been accounted for, the process of plotting the data and investigating it for thematic patterns can begin. The following table plots the corrections and reading marks in P46 according to the hand that made them and the folio on which they appear. The data is mapped out by recto and verso of each leaf of P46, shaded according to the intensity of corrections by each hand, and it is indicated whether or not reading marks are present on a particular page. Table 2. Distribution of Corrections by Folio Folio P46 Verses M1 1a–7b (unnumbered initial page–ιγ) missing 8a ιδ Rom 5:17–6:3 0 8b ιε Rom 6:5–14 0 9a–10b (ιϛ–ιθ) missing 11a κ Rom 8:15–25 0 11b κα Rom 8:27–35 0 12a κβ Rom 8:37–9:9 1 12b κγ Rom 9:9–22 2 13a κδ Rom 9:22–32 0 13b κε Rom 10:1–11 0 14a κϛ Rom 10:12–11:2 0 14b κζ Rom 11:3–12 0 15a κη Rom 11:13–22 0 15b κθ Rom 11:24–33 1 16a λ Rom 11:35–12:9 1 16b λα Rom 12:10–13:1 0 17a λβ Rom 13:2–11 2 17b λγ Rom 13:12–14:8 3 18a λδ Rom 14:9–21 1 18b λε Rom 14:22–15:10 0 19a λϛ Rom 15:11–19 0 19b λζ Rom 15:20–29 0 20a λη Rom 15:29–16:3 0 20b λθ Rom 16:4–13 1 21a µ Rom 16:14–23 0 21b µα Rom 16:23–Heb 1:7 0 22a µβ Heb 1:7–2:3 0 22b µγ Heb 2:3–11 2 52

M2

M3

M4

M5

ED

Marks

0 0

0 0

0 0

0 0

0 0

y y

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0

y y y y y y y y y y y y y y n n y y y y n y y y

Ebojo, “A Scribe and His Manuscript,” 303 and 304 n.62.

218 Folio 23a 23b 24a 24b 25a 25b 26a 26b 27a 27b 28a 28b 29a 29b 30a 30a 31a 31b 32a 32b 33a 33b 34a 34b 35a 35b 36a 36b 37a 37b 38a 38b 39a 39b 40a 40b 41a 41b 42a 42b 43a 43b 44a 44b 45a 45b 46a 46b 47a 47b 48a

Jacob W. Peterson P46 µδ µε µϛ µζ µη µθ ν να νβ νγ νδ νε νϛ νζ νη νθ ξ ξα ξβ ξγ ξδ ξε ξϛ ξζ ξη ξθ ο οα οβ ογ οδ οε οϛ οζ οη οθ π πα πβ πγ πδ πε πϛ πζ πη πθ ϙ ϙα ϙβ ϙγ ϙδ

Verses Heb 2:11–3:3 Heb 3:3–13 Heb 3:14–4:4 Heb 4:4–14 Heb 4:14–5:7 Heb 5:8–6:4 Heb 6:4–13 Heb 6:13–7:1 Heb 7:2–10 Heb 7:11–20 Heb 7:20–28 Heb 7:28–8:8 Heb 8:9–9:2 Heb 9:2–9 Heb 9:10–16 Heb 9:18–26 Heb 9:26–10:8 Heb 10:8–20 Heb 10:22–30 Heb 10:32–11:3 Heb 11:4–9 Heb 11:9–17 Heb 11:18–26 Heb 11:26–34 Heb 11:35–12:1 Heb 12:2–11 Heb 12:11–21 Heb 12:21–13:2 Heb 13:3–11 Heb 13:12–20 Heb 13:20–1 Cor 1:3 1 Cor 1:4–14 1 Cor 1:14–23 1 Cor 1:24–2:2 1 Cor 2:3–11 1 Cor 2:11–3:5 1 Cor 3:6–15 1 Cor 3:16–4:3 1 Cor 4:4–10 1 Cor 4:11–19 1 Cor 4:20–5:7 1 Cor 5:8–6:3 1 Cor 6:4–12 1 Cor 6:13–7:3 1 Cor 7:4–12 1 Cor 7:12–19 1 Cor 7:20–29 1 Cor 7:30–37 1 Cor 7:37–8:7 1 Cor 8:7–9:1 1 Cor 9:4–12

M1 0 2 0 0 0 0 0 0 0 1 0 0 0 2 1 2 0 0 1 1 1 2 1 1 0 2 1 1 0 0 1 0 0 0 0 1 0 2 1 1 0 0 0 1 0 2 0 0 1 3 0

M2 0 0 0 0 1 2 1 2 1 0 0 4 0 2 2 3 3 2 3 4 1 0 1 2 1 4 2 3 6 1 3 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 0 0 0

M3 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

M4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

M5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ED 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0

Marks y y y y y y y y y y y y y y y y y y y y y y y y y y y y y y n n n n n n n n n n n n n n n n n n n n n

219

Patterns of Correction as Paratext Folio 48b 49a 49b 50a 50b 51a 51b 52a 52b 53a 53b 54a 54b 55a 55b 56a 56b 57a 57b 58a 58b 59a 59b 60a 60b 61a 61b 62a 62b 63a 63b 64a 64b 65a 65b 66a 66b 67a 67b 68a 68b 69a 69b 70a 70b 71a 71b 72a 72b 73a 73b

P46 ϙε ϙϛ ϙζ ϙη ϙθ ρ – – ρα ρβ ργ ρδ ρε ρϛ ρζ ρη ρθ ρι ρια ριβ ριγ ριδ ριε ριϛ ριζ ριη ριθ ρκ ρκα ρκβ ρκγ ρκδ ρκε ρκϛ ρκζ ρκη ρκθ ρλ ρλα ρλβ ρλγ ρλδ ρλε ρλϛ ρλζ ρλη ρλθ ρµ ρµα ρµβ ρµγ

Verses 1 Cor 9:12–20 1 Cor 9:20–10:1 1 Cor 10:1–10 1 Cor 10:11–20 1 Cor 10:21–30 1 Cor 10:31–11:6 1 Cor 11:7–17 1 Cor 11:18–25 1 Cor 11:26–12:2 1 Cor 12:3–12 1 Cor 12:13–24 1 Cor 12:24–13:1 1 Cor 13:2–11 1 Cor 13:11–14:6 1 Cor 14:6–15 1 Cor 14:16–24 1 Cor 14:24–34 1 Cor 14:34–15:5 1 Cor 15:6–15 1 Cor 15:17–28 1 Cor 15:28–39 1 Cor 15:39–50 1 Cor 15:51–16:2 1 Cor 16:2–12 1 Cor 16:12–23 2 Cor 1:1–8 2 Cor 1:8–15 2 Cor 1:16–2:1 2 Cor 2:3–12 2 Cor 2:13–3:3 2 Cor 3:5–13 2 Cor 3:14–4:3 2 Cor 4:4–12 2 Cor 4:13–5:4 2 Cor 5:5–13 2 Cor 5:14–6:2 2 Cor 6:3–13 2 Cor 6:14–7:4 2 Cor 7:5–11 2 Cor 7:12–8:3 2 Cor 8:4–12 2 Cor 8:13–24 2 Cor 9:1–7 2 Cor 9:7–10:1 2 Cor 10:1–11 2 Cor 10:11–11:2 2 Cor 11:3–10 2 Cor 11:12–22 2 Cor 11:23–32 2 Cor 11:33–12:9 2 Cor 12:10–18

M1 0 0 1 0 2 0 0 0 1 0 0 1 0 1 1 1 0 1 0 3 1 0 1 2 0 1 2 0 0 0 2 1 1 0 0 0 1 1 1 1 0 1 0 0 3 1 0 0 0 1 0

M2 0 0 0 0 1 2 0 0 0 0 0 0 1 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0

M3 0 1 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

M4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0

M5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ED 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0

Marks n n n n n n n n n n n n n y y y y y y y y y y y n y n n n n n n n n n n n n n n n n n n n n n n n n n

220

Jacob W. Peterson

Folio P46 Verses 74a ρµδ 2 Cor 12:18–13:5 74b ρµε 2 Cor 13:5–13 75a ρµϛ Eph 1:1–11 75b ρµζ Eph 1:12–20 76a ρµη Eph 1:21–2:8 76b ρµθ Eph 2:10–20 77a ρν Eph 2:21–3:10 77b ρνα Eph 3:11–4:1 78a ρνβ Eph 4:2–14 78b ρνγ Eph 4:15–25 79a ρνδ Eph 4:26–5:6 79b ρνε Eph 5:8–25 80a ρνϛ Eph 5:26–6:6 80b ρνζ Eph 6:8–18 81a ρνη Eph 6:20–Gal 1:8 81b ρνθ Gal 1:10–22 82a ρξ Gal 1:23–2:10 82b ρξα Gal 2:12–21 83a ρξβ Gal 3:2–15 83b ρξγ Gal 3:16–29 84a ρξδ Gal 4:2–18 84b ρξε Gal 4:20–5:1 85a ρξϛ Gal 5:2–17 85b ρξζ Gal 5:20–6:8 86a ρξη Gal 6:10–Phil 1:1 86b ρξθ Phil 1:5–15 87a ρο Phil 1:17–28 87b ροα Phil 1:30–2:12 88a ροβ Phil 2:14–27 88b ρογ Phil 2:29–3:8 89a ροδ Phil 3:10–21 89b ροε Phil 4:2–12 90a ροϛ Phil 4:14–Col 1:2 90n ροζ Col 1:5–13 91a ροη Col 1:16–25 91b ροθ Col 1:27–2:7 92a ρπ Col 2:8–19 92b ρπα Col 2:22–3:11 93a ρπβ Col 3:13–25 93b ρπγ Col 4:3–12 94a ρπδ Col 4:16–1 Thess 1:2 94b ρπε 1 Thess 1:8–2:3 95a–96b (ρπϛ–ρπθ) missing 97a ρϙ 1 Thess 5:5–9 97b ρϙα 1 Thess 5:23–28

M1 0 0 0 1 1 2 0 0 0 1 1 0 0 1 2 2 1 1 0 0 2 1 1 0 0 2 1 0 0 0 2 1 0 0 1 1 0 1 0 0 0 0

M2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

M3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

M4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0

M5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ED 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0

Marks n n n n n n n n n n n n n n n n n n n n n n n n n y y n n n n n n n n n n n n n n n

0 0

0 0

0 0

0 0

0 0

0 0

n n

3.3.1 Which Hands Count? At the end of the last section, it was shown that Kenyon and Zuntz had attributed some of the reading marks to the hand that also inserted the stichoi and page

Patterns of Correction as Paratext

221

numbers, while Sanders and Ebojo attributed them to other hands. This problem of hand identifications for the page markers is part of a considerably more difficult problem with applying the patterns of correction method to P46. In particular, the method calls for eliminating the work of the main scribe and diorthotes from consideration. As the debate around reading marks hinted, the activity and role of the second hand presents a challenge. Not only is there debate over his role in inserting the reading marks, there is an equal divide over the authority of his role. As my ultimate conclusion on this matter will show, I believe the debate over the role of the second hand has been misguided. A recap of previous views will be helpful before covering my view of the second hand. Sanders identified the second hand as a contemporary corrector who was also responsible for the page numbers, but not for the reading marks. The stichoi are mentioned separately from the corrections so it is unclear if he intends to indicate that the hand responsible for the two are identical when describing the stichoi hand as contemporary. He leaves open the possibility that the stichoi were done by the “business manager” of a scriptorium.53 Kenyon only briefly mentioned the existence of a corrector in his second text volume on P46. Importantly, however, he distinguishes the “occasional” corrections by this hand from the person who inserted the stichoi, punctuations, and page numbers.54 Zuntz calls the second corrector the “ex-officio” and makes him also responsible for the page numbers, stichoi, reading marks, and punctuation.55 The second ANTF volume identifies the second hand as the one that added the page numbers, but only identifies the hand that inserted the reading marks as a later hand.56 Royse quotes Zuntz’s comments about “man 2” as a “useful summary” and elsewhere identifies the second hand as a corrector with access to the Vorlage of P46, thus implicating him as a diorthotes who added the page numbers and stichoi.57 Most recently, Ebojo has identified this hand as the “contemporary corrector” who “may have also been responsible for the inscription of the page numeration and the stichos notations.”58 He remarked that this hand “inscribed corrections with imposing authority” and, in quoting Zuntz’s ex-officio attribution, stated that “this sense of authority is demonstrated by what he did.”59 In sum, the majority view is of an authoritative corrector, likely a diorthotes, who was responsible for the page numbers and stichoi, but perhaps not the reading marks.

Sanders, A Third-Century Papyrus Codex of the Epistles of Paul, 15–7. Kenyon, The Chester Beatty Biblical Papyri, xiv–xv. 55 Zuntz, Text of the Epistles, 253, cf. the discussion above in my n.50. 56 Klaus Wachtel and Klaus Witte, Das Neue Testament auf Papyrus II: Die Paulinischen Briefe, Teil 2: Gal, Eph, Phil, Kol, 1 u. 2 Thess, 1 u. 2 Tim, Tit, Phlm, Hebr, ANTF 22 (Berlin: De Gruyter, 1994), L–LI. 57 Royse, Scribal Habits, 213–14, 237–39. 58 Ebojo, “A Scribe and His Manuscript,” 307–308. 59 Ebojo, “A Scribe and His Manuscript,” 308. 53 54

222

Jacob W. Peterson

If the model of scribe and diorthotes is accepted for the first two hands, they account for almost 85% of all corrections. The 66 corrections by the second hand are more than triple the collective amount of corrections by the third, fourth, and fifth hands.60 On a comparative basis, plus his role in adding the page numbers and stichoi, this might make the second hand look more like an official corrector. However, debates over the existence of scriptoria in the second–third century notwithstanding, a closer look at what the second hand actually did creates serious problems for such an identification.61 The tendency has been to describe this hand as devoting most of his time to the beginning of the codex, which potentially obscures that there are no corrections by this hand in the extant part of Romans. The first instance of a correction by the second hand, tellingly, occurs at Heb 1:1.62 This initial correction is followed by fifty-five more until they abruptly stop at the final verse of Hebrews (13:25). Thus, 85% of the corrections by this hand occur exclusively in Hebrews. The remaining ten corrections occur in 1 Corinthians (nine; first at 5:5) and 2 Corinthians (one at 7:13).63 Following the correction at 2 Cor 7:13, there are fully fifty-five (extant) pages without correction by this hand. In total, the sixty-six corrections by this hand occur on only 33 of the 172 extant pages (19.2%). Not only did this scribe incompletely correct the manuscript, the pages that were corrected were incompletely done and marred by the introduction of nonsense readings.64 Maintaining a view of this corrector as a professional requires the dismissal of a substantial amount of evidence to the contrary.65 The incomplete, inconsistent, and often nonsensical 60 Even if the eleven “ED” corrections are assigned to the third through fifth hands the disparity remains. 61 On the second–third century scriptorium debate, see, for example: Kim Haines-Eitzen, Guardians of Letters: Literacy, Power, and the Transmitters of Early Christian Literature (Oxford: Oxford University Press, 2000), 77–104 (esp. 83–96) versus Kurt Aland and Barbara Aland, The Text of the New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism, 2nd rev. ed. (Grand Rapids: Eerdmans, 1995), 70 versus Gordon D. Fee, “P75, P66, and Origen: The Myth of Early Textual Recension in Alexandria,” in Studies in the Theory and Method of New Testament Textual Criticism, ed. E. J. Epp and D. Fee (Grand Rapids: Eerdmans, 1993), 258–59; Zuntz, Text of the Epistles, 263–83. 62 Royse attributed corrections at Rom 9:25 and 12:16 to man 2. Ebojo has reassigned them; see “A Scribe and His Manuscript,” 302–304. 63 Royse attributed corrections at Eph 6:2, 6:20, Phil 1:20, and Col 3:3 to man 2. The correction at 6:20 is doubtful, while the others have been reassigned by Ebojo; see Ebojo, “A Scribe and His Manuscript,” 634, 636, 637. 64 Cf. Ebojo, “A Scribe and His Manuscript,” 309–12; Zuntz, Text of the Epistles, 253. 65 Though certainly impacted by the concentration of corrections in Hebrews where almost every page has reading marks, only seven of the sixty-six corrections by the second hand do not overlap with pages with reading marks. That is, fifty-seven corrections are on pages with reading marks and two more are on pages facing a page with reading marks. This correspondence also aids the identification of this hand with a later user (i.e. reader).

Patterns of Correction as Paratext

223

nature of this hand’s work combined with a clear interest in the text of Hebrews, especially the latter half, suggests a later user of the text rather than a diorthotes.66 3.3.2 Summarizing the Data Although there are 202 total corrections in the manuscript, this number drops to 97 when the 105 by the original scribe are removed and only the second through fifth hands are considered. Of the 172 pages in P46, corrections by these hands occur on only 54 pages (31.4%).67 The table above confirms that these corrections are not evenly spread, but heavily cluster in a few places with sporadic corrections elsewhere. The bulk of the corrections are found in the latter half of Hebrews. The reading marks are likewise clustered and are featured essentially uninterrupted from Romans through Hebrews, with a grouping at the end of 1 Corinthians and two pages in Philippians. In total, 66 of the 172 extant pages feature reading marks (38.4%). Corrections and reading marks have a relatively high degree of correlation as well. Only 26 of the 97 corrections occur on pages without reading marks. Five of those 26 are on pages opposite a page with reading marks. Thus, 78.4% of the corrections by later hands occur on pages or openings with reading marks. Looking at the table of corrections, we find the third through fifth hands were responsible for only twenty corrections. Their relationship to Hebrews is not nearly as marked as the second hand, but is still greater than might be expected. Together, these hands have six of their twenty corrections there.68 More interestingly, these hands continue to show a pronounced affinity toward pages with reading marks. Whereas 38.4% of pages were shown above to have reading marks, twelve of the twenty corrections by the third through fifth hands occur on 66 The reclassification of this hand deserves a chapter on its own. While most have identified this hand with the pages numbers and stichoi, this connection needs to be evaluated more fully. A look at folio 21→, where a page number, stichos notation, and a correction unanimously attributed to the second hand all occur, makes viewing a single hand as responsible rather suspect. If that hand is responsible for all these features, issues with the page numbers (blank opening between ρ and ρα) and the notoriously incorrect stichoi add additional, though not insurmountable obstacles to the professional credibility of the scribe. In my view, a theory of a later user adding page numbers for reference help and stichoi to give it a more professional appearance is less problematic than assuming a scribe-diorthotes working relationship (scriptorium?) in which the diorthotes has utterly failed at the task. 67 Ebojo’s overall distribution figures mentioned in §3.1 can also be updated with this data. While the 202 corrections on 172 pages results in a corrections-per-page rate of 1.17, corrections occur on only 107 pages (62%). 68 That is, six out of twenty corrections beats a purely random distribution that would see 3.95 corrections on the thirty-four pages containing Hebrews. Despite a 50% increase over expectation, modelling to determine if this is a statistically significant increase has not been performed. The more interesting point is the following one.

224

Jacob W. Peterson

pages or openings with reading marks (60%). Purely random corrections should roughly match that 38.4%, yet 60% correspond to pages with reading marks.69 Taking hands two through five together, 62 of the 86 corrections are in Hebrews and 71 of them overlap with pages or openings with reading marks. This high level of correlation evidences something other than viewing these correctors’ work as incomplete or merely random. The places where reading marks were added are indicative of where a community was reading and add to the available paratexts that are indicative of the receiving community. 3.4 Investigating the Patterns in P46 With the groundwork completed, the investigation into the corrections by the later hands in P46 can commence. In what follows I will argue that the work of these later hands suggests that the manuscript was used predominantly in a Jewish-Christian church context based on a marked interest in passages concerning Jewish-Christian identity.70 As a reminder, the focus is not on the corrections themselves, but the contexts in which they occur. Accordingly, when a verse is cited below as containing a correction, it should be read within the wider context of its page (Table 2 can be used as a reference). 3.4.1 Corrections on Pages without Reading Marks The corrections on pages or openings without reading marks can be generally grouped into two contexts. The first is what might be called general homiletical passages outlining Christian behavior and church order, which are to be expected in any church setting. In this first category, we find corrections around passages dealing with many-members-one-body language (1 Cor 12:20), generosity toward the saints (2 Cor 9:12, 14), bearing fruit (Col 1:7), spiritual wisdom and maturity (1 Cor 2:6, 15), sexual immorality (1 Cor 5:5, 6:14), not associating with unrepentant Christians and handling issues internally (1 Cor 6:2), and the promise and seal of God (2 Cor 1:19). The second context has corrections in passages dealing with Jewish-Christian identity and old covenant faithfulness. The topics addressed include eating meat sacrificed to idols, becoming all things to all people (1 Cor 9:22), the superiority of the new covenant (2 Cor 4:2), and eating meat from the market and meat sacrificed to idols (1 Cor 10:25). The

The per hand correspondence of corrections on pages with reading marks is: third hand 6/8 (75%); fourth hand 3/7 (42.9%); fifth hand 3/5 (60%). 70 As noted by Helmut Koester, the term “Jewish-Christianity” is notoriously problematic and is here intended in his “narrower” sense of a “congregation of followers of Jesus, who interpreted the law of Moses in a fashion that was not shared by the majority of Gentile Christians” and thus were concerned about issues over Torah observance such as dietary restrictions. cf. Helmut Koester, Introduction to the New Testament, Vol. 2: History and Literature of Early Christianity, 2nd ed. (Berlin: De Gruyter, 2000), 205, cf. 204–12. 69

Patterns of Correction as Paratext

225

context of 1 Cor 12:20 on many parts but one body could also be listed here as it covers the inclusion of Gentiles in the body. The two corrections on the page featuring 1 Cor 10:31–11:6 are difficult to classify. The correction at 10:25 on the previous page (discussed above) is by the same hand (M2) and is in the context of eating meat from the market and meat sacrificed to idols. As the text moves to the next page, the context continues into a discussion of doing everything for the glory of God and not causing offense to Jew or Greek. At 11:1–2 it shifts into discussing male headship and women’s head coverings. The two corrections on this page occur within the latter part of the text (11:3, 5). On one hand, these could be classed as relating to church order and Christian behavior. On the other, the second hand was just correcting a text on an issue of great importance to Jewish believers that spans the two pages, which also face one another in an opening. Thus, it is possible the corrector finished reading the section and then corrected the two quite obvious errors in the following text. In this case, these corrections likely should be included in the category of corrections on old covenant faithfulness. Neither option challenges the categories I have proposed, but uncertainty prevents saying anything conclusive about their presence in the text. One possible outlier among corrections on pages without reading marks occurs at 2 Cor 7:13. A possible explanation is that the context is offering a model in other churches to follow for obedience and righteousness during suffering. 3.4.2 Pages with Reading Marks The next item to look at are those pages with reading marks, noting particularly the seemingly random occurrences following Romans and Hebrews. The first instance covers 1 Cor 13:11–2 Cor 1:8, but notably skips the page containing the final greeting in 1 Cor 16:12–23. The large section from the end of 1 Corinthians covers a range of topics including speaking in tongues, prophecy, church order, the importance of the resurrection of Christ and the dead and its relationship to the current way of life. The single page from 2 Corinthians 1 addresses shared suffering and comfort. The second set of later pages with reading marks is an opening with readings from Phil 1:5–28, which encourages Christian love and righteousness and, for later readers, offers Paul as a model of suffering confidently to advance the Gospel. Overall, none of these sections reveal anything particularly interesting as the contents deal largely with Christian living.71

71 The correction at 1 Cor 13:5 can realistically be added to this set. It does not occur on a page with reading marks, but shares an opening with the start of the run of pages at the end of 1 Corinthians that do. Its context of love does not alter the conclusion of this section.

226

Jacob W. Peterson

3.4.3 Corrections on Pages with Reading Marks The final item to address are those corrections that occur on pages with reading marks. Five corrections are located in the sections with reading marks just discussed and share those themes (1 Cor 14:10, 14:29 bis; 16:7; Phil 1:20). With the exception of folio 21↓ and the final page of Hebrews, all of Romans and Hebrews feature reading marks.72 That Hebrews is so central to the activity in P46 naturally fits my suggestion of a Jewish-Christian context. However, there might even be internal clues within Hebrews in P46 that amplify this suggestion. The first part of Hebrews has numerous corrections that centre around Christ versus famous figures. There are corrections in the context of Christ’s superiority over the prophets (1:1), angels (1:9) and Aaron (5:9). Corrections occur around the text on Jesus as a priest in the order of Melchizedek (7:1 bis, 7:2) who is the ultimate high priest guaranteeing a better covenant (7:25). Three corrections happen in between those of chapters five and seven at 6:1, 6:2, and 6:6. This break represents a single opening in the manuscript and covers the writer of Hebrew’s pause in explaining Christ as a priest like Melchizedek to encourage maturity and warn against apostasy. These few corrections can be seen as more evidence for the pattern of corrections on homiletical topics, or as the later hands happening to fix these pages while working so much on the surrounding pages. The highest concentration of corrections, fifty-five in ten openings, spans Hebrews 8–13. This section begins with the argument that Christ is superior to the old covenant and proceeds with the practical outworking of Christ’s superiority with constant reference to Jewish themes and figures as illustrative of life under the new covenant. The few corrections in the surviving text of Romans further confirm the findings thus far. The first correction is at Rom 9:25 in the context of God extending his mercy to both Jew and Gentile. There is a correction at 12:16 in the context of guides for Christian life and community. Finally, there are two M3 corrections plus one ED correction in an opening at the end of chapter 15. These might be outliers in the overall pattern since the focus of the text is Paul’s travel plans, although the corrections do surround the portion about Gentiles sharing in the spiritual things of the Jews and their contributions to the saints in Jerusalem. 3.4.4 Corrections labelled “Extremely Difficult” Another area worth quickly mentioning are those corrections that could not be assigned a hand easily and were thus grouped into an ED class. Many of these overlap with reading marks, and have been included in previous discussion, but there are several that occur at seemingly random places in the manuscript. Yet, 72 Leaf 18 (Rom 14:9–15:10) also lacks visible reading marks, but since all that is preserved are a few letters at the margins it cannot be said whether the leaf contained reading marks.

Patterns of Correction as Paratext

227

these corrections also continue the themes proposed. For general homiletical issues, there are corrections in passages dealing with avoiding sin (2 Cor 12:19), competing with one another and boasting (2 Cor 10:12), and “the prize of the upward call of God in Christ Jesus” (Phil 3:15). On Jewish-Christian issues, there are corrections in passages about eating idol meat and watching out for the weaker brother (1 Cor 8:9) and becoming as one outside the law to reach those outside the law (“all things to all people;” 1 Cor 9:22). 3.4.5 A Possible Setting? The preceding discussion has revealed that, with rare exception, the pattern of corrections in the later hands in P46 suggests that the manuscript was used in a Jewish-Christian church context based on a marked interest in passages concerning Jewish-Christian identity as well as the expected pastoral topics. Before concluding, I want to very briefly take up whether this is a possible cultural milieu in the relevant time periods and locale. That is, I want to establish whether a third through perhaps fifth century, Jewish-Christian context is a possible cultural framework for P46 to exist within.73 The following attempt at plausibly situating the manuscript is based on the assumption that the Egyptian archaeological provenance of P46, whether from Lycopolis, Aphroditopolis, the Fayum, or elsewhere, is also generally the location of the community that used it.74 The assumption is that P46 was not used and moved around from some unknown place outside Egypt before ultimately being buried there, but was used roughly where it was buried.75 If this assumption is rejected it does not undo the previous work establishing the ideological context in which the manuscript was used. Rather, rejecting the assumption only prevents adding an additional layer of external verification to the conclusions.

P46 is traditionally dated to ca. 200 CE; cf. Pasquale Orsini and Willy Clarysse, “Early New Testament Manuscripts and Their Dates: A Critique of Theological Palaeography,” ETL 88/4 (2012): 458, 462, 470. Even if my proposal that the second hand is not a diorthotes is accepted, the hand must still be assigned to the third century on the basis that the third hand has been assigned to the late third century; cf. C. H. Roberts via Zuntz, Text of the Epistles, 253–54. To my knowledge, no dates have been offered for subsequent hands. My timeline has been stretched to consider up through the fifth century, while recognizing that the bulk of the corrections, those by the second and third hands, occurred in the earlier part of the date range. 74 For a recap of the various views and a suggested provenance, see Brent Nongbri, “The Acquisition of the University of Michigan’s Portion of the Chester Beatty Biblical Papyri and a New Suggested Provenance,” APF 60/1 (2014): esp. 103–12. 75 As an aside, Herman Hoskier repeatedly, albeit at times rather oddly, found numerous clues for the Egyptian-ness of P46 in the main text. See H. C. Hoskier, A Commentary on the Various Readings in the Text of the Epistle to the Hebrews in the Chester-Beatty Papyrus P46 (circa 200 A.D.) (London: Bernard Quaritch, 1938), 9–13, 31, 40, 49, 50, 56, 71. 73

228

Jacob W. Peterson

Since the status of Egyptian Christianity moves well beyond the confines of this study, I only want to quickly mention a couple of correlating pieces of evidence that suggest a Jewish-Christian milieu for P46. First is Bagnall’s explanation of the lack of papyrological record for the early church in Egypt as being due to the church being so numerically small and not spreading widely beyond Alexandria before the early third century.76 This would mean that if there were Jews in Egypt after their persecution by Trajan in the early second century they were not readily converting to Christianity, thereby delaying the necessary debates over how a Jew can be Christian. In a chapter on Jewish Christianity in Egypt, A. F. J. Klijn cites works by several early Christian sources, most importantly Clement and Origen, that demonstrate that “Christianity in Egypt was of a Jewish nature.”77 To mention just one of these, Eusebius records that Clement wrote a work titled “Ecclesiastical Rule against Judaizers.”78 One final piece of evidence is the popular consensus that the Beatty papyri were found buried in jars, which appears to be a Christian adoption of Jewish practices.79 This practice certainly could have been adopted widely so that it was used by any group of Christians (i.e. Gentile Christians), so it cannot be offered as positive evidence, but rather its shows that P46’s rumoured archaeological context does not damage the proposal of a Jewish-Christian context.

4. Conclusion This chapter is divisible into two interrelated, but ultimately distinct parts. The first part introduced patterns of corrections as a new type of paratext in manuscripts. This included an introduction to paratexts and the development of a procedure for selecting a manuscript, mining its data, and investigating its patterns of corrections for glimpses into the life-situations and readers who used the text in antiquity. The second part attempted to demonstrate the method by applying 76 Roger S. Bagnall, Early Christian Books in Egypt (Princeton: Princeton University Press, 2009), 5–6, 19–20, 23. Note also Bagnall’s somewhat offhand alternative explanation for the existence of so many LXX codices as potentially being “signs of the continuing presence of strongly Jewish traits in early Christianity” (p. 24). 77 A. F. J. Klijn, “Jewish Christianity in Egypt,” in The Roots of Egyptian Christianity, ed. B. A. Pearson and J. E. Goehring (Philadelphia: Fortress, 1986), 163–65. More recently, Gilles Dorival, “Un groupe judéo-chrétien méconnu: les Hébreux,” Apocrypha 11 (2001): 7–36. 78 Eusebius, Hist. eccl. 6.13.3. For many more primary sources discussing Jewish-Christian groups, see A. F. J. Klijn and G. J. Reinink, Patristic Evidence for Jewish-Christian Sects, NTSup 36 (Leiden: Brill, 1973). 79 C. H. Roberts, Manuscript, Society and Belief in Early Christian Egypt (London: Oxford University Press, 1979), 6–8. See again Nongbri, “The Acquisition,” on P46’s provenance.

Patterns of Correction as Paratext

229

it to one of the earliest manuscripts of the Pauline Epistles, P46. The patterns of correction in the manuscript revealed an attraction to passages that indicated its use in a Jewish-Christian context. There are many reasons to hesitate about these findings in P46, from methodological problems of assigning hands to correctors, my challenge of the role of the second hand, to what some might perceive as an insufficient data set. Indeed, as noted above, the results of the method can only be stated as confidently as the data allows. In the case of P46, the corrections fell into the two patterns discovered with rare exception, which inspires at least a modicum of confidence. However, more work needs to be done, especially on the role of the second hand, to establish the findings more securely. While this study has focused on the location of corrections, with additional input from reading marks, there are several other manuscript features to which the method could be applied in the future. Other potential paratexts include the location and frequency of candle-wax drippings, pages whose outer margins were darkened by dirt transfer from hands, pages where the parchment has become smooth or glossy from the reader’s repeated turning of those pages by grasping the edge between thumb and fingers, and the locations of marginal notes and apparatus. The poor quality of microfilm images has impeded such studies in the past, with the features in question only being observable by seeing the manuscript in person. The limited accessibility of manuscripts for these types of studies continues to change as various projects successfully digitize more manuscripts with increasingly sophisticated technology. High resolution images open up the possibility of applying this new method to a wider set of manuscripts. Accordingly, it is hoped that, even if the findings from P46 are found unconvincing, the new method for studying paratexts, and especially corrections, has provided a fruitful avenue for future studies on the use and reception of manuscripts. The end goal, by better understanding the cultural milieus of individual manuscripts, is a more complete picture of the development and expressions of Christianity in different times and places and the various means in which these communities interacted with their sacred texts.

Redactional Elements in the Text of Codex B Dirk Jongkind “It is high time that the bubble of codex B should be pricked.”1 Hoskier’s opening phrase of his two-volume work Codex B and Its Allies: A Study and an Indictment leaves little to the imagination. For Hoskier, the Vatican Codex (Vatican Library, Vat. Gr. 1209, henceforth B03), far from being the “neutral” text, is rather a representative of a text type which is the product of “general and constant revision.”2 When reading Hoskier it is always difficult to separate rhetoric from argument and fair representation from straw men, but the gist of his argument is that Westcott-Hort’s “Neutral text” contains little that is at all neutral.3 Though the majority of scholars have rejected the idea that B03 is the product of a thoroughgoing revision, the idea of an “Alexandrian recension” has been around for much of the twentieth century, though mostly as a possibility rather than a firm suggestion.4 At times the idea gained some traction, yet the outlines 1 H. C. Hoskier, Codex B and Its Allies: A Study and an Indictment (London: Bernard Quaritch, 1914), 1.i. 2 Hoskier, Codex B, 2.198. 3 For a comprehensive overview of the history of the idea of “text type” in New Testament textual criticism, see Eldon J. Epp, “Textual Clusters: Their Past and Future in New Testament Textual Criticism,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 520–41. 4 Perhaps Bousset has argued most forcefully for the recensional character of the “Neutral text” with B03 as its purest representative. He dated this recension by Hesychius to the late third century. “Der erste Grundsatz bei der Arbeit ist der, dass Kod. B eine Auktorität allerersten Ranges ist, da wo es gilt festzustellen, was denn eigentlich in der Hesychrecension gestanden. Natürlich wird man von ihm aus sehr oft auch auf die andern Vertreter der Gruppe rekurrieren müssen, aber im ganzen und grossen hat B bis auf Minutien, bis auf die Orthographie den Chrakter der Recension erhaltern.” Wilhelm Bousset, Textkritische Studien zum Neuen Testament, TU 11/4 (Leipzig: Hinrichs’sche, 1894), 74–110 (here 96). Lake suggests so much in 1902: “It is, therefore, certain that the second stage of the history of the text of Alexandria is the use of the ‘Neutral’ type of text (represented among manuscripts by ℵ B) which, so far as we know, was not used previously.” Kirsopp Lake, “The Text of the Gospels in Alexandria,” AJT 6 (1902): 82. Streeter emphasized the existence of local texts, and renamed Westcott-Hort’s “Neutral text” the Alexandrian. As for B03: “The natural conclusion to draw from this is that B represents approximately the oldest text of Alexandria.” B. H. Streeter, The Four Gospels: A Study of Origins, 4 th ed. (London: Macmillan, 1930), 57. For

232

Dirk Jongkind

of what such a recension entailed always remained unexpressed. Others, most notably Gordon Fee, rejected the idea, in part because no convincing historical context for such revision could be found. Still, the idea has never gone away entirely. It is the contention of this study that a detailed study of the peculiarities of B03 should give us pause not to reject the existence of any underlying editorial activity in the preparation of its text. To this purpose we will discuss five features in the New Testament text of B03 which suggest that, when it comes to the presentation of the text, this venerated manuscript shows an “editorial” hand. In other words, there are indications that the text underlying B03 was prepared using editorial principles and that this process has left discernible traces.5 If this thesis can be demonstrated, it would not only have considerable implications for our knowledge of one of the foundational manuscripts of New Testament textual criticism, but also for our understanding of how copying was carried out in the early centuries. Our set of tools to detect editorial activity would be expanded, as well as our appreciation of the level of textual scholarship in the first four centuries. If correct, future study of scribal habits and manuscript tendencies would have to consider not just the copyist’s clumsiness, but also their sophistication. For clarity’s sake, let it be clear what this study is not arguing. I have no reason to believe that B03 contains a bad text or is the result of any sort of systematic rewriting. In many aspects, and especially so in the gospels, B03 preserves an ancient text which often has all the hallmarks of preserving the original wording. But having said this, there are also elements in the text which point to a deliberate shaping of its wording. This leads to the superficially strange paradox that a good

Parsons, the question was still not solved by 1936: “…a doubt as to whether the Neutral text is not in reality the Alexandrian revision of earlier texts…these are some of the matters that are challenging the basic structure of the textual theory.” Ernest W. Parsons, “Review: A Third-Century Papyrus Codex of the Epistles of Paul by Henry A. Sanders,” Classical Journal 32 (1936): 114. More recently, Amphoux posited a number of early editions, dating the Alexandrian to 175 CE. C.-B. Amphoux, “Une édition ‘plurielle’ de Marc,” in The New Testament Text in Early Christianity: proceedings of the Lille colloquium, July 2000, ed. C.B. Amphoux and J. K. Elliott (Lausanne: Zèbre, 2003), 76. Finally, Metzger can be read as supporting the notion of the Alexandrian recension: “It is widely agreed that the Alexandrian text was prepared by skillful editors, trained in the scholarly traditions of Alexandria.” Bruce M. Metzger, The Text of the New Testament: Its Transmission, Corruption and Restoration, 3rd ed. (Oxford: Oxford University Press, 1992), 215. 5 There is a superficial similarity between the title of this essay and J. C. O’Neill, “The Rules Followed by the Editors of the Text Found in the Codex Vaticanus,” NTS 35 (1989): 219–28. O’Neill assumes that ancient editors played a role in the preparation of the text, this essay tries to establish that they acted as such.

Redactional Elements in the Text of Codex B

233

text can be the result of editorial work, though of course modern editors of ancient texts have subscribed to this principle for quite some time.6 In the ancient world, the scribe, or any scribal centre, has a natural role in the shaping of the text. In a context where the multiplication of texts can only be done manually, a serious copyist had to prepare their exemplar before copying. This means that a quality master document had to be selected and then worked over to ensure that any errors in that manuscript were removed before its use as a master-copy. In this way, when copied accurately, the next generation could be at least as good as the master-copy in its unprepared state, and possibly even better. The quality of the prepared master-copy will depend on the skill and the level of scholarship of the scribe, but also on the quality of the manuscripts available to them. In this study we argue that in the recent ancestry of B03 an excellent scholar with good manuscripts available created a “standard text,” a master copy, and that in B03 we can still see signs of this process.7 There is a terminological question. In the English language there are at least four words with overlapping meaning to describe the prepared manuscript, and their prototypical definitions lie in the world of the printed book, which is at best only an imperfect analogy for manuscript transmission. The words “edition,” “revision,” “recension,” and “redaction” all have overlapping definitions in the OED and may perhaps not be helpful in describing the phenomena involved in the preparation of a master-copy.8 Therefore none of these terms is used in a narrowly defined sense in this essay. Both “recension” and “edition” have been used to describe a putative “Alexandrian” issue of the New Testament, whilst

6 Compare Kenyon regarding the β-text (= Westcott-Hort’s Neutral text): “even if it is an edited text, it may be a well-edited text; and in the case of all ancient literature a well-edited text is the best that we can hope for.” Frederic G. Kenyon, The Text of the Greek Bible: A Student’s Handbook, 1 st ed. (London: Duckworth, 1937), 210 (new ed. 1949). 7 That B03 has scholarly characteristics has been stated often (see, for example, some of the references in n.4), but rarely been argued along the lines of this paper. A line of enquiry which may yield additional information about the scholarly background of B03 but is not explored here, might be the nature of the section numbering. The numbering in the Pauline corpus seems to assume that Hebrews was placed between Galatians and Ephesians, rather than after 2 Thessalonians, as it is in the actual manuscript. 8 Fee notes the ambiguity in the term “recension” and tries to distinguish between “revision” and “edition”: “Thus the term ‘recension’ may mean a ‘revision,’ implying both the creation of variants and the selection of similar readings where variation already exists, or it may mean an ‘edition,’ implying not emendation of the text but selection from good and bad manuscripts and/or good and bad readings.” Gordon D. Fee, “P75, P66, and Origen: The Myth of Early Textual Recension in Alexandria,” in Studies in the Theory and Method of New Testament Textual Criticism, ed. E. J. Epp and G. D. Fee (Grand Rapids: Eerdmans, 1993), 250 (originally published in 1974).

234

Dirk Jongkind

Carlo Martini uses the terms “a slight correction” and “a slightly revised text” interchangeably.9 If one approaches this problem from the perspective of the actual copying process, the following levels of editorial activity can be distinguished: 1. Shaping of the text. Decisions need to be made about quire and page size, numbers of columns, paratextual apparatus, paragraphing/line length, interpunction, and the marking of citations. Of these, the last three show already signs of interaction with the text itself (e.g. the decision to present a genealogy in list form). 2. Shaping of the wording. Decisions on abbreviations, ligatures, and nomina sacra, as well as spelling (with the exception of proper nouns). Among the spelling issues are itacisms, apocopation, and assimilation. A scribe can copy all these from their exemplar, but equally impose what is current in their context. 3. Choice of words. Decisions about the presence, absence, and order of words includes the spelling of proper nouns, textual changes such as harmonisation, but also the preparation of the exemplar with the help of other manuscripts. There is an argument to be made that also the subsequent correction of the copy by the scribe falls within this category. Of these three levels, the first is a necessary part of all copying; the physicality of the copying process could not be avoided in the ancient world. On the other hand, conscious activity on the second and third level encroach on preparing an “edition” of the text. An edition needs to contain a sufficient number of features and decisions to be called an edition, but at the very least an edition needs to have given conscious thought to these areas.10 In our investigation we will concentrate on how B03 shows features belonging to the second and third category, which, taken together, demonstrate editorial features. Each of the following five features can be found in other manuscripts, yet B03 occupies a special place amongst these witnesses by virtue of frequency or consistency. The five features are (1) the change from καθως to καθαπερ in the expression καθως γεγραπται, (2) the spelling ιωανης, (3) the order χριστος Ιησους, (4) the omission of the article in ο Ιησους, and (5) the representation of long /i/ with -ει-.

1. “As is written” The expression καθως γεγραπται is a fixed phrase in Romans and occurs 14 times (also occurring twice in each of 1 and 2 Corinthians). Ten times the phrase is

9 Carlo M. Martini, “Is There a Late Alexandrian Text of the Gospels?” NTS 24 (1977– 78): 295–96. 10 As such also a modern or ancient “diplomatic edition” is an edition: a conscious decision to not alter a text has been made.

Redactional Elements in the Text of Codex B

235

without any substantial variation, but B03 is involved in four variant units, and in each of these the variant reads καθαπερ. Apart from variants in the phrase καθως γεγραπται, there is also variation in 4:6 (καθαπερ και δαυειδ λεγει versus καθως in D06 F010 G012) and 12:4 (καθαπερ γαρ εν ενι σωµατι versus ωσπερ in D06 F010 G012), whilst outside Romans the main variant is 1 Cor 10:10, where καθαπερ is favoured (possibly correctly) over καθως by P46 ℵ01 B03 P025 88, but this example also falls outside that covered by the fixed phrase καθως γεγραπται. Further afield there is noteworthy variation in 2 Cor 3:18 (only B03 has καθωσπερ for καθαπερ) and 1 Thess 3:12 (καθωσπερ 69 1739 for καθαπερ Rel.). Romans 1:17 2:24 3:4 3:10 4:17 8:36 9:13 9:33 10:15 11:8 11:26 15:3 15:9 15:21

καθως all all Rel. all all all Rel. all Rel. Rel. Rel. all all all

καθαπερ

ℵ01 B03 Ψ044

B03 B03 ℵ01 B03 [ως: 049]

Table 1. καθως and καθαπερ in Romans

In general there is not much variation between καθως and καθαπερ in καθως γεγραπται, but clearly B03 is involved in more of these than any other witness, including two singular readings in Romans and two further readings with related support. The tendency towards καθαπερ might be an example of an Atticism in B03,11 though the source of the variation is not our interest here. It is important to our argument, though, that B03 shows a tendency in a minor philological point which is unique to its text. This is only a tendency, and some of it is reflected in related manuscripts, but as these variants all concern an identical change, it is warranted to treat them as a single phenomenon. That is, there is a common source behind this tendency that may reflect a moment in its ancestry where this

11 See BDF §453. So also Kilpatrick, who cites the condemnation by Phrynichus in “Atticism and the Text of the Greek New Testament,” in The Principles and Practice of New Testament Textual Criticism: Collected Essays of G. D. Kilpatrick, BETL 46, ed. J. K. Elliott (Leuven: Peeters, 1990), 26–7.

236

Dirk Jongkind

set was introduced in the text. A bold conclusion if this were the only instance of such set, but the following examples do provide further support.

2. Spelling of John One of the spelling curiosities found in certain early manuscripts is the spelling of the name “John” with a single rather than a double -ν-, ιωανης instead of ιωαννης. This is found often in B03, is the characteristic spelling of the name by scribe D of Codex Sinaiticus (e.g. in the correction of John 21:15), and figures also prominently in D05.12 P75 has the spelling with single -ν- throughout Luke until John 1:28, where the first hand corrects to -νν-, after which the spelling with -νν- continues with the exception of 3:27 (after a series of verses with -νν- in the immediate environment) and 10:40 (again corrected, followed by -νν- twice in 10:41). The situation in P75 may be explained as having been copied from an exemplar that has ιωανης throughout Luke and John, though the scribe (or the immediate exemplar) attempts to change to a more normal spelling from 1:28 onwards, with occasional interference from the spelling as read in the manuscript in front of the scribe. B03 has ιωανης throughout, except in 1 Esd 9:29,13 Luke 1:13, 60, 63 (which are the first three instances in Luke), and eight times in Acts (out of twenty-four instances).10 The origin of the spelling should probably be seen in confusion surrounding the doubling of Semitic -n- into Greek, and BDF rightly states that there is no reason why the -ν- should be doubled in the name ιωαννης.14 The consistency with which B03 presents the spelling ιωανης in the four gospels seems to point to a deliberate policy. The few exceptions of -νν- in B03 Luke and Acts seem to be a case of personal and contemporary scribal patterns resurfacing and suppressing the practice of the exemplar.15 That P75 and D05 often follow the same convention as Vaticanus suggest that the origin of this phenomenon is wider than just the master-copy of B03. Yet the level of implementation in B03 suggest an editorial policy. Interestingly, the nature of the change to a single -ν- is not a randomly chosen consistency, but a philologically more correct option that possibly reflects knowledge of the underlying Hebrew.

For a discussion of how the spelling of John has played a role in the discussion of D05 see D. C. Parker, Codex Bezae: An Early Christian Manuscript and its Text (Cambridge: Cambridge University Press, 1992), 109–10. 13 ιωανης in 1 Esd 8:38 [A02 ιωαννης] and 2 Par 28:12 [A02 ιωαναν]. 10 Acts 3:4 (possibly, the retracer may be responsible), 4:6, 13, 19; 12:25; 13:5, 25; 15:37. 14 BDF §40. 15 It is equally possible that the master copy itself was not consistent in applying the policy. 12

Redactional Elements in the Text of Codex B

237

3. Christ Jesus in Romans A third feature of B03, which like variants on καθως γεγραπται in the first example seems concentrated in the book of Romans, is the reversal of the order “Jesus Christ” to “Christ Jesus.” Christ Jesus is an idiom particular to the Pauline corpus and, as such, the change is not a change to general usage, but rather to Pauline custom. This would mean that an authorial peculiarity is strengthened in B03.16 There is no variation in the order χριστος ιησους in eight locations, and conversely, ιησους χριστος is without variation in seven instances in Romans.17 There are two places where there is only an add/omit variation of ιησους in χριστος ιησους.18 In twelve locations there is a word order variant: Romans 1:1 1:7 2:16 3:22

Jesus Christ Rel. Rel. Rel. Rel.

5:15 5:17 5:21 13:14

Rel. Rel. Rel. Rel. [τον κυριον ιησουν χριστον]

15:5 15:16

ℵ01 A02 C04* F010 1315 1505 2400 2495 D06 Ψ044 88 Rel.

16:25 16:27

Rel. Rel.

Christ Jesus P10 B03 81 1838 P10 ℵ01*vid B03 Εν χω ιυ A02 omit ιυ B03 1739 B03 B03 τον χριστον ιησουν B03 τον κυριον ιησουν 489 927 1739 1881 ιησουν χριστον τον κυριον ηµων P46 P46vid B03 C04corr D06 Ψ044 1739 Rel. ℵ01 A02 B03 C04 F010 G012 218 1505 1739 1962 2400 2495 B03 1739 B03

Table 2. “Jesus Christ” and “Christ Jesus” in Romans

16 Perhaps the tendency is not limited to Romans. See e.g. the singular reading of B03 in Matt 1:18, του δε χριστου ιησου η γενεσις ουτως ην. 17 χριστος ιησους 3:24; 6:3, 11, 23; 8:1, 2, 39; 16:3. ιησους χριστος 3:24; 6:3, 11, 23; 8:1, 2, 39; 16:3. 18 8:34 (NA28 includes ιησους in brackets; THGNT excludes ιησους but includes it as an alternative diamond reading), 15:17 (only P46 omits ιησου). See also 3:22 where only B03 omits ιησου, and 13:14 where an important group of minuscules leaves out χριστον in κυριον ιησουν χριστον. For a discussion of Gal 5:24, a case superficially analogous to Rom 8:34, but with a different attestation pattern, see Stephen C. Carlson, The Text of Galatians and Its History, WUNT 2/385 (Tübingen: Mohr Siebeck, 2015), 134–36.

238

Dirk Jongkind

The figure shows that there are five places where B03 has Christ Jesus where almost everyone else has Jesus Christ: 5:17, 21; 13:14 (“Christ Jesus” for “Lord Jesus Christ’”); 16:25 (together with GA 1739); 16:27. Additionally, there is 3:22 where B03 has another unique reading that avoids “Jesus Christ.” Further singulars are found in P10 at 1:7, A02 at 3:22, and 1739 in 5:15. The change from “Christ Jesus” to “Jesus Christ” is not limited to B03 alone; other manuscripts show the same change on occasion, though only those manuscripts that have some affinity with B03. The changes are not systematic or exhaustive by any means, but the frequency with which it shows up in B03 is remarkable. A tendency already found in the author is strengthened in B03. In combination with the previous two examples it seems that a pattern is emerging.

4. The Article with Jesus A fourth feature is, like the previous example, also restricted to a single book. The Johannine use of the article with the name ιησους has received attention in previous scholarship, often noting the confusion in the textual tradition.19 We pick up the discussion at Fee’s 1970 study. At the time of the publication of Fee’s article, UBS2 had been published, though UBS3 (= NA26) was still five years off. Fee seems worried about the changes in the use of the article in John as they appear to be “based chiefly on manuscript evidence, rather than on internal, especially stylistic, consideration.”20 Fee notes that B03 was considered to be divergent, but has now been exonerated by P66 and P75,21 citing Nevius in support. It is indeed remarkable that at some places where B03 was singular in its omission of the article before ιησους, it has now support of one or two of these papyri, though the papyri, especially P66, introduce a number of singular omissions as

19 T. F. Middleton, The Doctrine of the Greek Article: Applied to the Criticism and Illustration of the New Testament (Cambridge: J. & J. J. Deighton, 1833), 240–66 (though Middleton does not seem to comment on John’s peculiarities). Edwin A. Abbott, Johannine Grammar, Diatessarica 6 (London: A&C Black, 1906), 1967–970. Richard C. Nevius, “The Use of the Definite Article with ‘Jesus’ in the Fourth Gospel,” NTS 12 (1965): 81–5. Gordon D. Fee, “The Use of the Definite Article with Personal Names in the Gospel of John,” NTS 17 (1970): 168–83. For a Discourse Grammar explanation of the use of the article with names in John, see Stephen H. Levinsohn, Discourse Features of New Testament Greek: A Coursebook on the Information Structure of New Testament Greek, 2nd ed. (Dallas: SIL International, 2000), 155 and 159–60. For a different linguistic explanation see Ronald D. Peters, The Greek Article: A Functional Grammar of ὁ-items in the Greek New Testament with Special Emphasis on the Greek Article, Linguistic Biblical Studies 9 (Leiden: Brill, 2014), 247–51. 20 Fee, “The Use of the Definite Article with Personal Names,” 169. 21 Fee, “The Use of the Definite Article with Personal Names,” 169.

Redactional Elements in the Text of Codex B

239

well. Fee explains all this variation with the help of a number of methodological issues which nowadays are no longer undisputed assumptions. First, there is a problem with how Fee uses the notion of the “Neutral text type” as some sort of super-document. This enables him to label a singular omission of the article in P66 and another one in B03 both as omissions found in the “Neutral tradition,” and Fee urges to take these serious. Furthermore, Fee explains the presence/absence of the article mainly by means of structural considerations, that is, whether ιησους appears in a standard conversational turn or whether ιησους follows the verb or not. Such approach reduces grammar to mere structure. A more promising approach is demonstrated by Spielmann who gives an attractive discourse explanation of John’s use of the article.22 But in the end, though Fee manages to describe Johannine habits in general, he fails to give an answer as to the individual cases of variation where any of the manuscripts, and B03 in particular, give solitary testimony to the absence of the article. Should editors take this as “the witness of the early Neutral text type”? The following data are those as given by Fee.23 The name occurs 46 times outside the nominative, five times without the article, but in all of these cases “Jesus” has an attributive element.24 Fee notes only a single place among the 41 instances where there is “serious variation,” namely in 11:21, where the article is omitted by ℵ01 B03 and 213.25 Fee omits to mention that B03 alone omits the article at 12:3 and 19:38 (first occurrence), while there are two other places where some scattered minuscules do the same. Thus, of all early manuscripts, only B03 shows singular instances of the anarthrous form. This situation is not unlike that for the nominative. The cases where a manuscript is singular are a good indication as to the tendencies of the manuscript. In my count there are nineteen true singulars in this construction, and only the early manuscripts are considered here: 8x B03; 5x ℵ01; 3x D05; 2x P66 (once corrected); 1x A02. In addition there are readings with “random” support or with an omission within an alternative reading: 4x ℵ01 (6:24, 26; 8:31); 2x B03 (1:47; 10:7); 1x P75 (7:33); 1x D05 (20:18). From this it is clear that there is a tendency among ℵ01 and B03 to omit the article. It should be noted though that the singular readings of ℵ01 almost all occur in the earlier chapters of John.26 22 Kent Spielmann, “Participant Reference and Definite Article in John,” Journal of Translation and Textlinguistics 7 (1995): 45–85. 23 Fee, “The Use of the Definite Article with Personal Names,” 173–82. 24 1:17 δια Ιησου χριστου; 1:45 Ιησουν υιον του Ιωσηφ τον απο Ναζαρετ; 17:3 Ιησουν χριστον; 18:5 and 18:7 Ιησουν τον Ναζωραιον. 25 Fee, “The Use of the Definite Article with Personal Names,” 172. It is possible that C04 belongs in this list, as the text τον ιν κε ει ης is rewritten over the original hand. The image is unclear but there may even be an earlier interlinear addition of τον, now erased. 26 Gordon Fee also produced a study demonstrating the different character of ℵ01 in the first chapters of John in “Codex Sinaiticus in the Gospel of John: A Contribution to Methodology in Establishing Textual Relationships,” in Studies in the Theory and Method

240

Dirk Jongkind

To these numbers should probably be added the places where the omission is supported by two and three witnesses only. Two witnesses: B03 9x; P66 6x; ℵ01 3x; P75 2x; C04 2x; D05 1x; A02 1x.27 From this data it appears that anarthrous ιησους is found mostly in B03, but also in P66, ℵ01, and to a lesser degree in P75. P66 -ὁ +ὁ απεκριθη 13 1 Ιησους και ειπεν αυτ(ῳ)

P75 -ὁ +ὁ 12 0

03 -ὁ 12

+ὁ 1

01 -ὁ 9

+ὁ 4

02 -ὁ 8

+ὁ 3

05 -ὁ 6

+ὁ 5

032 -ὁ +ὁ 12 1

TR -ὁ +ὁ 4 9

απεκριθη 10 Ιησους απεκριθη 6 αυτ(ω) [ο] Ιησους

1

7

0

11

2

8

4

8

4

2

5

9

3

3

8

3

1

4

10

3

3

9

1

9

0

11

3

11

0

16

λεγει 1 (ειπεν) [ο] Ιησους

3

1

2

2

3

1

2

1

2

1

3

0

3

0

3

λεγει (ειπεν) αυτ(ω) [ο] Ιησους Ιησους Verb Verb…ο Ιησους Verb [ο] Ιησους

14

27

4

29

18

32

10

41

6

39

2

46

3

47

0

55

18

1

15

0

22

0

18

2

20

1

20

0

19

1

15

6

3

19

1

14

7

19

4

22

0

25

3

16

3

26

1

27

10

16

7

13

21

19

15

17

6

27

13

20

6

31

0

39

Total: 75 % 51 (rounded):

71 49

48 44

62 56

103 79 57 43

68 40

101 50 60 31

110 47 69 31

106 55 69 31

123 23 69 12

163 88

Table 3. Arthrous and Anarthrous forms of “Jesus” in John

With the wide-spread confusion surrounding this Johannine phenomenon, the method of looking at singular readings is not as helpful as it is in the case of καθως γεγραπται above. Simply giving the totals is sufficient to make the point. The figure above is taken from Fee with the only difference that an extra row is added to give the proportion of anarthrous and articulate occurrences of ιησους.28

of New Testament Textual Criticism, ed. E. J. Epp and G. D. Fee (Grand Rapids: Eerdmans, 1993), 221–43. 27 If we assume that at 20:29 the unrelated support of GA 729 is random, the numbers for B03 and P66 are one higher. 28 Fee, “The Use of the Definite Article with Personal Names,” 174.

Redactional Elements in the Text of Codex B

241

If we rank the manuscripts on the basis of how often the article is absent and assume even distribution even when manuscripts are incomplete, it becomes clear that B03 does so more than any other witness:

MS B03 P66 P75 ℵ01 A02 W032 D05 TR

Anarthrous Nominative Ιησους in John 56.6% 51.4% 43.6% 40.2% 31.3% 30.9% 30.7% 12.4%

Table 4. Percentage of Anarthrous Forms of “Jesus”

Fee notes rightly that the phenomenon of an anarthrous Ιησους is Johannine, but fails to point out that, relative to the other early manuscripts, B03 has by far the most anarthrous forms. Fee used these data to reject the idea that B03 showed any signs of being edited, since “the idiom is deeply entrenched in the Egyptian text-type and partly so in other MSS as well.”29 Leaving the issue of an Egyptian text type aside, Fee argues that B03 simply continues a Johannine idiom with a higher frequency than other early witnesses. This begs the question, however, of why this occurs in B03. Is it because in each of these cases B03 preserves the original or fell afoul of influence from the immediate context, or, alternatively, because it has intensified an already existing Johannine tendency? The simplest explanation of all the individualities of the various manuscripts is that each of them was affected by John’s irregular usage, yet the normally tightly controlled text of B03 was influenced more so than the others in this direction. As with the Pauline “Christ Jesus,” a reasoned eclectic approach to the situation in John is that the simplest explanation for the situation in B03 is that this manuscript has intensified an existing authorial phenomenon without imposing consistency.

5. Long /i/ Finally, a feature that functions on the second and third level of editorial shaping of the text is the way in which B03 deals with -ει- / -ι- interchange. In many manuscripts these itacisms appear random and show that their scribes were, what

29

Fee, “Myth of Early Textual Recension,” 269–70.

242

Dirk Jongkind

Caragounis called, anorthographoi, incapable of spelling correctly.30 However, the spellings in B03 demonstrate a regularity that has not been sufficiently recognised in previous scholarship on this manuscript.31 It is thanks to two of my co-editors of the Tyndale House Edition of the Greek New Testament, Patrick James and Peter J. Williams, that the scale of this feature was identified and I rely here on their findings and expertise.32 As an example, take the text of B03 in Matt 7:1–2: µη κρεινετε ινα µη κριθητε εν ω γαρ κριµατι κρεινετε κριθησεσθε. The difference between the present and future stem of the verb κρινω is that between a stem with a long /i/ -κρῑν- and a short /i/ -κρῐν-, which are then subsequently accented as κρίνω and κρινῶ. In the words from Matthew 7, the distinction between the two stems is signalled consistently by marking the long present stem with -ει-, while the future stem (and the short /i/ in the nouns) have simple -ι-. Though such marking of long and short /i/ is known from early inscriptions, it is remarkable to find this in a fourth century manuscript, a time in which the distinction between short and long vowels is supposed to be completely lost.33 A longer example where the same distinction holds true is the section 1 Cor 5:12–6:7. In over 309 cases of verbs and nouns relating to κρίνω, Patrick James did not find a single instance where -ει- was used for -ῐ- indisputably. Inversely, in cases of doubt, there is always a case to be made for the future. In a recent study, Williams extended the investigation into the representation of Hebrew vowel length in the genealogies of Matthew and Luke.34 Here prevocalic /y/ is always represented by -ι-, while stressed final /i/ is always -ει-. Furthermore, whenever -ει- occurs in names such as αµειναδαβ, there is secondary stress on this syllable in the underlying Hebrew. Also other Hebrew words in 30 Chrys C. Caragounis, The Development of Greek and the New Testament: Morphology, Syntax, Phonology, and Textual Transmission, WUNT 167 (Tübingen: Mohr Siebeck, 2004), 502. 31 BDF §23 notes though, “our earliest MSS treated the scholastic regulation much more freely than the later, i.e. they frequently wrote…(like Vaticanus and the great Hermas papyrus [Bonner 20]) ει for ῑ to distinguish it from ι.” 32 Patrick James presented his findings on 16 November 2016 in the Indo-European Seminar at the Classics Faculty, University of Cambridge under the title “Orthography, ? Spellings in Papyri, Uncials, and Tyndale House’s The New Testament in its Original Greek.” Peter J. Williams, “Semitic Long /i/ Vowels in the Greek of Codex Vaticanus of the New Testament,” in Studies in Semitic Linguistics and Manuscripts: A Liber Discipulorum in Honour of Professor Geoffrey Khan, Studia Semitica Upsaliensia 30, ed. N. Vidro et al. (Uppsala: Uppsala Universitet, 2018), 15–26. 33 “There is a very frequent interchange of ει and ι (whether long or short etymologically) in all phonetic environments throughout the Roman and Byzantine periods. This indicates the identification of the classical Greek /ei/ diphthong with the simple vowel /i/.” F. Gignac, A Grammar of the Greek Papyri of the Roman and Byzantine Periods: Volume I Phonology (Milano: Istituto Editoriale Cisalpino-La Goliardica, 1976), 189. 34 Williams, “Semitic Long /i/ Vowels,” 17–8.

Redactional Elements in the Text of Codex B

243

Greek seem to follow the same pattern. Of particular interest is the cry of dereliction in Matt 27:46 in B03, ελωει ελωει λεµα σαβακτανει, where both the form ελωει and σαβακτανει are unique to B03. ελωει is another example of a stressed final yod. The latter might be a better representation of ‫ שבקתני‬than the customary σαβαχθανει, and would be an additional argument in favour of intentional adaptation of the Greek to the Hebrew source of the expression. One important implication of this distinction is that we do not have to wait for the time of accented manuscripts to get an indication how the tradition, or more specifically a scribe, read the verb κρίνω. With B03 we find an early Greek source where the present and future are differentiated. Moreover, unlike the book-specific phenomena of the article with Jesus, the order of the names “Christ Jesus,” and the tendency affecting καθως γεγραπται, this deliberate representation of long /i/ occurs throughout the New Testament, and possibly also in the Greek Old Testament. But the consistency and scale of this phenomenon is not paralleled in other early New Testament manuscripts.

Conclusion We have looked at three phenomena in B03 that are concentrated in particular books: καθως γεγραπται and the order “Christ Jesus” in Romans, and the article with ιησους in John’s Gospel. The remaining two span the New Testament: the spelling of the name ιωαν(ν)ης and the representation of long /i/ by -ει-. Of these tendencies, numerically καθως is the weakest phenomenon and, though stronger in B03 than elsewhere, still relatively indistinct. The single -ν- spelling of ιωαννης is found elsewhere too and sometimes even quite strongly so (e.g. P75 in Luke). But in the other three arguments, B03 stands out noticeably from the rest of the tradition. Any description of B03 will have to take these characteristics into account, even though in general this manuscript has very few idiosyncrasies. But how to explain all this? Taking these five superficially unimportant signs together, it seems to me that the simplest explanation is that B03 goes back to a carefully prepared standard, or set of carefully prepared standards for different books, in which attention was given to authors’ style and philological purity. However, the impact of these considerations was constrained by the available manuscripts, though a certain strengthening bled through. Thus, the master-copy prepared sometime in the ancestry of B03 is not only well-prepared, but seems to have used good and ancient texts. Features such as the spelling of long /i/ and the name of John are not limited to single books and therefore may reflect a phase of editing that is later than the phenomena found in single books; they were implemented once the books were already together. However, various layers of editing are by no means a necessary hypothesis at this stage and, therefore, I assume a single authoritative master-copy for the moment, a master copy which has a

244

Dirk Jongkind

history of its own. Since this master-copy is dependent on previous manuscripts and only rarely seeks innovation, it comes as no surprise that some of these phenomena are to a degree found in other documents, sometimes because similar concerns are at stake, sometimes because they do reflect the Ausgangstext. However the particular frequency in B03 suggests an editorial hand. What is perhaps most surprising about the suggestion of the creation of an authoritative, high-quality text, is that it happened at all in the earliest centuries. Though we have the occasional indication that exemplars were created that functioned as a standard to collate against and which were used for correction, it has to my knowledge very rarely been argued that we have in B03 a careful copy of such prepared manuscript. At one point, Fee demanded that recensional elements can only be labelled as such when found throughout the New Testament.35 In my view that is a harsh, and perhaps even anachronistic, demand to make. After all, it is very well possible that B03 reflects a master-copy that had been prepared in the third, or possibly even in the early third century, a time from which we only have witnesses containing part of the New Testament corpus. If so, the complete New Testament corpus was created by bringing together corrected copies of parts of the New Testament. It is this material context in which one would expect editorial features that are restricted to a single book. Moreover, as it is also modern practice to take into account the style of the author, so it may have been then. So what about the “Alexandrian recension?” Is it time to revive this hypothesis? First of all, I am agnostic about the locality where the exemplar of B03 was prepared and, at this point, am unwilling to suggest any. Likewise, the label “recension” is problematic; especially when this is conceived of as a “published” edition. None of the features discussed above suggest a rewriting or retelling of the text. Whatever innovations were made to the text take place on the fringes. But that we have in B03 a copy of a text that was carefully prepared and done so with linguistic interest and competence seems clear. The master-copy was prepared using older manuscripts, which explains the shared features with other early manuscripts. But in its preparation it also seems that careful judgement calls were made, only leading sporadically to overemphasizing the unique language of John and Paul. Characterising B03 in this way does not bring us back to the original text (or whatever term one prefers for the oldest reconstructed wording), but it gives us access to work done by a colleague or colleagues with a similar interest in the details of the text as textual critics have nowadays. The fact that the preparation of the exemplar has left few, though clear traces is in itself an argument for the quality and restraint with which the work was carried out. In B03 we do not have a liturgised text, or a text that is set to remove harsh Greek formulations, or a text that aims to adapt the wording to that of a changing language. Rather it is a text sensitive to author-specific expressions 35

Fee, “Myth of Early Textual Recension,” 269.

Redactional Elements in the Text of Codex B

245

without imposing consistency, and quite likely it is a text prepared in a context with both Greek and Hebrew philological expertise.36 There remains a multitude of questions to be answered, especially as to the Hebrew philological interest noted with ιωαν(ν)ης. Is it present on a wider scale in the Greek Old Testament text and are there other links to Hebrew philology? Could the numbering of the Ten Commandments in Exodus be relevant here? The possibility that we have in B03 a document that shows evidence – even if only in a minor way – of Hebrew scholarship is an exciting prospect but needs much firmer support than raised in this essay. Besides the phenomena mentioned in this paper, which deserve a fuller treatment, what other editorial patterns can be detected? 37 What are the constraints that the creator of a master copy adhered to? Even bigger questions need to be raised too. If there was textual scholarship of a high level in the third or fourth century, was there perhaps a line of careful transmission within the centres where such scholarship was practised? Should we conceive of a circle around individual scholars who carried out this work? Is the impression of the simple, unsophisticated copying of texts that we might have learned from the Egyptian papyri perhaps incomplete? The notion that some of the contamination was carried out deliberately in creating an exemplar could help us to understand the distinctive elements of some of the tight textual families. It is inherently unlikely that the evolution of the New Testament text always happened gradually, in a plodding corruption from generation to generation. Part of it occurred in leaps and bounds. Much remains unknown even about our best and most famous manuscripts. And perhaps at times a scholarly bubble needs to be pricked. In the case of B03 however, it is clear that not only is the text copied in a controlled and precise manner, but also that it is copied directly or indirectly from a well-prepared text, and as such the weight and prestige that this witness has received over the past two centuries seems warranted.

This characterization of B03 is not unlike what Kenyon wrote about the Neutral text (a category no longer uncritically accepted): “Of the Neutral text…it can be affirmed with more confidence that, if it is the result of editorial handling, the editor was one who was seeking an original text. It is not harmonistic, it does not cultivate smoothness of phrase, it does not seek additions. It may be described as an austere text; Frederic G. Kenyon, Recent Developments in the Textual Criticism of the Greek Bible (London: British Academy, 1933), 85. 37 Room for expansion of editorial features may be found in phenomena such as the scope and items treated as nomina sacra and the treatment of numerals written as a Greek letter with overline. Cole’s data suggest that B03 is indeed remarkable vis-à-vis some of its contemporaries, and especially in comparison to P75. Zachary J. Cole, Numerals in Early Greek New Testament Manuscripts: Text-Critical, Scribal, and Theological Studies, NTTSD 53 (Leiden: Brill, 2017), 98–100. 36

The Garland of Howth (Vetus Latina 28): A Neglected Old Latin Witness in Matthew H. A. G. Houghton* Research on the textual tradition of the Latin New Testament has been slower to integrate digital approaches than comparable investigations of the Greek evidence. Bonifatius Fischer’s computer collation of substantial test passages in over four hundred and fifty Latin gospel manuscripts from the first millennium was a notable but isolated early achievement.1 It was only two decades later that this was used to identify new witnesses to the Old Latin textual tradition and the tables of overall statistical agreement were published.2 The Verbum Project, running at the University of Birmingham between 2002 and 2005, made full-text electronic transcriptions of manuscript witnesses to the Old Latin version of the Gospel according to John.3 Even so, these were not integrated into a single, automated workflow for producing the Vetus Latina edition of John. Other recent projects to produce volumes in this series have made only limited use of software tools.4

* The present study uses material prepared for the Irish Latin Gospel Books Transcription Project (www.insulargospels.net), and is offered in gratitude to Professor Martin McNamara. The author would also like to thank Dr Garrick Allen for the invitation to deliver this paper at the conference Herman Hoskier and the Future of Textual Scholarship on the Bible held in Dublin in August 2017. 1 Bonifatius Fischer, Die lateinischen Evangelien bis zum 10. Jahrhundert, AGLB 13, 15, 17, 18 (Freiburg: Herder, 1988–1991). 2 H. A. G. Houghton, “A Newly Identified Old Latin Gospel Manuscript: Würzburg Universitätsbibliothek M.p.th.f.67,” JTS 60 (2009): 1–21; Houghton, “The St Petersburg Insular Gospels: Another Old Latin Witness,” JTS 61 (2010): 110–27; J.-C. Haelewyck, “Un nouveau témoin vieux latin de Marc. Le ms. Durham, Cathedral Library A.II.10 + C.III.13 + C.III.20,” RBen 123 (2013): 5–12; Bonifatius Fischer†, “Die lateinischen Evangelien bis zum 10. Jahrhundert. Zwei Untersuchungen zum Text,” ZNW 101 (2010): 119–44. 3 For the separate electronic edition produced by this project, see http://www.iohannes.com/vetuslatina [accessed 23 March 2018]. 4 Transcriptions produced at the University of Mainz towards a Vetus Latina edition of the Acts of the Apostles were made directly into an Excel spreadsheet, released online in PDF form (http://nttf.klassphil.uni-mainz.de/179.php). There does not appear to be any electronic data underlying the Vetus Latina edition of Mark (J.-C. Haelewyck, ed., Vetus Latina.

248

H. A. G. Houghton

The present chapter combines the use of unpublished data from Fischer’s collations with a full-text electronic transcription of the Gospel according to Matthew made from new, high-resolution digital images and a computer-generated apparatus of variants from the standard text of the Stuttgart Vulgate in order to investigate the textual affiliation of a ninth-century gospel book copied in Ireland. In so doing, it represents a born-digital approach to the examination of Latin manuscripts of the New Testament, offering a paradigm for further work in this area. At the same time, given that the most recent study of this witness was produced by H. C. Hoskier, it allows for reflection on developments in the study of witnesses to the New Testament over the last century.

History of Research on the Manuscript In 1919, Hoskier published a New and Complete Edition of the Irish Latin Gospel Codex Usser[ianus] 2 or r2, otherwise known as ‘The Garland of Howth’ in Trinity College Library, Dublin.5 Hoskier’s edition of the surviving portions of this manuscript followed two works produced by Irish scholars in the preceding decades. The first was T. K. Abbott’s collation of this manuscript included as an appendix to his edition of Codex Usserianus Primus, the Old Latin gospel book known as r1 or VL 14.6 On the basis of his collation, Abbott identified the text of the Garland of Howth as Old Latin in Matthew, Vulgate in Mark, John and much of Luke, and a mixture of these traditions in Luke 2 and the latter part of the same gospel.7 A few years later, H. J. Lawlor refined Abbott’s conclusion by observing that the Old Latin section in Matthew only began in chapter 16: although Lawlor’s principal concern was to illustrate the phenomenon known as “block mixture,” when parts of the same manuscript are copied from different exemplars,

Die Reste der altlateinischen Bibel. 17. Evangelium Secundum Marcum. [Freiburg-im-Breisgau: Herder, 2013–]). 5 Although this is the description given on an initial page at the beginning of the volume, the title page itself has a slightly different formulation: H. C. Hoskier, The Text of Codex Usserianus 2. R2. (“Garland of Howth”). With Critical Notes to Supplement and Correct the Collation of the Late T.K. Abbott (London: Bernard Quaritch, 1919). It may be noted that, despite the characterisation of the volume as appearing in a series on “Old Latin Texts,” it is a one-off with no connection to the Oxford series of the same name. 6 T. K. Abbott, Evangeliorum Versio Antehieronymiana ex Codex Usseriano (Dublinensi), Adiecta Collatione Codex Usseriani Alterius (Dublin: Hodges, Figgis & Co., 1884). The appendix is on pp. 819–63: Abbott takes Codex Amiatinus as the base for his collation and uses symbols to mark agreements with Codex Usserianus primus and the Book of Kells. 7 Abbott, Evangeliorum Versio Antehieronymiana, xiv–xv. Abbott’s edition was used to cite this manuscript in the Oxford Vulgate. J. Wordsworth and H. J. White, Novum Testamentum Domini Nostri Iesu Christi Latine. 1. Evangelia. (Oxford: Clarendon, 1889–1898).

The Garland of Howth

249

he published a full transcription of this portion as an appendix in his Chapters on the Book of Mulling.8 Hoskier’s edition is what I have described elsewhere as a “textual facsimile.”9 The transcription matches the layout of the manuscript, but is mostly printed with normal type: special characters are used to render the insular g and symbols for autem and enim, certain capital letters including a and n, and the et digraph, in order to reproduce the appearance of the text as closely as possible. Although this is typographically impressive and assists with reading some damaged parts of the manuscript, the preservation of all the abbreviations makes this a difficult work for the non-specialist to use. At the end of the volume, Hoskier prints a supplement to Abbott’s collation of the manuscript in which some of the spelling conventions and abbreviations are explained. In fact, in his introduction, Hoskier describes the quality of Abbott’s work as “deplorable,” noting “one thousand errors of omission and commission” and a failure to report “over two thousand varieties of spelling.”10 Hoskier’s own transcription is generally very accurate, although it too can still be improved in a few places.11 In his brief introduction, however, he contents himself with listing just a dozen or so readings from each gospel which “seem to be of more than passing interest,” especially those which support his theory of the origin of the Latin version in a bilingual manuscript.12 There is no discussion of the position of the Garland of Howth within Latin textual tradition or an examination of its block mixture. The Garland of Howth was included in the inaugural register of Old Latin manuscripts issued by the Vetus Latina Institute in 1949, and assigned the number 28.13 Nevertheless, unlike similar manuscripts on either side, it was not excerpted onto the index cards which subsequently formed the basis of the Vetus Latina Database.14 It was also not included in any volumes of the Itala edition 8 H. J. Lawlor, Chapters on the Book of Mulling (Edinburgh: David Douglas, 1897), Appendix A, “The Old Latin Portions of ‘The Garland of Howth,’” is on pp. 186–201. 9 H. A. G. Houghton, “The Electronic Scriptorium: Markup for New Testament Manuscripts,” in Digital Humanities in Biblical, Early Jewish and Early Christian Studies, ed. C. Clivaz, A. Gregory, and D. Hamidović (Leiden: Brill, 2014), 31–60 (esp. 32, 43, 52). 10 Hoskier, The Text of Codex Usserianus 2, iii; the emphasis is original. 11 For example, he fails to spot the opening words of Matt 1:18 on the first surviving page of the codex. 12 Hoskier, The Text of Codex Usserianus 2, vi–x (quotation from vi); the verses he identifies as “polyglot-places” are Matt 20:30 and 26:3. 13 Bonifatius Fischer, Verzeichnis der Sigel für Handschriften und Kirchenschriftsteller, Vetus Latina 1/1 (Herder: Freiburg, 1949). It is also included in Patrick McGurk, Latin Gospel Books From A.D. 400 to A.D. 800 (Paris: Érasme, 1961) as number 85; in his brief description, McGurk observes the text of Matt 1:18 on fol. 1r missed by Hoskier. 14 VL 27 was excerpted from Rettig’s 1836 facsimile, while VL 30 was taken from Heer’s edition of 1910. For the Vetus Latina Database, see H. A. G. Houghton, The Latin New Testament: A Guide to its Early History, Texts, and Manuscripts (Oxford: Oxford University Press, 2016), 116.

250

H. A. G. Houghton

of Old Latin gospel manuscripts.15 The first examination of its biblical text since Hoskier was as part of Fischer’s computer collations published in the late 1980s. The table of overall agreements, published posthumously, indicated that the overall agreement of the Garland of Howth with the editorial text of the Stuttgart Vulgate was 78.9%.16 This places it among the twenty witnesses which differ most from this standard form. However, a breakdown of its affiliation in each passage and details of its closest relatives can be presented from the raw data files kindly provided to me by the publishers. The range of figures is given in Figure 1, which gives the number and extent of the passages, agreements with the Stuttgart Vulgate as both an actual number and a percentage, and the rank of dissimilarity from the Vulgate among all manuscripts in the sample. The most striking figures are those for the Gospel according Matthew: in passage 14, a selection from Matthew 26 and 27, the Garland of Howth exhibits the greatest difference from the Vulgate of all Latin gospel manuscripts copied in the first millennium, with an agreement of just 56.5%.17 Admittedly, the competition is reduced by Codex Bobiensis (VL 1) and Codex Palatinus (VL 2) not being extant here, but this represents a greater dissimilarity than well-known Old Latin witnesses such as Codex Vercellensis (VL 3), Codex Veronensis (VL 4) and Codex Bezae (VL 5). In passage 13, where there is a 64.1% agreement with the Vulgate, it comes in fourth place overall, behind Codex Palatinus, Codex Bezae and Codex Claromontanus (VL 12). In the passages from Mark and Luke the overall agreement is higher, consistent with texts displaying a mixture of Old Latin and Vulgate readings. In John the percentage drops again, although the manuscript is only extant for ten verses of passage 42 which may be too small a sample to be significant. In terms of identifying relatives, the evidence from Fischer’s raw data is inconclusive. The highest level of overall agreement with another manuscript in the collation is just 83.5%: this is with Codex Perusinus (P in the Oxford Vulgate), which is only extant in Luke.18

15 For Matthew, these are Adolf Jülicher, ed., Itala. Das Neue Testament in altlateinische Überlieferung. I. Matthäus-Evangelium (Berlin: De Gruyter, 1938); Adolf Jülicher, Walter Matzkow, Kurt Aland, eds., Itala. Das Neue Testament in altlateinische Überlieferung. I. Matthäus-Evangelium, 2nd rev. ed. (Berlin: De Gruyter, 1972). The indication in Houghton, Latin New Testament, 225, that this manuscript is included in the Itala is erroneous. 16 Fischer†, “Die lateinischen Evangelien,” 136. 17 This figure was verified from a tally of the relevant section of the collation in Fischer, Die lateinischen Evangelien I. Interestingly, in the portion from Matthew 26 the agreement is roughly 68% but in the portion from Matthew 27 this drops to 52%, although this may say more about the Vulgate text of this passage than VL 28. 18 For more on this manuscript, see Houghton, Latin New Testament, 274.

251

The Garland of Howth Passage 11 (Matt 2:19–4:17) 12 (Matt 8:2–9:8) 13 (Matt 16:9–17:17) 14 (Matt 26:39–58, 27:29–46) 21 (Mark 2:12–3:21) 22 (Mark 7:32–8:35) 23 (Mark 10:17–52) 24 (Mark 14:22–62) 31 (Luke 6:17–49) 32 (Luke 8:12–43) 33 (Luke 10:40–11:32) 34 (Luke 23:35–44, 24:8–13, 24–49) 41 (John 2:18–3:31) 42 (John 7:28–8:16) 43 (John 12:17–13:6) 44 (John 20:1– 21:4) Total

Agreements 173/270 144/255 314/366 282/332 282/340 264/334 210/245 306/359 277/342 57/84 2309/2927

% 64.1 56.5 85.8 84.9 82.9 79.0 85.7 85.2 81.0 67.9 78.9

Rank 4 1 18 20 21 21 19 16 18 [7]19 19

Fig. 1. Vulgate agreement and ranking of VL 28 from Fischer’s collations

All Old Latin witnesses appear among the lowest fifty relatives, showing less agreement with this text than more than 400 Vulgate manuscripts do: Codex Usserianus Primus, which Abbott notes shares a number of readings with the Garland of Howth, comes in 433rd place with an overall agreement of just 62.5% (1410 of 2255 variation units), slightly above Codex Veronensis and Codex Claromontanus. Codex Bezae and Codex Vercellensis are even lower, with agreements of 52.4% (1504/2869) and 51.2% (1365/2667) respectively. The only instances of a more marked agreement between the Garland of Howth and Old Latin witnesses are in the passages from Matthew, as shown in Figure 2. The agreement with Codex Usserianus Primus, the closest witness in both these passages, is still relatively low but exceeds the Vulgate agreement by 12.4% in passage 13 and 8.5% in passage 14. The correspondence with other Old Latin witnesses, particularly in passage 13, seems to confirm an Old Latin affiliation for the Garland of Howth in at least part of Matthew. Nevertheless, more extensive investigation is needed in order to determine the significance of these percentages.

19 The overall ranking tables exclude any witness which is extant in fewer than 100 variation units; this is the rank which would correspond to this percentage in this variation unit.

252

H. A. G. Houghton

Manuscript VL 14: Codex Usserianus Primus VL 8: Codex Corbeiensis VL 6: Codex Colbertinus VL 4: Codex Veronensis VL 11: Codex Rehdigeranus VL 16: Fragmenta Sangallensia VL 13: Codex Monacensis VL 12: Codex Claromontanus

Passage 13 76.5% (101/132) 73.0% (197/270) 72.2% (192/266) 71.9% (194/270) 66.7% (180/270) 66.2% (49/74) 63.6% (171/269) -

Passage 14 65.0% (147/226) 58.4% (149/255) 62.8% (164/261) 59.8% (156/261) 58.4% (149/255) 59.9% (157/262) 61.2% (161/263)

Fig. 2. Old Latin manuscripts agreeing most with VL 28 in Fischer’s collation of Matthew

Since Fischer’s collation, the Garland of Howth has been included in the Vetus Latina edition of John.20 It is assigned by the editors to Group 2B, indicating that its affiliation is with the Vulgate even though a few non-standard readings are shared with other Irish witnesses. However, the decision was taken to exclude it from the Vetus Latina edition of Mark due to the Vulgate nature of its text, based on a comparison of Hoskier’s edition with the Oxford Vulgate.21 As part of the Early Irish Manuscripts Project at Trinity College Dublin, new high-resolution digital images of the Garland of Howth were made freely available on the Library’s website in 2016.22 Conservation undertaken in conjunction with these images means that certain portions illegible to Hoskier can now be read more clearly. A fresh electronic transcription of Matthew, based on these images, was prepared by the Irish Latin Gospel Books Transcription Project in 2017, which forms the basis of the present study.23

Description of the Manuscript The Garland of Howth is a manuscript of the Latin Gospels in the standard Vulgate order copied in Ireland around the year 800. It is written in insular minuscule script, in a single column with normally 26 lines per page. The page size is 24.3

20 P. H. Burton, H. A. G. Houghton, R. F. MacLachlan and D. C. Parker, eds., Vetus Latina. Die Reste der altlateinischen Bibel. 19. Evangelium Secundum Iohannem (Freiburgim-Breisgau: Herder, 2011–). 21 See Haelewyck, Evangelium Secundum Marcum, 7. 22 See http://www.tcd.ie/library/early-irish-mss/launch-of-the-digital-garland-of-howth/ (blog post, 9 February 2016). The Digital Collections of Trinity College Dublin are accessed through http://digitalcollections.tcd.ie/home/ and a PDF of the Garden of Howth downloaded from http://digitalcollections.tcd.ie/content/1647/pdf/1647.pdf. 23 This transcription, by Alan Taylor Farnes, is available at http://epa pers.bham.ac.uk/3023; for more on the project, see Martin McNamara, “Irish Gospel Texts Publication Project,” Proceedings of the Irish Biblical Association 38 (2015): 85–98.

The Garland of Howth

253

by 17.5cm, with a text block of around 20 by 15cm.24 86 folios remain, containing portions from all four Gospels.25 A number of these parchment leaves have suffered damage such as tears, holes, and loss of margins as well as discoloration in certain places. There are two illuminated pages, Christi autem (Matt 1:18, 1r) and the opening page of Mark (22r). The text is divided into sense units by enlarged letters and there is minimal use of punctuation marks. Words are separated by blank space; prepositions are often attached to the following word, and there are occasional instances of unusual word-division across lines (e.g. g-entibus in Matt 24:14, f-actum in Matt 27:57). There is extensive use of abbreviation in addition to the standard nomina sacra: the insular symbols for autem, enim and eius are employed, as well as ÷ for est. Two different abbreviations are used for quae.26 A penultimate vowel is sometimes written underneath the following consonant at the end of a line, but the completion of a line in blank space above or below is rare.27 Although the insular g is used throughout, Hoskier notes that the insular s is uncommon, and only appears from Mark 7:18 onwards.28 At least four copyists worked on the manuscript.29 Corrections are few in number, and there are no marginalia. The orthography of the manuscript presents a variety of interesting features. Typical insular spellings are present throughout, such as the interchange between long i and e, the simplification of double consonants, and erroneous duplications (e.g. nissi for nisi, dixise for dixisse, sufocare for suffocare, occulos for oculos). There is also interchange between o and u (e.g. monus for munus, diabulus for diabolus), y and i, and ie for ei (especially in eicere). H is sometimes added in initial position (e.g. hira for ira), or reinforced by c (e.g. adprechendere for adprehendere, chipocritae for hypocritae). There is confusion between b and v, or b and p, as well as final t and d; s and z are sometimes exchanged, with sizania for zizania and even sabulo for zabulo (originally diabolo: e.g. Matt 25:41). Often, but not always, ae is simplified to e, although it is occasionally written as a

24 Figures from McGurk, Latin Gospel Books, 79. The manuscript is also included in E. A. Lowe, Codices Latini Antiquiores II. Great Britain and Ireland, 2nd ed. (Oxford: Clarendon, 1972), no. 72. 25 The current contents are: Matt 1:18–2:7; 4:24–5:29; 13:7–14:1; 16:13–18:31; 19:26– 26:18; 26:45–27:58; Mark 1:1–3:23; 4:19–5:36; 6:36–16:20; Luke 1:13–2:15; 3:8–6:39; 7:11–11:54; 12:45–14:18; 15:25–16:15; 17:7–19:10; 19:38–22:35; 22:59–23:14; John 5:12– 6:25; 8:7–10:3. 26 These consist of q with a superline (e.g. Matt 13:17) and q followed by a triangle of dots (e.g. Matt 13:16, 44); see further Hoskier, The Text of Codex Usserianus 2, iv–v. 27 One exception is found at the bottom of 54r (Luke 6:5). 28 Hoskier, The Text of Codex Usserianus 2, 59. 29 Hoskier notes that 26v is in a different hand and only contains 24 lines, while other scribes were responsible for folios 59–64, part of 74r, and 82–86 (The Text Of Codex Usserianus 2, ad loc.). In fact, there are at least three hands on 59r and fewer lines per page on the following folios.

254

H. A. G. Houghton

digraph (æ) or e-caudata (ę). The treatment of certain vowels, however, appears to be unique to the Garland of Howth: the third-person plural perfect -erunt is frequently written as -iarunt, and -iabant is sometimes found in place of -ebant (e.g. Matt 27:47); participles in -ens often gain an extra i, such as diciens and uidiens; long e is sometimes supplemented by a, with profeata commonly for propheta, secreato for secreto, habeabant for habebant and pleana for plena. In the other gospels, Hoskier notes the duplication of long vowels such as doo for do or paraa for para as well as the aspiration of t to th.30 A further unusual feature, for an insular manuscript at any rate, is interference in both directions between qu and c, with forms such as nesquitis for nescitis, relincimus for relinquimus, and accibus for a quibus.31 Although the non-standard spellings are sufficiently widespread to form a coherent system, there are also frequent scribal errors. Several lines are absent through homoioteleuton, and there are also examples of dittography.32 The omission of single words or short phrases is particularly common, sometimes resulting in a nonsense reading. The latter is also true of changes in number or case due to scribal inattentiveness; alternations between participles and finite verbs are also common and not always grammatical in context.33 There are a number of occasions on which a mistake may have arisen from the incorrect expansion of an abbreviation in the exemplar, including quicumque for quodcumque on four occasions, quasi for quia sic, quod fecit for profecit and uestros for nostros.34 The miswriting of honone for honore in Matt 13:57 suggests that the exemplar was written in insular script. This is confirmed by the incorporation of a gloss in Matt 27:5, where in place of argenteis the copyist writes arcadgabuthc, a garbled form of an Irish translation of argentum acceptum.35 A Latin gloss is incorporated in the following verse (hic est locum, Matt 27:6) and other doublets could have arisen this way, although they are a common feature of insular

30 Luke 4:6; 10:18; 14:3, 5; 17:8; 19:8; Mark 6:56: see Hoskier, The Text of Codex Usserianus 2, v. Double vowels, especially ii, are indicated by the insular addition of faint strokes above the letters. 31 Matt 17:24; 19:27; and 25:13. This alternation is more common in manuscripts of Italian origin; it appears in VL 7 which derives from a Roman pandect (see Houghton, Latin New Testament, 87). 32 Homoioteleuton: Matt 5:19; 17:16; 18:12–13; 19:28; 20:26–27; 21:44; 24:26; 25:5, 22–23 (possibly also 22:16); dittography: Matt 20:25; 22:30; 26:62; 27:30–31. 33 e.g. accediens…dicens in Matt 18:21, sedentes…audientes…et clamauerunt in Matt 20:30 and errantes nescientes in Matt 22:29. 34 Matt 16:19; 18:18; 26:54; 27:24, 26 respectively; in the last two cases, the abbreviations for quod and pro are very similar, while the similarity between n and u means that abbreviated first- and second-person plural pronominal adjectives are easily confused. 35 See Abbott, Evangeliorum versio antehieronymiana, xvii where the correct form of the Irish is given as airgid gabtha.

The Garland of Howth

255

tradition.36 Nevertheless, errors such as maiestate fratris for maiestate patris in Matt 16:27, nouisime censum for nomisma census in Matt 22:19 and de operibus fidelis in place of de xpo cuius filius in Matt 22:42 simply represent poor copying.37 Few of these errors have been corrected, and the consequence is that readings peculiar to this manuscript must be carefully weighed before being considered as genuine biblical forms which are no longer attested elsewhere.

The Text of Matthew Six portions of the Gospel according to Matthew are preserved in the Garland of Howth: three shorter passages from the first half of the Gospel on one or two folios each (Matt 1:18–2:7; 4:24–5:29; 13:7–14:1) and the majority of the latter half with a couple of interruptions on the remaining seventeen folios (Matt 16:13–18:31; 19:26–26:18; 26:45–27:58). It is surprising that the back of the page on which the text breaks off mid-verse is blank, suggesting that the conclusion of the gospel may never have been copied. In order to examine the textual affiliation of this witness, the electronic transcription was collated against the editorial text of the Stuttgart Vulgate.38 This produced a total of 1790 variants, which were compared with the Old Latin codices reported in the Itala and divided into three categories: purely orthographic variants; agreements with surviving Old Latin manuscripts; readings peculiar to VL 28 (i.e. without parallel in the Itala).39 The latter were subdivided into probable errors (e.g. omissions, nonsense readings) and possible alternative readings no longer preserved in direct tradition. The overall distribution is shown in Figure 3. The effect of the unusual orthography of the Garland of Howth on its agreement with the Vulgate (or, indeed, other witnesses) is immediately obvious: 29% of the variants are simply orthographic. When these are combined with the high proportion of unparalleled readings which are likely to represent copying errors (18%), they account for almost half of the differences from the Vulgate. E.g. sciscitabatur interrogabat at Matt 2:4, where ab eis is also repeated, or accesiarunt erigebant at Matt 17:23. Abbott, Evangeliorum versio antehieronymiana, xvii notes further doublets in Luke as well as the incorporation of a further Latin marginal gloss in Luke 23:12. 37 These are also selected as “examples of carelessness or foolishness” by Abbott, who provides similar instances from the other gospels (Evangeliorum versio antehieronymiana, xvi). 38 R. Weber, R. Gryson et al., eds., Biblia Sacra iuxta Vulgatam versionem, 5 th ed. (Stuttgart: Deutsche Bibelgesellschaft, 2007). The collation interface used was developed by Dr C. J. Smith of the University of Birmingham. 39 As with any analysis of this sort, the exact number of variants is dependent on the definition of the extent of each unit and there are some instances of overlap between classification. The figures given in the following tables should therefore not be taken as absolute, but should allow a margin of error in order to take into account the ambiguity of the data. 36

256

H. A. G. Houghton

Nevertheless, a substantial Old Latin element is clearly present, along with a group of other non-Vulgate readings which may include significant early forms. Type of variant Orthographic Old Latin parallel Unparalleled reading (probable error) Unparalleled reading (possible alternative) Total

Number 524 798 325 143

Proportion 29% 45% 18% 8%

1790

100%

Fig. 3. Classification of variants in VL 28

Breaking down these figures according to the surviving portions of text gives a more detailed picture of the textual affiliation. Figure 4 provides both the numbers for each type of variant and the frequency per hundred words, using the same variant categories as Figure 3:

(1)

1:18– 2:7 No Freq 15 6.1

4:24– 5:29 No Freq 41 8.6

13:7– 14:1 No Freq 44 5.1

16:13– 18:31 No Freq 67 5.3

19:26– 26:18 No Freq 266 6.4

26:45– 27:58 No Freq 91 6.6

(2) (3) (4) (5)

8 4 4 31

9 11 2 63

38 38 4 124

141 65 14 287

455 143 91 955

147 64 28 330

3.3 1.6 1.6 12.6

1.9 2.3 0.4 13.2

4.4 4.4 0.5 14.4

11.1 5.1 1.1 22.5

11.0 3.4 2.2 23.0

10.7 4.6 2.0 24.0

Fig. 4. Distribution of non-Vulgate readings in VL 28 by passage and frequency per hundred words. (1) = Orthographic; (2) = Old Latin parallel; (3) = Probable error; (4) = Possible alternative; (5) = Total.

There is a clear division of the gospel into two. In the first three portions, the frequency of variant readings is between 12.6 and 14.4 per hundred words, whereas from chapter 16 onwards this almost doubles. The proportion of orthographic variations is largely unchanged. Instead, the difference is largely due to a marked increase in variants paralleled in surviving Old Latin manuscripts, which occur at a consistent rate of around 11 per hundred words. In addition, readings characteristic of the Vulgate are almost entirely absent from the fourth portion.40 This confirms Lawlor’s identification of block mixture in the

These are readings only attested in the Itala by the Vulgate and manuscripts known to be partly Vulgate in their affiliation (e.g. VL 6, 7, 11, 15). There are 52 such readings in Matt 16:13–18:31, and the Garland of Howth differs from the Vulgate in 49 of then. The exceptions are the omission of propitius (esto) tibi in 16:22, obumbrauit rather than inumbrauit in 17:5, and the omission of illuc from 17:19. The omissions are not compelling 40

257

The Garland of Howth

manuscript, with the Old Latin portion beginning at some point in the gap between Matt 14:1 and 16:13. The consistency of the figures for the whole of the latter part of the gospel also bears out the impression given in Fischer’s testpassages from the beginning and end of this section. There are few variant readings with Old Latin parallels in the first three sections which are particularly notable. Nine involve an alternative rendering (quoniam for quia at 5:7 and 5:28; magnificant for glorificent at 5:16; eat for mittatur at 5:29; autem for ergo at 13:18; malus for malignus at 13:19; eis for illis at 13:24; absconsa for abscondita in 13:35; maligni for nequam in 13:38). There are several expansions, which are matched by some of the variants in the Garland of Howth not found in other surviving Old Latin gospels.41 Most of the other parallels involve minor changes and may be coincidental. Certain non-Vulgate forms correspond to Greek variants: the singular sine parabula[m] in 13:34; the additions of dicit eis and domine in 13:51; Iohannes rather than Ioseph in 13:55. Even though the first of these is not paralleled in direct Old Latin tradition, it has a strong claim to represent an Old Latin form. The most striking errors in this portion of text are the consistent substitution of the numeral XL in place of LX (Matt 13:8, 13:23), and temporibus in place of messoribus in 13:30 (prompted by tempore a few words previously). In order to classify the Old Latin element in the latter half of Matthew, readings allocated to this category were systematically compared with five Old Latin manuscripts.42 The figures for this agreement are shown in Figure 5: Agreements Percentage

VL 4 331/724 45.7%

VL 5 261/726 36.0%

VL 9 243/741 32.8%

VL 12 404/634 63.7%

VL 14 365/521 70.1%

Fig. 5. Overlap of Old Latin readings in VL 28 with other manuscripts

The overlap with VL 12 and VL 14 is noticeably higher than with the other three manuscripts. This is also demonstrated by no fewer than thirty-four readings shared with these two manuscripts alone of all those reported in the Itala. These often involve different renderings, such as pugnas for proelias (24:6), evidence for the influence of a Vulgate text, given many similar examples of the same types of variation in the manuscript, which leaves obumbrauit as the only distinctive Vulgate form. 41 Expansions paralleled in Old Latin manuscripts: homo iustus (1:19), essaiam profetam (1:22), una pars (5:29), parabulam tritici et sizaniorum (13:36), aures audiendi (13:43), genere piscium (13:47), filius ioseph (13:55), et ideo (13:58); expansions peculiar to VL 28: mariam coiugem (1:24), regis iudae (2:1), non occides non mechaberis (5:21). (Readings are given according to the orthography of VL 28, with abbreviations expanded.) 42 VL 12 and VL 14 were chosen on the basis of their clear parallels with VL 28; the other witnesses were intended to give a broader representation of Old Latin tradition (VL 2 and VL 3 were considered too fragmentary). As the whole text was available for comparison, it was not deemed necessary to re-evaluate the test passages reproduced in Figure 2.

258

H. A. G. Houghton

lamentabunt se for plangent (24:30), senserunt for cognouerunt (24:39), and suscepistis for collexistis (25:35).43 In addition, there are twenty-four readings which the Garland of Howth shares with VL 14 alone (including adprehendiens for tenens in 18:28, angulis uentorum for uentis in 24:31, uxores ducebant for et nuptum tradentes in 24:38, and duxerunt for suscipientes in 27:27), and twentytwo readings peculiar to the Garland of Howth and VL 12 (including dico uobis simile for adsimilatum in 18:23, maximum for magnum in 22:36, uero for autem in 23:8 and narrabitur for dicetur in 26:13). Some of the latter may also have been present in VL 14 where it is now lacunose, such as the distinctive reading fi(n)cta simulatione for hypocrisi in 23:28. The significance of the readings shared with VL 14 is underlined by Fischer’s collation, which shows that et exhiberet in 26:53 and apud uos eram in 26:55 occur only in these two witnesses among all Latin manuscripts surviving from the first millennium; similarly, spoliauerunt for exuerunt in 27:31 is peculiar to VL 12 and VL 28 of the entire collation, although VL 13 reads dispoliauerunt. Readings uniquely shared with other surviving Old Latin manuscripts are fewer in number: two each with VL 6, VL 10 and VL 15; three with VL 7; four each with VL 3 and VL 4; seven with VL 2; eight each with VL 8, VL 9 and VL 13; thirteen with VL 5.44 The various permutations of agreement with two or more surviving Old Latin witnesses are typical of the piecemeal attestation of readings in the pre-Vulgate tradition. The thoroughgoing Old Latin character of the Garland of Howth in the latter part of Matthew is also demonstrated by the consistency of its renderings and the presence of interpolations. For example, it always has coloni for γεωργοί (Vulgate agricolae; 21:33–41), sapientes for φρόνιµοι (Vulgate prudentes; 25:2–4), and puer or infans for παιδίον (Vulgate paruulus; 18:2–5). Quoniam is found several times for ὅτι where the Vulgate has quia (18:10; 23:10, 27; 24:33, 34), the demonstrative hic is added before mundus (16:26, 18:7) and in is present before place names (16:21; 17:23; 20:17).45 Other consistent non-Vulgate forms include accipere for δέχεσθαι (twice in 18:5) and suscipere for συνάγειν (25:35, 38, 43); εὐθέως is normally rendered by confestim (20:34; 21:2, 3; 26:49) or Fischer’s collation shows that only six manuscripts have abscidit rather than amputauit in 26:51 (a harmonisation to John 18:10): in addition to VL 12, 14, 28, these are the Rushworth Gospels (Hr), Bodmin Gospels (Hx), and New York Public Library 115 (Bl), all of insular origin. Similarly, ad turbas rather than turbis in 26:55 is just found in five of these six manuscripts (not the Rushworth Gospels) and three other insular witnesses (VL 30, VL 35, and London, BL Royal 1 A XVIII). 44 The comparative dataset for these is much smaller, as it is taken from the Itala rather than Fischer’s collation. Many similarities may be coincidental, such as the three omissions shared with VL 13 or the four shared omissions with VL 5 alone. The most compelling of the peculiar readings are uobis for illis in 18:19 (VL 2), datum for paratum in 20:23 (VL 3), sumpserunt for acceperunt in 25:4 (VL 7), and alius for unus in 24:40 (VL 8). 45 For other instances of quoniam, see 5:7; 16:28; 23:39; hic mundus is also found at 13:38, while at 24:21 saeculum is found in place of mundus. 43

The Garland of Howth

259

continuo (25:15; 26:74; 27:48), while adprehendere is twice found for κρατῆσαι (18:28; 22:6) and deludere renders ἐµπαίζειν on three occasions (20:19; 27:29, 41).46 The Garland of Howth has the interpolations present in several Old Latin witnesses at Matt 20:28; 23:14; 24:31 and 27:35, and also places 17:12b after 17:13. The only one of the common interpolations which is not present is at Matt 24:41; although comparable omissions elsewhere due to homoioteleuton might prompt the suggestion that this is a copying error, this is the only one of the five interpolations missing from VL 14, so its absence from VL 28 may well reflect its textual tradition. In addition, VL 28 has a long interpolation at the end of Matt 27:49 based on John 19:34: alius autem accepta lancia pupungit latus eius et exiit aqua et sanguis. Although this is not paralleled by any of the manuscripts reported in the Itala, it is also found in VL 30 and a number of Greek majuscule manuscripts, which demonstrates its antiquity.47 Among the Old Latin readings already known from other manuscripts, the additional support of the Garland of Howth is worth observing in several places. In Matt 17:15 (17:14 in the Vulgate), it has torquetur rather than patitur for πάσχει, a reading found in just five manuscripts from the first millennium, as well as aliquando (from ἔνιοτε rather than πολλάκις).48 In keeping with other Old Latin witnesses, it reads ignem rather than supplicium in 25:46 (a harmonisation to 25:41), has duodecim milia in 26:53, and includes unicum in 21:37, in eum quicquam in 26:60, audierunt in 26:61, in faciem eius in 27:30 and et postquam crucifixus est at the beginning of 27:45, none of which have any Greek support. The additions of tunicam purpureum in 26:60 and uenientem obuiam sibi in 27:32 are only found in Greek in Codex Bezae. VL 12 is the sole other instance of aduersus eum for ἐπ’ αὐτόν in 22:34, while in 24:42 VL 28 shares with VL 14 the conflation of both Greek readings, as qua die uel hora. VL 28 also reads stateres not argenteos in 26:15, begins 26:60 with et non inuenerunt, has filius dei uiui in 26:63, and Pharisaeis rather than senioribus in 27:41, all of which correspond to Greek alternatives. It is interesting that, while both VL 3 and VL 28 have the singular noun princeps sacerdotum in 27:6, which is probably a misreading, only VL 28 has the singular verb dixit.49 Another unusual

46 In contrast, tenere for κρατῆσαι occurs at 21:46; 26:4, 48 and 57, but on all four occasions the Old Latin tradition is invariant; the sole occurrence of statim in VL 28 is at 24:29, where it is the majority Old Latin reading; similarly, inluserunt is only found in VL 28 at 27:31. 47 For more on interpolations in the Latin tradition of Matthew, see Houghton, Latin New Testament, 158–60, which notes that 24:31 and 27:49 are among the five typical of insular manuscripts; the Garland of Howth is not extant for the others in Matt 8:24; 10:29 or 14:35. 48 The witnesses to torquetur reported in Fischer’s collation are VL 12, 14, 28, 35 and the Rushworth Gospels. 49 VL 28 also uniquely has princeps sacerdotum with a singular participle at 21:15; other Old Latin manuscripts have the same variation in 27:20.

260

H. A. G. Houghton

reading is non uenistis ad me in 25:43, shared with VL 8 and VL 9, where other Old Latin codices have non uisitastis me. Given the Old Latin affiliation of the Garland of Howth in the latter half of Matthew, some of the variants in this section which are not paralleled by manuscripts reported in the Itala may preserve pre-Vulgate readings. The collation gives a total of no fewer than 133 possible alternatives. The most likely are those which match renderings found elsewhere in Old Latin tradition or correspond to known Greek variants, while minor substitutions or changes in word order are less compelling. The interpolation at Matt 27:49 has already been mentioned above; another substantial variant reading with Greek support is found at Matt 21:29–30, where VL 28 reverses the order of the sons, with the first agreeing to go but then failing, as found in Codex Vaticanus, Family 13 and other Greek manuscripts. The following readings are also not transmitted in other Old Latin codices but find some correspondence in Greek tradition: 16:21 18:3 18:10 18:15 18:17 19:28 20:7 20:9 20:15 20:19 20:21 20:28 20:28 20:30 21:14 21:19 21:35 22:6 22:18 23:8 23:19 23:20 23:20 23:21 23:22

quia ] quod paruuli ] infantem (pro infantes) pusillis ] + qui in me credunt si ] quod si si autem et ecclesiam non ] si uero nec aeclisiam regeneratione ] generatione ista in ] ad ergo ] autem quod uolo facere ] + in eis (pro in meis) flagellandum ] ad flagillandum ait illi dic ] at illa dixit ei eminentioribus ] maioribus utilius ] gloria audierunt ] audientes sanauit ] sanabat illi ] ad eam ceciderunt ] coederunt adfectos ] adflictos cognita ] cognoscens uester ] + Christus an ] uel ergo ] enim in eo ] per eum eo ] eum super eum ] in ipso

23:25 23:32 23:39 24:3 24:18 24:22 24:31 25:10 25:12 25:39 25:43 25:46 26:5 26:9 26:57 26:65 27:28 27:29 27:35 27:40 27:42 27:46

inmunditia ] iniquitate implete ] adinpletis uobis ] + quoniam secreto ] in saecriato (i.e. in secreto) tunicam suam ] tonicas suas fieret salua ] saluasset eius ] suos dum autem irent ] euntes ait ] dicit aut ] uel nudus...infirmus ] nudus fui...infirmus fui aeternam ] perpetuam dicebant autem ] et dicebant multo ] pretio magna (pro pretiomagno) ubi ] in quo egemus ] opus uobis exuentes eum ] cum spoliassent eum uestiarunt plectentes coronam de spinis ] coronam de spinis texerunt et sortem mittentes ] miserunt sortes destruit ] distruas ei ] in eum et circa ] circa uero

In the interpolation at 20:28, the Garland of Howth is the only witness to have in locis maioribus for είς τοῦς ἐξέχοντας τόπους, while in reading gloria at the end of this addition it corresponds to the parallel passage at Luke 14:10 (δόξα rather than χρήσιµον). At Matt 23:25, iniquitate represents the Greek ἀδικίας rather

The Garland of Howth

261

than ἀκρασίας or ἀκαθαρσίας. Although euntes corresponds to the Greek participle at 25:10, in context this is not grammatical and could be a scribal error for euntibus. Hoskier sets much store by audientes in 20:30 (ἀκούσαντες), but this participle may also be erroneous since it creates an anacoluthon with et clamauerunt in the next phrase.50 While both pretio and multo are found in Old Latin manuscripts for πολλοῦ in 26:9, pretio magno is peculiar to VL 28. There is a cluster of unique readings in Matt 27:28–29: cum spoliassent to translate ἐκδύσαντες is also found in this manuscript and VL 12 two verses later, while uestiarunt may be a doublet (cf. uestientes for ἐνδύσαντες in VL 5); texerunt as a rendering of πλέξαντες is found in Old Latin manuscripts at John 19:2 but Fischer’s collation shows that the Garland of Howth is the sole witness to this form in Matthew.51 Most if not all of the readings in the Garland of Howth which are not paralleled in manuscripts reported in the Itala and do not have Greek support are secondary developments. Many of these are expansions, including amum in mari in 17:26, the addition of in saeculo isto after accipiet in 19:29, Essaiam profeatam in 21:4, resurrectione mortuorum in 22:30 (possibly harmonised to the next verse, where mortuorum is missing from VL 28), uocari ab hominibus in 23:8, uenire ad me in 24:48, emere oleum in 25:10, tradetur principibus sacerdotum in 26:2 and seniores populi in 27:20. The pronouns ipsi and hii in 25:44 and 25:46 are replaced by iniusti, while at 26:14 duodecim is substituted by discipulis suis. While a number of Old Latin witnesses have the erroneous ad dextris…ad sinistris in 27:38, in VL 28 this is corrected to ad dexteram…ad sinistram.52 Even sanguine iusti in 27:4 appears to be a simplification of the unusual original construction sanguine iustum (αἷµα ἀθῷον). A handful of odd readings deserve a brief mention. Ideo rather than et ego at the beginning of 16:18 has no Greek support, and may be a misreading: quomodo for quod in 22:31, and in eo for in dono in 23:18 are similar. The command defer ecclesiae in 18:17 is comparable to referes ecclesiae in VL 12, even though dic in other witnesses matches εἰπέ/εἶπον in Greek.53 Likewise, occident se hardly corresponds to µισήσουσιν ἀλλήλους in 24:10, yet VL 14 has occident inuicem here: this could derive from a misreading of odient inuicem (VL 2, 12), or it may somehow reflect the poorlyattested Greek variant παραδώσουσιν εἰς θάνατον.54 In oportuit ergo uenundari 50 Hoskier, The Text of Codex Usserianus 2, viii–ix; compare also the introduction of the participle deludentes in VL 28 for ἐνέπαιξαν in 27:29. 51 There is, in fact a doublet in VL 30 (et plectentes…texerunt), while VL 262 has intexentes. 52 On this variation, see P. H. Burton, The Old Latin Gospels (Oxford: Oxford University Press, 2000), 39. 53 Augustine Serm. 295.2 and Caesarius of Arles Serm. 28.3 have refer ad ecclesiam (cf. Jerome Ep. 125.19.6). 54 This is found in GA 043, and supported by both quotations from Arnobius in the Vetus Latina Database.

262

H. A. G. Houghton

pecuniam meam in 25:27, uenundari may be a corruption of te ut dares, as found in Augustine (Serm. 137.15; cf. te dare in VL 9, 12 and 14), or come from a doublet such as mittere uel dare in VL 30.55 At 27:48, VL 28 uniquely reads surgiens rather than currens (δραµών), which is difficult to explain other than a simple misreading.

Conclusion This investigation has confirmed the importance of the Garland of Howth for the Latin text of Matthew, as recognised by Hoskier and his predecessors. In the latter half of the gospel, it preserves a text with a consistent Old Latin affiliation. This displays marked similarities to VL 14 (Codex Usserianus primus) and VL 12 (Codex Claromontanus), manuscripts which are often characterised as sharing a “Gallo-Irish” text which probably had its origins in fourth-century Italy.56 The overlap of the extant portions of these codices in the latter part of Matthew facilitates their comparison: in addition to readings shared with one or both of these witnesses, the Garland of Howth also displays parallels with other Old Latin manuscripts. What is more, it preserves a number of readings not present in the principal Old Latin manuscripts but corresponding to Greek forms, most notably the ordering of Matt 21:29–30 and the interpolation in Matt 27:49. These may go back to an early strand of Latin biblical tradition. Other non-Vulgate forms appear largely to be secondary, featuring several expansions and grammatical adjustments. The manuscript is also characterised by a relatively high proportion of copying errors and very unusual orthography, especially for long vowels, which is worthy of a study in its own right. The use of electronic tools and digital data have made a distinctive contribution to the present study. The high-resolution online images were indispensable in the preparation of the transcription. Not only did they represent a significant improvement on earlier photographic prints from microfilm, but they enabled careful examination and re-examination of the manuscript without the restrictions of library opening hours: travel to Dublin was only required for the presentation of the results! The release of the complete proofread transcription online, through a centrally-maintained university repository, allows subsequent researchers to use the same electronic data for different forms of comparison.57 Compare the copying errors in Matt 16:27; 22:19 and 22:42 noted above. VL 12 was copied in Italy in the fifth century; an early Italian origin has also been suggested for VL 14 by D. N. Dumville, A Palaeographer’s Review: the Insular System of Scripts in the Middle Ages I (Osaka: Kansai University Press, 1999), 35–40. 57 Open access to this data is also ensured by the licensing of the file under a Creative Commons Attribution license, meaning that it may be freely re-used with acknowledgement to the original creators. 55 56

The Garland of Howth

263

For example, it is intended that it will in due course be redeployed in a synopsis or collation of manuscripts on the website of the Irish Latin Gospel Books Transcription Project. Similarly, the presence online of the full transcription means that the peculiar spellings of the manuscript may be discovered by historical linguists through search engines. The flexibility of this digital data is thus in contrast to Hoskier’s transcription and the printed collations which preceded it. In addition, the automatic generation of a list of differences spanning the whole text of Matthew removes the potential at this stage for the errors so criticised by Hoskier, although some may have been subsequently introduced when the readings were evaluated and the collation was edited. The principal departure from the digital medium was in the identification of parallels for the non-Vulgate readings. Once other Latin gospel texts are transcribed in their entirety, it should become possible to improve this stage of the process. As it is, the availability of such transcriptions looks set only to increase gradually through the creation of electronic resources in conjunction with studies such as the present chapter. Nevertheless, it is hoped that, now it has entered the digital realm, the Garland of Howth will no longer be neglected as a witness to the Old Latin gospel text.

Hoskier and His (Per)Version of the Ethiopic Curt Niccum The discovery of P46 captivated H. C. Hoskier. From his perspective, the manuscript’s early date and well-preserved state afforded a unique opportunity to probe the text of the Pauline Epistles prior to the period of the “great majuscules” and the Sahidic version’s influence on the transmission of the Greek New Testament.1 Because the papyrus promised to confirm previously articulated hypotheses, he rejected Kenyon’s negative assessment of the manuscript’s scribe.2 Instead, Hoskier granted the copyist a fair amount of ability and, correspondingly, accuracy.3 Thus rather than betraying an indifferent, inexact scribe, the numerous unique readings of P46 attested a different, more exact text. In order to rehabilitate the maligned scribe, Hoskier applied all of the evidence he could muster.4 Canvasing the manuscript, versional, and patristic evidence for the slightest agreements with P46, he tried to demonstrate that most of the scribal errors were attested in the material record and therefore should not be Hoskier preferred Kenyon’s early third century date over the slightly later date of Sanders. Cf. Frederic G. Kenyon, The Chester Beatty Biblical Papyri: Descriptions and Texts of Twelve Manuscripts on Papyrus of the Greek Bible, III/3 (Supplement): Pauline Epistles, Text (London: Emery Walker, 1936), xiv; and H. A. Sanders, A Third Century Papyrus Codex of the Epistles of Paul, University of Michigan Studies, Humanistic Series 38 (Ann Arbor: University of Michigan Press, 1935), 13. Hoskier’s reference to the Sahidic stems from his earlier hypothesis that all Greek manuscripts known at that time exhibited considerable influence from the versions. See Hoskier, Concerning the Genesis of the Versions of the New Testament (Gospels), 2 vols. (London: Bernard Quaritch, 1910–1911), and Hoskier, Codex B and Its Allies: A Study and an Indictment, 2 vols. (London: Bernard Quaritch, 1914). 2 Kenyon, Papyri, III/3, xix–xxi. 3 “[O]ur scribe was not a careless ignoramus, nor somnolent, nor inept, for most scribes, in my experience, are honourable copyists, and he is not an exception.” Hoskier, “A Study of the Chester-Beatty Codex of the Pauline Epistles,” JTS 38 (1937): 150. 4 Hoskier, Appendix to the Article on the Chester-Beatty Papyrus of the Pauline Epistles Known as P46 in the Journal of Theological Studies No. 150 Setting Forth Here in Supplementary Detail the Short Text of That Important Document (Oxford: Oxford University Press, 1937); Hoskier, A Commentary on the Various Readings in the Text of the Epistle to the Hebrews in the Chester-Beatty Papyrus P46 (circa 200 A.D.) (London: Bernard Quaritch, 1938); Hoskier, P46: Addenda et Corrigenda (Oxford: Oxford University Press, 1937); and Hoskier, “Chester-Beatty Codex.” 1

266

Curt Niccum

disregarded. Surprisingly, one of the most fruitful resources for his defense was the Ge‘ez version. He identified fifty-two variations in the thirteen letter Pauline corpus where the Ethiopic “practically alone” agreed with P46.5 He found seventy-four additional readings with support from the Ge‘ez and other witnesses.6 The labour Hoskier expended to defend this important document boggles the mind. He knew the readings of P46 in detail. He carefully collated Greek manuscripts and poured over versional and patristic evidence. And he accomplished this at a time when just obtaining the necessary sources was a monumental task. Presumably, Hoskier’s mind would be equally boggled if he could only observe how much easier technology has made these same labours 75 years later. Today’s technological advancements aside, Hoskier’s blunders should have been easily recognizable years ago. The numbers alone should have raised suspicions. In one publication, Hoskier lists 274 variants in P46 that mark “the shorter text” of the Pauline Epistles.7 In only 26 of these, or less than 10%, did he indicate support from the Ge‘ez version.8 By his reckoning, out of another 101 omissions enumerated in his Addenda et Corrigenda, the Ethiopic version shares the same or similar reading in only twenty-one.9 These low numbers were of no concern to Hoskier for they validated his premise. However, these paltry percentages should have alerted subsequent text critics to the limitations of Hoskier’s claims and possibly to the methodological flaws behind them. Instead, scholars neither attended to these questionable numbers nor took advantage of new developments and discoveries. As a result, the myth of a close kinship between the Ge‘ez and P46 was perpetuated based solely on Hoskier’s list of distinctive agreements. Thus, Arthur Vööbus and Bruce Metzger both mention Hoskier’s identification of “more than fifty readings.”10 Josef Hofmann, who had the wherewithal to critically assess Hoskier’s methodology, avers that Hoskier, Appendix, 5. Eighteen more are listed in Appendix, twenty in Addenda, and thirty-six from Hebrews in Commentary. 7 Hoskier, Appendix, 7–17. This list omitted the 60 variants in Hebrews that he would examine later in his Commentary. 8 This number includes all references to aeth and aeth?; occurrences of “cf. aeth” were not counted. 9 Hoskier, Addenda, 2–8. 10 Arthur Vööbus, Early Versions of the New Testament, PETSE 6 (Stockholm: Estonian Theological Society in Exile, 1954), 269; and Bruce Metzger, The Early Versions of the New Testament: Their Origin, Transmission and Limitations (Oxford: Clarendon Press, 1977), 232. Metzger adds the important caveat that Hoskier’s research was based on printed editions. This cautious approach was unfortunately not maintained in his introduction to New Testament textual criticism, for there he considers it a “fact” that the “version frequently agrees with p46 with little or no other support” (The Text of the New Testament: Its Transmission, Corruption, and Restoration, 2nd ed. [Oxford: Oxford University Press, 1968], 84; Bart Ehrman’s update retains this sentence verbatim in the book’s fourth edition [2005], 120). 5 6

Hoskier and his (Per)Version of the Ethiopic

267

Hoskier had “surprisingly established” the version’s agreement with P46 in these readings.11 And as recently as 2014, Tedros Abraha reiterated their unique relationship in his critical edition of the Corinthian Epistles.12 The only serious challenge to Hoskier came from an unexpected quarter: James Royse’s dissertation on scribal habits in the papyri.13 In an appendix, which was unfortunately not incorporated into the published version of the thesis, Royse systematically dismantled Hoskier’s work. Two important contributions in this appendix were (1) the identification of places where Hoskier outright erred, and (2) the observation that Hoskier relied too heavily on the Latin translation of the Ethiopic found in Walton’s polyglot. Royse’s response, though, was limited partially by the scope of his dissertation, which lent itself towards an abbreviated treatment and only required an examination of Hoskier’s list of unique readings, and partially by the lack of resources, for Royse, like Hoskier, only utilised printed editions of the Ge‘ez New Testament.14 Several advancements now make possible a more accurate and nuanced evaluation of Hoskier’s work. First, critical editions of New Testament books in Ge‘ez have started to appear. Josef Hofmann’s 1967 edition of Revelation was the first. It took another two decades before Rochus Zuurmond’s work on the Gospels appeared, but since then research on the Ethiopic version has advanced at a much faster pace. Recent editions provide greater clarity about the Pauline corpus’ transmission history and establish a firmer basis for reconstructing its earliest attainable text. Second, important manuscripts continue to be discovered. In some instances the discovery consists of a library or university publishing a catalogue of manuscripts or digitizing documents already in their possession. At other times, true 11 “Er hat überraschenderweise festgestellt, daß an die 50 Lesarten des äthiopischen Paulustextes mit dem griechischen Text des Chester Beatty-Papyrus 46 übereinstimmt” (J. Hofmann, “Das Neue Testament in äthiopischer Sprache,” in Die alten Übersetzungen des Neuen Testaments, die Kirchenväterzitate und Lektionare, ANTF 5, ed. K. Aland [Berlin: De Gruyter, 1972], 364). 12 Tedros Abraha, The Ethiopic Versions of 1 and 2 Corinthians (Rome: n.p., 2014). He refers to P46 on p. 9. On pp. 98–9 he lists a few readings that intersect with early Greek majuscules, including P46, but he does not attempt to test textual consanguinity. Abraha’s editions of Romans and Hebrews offer even less analysis: La lettera ai Romani: Testo e commentari della versione Etiopica, Äthiopistische Forschungen 57 (Wiesbaden: Harrassowitz, 2001) and The Ethiopic Version of the Letter to the Hebrews, Studi e Testi 419 (Vatican City: Biblioteca Apostolica Vaticana, 2004), but see p. 9 where he claims that P46 is his point of comparison in the Greek tradition. 13 James R. Royse, “Scribal Habits in Early Greek New Testament Papyri” (Th.D. diss., Graduate Theological Union, 1981), 694–724. 14 The printed editions consulted by Hoskier and Royse are Thomas Pell Platt, ed., Ethiopic New Testament (London: BFBS, 1830); and Tasfa Sion, ed., Novum Testamentum..., 2 vols. (Rome: V. and L. Dorici, 1548) reproduced in Brian Walton, ed., Biblia Sacra Polyglotta, 6 vols. (London: Thomas Roycroft, 1657).

268

Curt Niccum

discoveries are made, which through digitization are made publicly accessible while allowing the artefact to remain in situ. An example of the former is MS Ethiopic 1, a fourteenth century codex of the Pauline corpus at Harvard University. An example of the latter is the oldest known Ge‘ez manuscript of the Pauline Epistles, Sinai New Finds Ethiopic 2, a codex from no later than the thirteenth century located at St. Catherine’s monastery.15 It is a valuable supplement to the handful of witnesses that attest the earliest known Ge‘ez text. Third, the advancement of scholarship on the Ethiopic Bible as a whole is providing a wealth of details about translation technique.16 Although multiple Ethiopians tackled the earliest corpora of Christian literary works, they shared certain approaches. In addition, the idiosyncrasies of individual translators can sometimes be detected. For example, although Ge‘ez does not formally mark definite articles, a translator could create ways to represent them. Although generally eschewed by most, the translator of Revelation fastidiously employed such devices.17

Hoskier and the Ge‘ez Version As a result of the progress made in these areas, the deficiencies in Hoskier’s resources and methods become even more apparent. For the sake of this brief investigation, only the forty-two variations in 1 and 2 Corinthians that Hoskier claims have Ge‘ez support will be examined.18 After a survey of simple mistakes, Hoskier’s neglect of sources and translation technique will be documented.

15 Images of the manuscript have been made available to the Center for the Study of Ancient Religious Texts at Abilene Christian University thanks to the kind permission of Saint Catherine’s Monastery, Sinai, Egypt, and the generosity of Father Justin Sinaites. See Justin Sinaites, “Newly Recovered Manuscripts of the Scriptures from Saint Catherine’s Monastery, Sinai,” Restoration Quarterly 59 (2017): 5–24. The manuscript dates no later than the thirteenth century according to Ted Erho (private correspondence, 15 January 2018). 16 Michael Knibb, Translating the Bible: The Ethiopic Version of the Old Testament (Oxford: Oxford University Press, 1999), 55–112. See also Rochus Zuurmond, Novum Testamentum Aethiopice: The Synoptic Gospels, Part 1: General Introduction and Edition of the Gospel of Mark, Äthiopistische Forschungen 27 (Stuttgart: Franz Steiner, 1989), 49–56; and Josef Hofmann, Die äthiopische Johannes-Apokalypse kritisch untersucht, CSCO 297, Subsidia 33 (Leuven: CSCO, 1969), 67–105; and Hofmann, “Limitations of the Ethiopic in Representing Greek,” in Metzger, Early Versions, 240–56. 17 Although the gender of the biblical translators is unknown, historically they were probably male. Masculine pronouns will therefore be employed when referring to the translator of the Pauline corpus. 18 This essay is developed from a portion of “The Ethiopic Version of 1 and 2 Corinthians and Its Value for Text Criticism in Light of a New Witness,” presented at the Annual Meeting of the Society of Biblical Literature, Boston, 20 November 2017.

Hoskier and his (Per)Version of the Ethiopic

269

Errors Although a careful scholar, Hoskier was not immune to making mistakes. Royse has previously drawn attention to three particularly egregious ones: At 1 Cor 15:47 ሰማያዊ translates οὐράνιος; there is no Ge‘ez equivalent for πνευµατικός in this verse.19 In 2 Cor 5:19 the Ge‘ez repeats the phrase መልእክተ ሣህሉ, “message of mercy,” from the end of v. 18; it therefore diverges from all known Greek readings, including εὐαγγελίον in P46.20 And in 2 Cor 9:2 Hoskier equated ከመ with ὅτι.21 As often takes place in Ethiopian translation, a noun (προθυµίαν) is transformed into a subjunctive verb (ትጽህቁ) to which ከመ naturally belongs. To these must now be added three more blatant errors. The first concerns the number of occurrences of τὸ αὐτό in 1 Cor 10:3–4.22 Hoskier proposes that the Ethiopic lacks this phrase in both verses. He was confident the phrase did not appear in v. 4 (aeth), but he had some reservations about v. 3 (aeth?). The evidence actually supports the opposite.23 ዘውእቱ in v. 4, although not a felicitous translation of τὸ αὐτὸ, clearly renders the phrase. The lack of a parallel in v. 3 strongly suggests that the Vorlage lacked τὸ αὐτὸ there. Both verses therefore align not with P46, but with 04*, the manuscript with which the Ethiopic version of Paul’s letters most frequently agrees.24 The second error occurs with the supposed omission of ἢ in 2 Cor 3:1. Here the Latin translation led Hoskier astray, for the Latin treats the prefix ዘ as a demonstrative pronoun referring to መጻሕፍት, “letters,” i.e., etiam epistolis.25 Contrary to the Latin rendering, the position of ዘ makes clear that it functions as a relative pronoun; the text reads ዘእምኀቤነ,“which are from us,” not ዘመጻሕፍቲሆሙ, “those letters.” Dependent upon the faulty Latin, Hoskier could not avoid deducing that the Ethiopian translator had πρὸς ὑµᾶς ἐξ ἡµῶν instead of πρὸς ὑµᾶς ἢ ἐξ ἡµῶν before his eyes. But an eta undoubtedly stood in the Vorlage. Because the translator did not recognize that Paul was posing a series of questions, he was Royse, “Scribal Habits,” 700. Royse, “Scribal Habits,” 701. 21 Royse observes that this identification would require ὅτι to follow rather than precede οἶδα and thus would still disagree with P46 (“Scribal Habits,” 701). 22 The folio containing these verses is absent from the Sinai codex. 23 Royse argues that the awkward translation prohibits any confident assertions about the Vorlage, granting only that Pell Platt’s edition alone clearly supports an omission of the phrase in v. 4 (“Scribal Habits,” 709). 24 Royse grants that Pell Platt’s edition might support Hoskier’s claim (“Scribal Habits,” 709, 713). Unfortunately he does not also mention the strong Arabic influence on the edition that makes it unfit for determining the earliest attested text. On the consanguinity of the Ethiopic text and 04, see below and Niccum, “1 and 2 Corinthians.” 25 Royse first identified the cause of Hoskier’s error in “Scribal Habits,” 709, but his own treatment of the Ethiopic text is puzzling. Despite placing 2 Cor 3:1 in the category of passages where the Ge‘ez “alters the text considerably from that found in p46,” he then (incorrectly) reckons that the Ge‘ez shares P46’s nonsense reading. 19 20

270

Curt Niccum

unprepared for the word “or” (ἢ) in the context. As a result, the translator interpreted the letter as a relative pronoun (i.e., ἣ), which he “faithfully” rendered with ዘ. Finally, Hoskier inexplicably suggests that P46 and the Ethiopic version omit the second σάρξ in 1 Cor 15:39. The question mark that Hoskier places next to P46 hardly ameliorates the audacity of this claim.26 P46 lacks the text in question due to damage at the bottom of the folio, but the preceding line of text is complete and the tops of the very last words of the following line are clear enough to calculate how long the lacunose text might be. Hoskier does not reveal his reasoning for positing this omission, but it must have proceeded thus: (1) Based on a line length of 24 letters, which occurs a few times on this page of P46, the text is four letters shorter than that in the printed Greek New Testament; (2) the Ethiopic version does not translate the second σάρξ; (3) σάρξ is a four-letter word; (4) therefore, both P46 and the Ge‘ez omit σάρξ. However, in the lacunose portion there is an expected sequence of four fourletter words: αὐτὴ σάρξ ἀλλ᾽ ἄλλη. It is odd, therefore, to conclude that the one word that must be missing is σάρξ, especially considering the concrete possibility of parablepsis with ἀλλ᾽ ἄλλη. It is more than odd to resort to word counts on a page having some lines of 28 (and more!) characters, making the conjecture superfluous. A different type of error sometimes occurred with Hoskier’s employment of the Roman and Pell Platt editions, for neither transmits the earliest attested text. At 2 Cor 1:5, the Roman edition prints a later correction (ፍሥሓሁ, “comfort of”).27 The reading of the earliest attested text, ፍሥሐነ, “our comfort,” expresses the pronoun ἡµῶν absent in P46.28 Hoskier was also led to believe that the Ge‘ez lacked µέν at 2 Cor 10:10, but the earliest attested text actually has an equivalent, the suffix ሰ (መጽሕፍትሰ, “also [the] letters”).29 Translation Technique The majority of Hoskier’s inaccuracies stem from his failure to consider translation technique. As mentioned above, there are a number of characteristic

Hoskier, Addenda, 4. This error is also found in Ambrosiana B20 and therefore is included in the diplomatic text of Abraha’s critical edition. Although Hoskier only indicates that the Ethiopic version lacked ἡµῶν, Royse points out that the edition from which Hoskier laboured had ፍሥሓሁ ለክርስቶስ “the comfort of Christ,” which disagrees with P46 anyway (“Scribal Habits,” 700). 28 The disagreement with P46 may be apparent rather than real, for pronominal suffixes are often added (see below). Although the Ge‘ez cannot be used to determine the presence or absence of ἡµῶν, had Hoskier consulted the earliest text, he would not have included this variant in his list. 29 Because Royse, too, depended upon these later traditions, to challenge Hoskier he has to argue that µέν occasionally goes untranslated in Ge‘ez (“Scribal Habits,” 707). 26 27

Hoskier and his (Per)Version of the Ethiopic

271

activities shared by early Ethiopian translators. When exercised, they often produced different sentence structures, changed grammatical relationships, and altered wording. Therefore, where a translator opts for native word order, omits redundancies, or converts passive voice to active or indirect speech to direct, the Ge‘ez text almost always defies retroversion into Greek, let alone into Latin and then into Greek, Hoskier’s modus operandi. Transpositions Ge‘ez translations often follow the Greek slavishly, resulting in stilted language, but there is an obvious inclination toward adopting native syntactical patterns. To accommodate, translators often transposed sentence elements. As a general rule, if the Ethiopic text attests a different word order than that found in the Ausgangstext and also conforms to natural Ge‘ez syntax, conclusions about the underlying Greek cannot be made.30 Both Hoskier and Royse miss this point at 1 Cor 13:13. Ge‘ez sentence structure virtually demands placing the equivalent of τὰ τρία ταῦτα after µένει, thus agreement with P46 cannot be determined. The same problem occurs at 1 Cor 14:36.31 Omission of Redundancies Across the board, early Ethiopian translators seem uncomfortable with redundancy. As a result, the text critic must proceed with caution when dealing with omission of material that might have been perceived as unnecessarily repetitive. When the Ge‘ez supplies አጋዕዚኒ, “freeman,” for ὁ ἐλεύθερος κληθείς at 1 Cor 7:22, Hoskier assumes that the Vorlage, like P46, lacked an article. Understandably, Royse concluded that this interpretation “rest[ed] on nothing more than confusion about the syntactic resources available in Ethiopic” because Ge‘ez does not naturally express the definite article.32 But this particular reading is more Numerous studies have revealed that the Ge‘ez Bible, except in the Gospels, has affinities with Codex Sinaiticus or Codex Vaticanus, both of which have New Testament texts similar to the reconstructed Ausgangstext published in NA28 and UBS5 . Therefore, it is safer to begin with the Ausgangstext as the point of comparison when determining textual variation and assessing textual affinity. For an overview of the Ethiopic version and these codices in the Old Testament, see Armin Lange et al., eds., Textual History of the Bible, vols. 1–2 (Leiden: Brill, 2016–2018). For the New Testament see Curt Niccum, The Bible in Ethiopia: The Book of Acts, Ethiopic Manuscripts, Texts, and Studies 19 (Eugene: Pickwick, 2014), 47–68; Siegbert Uhlig and Helge Maehlum, Novum Testamentum Aethiopice: die Gefangenschaftsbriefe, Äthiopistische Forschungen 33 (Stuttgart: Franz Steiner, 1993), 49–59; Josef Hofmann and Siegbert Uhlig, Novum Testamentum Aethiopice: Die Katholischen Briefe, Äthiopistische Forschungen 29 (Stuttgart: Franz Steiner, 1993), 41–9; and Hofmann, Johannes-Apokalypse, 38–66. 31 Royse lists both of these passages as agreements without comment (“Scribal Habits,” 713). 32 Royse, “Scribal Habits,” 698. 30

272

Curt Niccum

complicated, for translators regularly transformed articular participles into finite verbs prefixed by ዘ, “the one who.” Because this construction appears earlier in the verse (ዘይትቀነይ = ὁ…κληθεὶς δοῦλος), the absence of ዘ in the second clause is all the more remarkable. Nevertheless, because the translator omitted the second occurrence of the participle κλῆθεις, likely due to the regular reduction of redundancies, the need to represent the participle’s determiner disappeared with it. Along the same lines, the omission of the second occurrence of Adam at 1 Cor 15:45 need not require an omission in the Vorlage. (Paul’s Adamic Christology here might also have eluded the translator’s understanding.) Hoskier and Royse are too confident in denoting this variant as a shared reading with P46.33 Likewise both scholars err in viewing the omission of δέ in 1 Cor 16:6 as significant.34 Far from signalling a distinctive agreement, the unusually stable reading of the Ge‘ez is caused by the removal of a redundancy, the second reference to Macedonia in the previous verse.35 The reframed wording made a conjunction awkward. “When passing through I will stay with you” flows more smoothly than “When passing through, and I will stay with you.” Both witnesses omit the conjunction, but for very different reasons. Direct Discourse When a translator encountered indirect speech, he frequently converted it into direct speech. This approach required some adjustment to syntax, but more importantly for evaluating Hoskier’s variants in the Corinthian epistles, it resulted in restricting the role of ὅτι to ὅτι recitativum.36 And, as Royse notes, in Ge‘ez, ὅτι recitativum is always left untranslated.37 As a result, the Ethiopic provides no unequivocal evidence for omissions of ὅτι at 1 Cor 15:12, 27, and 2 Cor 6:16.38 Hoskier’s treatment of 1 Cor 15:12 is particularly telling. The exact same phenomenon also occurs at the beginning of the verse, but it goes unmentioned by Hoskier (and Royse). There, too, a conversion to direct speech (and active voice) takes place. As expected with this switch, ὅτι recitativum disappears.

Hoskier, Appendix, 5; and Royse, “Scribal Habits,” 713. Hoskier, Addenda, 3; and Royse, “Scribal Habits,” 713. 35 In terms of stability, virtually all the older manuscripts omit ወ before ኀሊፍየ. 36 Royse points his readers to similar phenomena in Luke, Romans, and elsewhere in 1 Corinthians, but they do not exactly correspond. Three of the examples concern citations of scripture and one introduces direct speech (“Scribal Habits,” 704). 37 Royse, “Scribal Habits,” 704, 706. 38 Hoskier, Appendix, 10. Royse goes too far in stating that “eth does not support p46’s omission” (“Scribal Habits,” 704). Neither the presence nor the absence of ὅτι in the exemplar can be determined. 33 34

Hoskier and his (Per)Version of the Ethiopic

273

Divine Names In the earliest attested text of the Ge‘ez Bible, እግዚአብሔር renders both θεός and κύριος. Based on this common practice, Royse rightfully dismisses Hoskier’s claim that እግዚአብሔር must parallel θεός, the reading of P46 at 2 Cor 8:5.39 The use of እግዚአብሔር to translate θεός later in the verse (በከመ ፈቀደ እግዚአብሔር = διὰ θελήµατος θεοῦ) does not mitigate Royse’s indictment, for translators could distinguish between θεός and κύριος when desired, using እግዚአብሔር for κύριος and አምላክ for θεός. Therefore, the two uses of እግዚአብሔር could reflect any of the following combinations: κύριος/κύριος, θεός/θεός, κύριος/θεός, and θεός/κύριος. Paraphrase When translators encountered difficult vocabulary, complex syntax, or foreign concepts they sometimes resorted to paraphrase. As Hoskier scoured the Ge‘ez version for parallels to the Chester Beatty papyrus, he paid little attention to the complexity of the Greek or how the translator handled that complexity. Indeed his layout of the evidence conceals both the challenges the translator faced and the challenges of reconstituting the translated text. For 2 Cor 3:6 Hoskier curtly summarizes the unique agreement between the Ge‘ez and P46 as “ως pro ος.”40 But this and the preceding verse have been substantially abbreviated. Ἀλλ᾽ ἡ ἱκανότης ἡµῶν ἐκ τοῦ θεοῦ ὃς καὶ ἱκάνωσεν ἡµᾶς is condensed to እስመ እግዚአብሔር ረሰየነ, “because God appointed us.” The Greek text may have come under scrutiny because of the repetitious wording (ἱκανοί, ἱκανότης, ἱκάνωσεν), but it seems more likely that the awkward syntax served as the primary catalyst for creating a shorter, smoother reading. In any case, Royse is correct that እስመ, “because,” does not translate ὡς.41 Again at 1 Cor 14:18, Hoskier only notes the addition of ὑπέρ after εὐχαριστῶ τῷ θεῷ, but the translator clearly misunderstood the comparative genitive (πάντων ὑµῶν µᾶλλον) even though he recognized the individual lexemes. Familiar with the prayer formula, εὐχαριστῶ τῷ θεῷ, and seeing the often collocated genitive ὑµῶν, he replaced the unfathomable with the stereotypical.42 Whereas P46 can be translated “I thank God I speak in tongues more than all of you,” the Ethiopian translator produced a very different “I thank God for all of you when I speak in tongues.” Thus በእንተ, “concerning,” translates περί, not ὑπέρ, and the word was supplied by the translator, not found in the exemplar. Royse, “Scribal Habits,” 706–707. Hoskier, Appendix, 5. 41 Royse, “Scribal Habits,” 710. He may also be correct that it instead renders ἀλλ᾽ (v. 5), for it functions so in 1 Cor 1:27; 3:2; 4:4; and 12:24 as perhaps also in two other paraphrased passages, 1 Cor 10:29 and 33. 42 Compare 1:4, አአኵቶ ለእግዚአብሔር በእንቲአክሙ = εὐχαριστῶ τῷ θεῷ… περὶ ὑµῶν. 39 40

274

Curt Niccum

Consequently, both Hoskier and Royse incorrectly list this text as an agreement with P46.43 Finally, at 2 Cor 10:7 Hoskier claims that ክርስቶስ, “Christ,” instead of reflecting the more common genitive Χριστοῦ, must derive from the nominative form found in P46 (καθὼς αὐτὸς ὁ Χριστός).44 His focus on agreement in case, though, obfuscates the translator’s struggle to understand the larger context and the resulting paraphrase.45 One must despair of establishing the underlying Greek with such precision and regard the convergence in case alone as coincidental.

Issues in Translation and Transmission Certain characteristics shared by the translators reappear regularly in later transcribers. Especially problematic for Hoskier’s claims are the fluidity of conjunctions and the ubiquity of pronominal suffixes in Ge‘ez. Both in translation and transmission conjunctions appear and disappear without any warrant from the Greek and sometimes the context. Although pronominal suffixes are seldom removed, they are often added, similarly without recourse to text or context. Because the oldest known manuscript of Paul’s letters in Ge‘ez was copied at least 600 years after their translation, one cannot confidently assert much about the original Ge‘ez in these cases. At best one can speak in terms of possibilities. All probability is gone. Conjunctions One conceptual error impairing Hoskier’s analysis of the Ethiopic was the working assumption that the translator must have consistently rendered different Greek conjunctions. For example, he expected እስመ for every occurrence of γάρ, እንከ for οὖν, and ባሕቱ for ἀλλά. As a result he overlooked the wide semantic range and ubiquitous use of ወ, which the translator uses for these and most other Greek conjunctions. Consequently, Hoskier mistakenly assumes the Ge‘ez version omits οὖν at 1 Cor 14:11 when ወ substitutes for it, as it also does in 1 Cor 6:4; 8:4; 9:25; 10:31; 14:23; 15:11 and 16:11. In other places Hoskier fails to understand the role of the suffix ኒ. Thus there is no cause for positing an omission of καί at 2 Cor 4:3, for there is likely no better way to express εἰ δὲ καί than by ወእመኒ.46 Also, despite the added Royse, “Scribal Habits,” 713. Hoskier, Appendix, 5. Royse argues that the indeclinable ክርስቶስ, “Christ,” could be nominative or genitive (“Scribal Habits,” 710), but Hoskier is correct that grammatically it must be nominative in this verse. 45 Royse further remarks that በከመ ክርስቶስ, “just as Christ,” offers no specific equivalent for αὐτός in P46 (“Scribal Habits,” 710). 46 Hoskier, Appendix, 12. 43 44

Hoskier and his (Per)Version of the Ethiopic

275

complexity with እስመ (= ἐπειδή), the ኒ at the end of አይሁድኒ, “Jews also,” in 1 Cor 1:22 almost certainly attests the purportedly missing καί.47 Somewhat related to the above examples are Hoskier’s interpretations of ወይእዜኒ “and now” at 2 Cor 4:16 and 5:9. Here Hoskier illustrates the folly of his procedure for he contends that ወይእዜኒ equates to διὸ minus καί at 4:16 and διὸ plus καί at 5:9.48 Neither can be determined because ወይእዜኒ consistently translates both.49 There are four additional variants connected with conjunctions that are not so neatly reducible. Still, with each Hoskier goes beyond what the evidence can bear. The easiest to address is 2 Cor 7:13. The text is corrupt, so it is especially tenuous to assert that the exemplar lacked δέ.50 Slightly more challenging is the reading of 2 Cor 2:14. In Hoskier’s favour, all but one of the twenty-one manuscripts collated lack a conjunction in concert with P46.51 The unusually stable text notwithstanding, Royse has demonstrated that the Ge‘ez often ignores δέ in the phrase “thanks be to God,” so the formula may preclude translation of the conjunction.52 Among the evidence for the omission of οὖν at 2 Cor 5:20 Hoskier includes (aeth). He justifiably places aeth between parentheses, because the Ge‘ez lacks any of the usual words employed for translating οὖν, although it does have the conjunction እስመ, “because.”53 Although እስመ does not appear for οὖν elsewhere in the Corinthian letters, the translator, because of the context, may have chosen to translate it so. Alternately, the Vorlage may have had a different conjunction, one not attested among extant Greek manuscripts. Either way, the Ge‘ez version stands at odds with P46. Hoskier’s contention that ነቢያት, “prophets,” at 1 Cor 14:29 reflects an exemplar without δέ has some merit.54 The stability of the earliest text combined with phrasing that begs for a conjunction might allow one to cautiously consider the possibility of a shared omission. In light of the problems with conjunctions in Ge‘ez translation and transmission, however, the agreement should be dismissed or at most noted with the requisite caveats.

Hoskier, Addenda, 3. Hoskier lists aeth 1/2, but both editions read አይሁድኒ. Both passages are listed together, Appendix, 5. 49 See 1 Cor 12:3; 14:13; and 2 Cor 5:9. In other passages በዘ (2 Cor 4:13 bis) and ወባሕቱ (2 Cor 6:17; 12:7, 10) appear. On one occasion διό is misread as διά (2 Cor 1:20), and once it is omitted (2 Cor 2:8). In his critique of Hoskier, Royse shows that διό occurring by itself is also almost always rendered ወይእዜኒ in Acts. 50 Hoskier, Appendix, 12. 51 These include the twenty manuscripts collated by Abraha (1 and 2 Corinthians, 30– 50; Codex Raineri 127 is deficient here) and the Sinai codex. 52 Royse, “Scribal Habits,” 705. 53 Hoskier, Addenda, 5. The Sinai codex is deficient. 54 Hoskier, Appendix, 5 and 10–1. See also Royse, “Scribal Habits,” 713. 47 48

276

Curt Niccum

Κἀγώ in 1 Cor 16:10 presents a special case because it deals with the adverbial function of καί after ὡς. Whether because of a misunderstanding of this particular use of καί or through the limitations of Ge‘ez, the translator typically left καί unexpressed in this and similar expressions.55 One cannot conclude confidently, as Hoskier does, that the Greek exemplar only read ἐγώ.56 Pronominal Suffixes Although Ge‘ez, like Greek, has independent emphatic pronouns, translators often did not employ them, for the pronominal suffix sufficed. Furthermore, suffixes were added with impunity. Therefore, the Ge‘ez pronominal suffix cannot guarantee the presence of a pronoun in the Greek, let alone precisely identify its location in a sentence.57 Thus, at 1 Cor 16:7 ወኢይፈቅድ ይእዜ ኅሉፈ እርአይክሙ, “and I did not want now in passing to see you,” could translate the Ausgangstext (οὐ θέλω γὰρ ὑµᾶς ἄρτι ἐν παρόδῳ ἰδεῖν), the reading of P46, which transposes ἄρτι and ὑµᾶς, a reading with ὑµᾶς in some other location, or a reading that lacks ὑµᾶς altogether.58 At 2 Cor 12:6 P46 omits εἰς in the phrase µή τις εἰς ἐµὲ λογίσηται. The Ge‘ez has ኢይትሐዘቡኒ, “they should not regard me,” which Hoskier calculates as an agreement with the papyrus. Royse critiques Hoskier for not recognizing that the pronominal suffix ኒ sufficiently renders εἰς ἐµέ.59 It is particularly telling that Hoskier does not refer to the other two occurrences of the same prepositional suffix in the same verse: ርእዩኒ = βλέπει µε and ሰምዑኒ = ἀκούει ἐξ ἐµοῦ, the latter clearly demonstrating that the pronominal suffix can render a Greek prepositional phrase. Hoskier might be correct that the Ge‘ez and P46 agree in reading αὐτῆς at 1 Cor 11:14 (ወፍጥረታሂ, “and [indeed] her nature”). On the other hand there are grounds for reserving judgment: First, because of the frequent addition of pronominal suffixes, -ታ, “her,” may not signify an independent Greek pronoun. Second, the translator of Paul’s letters had difficulty with the intensive use of αὐτός. This can be seen from the discussion of 1 Cor 10:3–4 above, to some extent in the free ὡς καί, καθὼς καί, or οὕτως καί appear without a Ge‘ez equivalent for καί in 1 Cor 7:7; 11:12; 15:22; and 2 Cor 1:14. The phrase is omitted in 1 Cor 9:5 (parablepsis) and 2 Cor 8:11 (paraphrase). The exceptions are ወከመሁ at 1 Cor 11:25, which occurs at the beginning of a sentence where conjunctions are often added in translation, and ዘከመሰ at 1 Cor 13:12. 56 Hoskier, Appendix, 11. 57 There are exceptions to the rule. On rare occasions the translator’s mishandling of a text can only be understood as a product of one particular sequence of words in Greek. 58 Royse handles this passage well (“Scribal Habits,” 705), although some of the specifics need to be nuanced. For example, the reference to Hofmann’s discussion of the possessive pronoun in Revelation (Johannes-Apokalypse, 91–2) is unhelpful as ὑµᾶς is not a possessive pronoun and the translator of Revelation did not translate any of Paul’s letters. 59 Royse, “Scribal Habits,” 707. 55

Hoskier and his (Per)Version of the Ethiopic

277

paraphrase of 1 Cor 12:5, and certainly in the treatment of τὴν αὐτὴν ἀγάπην in Phil 2:2 (ወበውእቱ ተፋቅሮ, note the similar ዘውእቱ in 1 Cor 10:4).60 Therefore, if agreement between P46 and the Ethiopic version exists, it is likely accidental, the translator treating αὐτή as a possessive pronoun rather than actually spying αὐτῆς in his Vorlage. Miscellanea A couple passages are not easily categorized. Nevertheless, Hoskier’s mistreatment of the version remains constant. First, ብፅዕት, “blessed,” at 1 Cor 7:40, as Royse observes, is yet another example of Hoskier’s ignorance concerning “syntactic resources,” further declaring that “Ethiopic possesses neither comparative nor superlative forms of adjectives.”61 Such a claim goes too far, for the language does have the capacity to express them. Apparently translators felt no compunction to create corresponding comparative and superlative constructions. Second, P46 lacks εἰς τό before πάλιν in 2 Cor 13:2, and, as Hoskier perceives, there is also no literal Ge‘ez equivalent.62 Nevertheless ካዕበ encompasses the thought of the phrase even if not reproducing every word. The rarity of this construction should not be overlooked, for it may have stumped the translator, forcing him to conveniently ignore the prepositional phrase.63 Once more Hoskier holds the translator to unrealistic expectations.

Agreements with P46 It would be surprising if the Ge‘ez version shared no agreements with P46. Still, some of those identified by Hoskier may be accidental and therefore of no value for tracing a genetic relationship. First, the Ethiopic has no parallel for τὸ φθαρτὸν τοῦτο ἐνδύσηται ἀφθαρσίαν καί in 1 Cor 15:54, which could be due either to parablepsis in Greek or Ge‘ez.64 Second, it may also be that the Ge‘ez omits ἢ συγκρῖναι at 2 Cor 10:12, but the translation is so muddled that it is

60 Royse argues that the suffix -ሂ could translate the intensive pronoun here, but considering the challenges just noted, it is improbable (“Scribal Habits,” 699). The earliest manuscripts are divided between -ሂ and -ኒ. 61 Royse, “Scribal Habits,” 699, relying upon Hofmann’s “Limitations” in Metzger’s Early Versions, 243. 62 Hoskier, Appendix, 13. 63 A textual search for εἰς τό πάλιν in Thesaurus Linguae Graecae (http://stephanus.tlg.uci.edu) produced just twenty-nine other hits, thirteen of which are quotations of this passage. Only three of the twenty-nine predate the translation of the Bible into Ge‘ez (fourth century or earlier). 64 See Hoskier, Appendix, 11. If the Ge‘ez is indeed aligned with P46, the same reading is also found in ‫ *א‬088 0121a 0243 1175 1739* and other witnesses.

278

Curt Niccum

difficult to comprehend how one could tell.65 Assuming the words are omitted, parablepsis (ἐγκρῖναι^συγκρῖναι) or the removal of perceived redundancy are more plausible explanations for their absence than a mutual ancestor. Two other agreements that Hoskier detected should be postulated with a low level of confidence. Uncertainty arises at 2 Cor 12:14 because the translator did not understand the meaning of οὐ καταναρκήσω and, judging from the numerous Ge‘ez variants, subsequent scribes did not understand the hapax legomenon ኢ(ይ)ትፋጠነኒ, “and I will not be forced” (or its predecessor if the extant text is corrupt).66 Despite these uncertainties, the lemma would have changed had ὑµῶν been in the Vorlage (e.g., as in v. 13). Because all early witnesses, not just P46, omit ὑµῶν, an agreement would hardly be surprising. The Ge‘ez, with P46 and Ambrosiaster, perhaps omits τι at 1 Cor 8:2. The Ge‘ez reads ወእመቦ ዘይብል አእመርኩ ዓዲ ኢያእመረ ዘይደልዎ ያእምር, “and if there is one who says, ‘I know,’ he does not yet know that which he should know.” Royse objects, arguing that the translator’s rewording of the first part of the verse prevents confident retroversion in the second half.67 But the straightforward switch to direct discourse in the first half makes the omission of τι all the more glaring. Nevertheless, the chance remains that the translator opted to take τι with the second verb of knowing in which case it is translated by ዘ, “that which,” prefixed to ይደልዎ, “he should.” Therefore, agreement with P46 remains just a possibility. With a high level of confidence agreement between the Ge‘ez version and P46 can be assumed in the following passages, only one of which has the Ethiopic “practically alone” in support: – 1 Cor 7:39 om νόµῳ (P15[vid] P46 01* 02 03 05* 0278 6 33 81 1175 1241 1739 1881 lat samss Tert Cl Or Cyp).68 – 1 Cor 9:8 om ταῦτα1 (P46 1831). – 1 Cor 11:29 om ἀναξίως (P46 01* 02 03 04* 6 33 1739 pc co).69

Conclusion Hoskier identifies forty-two readings, mostly omissions, in 1 and 2 Corinthians shared by the Ethiopic version and P46. Upon closer examination, most of these evaporate. Six times Hoskier errs outright (1 Cor 10:3–4; 15:39, 47; 2 Cor 3:1; 5:19; 9:2). Twice his resources did not preserve the earliest attested text (2 Cor 65 Its complexity leads Royse to mistakenly associate ናጥብዕ, “we dare” (= τολµῶµεν), with ἐγκρῖναι (“Scribal Habits,” 701). 66 Hoskier thus correctly places aeth within parentheses (Addenda, 6). 67 Royse, “Scribal Habits,” 709. 68 Hoskier, Addenda, 3. 69 See Hoskier, Appendix, 10 and Royse, “Scribal Habits,” 713. Information on the manuscripts cited comes from Hoskier and/or the apparatus of NA28.

Hoskier and his (Per)Version of the Ethiopic

279

1:5; 10:10). In eleven instances Hoskier does not take into account translation technique (1 Cor 7:22; 13:13; 14:18, 36; 15:12, 27, 45; 2 Cor 3:5–6; 6:16; 8:5; 10:7). To these should be added the eleven variants involving conjunctions (1 Cor 1:22; 14:11, 29; 16:6, 10; 2 Cor 2:14; 4:3, 16; 5:9, 20; 7:13) and three variants involving pronominal suffixes (1 Cor 11:14; 16:7; 2 Cor 12:6). There are two additional passages where Hoskier incorrectly interprets the evidence (1 Cor 7:40; 13:2). Only seven possible agreements surfaced, and two of these could be accidental. Of Hoskier’s twenty-nine readings where the Ge‘ez agreed with P46 “practically alone,” only five have any merit, and of these, three are attested in at least all of the “great uncials.” Thus only two distinctive readings, 1 Cor 8:2 and 9:8, may exhibit a special link to the Chester Beatty papyrus. This is hardly noteworthy. An examination of the Ge‘ez version in the Teststellen of the Institute for New Testament Textual Research highlights the distance between these two witnesses.70 P46 and the Ge‘ez agree in 52% and 50% of the test passages in 1 and 2 Corinthians respectively. This differs significantly from 74% and 76% percent agreement with 04. Obviously other scholars have co-opted versional evidence for their own textual theories, but few have done so on such a grand scale and with such meticulous attention to detail.71 Hoskier, therefore, offers a case study for how not to mine the Ge‘ez version for evidence of early Greek readings. We should not abandon the project because of his failure. Instead, we should cautiously move forward. Indeed, despite the slow movement in Ethiopic studies in the decades after Hoskier wrote, great advancements in material, analysis, and technology have been made and more are on the horizon. Whereas Hoskier consulted only printed editions based on a limited number of codices, the number of known manuscripts has expanded exponentially. More importantly, digitized images of many documents can be accessed by anyone with an internet connection thanks especially to the Hill Monastic Manuscript Library,72 the Ethiopic Manuscript Imaging The Teststellen are taken from Text und Textwert der griechischen Handschriften des Neuen Testaments, II: Die paulinischen Briefen, vol. 2: Der 1. und der 2. Korintherbrief, ANTF 17 (Berlin: De Gruyter, 1991). For the methods employed in establishing agreements, see Niccum, “1 and 2 Corinthians.” 71 M.-É. Boismard and A. Lamouille rival Hofmann’s misguided labours in their studies of Acts in Le Texte occidental des Actes des Apôtres: Reconstitution et Réhabilitation, Synthèse 17, 2 vols. (Paris: Éditions Recherche sur les Civilisations, 1984), and Boismard, Le Texte occidental des Actes des Apôtres, rev. ed., Études Bibliques 40 (Paris: Gabalda, 2000); cf. Curt Niccum, “The Ethiopic Version and the ‘Western’ Text of Acts in Le Texte Occidental des Actes des Apôtres,” in Transmission and Reception: New Testament TextCritical and Exegetical Studies, ed. J. W. Childers and D. C. Parker (Piscataway: Gorgias, 2006), 69–88. 72 https://www.vhmml.org [all websites accessed 24 April 2018]. 70

280

Curt Niccum

Project,73 Mäzgäbä Se’elat,74 the Endangered Archives Programme,75 and projects funded by the EU and member countries such as Beta masāheft,76 EthioSPARE,77 and TraCES.78 Libraries, too, are slowly adapting to the changing sociology of scholarship made possible through the world wide web, with smaller libraries generally paving the way. Still, some of the more important repositories are beginning to digitize holdings and to make images available online, most notable among them being those of the Vatican Library,79 British Library,80 and Bibliothéque Nationale.81 Critical editions of the Ethiopic New Testament (Protestant canon) should be completed by 2030, including some needed updates of older works. These editions will facilitate reconstruction of the Greek Vorlage, provide data for more precisely identifying translation technique and idiosyncratic practices of individual translators, and aid in understanding the transmission history of religious texts in Ethiopia. With regard to analysis, online tools made available primarily through the Hiob Ludolf Centre will aid scholars in quantifying this information.82 Among the tools currently under development is an online, open source lexicon that should be fully functional by the time this essay is published.83 Biblical concordances are also in development at the Hiob Ludolf Center and at the Capuchin Franciscan Research and Retreat Center in Addis Ababa. The New Testament Virtual Manuscript Room (NTVMR) sponsored by the Institute of New Testament Textual Research could help bridge both presentation of data and analysis.84 Although originally envisioned as a workspace for studies of the Greek New Testament, its application to other religious literature and to other languages is being explored.85 Further, the textual history of the Ethiopian Bible seems to 73 Steve Delamarter has kindly made images available at George Fox University and the Hiob Ludolf Centre. 74 http://ethiopia.deeds.utoronto.ca/doCmd.jsp. 75 This programme is administered by the British Library (https://eap.bl.uk). 76 http://betamasaheft.eu. 77 https://www.aai.uni-hamburg.de/en/ethiostudies/research/ethiospare.html. 78 Both projects are housed at the Hiob Ludolf Centre for Ethiopian Studies, University of Hamburg (https://www.aai.uni-hamburg.de/en/ethiostudies.html). 79 https://opac.vatlib.it/iguana/www.main.cls?sUrl=homeMSS&language=eng. 80 https://www.bl.uk/collection-guides/christian-middle-east-collections. 81http://www.bnf.fr/en/collections_and_services/orient_eng/s.orient_manuscripts.html? first_Art=non. 82 https://www.aai.uni-hamburg.de/en/ethiostudies.html. 83 http://betamasaheft.eu/Dillmann. 84 http://ntvmr.uni-muenster.de. 85 For example, the Center for the Study of Ancient Religious Texts at Abilene Christian University in conjunction with the Münster Institute, St. Catherine’s Monastery at Mount Sinai, and the Museum of the Bible Scholars Initiative is using the NTVMR for work on a critical edition of the Greek and Syriac traditions of John Climacus’ Ladder of Divine Ascent.

Hoskier and his (Per)Version of the Ethiopic

281

be a prime candidate for the Coherence-Based Genealogical Method developed by Gerd Mink at the same Institute.86 Nevertheless, as the reception history of Hoskier’s work with the Ge‘ez version demonstrates, availability of resources and tools does not insure their use. Availability, likewise, does not prevent their abuse. Thus, even though the present state of research nearly precludes anyone from making the same mistakes that Hoskier made, it does not stop investigators from creating new ones. As scholarship moves forward and technology advances, certain caveats must remain in place. Perhaps such a warning co-opted from Hoskier’s predecessors may suffice: KNOWLEDGE OF TRANSLATORS MUST PRECEDE JUDGMENTS ON TRANSLATIONS!87

http://egora.uni-muenster.de/intf/projekte/gsm_aus_en.shtml. See also T. Wasserman and P. Gurry, A New Approach to Textual Criticism: An Introduction to the CoherenceBased Genealogical Method (Atlanta: SBL, 2017). 87 B. F. Westcott and F. J. A. Hort, The Bible in the Original Greek, 2 vols. (New York: MacMillan, 1881), 2.31 (capitalization theirs). 86

Ostraca and Talismans: The Story of Two Former Text-Critical Categories and What to do with Them Today Thomas J. Kraus On 8 October 2009 Tommy Wasserman began a blogpost on the Evangelical Textual Criticism Blog as follows:1 “Hooray! It is here! The digital Kurzgefasste Liste is on-line here in the Virtual Manuscript Room of Münster. The VMR has been up for a while, but mainly for test purposes. Today the ‘Handschriftenliste 1.0’ was released.” Without doubt the New Testament Virtual Manuscript Room (NTVMR)2 benefits every textual critic. One of its central features is that it is based on the Kurzgefasste Liste,3 whose list of majuscules is of central significance for the purpose and aim of this paper. Some entries in the NTVMR do not bear a date or any further pieces of information about their “manuscript details,” because they have been identified as part of other items already enumerated elsewhere in the Liste. Among these, however, we find two very interesting numbers that are not attributed to another majuscule manuscript in the catalogue: 0152 and 0153. For the first the NTVMR has “= [T1] Talisman” in the column “Content Overview”; for the second “[O1–20]” with “O” for ostracon. These general comments on both entries mark the starting point for further in-depth reflections about their history and importance. For 0152 the NTVMR provides additional details: “Additional Content of the ‘Kurzgefasste Liste’ 1994, 33: ‘von v. Dobschütz in der Fortsetzungsliste V, ZNW 32, 1933, 188, als Talisman (T1) registriert; die Liste der Talismane (fortgeführt bis T9, vgl. ZNW 32, 1933, 188) wurde nicht fortgesetzt.’” For 0153 we read “die Liste der Ostraka, geführt von O1–25 (vgl. ebda.) wurde nicht fortgesetzt.” In an blogpost entitled “Kurt Aland and Non-Continuous Manuscripts of the NT,” Brice C. Jones ponders why Aland eventually deleted the two categories

1 Tommy Wasserman, “Aland’s Kurzgefasste Liste 1.0 on-line!,” The Evangelical Textual Criticism Blog, https://evangelicaltextualcriticism.blogspot.de/2009/10 (accessed 1 April 2018). 2 http://ntvmr.uni-muenster.de/liste; “Manuscript Num.” for 0152 und 0153 [accessed 8 April 2018]. 3 Cf. Kurt Aland et al., eds., Kurzgefaßte Liste der griechischen Handschriften des neuen Testaments, ANTF 1, 2nd ed. (Berlin: De Gruyter, 1994); Kurzgefasste Liste (online edition [continuously updated] http://ntvmr.uni-muenster.de; “New entries to the ‘Liste’” and “Liste Supplement”; accessed 6 April 2018).

284

Thomas J. Kraus

“Talismans” and “Ostraca,” but he has not yet found a plausible explanation for the omission.4 What is left of these categories in the Liste are only the two numbers 0152 and 0153, accompanied by the aforementioned superficial remarks. In what follows, I do not intend to find a clear answer to Jones’ question, nor do I suggest to change the established, tried and tested, familiar classification of manuscripts of the New Testament for the sake of integrating non-continuous material, like talismans (and/or amulets), ostraca, and even epigraphical evidence. In David C. Parker’s words “0152 is a portal to a different category altogether,”5 and the same may be true for 0153. Consequently, the aim of this study is to utilise these two numbers as windows into a world of New Testament texts and their materiality beyond the items and details we find in the Kurzgefasste Liste. Furthermore, I attempt to reconstruct how these two Gregory-Aland numbers found their way into the Liste in the first place and how they became associated with the majuscules in particular, focusing specifically on how these artefacts necessitated text-critical categories of their own, and how these categories were then deleted from the Liste again, leaving only their number and short references about their identity. This analysis leads to some critical observations regarding the character and materiality of these categories. Whether or not these pieces of information are essential and useful for textual criticism, I cannot decide in this context. My major interest is not in text-critical questions per se, but in archaeological objects and their usefulness for further study, valuing them as witnesses to and representations of human beings, i.e., as artefacts. The value of this for reconstructing the text of the Greek New Testament is better left to the editors. But what I know is that such archaeological objects can tell us something that transcends certain readings and concerns for the correct form of a text. This type of study helps to understand manuscripts and the people behind them. A manuscript is much more than solely a written source whose text attracts the lion’s share of scholarly attention. Its material, the quality and performance of its scribe’s hand (including letter forms, line lengths, lines per page, margins), symbolic and iconographical elements, signs, pagination, and much more, are salient pieces of information and are pivotal features of what it has to offer. Its provenance, if known at all, other archaeological objects found together with it, and, of course, other texts it may preserve should be taken into consideration for describing, assessing and properly characterizing a manuscript, its use, purpose, and function. On the basis of such a wealth of hard facts the text that initially attracted attention can be evaluated in a more in-depth and reliable way. And this

4 Brice C. Jones, “Kurt Aland and Non-Continuous Manuscripts of the NT,” http://www.bricecjones.com/blog (entry dated 24 March 2014, accessed 31 August 2017). 5 David C. Parker, An Introduction to the New Testament Manuscripts and Their Texts (Cambridge: Cambridge University Press, 2008), 40.

Ostraca and Talismans

285

is what studies of manuscripts should be about, because they implicitly and indirectly hint at the people behind those artefacts.

1. “In and out again” – The Story of 0152 and 0153 as Entries and Categories In his Die griechischen Handschriften des Neuen Testaments, published in 1909, Caspar René Gregory reports about his efforts to establish a logical and useful classification of the New Testament manuscripts. He sent letters to respected textual critics seeking information about what they think of (1) contemporary systems of classifying manuscripts, such as the system proposed by von Soden, and (2) their thoughts on Gregory’s own suggestions of classifying manuscripts:6 Zu allererst will ich bemerken, dass ich selbst am liebsten alle hebräischen und griechischen Buchstaben bei den Grosschriften durch Zahlen ersetzt haben würde. Das habe ich in den ersten Briefen vorgeschlagen. Ich fand aber, dass die Mehrzahl der Kollegen das Aleph für den Codex Sinaiticus, und die griechischen Buchstaben beibehalten wollte…Doch stimmte die Mehrzahl eben für die Beibehaltung von ‫ א‬A–Z Γ–Ω. Eine Änderung in diesen Buchstaben hätte Verwirrung gestiftet.

According to his report most of the scholars who responded were rather fond of Gregory’s suggestion to categorize the manuscript evidence as follows: “Grosschriften” (uncials/majuscules), “Kleinschriften” (minuscules), and “Lesebücher” (lectionaries).7 His “Grosschriften” consist of two sub-groups: the parchment codices and the “Papyri.” For Gregory it seemed to be no problem to list 0152 as “eine Thonscherbe aus Megara” and 0153 as “20 griechische Ostraka” among the “Grosschriften.”8 In the third edition of his handy Einführung in das Griechische Neue Testament, Eberhard Nestle listed “Unziale” and “Papyri” as a first and “Minuskeln” as a second category, mentioning 0152 and 0153 among the uncials.9 Both Gregory and Nestle refer to the first editions of the ostraca. The situation changed completely with Ernst von Dobschütz, who received an order from the conference of New Testament scholars in Münster in 1923 to 6 Caspar René Gregory, Die griechischen Handschriften des Neuen Testaments (Leipzig: Hinrichs’sche, 1908), 13. 7 Gregory, Die griechischen Handschriften, 31–171. But see Gregory, Textkritik des Neuen Testaments, vol. 1 (Leipzig: Hinrichs’sche, 1900), in which he arranges the “Grosschriften,” “Kleinschriften,” and “Griechische Liturgische Bücher” according to the sequence of the writings in the New Testament (Gospels, Acts and Catholic Epistles, Pauline letters, and Revelation). 8 Gregory, Die griechischen Handschriften, 42–3. 9 Eberhard Nestle, Einführung in das griechische Neue Testament, 3 rd ed. (Göttingen: Vandenhoeck & Ruprecht, 1909), 88.

286

Thomas J. Kraus

bring the Gregory list of New Testament manuscripts up to date.10 He got a personal copy of the Einführung in die Griechischen Handschriften from Gregory’s wife, swiftly improved the “Liste,” and updated it with newly discovered and/or edited manuscripts, but changed the order of the categories to “Papyri,” “Majuskeln,” “Minuskeln,” and “Lesebücher” so that each of these four categories formed separate groups on the same hierarchical structural level. For 0153, he found Gregory’s handwritten note “Cairo, Inst. fr. …von drei verschiedenen Händen.”11 Even more important for the present context is that von Dobschütz was in charge of the fourth edition of Nestle’s Einführung in das Griechische Neue Testament published in 1924, in which he offers more information about 0152 and 0153. For the latter he repeats the details provided by Gregory12 and justifies why he did not count the ostraca among the talismans, a new category he established and marked by a capital Gothic “T,” “da sie mit ihren fortlaufenden Texten offenbar Lese-, nicht Zauberzwecken dienten.”13 Thus, the purpose and/or use of a manuscript was introduced as the decisive feature for its categorisation. As far as 0152 is concerned, von Dobschütz opens the new category “Talismans” consisting of two entries: T1, which he identifies with 0152, the ostracon from Megara, and T2 (which lacks a Gregory number), which he calls a wood tablet form Oxyrhynchus with Matthew 4:23, 24,14 but which turns out to have been written on parchment.15 Obviously, von Dobschütz recognized the problems that may arise from the integration of magical and non-continuous texts into the category of majuscules and uncials, necessitating the creation of a separate category, the talismans. In 1926 von Dobschütz published his second update “Zur Liste der Neutestamentlichen Handschriften” with two additional talismans, T3 and T4, which are two papyrus fragments. What is more illuminating is what he writes about the category itself: the group “T = Talisman” was introduced to standardize talismans written on different writing material.16 Two years later von Dobschütz enlarged the talisman category with a wood tablet (T5) and a papyrus amulet (T6)17 and, in the fourth update of the “Liste,” he not only extended the list of talismans with a Ernst von Dobschütz, “Zur Liste der Neutestamentlichen Handschriften [I],” ZNW 23 (1924): 248–65, here 248. 11 von Dobschütz, “Zur Liste [I],” 252. 12 Ernst von Dobschütz, Eberhard Nestle’s Einführung in das Griechische Neue Testament, 4 th ed. (Göttingen: Vandenhoeck & Ruprecht, 1923), 97. 13 von Dobschütz, Nestle’s Einführung, 86. 14 von Dobschütz, Nestle’s Einführung, 86. 15 Cf. the editio princeps by Arthur S. Hunt, P.Oxy. VIII 1077, 10–11 (TM 61805; LDAB 2959). 16 Cf. Ernst von Dobschütz, “Zur Liste der Neutestamentlichen Handschriften II,” ZNW 25 (1926): 299–306, here 300. 17 Ernst von Dobschütz, “Zur Liste der NTlichen Handschriften III,” ZNW 27 (1928): 216–22, here 218–19. 10

Ostraca and Talismans

287

long parchment scroll from the twelfth or thirteenth century (T7) and two papyrus amulets (T8 and T9), but added further items to the list of ostraca. Gregory number 0153 for O1–20 was no longer alone, but partnered with five additional entries (O1–25).18 Georg Maldfeld updated and corrected the list of Greek papyri in 1939 but could not publish it quickly, because he was sent to war. After 1945 it took some time to re-establish the Zeitschrift für die neutestamentliche Wissenschaft und die Kunde der älteren Kirche (ZNW); thus, Maldfeld’s work on the Greek manuscript fragments of New Testament papyri only appeared in print in 1949. Maldfeld’s main concern was to provide a meliorated version of the list Gregory had arranged in 1908,19 leading him to discuss relevant publications, new discoveries and editions, and to include an overview of the main lists published so far. In this publication and in his “Berichtigungen und Ergänzungen,” Maldfeld solely deals with papyrus fragments, as he calls them, because the quantity of new literature that had to be investigated and taken into account was too onerous.20 The situation changed completely when Kurt Aland took over the task of updating, correcting, and systematising the Liste. In his articles in 1953 and 1954 (Liste V)21 and the follow-ups in 195722 and 1969 (Liste VI and VII),23 he did not include the two categories “Ostraka” and “Talismans,” and 0152 and 0153 received no mention at all. Aland extensively described and explained changes and additions to the Liste, but he took no account of the two groups introduced by von Dobschütz and did not waste any words on why he deleted them. Following this decision, it was a natural step not to integrate ostraca and talismans (i.e., amulets) in the printed version of the Kurzgefasste Liste24 and Der Text des Neuen Testaments25 thereafter. As a result, in Aland and Aland’s introduction we 18 Ernst von Dobschütz, “Zur Liste der NTlichen Handschriften IV,” ZNW 32 (1933): 185–206. 19 Georg Maldfeld, “Die griechischen Handschriftenbruchstücke des Neuen Testaments auf Papyrus,” ZNW 42 (1949): 228–53. 20 Georg Maldfeld, “Berichtigungen und Ergänzungen zur Liste der griechischen Papyrusfragmente (P) des Neuen Testaments,” ZNW 43 (1950/51): 260–61. 21 Kurt Aland, “Zur Liste der griechischen neutestamentlichen Handschriften,” Theologische Literaturzeitung 8/9 (1953): 465–96, and, more extensively, Aland, “Zur Liste der griechischen Neutestamentlichen Handschriften V,” ZNW 45 (1954): 179–217. 22 Kurt Aland, “Zur Liste der Neutestamentlichen Handschriften VI,” ZNW 48 (1957): 141–91. 23 Kurt Aland, “Die griechischen Handschriften des Neuen Testaments, Ergänzungen zur «Kurzgefassten Liste» (Fortsetzungsliste VII),” in Materialien zur neutestamentlichen Handschriftenkunde 1, ed. K. Aland (Berlin: De Gruyter, 1969), 1–53. 24 Aland, Kurzgefasste Liste. 25 Barbara Aland and Kurt Aland, Der Text des Neuen Testaments. Eine Einführung in die wissenschaftlichen Ausgaben sowie in Theorie und Praxis der modernen Textkritik, 2nd ed. (Stuttgart: Deutsche Bibelgesellschaft, 1982).

288

Thomas J. Kraus

simply find something like “0152 = Talisman (aus der Liste zu streichen)” and “0153 = Ostrakon (aus der Liste zu streichen,”26 even though the latter consists of more than one object (O1–20). Due to this situtation, it is not surprising that introductions to New Testament textual criticism do not allow much space for these two Gregory-Aland numbers, nor do their authors consider ostraca and talismans (amulets) seriously as relevant manuscripts. Bruce M. Metzger, for instance, briefly gives a sort of an aside on the two former categories which, according to him, may “be described with only a few words.”27 The words and phrases he uses for the description do not leave any doubt about his opinion of the two categories, because he talks about the ostraca as “curious but unimportant source of our knowledge of the Greek text of the New Testament,” employs the phrase “superstitious use of talismans,” and ends the passage without any further reflections on the usefulness and value of the two categories. Léon Vaganay and Christian-Bernard Amphoux devote half a page to “the Ostraka and Talismans” in their classic text-critical introduction.28 Parker is rather neutral in his first mention of talismans and ostraca, i.e., 0152 and 0153, in his Introduction to New Testament Manuscripts and their Texts, simply reporting that von Dobschütz listed T1–9 and O1–25 as special cases.29 Later on, he evaluates the two categories together with “other documents not present in the Liste which may be textually valuable.”30 But why then did ostraca and talismans disappear from the Liste as independent categories? Although he could not find a reason provided by Aland for this situation, Jones identifies the starting point of deleting the categories as follows: As a general rule – a rule that is strictly enforced by the Institute in Münster – non-continuous text manuscripts are prohibited from being registered in the official Liste. It is not clear when this rule was actually formalized, but it was apparently established during the tenure of Kurt Aland.31

Despite Jones’ detective work, we do not know whether this rule pertaining to non-continuous manuscripts is decisive, especially when we remember that inconsistencies in the list of papyri and the status of non-continuous witnesses were openly addressed by the Alands themselves. The inconsistencies in the papyri are similar to the ones for 0152 and 0153 and the other ostraca and talismans. Moreover, there were the translations or versions about which von Dobschütz Aland and Aland, Der Text des Neuen Testaments, 131. Bruce M. Metzger, The Text of the New Testament. Its Transmission, Corruption, and Restoration, 3 rd ed. (New York: Oxford University Press, 1992), 33. 28 Léon Vaganay and Christian-Bernard Amphoux, An Introduction to New Testament Textual Criticism, trans. J. Heimerdinger, 2nd ed. (Cambridge: Cambridge University Press, 1991), 26. 29 Parker, Introduction, 26. 30 Parker, Introduction, 126. 31 Jones, “Kurt Aland.” 26 27

Ostraca and Talismans

289

primarily stated: “Diese liegen wohl außerhalb unserer Aufgabe. Gregory hat zur Vulgata erklärt, daß er nach Begründung der Bibelkommission durch Pius X auf die Fortführung der Liste verzichte.”32 Despite this perspective, von Dobschütz nonetheless lists and describes certain translations thereafter in his Liste I–IV.33 Aland, in his Liste V–VII,34 did not continue to catalogue the versions, which is consistent because the Liste is about “manuscripts” and not – as most of the translations are – an accumulation and/or a recension of manuscripts (consider the recensions relevant for the Göttingen edition of the Septuagint). Moreover, Kurt Aland did want to continue the work of von Dobschütz and the tradition he established, because he regarded his work as a consistent succession of von Dobschütz’s from exactly twenty years previous, who had himself continued Gregory’s Liste.35 But Aland’s aim was also to omit errata and inconsistencies and to establish a Liste of relevant, reliable, and practical information, for which he established the twenty well-known rubrics with dates, dimensions, content, and so on.36 Be that as it may, the main aim and purpose of this discussion is not to search for explanations, but to judge the results, to point out what sort of information may have gotten lost, and tentatively to suggest what these lost pieces of information would be useful for. In all the shuffle of manuscript lists in the twentieth century, Ostraca and Talismans were lost as a category, even though the difficulties with these categories and with categorizations generally were also experienced by the papyri and versions.

2. Category “O” (Ostraca) Ostraca are, according to Parker, items that should not be ignored and must not be judged at a first glance or just by taking over what others – sometimes rather ignorantly – claimed them to be. Let us take category “O,” the group of ostraca, and have a brief look at them. Potsherds, far from being carriers of short notes only, could have had longer and complete passages of texts, too, as Parker states for “a school exercise extending to 250 lines.”37 In addition, the use of an von Dobschütz, “Zur Liste [I],” 264. von Dobschütz, “Zur Liste [I],” 264; “Zur Liste II,” 306; “Zur Liste III,” 220–22; “Zur Liste IV,” 203–205. 34 Aland, “Zur Liste,” 465–96; “Zur Liste V,” 179–217; “Zur Liste VI,” 141–91. 35 Aland, “Zur Liste,” 466: “Zum letzten Mal ist 1933 aus der Feder von E. von Dobschütz eine Fortsetzung zu Gregorys Liste der griechischen Handschriften des Neuen Testaments erschienen…Wenn jetzt, 20 Jahre nach der letzten Publikation, wieder eine Fortsetzung erscheint…” 36 Aland, “Zur Liste,” 466; “Zur Liste V,” 181–82; “Zur Liste VI,” 155–56. 37 David C. Parker, Textual Scholarship and the Making of the New Testament (Oxford: Oxford University Press, 2012), 39–40. 32 33

290

Thomas J. Kraus

ostracon does not mean that the text of the object is of a lower status. What does “lower status” mean anyway?38 The potsherds O1–20, belonging together and probably written by three different hands, were numerated by Gregory as 0153 and put into a separate category by von Dobschütz. They present an interesting collection of diverse texts from the four canonical gospels.39 The ostraca are dated to the fifth or sixth century and were found somewhere in Upper Egypt in the late nineteenth century.40 One of them offers a rather extensive passage from Luke (22:40–71, and also 12:13–16 preceding it), but it lacks 22:43–44 as a number of other manuscripts also do.41 Is it then to be regarded as a non-continuous text? Isn’t its text somewhat longer than many papyri? If these ostraca had been written on papyrus, they would have probably become well-known and respected, and would not have been regarded as obscure and curious objects. Or in Peter Head’s words, who argued in favour of a more insightful treatment of ostraca, amulets, inscriptions, and other sources as relevant textual material, “there seems no reason why this collection of texts should not be regarded as a citable witness to the text of Luke at the relevant points.”42 This may or may not apply to the same degree to the other ostraca in the category made up by von Dobschütz. O21–25 represent interesting witnesses to certain verses or brief passages from the New Testament.43 O21 (= O.Crum 515), from the fourth–sixth century, is an ostracon with Luke 1:42 and 1:28 and a hymn to the Virgin on its recto. It was found in Deir elBahari, Djeme in Thebes west, and is kept in the British Museum in London.44

Parker, Textual Scholarship, 39. TM 61837; LDAB 2991. G. Lefebvre, “Fragments grecs des Évangiles sur Ostraka,” BIFAO 4 (1904): 1–15. According to Gregory’s notes, the ostraca were written by three different hands (cf. von Dobschütz, “Zur Liste [I],” 252). According to von Dobschütz the texts come from Matthew 27, Mark 5, 9, and 15, Luke 12, 22, and John 1, 18, and 19. Brice C. Jones, New Testament Texts on Greek Amulets from Late Antiquity (London: T&T Clark, 2016), 198, offers the exact verses: Matt 27:31–32; Mark 5:40–41; 9:17–22; 15:21; Luke 22:13–16, 40–71; John 1:1–9, 14–17; 18:19–25; 19:15–17. 40 Parker, Textual Scholarship, 39, 155; Cornelia Eva Römer, “Ostraka mit christlichen Texten aus der Sammlung Flinders Petrie,” ZPE 145 (2003): 183–201, here 186. Further see Peter M. Head, “Additional Greek Witnesses to the New Testament (Ostraca, Amulets, Inscriptions, and Other Sources),” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2012), 429–60, here 434–35 (and n.12–21). 41 Parker, Textual Scholarship, 39, 155. 42 Head, “Additional Greek Witnesses,” 435. 43 For relevant bibliographical data cf. Jones, New Testament Texts, 198–200. 44 TM 61710; LDAB 2862. 38 39

Ostraca and Talismans

291

O22 (= Wadi Sarga 5) is a potsherd from Wadi Sarga from the sixth or seventh century with John 2:1 on one side. It may be a school exercise, but this is just a suggestion.45 O23 (= O.Petr.Mus. 13, 15, and 16 [former O.Petrie 414]) consists of three ostraca and a small piece (O.Petr.Mus. 14) belonging together from the fifth century. They were found in Tentyris (Dendera) or, probably, in the surroundings of Thebes. The potsherds are opisthographic with texts from Acts, Romans, Galatians, James, 1 John, John, and liturgical texts. For the present purpose the fragments with 1 John 2:12–14 and 19–22, the small one with 2:23 and possibly some more words, 3:17–22 and 4:1–3, and 4:10–14 and 18–21 are of major relevance, because they offer a considerable amount of text from 1 John,46 which – as is generally known – is not that widely attested in this early period.47 These ostraca might become even more interesting when considered together with two other potsherds from the Petrie Museum in London that originally belonged together. The first is 8.5 cm high and 7.1 cm wide, the second measures 10.5 and 6.1 cm. They have writing on both sides, are from the fifth–sixth century, and come from Tentyris (Dendera). They preserve 1 John 2:13–16 and 18– 22, overlapping with O23 somewhat. Again the quality of the scribe’s hand is of average quality, it is rather fluent and not inexperienced, and there is some Coptic visible.48 O24 (= O.Camb. 129) dated to the fourth–sixth century, found somewhere in Egypt and now kept in Cambridge, has Rom 8:31 on its recto.49 There may be a connection to O.Camb. 29 from 20–21 CE with a receipt.50 O25 (= O.Crum 514), fifth–seventh century, from Egypt and now in the British Museum has Ps 117:27, 26 and Luke 1:28. O21 also probably preserves the same

45 TM 61675; LDAB 2826. Scott Bucking, “Christian Educational Texts from Egypt: A Preliminary Inventory,” in Akten des 21. Internationalen Papyrologenkongresses, ed. B. Krämer, W. Luppe, and H. Maehler (Leipzig: Teubner, 1997), 132–38, here 137. 46 TM 61646; LDAB 2796. Römer, “Ostraka,” 197–200 (nos. 23–25); Maria Serena Funghi, Gabriela Messeri, and Cornelia Eva Römer, eds., Ostraca greci e bilingui del Petrie Museum of Egyptian Archaeology (O.Petr.Mus.), vol. 1 (Florence: Edizioni Gonnelli, 2012), nos. 13–16. Images can be found in Römer, “Ostraka,” pl. 3, and at The Petrie Museum of Egyptian Archaeology (https://www.ucl.ac.uk/culture/petrie-museum: “Go to the Petrie Catalogue,” search for UC31897, UC62566, and UC62584; accessed 1 August 2017). 47 Head, “Additional Greek Witnesses,” 435–37; Jones, New Testament Texts, 199. 48 TM (= LDAB) 113378. Cornelia E. Römer, “Das zweisprachige Archiv aus der Sammlung Flinders Petrie,” ZPE 164 (2008): 53–62, here 58–9 (no. 29). Images can be found at The Petrie Museum of Egyptian Archaeology (https://www.ucl.ac.uk/culture/petrie-museum: “Go to the Petrie Catalogue,” search for UC62644 and UC62683; accessed 1 August 2017). 49 TM 61871; LDAB 3028. 50 TM 91867; cf. papyri.info (http://papyri.info/ddbdp/o.camb;;29; accessed 28 April 2018).

292

Thomas J. Kraus

Lukan text. Both, O21 and O25 originated from Deir el-Bahari (Djeme in Thebes west).51 If he were still alive and continued his category “O,” von Dobschütz would have to enlarge his list of ostraca extensively in relation to the material available in critical editions today, such as the ostraca O.Petr.Mus. with Acts, Romans, James, Jude, 1 John and the Lord’s Prayer among the biblical texts.52 Even if the ostraca are not integrated into a Liste in the future, they must be assessed as invaluable witnesses to the Christian textual tradition. What seems most important is that they must be recognised as valuable material resources and integrated more fully into academic discourse on the first centuries of Christianity. This engagement with ostraca must be undertaken without any prejudiced pre-assumptions that sometimes appear in text-critical discourse, like the idea they can be disregarded because they are perceived to represent the lower stratum of textual tradition and society, or because they only represent notes and non-continuous texts. Even if such determinations might be true for many ostraca,53 it is not automatically true for them all.

3. Category “T” (Talismans) The second and more diverse category is that of the talismans, established by von Dobschütz in Nestle’s Einführung in das Griechische Neue Testament in 1923 and then continued until his fourth Liste in 1933. Gregory only listed one talisman in his list of majuscules, 0152, which today is the only remnant in the Kurzgefasste Liste and in the NTVMR. Category “T” consisted only of nine objects that von Dobschütz put together because (a) in contrast to the ostraca 0153 they did not represent continuous texts, and (b) they were used as amulets, although (c) they were made of different material, including papyrus, parchment, wood, and pottery.54 All in all, the nine talismans T1–9 can be clustered in various ways according to diverse criteria. (1) They could be arranged according to their publication chronology following von Dobschütz’s model. (2) They could also be arranged by material used, since von Dobschütz’s talismans included an ostracon (T1), a wood tablet (T5), two parchment fragments (T2 and T7), and five papyri (T3 T4 T6 T8 and T9). But how far is the writing material significant? Why is T1 not

TM 62319; LDAB 3482. Ostraca greci e bilingui, nos. 1–20. Further see Römer, “Das zweisprachige Archiv.” 53 For instance, cf. Charikleia Armoni, James M. S. Cowey, and Dieter Hagedorn, eds., Die griechischen Ostraka der Heidelberger Papyrus-Sammlung, Veröffentlichungen aus der Heidelberger Papyrus-Sammlung, 11 (Heidelberg: Universitätsverlag Winter, 2005). 54 von Dobschütz, Nestle’s Einführung, 86; von Doschütz, “Zur Liste III,” 300. 51 52

Ostraca and Talismans

293

consequentially clustered together with O1–25 so that these form a group of ostraca, no matter what these have preserved as textual units? (3) The content and/or text the objects present may also be a reasonable criterion for clustering. In this case, T1 T3 T4 T5 T6 and T7 form a group of manuscripts with the Lord’s Prayer, or at least parts of it. In order to provide more insight into the use and application of the Lord’s Prayer, it may be of interest that it is preserved on an ostracon, a wood tablet, a parchment fragment, and papyri, and that there is one item with the Lord’s Prayer alone (T1), while the others have additional texts as well, e.g., a doxology (T3 T4 T5 and T6). But there are also other clusters possible. For example, those with Greek Psalm 90. T4 and T6 have this psalm together with the Lord’s Prayer, which we find on other papyri as well, while T8 has the psalm without the Lord’s Prayer. With such clusters we may be able to trace the tradition and use of certain texts or textual units in the first Christian centuries, probably also in relation to the forms of specific texts and variants. This is what I tried to do some years ago in an essay in which I grouped certain manuscripts with the Lord’s Prayer that have not become a part of the Liste.55 (4) Other criteria may involve page dimensions so that we can observe more about certain forms of books in a manner similar to Eric T. Turner’s definition of miniature codices, a category which includes a number of codices that are part of the Liste. But von Dobschütz and Maldfeld also made some interesting remarks about important palaeographical details and scholarly asides, including comments on miniature codices56 or Herman Hoskier’s work on Revelation.57 There might be additional essential palaeographical details, like abbreviations, features of punctuation, spelling and/or itacisms, or the scribe’s hand and writing style that could be a decisive trait for clustering manuscripts as well. Perhaps one 55 Thomas J. Kraus, “Manuscripts with the Lord’s Prayer – They are More than Simply Witnesses to that Text Itself,” in New Testament Manuscripts: Their Texts and Their World, TENTS 2, ed. T. J. Kraus and T. Nicklas (Leiden: Brill, 2006), 227–66. 56 von Dobschütz, “Zur Liste IV,” 187 n.6 about PSI X 1164 (a miniature codex with text from Jonah); Maldfeld, “Die griechischen Handschriftenbruchstücke,” 235, with critical and extremely useful reflections about the use and purpose of a miniature codex with Daniel 3 and Matthew 11 (Greek-Coptic), i.e., according to Maldfeld, “alt- und neutestamentliche Oden bzw. Hymnen.” Maldfeld regards as more likely that the codex (13 leaves) was not an amulet, but a copy of a Christian who had the financial background to own and use such a small miniature book. For miniature codices see Thomas J. Kraus, “Die Welt der Miniaturbücher in der Antike und Spätantike – Prolegomena und eine erste methodische Annäherung für eine Datensammlung,” SNTU 35 (2010): 79–110; Kraus, “Miniature codices in late antiquity – preliminary remarks and tendencies about a specific book format,” Early Christianity 7 (2016): 134–52. 57 Cf. von Dobschütz, “Zur Liste IV,” 197–99, with acknowledging notes about Herman Hoskier, Concerning the Text of the Apocalypse: Collation of All Existing Available Greek Documents with the Standard Text of Stephen’s Third Edition Together with the Testimony of Versions (London: Bernard Quaritch, 1929).

294

Thomas J. Kraus

of the reasons T was omitted as a category was because the material was too diverse to easily classify. For the purpose of this paper a few specific details will be considered more closely for some selected talismans from T1–9 in order to illustrate their potential critical value. T4 (= PSI VII 719) is a papyrus fragment of a single sheet from the sixth century, probably found at Oxyrhynchus, and today kept in the Biblioteca Medicea Laurenziana in Florence. The dimensions of 25 x 5.5 cm are rather awkward but point at a certain context of usage. We find textual elements as follows: cross; John 1:1; Matt 1:1; John 1:23; Mark 1:1; Luke 1:1; LXX-Ps 90:1; Matt 6:9 with doxology; crosses. The reverse side offers the rest of a protocol (top, bottom and right margin preserved) in a perpendicular Byzantine hand (similar to that of PSI I 65)58 that helps to date the papyrus to the sixth century with the protocol being the older part.59 The papyrus in its current state has one column of six very long lines (on average 69 letters along the fibres) and at least two vertical folds. The script is that of a non-calligraphic hand. After the initial cross the texts starts with Χριστὲ σῶτερ of which only χῶρ is visible (the rho looks like a staurogram; cf. T1 = O.Athens inv. 12227 = Pap.Graec.Mag. II O4). The text ends with ἀµήν and Χ(ριστός) and three crosses (no other nomina sacra are used; diaeresis over ι and υ and spiritus asper). Between Ps 90:1 and the Lord’s Prayer (Matt 6:9) twice καὶ τὰ ἑξῆς can be found. Apart from the interesting sequence of gospel incipits (without John), which is interesting for reflections on the canon and the ordering of gospels, as Joseph E. Sanzo has rightly pointed out in his extensive research into gospel incipits,60 the amulet witnesses the reuse of a papyrus manuscript for a new purpose, i.e., after its first use, it was prepared to be carried along, and is a proof of some of the typical elements of amulets: gospel incipits, Psalm 90-LXX, the Lord’s Prayer (with a doxology), crosses, and ἀµήν. The features of PSI VII 719 stand within a broader tradition. Moreover, although this papyrus is not in the Liste itself and with the deletion of the talisman category vanished completely from the text-critical listings, it is registered by Aland in his Reptertorium I, as Varia 31. Likewise Old Testament textual criticism does not have any problem with listing such witnesses in the Rahlfs catalogue of manuscripts: PSI VII 719 has a

Cf. R. Pintaudi, “Una nota a PSI I 65,” ZPE 56 (1984): 137–38. TM 61617; LDAB 2767. Kraus, “Manuscripts,” 245–46 (no. 12) and 244 (figs. 4 and 5); Theodore S. de Bruyn and Jitse H. F. Dijkstra, “Greek Amulets and Formularies from Egypt Containing Christian Elements,” BASP 48 (2011): 163–216, no. 38; Jones, New Testament Texts, no. 4. 60 Joseph E. Sanzo, Scriptural Incipits on Amulets from Late Antique Egypt, STAC 84 (Tübingen: Mohr Siebeck, 2014). Further see my review Thomas J. Kraus, “Sanzo, Scriptural Incipits,” TC: A Journal of Biblical Textual Criticism 20 (2015) http://rosetta.reltech.org/TC/v20/TC-2015-Rev-Sanzo-Kraus.pdf [accessed 28 April 2018]. 58 59

Ostraca and Talismans

295

Rahlfs number just as T8 (= P.Vindob.G 2312 = Stud.Pal. XX 294)61 is Rahlfs 2031, which is also listed by Aland in his Repertorium I (Varia 13). T8 is a folded amulet for a newly wed from the fourth–sixth century: there are seven asterisks (just like a seven-beam sun or seven sun-like stars), then the three biblical texts LXX-Ps 90:1–2, Rom 12:1–2, and John 2:1–2, followed by some Coptic letters and nomina barbara (Adonai; Sabaoth). The other side is blank; the folded packet measured about 2.5 x 2 cm. T6 (= P.Iand. I 6),62 another witness to the Lord’s Prayer and LXX-Psalm 90, from the fifth or sixth century and acquired in Hermopolis Magna, is a papyrus fragment that starts with “Gospel of Matthew,” followed by the Lord’s Prayer (Matt 6:9–13), and a verse that might be Luke 9:37. Then there is Luke 11:1b–2 followed by a doxology, an exorcism of Solomon against demons, all kinds of diseases, and wild animals (LXX-Ps 90:13, allusions to verses 5–6), and a protective incantation. The text starts with a cross and ends with φο (for ϙθ´, i.e., 99 representing ἀµήν as an isopsephistic for well-being and luck).63 Without doubt the papyrus served the purpose of an amulet: (a) the textual disorder is extraordinary and demands further reflection; (b) the papyrus (30 x 15.5 cm with 18 legible lines on the recto, verso blank) was folded seven times lengthwise and five times widthwise; (c) the sizeable letters are formed by an untrained and inexperienced hand; (d) the orthography is very erratic. How then can such a textual source be taken as a trustworthy witness to any text or fragment it preserves? T1 (= O.Athens inv. 12227 = Pap.Graec.Mag. II O4)64 is the fragment of a reddish brown clay tablet from the fourth century found in Megara, Greece, and now kept in the National Archaeological Museum in Athens (inv. 12227). It measures 12 x 13.5 cm, originally 18.5 x 22.5 cm, and is broken on the top and left. Interestingly, the clay was inscribed when soft and then fired to fix it. Adolf Deissmann, who claims to have possessed a plaster cast, reports that he had seen 61 TM 62325; LDAB 3488. De Bruyn and Dijkstra, “Greek Amulets,” 192–93 (no. 59); Jones, New Testament Texts, 161–67 (no. 21). 62 TM 64868; LDAB 6107. De Bruyn and Dijkstra, “Greek Amulets,” 188–89 (no. 36); Jones, New Testament Texts, 87–94 (no. 6); Peter Arzt-Grabner and Kristin De Troyer, “Ancient Jewish and Christian Amulets and How Magical They Are,” Biblische Notizen 176 (2018): 5–46, here 27–8. 63 Cf. Pap.Graec.Mag. II 17 or P.Giss.Lit. 5.4 for transcriptions. 64 TM 64372; LDAB 5594. Rudolf Knopf, “Eine Thonscherbe mit dem Texte des Vaterunser,” MDAIA 25 (1900): 313–24; Knopf, “Eine Thonscherbe mit dem Texte des Vaterunser,” ZNW 2 (1901): 228–33; Eberhard Nestle, “Zum Vaterunser von Megara,” ZNW 2 (1901): 347–49; Ulrich Wilcken, “Bibliographische Notizen und Mitteilungen,” APF 2 (1903): 161–80, here 166; Carl Schmidt, “Christliche Texte,” APF 2 (1903): 381–85, here 383 (no. 8); Otto Kern, Inscriptiones Graecae (Bonn: Marcus & Weber, 1913), pl. 50; Adolf Deissmann, Licht vom Osten. Das Neue Testament und die neuentdeckten Texte der hellenistisch-römischen Welt, 4 th ed. (Tübingen: J. C. B. Mohr, 1923), 43 n.2; Margherita Guarducci, Epigrafia Greca IV: Epigrafi sacre pagane e cristiane (Rome: Arbor sapientiae, 1978), 336–38; Kraus, “Manuscripts,” 237–38.

296

Thomas J. Kraus

it himself in Athens. Such a production process can indicate that the ostracon was specially produced to serve the purpose of protecting a house, its inhabitants and visitors.65 The text left on the fragmentary potsherd starts with ἐ]πιούσιον and ends with πονηροῦ (Matt 6:11–13). In the middle of the last line there is the vocative κυριέ (cf. P.Bad. IV 60; BGU III 954; P.Duk.inv. 778), followed by a form of the crux monogrammatica better known as the staurogram consisting of a ligature of the Greek letters tau and rho. However, it does not function here as a part of στραυός or σταυρόω (or its derivatives), but its apparent use parallels instead the one on sarcophagi, where it is an alternative form of the chi-rho ligature. The letters are upright and single without any ligatures or juxtapositions, the production process justifies that inscriptional letter style. What we have here is extraordinary in three ways. (1) At first glance, the ostracon should be put into von Dobschütz’s category “O”, because of its material nature. (2) The style and way of writing, however, justify the qualification of this ostracon as an epigraphic object. (3) The special purpose and use of this ostracon separates it from the other ostraca in von Dobschütz’s category “O”. T1 might have been placed in a location where it could be seen by others, and so it would have become effective communication. Consequently, this tablet or part of an ostracon should receive special attention and its interesting features must be pointed out more explicitly in the future. Whether or not its reading αφιοµεν in Matt 6:12 in agreement with D W and Θ (cf. αφηκαµεν) is really of text-critical value may be a matter of discussion.66 Does this then really support that it “does not have οτι σου εστιν, etc.” at the end and that it does “not support the doxology,” as James Snapp, Jr., claims in his blog?67 It might be doubted that it could be expected that this ostracon helps out in text-critical issues like these, because the inscribed text stops with πονηροῦ at the end of line 14. The line is completely filled, and the vocative κυριέ in the middle of the following line 15 indicates that rather clearly. The text was meant as it stands today. T2 (= P.Oxy. VIII 1077)68 and its appearance with folds, drawings, and cutouts are well-known. Jones,69 Sanzo,70 and most recently Peter Arzt-Grabner and

65 According to Knopf and Schmidt. More recently, the plain attribution of “magical” for the use of O.Athens inv. 12227 was challenged by Raquel Martín Hernández and Sofía Torallas Tovar, “The Use of the Ostracon in Magical Practice in Late antique Egypt: Magical handbooks vs. Material Evidence,” SMSR 80 (2014): 780–800, here 794. 66 Cf. James Snapp, “The Resurrection of T,” http://wwwthetextofthegospels.com (accessed 4 August 2017). 67 Cf. Snapp, “The Resurrection of T.” 68 TM 61805; LDAB 2959. 69 de Bruyn and Dijkstra, “Greek Amulets,” 186–87 (no. 19). Jones, New Testament Texts, 60–5 (no. 1). 70 Joseph E. Sanzo, “Wrapped up in the Bible: The Multifaceted Ritual on a Late Antique Amulet (P. Oxy. VIII 1077),” JECS 24 (2016): 569–97.

Ostraca and Talismans

297

Kristin De Troyer71 have written about it in detail. It is 11.5 x 6 cm when unfolded, dated to the sixth or seventh century, and preserves Matt 4:23–2472 introduced by ἰαµατικὸν εὐαγγέλιον κατὰ Ματθαίον (“Curative Gospel according to Matthew”).73 The texts of the three outer horizontal areas on the left and the right are encircled by crosses. The middle area is filled by the drawing of a person, presumably a woman, as Jones suggests.74 Again we have a similar production scheme for an amulet as we have seen before, though the original and its layout differs from T4 and T8 as delineated above. T2 is folded as long vertically and horizontally – here we see two horizontal and four vertical folds, but it was folded more times thereafter – achieving a rather small square packet of papyrus that can easily be taken along, especially when kept in a container or capsule or hung by a cord or string.75 Matt 4:23 is also present in T3 (= BGU III 954 = W.Chr. 133 = Pap.Graec.Mag. II P9),76 another papyrus that was folded into a square packet

71

Arzt-Grabner and De Troyer, “Ancient Jewish and Christian Amulets,” 10, 16–8, and

39. For this and other papyri and parchments with Matt 4:23 (and/or Matt 9:35) cf. Theodore de Bruyn, “Appeals to Jesus and the One ‘Who Heals Every Illness and Every Infirmity’ (Matt 4:23, 9:35) in Amulets in Late Antiquity,” in The Reception and Interpretation of the Bible in Late Antiquity: Proceedings of the Montréal Colloquium in Honour of Charles Kannengiesser, 11–13 October 2006, ed. L. DiTommaso and L. Turcescu (Leiden: Brill, 2008), 65–82. The eight papyri and parchments are P.Oxy. VIII 1077; P.Berol.inv. 6096 (TM 64853; LDAB 6091); P.Oxy. VIII 1151 (TM 61652; LDAB 2802); P.Turner 49 (= Suppl. Mag. I 31; TM 64846; LDAB 6084); P.Coll. Youtie II 91 (TM 69042; LDAB 10333); P.Köln inv. 2283 (= Suppl. Mag. I 33; TM 64874; LDAB 6113); BGU III 954 (TM 64990; LDAB 6231); P.Köln VII 340 (TM 61663; LDAB 2813). However, Arzt-Grabner and De Troyer, “Ancient Jewish and Christian Amulets,” 17 n.60, point out that Matt 4:23–24 is actually “found on only two certain amulets: BKT 6.7.1…and Suppl.Mag. 1.31.” 73 Jones, New Testament Texts, 60–5 (no. 1), has a full transcription and English translation: “Curative gospel according to Matthew. And Jesus went around all of Galilee teaching and preaching the Gospel of the kingdom and healing every illness and every illness [sic] and every infirmity among the people. And a report about him went out into all of Syria and they brought to him all who were sick and Jesus healed them.” 74 Cf. Jones, New Testament Texts, 60–5 (no. 1). According to Jones, this may indicate a connection between the ownership of an amulet by a woman and John Chrysostom, Hom. Matt. 72 (“gospels hung from their necks”). Jones refers to additional amulets folded to packets owned by women. 75 Cf. Arzt-Grabner and De Troyer, “Ancient Jewish and Christian Amulets,” 39: “…probable amulet against diseases, or text about Jesus’ healing power, artificially designed in an educational context or as a gift for another person.” Further see Theodore de Bruyn, Making Amulets Christian: Artefacts, Scribes, and Contexts (Oxford: Oxford University Press, 2017), 149–51. 76 TM 64990; LDAB 6231. De Bruyn and Dijkstra, “Greek Amulets,” 188–89 (no. 26); Kraus, “Manuscripts,” 246–47 (no. 13); Jones, New Testament Texts, 107–12 (no. 10); ArztGrabner and De Troyer, “Ancient Jewish and Christian Amulets,” 23–6. 72

298

Thomas J. Kraus

(here 2 x 1 cm and bound together with a red thread).77 Unfortunately, this papyrus from the sixth century from Herakleapolis Magna broke into tiny pieces in the process of unwrapping. The column of text is written in a careful hand of the Oxyrhynchus type with rounded letters slightly sloping to the right and typical nomina sacra (again κε, as well). There is an invocation, a prayer against demons and diseases, Matt 4:23 in lines 11–12, Matt 6:9–13 in lines 15–24 followed by a short doxology, and an allusion to the Nicaean symbolum. The Lord’s Prayer is introduced by τὴν εὔαγγελικὴν εὐχήν with ευ written above the line and the vocative κ(υρι)έ inserted before “deliver us…” For T9 (= P.Oxy. VIII 1151),78 a papyrus from the fifth century, we have a similar situation, as it was also originally tightly folded. Jones79 and AnneMarie Luijendijk80 have dealt with it in detail. They, for instance, pointed at the use of theotokos in “intercession of our lady the theotokos” (cf. P.Köln VIII 340 = Jones no. 18), a statement valuable in respect to Church history. In addition, this amulet was owned by Joannia, daugher of Anastassia, who probably had “commissioned her amulet at Oxyrhynchus, invoking the locally worshipped saints to help her recover from illness.”81 The amulet might have been ordered from a specialist. Its first editor, Arthur S. Hunt, reports that the papyrus was found “tightly folded, and tied with a string.”82 Unfortunately, the string is now lost. The damages indicate that it has been rolled up starting to the top, as Michael Zellmann-Rohrer, who studied the papyrus in situ, reports. It is probable that the rolled and tied papyrus was put into a container and suspended from the body.83 Again we may think of John Chrysostom’s observation that women hung gospels from their necks. Whether or not the amulet was copied from an actual manuscript, as Jones 77 Cf. Arzt-Grabner and De Troyer, “Ancient Jewish and Christian Amulets,” 25: “The papyrus was found folded and wrapped (size ca. 2 x 1 cm) with a brown cord that was most probably red in antiquity (which would match the instructions by magical books, cf. Pap.Graec.Mag. 1.5,385–388 [TM 64368]).” 78 TM 61652; LDAB 2802. De Bruyn and Dijkstra, “Greek Amulets,” 186–87 (no. 21); Arzt-Grabner and De Troyer, “Ancient Jewish and Christian Amulets,” 14, 29–32; de Bruyn, Making Amulets Christian, 107–109. 79 Jones, New Testament Texts, 134–40 (no. 17). 80 AnneMarie Luijendijk, “A Gospel Amulet for Joannia (P.Oxy. VIII 1151): Rhetoric and Religion on the Ground,” in Daughters of Hecate: Women and Magic in the Ancient World, ed. K. B. Stratton and D. S. Kalleres (Oxford: Oxford University Press, 2014), 418– 43. 81 Luijendijk, “A Gospel Amulet,” 421. 82 This is taken by Arzt-Grabner and De Troyer, “Ancient Jewish and Christian Amulets,” 32, under the heading “Magical markers use” and they conclude: “The production of the artefact as original amulet is at least highly probable if not certain, the use of it as amulet is certain.” 83 I am indebted to Michael Zellmann-Rohrer for making his unpublished paper “A Christian Amulet in Context: Report on a Re-edition and Study of P.Oxy. VIII 1151” available to me.

Ostraca and Talismans

299

supposes, depends on the idea of “manuscript” we employ: do we talk about any artefact with text, like we have here, as a “manuscript,” or do we think only of a proper and authoritative full-text manuscript? What criteria aid in a decision on such an issue? Furthermore, there is an interesting text-critical issue for John 1:3–4, which Zellmann-Rohrer refers to and which has to do with punctuation that differs from the printed NA28.84 The variant καὶ χωρὶς αὐτοῦ ἐγένετο οὐδὲ ἓν ὃ γέγονεν. ἐν αὐτῷ ζωὴ ἦν, supported by our text, is well represented among the early uncial manuscripts, and found favour in the so-called Byzantine majority text. So far no papyri have been cited in support of it, and indeed a continuous papyrus manuscript of the gospel would not necessarily explicitly indicate such punctuation. Kurt Aland already noticed this phenomenon in P.Oxy. VIII 1151, but supposed that there might be an individual and separate textual tradition different from that of the New Testament at work in this instance.85 In Zellmann-Rohrer’s words “it seems doubtful however that separate transmission of the citation in an amuletic formula would distort it significantly, given that the users will all have known their New Testament intimately.” T5 (= P.Bad. IV 60)86 is a wooden tablet from the seventh century found in a tomb in Qarara in 1914 that measures 42 x 16 cm, is horizontally cracked in two pieces, and has two round holes for strings. Its two sides are whitened with moulding. The recto has twelve lines in the left half of the tablet with Matt 6:9– 13 and a doxology separated from the blank right half by a line drawn vertically. There are wiped-out letters still faintly visible so that this artefact may be called a palimpsest. The reverse side offers some names with epithets. The scribe’s hand forms upright, single, and very irregular letters. Thus, the tablet probably served an educational purpose, a school text written from dictation, requiring the correction or wiping-off of text, and is not an amulet. The text of the Lord’s Prayer is written in Coptic letters (cf. MPER N.S. XV 184) and ends with the vocative κ(υρι)έ added as a nomen sacrum with a supralinear stroke (cf. O.Athens inv. 12227; BGU III 954; P.Duk.inv. 778). Is it possible that the Coptic alphabet was used in order to preserve the text in its original wording. Was it meant to transmit a certain content to Egyptians who did not know any Greek?

Zellmann-Rohrer, “A Christian Amulet.” Kurt Aland, “Über die Bedeutung eines Punktes. Eine Untersuchung zu Joh. 1,3.4,” in Neutestamentliche Entwürfe, Theologische Bücherei 63 (Munich: Kaiser, 1979), 351–91, here 369–70. Aland observes that the papyrus “hinter ὃ γέγονεν ein Satzzeichen hat,” but he continues that the punctuation there “scheint doch etwas gewagt” (369). 86 TM 65415; LDAB 6662. Kraus, “Manuscripts,” 248–50 and figs. 7 and 8 (no. 16); de Bruyn and Dijkstra, “Greek Amulets,” 208–209 (no. 157). 84 85

300

Thomas J. Kraus

T7 (= Chicago MS 125 = Greg. T7 = [formerly] Goodspeed Ms. Grk. 16; + Pierpont Morgan Library, New York, Ms. M. 499 [second portion]),87 the socalled D’Hendecourt Scroll, from Constantinople or Trabzon, Turkey (Trebizond), was written in the fourteenth century.88 The parchment scroll is 174.4 cm long and only 9.2 cm wide.89 Consequently, de Bruyn and de Bruyn together with Dijkstra did not include it in their lists of amulets or potential amulets. On the obverse, the scroll has 114 lines with Mark 1:1–8, a miniature with the evangelist Mark; Luke 1:1–7, a miniature with the evangelist Luke; John 1:1–17, a miniature with Christ in Majesty; a miniature with the evangelist John; Matt 6:9–13, a miniature with The Trinity; the Nicene Creed, a miniature with the Virgin and Christ Child, Psalm 68, a miniature with David, king of Israel. It is possible that a miniature of Matthew the evangelist is missing at the beginning of the scroll. The reverse has about 157 lines and offers a sequence of prayers which includes the petition, “O our Lord, protect us from Iblis (Satan)…” (l.31). The scroll’s one-time owner Suleyman ibn Sara is named in an invocation, several lines of which refer to the biblical account of Moses leading the Children of Israel through the desert. Accordingly it reads “…the rock [of which] water was brought forth that they drank [the Israelites]…likewise O Lord bring forth your hand filled with mercy upon you servant Suleyman b. Sara” (2.34–36). A drawing of a filigree cross precedes the prayers. The Greek text was written in minuscule script in dark brown ink (some of it fading), the Arabic in cursive script in ash brown ink (some fading and flaking). Its content and layout, especially its diminutive size, suggests that this manuscript is a fragmentary talismanic scroll. Maybe we can think of 11Q11 from Qumran with three hymnic compositions followed by Hebrew Psalm 91 (= LXX-Psalm 90), the small apotropaic scroll, as a parallel. But this is another special issue.

4. Where Does this Leave us? “Teach and Make Known” What do we do now with all of this? Is it just much ado about nothing? Definitely not. The list of majuscules within the Liste has its inconsistencies as we have seen in relation to 0152 and 0153. And the list of papyri – basically majuscules also – has its own, as well. As Parker phrases it: “In several ways the list of majuscules in the Liste is at least as problematic as the list of papyri.”90 According to Aland and Aland themselves, the papyri include talismans (P50 P78), 87 The following is from Kraus, “Manuscripts,” 250–51 (no. 17), supplemented and corrected by details provided by The Goodspeed Manuscript Collection (http://goodspeed.lib.uchicago.edu/ms/index.php?doc=0125; accessed 4 August 2017). 88 von Dobschütz, “Zur Liste IV,” 188, has twelfth or thirteenth century. 89 Kraus, “Manuscripts,” 248–50. 90 Parker, Textual Scholarship, 38.

Ostraca and Talismans

301

lectionaries (P2 P3 P44), excerpts (P43 P62), a manuscript with Odes (P42), hermeneia-manuscripts (P55 P59 P60 P63 P80), a writing exercise (P10), and an occasional note of a verse (P12).91 There is no need to discuss this statement any further, because some of its qualifications have already been discussed, challenged, and modified since then. So, what do we do with such an inconsistent Liste in terms of the majuscules and papyri on one side and the categories “O” and “T” on the other? The treatment of von Dobschütz’s categories, and other additional material, has been an ongoing matter of discussion since at least the late 1990s. (a) In 1999 Stuart R. Pickering complained about the non-use of and noninterest in ostraca, tablets, amulets, and other objects in textual criticism.92 Noncontinuous texts are treated as second rate evidence, though they are indispensable textual witnesses for reconstructing New Testament texts. Pickering chooses T8 (P.Vindob.G 2312) and the papyrus letter P.Abinn. 19 to demonstrate the value of such non-continuous manuscripts, which he thinks should be listed in a separate online and printed database or catalogue. (b) Stanley E. Porter took up Pickering’s criticism and plea and challenged the definition of a “continuous” and a “non-continuous” text, especially when we think of manuscript fragments that contain just a few lines or verses.93 For him the papyri, i.e., their material, have gained too much attention, even if they often preserve far less textual units than many of the lectionaries. Material should not matter as much as it does, leading him to propose two lists. A first list “would be given to those documents for which there is little or no doubt regarding their being New Testament manuscripts,” and a second “would include those documents for which there is some doubt, such as the papyri noted above (including lectionaries), the Apocryphal Gospels, as well as some other manuscripts.” This marks a shift from material to content. Later on Porter adds sub-lists to the second list which indicate the type of text or texts they exhibit – liturgical and apocryphal texts, amulets, excerpts, etc.94 (c) In an essay published in 2012, Peter M. Head writes about “Additional Greek Witnesses,” which are, according to him, ostraca, amulets, inscriptions, Aland and Aland, Der Text, 95. Stuart R. Pickering, “The Significance of Non-Continuous New Testament Textual Materials in Papyri,” in Studies in the Early Text of the Gospels and Acts. The Papers of the First Birmingham Colloquium on the Textual Criticism of the New Testament, ed. D. G. K. Taylor (Birmingham: University of Birmingham Press, 1999), 121–41. See the survey by Jones, New Testament Texts, 13–5. 93 Stanley E. Porter, “Why So Many Holes in the Papyrological Evidence for the Greek New Testament?” in The Bible as Book: The Transmission of the Greek Text, ed. O. O’Sullivan and S. McKendrick (New Castle: Oak Knoll, 2003), 167–86. See Jones, New Testament Texts, 15–8. 94 Stanley E. Porter, “Textual Criticism in the Light of Diverse Textual Evidence for the Greek New Testament: An Expanded Proposal,” in New Testament Manuscripts: Their Text and Their World, TENTS 2, ed. T. J. Kraus and T. Nicklas (Leiden: Brill, 2006), 305–37. 91 92

302

Thomas J. Kraus

and other New Testament excerpts.95 He challenges the qualification of manuscripts as “non-continuous” by referring to the Rahlfs-Fraenkel list of manuscripts of the Septuagint, which lists almost everything that has some bits of Septuagint text. Head describes every category, discusses selected example cases for it, and surveys the most important literature about them. With this, he demonstrates a practical way of developing further categories – and this means that inscriptions would be included as well. The separate lists, the catalogue, and the collection he suggests should serve as the basis for further discussions.96 (d) In his doctoral thesis, published in 2016, Jones aims at “the first systematic treatment of non-continuous manuscripts of the Greek New Testament.”97 Of course, such an agenda is too ambitious, leading him to specialise on amulets with New Testament texts. He defines “amulets” by various criteria (verso-recto; use of amulets; ἑρµηνεία; context of citation; specific context of New Testament citation) and is aware of the fact that a definition of what actually constitutes an amulet remains a tricky task, even if specific criteria are applied. Still more complex is it to define the term “non-continuous” in respect of fragmentary manuscripts with only small portions of text. In these examples it is hardly possible to decide whether or not these originally had continuous or non-continuous texts. His overall aim is “to remedy the neglect of non-continuous witnesses.” All in all, he ends up continuing von Dobschütz’s category “T” and adds those amulets that he examined that were not part of von Dobschütz’s lists. His list includes twenty-seven talismans.98 95

Head, “Additional Greek Witnesses,” 429–60. See Jones, New Testament Texts, 18–

20. 96 Head, “Additional Greek Witnesses,” 454: (i) the continuation of von Dobschütz’s lists in a separate list; (ii) selective manuscripts that are “likely to be cited in a critical apparatus to the New Testament text” collected in a separate list; (iii) “all possible additional witnesses to the New Testament text” accumulated in a separate list, similar to Porter’s proposal; (iv) cataloguing important papyri and establishing a database of transcriptions as Pickering proposed; (v) “a collection of relevant material compiled on a book-by-book basis through the New Testament.” 97 Brice C. Jones, “New Testament Texts on Greek Amulets from Late Antiquity and Their Relevance for Textual Criticism” (PhD diss., Concordia University, 2015), iv, 6 and 192. In the published version of his dissertation Jones phrases it as follows (New Testament Texts, 6): “This book, which represents the first systematic treatment of non-continuous artifacts, has two main objectives. The first objective is to define more closely the categories of continuous and non-continuous by formulating criteria for the identification of the latter. The second objective is to establish a comprehensive database of one category of non-continuous artifacts (amulets) and to provide a detailed analysis of both their texts and containers (i.e. the materials that bear the text).” 98 Jones, New Testament Texts, 195–98. Interestingly, he keeps T1, the ostracon with epigraphical features, and T7, the miniature scroll from the fourteenth century he did not want to integrate into his monograph because it was out of the time frame of “late antiquity,” within his list of talismans.

Ostraca and Talismans

303

All these efforts to take “Additional Greek Witnesses” seriously make scholars more aware of these artefacts’ value and use, and establishing a form of listing, cataloguing, or databasing for these witnesses are very welcome. A number of years ago I would have liked to have suggested something like that on my own, but I have never managed to do so. But at the same time some of the suggested additional lists or sub-lists create new problems and inconsistencies. Although there have been useful attempts to define amulets, it is still not always clear if a fragment was used as an amulet or not, something we see in the three groups proposed by de Bruyn and Dijkstra (probable, possible, improbable).99 How can we be sure about the intended purpose of an artefact? Additionally, what about apocryphal texts and fragments? Qualifications and suggestions are sometimes very tentative, especially when we think of fragments of unidentified content. And if we want to integrate epigraphic evidence – and we certainly should – what exactly should we accept as an inscription? Lintels, walls, and sarcophagi come to mind. But what about medallions, armbands, gems, or tablets with letters inscribed into the material? Even if we developed new categories or rekindled or revived old ones such as “O” or “T,” would students and scholars really accept them as first-rate witnesses to the text of the New Testament, even though they are archaeological objects that are, in their eyes, “obscure,” “mystic,” or “magical,” artefacts that are interesting but, according to too many scholars, represent primitive, syncretistic, or heretical beliefs in late antiquity. With these items there is even more to fight against than solely trying to enhance the importance of and attention given “noncontinuous” and other witnesses to the New Testament not in the Liste, especially when it comes to amulets. Every system of categorisation has its flaws. The following is not a plea for accepting flaws, but is instead an appeal to keep the Liste as it is, a traditional and accepted reference system that conveys a sort of trustworthiness and custom. Moreover, it is essential that scholars and students alike become aware of inconsistencies and flaws, to research them in order to (a) help to improve the Liste itself and (b) to teach critically but responsibly about everything that gets lost when scholars and students only take the Liste into account. In accordance with Head, I think that “there is an ongoing need for up-to-date catalogues of the [noncontinuous] material.”100 We need more publications like Cornelia Römer’s on O.Petr.Mus., Jones’ on amulets,101 de Bruyn’s on amulets,102 in addition to his

Cf. Arzt-Grabner and De Troyer, “Ancient Jewish and Christian Amulets,” 5–46. Head, “Additional Greek Witnesses,” 453. 101 Jones, New Testament Texts. 102 de Bruyn, Making Amulets Christian. 99

100

304

Thomas J. Kraus

catalogues,103 and those publications produced with Jitse Dijkstra.104 Of course, inscriptional evidence has to be considered as well. And why don’t we establish databases like the Leuven Database of Ancient Books and Trismegistos for ostraca, talismans, inscriptions, and all the items that have not been integrated into the Liste? If we establish a database for inscriptional witnesses to the New Testament, all available pieces of information about an entry should be collected, tagged, and electronically published. This process would enable palaeographical data, including observations on scribal hands, letter forms, short forms (abbreviations), punctuation, orthography, layout (including line numbers per page, letters per line, i.e., line lengths, margins, and so on),105 to be accumulated so that the people behind those Christian textual witnesses are more seriously taken into account. Perhaps the NTVMR could be a model for such a database if ways were devised that integrated more palaeographical information. But before I speculate and even fantasise too much and fall into daydreams, we need to consider both who could undertake such a project and how it would be accomplished. Who has the time and a position to carry out such a strenuous task? Until then, our task is to research, discuss and publish, and, above all, to teach about those archaeological objects in detail because they are indispensable for studying the first centuries of Christianity and are of the same value and importance as the manuscripts compiled in the Liste.

103 Theodore de Bruyn, “Papyri, Parchments, Ostraca, and Tablets Written with Biblical Texts in Greek and Used as Amulets: A Preliminary List,” in Early Christian Manuscripts: Examples of Applied Method and Approach, TENTS 5, ed. T. J. Kraus and T. Nicklas (Leiden: Brill, 2010), 145–90. 104 de Bruyn and Dijkstra, “Greek Amulets,” 163–216. 105 Kraus, “Manuscripts,” 230.

The Critical Apparatus of Stephanus’ Greek New Testament of 1550: Early Printed Editions and Textual Scholarship An-Ting Yi* In the introduction to his A Full Account and Collation of the Greek Cursive Codex Evangelium 604 (1890), Herman Hoskier expresses confidence in his collation, even seeing it as a replacement for the manuscript itself. He claims: It [the collation on Codex Evangelium 604] presents the whole codex to those at a distance, and does away with that never-ending re-examination of documents, which has been going on – at such a cost of precious time – ever since the days of Henry Stephens.1

It is noteworthy that Hoskier refers to the name of Henri Stephanus (i.e. Henry Stephens),2 who collated a number of Greek New Testament manuscripts in the royal library of France in 1546. The results contributed to a new edition published in 1550 by Henri’s father Robert Stephanus, an edition that contained the first critical apparatus of the Greek New Testament.3 In other words, Henri can be regarded as the first person whose collation became a critical apparatus in a Greek New Testament edition. It is this labour to which Hoskier refers. Perhaps due to its pioneering nature, this innovation has been criticised by many scholars. For instance, Tregelles in 1854 commented that “the various readings in the margin of Stephens’s folio edition seem rather to be appended as

* This study is written in memory of Prof Dr Tjitze Baarda (1932–2017), who kindly read an earlier draft and provided invaluable feedback. I also thank Phil Tolstead for correcting the English text. Any remaining errors are of course my own. 1 H. C. Hoskier, A Full Account and Collation of the Greek Cursive Codex Evangelium 604 (with Two Facsimiles) [Egerton 2610 in the British Museum] Together with Ten Appendices (London: David Nutt, 1890), xxviii (emphasis original). 2 Note that textual critics and historians refer in different ways to Stephanus’ name. The former group tend to use the Latin name, either the English version “Stephens” (e.g. Tregelles and Scrivener) or simply “Stephanus” (e.g. Metzger and Parker), but we usually read the French surname “Estienne” in historical works. I follow the convention of recent New Testament textual scholarship: Robert and Henri Stephanus. 3 Robert Stephanus, Τῆς καινῆς διαθήκης ἅπαντα. Novum Iesu Christi D. N. Testamentum. Ex bibliotheca regia, 3rd ed. (Paris: Robertus Stephanus, 1550). As will be shown below, critical apparatuses of the Latin text of the New Testament already appeared several decades earlier in Robert’s Latin Bible projects.

306

An-Ting Yi

an ornament to the text, than as giving it any real and fundamental utility.”4 However, is it appropriate to judge a sixteenth-century edition based on the standards of later generations? The present study initially arose from this query, and the starting point for what follows is a commitment that the evaluation of an edition needs to take its historical context into account. Such a belief fits the growing interest in studying the history of New Testament textual scholarship as valuable to the larger discipline, often referred to as the “historical turn.”5 Within this mode of analysis, New Testament manuscripts are no longer taken merely as tools for pursuing the “original text,” but they themselves are considered as valuable historical phenomena.6 Following the same path, I argue that historical editions should not be regarded only as outdated materials; instead, many intriguing stories can be revealed once thorough exploration of the given editions has been made. We could even say, in line with Hort’s famous maxim, that knowledge of editions should precede judgement upon the information they provide.7 This close evaluation of the rhetoric and composition of early printed editions represents a fruitful avenue for further scholarly engagement. This study explores the critical apparatus in the Greek New Testament of Robert Stephanus from a historical perspective. The exploration first sketches Stephanus’ historical and cultural context: mid-sixteenth-century France. This is followed by an examination of the edition and its critical apparatus, aiming to provide sufficient knowledge for a subsequent investigation that focuses on a specific witness in the apparatus – Codex Regius. This concentrated analysis allows 4 Samuel Prideaux Tregelles, An Account of the Printed Text of the Greek New Testament; with Remarks on its Revision upon Critical Principles. Together with a Collation of the Critical Texts of Griesbach, Scholz, Lachmann, and Tischendorf, with that in Common Use (London: Bagster, 1854), 31 (emphasis original); see a similar critique in F. H. A. Scrivener, Bezae Codex Cantabrigiensis, Being an Exact Copy, in Ordinary Type, of the Celebrated Uncial Graeco-Latin Manuscript of the Four Gospels and Acts of the Apostles, Written Early in the Sixth Century, and Presented to the University of Cambridge by Theodore Beza, A.D. 1581. Edited With a Critical Introduction, Annotations, and Facsimiles (Cambridge: Deighton, Bell, and co., 1864), x. 5 As the Amsterdam research group put nicely on New Testament conjectures: “this [historical] turn brings their historical context to the fore, instead of limiting their importance to their text-critical quality.” See Bart L. F. Kamphuis et al., “Sleepy Scribes and Clever Critics. A Classification of Conjectures on the Text of the New Testament,” NovT 57 (2015): 72–90, here 73. 6 See for example a now-classic work on a single manuscript by David Parker in his Codex Bezae: An Early Christian Manuscript and its Text (Cambridge: Cambridge University Press, 1992). 7 Brooke Foss Westcott and Fenton John Anthony Hort, The New Testament in the Original Greek. Introduction; Appendix (Cambridge: Macmillan, 1881), 31: “The first step towards obtaining a sure foundation is a consistent application of the principle that knowledge of documents [i.e. manuscripts] should precede final judgement upon readings” (emphasis original).

Stephanus’ Greek New Testament of 1550

307

us to determine the way in which the codex was used, and serves as a test case to evaluate the apparatus. Additionally, the historical influence of the apparatus – its reception history – is also examined, especially in the subsequent editions of Beza, Walton, Mill, Wettstein, and Griesbach. I conclude with remarks intended to demonstrate the advantages to analysing historical editions in the digital age.

The Historical Context of Robert Stephanus In Bruce Metzger’s handbook, Robert Stephanus (1503–1559) is introduced as “[t]he famous Parisian printer and publisher,” followed by a two-page description of his 1550 and 1551 Greek New Testament editions.8 Metzger also informs us that the 1550 edition was the first Greek edition to contain a critical apparatus and that the edition published the following year introduced a numeration system to distinguish verses. Yet, Stephanus’ achievements were much greater than this, as one historian of typography has commented: Stephanus was “a prolific author, a first-class editor, a successful printer, and a conscientious publisher.”9 One of the earliest publications prepared by Robert Stephanus was his first Latin Bible, published in two volumes in 1527 and 1528. This edition, as described by Scrivener, was “the first [Vulgate] edition of a really critical nature” because Stephanus had consulted three manuscripts to prepare the text.10 In fact, 8 Bruce M. Metzger, The Text of the New Testament. Its Transmission, Corruption, and Restoration, 3rd ed. (Oxford: Oxford University Press, 1992), 103–105; same in Bruce M. Metzger and Bart D. Ehrman, The Text of the New Testament. Its Transmission, Corruption, and Restoration, 4 th ed. (Oxford: Oxford University Press, 2005), 149–51. 9 Hendrik D. L. Vervliet, The Palaeotypography of the French Renaissance. Selected Papers on Sixteenth-Century Typefaces (Leiden: Brill, 2008), 97–104, here 97. Even this list does not exhaust Stephanus’ work. What can be added without hesitation is his role as a recognised scholar and lexicographer of classical Latin. As Anthony T. Grafton writes to indicate the contribution of Robert and Henri Stephanus: “In one book [i.e. Thesaurus linguae latinae] the Estienne [sic] revolutionized the field” (Joseph Scaliger. A Study in the History of Classical Scholarship. I: Textual Criticism and Exegesis [Oxford: Clarendon, 1983], 93). For a thorough and precise biography of Robert Stephanus, see Elizabeth Armstrong, Robert Estienne, Royal Printer. An Historical Study of the Elder Stephanus (Cambridge: Cambridge University Press, 1954). Its rev. ed. Armstrong, Robert Estienne, Royal Printer. An Historical Study of the Elder Stephanus (Appleford: Sutton Courtenay Press, 1986) contains a slightly corrected text with a treatment on new aspects of the studies on Stephanus between 1954 and 1984, as well as some other materials. 10 F. H. A. Scrivener, A Plain Introduction to the Criticism of the New Testament for the Use of Biblical Students, 2 vols., 4 th ed., ed. E. Miller (London: Bell, 1894), 2.62 (emphasis added); these words are not found in his first edition, A Plain Introduction to the Criticism of the New Testament for the Use of Biblical Students, 1 st ed. (Cambridge: Deighton, Bell, and co., 1861). The three manuscripts, according to Scrivener, are the Exemplar S.

308

An-Ting Yi

he began his scholarly career by collating Latin biblical manuscripts in 1524, when he was only twenty-one years old.11 The edition of 1527 and 1528 offered a revised Vulgate text together with glossaries and indices, which not only demonstrated the unique talent of this young scholar and printer, but also brought him great success. Since then, Latin Bibles were always the main project in Stephanus’ growing printing press. In 1532, a three-volume edition of the Latin Bible appeared with new Roman type as well as running headings, summaries, and marginal variant readings from examined manuscripts.12 Based on these antecedent works, Stephanus’ most significant examination of the Latin text was published from 1538 to 1540 – a three-volume folio edition of the Latin Bible.13 This edition includes readings from Erasmus’ Annotations as well as variants from more than twelve manuscripts and three editions in its margins.14 Once more in Scrivener’s words: “This edition is practically the foundation of the Modern Vulgate.”15 With its excellent scholarly quality and useful paratextual features, it has been called Stephanus’ masterpiece.16 In the same period, Stephanus’ reputation as a distinguished printer, wellknown within France and around Europe, had earned him a new position: the Royal Printer in Hebrew and Latin. This appointment not only allowed his work to have royal support, but also gave him the privilege of consulting manuscripts Germani parvum (Par. lat. 11937), the Corbey Bible (Par. lat. 11532–3), and the Bible of St. Denis (Par. lat. 2); cf. nos. 22, 20, and 10 respectively in Introduction, 2.68–70 (1894 ed.). 11 Stephanus himself reveals this detail in the preface to his first edition of the Latin Bible. An English translation can be found in Armstrong, Robert Estienne, 11 (1954 ed.). 12 Armstrong, Robert Estienne, 72 (1954 ed.) cites part of the preface to this edition: “In the first place I have compared the Vulgate translation of the Bible afresh so carefully with the ancient manuscript copies which in past years the monks of Saint-Denis and Saint-Germain and the sacred College of the Paris Theologians lent to me, that I think nothing has now been overlooked (for I admit I ‘nodded’ somewhat in the earlier edition, as certain of my friends have complained), and that our translation is now for the first time more or less as whole and complete as of old time the translator himself wrote it” (Armstrong’s translation; emphasis added). Such confidence reflects the Renaissance spirit of the time. See also Armstrong, Robert Estienne, 295–96 (1986 ed.); John Wordsworth, Old-Latin Biblical Texts: No. I. The Gospels according to St Matthew. From the St Germain MS. (g1), now numbered Lat. 11553 in the National Library at Paris (Oxford: Clarendon, 1883), xv. For a detailed description of the new Roman type Stephanus used, see Vervliet, Palaeotypography, 105–48, here 109–11, 114–19. 13 The first volume contains the Hebrew names and the index (1538), the second edition the New Testament (1539), and the third the Old Testament (1540). 14 Wordsworth, Old-Latin 1, xv–xvi and Appendix I (47–54); Armstrong, Robert Estienne, 72–5 (1954 ed.). 15 Scrivener, Introduction, 2.62 (emphasis original); again the first edition does not contain these positive words. To be sure, the standard edition used by the Roman Catholics for centuries is the so-called “Clementine Vulgate” of 1592, an edition that differs from Stephanus’ texts in many places; cf. Scrivener, Introduction, 2.64–5. 16 Armstrong, Robert Estienne, 298 (1986 ed.).

Stephanus’ Greek New Testament of 1550

309

in the collection of King Francis I.17 A few years later, perhaps not earlier than the spring of 1542, Stephanus also became the Royal Printer in Greek.18 Nevertheless, already in November 1540 he had participated in a project to produce a new Greek font called “les Grecs du Roy.”19 Eventually three different sizes of an elegant font appeared: “a Great Primer or Gros-romain (1543), a Pica or Cicéro (1546), and a Two-line Pica or Palestine (1550).”20 The first one was designed for the text of lavish folios and quartos, the second was for portable books or the notes of large editions, and the last for headings.21 With this royal font at hand, Stephanus published a considerable number of Greek classical works in excellent quality, including three Greek New Testament editions issued in 1546, 1549, and 1550.22 Already in the preface of his first two Greek New Testament editions, Stephanus announced his ambition of producing a text superior to Erasmus’.23 Such an ambitious project was based on Stephanus’ access to the royal library. There his son Henri collated several Greek New Testament manuscripts in 1546, the same year Robert launched his first edition of the Greek New Testament.24 In short, in Stephanus was appointed on 24 June 1539; see Armstrong, Robert Estienne, 117–23 (1954 ed.). 18 Although the official appointment had not been issued yet, Stephanus already took charge of Greek printing affairs after the death of his predecessor, Conrad Néobar, who held the position for only eighteen months and died in 1540; see Armstrong, Robert Estienne, 124–30 (1954 ed.). 19 This royal font was engraved by the famous punchcutter Claude Garamont, who reproduced the handwriting of the calligrapher Ἄγγελος Βεργέκιος (Angelo Vergecio). See Vervliet, Palaeotypography, 383–425, here 383–85; Armstrong, Robert Estienne, 52–4, esp. n.4 at 52 (1954 ed.). See e.g. Burney no. 97 in the British Library for Vergecio’s calligraphy: http://www.bl.uk/catalogues/illuminatedmanuscripts/record.asp?MSID=8292 [accessed 4 March 2018]. 20 Vervliet, Palaeotypography, 386. 21 Vervliet, Palaeotypography, 386. Descriptions and illustrations of each size can be seen respectively in Vervliet, Palaeotypography, 130–31 ([Gk 118]; cf. pp. 394–95), 133 ([Gk 80]; cf. pp. 392–93), and 134 ([Gk 150]; cf. pp. 396–97). 22 Armstrong, Robert Estienne, 129–30 (1954 ed.) suggests that the projects of Greek classical works were probably discussed cooperatively by Stephanus and Pierre du Chastel, the Royal Librarian. 23 Robert Stephanus, Τῆς καινῆς διαθήκης ἅπαντα. Novum Testamentum. Ex bibliotheca regia, 1st ed. (Paris: Robertus Stephanus, 1546), 3–4 and Robert Stephanus, Τῆς καινῆς διαθήκης ἅπαντα. Novum Testamentum. Ex bibliotheca regia, 2nd ed. (Paris: Robertus Stephanus, 1549), 3–4. 24 In fact, there are only two direct sources indicating that it was Henri Stephanus who actually did this set of collations: the preface of Henri’s own New Testament edition of 1587 and an additional remark in the preface of Beza’s second New Testament edition of 1565; see Jan Krans, Beyond What Is Written. Erasmus and Beza as Conjectural Critics of the New Testament, NTTS 35 (Leiden: Brill, 2006), 212 n.6. Although in 1546 Henri was only eighteen years old, the assignment was not an unthoughtful one, as he was well-known for his 17

310

An-Ting Yi

preparing his third and long-lived Greek New Testament, the elder Stephanus had the following advantages: a prominent printing press with advanced technologies, scholarly experience of editing Latin Bibles, the royal Greek font, and perhaps the most notable element for textual critics – privileged collations based on the royal manuscripts.

The Editio Regia and Codex Regius The “historical turn” appreciates not only acquaintance with historical backgrounds, but also requires comprehensive examination of a given subject. For this specific study, adequate knowledge of Stephanus’ 1550 edition and its critical apparatus should be obtained before giving judgement upon the information it provides. Let us start by describing the design and layout of this so-called Editio Regia, the 1550 edition.25 On the title page, the heading ex Bibliotheca Regia is present together with the printer’s mark used by Stephanus as the Royal Printer in Greek – an upright spear rounded with a serpent and a laurel plant. Below this mark is an inscription modified from Homer’s Iliad 3.179: Βασιλεῖ τ᾽ ἀγαθῷ κρατερῷ τ᾽ αἰχ µητῇ (“in honour of a noble king and a mighty spearman”; see Figure 1).26 Two separate prefaces follow the title page, the first in Greek and the second in Latin,27 in which Stephanus instructs his readers that he has consulted

prodigiousness in Greek. For instance his monumental lexicon of classical Greek, published in 1572, is called “a bombshell which quite transformed the possibilities of Greek studies in all fields” by Anna Carlotta Dionisotti, “From Stephanus to Du Cange: Glossary Stories,” RHT 14 (1984): 303–36, here 313. Additionally, his father Robert also had similar experience in examining Latin manuscripts in his youth. Thus it was reasonable for our experienced printer to assign this talented son to complete this specific task. 25 Although they were also produced with the royal support, Stephanus’ first two editions (both beginning the preface with “O mirificam Regis nostri…”) are known as editiones mirificae. 26 See Natasha Constantinidou, “Printers of the Greek Classics and Market Distribution in the Sixteenth Century: The Case of France and the Low Countries,” in Specialist Markets in the Early Modern Book World, ed. R. Kirwan and S. Mullins (Leiden: Brill, 2015), 275– 93, here 287, esp. n.46. This device was apparently applied to almost every publication of Stephanus as the Royal Printer in Greek. In other words, not only his Greek New Testament editions, but also other classical texts were produced from and based upon the manuscripts in the royal library, with only one exception. See also Antoine Coron, “The First Libraries: Blois, Fontainebleau, Paris,” in Creating French Culture. Treasures from the Bibliothèque nationale de France, ed. M.-H. Tesnière and P. Gifford (New Haven: Yale University Press, 1995), 151–66, here 159; Armstrong, Robert Estienne, 129 (1954 ed.). 27 Note that a preface written in Greek was extraordinary at that time. Compared to his first two editions containing only Latin prefaces, Stephanus introduced this new edition with

Stephanus’ Greek New Testament of 1550

311

manuscripts from the royal library, some from other locations, and the Complutensian Polyglot.28 The results are printed on the inner margins alongside the text, constituting the first critical apparatus in the history of Greek New Testament editions.

Fig. 1. Part of the title page of the 1550 edition, reproduced by kind permission of the VU University Library

The system of Stephanus’ apparatus has three elements: (1) the Arabic numeral or other symbols linking text and variant,29 (2) the marginal variant itself, and a Greek preface, which apparently was translated from the Latin one, in the elegant royal Greek font. 28 Stephanus, Novum Testamentum (1550 ed.), *iir . For the present study, apart from the digital images provided by the Bayerische Staatsbibliothek München (shelf mark: 999/2Script.67), I have also used a copy held by the University Library at the Vrije Universiteit Amsterdam (XC.05052). 29 More precisely, there are three ways to signal a textual variant: (a) an asterisk (*) for addition, similar to supralinear cross in NA28; (b) enclosed between a grave-accent-like symbol (`) and an apostrophe (’) for omission; and (c) beginning with an Arabic numeral and ending with an apostrophe for either substitution or transposition. Numbering restarts in each chapter, yet infrequently it (mistakenly) continues into the subsequent chapter (e.g. John 19–20). An infamous example of omission is the misplaced apostrophe after the phrase

312

An-Ting Yi

(3) the Greek sigla designating textual witnesses.30 In total this apparatus contains sixteen witnesses: the Complutensian Polyglot and fifteen Greek manuscripts.31 However, such an innovation has been seriously criticised by later textual scholarship. Notably, in 1707 Mill pointed out the incomplete and frequently incorrect information pertaining to readings in the Complutensian Polyglot, and this critique has been cited by subsequent scholars to illustrate the inadequacy of Stephanus’ apparatus. For instance, Tregelles notes that as the Complutensian text is often incorrectly cited in Stephens’s margin, we may conclude that the same thing is true of the manuscripts which were collated; for it would be remarkable if manuscripts were examined with greater accuracy than a printed book.32

Similarly, the mysterious occurrences of η′ (Codex Regius, a manuscript containing only the Gospels) in Acts (24:7 [2x]; 25:14; 27:1; 28:11) have also been used to illustrate the deficiency of Stephanus’ apparatus.33 A different point of ἐν τῷ οὐρανῷ in 1 John 5:7. It appears that this apostrophe should have been placed after ἐν τῇ γῇ in v. 8, that is, the variant signals the omission of the whole passage of the Comma Johanneum in the seven used manuscripts (δ′, ε′, ζ′, θ′, ι′, ια′, ιγ′). Yet the error (the omission only for the initial words, ἐν τῷ οὐρανῷ) came to support the presence of the Comma Johanneum for the centuries to come. Cf. Tregelles, Account, 32. 30 One may wonder whether Stephanus was the inventor of such a complicated system. David Parker points out that the system “is already found in editions by the Italian humanist Politian at the end of the fifteenth century” (An Introduction to the New Testament Manuscripts and Their Texts [Cambridge: Cambridge University Press, 2008], 193). Nevertheless, what had been used by Politian only addressed manuscripts with single (Latin) letters; cf. Michael D. Reeve, “Lucretius from the 1460s to the 17th Century: Seven Questions of Attribution,” Aev 80 (2006): 165–84, here 179. The origin of Stephanus’ system in his apparatus remains an open question. 31 The number of the collated manuscripts is wrongly assigned as “fourteen” in Metzger, Text, 104 (1992 ed., 150 in the 2005 ed.). A list of all these witnesses can be found in Krans, Beyond What Is Written, 337–38; see also J. K. Elliott, “Manuscripts Cited by Stephanus,” NTS 55 (2009): 390–95. Since Stephanus did not provide specific information, it took some time for the following generations to identify which manuscripts were actually being used. Eventually Wettstein was able to identify all the sigla in the 1550 edition. 32 Tregelles, Account, 31 (emphasis original). This citation is actually followed by the sentence I cited at the beginning of this study. More precise information is given by Scrivener (Introduction, 2.190, 300 in the 1861 ed.) on the basis of a thorough comparison: there are more than 2,300 differences between the text of the 1550 edition and of Complutensian but only 610 variants appear, and at least 56 of these are inaccurate. 33 See for example Scrivener, Introduction, 1.137; Elliott, “Manuscripts,” 393. Theoretically speaking, it is possible that Codex Regius did contain some portions of Acts when Henri Stephanus collated it. Nevertheless, a closer look at these five variants recorded in the 1550 edition shows that they are probably merely phantom readings: Acts 24:7 has two variants: (1) text ἀπήγαγε; variant ἀφείλετο (α′, η′); (2) text ἀπήγαγε; variant add καὶ πρός σε ἀπέστειλε (η′). Among Stephanus’ sixteen witnesses, the only one that actually contains these two variants is the Complutensian Polyglot (α′). Hence probably an error occurred: η′ has been put beside α′ in the first case and even replaced it in the second.

Stephanus’ Greek New Testament of 1550

313

view is found in Scrivener’s examination of the variants of β′ (Codex Bezae) in the 1550 edition. He indicates that the “error rate” of this set of data is less than one in fifteen.34 Not only is the percentage of accuracy relatively high, but some variants are extremely long and still recorded faithfully (see Figure 2, for example). Such opposing impressions call for a more detailed and consistent investigation of Stephanus’ critical apparatus. In the following I examine one of Stephanus’ witnesses – η′ (“the eighth one”), which denotes Codex Regius (L019 in the Gregory-Aland system).35 As a case study, this concentrated analysis focuses on all the occurrences of η′ and the instances of π. (stands for πάντες, viz. supported by all witnesses, including η′)36 in the Gospel of John. The results allow us to better assess the quality of the apparatus.

Acts 25:14: text διέτριβον; variant διέτριβεν (α′, ια′, η′). There are four witnesses that contain this variant reading (α′, θ′, ιγ′, ιε′; ια′ is now lost). Given the abnormal sequence of the witnesses (no. 1, no. 11, and then no. 8), a mistake might have happened in the typesetting process by placing η′ for either ιγ′ (GA 398) or ιε′ (GA 82). Acts 27:1: text παρεδίδουν; variant παρεδίδου (η′, ια′). All the remaining witnesses read παρεδίδουν here, albeit θ′ (GA 38), ι′ (GA 2298), and ιγ′ use a ligature to end this word. Perhaps Henri misinterpreted one of these manuscripts and then wrongly attributed it to η′. Acts 28:11: text ἀνήχθηµεν; variant ἤχθηµεν (α′, ε′, η′, ιγ′). Apart from the appearance of η′, the information is correct but partial since θ′ and ιε′ also read ἤχθηµεν. Confusion might have occurred during the process of collation. 34 According to Scrivener, in a total of 389 appearances of D05 there are 25 places that “β′ quite differs from Cod. D” (Bezae Codex, ix). But note that Scrivener’s calculation excludes the sloppy use of π. About half of the instances “loosely or falsely” record the readings of D05 (n.3 at the same page). 35 In fact, this is how a librarian of the royal library in Paris signified Codex Regius. A noteworthy mark “Roberto Stephano Η” can still be observed in the present form of the codex (see the first line of this folio: http://gallica.bnf.fr/ark:/12148/btv1b53045883v /f8.item, accessed 4 March 2018). This mark, already noticed in Scrivener’s time (Introduction, 109 [1861 ed.]; Introduction 1.138 [1894 ed.]), must have been added after Wettstein’s contribution of identification. Furthermore, it would be interesting to know when this codex was entered the royal library. Although no record has been found, a historical event might shed some light on this issue. In 1538, a policy was issued by King Francis I of France to search for and acquire Greek manuscripts abroad. At that time these were extremely scarce in his royal collections. This policy obligated French ambassadors to collect as many manuscripts with Greek characters as they could. As a consequence, some five hundred Greek manuscripts were brought to the royal library between 1540 and 1550. It could be the case that Codex Regius was one of those treasures. Cf. Vervliet, Palaeotypography, 384–85; Coron, “Libraries,” 156–59. 36 Note that this πάντες does not include all the sixteen witnesses but only all the witnesses that contain the present portion. Concerning the portion of the Gospels, the list of witnesses includes ten manuscripts (β′, γ′, δ′, ε′, ϛ′, η′, ζ′, θ′, ιβ′, ιδ′) and of course also the Complutensian Polyglot. Cf. Krans, Beyond What Is Written, 337–38.

314

An-Ting Yi

Fig. 2. A long variant at the end of John 6:56 (p. 174): after κἀγὼ ἐν αὐτῷ (l. 3), an asterisk notifies the addition of β′, reproduced by kind permission of the VU University Library

Before moving to detailed discussions of all the attested variants, a classification of different kinds of errors is required. Scrivener’s rigorous examination on the occurrences of β′ in the 1550 edition is used as the point of departure. According to Scrivener, the errors attested in Stephanus’ apparatus can be divided into three kinds: (1) variants that are recorded loosely, especially in those places where β′ is joined with other witnesses; (2) variants that misidentify scribal corrections as the first hand; and (3) other incorrect variants.37 Developing further Scrivener’s taxonomy, I record below the following types of readings in η′ and π.: (1) correct variants, denoting those places where the apparatus correctly corresponds to the readings in the manuscript, including different spelling and punctuation; (2) imprecise variants, that is, the given variants are recorded though with imprecise information (for instance, different word order or misplaced location); (3) inaccurate attribution, which includes the variants of other witnesses inaccurately attributed to our codex; and (4) other incorrect variants.38 In the Fourth Gospel of Stephanus’ Editio Regia, there are seventy-nine variants that concern Codex Regius (η′ and π.).39 After comparing with my own transcription, fifty-seven readings precisely record what our codex contains.40 Additionally, ten readings are correct, but either contain normalised spellings or different punctuation.41 As a result, sixty-seven out of seventy-nine belong to the first category (nearly 85%). Scrivener, Bezae Codex, ix. Compared to D05, L019 has fewer scribal corrections, and none of these appears in the variants of η′ in John. Thus I have modified Scrivener’s first two categories. 39 See the Appendix for a full list of all the attested variants. 40 Yet the variant at John 8:9 is interesting: it is recorded that η′ (and α′) omits the clause καὶ ὑπὸ τῆς συνειδήσεως ἐλεγχόµενοι, but actually the whole pericope adulterae is absent in L019. 41 These variants are: John 1:39 ὄψεσθε (ε > αι); 2:17 καταφάγεται (αι > ε); 6:40 γάρ ἐστι (omit final ν); 8:38 ἃ ἠκούσατε παρὰ τοῦ πατρός (ε > αι); 13:24 καὶ λέγει αυτῷ, εἰπὲ τίς ἐστι 37 38

315

Stephanus’ Greek New Testament of 1550

The twelve variants that belong to the other three categories are the following: Table 1. Variants of η′ in John: categories 2, 3, and 4 John

Text

Witnesses

Variant

Cat.

omit

Correct reading of Regius as text

1

6:15

πάλιν

π.

2

8:9

ἑστῶσα

3

8:49

ἀτιµάζετε

π.

οὖσα

omit

3

η′

ἠτιµάσατε

as text (ε > αι)

4

4

12:31

τούτου

β′, ε′, η′

omit

as text

3

5 6

14:22

τί

π.

add καί

as text

3

15:1

ἀληθινή

δ′, η′

as text

3

7

16:3

ὑµῖν

π.

add ὑµεῖς τὰ κλήµατα omit

as text

3

8

18:29

αὐτούς

η′

add ἔξω

misplaced

2

9

19:13 21:15

α′, γ′, ε′, η′, ιβ′
 η′

τούτων τῶν λόγων Ἰωαννᾶ

τὸν λόγων τούτων Ἰωάννου

2

10

τοῦτον τὸν λόγον Ἰωνᾶ

11

21:16

Ἰωνᾶ

η′

Ἰωαννᾶ

lacunose

2

12

21:17

Ἰωνᾶ

η′

Ἰωαννᾶ

lacunose

2

3

2

Category 4 (“other wrong variants”) has only one occurrence: at John 8:49 Codex Regius reads ἀτιµάζεται, not ἠτιµάσατε.42 Next, the third category (“inaccurate attribution”) contains six variants. Those at John 6:15; 8:9; 14:22 and 16:3 are all due to inaccurate use of π. (πάντες), and among them the note at 8:9 is of particular interest. Henri Stephanus might have forgotten that, unlike other witnesses he consulted, Codex Regius merely has a large blank space in the place where John 7:53–8:11 would occur in many manuscripts.43 Furthermore, there are two more inaccurate accidents where η′ is juxtaposed to other witnesses. At John 12:31 the codex does not omit the word τούτου, and similarly it follows the text without addition at John 15:1.44 The second category has the remaining variants. First, at John 18:29 we are informed that η′ has an additional adverb ἔξω after πρὸς αὐτούς. Yet the actual (omit final ν); 15:8 γενήσθε (ε > αι); 20:11 τῷ µνηµείῳ (ει > ι); and 20:31 add αἰώνιον (ι > η). Additionally, the two variants at 13:30–31 and 17:12 have punctuation that differs from L019. 42 As far as I know, no witness in Stephanus’ edition has such a variant. 43 L019 219v–220r; see the discussion in the “Notes on Select Readings” section of the appendix in Westcott and Hort’s Introduction, 83. 44 In fact there are three witnesses that omit τούτου at John 12:31: α′, β′, and ε′. And only δ′ (GA 5) contains the addition of ὑµεῖς τὰ κλήµατα at John 15:1.

316

An-Ting Yi

location of this word is before πρὸς αὐτούς. Such a discrepancy may probably be made either in the collation work or during the typesetting process. Second, the error at John 19:13 seems due to confusion with readings in other witnesses, perhaps in particular ε′ (GA 6).45 A possible scenario could be that after collating the reading from ε′ (assuming that the procedure follows the sequence of the sigla), Henri noticed a similar reading in Codex Regius – a bizarre one indeed – and then recorded it beside τούτων τῶν λόγων. Third, the three variants at John 21:15–17 should be considered together. Given that the present text of Codex Regius ends with the first several words of v. 15, it is impossible to know the exact wordings of vs. 16–17 of the codex. However, there is a variant reading at v. 15, which concerns the name of Peter’s father. Our codex reads Ἰωάννου (genitive of Ἰωάννης) and differs from the Textus Receptus reading, the core of which is preserved in the text of Stephanus’ 1550 edition.46 In contrast, we read Ἰωνᾶ (genitive of Ἰωνᾶς)47 with a marginal reading “Ἰωαννᾶ. η” three times in this very edition, as shown in Figure 3. At first glance, this “error” appears to be a simple mistake, yet a tentative reconstruction of how Henri prepared his collations provides evidence for our exploration. When he collated the manuscripts from the royal collection in 1546, the best edition Henri could use was Erasmus’ last edition, either the 1535 edition or the 1540 Basel edition. Notably the Erasmian text read Ἰωαννᾶ, the same reading as Stephanus’ variant,48 and thus Henri could have noted that Codex Regius supported the edition he used for collation, although these two words were slightly different in actuality.49 If this is the case, then the occurrences of the Among the cited witnesses, the information about ε′ is correct, and ιβ′ (GA 9) apparently has τω λογων τουτον, although it is hard to identify. The other two (α′ and γ′) follow the text exactly. 46 See Krans, Beyond What Is Written, 143 n.165: “Strictly speaking, the Textus Receptus is not a single text, but comparable to a text-type. Stephanus, Beza, and the Elzevirs’ texts belong to its core, while Erasmus’ texts can be seen as part of the Textus Receptus, but with many idiosyncrasies.” See also Metzger, Text, 106 (1992 ed., 152 in the 2005 ed.). 47 The reading of L019 (Ἰωάννου) is equivalent to NA28. Additionally, the reading of Ἰωνᾶ is probably an assimilation to Matt 16:17; cf. Bruce M. Metzger, A Textual Commentary on the Greek New Testament, 2nd ed. (Stuttgart: Deutsche Bibelgesellschaft, 1994), 220, 172. 48 Desiderius Erasmus, Novum Testamentum iam quintum accuratissima cura recognitum a Des. Erasmo Roter. cum Annotationibus eiusdem ita locupletatis, ut propemodum opus novum videri possit, 5 th ed. (Basel: Froben, 1535). This was a well-printed folio edition, so it was ideal for collation. Stephanus’ first edition of 1546, on the other hand, was in sextodecimo size which made it difficult to write anything apart from the text. Aside from this, Stephanus’ first two editions read Ἰωνᾶ in John 21:15–17 as well. Tregelles, Account, 30 also suggests that Henri probably made collations of Erasmus’ last edition, although he simply asserts this proposal. 49 Interestingly, at John 1:42 a variant of the same kind – Ἰωάννου rather than Ἰωνᾶ – was precisely recorded in the 1550 edition. Erasmus’ last edition read Ἰωνᾶ there, just like 45

Stephanus’ Greek New Testament of 1550

317

variants at vs. 16 and 17 can be explained as either a logical conjectures based on the presumed agreement between these three places or a mistake made after this codex was no longer available to Henri.50

Fig. 3. The three variants on the name of Peter’s father, reproduced by kind permission of the VU University Library

Apart from the above categories that concern the quality of Stephanus’ variant readings, the following analysis of John 1 tests the quantity of the information provided in this apparatus. In other words: how comprehensive is Stephanus’ apparatus in John 1 in capturing variation between Codex Regius and the main text? The 1550 edition records four (correct) variants in the chapter, at 1:28, 39 (2x), and 42.51 However, our codex actually differs in forty-one places from the

Stephanus’ text. It could well be that Henri observed that particular variant at John 1:42 and wrote it down apart from the word Ἰωνᾶ in his collation book. The inconsistent text in Erasmus’ edition might also cause these imprecise variants in John 21:15–17. 50 Viewed in this light, although it cannot be totally excluded, the possibility that Henri indeed read the (now lost) last folio of John in the codex and collated the two variants accordingly is relatively low. 51 Note that the apparatus of this edition is actually a selection of Henri’s collation book, albeit with a majority of the variants; see Krans, Beyond What Is Written, 212–13.

318

An-Ting Yi

text of Stephanus, even excluding orthographical differences, diverse spellings of proper names, and obvious scribal errors.52 First, there are seven additions in Regius vis-à-vis the main text, none of which are recorded in the apparatus. Four of these additions are definite articles. For instance, Codex Regius reads καὶ ὁ θεός in John 1:1c.53 Similar to the additions, each of the three instances of transposition are absent in the apparatus.54 And there exist nineteen variants of omission, but only the one at John 1:39 is found in the margins of the 1550 edition. Although most differences are minor, a few of these are indeed significant. Notably the clause ὃς ἔµπροσθέν µου γέγονεν in John 1:27 is omitted in the codex, agreeing with the text of NA28. There are also twelve variants of substitution, and Stephanus notes three of them: Βηθανίᾳ for Βηθαβαρᾶ at 1:28 (supported by π.),55 ὄψεσθε for ἴδετε at 1:39,56 and Ἰωάννου for Ἰωνᾶ at 1:42 (both the second and the third are supported by η′ alone). Yet, two other neglected readings deserve to be mentioned: µονογενὴς θεός for ὁ µονογενὴς υἱός at 1:18 and µεθερµηνευόµενον for ἑρµηνευόµενον at 1:38. Once again our codex offers readings in agreement with two of the best witnesses to reconstruct the text of this gospel: P66 and Codex Vaticanus.57 52 Such a significant number of differences is not hard to imagine, since L019 has a text close to B03. In John 1–10, for instance, 65% of the test passages selected by the Münster team are in agreement between B03 and L019. Cf. Kurt Aland, Barbara Aland, and Klaus Wachtel, eds., Text und Textwert der griechischen Handschriften des neuen Testaments V. Das Johannesevangelium. 1. Teststellenkollation der Kapitel 1–10, vol. 1/2, ANTF 36 (Berlin: De Gruyter, 2005), 2*. Also, according to their “Manuscript Clusters” (http://intf.unimuenster.de/TT_PP/), L019 is the second closest manuscript to B03 in the text of John, only lower than P75. 53 Additionally, three other additions (marked in italics) are (1) οὖν καὶ παρʼ αὐτῷ ἔµειναν τὴν ἡµέραν ἐκείνην (v. 39c); (2) καὶ εἷς ἐκ τῶν δύο τῶν ἀκουσάντων παρὰ Ἰωάννου (v. 40); and (3) καὶ λέγει αὐτῷ ὁ Ἰησοῦς (v. 43b). 54 These readings are (1) ἐγὼ οὐκ εἰµί for οὐκ εἰµὶ ἐγώ (v. 20); (2) οἱ δύο αὐτοῦ µαθηταί for αὐτοῦ οἱ δύο µαθηταί (v. 37); (3) σὺ βασιλεὺς εἶ for σὺ εἶ ὁ βασιλεύς (v. 49; this one actually belongs to the “combination” type, for it also omits the article). 55 Cf. Metzger, Textual Commentary, 171. Unlike later editions of the Textus Receptus, in Stephanus’ text the iota subscript is absent. Additionally, Origen proposes an interesting conjecture (cj10521) supporting Βηθαβαρᾷ; see http://ntvmr.uni-muenster.de/nt-conjectures?conjID=cj10521 [accessed 4 March 2018]. 56 Note that L019 actually has ὄψεσθαι here. 57 At John 1:18, ℵ01 also supports this reading, but P75 has ὁ µονογενὴς θεός instead. At John 1:38, the situation reversed: P75 has the same reading of the text of L109 but ℵ01 reads λέγεται ἑρµηνευόµενον. The judgement on the former case is indeed difficult to make, especially before the discovery of P66: Tischendorf’s text (Novum Testamentum Graece. Ad antiquissimos testes denuo recensuit, apparatum criticum omni studio perfectum apposuit, commentationem isagogicam praetexuit…Editio octava critica maior, vol. 1 [Leipzig: Giesecke & Devrient, 1869], 752, at v. 39) stands with ℵ01 (and the Textus Receptus) here; yet both Tregelles (The Greek New Testament. Edited from Ancient Authorities, with Their

Stephanus’ Greek New Testament of 1550

319

This data demonstrates the conflicting nature of Stephanus’ apparatus: the quality of its attested variants is acceptable, but the quantity is nevertheless disappointing. On the one hand, among the seventy-nine variants concerning η′ in John in the 1550 edition, five of them are loosely recorded and seven others either wrongly attribute a reading to other witnesses or are simply incorrect. Furthermore, if the sloppy use of πάντες is excluded (as Scrivener’s analysis does), then the error rate drops to about ten percent. In the light of the pioneering work of an edition such as this, Stephanus’ apparatus, at least as far as η′ in John is concerned, has been prepared with considerable care. On the other hand, the quantity of the recorded information is very low. Given that the text of the codex is close to the critical text we now prefer, one would expect some distinctive readings to be recorded. Yet, only very few variants are put into the margins of Stephanus’ edition, but many notable readings of our codex have been overlooked. These results might have been due to the infancy of collating New Testament manuscripts, and as the next section will show, Stephanus’ context played a significant role in his working method.

Reassessing Stephanus’ Apparatus in Context In light of the preceding investigation, we can now restate the question: on what grounds shall Robert Stephanus’ innovative product be evaluated? A primary observation that needs to be stated at the outset is that this apparatus undoubtedly fails to meet the standard set by later generations of textual critics. It contains only a limited number of variants without any particular articulated principle of selection.58 Moreover, concerning the function of this variant collection, it seems to have had little influence on Stephanus’ text, as the following citation from Scrivener shows: “I would not say with some that these authorities stand in the margin more for parade than use, yet the text is perpetually at variance with the majority of them, and in 119 places with them all.”59

Various Readings in Full, and the Latin Version of Jerome. Part II. Gospels of St. Luke and St. John [London: Bagster, 1861], 381) and Hort (“Notes,” 74) follow the reading we now prefer. Cf. Metzger, Textual Commentary, 169–70. 58 See for example Tregelles’ evaluation: “The readings were selected by Henry Stephens, the editor’s son, on no very particular principle apparently, and with but little exactitude” (Thomas Hartwell Horne and Samuel Prideaux Tregelles, An Introduction to the Critical Study and Knowledge of the Holy Scriptures. Volume IV. An Introduction to the Textual Criticism of the New Testament; with Analyses, etc., of the Respective Books, and a Bibliographical List of Editions of the Scriptures in the Original Texts and the Ancient Versions, 10th ed. [London: Longman, Brown, Green, Longmans, & Roberts, 1856], 123). 59 Scrivener, Introduction, 2.190 (300 in the 1861 ed.).

320

An-Ting Yi

The variant in the 1550 edition at John 1:28 – Βηθανίᾳ for Βηθαβαρᾶ60 – is a good example to illustrate Scrivener’s point. This variant was not only supported by all the consulted witnesses, but it had also been the text in Stephanus’ previous editions. Nevertheless, in his third edition he turned back to the reading offered by Erasmus, and Βηθαβαρᾶ/Βηθαβαρᾷ eventually became the Textus Receptus reading for centuries to come. In other words, instead of using his son’s collation to improve the text, Robert followed a poor text in his principal edition of 1550. Given that he had access to Codex Regius, which essentially contains a better text, it is not surprising that his work has been regarded as uncritical. However, this apparatus and the collation work that stands behind it demonstrates a certain level of text-critical acumen. Variants were collated with considerable care and presented, for the most part, accurately. Moreover, my tentative reconstruction of the process of Henri’s collations should also be taken into account, especially the fact that the edition he collated against was likely Erasmus’ last edition. As a result, in order to preserve the collations as precisely as possible, the Erasmian text was adopted as the base text of Stephanus’ 1550 edition. Collation thus influenced the main text in a direct manner. This was the easiest way to avoid errors and it also saved considerable time for the preparation of typesetting. Such a particular motivation – indeed a very practical one – can explain the recurrence of the πάντες notation as well. This practical motivation also helps us to bring Robert Stephanus back to his own historical context. As the King’s Printer, his highest concern was not necessarily the publication of a scholarly edition that met the highest crucial standards. Much more than that, this Editio Regia was dedicated, first and foremost, to glorifying his patron, King Henry II. What could be a more perfect gift than a new edition of the sacred text in its original language by using the royal font, based on the manuscripts from the royal library? Except for the critical apparatus discussed in this study, the paratextual emphasis of the Greek materials in the edition is another good example. As I noted, the 1550 edition begins with a unique Greek preface. It is followed by a wealth of supplemental materials in Greek: John Chrysostom on the Gospels, the Eusebian canons, the kephalaia and their titloi, and even a Greek poem.61 These materials allowed Stephanus to put as much Greek – in the royal font no less – in the edition as possible, aiming to impress his readers (including the French King) from the very first pages (see Figure 4). Thus, if we examine this edition from such a political perspective, many elements of the apparatus come into clearer focus. 60 Stephanus does not give Βηθαβαρᾷ but Βηθαβαρᾶ in his 1550 edition, yet the iota subscript is still present in Erasmus’ final edition. 61 See Scrivener, Introduction, 2.191–92 for a summary of these supplements. The poem, superscribed with Ἑρρίκος ὁ Ροβερτου Στεφάνου, φιλοθέῳ παντί (“Henri, son of Robert Stephanus, to all who love God”), was composed by Henri Stephanus who had great aptitude for Greek composition.

Stephanus’ Greek New Testament of 1550

321

Viewed in the same light, the critical apparatus can even be taken as a diplomatic exhibition, an exhibition for demonstrating the readings not only drawn from the royal manuscripts but also printed in the royal font.62 Once we begin to contextualise Stephanus and his edition, puzzling features of the edition become more intelligible. Armstrong puts this nicely: [Stephanus] lived by selling his editions, mainly to a critical professional-class public in an ever increasingly competitive international booktrade. His outstanding excellence as a printer was an asset. So was his scholarship. But he had a talent without which these might have done no more than earn him a livelihood and an honourable reputation: he was a practical man of business.63

Therefore, somewhat ironically, the first critical apparatus containing textual variants of Greek manuscripts was not prepared in a purely academic setting; a wider array of political and economic concerns are also present. Nevertheless, under certain historical circumstances, this apparatus became the only accessible information on Codex Regius in the hundred years after 1550. For a long period of time, every editor of the Greek New Testament had to rely on the variant readings in Stephanus’ edition to glimpse the text of this codex. The next section explores some remarkable stories of this history of usage.

Fig. 4. Part of p. 201 (John 19:16–19), with permission of the VU Library

62 I would therefore argue that Tregelles’ observation cited at the beginning of this study (“the various readings in the margin of Stephens’s folio edition seem rather to be appended as an ornament to the text”) could be seen as a correct judgement, although I have arrived at this conclusion from a different angle. 63 Armstrong, Robert Estienne, 306 (emphasis added, 1986 ed.); see also 294–302 for Stephanus’ strategy in publishing his Bible projects, esp. 294: “But all his piety nor wit would not have availed him to sustain such a costly publishing programme had he not also exercised a well thought out strategy, constantly revised and renewed, to ensure that his editions brought in a reasonable profit.”

322

An-Ting Yi

Stephanus’ “Borrowers”: From Beza to Griesbach The Stephanus apparatus is not just noteworthy because it is the first one that records Greek manuscripts, but because of its wide-spread history of usage, particularly by later editors of the New Testament. Evaluating the reception history of the apparatus is especially important because the text of Regius was only accessible through the 1550 edition for more than a century and a half. It could even be said that Stephanus’ apparatus was “on loan,” and that the loan period continued until the eighteenth century. The following discussions focus on some notable “borrowers” of Stephanus. The first borrower was Theodorus Beza (1519–1605). Apart from his role as Calvin’s successor at Geneva, Beza also succeeded Stephanus as an editor of New Testament editions in the second half of the sixteenth century.64 His first folio edition, finished in 1556 and published in 1557, actually belonged to Stephanus’ last Bible project. It contained Stephanus’ Vulgate text as well as Beza’s new translation with annotations. From Beza’s second edition of 1565 onwards, the content also included a Greek text of the New Testament.65 Moreover, Beza had made use of both the critical apparatus of Robert Stephanus’ 1550 edition and Henri’s collations in his annotations to the text.66 To put it more precisely, in some portions of his first edition, Beza referred to the manuscripts with ordinal numbers, which generally agree with the Greek sigla in Stephanus’ apparatus. The relationship between the editions is readily perceptible in the following example. As I noted above, at John 1:28 Stephanus correctly records a variant (Βηθανίᾳ for Βηθαβαρᾶ) supported by all the examined witnesses. In his first edition, Beza followed the steps of his predecessor faithfully and made an annotation on “in Bethabara” of the Latin text. He states that an alternative reading “in Bethania” 64 A piece of historical information might be of interest. Due to a series of bans on his Bible editions issued by the censors in Paris, Stephanus emigrated to Geneva in November 1550, the same year that the Editio Regia appeared. At that time he became a Protestant and lived there until his death. Stephanus also transferred his printing career to Geneva and published his fourth edition of the Greek New Testament (1551) in the same city. Additionally, Calvin’s definitive edition – the Institutio Christianae religionis of 1559 – was also printed by Stephanus’ publisher. Cf. Armstrong, Robert Estienne, 147–54, 211–20 (1954 ed.). 65 See Krans, Beyond What Is Written, 202–206. 66 For a thorough discussion on this topic, see Krans, Beyond What Is Written, 211–16. In the preface of his first edition (Theodorus Beza, Novum D. N. Iesu Christi testamentum [Geneva: Robertus Stephanus, 1556], 2r), Beza himself states that he has consulted Henri’s collations directly. Interestingly, Beza seems to give the impression that he did examine the manuscripts himself. Cf. Johann Jakob Wettstein, Ἡ Καινὴ Διαθήκη. Novum Testamentum Graecum editionis receptae cum lectionibus variantibus Codicum MSS., Editionum aliarum, Versionum et Patrum nec non commentario pleniore ex Scriptoribus veteribus Hebraeis, Graecis et Latinis historiam et vim verborum illustrante. Tomus I. Continens quatuor Evangelia (Amsterdam: Dommeriana, 1751), 148; Krans, Beyond What Is Written, 213, esp. n.10.

Stephanus’ Greek New Testament of 1550

323

is found in the Vulgate and ἐν Βηθανίᾳ appears in “all the old manuscripts.”67 Similarly, in Stephanus’ edition the variant at John 1:42 (Ἰωάννου for Ἰωνᾶ) notes that η′ is the only witness. Beza thus annotates: “It was written Ἰωάννου in the eighth manuscript.”68 The inaccurate information of the Stephanus edition at John 15:1 (the addition of ὑµεῖς τὰ κλήµατα, supported by δ′ and η′) has also been preserved in Beza’s annotations.69 As usual, only a vague expression is offered: “It was added ὑµεῖς τὰ κλήµατα in two copies.”70 These examples show that Beza solely relied on Stephanus’ sources. As a result, both the correct and the inaccurate readings have been transmitted. Additionally, through Beza’s first edition, a number of Stephanus’ variant readings entered the Geneva New Testament of 1557 as marginal notes, which were generally followed by the Geneva Bible published three years later.71 Stephanus’ apparatus was indeed influential. The so-called London Polyglot Bible, prepared by Brian Walton (1600–1661), is the next borrower that deserves attention. The fifth volume contains the New Testament text in six languages – Greek, Latin, Syriac, Ethiopic, Arabic, and Persian (only in the Gospels)72 – and its Greek text usually agrees with Beza, Novum testamentum, 109r (same in Novum D. N. Iesu Christi testamentum, sive Novum Foedus, 2nd ed. [Geneva: Henricus Stephanus, 1565], 360): “in omnibus vetustis codicibus.” In fact the first one of these witnesses, the Complutensian Polyglot, is neither old nor a manuscript, yet Beza seems to treat it as equal to other witnesses. 68 Novum testamentum, 109v: “In octauo codice scriptum erat Ἰωάννου.” This annotation is actually assigned to v. 41. In the second edition, the annotation (which is now at v. 43) has been modified: “We read Ἰωάννου in one manuscript” (“In uno codice scriptum legimus Ἰωάννου,” 362). 69 Six other variants that belong to my third and fourth categories are absent in Beza’s work. 70 Novum testamentum, 128r (427 in the 1565 ed.): “In duobus exemplaribus additum erat ὑµεῖς τὰ κλήµατα.” The impreciseness of Beza’s information is criticised by many scholars; cf. Krans, Beyond What Is Written, 216–18. 71 According to Metzger, there are twenty-one variant readings in the margins of the New Testament of the Geneva Bible of 1560 (“The Influence of Codex Bezae upon the Geneva Bible of 1560,” NTS 8 [1961]: 72–7, here 73–5). Among them only two readings concern the Gospels, one at Luke 17:35 (v. 36 of the modern versification system) and another at John 8:59. The former originates from an addition in the critical apparatus of Stephanus (supported by α′, β′, γ′, ϛ′, ζ′). The latter, although absent in Stephanus’ apparatus, can be found in Beza’s annotation in his first edition (Novum testamentum, 121v: “After these words it is placed in the old copies, διελθὼν διὰ µέσου αὐτῶν, καὶ παρῆγεν οὕτως” [“Post haec verba subiicitur in vetustis exemplaribus, διελθὼν διὰ µέσου αὐτῶν, καὶ παρῆγεν οὕτως”]; same in the 1565 ed., 403). Our codex is actually one of the “old copies” mentioned by Beza, who must have received this piece of information from Henri’s collations. At this point Metzger has failed to notice the presence of Beza’s annotations in his edition of 1556 (“Geneva Bible,” 75–7). Cf. Krans, Beyond What Is Written, 197 (esp. n.7) and 205 n.31. 72 Brian Walton, Bibliorum sacrorum tomus quintus: sive Novum D. N. Jesu Christi Testamentum (London: Roycroft, 1657). For a discussion of Walton’s significant volumes as well as his historical context, see Peter N. Miller, “The ‘Antiquarianization’ of Biblical 67

324

An-Ting Yi

Stephanus’ 1550 edition. For textual critics, a more notable element is that variant readings from Codex Alexandrinus are given below the Greek text.73 The sixth volume of appendices includes a thirty-six-page critical apparatus, assembling a considerable number of variants in the whole Greek New Testament.74 By collecting the work done by previous scholarship as well as his contemporaries, Walton was able to synthesise the stock of existing collations into one collection, in which Stephanus’ variant readings were positioned on the top of the list.75 Just as Beza himself did not examine (most of) the manuscripts used by Stephanus,76 neither did Walton. What Walton did was to draw some variants from Stephanus’ edition and present them in his collection. For instance, three of the four readings concerning η′ in John 1 are found in Walton’s Polyglot.77 Likewise, almost all imprecise or inaccurate readings noted above are preserved in Walton’s collations, just like Beza in his annotations.78 Beyond Beza and Walton, even John Mill (1645–1707), whose work “has been said to commence the age of manhood in the criticism of the Greek Testament,”79 also borrowed the variant readings of Codex Regius from Stephanus. In his foundational edition Novum Testamentum of 1707, Mill aggregated some ninety manuscripts as well as ancient versions and patristic citations.80 Scholarship and the London Polyglot Bible (1653–57),” JHI 62 (2001): 463–82. See also Alastair Hamilton, “In Search of the Most Perfect Text: The Early Modern Printed Polyglot Bibles from Alcalá (1510–1520) to Brian Walton (1654–1658),” in The New Cambridge History of the Bible. Volume 3: From 1450 to 1750, ed. E. K. Cameron (Cambridge: Cambridge University Press, 2016), 138–56, here 151–54. 73 Cf. Scrivener, Introduction, 2.197–98 (312–13 in the 1861 ed.). 74 Brian Walton, Ad Biblia sacra polyglotta appendix. In quo varii Tractatus, Annotationes, Lectiones variae, Hebr. Graec. Lat. Samarat. Chald. Syr. Arab. Aethiop. Pers. cum Indicibus, etc. quae tomum sextum constituunt (London: Roycroft, 1657), “Variantes Lectiones Graecae Novi Testamenti” (numbered separately). 75 Walton, “Variantes Lectiones Graecae,” 1; see Scrivener, Introduction, 2.197–98 (312– 13 in the 1861 ed.) for the identification of the sixteen collections. 76 Although he was unaware of it, Beza actually had kept one of the manuscripts (Stephanus’ β′) for many years. This is why it has been called Codex Bezae. On this topic see Krans, Beyond What Is Written, 228–34 for a detailed discussion. 77 Walton, “Variantes Lectiones Graecae,” 17. The second variant at John 1:39 (δέ; omitted by π.) is missing. 78 Only one variant is left out in Walton’s appendix: the addition of καί at 14:22 (no. 5 in Table 1 above). 79 Tregelles, Account, 41. See also Scrivener, Introduction, 2.201: “Dr. Mill’s services to Biblical criticism surpass in extent and value those rendered by any other, except perhaps one or two men of our own time” (361 of the 1861 ed. has “…except perhaps one man yet living”). 80 John Mill, Novum Testamentum. Cum lectionibus variantibus MSS. exemplarium, versionum, editionum, SS. Patrum et Scriptorum ecclesiasticorum; et in easdem notis. Accedunt loca Scripturae parallela, aliaque ἐξηγητικά et appendix ad variantes lectiones. Praemittitur dissertatio, in qua de libris N.T. et Canonis constitutione agitur, historia S. Textus N.

Stephanus’ Greek New Testament of 1550

325

Nevertheless, although he did collate a sizeable quantity of manuscripts himself, Mill was unable to consult the manuscripts in the royal library of France. Consequently, after carefully examining the presence of η′ in Stephanus’ margins, Mill stated – somewhat hesitantly – that this very codex contains not only the Gospels but also the book of Acts.81 At the same time, due to Regius’ inaccessibility, Mill simply reproduced the same information given by Stephanus in his own apparatus. Concerning Codex Regius in John, Mill precisely recorded almost all variant readings of Stephanus’ η′, including the twelve incorrect cases mentioned above.82 Additionally, although he provided many variants on the codex, Mill seldom made his own judgement on its quality.83 In sum, regardless of the reputation of preparing an epoch-making edition, Mill was still a borrower of Stephanus insofar as it came to Codex Regius.84 Foederis ad nostra usque tempora deducitur, et quid in hac editione praestitum sit, explicatur, 1 st ed. (Oxford: Theatrum Sheldoniacum, 1707). Three years later Ludolf Küster published the second edition (Novum Testamentum Graecum, cum lectionibus variantibus MSS. exemplarium, versionum, editionum, SS. patrum et scriptorum ecclesiasticorum; et in easdem notis. Accedunt loca scripturae parallela, aliaque exegetica. Praemittitur dissertatio de libris N.T. et canonis constitutione, et s. textus N. Foederis ad nostra usque tempora historia. Studio et labore Joannis Millii S.T.P. Collectionem Millianam recensuit, meliori ordine disposuit, novisque accessionibus locupletavit Ludolphus Kusterus [Rotterdam: Fritsch & Böhm, 1710]), rearranging the information in the appendix to its proper places, adding variant readings of twelve new manuscripts, and inserting the section number to the prolegomena. Cf. Scrivener, Introduction, 2.203–204. 81 John Mill, “Prolegomena,” in his first edition, cxviii (§1168 in the 1710 ed.). See the discussion above on the five “phantom variants” in Acts. Additionally, Mill considers that Stephanus regarded this codex (and also β′) as an inferior one (“πονηροῦ κόµµατος,” perhaps meaning “of a bad stamp”), for it has been put in the margins many times against the main text. Such an impression may have been drawn from Mill’s misunderstanding that Stephanus’ critical apparatus was first and foremost a text-critical one, as in Mill’s own edition. 82 As far as I know, among the seventy-nine variants only the omission of οὖν at John 12:21 (η′) and the substitution (ἴσχυον for ἴσχυσαν; γ′ and η′) at John 21:6 are omitted. 83 In the Fourth Gospel, Mill discusses the variants that relate to Stephanus’ η′ in twelve places, yet most of them merely contain a summary of previous scholarly opinions, either Erasmus, Grotius, or Lucas Brugensis. As regards the substitution at John 13:30–31 (νύξ. Ὅτε οὖν ἐξῆλθε for νὺξ ὅτε ἐξῆλθε), he refers to Grotius’ judgement that this reading was derived from a Latin exemplar (Mill, Novum Testamentum, 287 n.t) due to insufficient supports. Additionally, twice (the incorrect addition at John 15:1 and the correct one at 20:31) he indicates that these readings have been inserted from other places in the Gospel; see Novum Testamentum, 290 n.t and 306 n.k. Only at one place does he discuss extensively where John baptised (i.e. John 1:28), but η′ has little, if any, significance in that discussion (Novum Testamentum, 245 n.i). 84 To be sure, such a limitation was unavoidable for an edition that tried to gather as many of the variant readings that had been previously noticed as possible. In order to achieve this goal, Mill must have relied on other works for manuscripts he could not examine personally. Thus, similar errors might have occurred due to the imprecise sources to which Mill had access.

326

An-Ting Yi

Codex Regius, however, became available once again to critics from the beginning of the eighteenth century onwards. Two esteemed scholars examined it personally and made use of its variant readings in their editions. In 1715, Johann Jakob Wettstein (1693–1754) went to Paris to collate some Greek manuscripts of the New Testament, including Codex Regius.85 The new readings gained from this activity were incorporated in the first volume of his Greek New Testament edition (1751–1752). According to my examination, Wettstein’s edition records 481 variant readings in the John portion of Codex Regius, which far surpasses the variants given by Stephanus.86 However, although he collated the codex himself, a number of readings recorded in Wettstein’s edition resembled the incorrect information of the Editio Regia of 1550. More precisely, four readings (John 8:49; 12:31; 15:1; 19:13) record exactly the same errors as found in Stephanus’ apparatus. In particular, at John 15:1 after mentioning the addition of ὑµεῖς τὰ κλήµατα in Codex Regius, Wettstein expressed his hesitation with an intriguing remark: “if the numbers have been placed properly by Robert Stephanus.”87 They had not been. In the second half of the eighteenth century, Johann Jakob Griesbach (1745– 1812) also personally scrutinised Codex Regius.88 As a result, he inserted hundreds of readings into the second edition of his Greek New Testament.89 By re-

85 Wettstein states confidently in Prolegomena ad Novi Testamenti Graeci editionem accuratissimam, e vetustissimis codd. MSS. denuo procurandam; in quibus agitur de codd. MSS. N. Testamenti, Scriptoribus Graecis qui N. Testamento usi sunt, versionibus veteribus, editionibus prioribus, et claris interpretibus; et proponuntur animadversiones et cautiones ad examen variarum lectionum N. T. necessariae (Amsterdam: Wettstein & Smith, 1730), 19: “I collated it [Codex Regius] hastily but nevertheless much more accurately than Stephanus” (“festinanter contuli, sed multo tamen accuratius Stephano”). The year in which he made the collation can be found in the revised “Prolegomena” (Wettstein, Novum Testamentum Graecum, 1.41). 86 Among these, 446 (ca. 93%) are correct, including 24 with different spellings. 22 variants are recorded imprecisely (ca. 4.6%). In the remaining 12 places (ca. 2.5%), the text of the codex contains no variant and thus should be considered inaccurately recorded. 87 Novum Testamentum Graecum, 1.936: “si numeri a R. Stephano recte positi sunt.” Additionally, GA 5 (Stephanus δ′) is also present in Wettstein’s apparatus, just as in Stephanus’ margins. 88 “We have examined the whole codex afresh, with the greatest possible care” (“Nos totum codicem denuo excussimus, et summa qua potuimus cura”), cited from Johann Jakob Griesbach, Symbolae criticae ad supplendas et corrigendas variarum N. T. lectionum collectiones. Accedit multorum N. T. codicum Graecorum descriptio et examen. Tomus prior (Halle: Curtius, 1785), lxxii. Its context may be of interest: this sentence occurs subsequent to a citation from Wettstein’s admission of his hasty collation. 89 To my knowledge, this edition (Novum Testamentum Graece. Textum ad fidem codicum versionum et patrum recensuit et lectionis varietatem adiecit…Volumen I. IV Evangelia complectens. Editio secunda emendatior multoque locupletior [Halle: Curtius, 1796]) contains 550 variant readings concerning L019 in John, including 168 “new” ones. Interestingly,

Stephanus’ Greek New Testament of 1550

327

examining the codex itself, Griesbach was able to correct many errors in previous editions.90 Yet, his critical apparatus was still not a perfect product: at least ten errors are preserved concerning Codex Regius in John, seven of which also appear in Wettstein’s 1751 edition.91 These pieces of imprecise information could have been due to the way in which Griesbach collated Codex Regius. It is likely that he made his collation on the basis of Wettstein’s critical apparatus. Therefore, except for those he had already corrected during his examination of the codex, Griesbach was unable to detect any other errors that Wettstein had made.92 This scenario explains why the two-hundred-year-old phantom reading ὑµεῖς τὰ κλήµατα at John 15:1, first “attested” by Stephanus, is still retained in Griesbach’s edition, an edition that was based on first-hand examination of the codex. To summarise: no one after Henri Stephanus examined Codex Regius for textcritical use for over one and a half centuries. Thus, every editor of the New Testament who wanted access to the codex needed to consult the critical apparatus in Stephanus’ 1550 edition. Even during the eighteenth century when the codex became available again, a number of errors from Stephanus still found their way into the rigorous editions prepared by Wettstein and Griesbach. Inaccuracies persist from edition to editions and are transmitted from one edition to another (see Figure 5). It was until the publication of a whole transcription of Codex Regius in 184693 that these transmitted errors were eventually corrected. With this transcription in hand, scholars were finally able to examine the text of the codex without visiting Paris.

approximately one hundred readings recorded by Wettstein find no place in the critical apparatus of Griesbach’s edition. 90 Note that lengthy remarks by Griesbach on Wettstein’s critical apparatus can be found in the first volume of Symbolae criticae. On the basis of these corrections, the error rate of Griesbach’s Novum Testamentum Graece is extremely low: in John more than 98% of the variants on L019 are correct, according to my examination. 91 These imprecise variants are at John 4:2; 6:23; 8:49; 12:31; 14:7; 15:1; 19:13. 92 It is necessary to take the issue of accessibility into account, especially before the reproduction of a given manuscripts appeared. For Griesbach had to decide how to manage the limited time during his stay in Paris. To use Wettstein’s extensive list of variants as the basis was undoubtedly a workable solution, and perhaps even the best way that Griesbach could have conjured. 93 Constantin von Tischendorf, Monumenta sacra inedita sive reliquiae antiquissimae textus Novi Testamenti graeci ex novem plus mille annorum codibus per Europam dispersis eruit atque edidit… (Leipzig: Tauchnitz, 1846), 15*–24* (“Prolegomena”) and 57–400; our codex is entitled “Codex Parisiensis Regius NR. LXII. (Evangeliorum L.).” Tischendorf’s transcription is actually a “textual facsimile,” that is, not only the text but also the layout (segmentation, punctuation, ekthesis, and the Eusebian Apparatus, etc.) are reproduced as precisely as possible.

328

An-Ting Yi

Regius H. Stephanus 1546 R. Stephanus 1550 Beza 1556-1598 Walton 1657 Mill 1707

Wettstein 1715 Wettstein 1751 Griesbach 1785 Griesbach 1777/1796

1500

1550

1600

1650

1700

1750

1800

Fig. 5. The use of Codex Regius between the sixteenth and the eighteenth centuries

Concluding Remarks The preceding discussion examined the 1550 Greek New Testament edition of Robert Stephanus, the first critical user of Codex Regius in the history of New Testament textual scholarship. By investigating his historical context I argued that, for the Royal Printer Stephanus, although it contains considerable text-critical elements, the critical apparatus of the 1550 edition primarily served to express his gratitude to the French King. Moreover, the examination of the variants regarding Codex Regius in John enables us to evaluate both the quality and the quantity of the apparatus. In particular, by tracing those variants that are inaccurate representations, this examination also provides crucial evidence for knowing of the way in which the codex was actually collated. We are able to reconstruct, in part, the process by which Henri’s collations made their way into Robert’s edition. Furthermore, this study further explores how Regius’ readings were used – through Stephanus’ apparatus – by a number of later editors of the Greek New Testament. Such an exploration reveals a unique history of the reception of Codex Regius between the sixteenth and the eighteenth centuries. It could even be said, in a metaphorical way, that through an incomplete replication made by

Stephanus’ Greek New Testament of 1550

329

Robert Stephanus in 1550, Codex Regius did travel around Europe over the next two centuries. Editions are important arbiters of the Greek text of the New Testament. Altogether, this study contributes to the current “historical turn” in textual scholarship. Just as the reorientation towards manuscripts has brought insights into the history of early Christianity, explorations of historical editions also uncover many significant parts within the history of New Testament textual scholarship. From this historical perspective, variants in the critical apparatus of given editions are not merely trivial pieces of information. On the contrary, they are windows that allow us to glimpse the editors’ historical contexts, their ways of collecting data, their judgements, and even hesitations.94 More than a century ago Hoskier mentioned the burdensome labour for collating New Testament manuscripts, an unavoidable praxis for textual critics beginning with Henri Stephanus. Given that the working environment was usually limited to a specific reading room in libraries within a restricted period of time, such tasks were indeed laborious. However, thanks to the digital age, now examinations of manuscripts have become much easier and more convenient through high-resolution digital images that have been made available online.95 Furthermore, the rapidly growing digitisation of rare books in the past decade makes the present study possible: researchers no longer need to travel around the world to consult historical scholarly works; the digital form of almost all the works can now be reached through the internet. An ever-growing body of material is being continually enhanced by new digitisation projects. By means of these digital supports, the modern scholarly community is now equipped to initiate more historical studies based on data-driven analysis that I have tried to model. For instance, similar explorations could focus on other manuscripts, in particular those with extensive use in their reception history.96 By exploring the use of a given manuscript, the massive jungle of critical apparatuses in the printed editions – one of the most complicated yet often neglected subjects – can eventually be elucidated.

94 The metaphor of “windows” is of course inspired by Bart Ehrman, who argues that manuscript texts can serve as windows to discover the social world of the scribes who prepared those manuscripts; see Ehrman, “The Text as Window: New Testament Manuscripts and the Social History of Early Christianity,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. B. D. Ehrman and M. W. Holmes (Grand Rapids: Eerdmans, 1995), 361–79, republished in updated version in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 803–30. 95 The digital images of Codex Regius were put online by the Bibliothèque nationale de France on 30 December 2013. They can be explored now via http://gallica. bnf.fr/ark:/12148/btv1b53045883v. 96 Potential candidates are A02 B03 C04 and D05 to name a few.

330

An-Ting Yi

In the light of the vast amount of materials we currently have access to, the future directions of New Testament textual scholarship should include the attempts to understand our past in a better way. Indeed, the time is now ripe for us to start tracing, describing, and narrating many undiscovered but notable stories in the history of our field.

Appendix: Variant Readings of Codex Regius in the Gospel of John in Stephanus’ Edition The following table is a complete list of the variant readings concerning η′ (including π.) in John in Robert Stephanus’ Greek New Testament edition of 1550. Except for those correctly represented variants (category 1), incorrect information is filled in grey and its verse number is put in bold with the following categorisation: imprecise variants (category 2); inaccurate attribution variants (category 3); and other wrong variants (category 4). No.

John

Text in St.

Variant of η′ in St.

Remarks

1

1:28

Βηθαβαρᾶ

Βηθανίᾳ

St. reads Βηθαβαρᾶ instead of Βηθαβαρᾷ

2

1:39

ἴδετε

ὄψεσθε

L019: ε > αι

3

1:39

δέ

omit

4

1:42

Ἰωνᾶ

Ἰωάννου

5

2:15

ποιήσας

add ὡς

6

2:17

κατέφαγε

καταφάγεται

7

2:22

αὐτοῖς

omit

8

3:25

Ἰουδαίων

Ἰουδαίου

9

4:25

Οἶδα

οἴδαµεν

10

4:35

ἔτι

omit

11

4:35

τετραµηνόν

τετράµηνος

12

5:5

ἀσθενείᾳ

add αὐτοῦ

13

5:10

οὐκ

add καί

14

5:16

καὶ ἐζήτουν αὐτὸν ἀποκτεῖναι

omit

15

5:30

πατρός

omit

16

6:11

τοῖς µαθηταῖς, οἱ δὲ µαθηταί

omit

17

6:15

πάλιν

omit

18

6:17

οὐκ

οὔπω

19

6:22

ἐκεῖνο εἰς ὃ ἐνέβησαν οἱ µαθηταὶ αὐτοῦ,

omit

L019: αι > ε

category 3

331

Stephanus’ Greek New Testament of 1550 No.

John

Text in St.

Variant of η′ in St.

20

6:24

καί

omit

21

6:28

ποιοῦµεν

ποιῶµεν

22

6:40

δέ ἐστι

γάρ ἐστι

L019: moveable -ν

23

6:45

ἀκούσας

ἀκούων

St. correctly indicates that L019 reads as his text.97

24

6:55

ἀληθῶς

ἀληθής

25

6:55

ἀληθῶς

ἀληθής

26

6:63

λαλῶ

λελάληκα

27

6:69

τοῦ ζῶντος

omit

28

7:8

ταύτην

omit

29

7:26

ἀληθῶς

omit

30

7:31

τούτων

omit

31

7:33

αῦτοῖς

omit

32

7:39

πισεύοντες

πισεύσαντες

33

7:40

τὸν λόγον

των λόγων τούτων

34

8:9

καὶ ὑπὸ τῆς συνειδήσεως ἐλεγχόµενοι

omit

35

8:9

ἑστῶσα

οὖσα

36

8:29

οὐκ

add καί

37

8:38

ὃ ἑωράκατε παρὰ τῷ πατρί

ἃ ἠκούσατε παρὰ τοῦ πατρός

L019: ε > αι

38

8:49

ἀτιµάζετε

ἠτιµάσατε

category 4

39

8:53

σύ

omit

40

9:8

τυφλός

προσαίτης

41

9:9

Ἄλλοι δέ, Ὅτι ὅµοιος

ἄλλοι ἔλεγον, Οὐχὶ, ἀλλ’ ὅµοιος

42

9:17

Λέγουσι

add οὖν

43

10:14

γιὼσκουµαι ὑπὸ τῶν ἐµῶν

γινώσκουσί µε τὰ ἐµά

44

10:34

Ἐγώ

add ὅτι

45

11:19

καὶ πολλοί

πολλοὶ δέ

46

11:19

τὰς περὶ Μάρθαν

τὴν Μάρθαν

47

11:31

λέγοντες

δόξαντες

48

11:41

οὗ ἦν ὁ τεθνηκὼς κείµενος

omit

49

11:54

Ἐφραΐµ

Ἐφρέµ

50

12:1

νεκρῶν

add ὁ Ἰησοῦς

97

Remarks

category 3

In the margins we read πάντα πλὴν τοῦ γ, καὶ τοῦ η (all except γ′ and η′).

332 No.

An-Ting Yi John

Text in St.

Variant of η′ in St.

51

12:7

εἰς τὴν ἡµέραν τοῦ ἐνταφιασµοῦ µου τετήρηκεν

ἵνα εἰς τὴν ἡµέραν τοῦ ἐνταφιασµοῦ µου τηρήσῃ

52

12:17

ὅτε

ὅτι

53

12:19

κόσµος

ὅλος

54

12:21

οὖν

omit

55

12:31

τούτου

omit

56

12:35

µεθ

ἐν ὑµῖν

57

13:24

πυθέσθαι

καὶ λέγει αυτῶ, εἰπὲ τίς ἐστι

58

13:27

τότε

omit

59

13:30– 31

νὺξ ὅτε ἐξῆλθε. Λέγει

νύξ. Ὅτε οὖν ἐξῆλθε, λέγει

L019 has a paragraphus-like symbol (⸭) here.98

60

14:22

τί

add καί

category 3

61

14:30

τούτου

omit

62

15:1

ἀληθινή

add ὑµεῖς τὰ κλήµατα

63

15:7

αἰτήσεσθε

αἰτήσασθε

64

15:8

γενήσεσθε

γενήσθε

L019: ε > αι

65

16:3

ὑµῖν

omit

category 3

66

16:16

οὐ

οὐκέτι

67

17:11

οὕς

ᾧ

68

17:12

σου· οὓς δέδωκάς µοι ἐφύλαξα

σου, ᾧ δέδωκάς µοι καὶ ἐφύλαξα

69

17:20

πιστευσόντων

πιστευόντων

70

18:29

αὐτούς

add ἔξω

71

19:3

καί

add καὶ ἤρχοντο πρὸς αὐτόν

72

19:13

τοῦτον τὸν λόγον

τούτων τῶν λόγων

category 2

73

20:11

τὸ µνηµεῖον

τῷ µνηµείῳ

L019: ει > ι

74

20:24

ἦν

add ἐκεῖ

75

20:31

ζωήν

add αἰώνιον

76

21:6

ἴσχυσαν

ἴσχυον

77

21:15

Ἰωνᾶ

Ἰωαννᾶ

category 2

78

21:16

Ἰωνᾶ

Ἰωαννᾶ

category 2

79

21:17

Ἰωνᾶ

Ἰωαννᾶ

category 2

98 99

F. 238v, b18; this symbol appears to be equivalent to a full stop. F. 246v, a18–20: σου | ω δεδωκας µοι ⸭ | και εφυλαξα.

Remarks

category 3 L019: moveable -ν

category 3

The reading is correct, yet L019 has different punctuation.99 category 2

L019: ι > η

Methods of Evaluating Textual Relationships: From Bengel to the CBGM and Beyond Tommy Wasserman For nearly 300 years, the category of text types has prevailed as a means of grouping and evaluating the vast number of textual witnesses to the New Testament. This is rapidly changing. Although the theory of text types is still alive and well in current text-critical practice, some scholars have recently called to abandon the concept altogether in light of new computer-assisted methods for determining textual relationships in a more exact way. To be sure, there is already a consensus that the various geographic locations traditionally assigned to the text types are incorrect and misleading, and some scholars prefer to refer to the text types as “textual clusters.” This chapter gives an historical overview of the different methods of evaluating textual relationships, from the time when scholars began to categorize witnesses into different text types to the most recent Coherence-Based Genealogical Method (CBGM), used in contemporary practice to construct the major critical editions of the Greek New Testament.1 The final part of the chapter describes how the CBGM is changing the discipline and discusses limitations and future improvements of the method.

1. The Birth of the Text Types Over three hundred years ago, in 1707, John Mill published an edition of the New Testament in its original Greek language that listed some 30,000 textual differences between the surviving manuscripts that informed his edition.2 The existence of variations and incongruities among the multiple sources for the New

1 For reasons of space I have omitted methods of evaluating papyri from this survey. See, however, Lonnie Bell, The Early Textual Transmission of John: Stability and Fluidity in its Second and Third Century Greek Manuscripts, NTTSD 54 (Leiden: Brill, 2018), 1–17. 2 John Mill, Novum Testamentum Graecum, cum lectionibus variantibus MSS. exemplarium, versionum, editionum SS. patrum et scriptorum ecclesiasticorum, et in easdem notis (Oxford, 1707).

334

Tommy Wasserman

Testament would drive a lively scholarly interest in determining the ultimate, most authentic, most accurate text of the Greek New Testament. In light of the ever-increasing new textual evidence, J. A. Bengel (1687–1752) soon took on the task to sort out the mass of materials being made available – Greek manuscripts, ancient translations, and citations of the New Testament by church fathers. Since the relationship between individual textual witnesses was (and is) so complex, Bengel proposed that the text be studied in terms of textual groupings and divided the textual witnesses into “families, tribes and nations.”3 At first he assigned manuscripts with shared readings to three recensional texts originating from Alexandria-Egypt (associated with Hesychius); Constantinople-Antioch (associated with Lucian); and Jerusalem-Palestine (associated with Origen, Pamphilus, and Eusebius). In addition, he assumed that Rome and the West had their Latin text translated from Greek manuscripts that were more ancient than these three recensions.4 Later in the eighteenth century, Bengel’s scheme was refined in turn by J. S. Semler and J. J. Griesbach, who connected the text types (or groupings) to geographical locales where the text might have been revised at some point (apart from the normal process of copying): Alexandrian (derived from Origen), Western (Latin versions and fathers), and Eastern (used by Antiochian and Constantinopolitan churches). Griesbach added Codex Vaticanus, as well as other Alexandrian fathers to the Alexandrian text type.5 The two Cambridge professors B. F. Westcott and F. J. A. Hort preferred to label the earliest stage of the Alexandrian text type as a “Neutral text” (i.e., unchanged), which successively became more corrupt and thus labelled “Alexandrian text.” The “Neutral text,” chiefly represented by the fourth-century Codices Vaticanus (B03) and Sinaiticus (ℵ01), formed the basis of their ground-breaking edition of The New Testament in the Original Greek (1881).6 3 J. A. Bengel, Apparatus Criticus ad Novum Testamentum, 2nd ed. (Tübingen: J. G. Cotta, 1763 [1734]), 20 (Introduction, §29): “Posset variarum lectionum ortus, per singulos codices, per paria codicum, per syzygias minores majoresque, per familias, tribus, nationesque illorum, investigari et repraesentari.” 4 Bengel, Apparatus Criticus (Introduction §32). Subsequently, however, Bengel reduced the text types to two main groups: the “African” and the “Asiatic,” with various subgroups (p. 669). 5 Eldon J. Epp, “Textual Clusters: Their Past and Future in New Testament Textual Criticism,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 528– 31. 6 Brooke Foss Westcott, and Fenton John Anthony Hort, eds., The New Testament in the Original Greek, 2 vols. (London: Macmillan, 1881). This edition – which in Westcott and Hort’s view now represented the most accurate and authentic version of the New Testament in the original language – furnished the death blow to the Textus Receptus, which had dominated Greek editions and, indirectly, Bible translations (e.g., the KJV) ever since Erasmus published the first edition of the Greek New Testament in 1516.

Methods of Evaluating Textual Relationships

335

Westcott and Hort viewed ℵ01 as very close to B03, but clearly gave priority to the latter: “B must be regarded as having preserved not only a very ancient text, but a very pure line of very ancient text.”7 Further, they proposed that at the beginning of the third century, the three textual streams, Neutral, Alexandrian, and Western, flowed into what became the Syrian text later labelled “Byzantine,” since it came to dominate the Eastern Church. This text type, which they considered inferior, is reflected in the vast majority of manuscripts (hence the “Majority Text”). It did not take long before Westcott and Hort were criticised on various points. In 1897, George Salmon (Trinity College Dublin) suggested that the name “Neutral” be replaced by the term “Alexandrian” in its traditional sense as coined by Griesbach, which was not as pretentious as “Neutral” – a term based on a judgment of its pure textual character.8 But a name founded on the quality of the text presupposes that the text has been examined, and a decision pronounced on it which those who adopt the nomenclature cannot consistently reverse. The name “neutral” presupposes the establishment of WH’s theory that all additions and alterations in this neutral text are due to later corruptions. But little mischief would have been done by the substitution of a new name for the old name “Alexandrian,” if Hort had not, in order to shield his “neutral” text from the danger of being dishonoured by the epithet “Alexandrian,” appropriated that title to another family of readings.9

2. Hermann von Soden: Gathering and Classifying the Manuscripts During the years between 1902–1913, Hermann von Soden’s monumental fourvolume text-critical work, Die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt hergestellt auf Grund ihrer Textgeschichte, appeared.10 Von Soden and his colleagues had collated numerous Greek manuscripts, which were presented in the edition with comments on various readings. Moreover, he Westcott and Hort, New Testament, 2.251. Already Constantin Tischendorf’s monumental eighth edition had favoured ℵ-Β, but with a strong bias in favour of the former, owing, no doubt, to his discovery of the manuscript in St Catherine’s Monastery and the hype that accompanied its appearance to the world. See Eldon J. Epp, “The Late Constantin Tischendorf and Codex Sinaiticus,” in Studies on the Text of the New Testament and Early Christianity: Essays in Honor of Michael W. Holmes, NTTSD 50, ed. D. M. Gurtner, J. Hernández Jr., and P. Foster (Leiden: Brill, 2015), 47. 8 George Salmon, Some Thoughts on the Textual Criticism of the New Testament (London: Murray, 1897), 48–55. 9 Salmon, Some Thoughts, 49–50. 10 Hermann von Soden, Die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt hergestellt auf Grund ihrer Textgeschichte, 2 parts in 4 vols., 2nd unchanged ed. (Göttingen: Vandenhoeck & Ruprecht, 1911–1913). 7

336

Tommy Wasserman

classified nearly all Greek manuscripts known at the time (including some 1260 minuscules) according to three text types associated with certain geographical locations: the Hesychian (H – Alexandrian or Egyptian) represented by ‫ א‬and B; the Jerusalemite (I – Palestine), represented by D, the Old Latin version and the Syriac Sinaiticus; and the Koiné (K – Antiochian or Syrian) represented by K and Π. Thus, he assigned each witness by collating them against these major texts in selected, but mostly unidentified passages (Stichkapitel). However, he wanted to go further and determine more precise textual relationships, by subdividing the two latter groups. As Bart D. Ehrman observes, von Soden did not sufficiently account for his methodology, but, nevertheless, three factors can be ascertained, on which basis he made his classification: (1) the general affinities of text as determined by collations; (2) the forms of the pericope adulterae (seven forms µ1–7); and (3) the textual “equipment” of the manuscripts (Eusebian canons, Euthalian apparatus, etc.).11 On the whole, von Soden’s venture proved to be a failure in several respects: first, his theoretical presuppositions regarding the text types proved to be false, as did many of his manuscript classifications, especially in regard to his so called I-text.12 Secondly, his very complicated system of symbols representing manuscripts, as well as his divided apparatuses have provoked much misunderstanding.13 Finally, the information in the apparatus is often incomplete and

Bart D. Ehrman, “Methodological Developments in the Analysis and Classification of New Testament Documentary Evidence,” NovT 29/1 (1987): 30. 12 In regard to Jude, Sakae Kubo tested von Soden’s classification by using more recent methodology (see further below). Kubo analysed 37 manuscripts, including 31 that had been classified by von Soden, and concluded concerning the purported I-text, “the large number of manuscripts classified under I by von Soden cannot be distinguished from the manuscripts of the K text” (“Textual relationships in Jude,” in Studies in New Testament Language and Text, NTSup 44, ed. J. K. Elliott [Leiden: Brill, 1976], 280). Cf. J. Tim Gallagher, “A Study of von Soden’s H-Text in the Catholic Epistles,” AUSS 8 (1970): 97–119; and W. L. Richards, “A Critique of a New Testament Text-Critical Methodology: the Claremont Profile Method,” JBL 96 (1977): 555. 13 Indeed von Soden’s edition is almost impossible to use without keys and transposition tables, such as are available in Friedrich Krüger, Schlüssel zu von Soden’s die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt hergestellt (Göttingen: Vandenhoeck & Ruprecht, 1927), and in Kurt Aland, Michael Welte, Beate Köster, and Klaus Junack, eds., Kurzgefasste Liste der griechischen Handschriften des Neuen Testaments, ANTF 1, 2nd rev. and enl. ed. (Berlin: De Gruyter, 1994), 390–427. Von Soden’s own notes on format and apparatus are not sufficient. A brief introduction to the edition is available in Kurt Aland and Barbara Aland, The Text of the New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism, trans. E. F. Rhodes; 2nd rev. and enl. ed. (Grand Rapids: Eerdmans, 1995), 40–3. 11

Methods of Evaluating Textual Relationships

337

unreliable, so that the attestation of its readings needs constant verification elsewhere.14 By the time von Soden’s magnum opus was published, the British text critic Herman C. Hoskier had achieved a reputation as “an indefatigable and scrupulously accurate collator of manuscripts and Versions” and was one of few in the position to review von Soden’s work, which he had consulted for his own examination of ‫ א‬and B to which we will turn below.15 The devastating review article that appeared in Journal of Theological Studies was published just after the news of von Soden’s sudden death in a traffic accident, so that one of the editors of the journal, William Sanday, felt compelled to write an apologetic two-page preface to the review.16 Hoskier opened his review by going straight to the point: I had looked for great things from Dr von Soden’s final volume of the Text. The earlier volumes were very heavy reading, but I expected that his Text and critical notes would fill a gap in our studies. Alas, he has but complicated our problems, and instead of writing a eulogy on his work I regret to have to condemn it strongly…the apparatus is positively honeycombed with errors, and many documents which should have been recollated have not been touched, others only partially, and others again have been incorrectly handled.17

After nineteen pages of severe criticism, Hoskier concluded his scathing review by saying, “But to state these matters is only to make a partial impression on my readers of the grievous state of things in this latest book on a most intricate subject. Es ist zum Weinen.”18 In spite of this and other negative reviews, von Soden is counted among the giants of New Testament textual criticism, being a pioneer in collecting the evidence from a large number of hitherto unknown manuscripts, especially minuscules. Moreover, remarkably many of von Soden’s genealogical classifications in sub-groups of the Byzantine text (his Koiné) have actually been confirmed by more recent methods, including the Claremont Profile Method (see below).19 A quite illustrative example is C. A. Albin’s report of 11 errors in von Soden’s account for Codex Ψ (044) in Jude alone. See Albin, Judasbrevet: Traditionen, Texten, Tolkningen (Stockholm: Natur och Kultur, 1962), 275 n.345. 15 Henry St. John Thackeray, review of Codex B and its Allies: A Study and an Indictment (by Herman C. Hoskier), The Journal of Hellenic Studies 35 (1915): 275–76. By this time, Hoskier had published A Full Account and Collation of the Greek Cursive Evangelium 604 (London: David Nutt, 1890); The Golden Latin Gospels: JP in the Library of J. Pierpont Morgan (New York: Private Printing, 1910); and “Evan. 157 (Rome. Vat. Urb. 2),” JTS 14 (1912–1913): 78–116; 242–93; 359–84. 16 H. C. Hoskier, review of Die Schriften des Neuen Testaments (by Hermann von Soden), JTS 15 (1914): 307–26 (William Sanday’s preface, “Baron Hermann von Soden,” 305–306). 17 Hoskier, review of Die Schriften, 307. 18 Hoskier, review of Die Schriften, 326. 19 Eldon J. Epp offers a brief description and a summary of the results utilizing the Claremont Profile Method (CPM) for the IGNTP project and concludes: “On the other hand, many of von Soden’s groupings and classifications have been strikingly confirmed, 14

338

Tommy Wasserman

3. Herman Hoskier’s Indictment against Codex B and Its Allies Hoskier was also one of the loudest critics of Westcott and Hort in the twentieth century. In 1914 he set out to show that the “Neutral text” of Westcott and Hort was neither free from corruption, nor homogenous, presenting an “indictment” against Codex B and its allies.20 The indictment was just as much against Hort – if not more so – as evident from the preface: It is high time that the bubble of codex B should be pricked. It had not occurred to me to write what follows until recently. I had thought that time would cure the extraordinary Hortian heresy, but when I found that after a silence of twenty years my suggestion that Hort’s theories were disallowed today only provoked a denial from a scholar and a critic who has himself disavowed a considerable part of the readings favoured by Hort it seemed time to write a consecutive account of the crooked path pursued by the MS B, which – from ignorance I trow – most people still confuse with purity and “neutrality.” I proceed to “name” the aforesaid scholar, since he has challenged me.21

Hoskier wanted to prove that Vaticanus and its relatives were inferior to the Byzantine text as represented in the Textus Receptus, and that the latter had largely formed the basis which Vaticanus had tampered with. After an extensive cross examination of Vaticanus the attorney demanded a verdict: The verdict asked is whether B represents a “neutral” text or not. The claims put forward by us are that B does not exhibit a “neutral” text, but is found to be tinged, as are most other documents, with Coptic, Latin and Syrian colours, and its testimony therefore is not of the paramount importance presupposed and claimed by Hort and by his followers. That B is guilty of laches, of a tendency to “improve,” and of “sunstroke” amounting to doctrinal bias. That the maligned Textus Receptus served in large measure as the base which B tampered with and changed, and that the Church at large recognized all this until the year 1881 – when Hortism (in other words Alexandrism) was allowed free play – and has not since retraced the path to sound traditions.22

Hoskier was unable to prove his theory that the text of Codex Vaticanus, in particular, had been influenced by the versions through recension and bilingual

including designations of groups, subgroups and individual MSS” (“The Claremont Profile Method for Grouping New Testament Minuscule Manuscripts,” in Studies in the Theory and Method of New Testament Textual Criticism, ed. E. J. Epp and G. D. Fee [Grand Rapids: Eerdmans, 1993], 220). For a brief description and evaluation of the Claremont Profile Method, see Thomas C. Geer Jr. and Jean-François Racine, “Analyzing and Categorizing New Testament Greek Manuscripts,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 502–505. 20 H. C. Hoskier, Codex B and Its Allies: A Study and an Indictment, 2 vols. (London: Bernard Quaritch, 1910–1911). 21 Hoskier, Codex B, 1.i. 22 Hoskier, Codex B, 1.465.

Methods of Evaluating Textual Relationships

339

ancestors.23 Further, as one reviewer remarked, he alienated critics and showed an “attitude of scathing contempt which he takes up towards his distinguished predecessors and co-workers in the same field.”24 On the other hand, by doing what he was good at – collating Greek manuscripts – he was able to tabulate no less than 3,036 differences between Vaticanus and Sinaiticus in the Gospels apart from minor errors (e.g., spelling differences), from which he concluded, “In the light of the following huge lists let us never be told in future that either ‫ א‬or B represents any form of “Neutral text.”25 Three thousand differences may sound a huge number, but Hoskier did not provide a corresponding number of places where the manuscripts agree against other manuscripts where there is textual variation. This would require large collations. Now, with the Editio Critica Maior (ECM), we have access to such large collations – not yet in the Gospels but in other parts of the New Testament – and with the accompanying computer software tools we are able to put these figures into a larger perspective.26 Over two hundred manuscripts were collated for the ECM of the Catholic Letters wherever they were extant, which resulted in over three thousand places of textual variation in this corpus.27 A collation of Vaticanus against Sinaiticus in 3,003 places of comparison shows that they have the same text in 2,613 variation units (87%), and, thus, differ in 390 variation units (13% disagreement). This is a purely quantitative measurement. When the ECM editors examined these 390 places further in the process, they decided that in 250 cases Vaticanus had the prior reading, in 89 cases Sinaiticus had the prior reading, and in 51 places the relationship was unclear (these decisions are related to Hoskier stated his charge at the outset, “And I charge B with being the child of a Graeco-Latin recension, and by its scribe or by its parent of being tremendously influenced by a Coptic recension or by a Greco-sahidic and/or Greco-bohairic MS” (Hoskier, Codex B, 1.7). 24 In his review of Hoskier’s two-volume monograph, Thackeray pointed out several problems, the most serious being Hoskier’s terrible attitude to other scholars: “Above all, he alienates critics and mars his work by the attitude of scathing contempt which he takes up towards his distinguished predecessors and co-workers in the same field. Previous critics of the Westcott-Hort text, like the late Dr. Salmon, have ventured with diffidence to disagree with the conclusions of those master-workers. Mr. Hoskier has no such scruples and outBurgons Burgon in his withering ‘indictment.’ ‘This old bosh about a “Syrian” text!’ (i. 270) ‘this “Q” business’ (i. 41, note), in such terms does he refer to theories which have now gained wide acceptance” (Thackeray, review of Codex B and its Allies, 275). 25 Hoskier, Codex B, 2.1. 26 Currently, work on the ECM of the Gospel of Mark is being done by the Institut für neutestamentliche Textforschung in Münster, but no data have been published yet. 27 The number of variant passages in the total corpus is 3,043 in the latest revision of the ECM. The number of included manuscripts in the published ECM varies between different books (James: 186; 1–2 Peter: 147; 1 John: 145; 2–3 John and Jude: 151). See Barbara Aland et al., eds., Novum Testamentum Graecum: Editio Critica Maior IV; Catholic Letters; Part 2: Supplementary Material, 2nd ed. (Stuttgart: Deutsche Bibelgesellschaft, 2013), 5–9. 23

340

Tommy Wasserman

“genealogical coherence” – see below). Based on these decisions in local variation-units, Vaticanus turned out to be the closest “potential ancestor” of Sinaiticus. If we assume, for the sake of convenience, the same rate of variation (13%) per number of words in the Gospels, then we can get a better idea of the significance of the three thousand differences between ‫ א‬or B that Hoskier accounted for. In NA28 there are 7,591 words in the Catholic Letters and 64,767 words in the Gospels, which means that 390 differences in 3,003 variation units corresponds to 3,328 differences in 25,622 variation units the Gospels.28 This means that the two manuscripts are probably equally close in the Gospels (around 88%). In the “Potential ancestors and descendants” tool it is possible to look for the closest relatives of any included manuscript. Codex Vaticanus is the closest relative (and potential ancestor) of Codex Sinaiticus in the Catholic Letters (87.1% agreement).29 If, on the other hand, we examine the closest relatives of Vaticanus we find twenty-four manuscripts which are closer than Codex Sinaiticus, with Codex Ephraemi Rescriptus (04) at the top with 89.4% agreement. These levels of agreement may be higher than Hoskier had anticipated, but in light of all the results, they do not prove the existence of the Alexandrian text type as a distinct group of manuscript which can easily be separated from other groups. For example, the Byzantine uncial 020 shows 84.5% agreement with Sinaiticus in the Catholic Letters. Nevertheless, Hoskier’s critique showed that the representatives of the “Neutral” text were not unaffected by textual changes, and the name “Neutral” was soon replaced by “Alexandrian,” as Salmon had suggested, so that Bengel’s original division into three major groups, or “text types” to use a later term, was reinstated. A final word about Hoskier’s view of the history of the text is in place since he has been regarded by many as a champion of the Byzantine Majority Text, and, as such, the heir of John W. Burgon who defended the Majority Text until his death in 1888.30 However, as Daniel Wallace has demonstrated, Hoskier did

28 In this connection, it is necessary to distinguish between variation units and the number of variants (one unit of variation has two variant readings or more). In my own edition of Jude, I counted over thirty readings in Jude 23a, which is an extreme case. According to the best estimate (based on a sound method), there are 591,044 variant readings in the whole New Testament. See Peter J. Gurry, “The Number of Variants in the Greek New Testament: A Proposed Estimate,” NTS 62/1 (2016): 111. 29 The “Potential ancestors and descendants” tool is found at http://intf.uni-muenster.de/cbgm2/PotAnc5.html. 30 See Edward Meyrick Goulburn, John William Burgon, Late Dean of Chichester: A Biography with Extracts from his Letters and Early Journals, 2 vols. (London: John Murray, 1892).

Methods of Evaluating Textual Relationships

341

not espouse all the views of his friend Burgon.31 Wallace mentions five points of differences between the views of Hoskier and Burgon: [Hoskier] (1) embraced the internal canons of the harder and shorter readings; (2) adopted readings found in a small minority of MSS, especially of the Western strain; (3) accepted the probability of a Lucianic recension behind the Byzantine text; (4) thought that the Majority of MSS – not just ℵ and B – were coloured by versional tamperings; (5) explicitly argued against the value of the Byzantine cursives; (6) believed that many portions of the NT were originally written in Aramaic.32

4. The Battle over the Papyri The criticism against Westcott and Hort’s “Neutral” text continued in the decades that followed. In particular, new papyrus discoveries from the 1930s onward caused many scholars to further question whether the “Neutral” text really represented a pure line of transmission from the earliest time, as Westcott and Hort had assumed. Some papyri did not align clearly with any of the established text types and, thus, reflected a more diverse and fluid state of transmission than expected. Therefore, Frederic G. Kenyon suggested that the “Neutral text” of Codex Vaticanus must be the product of a scholarly recension that probably took place in Alexandria in the fourth century.33 However, with the discovery and publication of P75 in 1961 the question of an Alexandrian recension came into a new perspective. The first editors, Victor Martin and Rodolphe Kasser, assigned it a date between 175 and 225 CE based on a palaeographic assessment.34 Subsequent studies of P75 in Luke by Carlo M. Martini and in John by Calvin L. Porter demonstrated that the text of P75 was

31 Daniel B. Wallace, “Historical Revisionism and the Majority Text: The Cases of F. H. A. Scrivener and Herman C. Hoskier,” NTS 41 (1995): 284–85. 32 Wallace, “Historical Revisionism,” 285. Cf. Wilbur Pickering, The Identity of the New Testament Text, rev. ed. (Nashville: Thomas Nelson, 1980), 30, 106. 33 Frederic G. Kenyon, “Hesychius and the Text of the New Testament,” in Mémorial Lagrange, ed. L.-H. Vincent (Paris: J. Gabalda, 1940), 250. Cf. Kenneth W. Clark, “The Effect of Recent Textual Criticism upon New Testament Studies,” in The Background of the New Testament and Its Eschatology, ed. W. D. Davies and D. Daube (Cambridge: Cambridge University Press, 1954), 37; Günther Zuntz, The Text of the Epistles: A Disquisition upon the Corpus Paulinum. The Schweich Lectures of the British Academy, 1946 (Oxford: Oxford University Press, 1953), 271–72; Werner G. Kümmel, Introduction to the New Testament, 14 th rev. ed. of P. Feine and J. Behm (Nashville: Abingdon, 1966), 384; Bruce M. Metzger, The Text of the New Testament: Its Transmission, Corruption and Restoration (Oxford: Clarendon, 1964), 215–16. 34 Victor Martin and Rodolphe Kasser, Papyrus Bodmer XIV: Évangile de Luc chap. 3– 24 (Geneva: Bibliotheca Bodmeriana, 1961); and Martin and Kasser, Papyrus Bodmer XV: Évangile de Jean chap. 1–15 (Geneva: Bibliotheca Bodmeriana, 1961).

342

Tommy Wasserman

almost identical to the text of Codex Vaticanus.35 The close relationship of P75 and Vaticanus demonstrated the stability of that kind of text during at least a century and a half in an era of textual transmission that was presumably uncontrolled. Thus, the central question whether the “Neutral” (or Alexandrian) text or text type is the result of a recension or of a strict transmission was pushed back into the second century.36 As Epp explains, “the long-standing conviction of a fourth-century recension of what had been called the B-text was freely given up – no struggle, no strife.”37 More recently, however, Brent Nongbri has attempted to re-open the case by challenging the accepted dating of P75 and, thus, the central place of the manuscript in the history of the text.38 As I have shown, elsewhere, however, P75 is not our only secure evidence that the kind of text reflected in Codex Vaticanus was circulating around 200 CE.39 In fact, P4 (Luke) and P64+67 (Matthew), dated to the second century by Pasquale Orsini and Willy Clarysse for good reasons – they belong to the initial phase of the biblical majuscule – attest to a text very close to Codex Vaticanus.40 In any case, the “battle over the papyri,” i.e., the rival assessments of their relative worth for the history of the New Testament text will likely continue.41

35 Calvin L. Porter, “A Textual Analysis of the Earliest Manuscripts of the Gospel of John” (PhD diss., Duke University, 1961); Porter, “Papyrus Bodmer XV (P75) and the Text of Codex Vaticanus,” JBL 81/4 (1962): 363–76; Carlo M. Martini, Il problema della recensionalità del codice B alla luca del papiro Bodmer XIV (Rome: Biblical Institute Press, 1966). 36 Eldon J. Epp, “The Twentieth Century Interlude in New Testament Textual Criticism,” JBL 93/3 (1974): 393. 37 Eldon J. Epp, “Decision Points in Past, Present, and Future New Testament Textual Criticism,” in Studies in the Theory and Method of New Testament Textual Criticism, ed. E. J. Epp and G. D. Fee (Grand Rapids: Eerdmans, 1993), 42. 38 Brent Nongbri, “Reconsidering the Place of Papyrus Bodmer XIV–XV (𝔓75) in the Textual Criticism of the New Testament,” JBL 135/2 (2016): 405–37. 39 Tommy Wasserman, “Was There an Alexandrian Recension of the Living Text of the Gospels?” in Liturgy and the Living Text of the New Testament, ed. H. A. G. Houghton (Piscataway: Gorgias, 2018), 1–22. 40 Pasquale Orsini and Willy Clarysse, “Early New Testament Manuscripts and Their Dates: A Critique of Theological Paleography,” ETL 88/4 (2012): 470. In their case study they state, “𝔓64+67+4 is written in a biblical majuscule belonging to the early phase of the canon. The writing angle is still uncertain, so that sometimes no shading is visible. This writing is similar to that of P. Vindob. G 29768 (late II–early III; LDAB 2761), as noted by Skeat, and may be attributed, therefore, to a period between the second and third centuries” (p. 461). See also Pasquale Orsini, Manoscritti in maiuscola biblica, Materiali per un aggiornamento (Cassino: Università di Cassino, 2005), 85–6, where P.Vindob. G 29784 is offered as an additional comparison (end of second century). 41 Epp, “Decision Points,” 43.

Methods of Evaluating Textual Relationships

343

5. Ernest C. Colwell and E. W. Tune: Quantitative Analysis In 1945 Bruce M. Metzger raised an important issue regarding the dominant method of evaluating textual relationships in regard to the so-called “Caesarean text”: Is it licit to reconstruct the ancient “Caesarean text” from ofttimes late documents merely by pooling the non-Byzantine variants?...the accepted method of determining the “Caesarean text” cannot but fail to discover all “Caesarean” readings, for certain “Caesarean” readings have undoubtedly passed into the Byzantine text and therefore are not disclosed when “Caesarean” manuscripts are collated against a Byzantine text. On the other hand, must it not be acknowledged that we have but Hobson’s choice in the matter; is any other more satisfactory method for determining the “Caesarean text” at our disposal?42

Metzger identified a deficiency in the method of establishing genealogical relationships between New Testament manuscripts that had prevailed ever since the time of Karl Lachmann (with von Soden as a notable exception) – the custom to compare any manuscripts by counting the number of common variants against an extrinsic standard, namely the Textus Receptus (“a Byzantine text”).43 This method of categorizing manuscripts went hand in hand with the custom to collate them against the “Standard text” in the first place.44 In light of the work of Westcott and Hort, however, it had become untenable to use the Textus Receptus as a standard text, but there was yet no alternative. In a series of ground-breaking articles published during the late 1950s and 42 Bruce M. Metzger, “The Caesarean Text of the Gospels,” JBL 64/4 (1945): 486–87. The methodology had been challenged as early as in 1814 by Archbishop Richard Laurence, who specifically criticised Griesbach’s classification of Codex Alexandrinus in the Pauline Letters. Later on, in the nineteenth century, Westcott and Hort concluded that since the “Syrian” (Byzantine) editors utilised the earlier types of text in their text, this text must contain both “Western” and “Neutral” elements; therefore, being a mixed text, it could function as a standard of textual comparison. However, their proposal for the detailed comparisons of all witnesses to one another had no impact on their contemporaries, in that it was connected with insuperable practical difficulties, reflected also in the fact that they themselves did not collate a single manuscript. See Ehrman, “Methodological Developments,” 26–9. 43 At first, the Textus Receptus was thought to adequately represent the original text of the NT, and departures from it represented corruption. Common corruption then could form a basis for groupings. When scholars gradually became aware that the Textus Receptus was corrupted, then the departures from the Textus Receptus were understood as earlier forms of the New Testament text, so that again manuscripts agreeing in variations from the Textus Receptus embodied the non-Byzantine text types. See Bart D. Ehrman, “A Problem of Textual Circularity: The Alands on the Classification of New Testament Manuscripts,” Biblica 70 (1989): 378. 44 See H. C. Hoskier, Concerning the Text of the Apocalypse: Collation of All Existing Available Greek Documents with the Standard Text of Stephen’s Third Edition Together with the Testimony of Versions, Commentaries and Fathers, 2 vols. (London: Bernard Quaritch, 1929), 1.160.

344

Tommy Wasserman

1960s, Ernest C. Colwell and Ernest W. Tune developed a sounder method of determining textual relationships between manuscripts and identifying various text types – so-called “quantitative analysis.”45 Significantly, Colwell argued that in order to locate a manuscript’s genealogical relationship to other manuscripts it must ideally be “compared completely [in all places of textual variation] with all other manuscripts.”46 In practice, however, they realised that it was impossible to make a comparison with such a large number of witnesses, at least until textual criticism began to use computers, and, therefore, in one of Colwell and Tune’s ground-breaking studies, it became necessary to use a selection of representative control witnesses from each major textual grouping and to count all textual variation between the included manuscripts.47 In this study, Colwell proposed a quantitative criterion of a text type relationship which was to become widely used: “the quantitative definition of a text-type is a group of manuscripts that agree more than 70 per cent of the time and is separated by a gap of about ten percent from its neighbors.”48 Further, Colwell and Tune stated that variants should be “weighed” before counting, so that the quantitative measurement was based only on those variants that were “genetically significant.”49 Colwell and Tune regarded a textual variant to be significant for determining genealogical relationships only if it had support from at least two Greek manuscripts, and did not consist of a nonsense reading or other scribal error. Through this procedure, Colwell and Tune were able to show the statistical relationship of agreement between the major groups of the manuscript tradition and any new manuscript could be located within this established grid of text types. The new methodology proved to be a great advance on previous work and has been used extensively in many subsequent studies.50 45 These articles are conveniently collected in Ernest C. Colwell, Studies in Methodology in Textual Criticism of the New Testament (Leiden: Brill, 1969). Two of the articles were co-written with Ernest W. Tune. For an account of the development of methods for determining manuscript relationships from the time of Colwell onwards, see Geer Jr. and Racine, “Analyzing and Categorizing New Testament Greek Manuscripts,” 497–518. 46 Ernest C. Colwell, “Method in Locating a Newly-Discovered Manuscript within the Manuscript Tradition of the Greek New Testament,” in Studia Evangelica, ed. K. Aland et al., TU 73, vol. 1 (Berlin: Akademie-Verlag, 1959), 757 (repr. in Studies, 26). 47 Cf. Ernest C. Colwell and Ernest W. Tune, “The Quantitative Relationships Between MS Text-Types,” in Biblical and Patristic Studies in Memory of Robert Pierce Casey, ed. J. N. Birdsall and R. W. Thomson (Freiburg im Breisgau: Herder, 1963), 25–32; repr. as “Method in Establishing Quantitative Relationships between Text-Types of New Testament Manuscripts,” in Studies, 56–62. 48 Colwell, Studies, 59. 49 Colwell, Studies, 58. 50 For some applications, see René Kieffer, Au delà des recensions? L’évolution de la tradition textuelle dans Jean VI, 52–71 (Lund: C.W.K. Gleerup, 1968); Kubo, “Textual relationships in Jude,” 276–82; W. L. Richards, The Classification of the Greek Manuscripts

Methods of Evaluating Textual Relationships

345

In particular, Gordon D. Fee has been a strong advocate of the quantitative method and has refined it in several of his studies of P66, Codex Sinaiticus, and the text of several church fathers.51 One area where Fee offered an improvement was in the classification of textual variants and in determining the extent of a variation-unit with a set of variants.52 In this connection, Fee suggested that all variation-units (except orthographic differences) where at least two manuscripts agree against the rest, should be included in the initial quantitative analysis and only after this “counting,” a process of “weighing” variation-units as to their genealogical significance could occur.53 The necessity of both counting and weighing relate to the fact that the genealogical significance of a textual variant has one static and one dynamic side. On the one hand, the genealogical significance depends on the nature of the variant itself (the static side): It is a truism of our discipline that some agreements in variation by their very nature are just as likely to be the result of independent scribal activity as others are almost impossible to explain apart from some kind of dependence on exemplars from the same family or text-type. 54

On the other hand, there is a dynamic side to genealogical significance: The closer MSS are to one another in actual point of origin, the closer will usually be the textual relatedness between them in all the classifications of variation. Therefore, at the highest level of manuscript relationships (over ninety percent agreement in total variations) all variants take on genetic significance.55

of the Johannine Epistles, SBLDS 35 (Missoula: Scholars, 1977); Larry Hurtado, Text-Critical Methodology and the Pre-Caesarean Text, SD 43 (Grand Rapids: Eerdmans, 1981); Carroll D. Osburn, “The Text of the Pauline Epistles in Hippolytus of Rome,” Second Century 2 (1982): 97–124; Tommy Wasserman, “The Patmos Family of New Testament MSS and Its Allies in the Pericope of the Adulteress and Beyond,” TC: A Journal of Biblical Textual Criticism [http://purl.org/TC] 7 (2002): pars. 1–59. 51 Gordon D. Fee, Papyrus Bodmer II (P66): Its Textual Relationships and Scribal Characteristics, SD 34 (Salt Lake City: University of Utah Press, 1968); Fee, “Codex Sinaiticus in the Gospel of John,” NTS 15 (1968–1969): 23–44; Fee “The Text of John in Origen and Cyril of Alexandria: A Contribution to Method in the Recovery and Analysis of Patristic Citations,” Biblica 52 (1971): 357–94; Fee, “Origen’s New Testament and the Text of Egypt,” NTS 28 (1982): 348–64; Fee, “P75, P66, and Origen: The Myth of Textual Recension in Alexandria,” in New Dimensions in New Testament Study, ed. R. N. Longenecker and M. C. Tenney (Grand Rapids: Zondervan, 1974), 19–45; Fee, “The Text of John and Mark in the Writings of Chrysostom,” NTS 26 (1979–1980): 525–47. 52 Gordon D. Fee, “On the Types, Classification, and Presentation of Textual Variation,” in Studies in the Theory and Method of New Testament Textual Criticism, ed. E. J. Epp and G. D. Fee (Grand Rapids: Eerdmans, 1993), 62–79. 53 Fee, “On the Types,” 67–8. 54 Fee, “On the Types,” 67. 55 Fee, “On the Types,” 67–8. Fee was not the first to make this observation. For example, Kirsopp and Silva Lake say, “In this list those variants which are purely misspellings would not be significant if they did not fit in so well with the grouping based on more secure

346

Tommy Wasserman

The weighing of variants was not a part of the quantitative analysis proper. Fee’s desideratum of both counting and weighing variants in order to determine manuscript relationships anticipates the feature of “connectivity” utilised in the more recent CBGM which integrates these two procedures in a refined way. The quantitative method has certain other limitations. First of all, the method can only provide information on how similar two given texts are, i.e., how often they agree or disagree. The method cannot show which manuscript or textual variant is prior to the other – it cannot yield any genealogical data. A complicating factor in this connection is the phenomenon known as “contamination,” i.e., when one manuscript’s text is derivative of multiple other manuscripts. One way this can happen is when scribes, after copying an exemplar, correct the copy against a different manuscript. This means that a different (earlier or later) reading than the reading in the exemplar can be copied into a manuscript. In this way, any manuscript contains a mix of prior and posterior readings. It is therefore impossible to make a neat genealogical stemma of manuscripts in the New Testament textual tradition in order to find a single superior document to always follow.56 Instead critics have had to evaluate the readings in individual passages based both on external and internal evidence – this procedure is known as the local-genealogical method.57 Secondly, Colwell and Tune used two short cuts in their own quantitative analysis: (a) only representative portions of text were collated; and (b) representative witnesses of pre-defined textual groups were selected. Only a full collation of a document can allow for a final judgment of its textual character.58 The second shortcut of choosing representative witnesses of pre-defined groups is also problematic, since the identity of any group must be defined on some ground, and then it must be established that a certain witness is a good representative of the group. Finally, it became apparent that the Colwell-Tune guideline that manuscripts of the same group will agree in 70% of all significant variation and be separated from others by at least 10% is only applicable to some groups and only when examining few witnesses.59 evidence” (Family 13 [The Ferrar Group]: The Text according to Mark [London: Christophers, 1941], 25). 56 As Colwell explains, Lachmann’s genealogical method was therefore inapplicable because it presupposes a textual tradition without contamination (or “mixture”). See Ernest C. Colwell, “Genealogical Method: Its Achievements and Its Limitations,” JBL 66 (1947): 114– 16. 57 Aland and Aland, The Text of the New Testament, 34, 281. 58 Cf. Klaus Wachtel who points out, “objectivity can only be reached, if all available evidence is taken into account” (“Colwell Revisited: Grouping New Testament Manuscripts,” in The New Testament in Early Christianity: Proceedings of the Lille Colloquium, July 2000, ed. C-.B. Amphoux and J. K. Elliott [Lausanne: Zèbre, 2003], 39). 59 This was observed by Fee, “Codex Sinaiticus in the Gospel of John,” 226–27; Richards, Classification, 33–41. More recently, Eldon Epp admits: “the tricky issue, of course, is

Methods of Evaluating Textual Relationships

347

6. Representing the Minuscules: The Claremont Profile Method Due to the lack of interest and method, no one from von Soden up to the 1970s had seriously attempted a major classification of the mass of minuscules. Since they had “a very high percentage of agreement among themselves,” W. L. Richards observed, “most textual critics have been content through the years simply to lump these closely related MSS together.”60 A growing awareness of the inadequate representation of the minuscule manuscripts in contemporary editions, however, led to the launching of the International Greek New Testament Project (IGNTP). The committees behind the project expressed their hope to produce a more balanced apparatus, which would give an adequate representation of “every known text-type, family or subfamily, as well as to any such groups as may be discovered in the course of our work.”61 Eventually, two graduate students at Claremont Graduate School, Frederik Wisse and Paul McReynolds, succeeded in developing a new method, the Claremont Profile Method (CPM), for selecting which minuscules that would be included in the IGNTP apparatus of Luke, which was the first projected volume. Colwell had once emphasized that “a group cannot be a group unless it has unique readings,”62 but Wisse and McReynolds, searching for such unique features of groups, observed that unique readings were relatively few in number and should not be the focus in group identification.63 On the other hand, they concluded that groups of manuscripts did have unique combinations of readings. These so-called “group profiles” emerged during the collation process and were adjusted during the course of work, but proved to be definite after some 200 manuscripts had been collated.64 After establishing the group profiles they could readily be displayed on graphs, and any unclassified manuscript agreeing roughly (in at least two-thirds of the readings) with a group profile could be assigned to that group.65 Eventually, 1,385 minuscules were collated in Luke 1, 10, and 20, and classified according to group profiles. This method too has been criticised, notably by Ehrman who pointed out several weaknesses: (1) the CPM uses only spot collations, with the risk of not determining, in percentage terms, what extent of agreement in readings joins members into a group, and what degree of separation in agreements determines the existence of a separate group” (“Textual Clusters,” 571). 60 Richards, “A Critique,” 562. 61 Merril M. Parvis, “The International Project to Establish a New Critical Apparatus of the Greek New Testament,” Crozer Quarterly 27 (1950): 307. 62 Colwell, Studies, 30. 63 Frederik Wisse, The Profile Method for the Classification and Evaluation of Manuscript Evidence as Applied to the Continuous Greek Text of the Gospel of Luke (Grand Rapids: Eerdmans, 1982), 35. 64 Wisse, Profile Method, 42. 65 Wisse, Profile Method, 126–33 (Appendix II).

348

Tommy Wasserman

detecting “block-mixture” (when the copyist changes exemplar); (2) one cannot be certain that group members have a high proportion of agreement in total variation, not only in readings previously determined significant for group profiles. For example, Codex Bezae was placed in group B (Alexandrian) along with ℵ and B among others.66 Ehrman has attempted to offer a refinement, combining the CPM with quantitative analysis in his “Comprehensive Profile Method.”67

7. The Coherence-Based Genealogical Method In 1982, Gerd Mink of the Institute for New Testament Textual Research (INTF) published an essay that described a new method for surveying the genealogical structure of the manuscript tradition.68 Since then, the method, subsequently known as the Coherence-Based Genealogical Method (CBGM), has been further developed and described.69 The method was adopted by the INTF in their ongoing work on a major critical edition, the Editio Critica Maior (ECM). So far, only the Catholic Letters and Acts have been newly edited in the ECM, but work is currently being done on Mark, John, the Pauline Letters, and Revelation. In the years to come, each of these books, along with the remaining books of the New Testament, will be completely revised and published in new volumes of the ECM and these in turn will be used to produce new editions of the NA and UBS. The CBGM has been designed to overcome the most serious limitation of the genealogical method – contamination. The method is based on some essential assumptions about the normal conditions of the New Testament textual transmission which are considered more probable than their opposites (following the rule of parsimony): (I) a scribe wants to copy a manuscript with fidelity; primarily the scribe does not want to create new readings; (II) if the scribe introduces other readings, they come from another source

Ehrman, “Methodological Developments,” 43–4 (Wisse, Profile Method, 52). See Bart D. Ehrman, “The Use of Group Profiles for the Classification of New Testament Documentary Evidence,” JBL 106/3 (1987): 465–86. The method has been utilised in a few studies, e.g., Bart D. Ehrman, “Heracleon and the ‘Western’ Textual Tradition,” NTS 40/2 (1994): 161–79. 68 Gerd Mink, “Zur Stemmatisierung neutestamentlicher Handschriften,” in Bericht der Hermann Kunst-Stiftung zur Förderung der neutestamentlichen Textforschung für die Jahre 1979–1981 (Münster: Institut für neutestamentliche Textforschung, 1982), 100–14. 69 For an extensive introduction to the CBGM including an annotated bibliography, see Tommy Wasserman and Peter J. Gurry, A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method (Stuttgart: Deutsche Bibelgesellschaft, 2017). 66 67

Methods of Evaluating Textual Relationships

349

(normally a manuscript); (III) if the scribe uses more than one source, few rather than many sources will be used; and (IV) the source copies have closely related texts rather than less related ones.70

It follows that nearly all manuscripts have close relatives (even if some are not extant), and nearly all readings of a manuscript are found in its exemplar and/or among its closest relatives. Thus, there is a sufficient degree of coherence in the manuscript tradition that allows the tracing of the genealogical relationship between the texts, in spite of exceptional factors that disturb the coherence, i.e., when scribes have introduced readings from other manuscripts (contamination) or created new readings, which may result either in accidental agreements between witnesses or singular readings. Because of the basic coherence, the CBGM detects these exceptional cases – no case is excluded. The CBGM focuses primarily on texts, not on manuscripts (although GregoryAland numbers are used for both). This is because many manuscripts in the textual tradition are now lost, and a text may be older than the manuscript that carries it. Moreover, contamination implies that a number of the variants of an ancestor manuscript may in fact be posterior to the variants of a descendant manuscript. Therefore, a stemma of the textual tradition, from the viewpoint of this method, can only reflect a general textual flow originating from the reconstructed text from which the manuscript transmission started, the “initial text,” or Ausgangstext (see section 8 below).71 Since contamination is assumed to have happened in small steps, exemplars and copies will normally have closely related texts. Therefore, the CBGM takes as a starting point the degree of textual agreement between witnesses based on a full quantitative analysis of all the included manuscripts. A similar preliminary step is also used to eliminate from further consideration the majority of Byzantine manuscripts that witness to the late Byzantine text. This step, however, is based on a smaller selection of test passages (Teststellen) in each book.72 The overall degree of agreement between the texts corresponds to the degree of probability that a single textual agreement between two witnesses indicates a genealogical relation between them. If two witnesses are textually distant on the whole, then their common attestation of a textual variant is more likely coincidental, i.e., the readings have probably emerged independently in the two

70 Gerd Mink, “Problems of a Highly Contaminated Tradition: The New Testament: Stemmata of Variants as a Source of a Genealogy for Witnesses,” in Studies in Stemmatology II, ed. P. van Reenen, A. den Hollander, and M. van Mulken (Philadelphia: John Benjamins, 2004), 25. 71 For Mink’s definition of “initial text” as opposed to “autograph” and “archetype,” see “Problems,” 25–7. 72 See Aland and Aland, The Text, 317–37.

350

Tommy Wasserman

witnesses. In this way, shared textual variants are regarded as more or less connective in terms of genealogical significance.73 The connectivity, however, depends not only on the degree of overall agreement between the attesting witnesses, but also on the character of the variant.74 Some variants are, by their nature, less likely to occur by chance. Hence, variants are both counted and weighed, as Fee suggested.75 The degree of agreement between witnesses (based on the quantitative analysis) and the connectivity of variants is regarded as “pre-genealogical evidence,” taken into account prior to any decision about the historical development of the variants. In light of the “pre-genealogical evidence” local stemmata of variants are drawn up, first in passages where the genealogical relationship is relatively clear, by applying the generally accepted methods of internal and external criteria. The latter involves “provisional assessments of the age [of the text] and quality of a number of well-known witnesses.”76 This first phase was inadequately described for the ECM Catholic Letters, but has now been accounted for in detail for Acts. Klaus Wachtel explains the preliminary steps for the application of the CBGM in Acts: In phase 1, in addition to the computation of pre-genealogical coherence, a first assessment of the text historical importance of manuscripts was conducted based on the rates of agreement with NA28…The following 24 witnesses agree more often with this hypothesis about the Ausgangstext (from now on represented by A) than with all witnesses that do not feature A as their closest relative: 03 (96,5), P74 (95,1), 01 (94,7), 81 (94,7), 02 (94,5), 04 (92,6), 1175 (92,4), 1739 (92,0), 33 (91,5), 307 (91,1), 610 (90,9), 453 (90,8), 181 (90,5), 2818 (90,4), 1678 (90,1), 623 (89,8), 1409 (89,8), 2344 (89,8), 5 (89,3), 1875 (89,3), 1642 (89,1), 180 (88,4), 94 (88,3), L1188 (86,9).77

The manuscripts in bold were consistently cited witnesses in NA28, so there is an obvious circularity here, their proximity to A is partly due to the fact that the reconstruction of the initial text was guided by these same manuscripts. However, this is not the case for those following after GA 33. Next to the “queen of minuscules” (GA 33) is 307 – a Byzantine catena manuscript. Wachtel further explains that another group of manuscripts was given special consideration for its affinity to the “Western text.” This group “comprises primarily P127, 05, 08, 1884, and the smaller papyrus fragments P29, P38, P48.

Mink, “Problems,” 28–9. Mink, “Problems,” 29, 54–5. 75 Fee, “On the Types, Classification, and Presentation of Textual Variation,” 67. 76 Mink, “Problems,” 16. 77 Klaus Wachtel, “Notes on the Text of the Acts of the Apostles,” in ECM Acts 1.1, 28*– 73 74

9*.

Methods of Evaluating Textual Relationships

351

Moreover, 614, 1292, 1505, 1611, 1890, and 2138 were assigned to this category.”78 The external evidence is applied with great reservation in the initial stage, since the role of most witnesses in the history of transmission is largely unknown at this point. The procedure is repeated in an iterative process moving from clear cases to more complex variation-units. The accumulated data contributes to the successive revision of local stemmata, and, ultimately, to the solution of more difficult cases of textual variation.79 In this way, the external evidence, that reflects the most probable place of each witness in the general textual flow, successively becomes established and can be applied with increasing confidence. The graph of the most probable relationships between the witnesses (or texts) in a book or corpus shows the “predominant textual flow” in that book or corpus. The potential ancestor(s) of a witness, placed at a higher level in the global stemma, exhibit a greater number of readings with priority (as decided by the critic) than the descendant witness. Whenever a new local stemma is to be drawn up or revised, one can use the genealogical database to assess various potential stemmata, to see if the reconstruction of readings in the single variation-unit is coherent with the accumulated genealogical data for the book or corpus.80 In the ideal case there is a perfect genealogical coherence in and between the attestations of the variant readings, i.e., a full correspondence between the textual flow in the variation-unit and the textual flow in the book/corpus, based on all previous decisions about local stemmata. Because of contamination and coincidental emergences of readings one might have to allow for imperfect coherence to some extent. It is up to the critic to adjust in each variation-unit the level of tolerance, so to speak. The graph showing the textual flow in a particular variation-unit is termed “textual flow diagram.” Whereas a local stemma represents the genealogical relationship between variants in one variation-unit, the global stemma of a book or a corpus is a representation based on all local stemmata, reflecting the most likely genealogical relationships between the texts that the manuscripts carry. The witnesses in the global stemma are arranged in a database according to the levels of probability that one is the ancestor of the other, and this data can be graphically represented according to these levels. The global stemma is the final step in the CBGM building on all the work that has been done before. The current way to construct this

Wachtel, “Notes,” 29*. Some cases will remain unsolved. In the Catholic Letters, the editors left the decision open in forty-three places, and in Acts, 155 textual decisions were left open (the guiding line is split in the ECM edition in these variation units). 80 These tools are available on-line in the modules “Coherence in Attestations” and “Coherence at Variant Passages” in the software “Genealogical Queries 2.0. Online: http://intf.uni-muenster.de/cbgm2/GenQ.html [accessed 15 April 2018]. 78 79

352

Tommy Wasserman

stemma, however, takes a lot of labour, and therefore no complete global stemma yet exists, although there might be a way to automate its construction.81

8. How the CBGM is Changing Textual Criticism82 First of all, an important shift has come in terms of the editors’ primary goal. Whereas textual critics have traditionally (though not always) pursued the “original text” of the New Testament, there has been a growing awareness about the ambiguity of the term.83 Some believe that the term “original text” is simply too vague to be meaningful and is therefore a cause of confusion. Others go further and argue that the very notion of a single, authorial text for the New Testament is indefensible.84 In the midst of the debate, the CBGM has brought with it a new goal and a new term to go with it. This new goal is to reconstruct, not the “original text,” but the Ausgangstext (“starting point text”), which is often represented with a capital A. The meaning of the term is “the reconstructed form of the text from which the manuscript transmission started.”85 From this definition it follows that the initial text may refer to the author’s text or it may refer to something later. That is because it is the text from which the manuscript tradition as we know it started. The editors of the Catholic Letters in the ECM are quite clear in stating that “following the most simple assumption, we claim that the present reconstruction is a hypothesis about the texts of the authors.”86 Nevertheless, the new term, initial text, has the benefit that scholars can often work together on significant text-critical projects even if they do not agree exactly in their views of textual history or about what text they have arrived at once they agree on the form of the initial text itself.87 See Wasserman and Gurry, New Approach, 98–102. This section and the following summarize Wasserman and Gurry, New Approach, 5– 13, with some additions. 83 On this point, see Eldon J. Epp, “The Multivalence of the Term ‘Original Text’ in New Testament Textual Criticism,” HTR 92/3 (1999): 245–81. 84 For the best introduction to the debate, see Michael W. Holmes, “From ‘Original Text’ to ‘Initial Text’: The Traditional Goal of New Testament Textual Criticism in Contemporary Discussion,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 637–88. 85 Gerd Mink, “Contamination, Coherence, and Coincidence in Textual Transmission: The Coherence-Based Genealogical Method (CBGM) as a Complement and Corrective to Existing Approaches,” in The Textual History of the Greek New Testament: Changing Views in Contemporary Research, ed. K. Wachtel and M. W. Holmes (Atlanta: SBL, 2011), 143 (emphasis added). 86 ECM Catholic Epistles 1, 30*. 87 For example, compare the different views of the textual history reflected in the respective essays of the two ECM editors Holger Strutwolf and David Parker in The Textual History 81 82

Methods of Evaluating Textual Relationships

353

A second, perhaps more apparent result of the CBGM is a number of changes to the most popular editions of the Greek New Testament, the NA and UBS editions. In the Catholic Letters, there are a total of thirty-four such changes and in Acts there are fifty-two changes.88 Along with the changes to the text, there has also been a slight increase in the ECM editors’ uncertainty about the text, an uncertainty that has been de facto adopted by the editors of NA/UBS in the Catholic Letters, where these passages are marked with a diamond (♦).89 This symbol marks places where the ECM editors “formally refrain from any rating” as to which reading is preferred, and in these instances the text of the primary line is split in the ECM edition.90 A third related change is a significant re-evaluation of the Byzantine text, which is given more weight than in the past. When the CBGM was first used on the Catholic Letters, the editors found that a number of Byzantine witnesses were surprisingly similar to their own reconstructed text. This unexpected discovery encouraged a second look and led to a renewed appreciation for these manuscripts and their shared text. This, in turn, led them to revise all their earlier decisions where they had chosen against this shared Byzantine text. As a result, ten of twelve changes between their first use of the CBGM on the Catholic Letters and their second use are in favour of the Byzantine text91 and they now consider it to be “an important witness to the early text” overall.92 The situation in Acts is similar. There were fifty-two changes to the critical text. In thirty-six cases the changes were made in conformity with the Majority Text and in only two cases against the Majority Text. Further, in 105 of the 155 passages where the editors leave the decision open about the initial text, the Byzantine witnesses attest to the reading deemed to be of equal value to variant a (= NA28). In twenty of the 155 passages the Byzantine witnesses side with variant a.93 Thus, the Byzantine manuscripts have early roots and this has put them in a position, in some cases, to preserve the earliest reading in isolation from the rest of the Greek New Testament (chapters 1 and 2). In my opinion, it makes little sense to appeal to authorial style or theology if one does not think we have the approximate author’s text. See Wasserman, “Was There an Alexandrian Recension,” 18–21. 88 These are listed in the introductions of each edition. 89 It should be noted, however, that textual changes will not automatically be transferred from the ECM to the future NA/UBS editions. The NA/UBS editorial committee will assess the textual changes including the passages marked with a diamond in the ECM and decide independently (as a committee) which text to adopt and which passages to mark as uncertain. 90 ECM Catholic Epistles 1, 34*. The editors prefer to speak of “guiding line” at these passages where the primary line is split and displays two alternative variants (ECM Catholic Epistles 1, 37*). 91 See Jas 1:20; 2:4, 15; 4:10; 1 Pet 5:1; 2 Pet 2:18, 20; 2 John 5; 12; 3 John 4. In each of these, there is support against the newly-adopted reading from manuscripts such as P72, Sinaiticus (01), Alexandrinus (02), Vaticanus (03), and Ephraemi Rescriptus (04). 92 ECM Catholic Epistles 2, 10. 93 ECM Acts 1.1, 31*, 34*–7*.

354

Tommy Wasserman

of the tradition, as Günther Zuntz suggested regarding the Pauline corpus.94 This is something all New Testament scholars will need to reckon with in Acts, the Catholic Letters, and beyond if the same data holds true elsewhere. A fourth change concerns the text types to which textual critics from Bengel onwards have referred in order to group and evaluate New Testament manuscripts. If one reads Metzger’s widely used textual commentary that accompanies the UBS GNT, the notion of text types is absolutely essential to his explanation of the history of the New Testament text and, with it, to the practice of textual criticism itself.95 The UBS committee’s own explanations for its decisions are so regularly couched in these categories that it is hard to imagine Metzger’s commentary without them. Although few scholars today would still associate these text types with distinct locales, most do still associate them with distinct levels of importance. The Alexandrian is typically considered the most reliable text type with the “Western,” (sometimes the Caesarean), and Byzantine generally following in that order. Epp himself prefers to refer to the text types as “textual clusters.”96 However, the application of the CBGM has convinced the ECM editors to abandon the concept of text types altogether. For many, the practice of New Testament textual criticism can hardly be conceived of without these comfortable categories. But the ECM editors have replaced the relationships and value of text types with the relationships and value of individual witnesses – well over onehundred of them in the Catholic Letters. Because the computer can keep track of all these witnesses and their place in the transmission, there is no need to group them into a few text types. Moreover, by focusing on individual witnesses, the difficult problem of defining text types and their boundaries is bypassed. One definite exception here is that the editors still recognize the Byzantine text as a distinct text form in its own right. This is due to the remarkable agreement that one finds in late Byzantine manuscripts. Their agreement is such that it is hard to deny that they should be grouped. In fact, the editors using the CBGM do group them together, subsuming them in the apparatus under the symbol Byz. On the other hand, the editors want to avoid the term “text type” to describe the Byzantine text because it brings with it the notion of a textual revision (or “recension”), a notion that persists in spite of the attempt to redefine text types as a

94 Zuntz, Text of the Epistles, 254. Cf. Michael W. Holmes, “Reasoned Eclecticism in New Testament Textual Criticism” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 789. 95 Bruce M. Metzger, A Textual Commentary on the Greek New Testament, 2nd ed. (New York: United Bible Societies, 1994). 96 Epp, “Textual Clusters,” 556. For Epp’s most recent attempt to identify two distinct texts in Acts, see “Text-Critical Witnesses and Methodology for Isolating a Distinctive DText in Acts,” NovT 59/3 (2017): 225–96.

Methods of Evaluating Textual Relationships

355

process.97 In other words, the Byzantine textual tradition should be regarded as the result of a long process, albeit one that produced a distinct text form preserved in a huge group of similar manuscripts.98 As for the “Western” text type, it is also clear that there are unique “Western” readings in Acts, but, at the same time, the mutual agreement of their attesting witnesses is relatively poor. They disagree with one another at the same level as with non-Western witnesses. Therefore, the ECM editors are reluctant to identify the witnesses as belonging to a text type and prefer to speak instead about a “‘Western’ cluster of variants” (rather than witnesses), a stratum of the New Testament textual transmission which requires other methods than the CBGM to further explore because (a) the Greek witnesses attesting to “Western” readings lack coherence; and (b) these readings are attested partly by versional witnesses.99 A final comment on the “Western” witnesses concerns the close relationship between GA 614 and 2412, the former of which is consistently cited in NA28. Barbara Aland considered 614 as attesting to a pre-D form of the “Western” text lacking the more typical Western readings, whereas Christopher Tuckett regards this form as “later influenced by the Byzantine text.”100 The close relationship of these two manuscripts has been noted ever since Kenneth W. Clark provided a collation of 2412 and suggested that the older 2412 could be the exemplar of 614.101 The CBGM results for Acts, however, list 2412 as the closest potential ancestor of 614. Although the manuscripts are likely siblings, these results On this point, see Wachtel, “Colwell Revisited,” 31–43. In this connection, Peter J. Gurry’s recent evaluation of the CBGM suggests that Wachtel’s view of the development of the Byzantine text needs slight revision in the Catholic Letters. See Peter J. Gurry, “The Harklean Syriac and the Development of the Byzantine Text: A Historical Test for the Coherence-Based Genealogical Method (CBGM),” NovT 60/2 (2018): 183–200. Cf. Wasserman and Gurry, A New Approach, 102–108. 99 As we have seen, however, the ECM editors have made sure to include witnesses traditionally affiliated to the “Western” text in the initial phase of the CBGM. For further discussion of a new approach to the “Western” text in Acts and the limitation of the CBGM in relation to this phenomenon, see ECM Acts 1.1, 31*–2*, Georg Gäbel’s special study, “‘Western Text’, ‘D-Text Cluster,’ ‘Bezan Trajectory,’ Or What Else? – A Preliminary Study,” in ECM Acts 3, 83–136; and Klaus Wachtel’s special study, “On the Relationship of the ‘Western Text’ and the Byzantine Tradition of Acts – A Plea Against the Text-Type Concept,” in ECM Acts 3, 137–48. 100 Barbara Aland, “Entstehung, Charakter und Herkunft des sog. westlichen Textes untersucht an der Apostelgeschichte,” ETL 62/1 (1986): 27–8; Christopher Tuckett, “The Early Text of Acts,” in The Early Text of the New Testament, ed. C. E. Hill and M. J. Kruger (Oxford: Oxford University Press, 2012), 164 n.33. Epp, “Text-Critical Witnesses,” 243, includes both 614 and 2412 in the “614 group” among “secondary witnesses” to the D-Text in Acts. 101 Kenneth W. Clark, Eight American Praxapostoloi (Chicago: University of Chicago Press, 1941), 35–7. 97 98

356

Tommy Wasserman

confirm Clark’s proposal (which was based on the age of the manuscripts) that 2412 is the head of “group 614.” At least in Acts, I propose that 2412 should be cited instead of (or along with) 614 in future editions of NA.102 The rejection of the concept of text types as a means of understanding the history of the text is significant. What fruit this will bear in the long term remains to be seen. It also remains to be seen whether New Testament scholars more generally will accept the CBGM as a viable replacement to text types. The issues are important and will, no doubt, be debated for some years to come. Finally, the adoption of the CBGM will have consequences for the practical application of external and internal criteria for evaluating readings.103 Elsewhere, in a discussion about the CBGM and the future of criteria, I have suggested, in relation to external evidence, that a reading is to be preferred if “supported by witnesses that have the initial text as their closest potential ancestor;” and if “the resulting local stemma is coherent with the predominant textual flow in the book or corpus.”104 These external criteria are incompatible with the text type categories since manuscripts from different traditional text types show up in textual flow diagrams in unexpected places, indicating textual interaction and ancestry beyond the traditional borders of the text types. Further, in relation to internal evidence, specifically transcriptional probability, I have proposed, that “a reading with imperfect genealogical coherency among its attesting witnesses is more likely the creation of scribes, since it seems to have arisen several times in the tradition by coincidence.”105 The case for multiple independent emergence of a reading is strengthened if its emergence can be explained on philological grounds, e.g., as a harmonization to a parallel, or as a copying mistake on palaeographical grounds. Thus, this new criterion complements traditional internal criteria.106 102 In spite of the results, Georg Gäbel and Klaus Wachtel continue to refer to 614 as head of a group of minuscules including 383 1292 1501 1505 1611 2138 2147 2412 and 2652 (Barbara Aland also included 206 915 1838 1891 and 2495). See Gäbel, “Western Text,” 83–106 (see e.g., 86); Wachtel, “A Plea Against the Text-Type Concept,” 137–48 (e.g., 142). In an earlier study, Gäbel also included 1751 in the group. See Georg Gäbel, “The Text of P127 (P.Oxy. 4968) and Its Relationship with the Text of Codex Bezae,” NovT 53/2 (2011): 113–14, 142–43, 149. 103 See Tommy Wasserman, “Criteria in Evaluating Readings in New Testament Textual Criticism,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, NTTSD 42, ed. B. D. Ehrman and M. W. Holmes (Leiden: Brill, 2013), 579–612. 104 Wasserman, “Criteria,” 605. 105 Wasserman, “Criteria,” 606. 106 In my work on Jude I depended on the CBGM results in my evaluation of external criteria (at the time, the online interface was not available). See Tommy Wasserman, The Epistle of Jude: Its Text and Transmission, ConBNT 43 (Stockholm: Almqvist & Wiksell, 2006), 234–339 (esp. 237–38). For application of the new transcriptional criterion, see Wasserman, “Historical and Philological Correlations and the CBGM as Applied to Mark 1:1,”

Methods of Evaluating Textual Relationships

357

9. Limitations and Future Improvements of the CBGM The CBGM has not completely resolved the problem of contamination, despite claims to the contrary.107 The CBGM tries to address the problem by relating texts rather than manuscripts, by allowing for multiple ancestors, and by using coherence to detect where contamination has happened. However, there are still cases where contamination could go undetected in the CBGM and the result would be that proper ancestor-descendant relationships would be inverted.108 Thus far, the CBGM is the best method to deal with contamination, but there are still certain problematic scenarios and the loss of witnesses plagues all methods at some point. Only a selection of the manuscripts that once existed have been preserved, and the earliest manuscripts are often incomplete. For the CBGM the selection of manuscripts is made still smaller in that the method in its present application incorporates only about one-third of our extant Greek manuscripts.109 The selection includes, to be sure, the most important and representative ones but they are still a selection not only of what once was but even of what we now have. Beyond this selection of Greek witnesses, the CBGM does not currently include any versional or patristic data. Their testimony can, of course, be used when constructing local stemmata, but after that point, we cannot see them in the textual flow diagrams or the global stemma nor can we compare them using pre-genealogical coherence. Hence, the developers of CBGM should seek ways to incorporate all the textual evidence currently cited in the ECM.110 Another improvement along these lines would be the possibility of better handling witnesses that underwent systematic correction in a special way. For example, GA 424 is a very peculiar manuscript which has been extensively corrected

TC: A Journal of Biblical Textual Criticism 20 (2015): 1–11; and Wasserman, “The Coherence Based Genealogical Method as a Tool for Explaining Textual Changes in the Greek New Testament,” NovT 57/2 (2015): 206–18. 107 Cf. David C. Parker, Textual Scholarship and the Making of the New Testament (Oxford: Oxford University Press, 2012), 84. 108 For examples, see Peter J. Gurry, A Critical Examination of the Coherence-Based Genealogical Method in New Testament Textual Criticism, NTTSD 55 (Leiden: Brill, 2017), 151–54; Mink, “Problems,” 49–59, 63–7. 109 There is nothing in theory that limits the number of manuscripts to which the method can be applied. In my comprehensive study of 560 manuscripts in the Letter of Jude, I have suggested that the ECM editors could test the validity of their selection of 134 manuscripts in Jude by comparing their global stemma to a global stemma for the complete manuscript tradition of Jude using my data. See Wasserman, Epistle of Jude, 29. 110 Patristic and versional evidence, as well as evidence from Greek lectionaries could be included in the CBGM in the same way that fragmentary witnesses already are. The user could then have the option to include or exclude them, and they probably would need to be excluded from the global stemma.

358

Tommy Wasserman

against at least one other manuscript, so that 424* and 424C (or 424Z) could arguably be viewed as two distinct witnesses. Thus, in a case like this, an option to treat the corrected text as a different witness would be desirable (although it may be problematic to count all corrections as made from another manuscript).111 The bilingual GA 629 is another unusual manuscript, one in which the text of the Latin column has heavily influenced the Greek column. (The influence from Latin includes the first hand.) Another possible candidate is 01C2, a corrector of Codex Sinaiticus.112 Another way to improve the CBGM would be to create a means of weighing different agreements differently in pre-genealogical coherence. Currently, agreements are only “weighed” at the stage where genealogical coherence is used, and at that point, the connectivity of the variant is taken into consideration. The most flexible option would be to allow individual users of the CBGM to categorize agreements using their own schema. This could include an option to exclude certain singular readings from the count of agreements (pre-genealogical evidence).113 Further, in regard to variants, it would be useful to be able to include certain “error” readings in the method as distinct readings rather than the current method of counting them as their “intended” reading. The reason is that such errors convey genealogical information. Sometimes, on a closer inspection, the “errors” may even represent meaningful (mis)readings or explain other readings created to correct them.114 Currently, every time the ECM identifies an error reading with its intended reading, it is doing exactly what the editors have done throughout the CBGM, which is connecting variants with their source. This information should therefore be included as such in the CBGM rather than discarded by

111 My examination of the twenty-five corrections of 424 in Jude shows that they are shared with GA 6 in nineteen cases, suggesting that it may be possible to identify the source texts used in corrections. 112 Klaus Wachtel confirms in personal communication that this function is on the agenda for future development of the CBGM software. Besides error readings, future versions might also provide an option to include orthographica (i.e., spelling differences). In cases where witnesses are closely related, e.g., P75 and 03, even these textual minutiae are genealogically significant. 113 While some think that no singulars should be excluded on principle (e.g., Gurry, Critical Examination, 188–92; Wachtel, “Colwell Revisited, 34–5), perhaps the best route is to make a decision on a case-by-case basis since the singulars in P75 , for example, are more likely inherited than those in, say, P45. David Parker, who is editing ECM John, has informed me in personal communication that he is doing some testing along these lines. 114 On this point, see Tommy Wasserman, “Theological Creativity and Scribal Solutions in Jude,” in Textual Variation: Theological and Social Tendencies? Papers from the Fifth Birmingham Colloquium on the Textual Criticism of the New Testament, ed. by H. A. G. Houghton and D. C. Parker (Piscataway: Gorgias, 2008), 75–84; and the discussion of nonsense readings in Gurry, Critical Examination, 195–99.

Methods of Evaluating Textual Relationships

359

treating both error and intended reading the same. And this could be extended so that a certain variant could be connected with an error as its immediate source.115 The final improvement is the most obvious and, for that very reason, also the easiest to overlook. What scholars and student most need now is a customizable version of the CBGM that allows them to edit the local stemmata and create their own database of genealogical data.116 Currently, both the CBGM for the Catholic Letters and the updated version for Acts are closed systems; their underlying data cannot be changed.117 Providing a customizable option would mean creating a version that allows each user to have his or her own editable database. Fortunately, Klaus Wachtel has reported in personal communication that the INTF and the Cologne Center for eHumanities (CCeH) are currently developing a version of Genealogical Queries of Acts with editable local stemmata. This is a very welcome development indeed. Finally, a customizable version should also allow the user to divide the text into variation units and set the variants based on their own philological and textcritical considerations of what goes together.118

Summary and Conclusion In this chapter I have surveyed the various methods of evaluating textual relationships from the birth of the text types, introduced by Bengel, to the highly sophisticated computer-assisted method in current use, the CBGM. First, we have seen how Bengel’s original scheme of text types was refined by Semler and Griesbach and reached a peak through the monumental work of Westcott and Hort towards the end of the nineteenth century. Their theories about the history of the Greek New Testament text and the principles they used to establish it came to play a major role in the text-critical practice of the twentieth century, as reflected not least in the work of the UBS committee and the two editions of Metzger’s widely used Textual Commentary. Cf. Martini, Il problema, 86–122. Cf. Wasserman, “Criteria,” 607: “A desideratum for the future is an interactive interface that will enable users to pursue the complete critical process: to create their own local stemmata of variants, build up a genealogical database, and successively evaluate the consequences of their textual choices. Perhaps we will see the emergence of multiple genealogical databases reflecting different editorial orientations.” 117 The interface for the Catholic Letters is online at: http://intf.uni-muenster .de/cbgm2/GenQ.html; the material from Acts, the latest interface available online at http://ntg.cceh.uni-koeln.de/ph4/ [accessed 18 April 2018], has been significantly upgraded and is more responsive and informative than the older interface for the Catholic Letters. 118 On the problem of dividing the text of Jude into variation units, see Wasserman, Jude, 128, 231–32, 256, 319; cf. Gäbel on the evaluation on long, expanded readings in the “Western” text (ECM Acts 3, 124–33). 115 116

360

Tommy Wasserman

At the beginning of the twentieth century, von Soden took on the formidable task to collate, classify, and present the texts of numerous hitherto unknown Greek manuscripts in his large four-volume work, which, unfortunately, was riddled with problems, something which his contemporary, Hoskier, pointed out in a scathing review. Hoskier was another scrupulous collator of manuscripts, but his attempt to determine their place in the textual history failed too. On the other hand, he succeeded to demonstrate that the “Neutral” (later Alexandrian) text type was neither unaffected by textual changes, nor as homogenous as once thought – he tabulated thousands of differences between Vaticanus and Sinaiticus in the Gospels. In lacking exhaustive collations, however, it was not yet possible to put this figure into a larger perspective. The similarity of these two early witnesses, as I have demonstrated, is nearly 90% in the Catholic Letters in spite of a similar relative ratio of disagreements. Hoskier and others had tried to classify manuscripts by counting their common deviations from the Textus Receptus, which was problematic from a methodological standpoint. Later in the twentieth century, Colwell and Tune developed a sounder method for determining textual relationships between manuscripts and identifying text types, called quantitative analysis, which compared manuscripts with each other instead of an external standard. The main limitation of this method, however, was that it could not yield any genealogical data, but only demonstrate the textual proximity of manuscripts (or distance between them). Further, their collations were based on a very limited selection of witnesses and test passages. In 1982, Gerd Mink of the INTF in Münster presented a new and groundbreaking method, now known as the CBGM, which is a computer-assisted method for determining textual relationships and establishing the initial text, as printed in the major ECM editions, and in the widely used NA/UBS hand editions of the Greek New Testament. I have described how the adoption of this new method has brought about several changes, not only to the critical text and apparatus, but to the way in which textual criticism is practiced. Significantly, the proponents of the method have proposed to abandon the text type categories, which have been used for nearly 300 years, because the method allows for larger precision in the evaluation of textual relationships. In the final part of this survey, I have pointed to some limitations of the CBGM and proposed a number of improvements for the future. It is important to keep in mind that this method is far from perfect, it does not solve the problem of contamination once and for all, but it is currently, in my opinion, the best way of dealing with it. In this connection, it is necessary to emphasize that the ultimate result of the CBGM, the global stemma, is not a stemma of the extant manuscripts, but it is a graphic representation of the simplest hypothesis about how the text of our manuscripts developed. As Richard Evans reminds us, our historical knowledge is always contingent on “the extent to which it is possible to

Methods of Evaluating Textual Relationships

361

reconstruct the past from the remains it has left behind.”119 What is left behind to text critics are literally fragments, chance survivals from the past – we are trying to piece together the puzzle with only some of the pieces.

119

Richard J. Evans, In Defense of History (London: Granta Books, 1997), 110.

Thoroughgoing Eclectic Textual Criticism: Manuscripts and Variants of Revelation J. K. Elliott 1. Manuscripts The textual situation in the Book of Revelation differs in many respects from the manuscript heritage for other parts of the New Testament. As a consequence, evaluating textual variation in Revelation differs too. There are comparatively fewer manuscripts of Revelation than of other New Testament books and there is a greater necessity for editors to depend on minuscule witnesses. Also, there is a recognition that this book was often accompanied by a commentary, as well as an awareness that Revelation was sometimes transcribed separately from other books of the New Testament and was even accompanied in some manuscripts by non-biblical texts. The Greek text of Revelation has been subjected to authoritative and thorough academic scrutiny in the past century, resulting very recently in a shaking of the foundations of the previously firm understanding of its textual character.1 The text of Revelation is being given much scholarly attention at the moment, primarily because a research team, under Professor Martin Karrer’s direction, is undertaking a large project in Wuppertal. This will end with the publication of the Editio Critica Maior (ECM) for that book and its text will become the text of the Nestle-Aland and United Bible Societies’ hand editions. To start its work the team collected a full range of photographs or digitized forms of all the manuscripts of Revelation, and from those a goodly number of Teststellen (test passages) have been prepared as part of the Text und Textwert series (TuT); the total being 123, proportionately greater than the number of Teststellen isolated in Münster and prepared for the other books of the New Testament. Recent articles by Karrer’s colleagues about the manuscript heritage of this book, as well as the tables listed by David Aune in the introductory matter to his commentary on Revelation, have rendered some of the details in the standard printed register of

1 As, for instance in the article by Juan Hernández Jr., “The Creation of a Fourth-Century Witness for the Andreas Text Type: A Misreading in the Apocalypse’s Textual History,” NTS 60 (2014): 106–20.

364

J. K. Elliott

Greek New Testament manuscripts, the Kurzgefaßte Liste (hereafter Liste),2 out of date. The time is ripe to revisit the official register of Greek New Testament manuscripts.3 Additional manuscripts have been registered since the second edition of the Liste; in other cases some manuscripts have been given new numbers. This is ongoing work, but text critics should always use the latest registered GA (= Gregory-Aland) numbers for manuscripts. The Teststellen chosen for the TuT volume display a greater number of variants than is conventionally printed in the apparatus criticus in a hand edition of a Greek New Testament. I examine below information from the recently published Teststellen, focusing on a number of important new issues raised from this project.4 Specifically, I deal with the following matters: (1) The re-numbering of certain portions of manuscripts where the book of Revelation has now been allocated a new and distinctive Gregory-Aland siglum; (2) Manuscripts previously listed as Abschriften, i.e. copies of other manuscripts, now re-numbered; (3) Lists of manuscripts which scholars judge ought never to have been allocated a Gregory-Aland registration number and some borderline cases that need reconsideration; (4) The use of brackets around manuscript numbers in the Liste. 1.1 Manuscripts in the Liste allegedly containing Revelation According to the statistics collected by Markus Lembke, 307 manuscripts allegedly contain Revelation.5 TuT has 310 in its listing. My own count is: Published in print as Kurt Aland, ed., Kurzgefaßte Liste der griechischen Handschriften des Neuen Testaments, ANTF 1, 2nd ed. (Berlin: De Gruyter, 1994) and nowadays regularly updated online. The online version is accessible at http:www.uni-muenster.de/liste, or via the INTF website. 3 The facts and figures will also require adjustments to the listings in J. K. Elliott, “The Distinctiveness of Greek Manuscripts of the Book of Revelation,” JTS 48 (1997): 116–24. Also see M. Lembke, Darius Müller and Ulrich B. Schmid, eds., Text und Textwert der griechischen Handschriften des Neuen Testaments. VI. Die Apokalypse, ANTF 49 (Berlin: De Gruyter, 2017), TuT hereafter. 4 Any changes from the current NA28 that may occur in the ECM Revelation under preparation should be discussed in full in the Wuppertal team’s third part of their ECM edition. 5 M. Lembke, “Beobachtungen zu den Handschriften der Apokalypse des Johannes,” in Die Johannesoffenbarung. Ihr Text und ihre Auslegung, ABG 38, ed. M. Labahn and M. Karrer (Leipzig: Evangelische Verlagsanstalt, 2012), 19–70 (esp. 21), and Martin Karrer’s introductory chapter, “Der Text der Johannesapokalypse,” in Die Johannesapokalypse, WUNT 287, ed. J. Frey, J. A. Kelhoffer and F. Tóth (Tübingen: Mohr Siebeck, 2012), 43– 78. Cf. also Ulrich Schmid, “Die Apokalypse, überliefert mit anderen neutestamentlichen Schriften – eapr-Handschriften,” in Studien zum Text der Apokalypse, ANTF 47, ed. M. Sigismund, M. Karrer and U. Schmid (Berlin: De Gruyter, 2015), 421–41. Lembke’s total of 307 seems rather large, but, as we see, precise totals are hard to achieve, given ongoing work on manuscripts. 2

Thoroughgoing Eclectic Textual Criticism

365

(1) Seven papyri: P18 P24 P43 P47 P85 P98 P115 – all containing only Revelation; (2) Twelve majuscules: ‫א‬01 A02 C04 P025 046 051 052 0163 0169 0207 0229 0308, the final eight of which contain only Revelation; (3) The remainder are minuscules,6 and these date from the ninth to the nineteenth century. The oldest minuscules are 1424 1841 and 1862. Modern editors of a critical edition of the Greek text of Revelation are largely dependent on minuscules with a higher than usual age. Some contain only Revelation; the rest combine Revelation with other parts of the New Testament.7 The total of minuscules reaches 283, but with 2924 the figure becomes 284. The minuscules can be further subdivided based on their juxtaposition to other New Testament works. 52 minuscules (and majuscules ‫א‬01 A02 C04) contain the whole of the New Testament (ea/cpr = Gospels, Acts/Catholic Letters, Paul, and Revelation). The number reduces to 50 if we delete 339 (burned) and 1785r (lost). The Liste has now re-numbered the sections containing Revelation in the following otherwise complete copies of the New Testament: manuscripts: 180 = 2918, 209 = 2920 and 1668 = 2909. 2041 is the number now given to the section of Revelation previously catalogued within 1040. 73 minuscules (plus majuscule P025) containing a/cpr in the Liste is a total that needs to be reduced to 72 because 1757 is invalid (see below).8 However, recent information suggests that 1768 (previously 1768 a/cp) also contains Revelation and so should be added here. So now we increase the total back to 73. The following sections containing Revelation (alongside a/cp) have now been re-numbered for the following manuscripts: 181 = 2919, 429 = 2921, 1140 = 2922, 1857 = 2923, 1894 = 2926.9 For 94/2917 see TuT, 4 and its listings. One minuscule, 886, contains eapr but not c.10 Ten minuscules contain er (possibly seen to be an unusual pairing of two Johannine writings): 792 1006 (1064?) 1328 1551 1685 2323 2643 2656 2794. 6 David E. Aune, Revelation, WBC, 3 vols. (Dallas: Word, 1997–1998), cxl–cxlviii lists 293 minuscules but he includes 1r and 2814 which refer to the same manuscript and he includes 60r and 2821 which are the same manuscript. Likewise 1352b and 2824, referring to the same manuscript, are listed here. Item 72 in his list is blank. He includes 1277 even though it lacks a text of Revelation; 598e (!) is included. Thus he actually has 287 minuscules. 7 Schmid, “eapr-Handschriften,” 436–37 lists the manuscripts that combine Revelation with other New Testament books. 8 Schmid’s list (“eapr-Handschriften,” 436–37) includes in this category, no. 1: 49 minuscules and 7 “uncertain” witnesses; but he admits 180 into his list; in list 2 he includes: 1757 and 1857 (this is now 1857ap; the erstwhile 1857r is numbered 2923). His category 5 should have added 2004 = 1835. 9 See Schmid, “eapr-Handschriften,” 439. 10 See H. C. Hoskier, Concerning the Text of the Apocalypse, 2 vols. (London: Bernard Quaritch, 1929).Vol. 1 contains collations; vol. 2.389 is where 886 is pronounced “not a proper text of Revelation.” Lembke, “Beobachtungen,” 24 allows it on the basis of its few surviving verses of Revelation.

366

J. K. Elliott

Two minuscules exist which may be described as “Johannine.” These manuscripts, 368 and 743 (cf. 2419), contain only Johannine compositions (John, 1–3 John, Revelation). Four minuscules contain a/cr: 1859 2186(no c) 2619(no a) 2847. We could add the supplement (= two leaves) added to the Revelation section of 1835a/cr (cf. 2004). This has been numbered 2924.11 Seven minuscules, including 1957,12 contain pr. Circa 137 minuscules contain only Revelation (including 2041), although the total number will soon need to change. See below in the next two sections. Taking the lower of the two figures in these various arrangements, the total number of manuscripts (303) agrees with my count. 1.2 Re-numberings of manuscripts of Revelation Note above that ten minuscule manuscripts have new GA numbers.13 Usually there is a clear indication (handwriting, for example) to reveal that some other composite manuscripts have Revelation by a different or later hand, indicating that this book was added at a later stage in the manuscript’s history.14 The following information, too, may merit a separate, distinctive registration number: 61 eapcr, Aune says that r was written by a different copyist;14 88 acpr, Aune says that r was added by a later copyist;15 172 acpr, but r is by a different hand; 385 eacpr, r was added later and follows works by Chrysostom,16 which calls for re-examination; 632 acpr, r was written by a different copyist from those of the Praxapostolos. Ulrich Schmid also asks if Revelation in the following composite manuscripts should not be given a new, separate number: 35 61 175 757 1072 1652 1704 2201 2494 2554.17 These and other questionable numberings require further reexamination. 11 See

Schmid, “eapr-Handschriften,” 431–32. 1957 is the number given to the (minuscule) supplement to B03, and cf. 2924. 2075, a manuscript of Revelation, also contains a supplement. Other supplements, none containing Revelation, are to be located within the following manuscripts: 323 892 1241 1739. Logically those supplemented sections ought to be allocated a separate GA number too. 935 contains r and is a supplement (as too is 2004S = 2924). 13 Lembke, “Beobachtungen,” 21, 62–9 cites nine of these manuscripts by their old Gregory-Aland numbers, the exception being 1668r, cited under its new number (2909). 14 Less usually in the case of 180 it is the sections apcr (dated 1273) that were added to e (twelfth century). 15 Aune, Revelation, 1.cxl–cxlviii. 16 For the inclusion of Revelation amid non-biblical works see later (section 1.8). 17 35 757 and 1072 belong to the Complutensian group, E3 (see Aune, Revelation, 1.clvii– clviii). All the other manuscripts in that group, barring 2041r (= 1040 eacpr), are likewise manuscripts containing only Revelation. Schmid, “eapr-Handschriften,” 432–36. 12

Thoroughgoing Eclectic Textual Criticism

367

1.3 Manuscripts that are Probable Candidates for Deletion from the Liste Of the minuscule manuscripts allegedly containing Revelation it is clear from the Liste that certain among them ought never to have been allocated a number in a listing of manuscript copies. For example, some of the following seem to have been copied from a printed edition of the Greek text of Revelation and thus cannot represent part of the normal manuscript tradition. Other issues include the fact that some are in Modern Greek, some are not a continuous text of Revelation, and a few are now lost or destroyed. For example: 296 copied from a printed edition (If this is so, then it qualifies as an Abschrift in the Liste – as, too, ought the other manuscripts in the list said to have been copied from a printed edition.) 339 burned 1064 probably copied from a printed edition 1757 only a commentary without any text of Revelation 1785 lost 1806 lost 1824 copy of 2062 1903 copied from a printed edition 2039 burned 2049 copied from a printed edition (bracketed in Liste) 2063 only a commentary without any text of Revelation (bracketed in Liste) 2066 copied from a printed edition (bracketed in Liste) 2072 copied from a printed edition (bracketed in Liste) 2075 copied from a printed edition 2114 in Modern Greek (bracketed in Liste) 2116 copied from a printed edition (bracketed in Liste). Its present location not known 2136 copied from a printed edition (Textus Receptus) 2402 in Modern Greek (bracketed in Liste) 2408 is now seen not to be a New Testament manuscript but a text of Origen 2433 only preserves a commentary without any text of Revelation (bracketed in Liste) 2449 in Modern Greek (bracketed in Liste) 2619 possibly copied from a printed edition (this needs to be checked) 2926 (formerly 1894) is based in part on the Erasmus or the Aldine editions according to Hoskier.18 The manuscripts said to have been copied from a printed edition need to be examined. I gather that 2926 (the former 1984) was copied from the Textus

18

Hoskier, Concerning the Text, 1.610.

368

J. K. Elliott

Receptus only at the beginning of the manuscript. The same seems to be true of 1903. Only the supplement to 2075 was influenced by a printed edition.19 In a similar vein, Lembke notes fourteen manuscripts not being used in the Teststellen programme. Among these fourteen manuscripts Lembke, at the time of his writing (i.e. before 2012), lacked their photographs, possessed only incomplete photographs, or in one case (2648) the photograph was illegible. The transience of that information makes one wonder why we were told about these manuscripts. The accessibility of these manuscripts will surely change, and indeed in some cases has already changed.20 The following manuscripts are enclosed in square-bracketed in the Liste, implying they should not be included there: 1776 1777 (both are copies of the Textus Receptus) and 2087. These merit re-examination. The TuT data also intimates that 1064 1775, for instance, manuscripts not bracketed in the Liste, may be eliminated. All in all, therefore, the total number of manuscripts needs to be reduced. Thus the already relatively smaller number of manuscripts of Revelation vis-àvis the rest of the New Testament will be further reduced. 1.4 Abschriften Among the manuscripts in the printed Liste marked as Abschriften, only two contain Revelation. These are 205abs and 2036abs. They have now been re-numbered 2886 and 2891 respectively and are included in Lembke’s list.21 Incidentally, it is worthwhile querying the wisdom of these re-numberings because they obliterate the link with the manuscript from which they are copied. For instance 205 has a direct copy previously known as 205abs and the latter has received a new number (2886 in this case), meaning that the link between these witnesses is no longer clear.

19 Aune, Revelation, 1.cxl–cxlviii notes several more manuscripts which are copies or near copies of others. They too should therefore be considered Abschriften. These include 2082 a copy of 2043, 2078 from 2436, and 2258 from 2076. See further Darius Müller, “Abschriften des Erasmischen Textes im Handschriftenmaterial der Johannesapokalypse,” in Studien zum Text der Apokalypse, ANTF 47, ed. M. Sigismund, M. Karrer, and U. Schmid (Berlin: De Gruyter, 2015), 165–268. 20 TuT contains further up-to-date information on the manuscripts and on text groups as well as information on progress towards the ECM. On the original information cf. Lembke, “Beobachtungen,” 25–6, 29. 21 Lembke, “Beobachtungen,” 21.

Thoroughgoing Eclectic Textual Criticism

369

1.5 Dates22 It is well known that dating manuscripts is not an exact science, and several may change their dates from those given in the Liste and elsewhere. However, the oldest fragment seems to be P98 in Cairo, which is dated to the second century. P47 and P115 are third century. P18 is usually dated third to fourth century. The oldest complete text is ‫א‬01 (fourth century). Other fourth century manuscripts are P24 0169 0207 0308. Tobias Nicklas, in a recent book on the early text of the New Testament, introduces the earliest texts containing Revelation: P18 P24 P47 P98 P115 0169 0207 (but he does not refer to 0308).23 The later popularity of Revelation shown by the high number of Greek manuscripts dating from the Middle Ages is matched in the Latin tradition where remarkable illuminated copies of Revelation were also highly popular in this later period. The decimation of European society by the Black Death and similar pestilences inevitably made a book that looked to a New Jerusalem particularly appealing. Among illustrated copies of the Apocalypse, Roger Gryson gives examples from Trier, Cambrai, and Valenciennes.24 In addition to the commonplace observation that the closer we come to modern times then inevitably it is likely that more manuscripts are extant, we may say here that the proliferation of copies of Revelation in Greek and especially in Latin from the mediaeval period may well also be due to the relevance of its subject-matter to readers of the Middle Ages. 1.6 Commentary Manuscripts Most minuscule manuscripts of Revelation are accompanied by a commentary, and these represent an unusually high proportion of witnesses. Revelation was clearly a book that needed to circulate with patristic explanations. In many cases the text represented by the commentary is older than the (often very late) manuscript in which it is found. The main Greek commentaries are by Oecumenius, possibly from the sixth–seventh century, now found in nine minuscules and twenty-three partial witnesses, and by Andreas of Cappadocian Caesarea from the sixth–seventh century.25 Some thirteen manuscripts contain only an abbreviated form of Andreas’ commentary and fifteen contain scholia (i.e. marginal 22 D. C. Parker, An Introduction to the New Testament Manuscripts and their Texts (Cambridge: Cambridge University Press, 2008), 233 (Table 7.1) lists only three hundred manuscripts by date; Aune, Revelation, 1.cxl–cxlviii has two hundred and ninety three. Both offer differing dates as well as different totals. 23 Tobias Nicklas, “The Early Text of Revelation,” in The Early Text of the New Testament, ed. C. E. Hill and M. J. Kruger (Oxford: Oxford University Press, 2012), 225–38. 24 Roger Gryson, ed., Apocalypsis, Vetus Latina 26/2 (Freiburg: Herder, 2000–2003), 77. 25 See Juan Hernández Jr., “Andrew of Caesarea and his Reading of Revelation: Catechesis and Paranesis,” in Die Johannesapokalypse, WUNT 287, ed. J. Frey, J. A. Kelhoffer, and F. Tóth (Tübingen: Mohr Siebeck, 2012), 755–74.

370

J. K. Elliott

notes based on the commentary).26 Another commentary is by Arethas of Caesarea (ninth–tenth century), found in fifteen manuscripts.27 Some later Ottoman commentaries are also found in manuscripts of Revelation. For example, the commentary by Maximus the Peloponnesian is preserved in 2114 and 2402.28 These two manuscripts are in Modern Greek. 1.7 Text types29 David Parker is correct to remind his readers that each of the sections of the New Testament (ea/cpr) has its own history and the application of text types like “Western,” “Byzantine,” “Alexandrian,” and (in some quarters) “Caesarean,” which may be relevant in the Gospels, cannot nor should be used outside that fourfold collection.30 Most text-critics today use the names of these text types hesitatingly, even in the Gospels, or apply terms such as “Western” etc. only in the most general way, because it is now being recognised that (1) on the one level these are too broad-brush in their application when what is demonstrable are only smaller families and groupings (such as are evident par excellence in Josef Schmid’s work on Revelation, but theoretically possible elsewhere). Very few manuscripts have enough errors in common with other witnesses to merit their being placed under a common umbrella title. And (2), on another level, the terms are too narrowly restrictive, being inclusive of only a small number of witnesses. For example, it is now widely recognised that even most of the so-called Alexandrian witnesses do not cohere. The general, geographical terms to which all extant manuscripts supposedly belong are false especially in relation to Revelation. (The “Alexandrian” B03 for instance is no longer extant for this book.) Section 2 below explains what is more appropriate when coping with variants in Revelation. The Münster Institute (INTF) has virtually dispensed with such outdated concepts, acknowledging that they were appropriate only when scholars had a small number of witnesses available. Nowadays, a greater number of witnesses that can be ordered electronically (e.g. by CBGM, see below) creates a situation that throws up a huge admixture of allegiances. The INTF’s TuT series

26 Aune, Revelation, 1.cxl calculates that only 98 manuscripts contain a commentary (not always in its entirety). 27 Marcus Sigismund, “Quaestiones Aretae I,” in Studien zum Text der Apokalypse II, ANTF 50, ed. M. Sigismund and D. Müller (Berlin: De Gruyter, 2017), 411–32. 28 A. Argyriou, Les exégèses grecques de l’Apocalypse à l’époque turque (1453–1821). Esquisse d’une histoire des courants idéologiques au sein du people grec asservi, Εταιρεια Μακεδονικων Σπουδων (Thessaloniki: Kronoz, 1982). 29 Josef Schmid, Studien zur Geschichte des griechischen Apokalypse-Textes (Munich: Karl Zink, 1955–1956). Aune, Revelation, 1.clvi–clviii lists manuscripts of Revelation by text type. 30 Parker, Introduction, 6–7.

Thoroughgoing Eclectic Textual Criticism

371

abandoned the text type categories and they made no use of theories based on text types when establishing the ECM for the Catholic Epistles. Except for the dwindling number of scholars prepared to print automatically readings supported by the old Byzantine or Alexandrian text types, I suspect few readers will feel aggrieved by the new, more open, policy adopted by the editors of ECM volumes. The term “transparency” may have become a journalist’s cliché for reporting political and public-service manoeuvring, but the term is more properly used in currently applied text criticism in Münster and elsewhere. 1.8 Hybrid Manuscripts Another area that requires more attention is the description of the contents of manuscripts containing Revelation alongside non-canonical texts. This phenomenon may say something about how this book was regarded and used, especially by a Greek church that only belatedly accepted Revelation into its canon and which excludes its readings from the lectionary. The manuscripts that contain only Revelation from the New Testament canon, but which also include the following writings are to be found in an article by Bruce M. Metzger published in 2003.31 The writings are diverse, including Gregory of Nyssa, Gregory of Nazianzus, Job, Justin Martyr, Exhortation to the Greeks, Basil, Maximus the Peloponnesian, Theodoret, Peter of Alexandria and Chrysostom. Probably readers of Revelation felt a need to read this enigmatic text alongside commentaries by respected Church Fathers, but also wished to compare Revelation with other biblical and Patristic writings to locate an appropriate context for it. 1.9 Printed Editions Moving on from individual manuscripts to printed editions, it is significant to note that there had been no satisfactory edition of Revelation before Tregelles’ edition of 1850. Prior to that date only the faulty Textus Receptus had been printed.32 Tregelles in his 1872 edition of the New Testament used five majuscules, including ‫א‬01, A02, and C04, nearly one hundred minuscules largely based on collations made by Matthaei and by Wettstein, and seven versions. Tischendorf in the same year produced his edition that included the evidence of some forty minuscules. When Westcott and Hort were preparing their edition, they were obviously without their beloved B03 in Revelation. ‫א‬01, their other

31 See B. M. Metzger, “The Future of New Testament Textual Studies,” in The Bible as Book: The Transmission of the Greek Text, ed. S. McKendrick and O. O’Sullivan (London: The British Library, 2003), 201–208, esp. 205. 32 F. H. A. Scrivener, A Plain Introduction to the Criticism of the New Testament (London: Bell, 1894), 2.184 lists Erasmus’ readings in Rev 22:16–21 taken from the Latin to compensate for his lacunose Greek manuscript. There are other readings elsewhere where Erasmus has translated a Latin reading into the Greek text of Revelation.

372

J. K. Elliott

favoured manuscript, seemed to them to have what they considered to be an inferior text. They preferred instead as their “best” manuscripts other witnesses, notably A02 and C04. The newly published Tyndale New Testament states that originally it was to have been a revision of Tregelles’ text.33 With the preparatory work for ECM Revelation making good and successful progress, we are shortly to be entering into a new era with a different critical edition of this enigmatic and challenging New Testament book. The wider and arguably more scientificallybased methodology, i.e. CBGM, brings us into a new era in the text-critical study of Revelation. In contrast to the text produced in NA/UBS we shall soon have a different text that ought to change much of the thinking about this enigmatic tradition.

2. Variants In 2014 I published a short textual commentary on Revelation in the journal Novum Testamentum34 in the light of the then recently published 28th edition of the Nestle-Aland Greek New Testament.35 I selected a number of variants from that edition to reach decisions on what I considered to be the Ausgangstext. As we know, the book of Revelation has been treated differently by text-critics from other books in the New Testament. Partly, conservative exegetes find that their preferred Majority text type36 (often equal to the bulk of the Byzantine manuscripts) is divided in Revelation; some editions designate the divisions as MA and MK. Similarly, the revolutionary text inaugurated under Westcott and Hort’s editorships in 1881 resulted in their dependence on the readings in Codices Sinaiticus and Vaticanus – except in Revelation where, of course, Codex Vaticanus is no longer extant. I am aware in Revelation that “traditional” text-critics (be these Majority text advocates or “mainstream” eclectics) are unable to rely on their favourite 33 Dirk Jongkind, ed., The Greek New Testament (Cambridge: Cambridge University Press, 2017) uses Tregelles’ edition as its starting-point (see pp. 505–506, 525) despite its having “relaxed” its rules for Revelation. 34 J. K. Elliott, “A Short Textual Commentary on the Book of Revelation and the New Nestle,” NovT 56 (2014): 68–100. 35 Barbara Aland, Kurt Aland, Johannes Karavidopoulos, Carlo M. Martini, and Bruce M. Metzger, eds., Nestle-Aland: Novum Testamentum Graece, 28 th ed. (Stuttgart: Deutsche Bibelgesellschaft, 2012). 36 Modern examples of the majority text type are Zane C. Hodges and Arthur L. Farstad, The Greek New Testament according to the Majority Text, 2nd ed. (Nashville: Nelson, 1985), esp. xxxii–xli; and Maurice A. Robinson and William G. Pierpont, The New Testament in the Original Greek: Byzantine Text-Form (Southborough: Chilton, 2005). These demonstrate the fissiparous Koine text types.

Thoroughgoing Eclectic Textual Criticism

373

witnesses from, say, the Gospels. In Revelation most text-critics pay lip service to the “reliability” of A02 and C04 over against P47 and Sinaiticus (as far as the so-called “Alexandrian” witnesses are concerned) or in 046 051 052 plus many minuscules for the Byzantine witnesses. But, if we study the texts produced by the different critics, it can readily be demonstrated how erratically their respective manuscripts have been treated by them. My own methodological approach to variation in New Testament manuscripts is dubbed thoroughgoing textual criticism. I prefer to make decisions about the Ausgangstext based not on any particular manuscripts or group of manuscripts, nor on a manuscript’s antiquity or alleged textual affiliation but on its readings, variant by variant. Common principles that are applied when assessing textual variation in manuscripts are to accept as original those readings that agree with the writing’s proven style, or to reject readings that seek to improve on perceived infelicities in the language or grammar of the original, perhaps avoiding Semitisms or Hellenistic Greek. Such principles are enumerated in standard handbooks like Aland and Aland37 and Metzger’s Textual Commentary.38 In the case of the United Bible Societies’ edition (UBS) the principles are often all-too-readily jettisoned by the UBS committee if a majority on its committee deemed the supporting manuscript testimony to be “weak”! Aland and Aland tried to classify all manuscripts into text types, but their lack of success in applying these principles to the editing of the text of Revelation may be seen in NA28, where the preferred text type for Revelation (A02 C04) has not resulted in their accepting all the readings of those combined witnesses throughout Revelation. It is worth noting places where their printed text is found in only a few or in different manuscripts. A rough summary taken from the NA28 apparatus shows the way in which UBS/NA behave: with A02 against C04: 12:5, 12; with only A02: 5:1, 9 (with versions); 12:10; 13:10, 18 (where C04 is absent); 22:21; with A02 P025 against C04: 12:3; with ‫א‬01 P025 against A02: 14:8; with M against ‫א‬01: 9:13; with C04 against A02: 13:8; 14:8; with P025 MK against ‫א‬01 A02: 19:14; with 051 M against ‫א‬01 A02: 20:11; with ‫א‬01 against A02 C04: 16:5; with ‫א‬01* 1854 MK against ‫א‬01c A02: 21:27; 37 K. Aland and B. Aland, The Text of the New Testament, 2nd Eng. ed. (Leiden: Brill, 1989) (hereafter Aland and Aland). 38 Bruce M. Metzger, A Textual Commentary on the Greek New Testament (Stuttgart: Deutsche Bibelgesellschaft, 1994).

374

J. K. Elliott

with 1006c 2329 against ‫א‬01 A02 C04: 18:3; with C04 against ‫א‬01 A02: 18:22. In many ways these examples clearly demonstrate the inability of scholars to apply to Revelation the methodology that lies behind books in the rest of the New Testament. In Revelation the privileging of certain preferred manuscripts will not work. A thoroughgoing eclectic approach seems to be what is needed; the CBGM should be similar. The methodology I adopt when confronted by textual variation throughout the Greek New Testament (not only in Revelation) is to see first if I can discern which of several readings is likely to have been the originator of the others. Sometimes that may be possible if one alternative poses particular theological or syntactical or grammatical problems that may have invited emendation, correction, or adaptation. This is the old principle of difficilior lectio potior. At other times one may be able to accept as the earliest recoverable reading a longer text, especially when shortened versions can reasonably be demonstrated to have come about accidentally through parablepsis, homoioteleuton,39 or the like. Recent work on scribal habits reinforces the rule that the longer text is more likely to be original, other things being equal.40 Thus the principle brevior lectio potior no longer applies in a mechanical way. In places where the two criteria above are inappropriate, other checks may be applied. I am usually tempted to accept as the earliest reading (1) one whose language conforms to Semitic usage, if the alternative(s) present “better” Greek; or (2) a Koine expression or word as opposed to an Attic alternative, because scribes beyond the first Christian century, often the educated men of their day, avoided “unGreek” or unclassical words. Above all, I tend to accept as “original” a reading that conforms to the language, style, vocabulary (and, indeed, the theology) of the earliest recoverable text. One can plot and establish each biblical author’s style and usage from the many “safe” places where all extant manuscripts are in agreement, there being no reported variant. Having established the usage from the secure places, variants that concern a feature agreeing with the author’s practice elsewhere are likely to represent the earlier text. Obviously, it may transpire that some hitherto firm readings may be challenged by subsequent collations, but, in general, it often seems that one is able to establish an author’s preferences and see which variants conform to and which readings disagree with that usage. My thoughts on these issues have changed little since I published my aforementioned article in Novum Testamentum in 2014, but now I begin to develop these principles further.

39 I use A. C. Clark’s term “hom.” “Homoioteleuton” and “homoioarcton” are not helpful when referring to manuscripts written in scriptio continua. 40 E.g. James R. Royce, Scribal Habits in Early Greek New Testament Papyri (Leiden: Brill, 2008).

Thoroughgoing Eclectic Textual Criticism

375

In the last few years a new-fangled methodology called the Coherence Based Genealogical Method (CBGM) was designed that would allow the readings in the overwhelming numbers of Greek manuscripts to be considered fully; it was devised and promulgated by Gerd Mink and his colleagues at the University of Münster.41 It, therefore, began life in the INTF to facilitate ongoing work on the ECM Catholic Epistles. The CBGM encourages editors to discuss each and every variant throughout a given text with a view to reaching editorial decisions about which of several readings in a variation unit was likely to have been original. Both methodologies seek to find the Ausgangstext that best fits the author, the theology, and context rather than the blind following of the text in certain favoured witnesses. A term like “the cult of the best manuscript” is démodé. A subsequent publication, ECM Acts,42 changes the text of NA28 52 times, but even more remarkably, it provides some 155 alternatives as split lines (usually with two possible readings although there are three places when three choices are given as alternatives in split lines). This latter convention, commendable as it is, shows itself to be realistic, because editors cannot always readily decide which of several readings is original and which secondary. By harnessing a reservoir of readings and the manuscripts that contain them, the CBGM similarly allows readers to reach their own conclusions. Democratic ideals allow readers, rather than editors, a say in which text is most likely to be original. I have not been privy to any such editorial meetings at which the running (or guide) text is being decided upon, but Münster has now published ECM Acts and their decisions appear in an article by one of its editors, Klaus Wachtel.43 Comparable decisions were to have been published for their earlier ECM edition of the Catholic Epistles,44 but the promised part 3 (Studien/Studies), which would have included this work, has been jettisoned, a decision which many of us find unconscionably preposterous, given the earlier promises and puffs. As stated earlier, ECM Revelation is on its way. ECM John, being edited in Birmingham, should follow shortly afterwards. Commentaries on a selection of the textual variants in those volumes are expected.

41 Discussions and descriptions of Mink’s methodology may be seen in Peter J. Gurry, A Critical Examination of the Coherence-Based Genealogical Method in New Testament Textual Criticism, NTTSD 56 (Leiden: Brill, 2017) and Tommy Wasserman and Peter Gurry, A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method (Atlanta: SBL, 2017). 42 Holger Strutwolf, Georg Gäbel, Annette Hüffmeier, Gerd Mink and Klaus Wachtel, eds., Novum Testamentum Graecum: Editio Critica Maior. III. Die Apostelgeschichte (Stuttgart: Deutsche Bibelgesellschaft, 2017). 43 Klaus Wachtel, “Text-Critical Commentary,” in ECM Acts 3, 1–38. 44 Barbara Aland, Kurt Aland, Gerd Mink, Holger Strutwolf, and Klaus Wachtel, eds., Novum Testamentum Graecum: Editio Critica Maior. IV. Die Katholischen Briefe, 2nd ed. (Stuttgart: Deutsche Bibelgesellschaft, 2013).

376

J. K. Elliott

In Revelation, where the manuscript tradition is quite different, we may expect that Karrer’s team will produce a text with an above-average number of changes from NA28 and UBS5. That work is a few years away from publication, but in the meanwhile we do have TuT to hand. The work in TuT is what Hoskier started a century ago, namely full collations of each manuscript of Revelation. Thoroughgoing textual criticism and the CBGM may agree on the texts produced. When I composed my short commentary one of my divisions was numerals, where I dealt with the variants involving the many numbers found in Revelation. Perhaps the most famous variant – certainly the most notorious – is that concerning the number of the beast at Rev 13:18. Is the numeral 666 or 616 or what? Then I was inclined to argue in favour of the originality of 616, influenced in part by the recently published P115 which reads 616. It is a reading known also to Irenaeus.45 I concluded my discussion of the variant there with the following paragraph: Gematria and the symbolic significance accorded to names and their numerical equivalents gave rise to ingenious word/number play in antiquity. Here the Greek form Neron Caesar gives (in Hebrew) the number 666. The Latin Nero Caesar is 616 in Hebrew. Either number therefore could be applied to the same bestial character. 616 could also be read as Gaius (Caligula). But 666 is the more dramatic, being a threefold repetition of a number that falls just short of the perfect number, 7. But, as such, 666 could be a secondary improvement. Since the TuT volume on Revelation has been published one sees at (the aptly numbered) Teststelle 66 that there are more variants than I had dealt with (taking my small selection from the NA28 critical apparatus containing five variants). TuT has over 20 variants here, most of which have the numeral in figures rather than words.46 One may of course now extend the apparatus in NA28 considerably, but I would currently hesitate to favour 616, or any number other than 666. If the figure were written in numerals, careless scribes or scribes working from a poor condition manuscript could readily and accidentally put in any letter resulting in the consequent differing numbers. That is but one example of where we may choose to adjust our decisions and resultant text. Obviously the adoption of CBGM will eclipse the readings of the 123 Teststellen in TuT, but the latter book will of course influence the Wuppertal team’s use of which manuscripts they choose to display in their apparatus. TuT will often merely amplify and bulk up v.ll. already known. Additional collations merely expand currently published apparatus. Obviously, silly errors 45 This variant is discussed with appropriate citations by Parker, Introduction, 242–44. See also J. N. Birdsall, “Irenaeus and the Number of the Beast: Revelation 13,18,” in New Testament Textual Criticism and Exegesis, BETL 161, ed. A. Denaux (Leuven: Leuven University Press, 2002), 349–59. 46 See Zachary Cole, Numbers in Early Greek New Testament Manuscripts: Text-Critical, Scribal, and Theological Studies (Leiden: Brill, 2017).

Thoroughgoing Eclectic Textual Criticism

377

previously not identified in certain manuscripts will appear in TuT, but my hunch is that very few genuinely new readings will be exposed. In Wachtel’s commentary on the new and differing readings printed in ECM Acts the notes are divided into two sections introduced by TC (= Transcriptional Probability) and by GC (= Genealogical Coherence, i.e. CBGM). Clearly, CBGM has had a significant role to play in the decision-making, but seems not to have been decisive everywhere. That is as things should be, but I note with satisfaction that the method of textual criticism that I have espoused is close to the CBGM. In Münster’s case they seem to discuss the likeliest Ausgangstext, and, after comparing this with variant readings, conclude that x is more likely to have derived its secondary reading from y rather than vice versa. Manuscripts are treated as “tradents” to use their jargon and neologism, and the relative ages of these witnesses are irrelevant. Thus for the sake of argument, a twelfth century manuscript may today carry a reading that influenced an eleventh century witness. The sheer chance of survival of original and secondary readings is arbitrary. And that conclusion chimes well with thoroughgoing textual criticism which may occasionally prefer a reading that agrees with the biblical author’s language, style, theology but which by the sheer chance of survival is extant today in only a couple of mediaeval manuscripts. In Revelation it is significant that NA is prepared to print as Holy Writ readings that may appear in very few witnesses. We wait with bated breath to discover what emerges in ECM Revelation. Meanwhile we congratulate Karrer and his Mitarbeiter and wish them all power to their collective elbows as they prepare this long awaited addition to the ECM series. The textual criticism of Revelation is unlikely to be the same again once their edition is published and thereafter becomes the text of the NA and UBS popular pocket editions.

Speaking in Tongues: Collecting the Chester Beatty Biblical Manuscripts Jill Unkel To [Beatty] collecting Oriental manuscripts has been something of a challenge. He accepted the challenge, for he wanted to know if the ancient writings could reveal more of the hidden secrets of the East, the cradle of Christianity and civilisation. He wanted to explore and probe the Oriental mysteries like a mining engineer searching in the bowls of the earth for valuable metals and treasures. In a sense, when collecting he was still a prospector.1

The Chester Beatty collection includes manuscripts produced throughout the Christian East and West, from Egypt to England and from Ethiopia to Armenia. It is therefore represented not only by the early Christian biblical languages of Greek, Latin, and Syriac, but also by Coptic, Ge’ez, Armenian, and Church Slavonic. If printed books are added to this list, a myriad of languages into which the Christian Bible has been translated unfurls, including but not limited to Amharic, Arabic, Bengali, Chinese, Dutch, English, French, German, Italian, Mongolian, Manchu, Panjabi, Persian, Russian, Spanish, Tahitian, Tamil, and Tibetan. The list would expand further if it were to include not just biblical texts but any other Christian text. The variety of languages illustrates not only the diversity of the Christian church, but also the diversity of Chester Beatty’s interests. Indeed, part of his strategy when collecting South Asian manuscripts was specifically to acquire representations of different languages.2

From Denver to London to Cairo Chester Beatty’s interest in collecting began before he moved to London in 1913, but it was subsequent to this move that his passion truly blossomed and was

1 John Murdoch, unpublished biography of A. Chester Beatty (commissioned in 1958), 413, CBL Archives. 2 Hyder Abbas, “‘We want quality and condition’: the Formation of Chester Beatty’s South Asian Manuscripts and Miniatures Collection,” Original Intentions: Essays on Production, Reproduction, and Interpretation in the Arts of China, ed. N. Pearce and J. Steuber (Gainesville: University Press of Florida, 2012).

380

Jill Unkel

encouraged.3 Beatty made his initial fortune working for the Yukon Gold Company under the Guggenheims, but left his lucrative contract to start his own mining consultancy company, Selection Trust (1914) which eventually earned him the nickname “Copper King” (Figure 1).4 It was following the sudden death of his first wife, affectionately called Ninette (Grace Madeline Rickard), from typhoid fever in 1911 that he relocated to London.

Fig. 1. Chester Beatty at the Dolores gold mine, Mexico, 1903, CBL Archives, PH 13

Beatty purchased 24 Kensington Palace Gardens (Baroda House) in 1912, had it completely “modernised” by adding indoor plumbing and heating, and redecorated the interior to his taste. Together with his two children and his architectbrother, William Gedney, Beatty sailed for England the following spring and that June married his second wife, Edith Dunne. Around this time Beatty’s doctors advised him to escape the cold, damp London winter for a warmer climate as a 3 Beatty met the well-known bibliophile and Director of the Fitzwilliam Museum, Sydney Cockerell, in 1915, who encouraged his collecting and introduced him to collections that helped to raise his standards. Cf. Laura Cleaver, “The Present,” in Miniature Masterpiece: The Coëtivy Hours, ed. J. Unkel (Dublin: Chester Beatty Library, 2018), 104. 4 A. J. Wilson, The Life and Times of Sir Alfred Chester Beatty (London: Cadogan, 1985). See also, Fionnuala Croke, ed., Chester Beatty’s A–Z: From Amulet to Zodiac (Dublin: Chester Beatty Library, 2014). A copy of his contract with Guggenheim is kept in the CBP, K98.

Speaking in Tongues

381

way to alleviate the symptoms of the silicosis infecting his lungs – a condition he had acquired during his early mining days. Heeding their advice, Beatty and Edith, together with “little” Ninette and Chester Jr, travelled to Cairo in February 1914 (Figure 2). In fact, most subsequent winters up to the Second World War (excluding the First World War during which they travelled to China and Japan) were spent in the relative warmth of the Egyptian winter. They eventually built the Blue House (Beit el Azrak) in the “shadow of the Great Pyramids” complete with orange and lemon groves. Years later, in an interview with Sheila Powerscourt, Beatty attributed his bibliomania to that first trip to Egypt: “This is how I began collecting. When my wife and I took a trip to Egypt in 1913, I spent a lot of time in the souks and bought a few papyri that turned out to be important. I also saw some decorated Qur’ans that seemed remarkable.”5 It was this annual retreat that had a profound impact on his collecting interests and the eventual composition of his Library.

Fig. 2. Ninette and Chester Beatty Jr. in Egypt, 1914, CBL Archives, PH 14

Cairo and the Biblical Papyri Beatty’s connections in Cairo not only resulted in his world-renowned collection of illuminated Qur’ans and Arabic manuscripts, but also shaped what is now the Western collection. Already in the 1920s he had acquired biblical texts in Coptic, 5

Sheila Powerscourt, Sun Too Fast (London: Geoffrey Bles, 1974), 230.

382

Jill Unkel

Armenian, and Syriac through dealers in Cairo.6 But it was a collection of manuscripts written in Greek that made Beatty internationally renowned. In the winter of 1929/30, a phenomenal collection of papyrus codices came to his attention through Egyptian antiques dealers.7 Always concerned with quality and value he asked his advisors at the British Museum in London to ensure the antiquity and authenticity of the manuscripts. Eric G. Millar (together with Harold Idris Bell) examined a few examples from the “find” and returned a coded-cable that read: SILVER MINE VERY RICH HAS 3 SHAFTS (STOP) GOLD MINE RICH HAS FOUR SHAFTS (STOP) SHOULD BUY BOTH WITHOUT FAIL ESPECIALLY SILVER MINE. 8

A follow-up letter dated 10 February explained the code: the silver mine was probably third century (3 shafts) and contained the text of Daniel (CBL BP X), (Figure 3) and the gold mine he dated to the fourth century (4 shafts) and this book contained the text of Genesis (CBL BP IV). Millar also expressed his excitement, referring to Daniel as a “real find” and Genesis as “bound to be worth having.”9 The manuscripts in this find, which came to number eleven (or twelve by collection number), are now called collectively the Chester Beatty Biblical Papyri.10 When announced in The Times on 19 November 1931, it caused a sensation. The discovery changed the existing understanding of pre-Constantinian textual history. With the New Testament books (Gospels and Acts, Pauline Epistles, and Revelation) all dated to the third century, these documents were not only surprising for having survived the Diocletian persecutions, but the dating moved New Testament scholarship back by at least one hundred years. The discovery of P45 (CBL BP I) specifically altered the understanding of when Christians accepted the four gospels as canonical to earlier than had previously been presumed.

6 A Harklean Gospels (CBL Syc 703) dated to 1177 was acquired in Cairo in 1929; a latetwelfth century Armenian Gospel book (CBL Arm 556) was acquired in Cairo 1928; and three Coptic manuscripts (CBL Cpt 813–815) containing gospel and other biblical texts were acquired in Cairo in 1924/25. 7 For a more detailed history of the acquisition see, Charles Horton, “The Chester Beatty Biblical Papyri: A Find of the Greatest Importance,” in The Earliest Gospels: The Origins and Transmission of the Earliest Christian Gospels; The Contribution of the Chester Beatty Gospel Codex P45, ed. C. Horton (London: T&T Clark, 2004), 149–60. 8 Letter confirming telephone message, in which Millar asked Wooderson to send the telegram to Beatty (Eric Millar to John Wooderson, 3 February 1930, CBP 911). Eric Millar and Howard Idris Bell were in the Department of Manuscripts at the British Museum. 9 Although they had yet to test the readings for Genesis (Millar to Beatty, 10 February 1930, CBP 911). 10 Eight bifolios of Ezekiel and Esther were originally labelled CBL BP IX and thirteen folios of Daniel were marked as CBL BP X. It was later determined that the same hand was responsible for both Daniel and Esther (Ezekiel written in a different hand) and that Daniel was therefore originally included between Ezekiel and Esther in a single manuscript.

Speaking in Tongues

383

Fig. 3. Daniel from Ezekiel, Daniel, Esther, Egypt, ca. 200–250 AD, CBL BP X f.78

The Chester Beatty Biblical Papyri are some of the most important examples of early Christian artefacts. From the moment The Times announcement appeared, the Director of the British Museum, Frederic G. Kenyon’s letters to Beatty are filled with an anxiety to publish, with inquisitive scholars anxious to get a closer look. In 1937 Joan Kingsford Wood, Beatty’s librarian, facilitated perhaps the first academic seminar on the Beatty Papyri for a group of students from the University of London.11 Indeed, these manuscript continue to be important as both religious and historical documents, and are the subject of innumerable articles, books, theses, and documentaries. Beatty was always keen to ensure that his collection would be enjoyed by scholars and the public alike: “I am a tremendous believer that if you have these things you should show them, so that other people can enjoy them as well as yourself.”12 Beatty’s sentiment must surely have fed into his decision to bequeath his collection to Ireland, making the museum perhaps the greatest cultural gift given to the nation by a single individual. In a letter to Beatty dated 9 July 1936, Kenyon told him: “You have certainly written your name large in the history of the Bible text by your acquisition of 11 These appear to have been students of Vincent McNabb, Irish scholar and priest who lectured through the University of London’s External Lecture’s scheme (JKW to Kenyon, 8 December 1937, CBP 710). 12 Beatty to F. Richter, 4 December 1944, CBP 1120.

384

Jill Unkel

such a wonderful group of early MSS, which have so greatly extended knowledge of the way in which the books were written and circulated.”13 It was this notoriety, this fame and accolade that clearly influenced the subsequent direction of Beatty’s collecting, but there were other, mitigating factors.

Europe and Beyond Like many American collectors in the early twentieth century, some of Beatty’s earliest acquisitions were medieval European illuminated manuscripts written in Latin.14 Many of these were French Books of Hours acquired around the time of his move to London. He continued to collect European manuscripts steadily until 1929 when these acquisitions came to an abrupt halt. Not only did he discontinue his acquisition of these manuscripts, he also chose to sell a significant portion of this collection in the early 1930s, even while still paying Eric Millar to produce the third volume of his Western Manuscripts catalogues.15 As a result of Beatty’s decision to sell, Millar decided not to continue his work on the volume. Beatty never explicitly elucidated his reasons but many have drawn a connection between his decision and the Great Crash of 1929. Belle de Costa Greene, Director of the Pierpont Morgan Library, wrote to Eric Millar to discuss Beatty’s forthcoming sales but lamented the “apocalyptic beast called Depression” which she felt would hinder her purchasing abilities for years to come, in addition to the unfavourable exchange rates between the British pound and American dollar.16 The dealer, Bernard Quaritch, noted his pleasure with the “bargains” he acquired in the Beatty sale but also that Beatty was forewarned of the outcome, noting that “Mr. Beatty is a very disappointed man, 13 Kenyon to Beatty, 9 July 1936, CBP 451. Kenyon noted to Beatty in 1938, “I think ‘the Chester Beatty Papyri’ now hold an assured position among the most important documents of Biblical criticism” (Kenyon to Beatty, 26 February 1938, CBP 451). 14 For a history of Chester Beatty’s collection of European manuscripts see, Laura Cleaver, “The Western Manuscript Collection of Alfred Chester Beatty (ca. 1915–1930),” Manuscript Studies 2/2 (2017): 445–82. 15 Only two of the proposed five sales were held and Beatty chose to keep the remainder of the collection which he gave to his wife Edith and were returned to his collection after her death in 1952. Catalogue of the Renowned Collection of Western Manuscripts, The Property of A. Chester Beatty, Esq.: The First Portion, Sotheby’s sale, 7 June 1932 (London: Sotheby & Co., 1932); Catalogue of the Renowned Collection of Western Manuscripts, The Property of A. Chester Beatty, Esq.: The Second Portion, Sotheby’s sale, 9 May 1933 (London: Sotheby & Co., 1933). 16 Belle de Costa Greene to Millar, 21 February 1933, Morgan Library Archives, Director’s Files, Series 1, Box 1369, Eric Millar. Morgan later purchased the Mostyn Gospels from Beatty’s first sale (via Quaritch). Quaritch spent £1500 at the sale, Morgan paid him £2000 (cables between Bernard Quaritch and Morgan 7 and 8 June 1932, Morgan Library Archives, ARC 1310).

Speaking in Tongues

385

although of course he did not expect too much.”17 Not only did the sales produce poor results but it left people making assumptions about the state of Beatty’s finances. The fact that many people took a significant hit in the world financial crisis that followed the Crash led to some of his contemporaries to suggest that Beatty did not escape the Crash unscathed. In a letter to Greene in 1932 Fergusson wrote [Beatty] told me that the reason for the sale was his wishing to provide for death duties in America and England by spreading the sale of his MSS over four or five years. This I tell you in confidence, but there is no doubt much truth in the fact that he must have lost considerable sums of money lately. 18

In fact, Beatty may have been purposefully less than forthcoming with Ferguson, considering his indiscretion.19 Insulted by the poor results, and the gossip, Beatty cancelled the last three proposed sales. While the 1929 Crash and the subsequent Depression would have had some effect on his portfolio, Beatty was diversified, so rumours of his financial losses were greatly exaggerated. In addition, while Beatty often reinstated a “library budget,” he was always willing to break his own rules for a particularly great book.20 Furthermore, he did not cease collecting, so the Crash cannot explain why he decided to stop collecting European manuscripts, or indeed to sell them. As already mentioned, the bulk of the Chester Beatty Biblical Papyri was acquired in 1929/30. Furthermore, he purchased an important collection of Manichean manuscripts the following winter, also from dealers in Egypt. In addition to the purchase price, Beatty paid for the conservation, photography, and publication of a considerable portion of his everincreasing collections.21 As Beatty continued to spend money on his collection throughout the early 1930s, his decision to stop acquiring and to sell his European manuscripts cannot be explained by financial constraints. Throughout his collecting career Beatty’s interests often waxed and waned; and in the late 1920s, as Greene herself pointed out, “Beatty had rather lost interest in his European manuscripts.”22 Illuminated European manuscripts were extremely popular collectors’ items, and already before the Crash Beatty had failed to acquire manuscripts sold at auction when Quaritch to Greene, 10 June 1932, Morgan Library Archives, ARC 1310. F. S. Fergusson to Greene, 4 April 1932, Morgan Library Archives, ARC 1310. 19 Beatty later amended his Irish will to leave the majority of his remaining European manuscripts to his personal estate which were subsequently sold at auction and the proceeds likely in part used to cover death duties. The codicil was added in 1966 leaving his Western manuscripts to his personal estate with the exception of the forty “currently on display” (CBL Arc 1.1; The property of the late Sir A. Chester Beatty sold by order of the executors, parts 1 & 2, Sotheby’s, London, 3 December 1968 and 24 June 1969). 20 Abbas, “‘We want quality and condition.’” 21 Beatty paid Kenyon £600 for his work on the transcriptions and publications of the Beatty Biblical Papyri (Beatty to Kenyon, 7 July 1936, CBP 451). 22 Greene to Ferguson, 23 March 1932, Morgan Library Archives, ARC 1310. 17 18

386

Jill Unkel

other interested parties were happy to spend over the estimate, and Beatty’s top offer. Throughout his career he commented on excessive bidding (something he would have relished had it happened for his own sales). Beatty was likely frustrated with a market he saw as overinflated and saturated. His attitude to a cost to value ratio is best illustrated by his opinion of the British Museum’s acquisition of the Codex Sinaiticus in 1933 (two years after Beatty’s sensational acquisition of Biblical Papyri had been made public). “Pfoo! £100,000 the British Museum paid. And that was in 1933. Absolutely crazy and wrong – it was worth nothing of the sort.”23 Although we may disagree with his opinion today, Beatty had good reason to have it. Ever the shrewd businessman, Beatty was adamant that he never pay more than he felt an object was worth. He told Kenyon, in 1932, that he paid about £40 per leaf (so perhaps even less) for the portion of the Beatty Papyri he acquired in 1931.24 And in 1934 when he had the opportunity to acquire additional leaves from the Ezekiel-Daniel-Esther manuscript he rejected the offer, telling Kenyon that the owner of the leaves “has the idea they are worth more than the whole of the Testament. I think he is asking about £120 a leaf. However, next year, I may be able to get them at a fair price.”25 Although the exact amount Beatty paid for the eleven biblical manuscripts is unclear, at the prices he suggested, Beatty acquired the collection for less than a quarter of what the British Museum spent on the Codex Sinaiticus. No wonder his reaction to the price they paid, especially when the then retired director of the British Museum, Kenyon, wrote Beatty, “your papyrus really mark an epoch in the history of the Bible text, comparable to the discovery of the Codex Sinaiticus. They have filled a gap of two centuries in the record, and more take an important place in all works on testament criticism for the future.”26 After all, it contains the earliest copy of all four Gospels and Acts contained in one codex, the earliest copy of any portion of Mark, the earliest surviving copy of the Pauline Epistles, and the Book of Numbers (dated to the mid-second century CE) was at the time considered the earliest copy of any portion of the Bible (superseded by the discovery of the Dead Sea Scrolls in the late 1940s). Beatty must have thought, if

Powerscourt, Sun Too Fast, 240. “I think they [University of Michigan] are paying nearly £100 per leaf and I believe I paid about £40” (Kenyon to ACB, 10 June 1932, CBP 451). The letter is in reference to Michigan’s portion of the find, of which they were negotiating exchanging or selling to Beatty fragments they were in possession of from Beatty’s portion of the find, including fragments of Acts. They were willing to sell the fragments to Beatty at cost (£80), which is what eventually happened, but for a while Beatty considered an exchange of the fragments for 3 or 4 leaves from the Pauline Epistles in his possession; suggesting he in fact paid somewhat less than £40 per leaf. 25 Beatty to Frederic Kenyon, 30 April 1935, CBP 451. Beatty was unable to come to a deal with the owner that satisfied both parties and the leaves are now at Princeton University. 26 Kenyon to Beatty, 17 March 1938, CBP 451. 23 24

Speaking in Tongues

387

the Beatty Papyri and the Codex Sinaiticus are comparable, surely so was their value. Therefore, Beatty concluded, the British Museum paid too much.

Fig. 4. Coëtivy Hours, Dunois Master, Paris, 1443–1445, CBL W 082

In this context, his waning interest in European manuscripts is easily explained. Why continue to acquire expensive European manuscripts – the Coëtivy Hours (Figure 4) was acquired for £4000 in 1919 – when more valuable and value-full deals could be had. The whole collection of Biblical Papyri were probably acquired for less than £20,000 in the 1930s. In addition, while any rich individual could outbid his rivals to acquire a desirable book at auction, Beatty took pleasure in finding deals and being ahead of the curve. Later in life he recalled with pleasure his fortuitous acquisition of his surimono (Figure 5) and nara-ehon collections: People complain that I’ve got all the surimono. If one does come on the market, it fetches a ridiculous price. But would anyone buy them more than forty years ago as I did, because I thought them beautiful, and when they cost next to nothing? Not on your life! Like those Tosa albums. People only wanted something they’d heard about so they went for the prints, and the

388

Jill Unkel

Fig. 5. Three lucky dreams: Aubergines, Totoya Hokkei, Japan, ca. 1828, CBL J 2129 Tosa painters were ignored. I think I’ve a really fine collection of those Japanese painted books.27

If too late to acquire the best European manuscripts at good value, Beatty was much luckier elsewhere. In the 1950s when rich American museums entered the market and were prepared to pay “crazy prices” for Mughal miniatures he noted, with a sigh of relief, how fortunate he was to enter that field at the right time.28 While his interest in European manuscripts faded, his acquisition of biblical and other Christian manuscripts continued. His collection no longer included Latin manuscripts as a priority (although Latin is still well-represented). Instead Christian texts in languages including Armenian, Syriac, Ge’ez, Slavonic, Coptic, and Greek, many of which were beautifully illuminated, continued to be added to his Library. James White, Director of the National Gallery of Ireland, noted in 1968, “[Beatty] and his assistants searched for authentic representation of the growth

Powerscourt, Sun Too Fast, 241–42. Beatty to Wilkinson, 3 February 1952, CBP 1130.8; Beatty to Wilkinson, 22 October 1953, CBP 1130.9. 27 28

Speaking in Tongues

389

Fig. 6. Miracles of Jesus (Ta’amra Iyasus), Ethiopia, eighteenth century, CBL W 913

of man’s capacity to express himself in China, other far eastern countries, Egypt, Persia and Greece.”29 Much of his Ethiopian collection was acquired in the 1950s and 1960s when they became available through public auctions. In fact, he started collecting them much earlier. Beatty purchased an illuminated Ethiopian prayer book (CBL W 911) in the winter of 1937 from a dealer in Cairo and that May acquired an illuminated copy of the Miracles of Jesus (CBL W 913) (Figure 6) from the Londonbased dealer Luzac & Sons. Of this “important little collection,” Beatty said, “We are very wise, I think, to cover all these ranges…if people are interested in Ethiopic art, we have an enormous number of fine miniatures scattered through the manuscripts.”30 As Beatty’s collecting evolved, as with Selection Trust’s acquisition of mines, he concentrated his library collecting on items of “rarity and interest, particularly on rare and unknown lines.”31 The Chester Beatty Biblical Papyri (Figure 7) was one of the most important biblical discoveries ever made. Its acquisition may well have been a turning point in Beatty’s collecting. Having lost interest in the popular and expensive European illuminated manuscripts he took chances on, and pleasure from, acquiring gems that were not yet understood nor desired on the open market. Beatty wanted to acquire collections representing as many cultures as he came in contact with, often using language as a measure of

Sunday Press, 21 January 1968. Beatty to R. J. Hayes, 23 March 1959, CBP 1086.3. 31 Abbas, “‘We want quality and condition.’” 29 30

390

Jill Unkel

cultural diversity. He was able to understand the value of his collection even without the ability to understand the languages. If it was deemed beautiful before, it will be again, because tastes are circular and people are predictable, so much so that he noted: I was with these Egyptologists and they showed me a papyrus they told me was from a commander-in-chief to his Pharaoh. I said: “Why, I can read this quite easily. The general is asking for many more troops and also more money.” They were a pretty surprised bunch. “We didn’t know you could read hieroglyphics!” they said. They were right. I couldn’t.32

Fig. 7. Luke from Four Gospels and Acts, Egypt, ca. 200–250 CE, CBL BP I ff.13–14

Going Forward A major part of Beatty’s stratagem was the cataloguing and publication of his collections. The publication of the Chester Beatty Biblical Papyri was planned almost immediately, with each manuscript given its own edition in a series that was produced almost annually from 1932 to 1941; a significant portion of the Coptic biblical papyri was published in the 1940s. Focus has tended to lean toward the papyri, slightly favouring New Testament texts over the Septuagint and Greek over Coptic, leaving plenty of room for additional examination of the biblical papyrus collection. Indeed, many of the numerous fragments have been virtually untouched by scholars and may yet have more to reveal. It is always possible that biblical fragments in the collection belonging to folios already published and, once identified, could be reinserted into their proper place. In addition, the museum’s biblical manuscripts are so much more than the biblical papyri.

32

Powerscourt, Sun Too Fast, 224.

Speaking in Tongues

391

As mentioned above, the languages represented in the collection are numerous and its cultural heritage diverse. Relatively little attention has been paid to the exquisite collection of illuminated (and non-illuminated) Armenian manuscripts since the publication of the catalogue in the late 1950s, although it remains a rich resource.33 Neither has the Ethiopian collection of manuscripts garnered the notice it truly deserves; its brief catalogue offers only a taste of the richness within their covers.34 In addition, no published catalogue exists for the Syriac, Greek, Coptic, and Church Slavonic parchment manuscripts, although some individual manuscripts have been published. Apart from their unique bindings (which have been catalogued but remain unpublished), Beatty’s collection of early printed books are an underutilised and untapped resource that could lead scholars in any number of directions. The print collection also offers insight into the biblical verses employed by artists when creating their devotional images and the collection of Icons has plenty to say about the nature of devotion. In fact, the biblical papyri have received the most attention of the entire collection of Christian texts, leaving so many rich veins under- or even unexplored. Sharing the collections and accessibility lie at the heart of the Chester Beatty’s ethos, and a major project to allow greater access to the museum’s treasures is well underway. The museum is committed to the digital publication of its collection and from 2017 the Chester Beatty developed an entirely new digital department. With digitisation, it will become easier for scholars to examine the manuscripts and identify unpublished texts, as well as those that deserve a second or even third look. The digital department, working in conjunction with the curatorial department, will enhance the visual catalogue of the collection as well as enable greater online access. In the autumn of 2018 the museum will launch a new website and the first phase of the Collections online, a searchable resource which will allow anyone to explore, discover, and research the collection at any time and from anywhere. This online resource will reveal the breadth of the Collection and enable the public to access its depths at the touch of a screen or the click of a mouse, both encouraging new scholarship and allowing research to start at home. Furthermore, the Chester Beatty hopes the enhanced access will aid the development of new collaborations and research projects as well as encourage both young and established scholars to delve into the depths of the biblical collections.

Shiarpie der Nersessian, The Chester Beatty Library: A Catalogue of the Armenian Manuscripts, with an Introduction on the History of Armenian Art, 2 vols. (Dublin: Hodges and Figgis, 1958). 34 Enrico Cerulli, I manoscritti etiopici della Chester Beatty Library in Dublino (Roma: Accademia Nazionale dei Licei, 1965). 33

Editing the New Testament in a Digital Age

The Future of the Critical Edition D. C. Parker The more common question to have been addressed by those involved in making digital editions is “what is the critical edition of the future?” This has now been under discussion for long enough – a good twenty years – for some of the editions of the future to have become editions of the past. I would be reluctant to address it, since I have already discussed it several times.1 The title I am addressing is a more fundamental one, for it begs the question: “does the critical edition have a future?” I shall accept that it does, on the grounds that many of us are involved in making them, with the enthusiastic support of publishers and the recognition of their importance by readers of the text. Accepting that, we move on to another important question: “what sort of a future is it?” I believe that this question lies at the heart of our understanding of the discipline of textual criticism. In considering it, I have reached the view that when we attempt to answer it, we are talking in fact not so much about the future as about the present. And to understand our present, we need equally to consider the past, and in particular the development of the way in which the critical edition has been viewed, of the expectations placed upon it. This part of my study begins with a wide historical view of opinions regarding the status of the text of critical editions, and then turns to a study of the impact of text-critical research on the critical text since the middle of the twentieth century.

The Status of the Text of Critical Editions of the Greek New Testament The following section takes a number of samples from well-known scholars. It is a small selection from a broad span of time and the choice is based on the significance of the writers and not on any attempt to be representative. 1 Most recently, the last of my Lyell lectures in Oxford back in 2011 had exactly this theme, under the title “The New Testament of the Future”: D. C. Parker, Textual Scholarship and the Making of the New Testament. The Lyell Lectures Oxford. Trinity Term 2011 (Oxford: Oxford University Press, 2011), 125–42.

396

D. C. Parker

We start three hundred years ago with a short but very important work by Richard Bentley, who set out to make an edition of the Greek New Testament to go alongside his other editions, but never completed the task. He describes his views in a letter to Archbishop Wake, which was published as Proposals for Printing a New Edition of the Greek Testament. He concludes: I am already tedious, and the post is a-going. So that, to conclude – in a word, I find that by taking 2000 errors out of the Pope’s Vulgate, and as many out of the Protestant Pope Stephens’, I can set out an edition of each in columns, without using any book under 900 years old, that shall so exactly agree, word for word, and, what at first amazed me, order for order, that no two tallies, nor two indentures, can agree better.2

We continue with S. P. Tregelles (1813–1875), who produced a highly-regarded critical edition of the Greek New Testament (1857–1872). In the Preface to his 1854 book An Account of the Printed Text of the Greek New Testament (which was the standard introduction of its day), he wrote: The more we rightly regard Holy Scripture as the charter of that inheritance to which we look forward, and which we know as given at the price of the Saviour’s blood, the more shall we be able to estimate the importance of TEXTUAL CRITICISM, by which we know, on grounds of ascertained certainty, the actual words and sentences of that charter in the true statement of its privileges, and in the terms in which the Holy Ghost gave it.3

The work which took Tregelles’ place was F. H. A. Scrivener’s Plain Introduction, which appeared in four editions between 1861 and 1894. Scrivener wrote slightly more cautiously: By collecting and comparing and weighing the variations of the text to which we have access, it aims at bringing back that text, so far as may be, to the condition in which it stood in the sacred autographs; at removing all spurious additions…at restoring whatsoever may have been lost or corrupted or accidentally changed in the lapse of eighteen hundred years. We need spend no time in proving the value of such a science, if it affords us a fair prospect of appreciable results, resting on grounds of satisfactory evidence. Those who believe the study of the Scriptures to be alike their duty and privilege, will surely grudge no pains when called upon to separate the pure gold of God’s word from the dross which has mingled with it through the accumulation of so many centuries.4

One last writing from the Victorian era: the opening words of Westcott and Hort’s The New Testament in the Original Greek (1881): 2 Bentley, Proposals for Printing a new edition of the Greek Testament and St. Hierom’s Latin Version, with a full answer to all the remarks of a late pamphleteer, By a Member of Trinity College in Cambridge (1721) in A. Dyce, ed., The Works of Richard Bentley, D.D. (London: Francis Macpherson, 1838; repr. New York: Georg Olms, 1971), 3.475–538 (here 479). 3 S. P. Tregelles, An Account of the Printed Text of the Greek New Testament (London: Samuel Bagster and Sons, 1854), viii. 4 F. H. A. Scrivener, A Plain Introduction to the Criticism of the New Testament, 4th ed. (London: George Bell and Sons, 1894), 1.5.

The Future of the Critical Edition

397

This edition is an attempt to present exactly the original words of the New Testament, so far as they can now be determined from existing documents…The office of textual criticism, it cannot be too clearly understood at the outset, is always secondary and always negative. It is always secondary, since it comes into play only where the text transmitted by the existing documents appears to be in error…[It] is always negative, because its final aim is virtually nothing more than the detection and rejection of error.5

Moving more swiftly through the next generation, we take a well-known sentence from a little book by Kirsopp Lake (1872–1946), The Text of the New Testament, which went through six editions between 1900 and 1928: “The object of all textual criticism is to recover so far as possible the actual words written by the writer.”6 Continuing into the second half of the twentieth century, we encounter – inevitably – the opening words of the book by B. M. Metzger (1914–2007), The Text of the New Testament: Its Transmission, Corruption and Restoration. It was first published in 1964, and the words appear unchanged in the fourth edition, which was co-written with Bart Ehrman and published in 2005: The necessity of applying textual criticism to the books of the New Testament arises from two circumstances: (a) none of the original documents is extant, and (b) the existing copies differ from one another. The textual critic seeks to ascertain from the divergent copies which form of the text should be regarded as most nearly conforming to the original.7

Last in this survey are taken the closing words of my first publication, an article that came out in New Testament Studies in 1977 with the title “The Development of Textual Criticism Since B.H. Streeter.” I concluded there by citing words from the first writing I quoted in this whistlestop tour, Richard Bentley’s 1721 writing Proposals for Printing a New Edition of the Greek Testament: Let us look forward, at however close or distant a date, to the situation at which Bentley hoped to have arrived, ‘so that that book, which, by the present management, is thought the most uncertain, shall have a testimony of certainty above all other books whatever; and an end be put at once to all var. lectt. now or hereafter.’8

5 B. F. Westcott and F. J. A. Hort, The New Testament in the Original Greek, corrected ed. (London: Macmillan, 1896), 1–3. 6 K. Lake, The Text of the New Testament, 6 th ed., rev. by S. New (London: Rivingtons, 1943), 1. 7 B. M. Metzger and B. D. Ehrman, The Text of the New Testament: Its Transmission, Corruption and Restoration, 4 th ed. (Oxford: Oxford University Press, 2005), xv. 8 D. C. Parker, “The Development of Textual Criticism Since B.H. Streeter,” NTS 24 (1977): 149–62, (here 166); repr. in Manuscripts, Texts, Theology. Collected Papers 1977– 2007 (Βerlin: De Gruyter, 2009), 166. This has been selected not in any view that I match the other writers quoted in significance, but because it and the next of my works which I quote provide an illustration of the development of certain new ideas which has taken place in recent decades.

398

D. C. Parker

Coming full circle, we may see that the prevailing narrative has been that textual criticism is the removal of uncertainty so as to reach a point where the “original text” may be recovered. We can see certain changes. Notably, the pious language of Tregelles and Scrivener was not present in Bentley and disappears from Hort onwards. Hort offers a sharp contrast, writing rather in the fashion of a nineteenth-century natural historian (which he also was). His style strikes one as similar to Darwin’s. The twentieth century largely follows his lead in writing without attributing specific characteristics such as divine inspiration to the New Testament text. One may also detect – and this is the important point – a growing circumspection, from Bentley’s reconstructions that “exactly agree, word for word” through to Lake’s “so far as possible.” The 1970s was a point in the twentieth century when it was appeared to many that the task of textual criticism of the New Testament was largely done. The great critical editions of Lachmann, Tischendorf, and Westcott and Hort had laid down a form of text which was broadly agreed to be largely indisputably original. It is certainly true that some researchers, such as Dean Burgon and Hoskier, argued in favour of the text found in the majority of witnesses. There were other scholars who pointed to readings in “western” witnesses as possibly authentic. But the majority of scholars, including the most influential, believed that the Hortian text was broadly correct. Given this success, what was left was a sort of hoovering up exercise, in which a small number of less significant uncertainties could be resolved. We can understand this as a reasonable point of view, if we consider the huge strides taken by the nineteenth century editors. Their achievement was to overthrow the Textus Receptus, the printed form of text which prevailed without challenge from Erasmus’ first edition of 1516 onwards. The belief that the Greek text was per se better than the Latin Vulgate meant that deficiencies in the Greek text that came to be accepted were not obvious. In any case, it was to be a century before scholars began to recognise and get regular access to older copies that helped them to understand more ancient forms of text. By older I mean predating the ninth century. Once they did, Bentley was able to present his Proposals for Printing, and editors from Bengel onwards to propose alternative readings to the Textus Receptus. It was Lachmann who first made an edition using only the oldest manuscripts. In his time this was A B C D.9 At a similar period, Tischendorf began his journeys and the great series Monumenta sacra inedita, so that many older manuscripts, most notably Codex Sinaiticus, became known to western researchers. The result was that he and Hort and others were able to make editions that used fourth and fifth century manuscripts as the basis for their editions. This represented a huge change from the Textus Receptus. Gone were the Johannine Comma, the Pericope Adulterae, the Longer Ending of Mark. Gone too were so many harmonisations and alterations 9 Codices Alexandrinus (A02), Vaticanus (B03), Ephraemi Rescriptus (C04), and Bezae Cantabrigiensis (D05).

The Future of the Critical Edition

399

in the text of Mark that the new editions produced what by comparison with the Textus Receptus was a new version of the Gospel. A new approach to the Synoptic Problem and the influential theory of the Messianic Secret were just two developments that would never have been possible using the Textus Receptus. The new research opportunities offered to New Testament scholars may have been another reason for the success of the new text. This dramatic change was, in their day, sufficient to justify the bold claim by Westcott and Hort that their edition was The New Testament in the Original Greek. As a matter of fact, the manuscript they considered the most reliable, Codex Vaticanus, was probably copied in the middle of the fourth century. So there was still a gap of three centuries between the oldest manuscript and the oldest New Testament writings. But compared to the Textus Receptus, that was an improvement by seven hundred years and more.

Twentieth Century Text-Critical Research on the Critical Text Today, the belief that the task of recovering the oldest text of the New Testament is done appears rather quaint. Indeed, the picture looks remarkably different. Twenty years after my enthusiasm for putting an end to all variant readings for evermore, I was writing quite differently. The attempt to discern earlier forms of text, from which those known to us are descended, is an essential task in the critical study of Christian origins. It does not follow that it is also necessary to recover a single original text…The quest for a single original text of the Gospels is driven by the same forces that have sought a single original saying of Jesus behind the different texts of different Gospels. Both are dubious…Is the tradition of Jesus’ sayings on divorce best understood by attempting to recover a single original text of each of the four passages, and then from those four a single original saying of Jesus? Or is it best understood by recognising the role of all the text forms as interpretations of the tradition?10

For me, this step was a decade in coming, but it had been long anticipated by other writers, to whom I could have paid more attention if I had not been as accepting as most other people of the dominant narrative. Eldon Epp has chronicled the significance of the work of the Chicago School, especially E. C. Colwell and K. W. Clark in recognising the presence of theologically motivated variation in the witnesses to the text of the New Testament.11 Another approach was taken 10 D. C. Parker, The Living Text of the Gospels (Cambridge: Cambridge University Press, 1997), 208–209. 11 E. J. Epp, “The Multivalence of the Term ‘Original Text’ in the New Testament Textual Criticism,” HTR 92 (1999): 245–81, esp. 271–75, repr. in Epp, Perspectives on New Testament Textual Criticism: Collected Essays 1962–2004 (Leiden: Brill, 2005), 551–93, esp. 581–86. See also e.g. Epp, “Anti-Judaic Tendencies in the D-Text of Acts: Forty Years of Conversation,” in The Book of Acts as Church History: Text, Textual Traditions and Ancient

400

D. C. Parker

by Ehrman in The Orthodox Corruption of Scripture.12 There are significant differences between our two approaches, but they share something in common, namely the argument that some textual variation is the result of theological discussion, and is much more interesting to study than Scrivener thought (remember that he described secondary readings as dross to be discarded once the gold has been extracted). One of the differences is that Ehrman’s approach remains polarising in that it treats one reading as original and the other as a corruption. I argued in The Living Text of the Gospels that the evidence of four gospels and significant textual variation between the manuscripts provides powerful evidence that early Christians were not that interested (perhaps even not at all) in authoritative writings and single unchanging forms of text. Rather, within the broader compass of an oral tradition and the conviction that they were inspired by God’s spirit, they sometimes changed the letter to retain what they believed to be the spirit of the idea expressed. So the textual variation remains significant for two reasons: first, because it provides important primary material in the study of early Christianity; secondly because the quest not only for the original text of the Gospels but also for the ipsissima vox Iesu cannot be made on the basis of a modern eclectic text such as the Nestle-Aland, but must continue to engage with the real multiplicity of forms in which the tradition survives. I stress the significance of work from the Chicago School onwards because the concept of a dominant narrative in scholarship is always a dangerous one to trust. In fact, research always has many potential developments within it. The picture I have offered of a clear trajectory from Bentley into the twentieth century is itself likely to overlook other ideas which may prove fruitful at another time. To return to the future of the critical edition: we can see that regardless of the detail of the argument, one of the main historical functions of the critical edition has been that, starting with a really bad text based on the witnesses at the young end of the tradition it provided a vehicle for developing some rules to make a better text, using the witnesses at the old end. The basis of this was the compilation of evidence. Once Bentley was able to study enough collations of manuscripts, he was able to set out a theory about the way of reconstructing the text of the fourth century.13 This process was essential, and in fact there have been two periods when it changed textual criticism dramatically. The first was when

Interpretations / Apostelgeschichte als Kirchengeschichte: Text, Texttraditionen und Antike Auslegungen, BZNW 120, ed. T. Nicklas and M. Tilly (Berlin: De Gruyter, 2003), 111–46, esp. 114–17, repr. in Perspectives on New Testament Textual Criticism, 699–39, esp. 703– 705. 12 B. D. Ehrman, The Orthodox Corruption of Scripture (Oxford: Oxford University Press, 1993). A second edition in 2011 is different only in an Afterword which, sadly, does not engage with the lively debate surrounding the work. 13 This depended upon the nascent discipline of palaeography, which was able to demonstrate the age of the manuscripts.

The Future of the Critical Edition

401

editors had both Codex Sinaiticus and Codex Vaticanus available.14 The most important outcome of this was the edition of Westcott and Hort. The second period began at the end of the nineteenth century, when the first papyrus copies were recovered. Subsequent decades saw the publication of one document, P52, that was dated to the middle of the second century and whose date was in various writings mysteriously moved into the second quarter of the century. Its existence was even used to counter arguments for a late date for the composition of the Gospel of John. P52 is only a scrap, but the post-war years saw the publication of two extensive copies, P66 of John and P75 containing Luke and John. The former was dated to about the year 200 and the latter to the beginning of the third century. Scholars were immediately struck by the close similarities between P75 and Codex Vaticanus. It was an easy step to claim therefore that the editorial text of Westcott and Hort, tweaked in some ways by the time that the first UBS edition came out in 1966 but still closer to Codex Vaticanus than to any other copy, could be proven to be 150 years older than had previously been thought. The gap between the oldest recoverable text and the dates of the writings had been halved. It seemed rather easy then to argue that the oldest copies must be based on similar copies, which took one back a little further. In fact by the 1970s there seemed even to be a feeling among textual critics that all the important work had been done. In effect, although there had been no new editio maior since von Soden (which because of various problems has only ever been considered useful for the large amounts of material gathered in preparation for the work), it was believed that the available hand editions were a pretty good approximation to the best that would ever be achieved. Thus, on the positive side, the critical editions had produced scientifically reasoned replacements for the haphazard Textus Receptus. Negatively, they became too successful in convincing everyone, including the editors themselves, that because they were so much better they were therefore the best and only text that could be attained. Textual criticism seemed even more of a mopping-up exercise in the light of Eldon Epp’s influential article “The Twentieth-century Interlude in New Testament Textual Criticism,” published in 1974. Epp wrote that this characterization of interlude is, on the one hand, to suggest something negative: it affirms that the critical work of the period is not a main feature, but a subsidiary or a secondary and minor performance following a

14 See E. J. Epp, “Codex Sinaiticus: Its Entrance into the Mid-Nineteenth Century TextCritical Environment and Its Impact on the New Testament Text,” in Codex Sinaiticus: New Perspectives on The Ancient Biblical Manuscript, ed. S. McKendrick et al. (London: British Library, 2015), 53–90.

402

D. C. Parker

portion of a main event. On the other hand, there is a positive aspect, for interlude implies – if not demands – that another major act is to follow. 15

The faithful few who kept the candle burning in this period were certainly hopeful that a new generation would bring a fresh impetus, and this proved the case, although the faithful few were not always happy with the directions that their successors took. The fact is that much of the textual criticism of the last quarter of a century has turned the traditional concepts of textual criticism on their head. Fundamentally, what has often been called Narrative Textual Criticism does not discard variant readings as dross once they have been shown to be secondary, but studies them as of continuing interest in understanding the different forms of text as they existed at different stages. I have followed this to one possible conclusion recently in challenging the widespread negative view of the text of the Byzantine world.16 This view was expressed by Hort in the following words: Entirely blameless on either literary or religious grounds as regards vulgarised or unworthy diction, yet showing no marks of either critical or spiritual insight, it presents the New Testament in a form smooth and attractive, but appreciably impoverished in sense and force, more fitted for cursory perusal or recitation than for repeated or diligent study.17

Against this I argue that there are three good reasons for taking the Byzantine text forms more seriously. One is the Orthodox view that the church’s text of Scripture is the text that it became, not the text as it was at the beginning. The second is that the old concept of text types with the good Alexandrian and the bad Byzantine at opposite ends of the development can no longer be sustained. Using complete transcriptions and analysing databases of variation, instead we make diagrams which display possible textual flows of the manuscript tradition. The third reason is that the view of a unified Byzantine text is rather hard to sustain in detail. Actually, the tradition is as broad as it is long – there are as many differences between Sinaiticus and Vaticanus as there are between the Initial Text and the late witnesses at the other end of the textual flow.18 15 E. J. Epp, “The Twentieth-Century Interlude in New Testament Textual Criticism,” JBL 93 (1974): 386–414, here 387; repr. in E. J. Epp and G. D. Fee, Studies in the Theory and Method of New Testament Textual Criticism (Grand Rapids: Eerdmans, 1993), 83–108, here 84; and in Epp, Perspectives on New Testament Textual Criticism, 59–100, here 60. In 1979 Epp was to write “New Testament Textual Criticism in America: Requiem for a Discipline,” JBL 98 (1979): 94–8, followed a year later by “A Continuing Interlude in New Testament Textual Criticism,” HTR 73 (1980): 131–51. Both are reprinted in the collections already mentioned. 16 D. C. Parker, “New Testament Textual Traditions in Byzantium,” in The New Testament in Byzantium, ed. D. Krueger and R. S. Nelson (Washington, D.C.: Dumbarton Oaks, 2016), 2–32. 17 Westcott and Hort, The New Testament in the Original Greek, 116, 133, 134–35. 18 For more information, see Parker, Textual Scholarship and the Making of the New Testament, 93.

The Future of the Critical Edition

403

Here we come to a significant challenge to the traditional concept of the critical edition. Once it was believed to be all about reconstructing the “original text” – the text critics doing themselves out of a job in effect, since once this “original text” had been made – what else was there to do? Instead today we see scholars with many different attitudes. One might single out Jennifer Knust and Tommy Wasserman, in their book on the Pericope Adulterae. Their approach has been to study the different forms and presentations of this passage as a window into different modes of interpretation at different epochs. In such work, the value of the critical edition is twofold. First, it is a sourcebook of information, providing the various forms in which the work has existed at different stages (or at least with a representative selection). Second, by reconstructing a critical text and in the shape of its apparatus (and in the Coherence-Based Genealogical Method with the textual flow), it provides a reconstruction of the ways in which these forms developed and how they are related to each other.

The Future of the Critical Edition So far I have suggested or implied that the critical edition has been of value (1) in replacing a very poor text with a much more systematically constructed one using a better selection of witnesses; (2) in bringing users up to date with fresh evidence; and (3) in providing the material for understanding the history of the text. I see no reason why critical editions should not continue to fulfil these functions in the future, and why they should not remain of interest to users. There are, however, further issues to be discussed regarding the concept of the critical edition and its future role. I have suggested that traditionally it has been regarded as a means of studying all the evidence to recover the original text. Today we attempt a more nuanced approach, and agree that the goal of textual criticism is to recover the Initial Text, the oldest form of text that we can reconstruct from the available witnesses using philology. But there are still those who then wish to argue that this Initial Text is virtually indistinguishable from the authorial text which we believe to lie behind it. Dependent upon the theory that the Initial Text is also the authorial text is an entire industry of scholarship, out of which one might single out the study of the Synoptic Problem and the attempt to reconstruct the hypothetical document Q as a noteworthy example. Dependent also upon the theory are various forms of belief surrounding the character of the biblical writings. Will the critical edition continue to provide certainty? Or is its role also to provide a guide to uncertainty, a confession of ignorance?19 19 For the suggestion that in contrast to most of the previous 200,000 years, human thinking is now based on the willingness to admit ignorance rather than the claim to knowledge, see Y. N. Harari, Sapiens. A Brief History of Humankind (London: Penguin, 2014), 279. For a similar view on the nature of science, see S. Firestein, Ignorance: How it Drives Science

404

D. C. Parker

What is the function of the critical edition? Is it to achieve certainty, or perhaps we should say, how far do the editors and users believe that it is achieving certainty? Is it also indicating the degree of uncertainty, the things we do not know and do not understand? Is the traditional format designed to achieve the one or the other? How far does the use of split lines in the latest volume of the ECM Acts raise our recognition of the level of uncertainty, or does it paradoxically have the effect of making us too confident in the security of all the rest? We might also in this regard consider the debate about the use in the UBS Greek New Testament of the letters A to D indicating the confidence of the editors. The level of uncertainty attracted a good deal of attention, not least for the high proportion of C ratings, and continues to be a subject for discussion. I simply invite consideration as to why that was. Were readers expecting a stronger guidance from the editors? What may different users want of the critical edition in the future? Will they want it to be a piece of provisional research pointing to new research questions and fresh confessions of ignorance and new ways of making further critical editions? Or to be answers complete in themselves? I cannot answer that for other people. I do know that given the tiny number of manuscript copies to have survived and the difficulties of developing and using ways of comparing them, for me accepting ignorance has become more important. Finally, I turn to the growth in complexity of the critical edition of the Greek New Testament, simply to draw attention to the matching growth in the number of people involved. Compared to the single name on the title page of a Tregelles or a Tischendorf, or even the duo Westcott and Hort, the title page of the ECM includes half a dozen names. The editions of the International Greek New Testament Project (IGNTP) list many contributors. And the members of these complex teams are drawn from many places, cultures, and countries. The IGNTP was in its founding already a transatlantic organisation. Now in its partnership with Münster it works together with scholars from across Europe. For some, the process of working on it is as strong a justification for the modern critical edition of the New Testament as any other. For Research Councils, a work which brings together researchers from different places and enables them to exchange ideas, methods and ways of thinking can be in itself a justification for funding it. The benefits go beyond the task in hand. It is not hard to see that human beings can achieve far more working together than they can in isolation, and it is hard not to believe that in its social complexity the ECM, to take but one example, will continue to prove this. To conclude: recent decades have seen the growth of new theories of textual criticism, and it is no longer possible to maintain that the critical edition has a (Oxford: Oxford University Press, 2012), 17: “Being a scientist requires having faith in uncertainty, finding pleasure in mystery, and learning to cultivate doubt. There is no surer way to screw up an experiment than to be certain of its outcome.”

The Future of the Critical Edition

405

single purpose, or that its users will be agreed on its status or character. Instead it may function as guide to the history of the text; reconstruction of one or more possible forms of text older than others, whose relation to an authorial text is disputed and in any case is likely to vary from work to work; a statement of the forms in which the text has existed at different times; a source for further research of various kinds of information. If this is true, then it is even more certain that the critical edition has a future, since if one use fails another will remain.

Old Wine, New Wineskins: Digital Tools for Editing the New Testament Catherine Smith The use of computers in New Testament textual criticism goes back almost as far as the first use of a computer for any kind of text processing. That first project is generally agreed to be the Index Thomisticus begun in 1949 by Roberto Busa.1 Busa’s own account of the work has become something of a founding myth of digital humanities.2 Part of the story recounts his interaction with another pioneer of the discipline, John W. Ellison. Ellison was the first person to use a computer to address problems of New Testament textual criticism and began his work only a year after Busa.3 While technology has changed considerably in the intervening decades, the themes and motivations that were key in Ellison’s work are still present in the discussion of our use of digital tools today. This chapter discusses the development of computing in the discipline from this early adoption to our increased reliance on it today. It examines the growing expectations of software and some of the challenges that must be addressed if our data and tools are to continue to adapt to the ever-changing digital environment.

1 Susan Hockey, “The History of Humanities Computing,” in A Companion to Digital Humanities, ed. S. Schreibman, R. Siemens, and J. Unsworth (Malden: Blackwell, 2004), 4; Steven E. Jones, Roberto Busa, S.J., and the Emergence of Humanities Computing: The Priest and the Punched Cards (New York: Routledge, 2016). 2 Roberto Busa, “The Annals of Humanities Computing: The Index Thomisticus,” Computers and the Humanities 14/2 (1980): 83–90; Jones, Roberto Busa, 1–3. 3 Robert A. Kraft, “The Use of Computers in New Testament Textual Criticism,” in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis, ed. B. D. Ehrman and M. W. Holmes (Grand Rapids: Eerdmans, 1995), 268.

408

Catherine Smith

A Selective History of Electronic Editing The magic of myth and legend has come true in our time. One types the correct incantation on a keyboard, and a display screen comes to life, showing things that never were nor could be (1975).4

Ellison’s doctoral work focused on New Testament textual criticism but he is probably best known for his work on the first computer-generated concordance of the Bible, and it is in this role that he and Busa crossed paths.5 They began their work independently and, as far as I can tell, each without knowledge of the other; as Busa himself says, these ideas tend to arise from the milieu.6 They were also working with different technologies. Busa was using IBM punch card machines with all data entry and manipulation performed using punched cards.7 He initially estimated that his project would require a total of 13 million punched cards (weighing about 500 tonnes and taking up 108 cubic metres of space). Ellison used the Remington Rand Univac machine and, while he did use punched cards for some of the initial input verification, the main data entry, processing instructions, and storage all used magnetic tape.8 The story goes that Busa read

4 Frederick P. Brooks Jr., The Mythical Man-Month: Essays on Software Engineering, anniversary ed. (Boston: Addison-Wesley, 1975); (repr., 1995 with four new chapters), 7– 8. 5 Meredith Hindley, “The Rise of the Machines. NEH and the Digital Humanities: The Early Years,” Humanities: The Magazine of the National Endowment for the Humanities 34/4 (2013), https://www.neh.gov/humanities/2013/julyaugust/feature/the-rise-the-machines [accessed 9 March 2018]. Depending on whether or not punched card machines are classified as computers, Ellison could have been the first person to produce the first computer generated concordance of anything. Oakman cites Ellison’s concordance as the starting point for humanities computing (Robert L. Oakman, Computer Methods for Literary Research, 2nd ed. [Athens: University of Georgia Press, 1984], ix), which is also noted by Edward Vanhoutte, “The Gates of Hell: History and Definition of Digital | Humanities | Computing,” in Defining Digital Humanities : A Reader, ed. M. Terras, J. Nyhan, and E. Vanhoutte (Farnham: Ashgate, 2013), 131. Hindley specifically credits Ellison with making “the crucial jump from accounting machines to computers” (Hindley, “The Rise of the Machines”). Along these lines Clair M. Cook, “Automation Comes to the Bible,” Computers and Automation 7/3 (1958): 18 (repr. of the same article in Christian Century 54/30 [1957]) reports that Ellison experimented with punched card machines for several years before deciding they were not suitable for his text-critical work and turning to the electronic computer. 6 Busa, “The Annals of Humanities Computing,” 84. 7 Busa worked in collaboration with H. Paul Tasman, a Senior Engineer at IBM. Jones, Roberto Busa, 18. 8 “Bible Labor of Years is Done in 400 Hours,” Life (February 1957): 92. Gardner Soule, “The Machine that Indexed the Bible,” Popular Science (November 1956): 174. Ellison’s work on the RSV concordance was completed alongside Remington Rand engineers John Graham (algorithm design) and Al Bosgang (programming).

Digital Tools for Editing the New Testament

409

about Ellison’s work in a newspaper which led to the two meeting.9 Busa’s own account is as follows: One day I learned from the newspapers that an Episcopalian minister, Rev. John W. Ellison who was preparing a concordance of the Revised Version of the Bible…had used Remington magnetic tapes…I went to shake hands with him and said: “You are a great ally of mine!” Immediately after I went to IBM: “See what Remington is doing?” Since that time the processing of the IT [Index Thomisticus] has been done mainly by computers and punched card equipment was used only peripherally. 10

Recent research by Steven Jones suggests that the relationship between the two was more significant than this summary account implies. The Busa archive has an entire folder relating to Ellison which contains documents and correspondence that illustrate what Jones describes as “a true intellectual collaboration” including sharing details of their methodologies.11 Ellison’s concordance of the RSV was completed in 1955 but it took two years to typeset and was published in 1957.12 Busa’s completed concordance of Thomas Aquinas’ writings was published in 56 volumes between 1974 and 1980 and only required six million punched cards thanks to the use of magnetic tape.13 9 Thomas Nelson Winter, “Roberto Busa, S.J., and the Invention of the Machine-Generated Concordance,” The Classical Bulletin 7/1 (1999): 13 suggests that this meeting might have happened in 1954, and he is followed in that by Vanhoutte ( “The Gates of Hell,” 127). Winter seems to assume this date from a subscript in a paper by Busa, which lists a number of articles in the press that relate to Ellison’s work, the earliest being from 1954 (Roberto Busa, “The Use of Punched Cards in Linguistic Analysis,” in Punched Cards: Their Applications to Science and Industry, ed. R. S. Casey et al., 2nd ed. [New York: Reinhold Publishing Corporation, 1958], 373). However, this article in Time (“According to Mark IV,” Time [9 August 1954]) concerns Ellison’s doctoral work on Textual Criticism and does not mention Remington Rand (since he was using the Harvard Mark IV), magnetic tapes, or his work on the concordance. Recent research by Jones using the Busa archive suggests that the men had no contact until 1956 with the initial contact made in May of that year, their correspondence continuing through the Autumn (Jones, Roberto Busa, 101). There were a few articles in the press in early 1956 that could have been the trigger for contact. A very brief announcement of the concordance project appeared in the New York Times in February, which mentions the Univac (“Bible Index Computed,” The New York Times, 17 February 1956), and a technically detailed article in the IBM Systems magazine in March–April (William R. McCully, “Univac Compiles a Complete Bible Concordance,” Systems Magazine [March– April 1956]: 22–3). 10 Busa, “The Annals of Humanities Computing,” 85. 11 Jones, Roberto Busa, 101. 12 The typesetting was not done by the Univac. The output was printed with a Uniprinter (which took about 1,000 hours) and the pages were then sent to the typesetters who “follow[ed] the usual procedure for the manufacture of a book” according to McCully, “Univac Compiles a Complete Bible Concordance,” 23. See John W. Ellison, Nelson’s Complete Concordance of the Revised Standard Version Bible (New York: Thomas Nelson, 1957). 13 Roberto Busa, “Concluding a Life’s Safari from Punched Cards to World Wide Web,” in The Digital Demotic: Selected Papers from Digital Resources in the Humanities 1997,

410

Catherine Smith

Ellison’s work on textual criticism started before his work on the concordance. In 1950, Ellison approached Howard Aiken at the Harvard Computational Laboratory to ask for processing time on the university computer for his project, apparently the first humanities scholar to do so.14 In 1951 Ellison was granted time on the (still not quite operational) Harvard Mark IV computer and began learning how to program it.15 He had a vision to transform the way data was collected, processed, and shared by textual critics. He observed that while “[t]he differences between any given pair of manuscripts are finite and immutable”16 each scholar who studied a manuscript started from scratch by comparing it to a standard printed text and a selection of other manuscripts before publishing the results.17 He suggested that, in order to deal with the large volume of New Testament manuscripts, data could instead be collected collaboratively according to agreed methods and categories and stored in look-up tables akin to the logarithm tables used for mathematics.18 He also wanted to see whether using all the points of variation in a small selection of verses could find the same family groupings that had been established by traditional methods using the “significant” readings from the full text.19 The ideas behind this computer-assisted method were that the results would be reproducible across a range of New Testament texts and that the work could be built on by adding the data from other manuscripts to those already studied without having to start the entire comparison process again. In keeping with his principles of data reuse Ellison did not make any manuscript collations himself, but instead obtained permission to use those being prepared by the American committee of the IGNTP for their edition of Luke.20 In addition to the 81 IGNTP collations, Ellison also used 200 collations of Luke 10:1–15 prepared by Merrill Parvis and it is this additional data that determined the selection of these 15 verses for his research.21 Each of these collations would have recorded, for a Office for Humanities Communication Publication 10, ed. L. Burnard, M. Deegan, and H. Short (London: Office for Humanities Communication, 1998), 3. 14 John W. Ellison, “Computers and the Testaments,” in Computers in Humanistic Research, Readings and Perspectives, Prentice-Hall Series in Automatic Computation, ed. E. A. Bowles (Englewood Cliffs: Prentice-Hall, 1967), 163. 15 While Busa and Ellison both worked with programmers on their concordances, Ellison wrote his own algorithms for his text-critical work. Ellison, “Computers and the Testaments,” 163–64. 16 John W. Ellison, “The Use of Electronic Computers in the Study of the Greek New Testament Text” (PhD diss., Harvard Divinity School, 1957), summary 6. Ellison’s thesis is quoted in this article by the kind permission of his next of kin. 17 Ellison, “Computers and the Testaments,” 162. 18 Ellison, “The Use of Electronic Computers,” 124, summary 6. 19 Ellison, “Computers and the Testaments,” 162. 20 This was later published as IGNTP, The Gospel according to Luke, The New Testament in Greek 3 (Oxford: Clarendon, 1983–1987). 21 Ellison, “The Use of Electronic Computers,” 63–4, 67.

Digital Tools for Editing the New Testament

411

single manuscript, all of the places where its text differed from the Textus Receptus (TR).22 Ellison encoded all of the differences in what was effectively a spreadsheet:23 the first four columns pinpointed the unit number and the position of the variant within the unit.24 The fifth column identified the variant reading and the final two columns were used to classify the variants into one of eight categories: substitutions, omissions, additions, inversions, spelling of proper nouns, changes of inflectional form, itacisms, and other spelling variations.25 The computer was then used to compare each pair of manuscripts to see whether they agreed or disagreed in their deviation from or agreement with the TR, and to count the number of times they did in each of the variant categories. The output from the computer took the form of a table for each manuscript showing the number of differences between it and all of the other manuscripts subdivided by the nature of the variation.26 By analysing the patterns of agreements and disagreements in these tables Ellison found that 307 of the 309 manuscripts used in his study fell into the same groupings established by the older methods and that the “‘insignificant’ readings seemed to reveal as much as the ‘significant’ ones when they were all studied.”27 Ellison’s work is often cited in literature discussing the use of computers for textual criticism,28 but it does not seem to have been taken into consideration in the development of the profiling methods that were later developed for New Testament manuscripts,29 perhaps to Ellison’s frustration. Several years after the The IGNTP used reproductions of an edition published by Oxford University Press in 1873. 23 The technology available meant that it was actually encoded and stored as a seven digit number. Ellison, “The Use of Electronic Computers,” 43. 24 The section of text had been split into 65 units by Parvis, and this was adopted by Ellison for his input scheme. Ellison “The Use of Electronic Computers,” 67. 25 Ellison also had a system for recording variations within variations which is left out here for simplicity. Ellison, “The Use of Electronic Computers,” 42–4. 26 Ellison, “Computers and the Testaments,” 164. The data calculated by Ellison, and on which his conclusions are based, is essentially the same data that would be used in the Coherence Based Genealogical Method (CBGM) to establish pregenealogical coherence; the categories used by Ellison would even allow spelling variants to be excluded from the calculation. See Tommy Wasserman and Peter J. Gurry, A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method (Atlanta: SBL, 2017), 27– 42. 27 Ellison, “Computers and the Testaments,” 165–66. 28 Kraft, “The Use of Computers,” 268, 274; Bruce M. Metzger and Bart D. Ehrman, The Text of the New Testament: Its Transmission, Corruption, and Restoration, 4 th ed. (Oxford: Oxford University Press, 2005), 240. 29 It is not possible to say for certain why this is the case. Part of the agreement made with the IGNTP regarding use of their data was that Ellison was not allowed to publish his results until the edition of Luke had been published (Ellison, “The Use of Electronic Computers,” 63–4). In 1950, around the time Ellison was starting his work the expected publication date for the Luke volume was 1954 (American Editorial Committee of the IGNTP, 22

412

Catherine Smith

completion of his doctoral work he writes that “the scholars who are most involved in textual studies at present feel that the subjective intuition of the scholar is going to make the greatest discoveries. They are still at work with the older methods, essentially counting on their fingers!”30 Nevertheless, the themes present in Ellison’s work (collaboration, agreed standards, data sharing and reuse, and computational techniques) are still key to the IGNTP today in its work towards the Editio Critica Maior (ECM). Computers have played an important part in the preparation of the ECM since its inception, at least as early as 1970.31 There are definitely similarities with Ellison’s work but I have not been able to establish a direct line of influence. Under the direction of Kurt Aland at the Institut für Neutestamentliche Textforschung (INTF), and with support from Wilhelm Ott from the Tübingen computing centre and Bonifatius Fischer from the Vetus Latina Institute, work began on selecting the manuscripts for inclusion in the ECM based on collations of selected test passages.32 Although the manual selection of passages was what Ellison was trying avoid, the computational task was very similar. The selected passages were manually collated and the results were then encoded for processing by the computer. Each variant reading for each of the 1,000 selected passages had a punched card which contained the identification number for the

“Proposed Publication of the Manuscript Evidence for the Text of the Greek New Testament,” The Bible Translator 1 [1950]: 171), but the volume was not actually completed until 1987 which perhaps means that Ellison’s work was not as widely known as it might otherwise have been. An article originally published in the same year Ellison was awarded his doctorate seems to suggest that the method Ellison used was copyrighted; Cook, “Automation Comes to the Bible,” 16 notes that “Mr Ellison’s copyrighted Methods of Using Digital Computers.” This fact also may have hampered its further use, although this seems to go against Ellison’s overall philosophy. Of course it could also be the case that the discipline was simply not ready to trust this approach, in particular the use of a full section of the text rather than passages carefully selected by experienced scholars. However, doctoral research by Andrew Edmondson at the University of Birmingham suggests that Ellison may have been correct in his proposition. Using phylogenetic software, Edmondson has shown that family groups can be established using collation data from chapters that do not contain any of the readings considered to be significant in defining the family group. See Andrew Edmondson, “Discovering New Family Trees of Biblical Manuscripts,” Poster presented at the University of Birmingham Research Poster Conference (16 June 2015). 30 Ellison, “Computers and the Testaments,” 166. 31 Bonifatius Fischer, “The Use of Computers in New Testament Studies, with Special Reference to Textual Criticism,” JTS 21/2 (1970): 307–308; Kurt Aland, “Novi Testamenti Graeci Editio Maior Critica,” NTS 16/2 (1970): 163–77. 32 There is no published record of exactly how these passages were selected. Richards describes the details he managed to find in W. Larry Richards, “An Analysis of Aland’s Teststellen in 1 John,” NTS 44 (1998): 27–30.

Digital Tools for Editing the New Testament

413

passage, the text of the reading and all of the witnesses that attest the reading.33 As with Ellison’s work, the computer was used to calculate the similarities between the text of the different manuscripts and these calculations form the basis of the selection of witnesses for inclusion in the ECM.34 A similar workflow was originally planned for the preparation of the final apparatus, although by the time the processing was actually done, desktop computers and databases were available and were used instead.35 Several computer programs existed by the early 1970s which could handle machine collation of texts, but none could have coped with the complexity of the New Testament tradition.36 As early as the 1950s, Ellison proposed a system in which several component collation programs could be linked together and run in a single operation.37 Ott took a similar approach in the 1970s,38 but it was not really until personal computers began to take over from main frames in the 1980s that the idea of fully-integrated software support for critical editing could be considered. Ott suggests that it is important “that the software used [1] cover all the steps of critical editing [2] is flexible enough to be adapted to the user’s respective requirements [and 3] be a tool that can be handled easily and safely by people who are…non-programmers.”39 33 Wilhelm Ott, “Computers and Applications in Textual Criticism,” in The Computer and Literary Studies, ed. A. J. Aitken, R. W. Bailey, and N. Hamilton-Smith (Edinburgh: Edinburgh University Press, 1973), 201. 34 For examples of the output formats see Ott, “Computers and Applications,” 207–23. The full results of the analysis are also published in the volumes of Text und Textwert. The data processing of the test passages was begun by Ott at the Tübingen computing centre, but was later transferred to the Münster computer centre with the programming completed by Annette Benduhn-Mertz under the academic direction of Gerd Mink. See Kurt Aland, “Vorwort,” in Text und Textwert der griechischen Handschriften des Neuen Testaments, I. Die Katholischen Briefe, Band 1: Das Material, ed. K. Aland (Berlin: De Gruyter, 1987), VII. 35 Ott, “Computers and Applications,” 202; Fischer “The Use of Computers,” 307; Klaus Wachtel, “Editing the Greek New Testament on the Threshold of the Twenty-First Century,” Literary and Linguistic Computing 15/1 (2000): 43–4 and personal communication. 36 For a summary of the programs available, see Penny Gilbert, “Automatic Collation: A Technique for Medieval Texts,” Computers and the Humanities 7/3 (1973): 139–47; for an idea of some of the problems with the programs see Robert L. Oakman, “The Present State of Computerized Collation: A Review Article,” Proof: The Yearbook of American Bibliography and Textual Studies 2 (1972): 333–48. 37 Ellison, “The Use of Electronic Computers,” 123–24 38 Julianne Nyhan and Andrew Flinn, “The University Was Still Taking Account of Universitas Scientiarum: Wilhelm Ott and Julianne Nyhan,” in Computation and the Humanities: Towards an Oral History of Digital Humanities, ed. J. Nyhan and A. Flinn (Berlin: Springer, 2016), 66–7. 39 Wilhelm Ott, “Software Requirements for Computer-aided Editing,” in Editing, Publishing, and Computer Technology : Papers given at the Twentieth Annual Conference on Editorial Problems, University of Toronto, 2–3 November 1984, ed. S. Butler and W. P. Stoneman (New York: AMS Press, 1988), 103.

414

Catherine Smith

Peter Shillingsburg suggests similar system requirements, writing that a good computer assisted process should be “an integrated system in which the output from any one stage of computer work is usable as input for the next stage”; that it should cover all stage of editing from examining source texts to typesetting and printing; that all the mechanical and repetitive work is done by the computer; and that “every stage in the process is interruptible, reviewable, revisable, and, if necessary, repeatable.”40 This last point is equivalent to Ott’s second point concerning flexibility, as Shillingsburg emphasises: Editing is an art, not a science, and computers are only machines. As editors, we do not want our decisions to be influenced by the limitations of the tools we use. Any artist or craftsman worth his salt who lacks a tool to do a certain thing gets or makes one that will. 41

To summarise: what is required is a fully integrated tool which is completely flexible and configurable by each editor to suit their particular editorial preferences without the need to do any programming, and that can deal with any textual tradition and support all stages of the editorial process from analysing the source texts to typesetting the final apparatus. That is a wonderful concept, but (although not everyone has given up on it) it is an ideal, which I and others do not consider to be achievable.42 Pierazzo suggests that “it is impossible to develop a single framework that will satisfy all the possible use cases that characterize the variegated panorama of textual scholarship.”43 And, [w]hile it is (relatively) simple to model the way one scholar works, or even a small group, it is much more complicated to model the working methods of a large community of textual scholars, due to the different theoretical approaches to their workflow, the different types of editorial product they aim to produce, and their national and disciplinary habits and idiosyncrasies.44

Joris van Zundert suggests that such frameworks are “waiting for a horde of uniformly behaving humanities scholars that will never come”45 and that “the very wish to cater to everyone pushes the designers toward generalization, and thus necessarily away from delivering data models specific enough to be useful to

40 Peter L. Shillingsburg, Scholarly Editing in the Computer Age : Theory and Practice, 3rd ed. (Ann Arbor: University of Michigan Press, 1996), 138–39. 41 Shillingsburg, Scholarly Editing in the Computer Age, 139. 42 Peter Robinson’s Textual Communities platform seems to be aiming for something along these lines. See Peter Robinson, “Some Principles for Making Collaborative Scholarly Editions in Digital Form,” Digital Humanities Quarterly 11/2 (2017): §38. 43 Elena Pierazzo, Digital Scholarly Editing: Theories, Models and Methods (Farnham: Ashgate, 2015), 116. 44 Pierazzo, Digital Scholarly Editing, 111. 45 Joris van Zundert, “If You Build It, Will We Come? Large Scale Digital Infrastructures as a Dead End for Digital Humanities,” Historical Social Research/Historische Sozialforschung 37/3 (2012): 165–66.

Digital Tools for Editing the New Testament

415

anyone.”46 This tension is commonly found when computers are introduced into the humanities. It is difficult to strike a balance between the usability and reusability of technical components and the unique character and requirements of each project, between generalizability and specificity,47 between standards and innovation. In a book review from 1980, Lou Burnard observes that one of the first discoveries “the novice in the art of literary computing always makes is that it has all been done before. The reinvention of the wheel, and its proud presentation at international symposia, is a familiar event.”48 And that is, of course, exactly what I am doing now. The paradigm shift for the IGNTP and the ECM happened long before any of the interfaces discussed below existed. Since 1997 the editorial teams have been committed to a digital workflow based on full text transcriptions. For transcription and collation they used Collate2, one of the few integrated systems that attempted to strike the balance between the general and the specific.49 In 1997 Collate had been under development for about a decade and Collate2 had been released the previous year, yet it still had to be extended to accommodate the requirements of the New Testament.50 With support from Peter Robinson, the INTF team successfully used Collate2 to produce the ECM of 1 John (2003), the other Johannine Epistles and Jude (2005), and Acts (2017).51 When it became clear that Collate2 would need to be replaced, a number of projects were initiated for the purpose. CollateX was developed as part of the Interedition project to replace the core collation functionality, and the Workspace for Collaborative Editing project addressed the interface elements required for the ECM.52

van Zundert, “If You Build it,” 172. Max Kemman and Martijn Kleppe, “User Required? On the Value of User Research in the Digital Humanities,” in Selected Papers from the CLARIN 2014 Conference, October 24–25, 2014, Soesterberg, The Netherlands (Linköping: Linköping University Electronic Press, 2015), 71. 48 Lou D. Burnard, “At Home with the Hardware,” The Times Literary Supplement 4024 (1980): 533. 49 Peter Robinson, Collate: Interactive Collation of Large Textual Traditions, Version 2 (Computer Program distributed by the Oxford University Centre for Humanities Computing: Oxford, 1994). Wilhelm Ott’s TUSTEP would be another example. See https://tustep.unituebingen.de/tustep_eng.html [accessed 25 July 2018]. 50 D. C. Parker, “Electronic Religious Texts: The Gospel of John,” in Electronic Textual Editing, ed. L. Burnard, K. O’Brien O’Keeffe, and J. Unsworth (New York: Modern Language Association of America, 2006), 202–203. 51 H. A. G. Houghton et al., “The Editio Critica Maior of the Greek New Testament: Twenty Years of Digital Collaboration,” Digital Philology (forthcoming). 52 Interedition was funded by a European Science Foundation COST action from 2008– 2012 and involved many partners including ITSEE and the INTF. The Workspace for Collaborative Editing was jointly funded by the AHRC and the DFG from 2010–2013. 46 47

416

Catherine Smith

In what follows I will focus on the transcription and collation stages of the digital workflow, since these are the stages that were most transformed by its adoption.53 When the IGNTP produced its edition of Luke’s Gospel transcription and collation were a single task. The only words of a manuscript that were transcribed were those that differed from the text with which they were being collated.54 These 250 or so collations then had to be combined to make the edition, not an easy task on paper even with some actual cut-and-paste.55 Since 1997, there have been two distinct stages: first, each manuscript has been transcribed in full (although this actually entails adapting a base text to reflect the text in the witness) and then all of the transcriptions are automatically collated and the result edited for publication. The tools used to support these two stages of the workflow will be discussed in this paper. Each has taken a different approach to resolving the tension between the editorial task and the computer. Their introduction has not changed the workflow; they are just new wineskins for the same old data.

Transcription The use of computers in classics, although it has made great strides, is still faced with the fact that before almost any kind of work can be done, a corpus must be prepared so that a machine can understand it; optical scanners hold hope for the future, but at the present, we are still at a stage where much manual labor is necessary. It is to be hoped that duplication of such comparatively tedious work can be avoided as far as possible, leaving time free for developing more interesting and imaginative applications of computers to classical studies (1968).56

Transcription is still the first task for many digital editing projects. While the optical scanners predicted above are now widely used for printed texts, nearly 50 years on from this quotation the situation for manuscripts is not radically different.57 Data entry is at least easier now than it once was. If errors were made in

53 For a description of the full workflow see H. A. G Houghton and C. J. Smith, “Digital Editing and the Greek New Testament,” in Ancient Worlds in Digital Culture, Digital Biblical Studies 1, ed. C. Clivaz, P. Dilley, and D. Hamidović (Leiden: Brill, 2016), 110–27. 54 For a detailed description of how manual collations are made see D. C. Parker, An Introduction to the New Testament Manuscripts and Their Texts (Cambridge: Cambridge University Press, 2008), 95–100. 55 Ulrich Schmid, “Transmitting the New Testament Online,” in Text Comparison and Digital Creativity: The Production of Presence and Meaning in Digital Text Scholarship, ed. W. Th. van Peursen, E. Thoutenhoofd, and A. van der Weel (Leiden: Brill, 2010), 193– 95. 56 Stephen V. F. Waite, “Computers and the Classics,” Computers and the Humanities 3/1 (1968): 29. 57 The READ project (https://read.transkribus.eu) is working with many partners to improve Handwritten Text Recognition (HTR) using the tool Transkribus. While this has potential to change the situation in the future, the large amount of training data required means

Digital Tools for Editing the New Testament

417

transcriptions on paper tape, the options available for correcting them were very similar to those used by scribes in manuscripts. The typists could either move back to the problem character and overwrite it with a special delete character that the computer would ignore (rather like adding a dot above a letter) or type a special character at the end of the line to show that the whole line should be ignored, followed by a number to ignore multiple lines (rather like the inverted commas that can be found in some manuscripts marking complete deletion).58 With electronic data entry and storage the practical challenges of transcribing have eased, but at the same time our expectations of transcriptions, and by extension transcribers, have grown. TEI has become the de facto means of encoding text transcriptions in the humanities and it encapsulates the tension between standardisation and flexibility. People have pointed out that – despite its subtitle – the flexibility allowed within the TEI guidelines means that it does not work as a format for interchange or interoperability and also makes it complex to use as an encoding tool.59 The TEI guidelines encompass such a wide range of possibilities that, as John Lavagnino says, if you chose the TEI for your project, that is not the end of choice; “there are further questions about exactly how you use it.”60 Not only must each project decide what features of their manuscripts to encode but also define the subset of TEI they will use to encode them and how that subset is to be applied. Andrews describes an Interedition-sponsored thinktank at the 2011 TEI Members’ Meeting about interoperability within the TEI. She reports: It became clear over the course of the meeting that flexibility and customizability is currently much more important to textual scholars than the sort of standardization that would allow for true progress toward digital critical editions.61

TEI is often referred to as a standard. Indeed, this is mentioned in the first of the Poughkeepsie principles which formed the basis of the TEI, which states: “The guidelines are intended to provide a standard format for data interchange in

it is unlikely to be a practical solution for the New Testament where we already have a reliable text on which to base our transcriptions. 58 Wilhelm Ott, “Transcription and Correction of Texts on Paper Tape. Experiences in Preparing the Latin Bible Text for the Computer,” Revue Informatique et Statistique dans Les Sciences Humaines 2 (1970): 57–8. 59 See for example Desmond Schmidt, “Towards an Interoperable Digital Scholarly Edition,” Journal of the Text Encoding Initiative 7 (2014), https://journals.openedition.org/jtei/979 [accessed 9 March 2018] and Martin Holmes, “Whatever Happened to Interchange?” Digital Scholarship in the Humanities 32 suppl_1 (2017): i63–i68. 60 John Lavagnino, “When Not to Use TEI,” in Electronic Textual Editing, ed. L. Burnard, K. O’Brien O’Keeffe, and J. Unsworth (New York: Modern Language Association of America, 2006), 335. 61 Tara L. Andrews, “The Third Way,” Variants 10 (2013): 63.

418

Catherine Smith

humanities research.”62 However the ability for scholars to customise the guidelines is also one of the founding design principles.63 Ide and Sperberg-McQueen (an editor of the initial TEI guidelines), describe the situation as follows: The TEI made an early commitment to formulating its encoding scheme as a set of guidelines, rather than as a standard, for encoding literary and linguistic materials. This reflects first of all a commitment to preserving the intellectual autonomy of researchers who encode texts electronically: by constraining the allowable forms and intellectual content of electronic texts, a strictly normative standard would risk constraining their ability to reflect the particular intellectual commitments of individual researchers, and thus in some ways the types of intellectual work which can conveniently be undertaken. A wholly permissive encoding scheme, on the other hand, risks failing to provide a usable basis for the sharing and reuse of expensive textual resources. It also encourages pointless variation in encoding methods arising not from different intellectual approaches to textual research, but from merely random differences in the use of the scheme.64

They go on to say that “[t]he TEI has tried to steer a middle course.”65 It could be argued that this middle course has still failed when it comes to interchange, but the breadth and flexibility offered by the guidelines is surely the reason that TEI has become so widely used. Before the ECM teams began using TEI, the importance of flexibility within agreed parameters as a means to facilitate collaboration had already been recognised, as David Parker notes: The use of agreed computer methodologies makes it possible for separate projects to work together without losing their identity. Thus, IGNTP and INTF have been able to share certain common objectives without losing their individuality. This balance is important, for if everyone did things exactly the same way, textual editing would be greatly impoverished.66

While the customisation options of the TEI are an important factor in its popularity within the humanities, they do cause problems for the creation of generic tools and software that can support its use.67 The use of XML as the current expression of TEI means that all of the standard XML editing tools can be used: these tools can make XML easier to work with by providing syntax highlighting and validation support. However, using them still requires a very high level of XML and TEI knowledge and a lot of discipline on the part of the transcriber.

62 James Cummings, “The Text Encoding Initiative and the Study of Literature,” in A Companion to Digital Literary Studies, ed. R. Siemens and S. Schreibman (Oxford: Blackwell, 2008), 453. 63 Cummings, “The Text Encoding Initiative,” 472. 64 Nancy M. Ide and C. Michael Sperberg-McQueen, “The TEI: History, Goals, and Future,” Computers and the Humanities 29/1 (1995): 8. 65 Ide and Sperberg-McQueen, “The TEI,” 8. 66 Parker, “Electronic Religious Texts,” 205. The importance of the electronic workflow, in particular the transcription format, for aiding collaboration is echoed in Wachtel, “Editing the Greek New Testament,” 48. 67 Pierazzo, Digital Scholarly Editing, 112, 119–20.

Digital Tools for Editing the New Testament

419

When a critical edition is created electronically, the transcriptions provide the foundation on which everything else is built: it is impossible to create a good quality edition from poor quality transcriptions. However, as this is one of the most time consuming phases it is also the most likely to be crowd-sourced and that requires careful consideration of the tools used. Transcribe Bentham, the flagship crowd-sourcing project in this area, has been actively engaging volunteers to make transcriptions of Bentham’s writings, including many unpublished manuscripts, since 2010.68 While not compulsory, the use of basic TEI markup in the transcriptions is encouraged and a transcription toolbar is provided to help users add the appropriate tags.69 The toolbar makes it easier for transcribers to add tags and the team found that many people did choose to add the TEI annotation, in particular the “super-transcribers” (the small percentage of volunteers that produce the vast majority of the work).70 However, they also found that the visual presence of the tags in the transcription interface was putting off potential volunteers. Consequently, as part of the READ project, the Transcribe Bentham team are working on a “what-you-see-is-what-you-get” (WYSIWYG) editor to replace the existing toolbar so that in future the XML tags will be hidden.71 They predict that having such an editor will increase the rate at which transcriptions are made as well as improving recruitment and retention of volunteers and therefore the investment in creating the editor will be worth it.72 This idea of a payoff of investment is an important point and goes hand in hand with the tension around specific versus general requirements. The use of the toolbar and the introduction of the WYSIWYG editor necessarily restricts the tags that can be used but, for Transcribe Bentham, the restriction of tags and the time invested to make the editor are both compromises worth making in order to achieve their goals. Another system currently in development that takes a very different approach to transcription is the Textual Communities platform.73 Textual Communities is Details of the Transcribe Bentham project can be found at http://blogs.ucl.ac.uk/transcribe-bentham [accessed 4 March 2018]. 69 Tim Causer and Melissa Terras, “Crowdsourcing Bentham: Beyond the Traditional Boundaries of Academic History,” International Journal of Humanities and Arts Computing 8/1 (2014): 52. 70 Causer and Terras, “Crowdsourcing Bentham,” 49, 52. 71 Tim Causer and Valerie Wallace, “Building a Volunteer Community: Results and Findings from Transcribe Bentham,” Digital Humanities Quarterly 6/2 (2012): §62–4; Louise Seaward, “The Crowd, the Volunteers and the Supertranscribers. Building and Supporting an Online User Community for the Bentham Edition,” presentation read at the co:op READ Convention Marburg 2016. Available online at https://coop.hypotheses.org/224 [accessed 4 March 2018]. 72 Tim Causer, Justin Tonra, and Valerie Wallace, “Transcription Maximized; Expense Minimized? Crowdsourcing and Editing The Collected Works of Jeremy Bentham,” Literary and Linguistic Computing 27/2 (2012): 131. 73 Available at http://www.textualcommunities.org [accessed 25 July 2018]. 68

420

Catherine Smith

designed to be a general platform to support all stages of the editorial process allowing anyone to make a digital edition without specialist technical support.74 When discussing transcription, Robinson explains that “this system does not attempt to hide the XML: we think it helpful for the editor and transcribers to see exactly what encoding is being applied to the document.”75 Despite this positive spin, the prominence of the XML in the editor is part of the compromise inherent in trying to develop a transcription tool that works for any edition using the full range of TEI.76 In order to use this generic tool successfully projects must invest time and resources into mastering TEI-XML and passing these skills on to their transcribers. An excellent example of this is the work of the Estoria de Espanna (Estoria) project, one of the six projects that began using the Textual Communities system as a trial before its formal release.77 In addition to the transcriptions made by members of the Estoria team, they had some success with a small crowd sourcing trial using the Textual Communities transcription interface.78 While the team did find that the XML initially put off some potential volunteers, carefullystaged training materials that gradually exposed volunteers to the XML annotations required were successful in encouraging participation.79 The materials produced to support this endeavour are extensive and represent a considerable investment by the project.80 In addition to the development of the supporting material, time also had to be invested in providing feedback to volunteer transcribers, which the project team admits may cause problems if the crowd-sourcing scheme were to be expanded in the future.81 The IGNTP was using volunteers to make transcriptions long before the term crowd-sourcing even existed. Volunteers helped produce the IGNTP edition of Luke’s Gospel (published in two volumes in 1984 and 1987) by making collations of manuscripts. Photographic images of the manuscripts would be posted 74 Peter Robinson, “Some Principles for Making Collaborative Scholarly Editions,” §38. The project was also partly inspired by the growing interest in web-based collaborative editing and crowd sourcing. See Peter Robinson “The background to the Textual Communities Project,” accessible at http://wiki.usask.ca/display/TC/The+Background+to+the+Textual+ Communities+Project [accessed 25 July 2018]. 75 Robinson, “Some Principles for Making Collaborative Scholarly Editions,” §40. 76 In order to support the full scope of TEI Robinson supports any XML in the Textual Communities transcription interface. Robinson, “Some Principles for Making Collaborative Scholarly Editions,” §40. 77 Robinson, “Some Principles for Making Collaborative Scholarly Editions,” n.17. The Estoria de Espanna was led by Aengus Ward at the University of Birmingham (http://estoria .bham.ac.uk, accessed 4 March 2018). 78 Polly Duxfield, “Transcribing the Estoria de Espanna Using Crowdsourcing: Strategies and Aspirations,” Magnificat Cultura i Literatura Medievals 2 (2015): 131. 79 Duxfield, “Transcribing the Estoria de Espanna,” 142–44. 80 The canvas course created by the team on the Canvas platform can be seen here: https://canvas.bham.ac.uk/courses/6673 [accessed 4 March 2018]. 81 Duxfield, “Transcribing the Estoria de Espanna,” 145.

Digital Tools for Editing the New Testament

421

out to the volunteers who would make a collation of the witness against the agreed base text and then post their collations back to the editorial team. When full text electronic transcriptions replaced individual collations, the IGNTP volunteers along with a team of paid students and staff in Münster and Birmingham were able to make transcriptions in plain text files using the relatively simple Collate2 markup scheme.82 However, with the move to XML both the IGNTP and INTF teams felt that it was too much to ask transcribers to be able to work directly in XML and that the quality of the transcriptions would be adversely affected by the added visual distraction of XML tags. As the New Testament has such a long transmission history and so many transcriptions remain to be made for the ECM, it was worth investing the time and resources to create a bespoke transcription tool that removed the need for transcribers to learn XML. As part of the Workspace for Collaborative Editing project developers Martin Sievers and Gan Yu from the Center for Digital Humanities at the University of Trier produced the Online Transcription Editor (OTE). A screenshot of the OTE within the Workspace can be seen in Figure 1. The OTE was not written from scratch but is an extension to the open-source rich text editor TinyMCE.83 It is possible to transcribe the full text into the editor but for the New Testament we always use a base text which the transcriber modifies to match the text of the manuscript.84 For simple annotations such as adding line breaks shortcuts are included: whenever the return key is pressed a line break will be added. For more complex annotation, transcribers use a series of menus and submenus to add the annotation to the text without ever encountering any XML. Mouseovers inform the user of any additional details: for example, the text that can be seen in square brackets in Figure 1 is text that has been supplied by the transcriber. The mouseover (shown in Figure 2) informs the user of the reason the text has been supplied and the source from which the supplied characters were taken. To add the annotation, the user highlights the section of text involved and then clicks on the appropriate menu.

The transcription guidelines for the Collate format used by the Principio Project are available online. D. C. Parker et al., The Principio Project: Procedures for Transcribing Witnesses (2001), http://epapers.bham.ac.uk/2948, 9–17 [accessed 4 March 2018]. 83 https://www.tinymce.com [accessed 4 March 2018]. 84 H. A. G. Houghton, “Electronic Transcriptions of New Testament Manuscripts and their Accuracy, Documentation and Publication,” in Ancient Manuscripts in Digital Culture: Visualisation, Data, Communication, Digital Biblical Studies 3, ed. D. Hamidović, C. Clivaz, and S. Savant (Leiden: Brill, forthcoming). 82

422

Catherine Smith

Fig. 1. Screenshot of the OTE within the Workspace for Collaborative Editing

Fig. 2. Example of mouseover in the OTE

This will sometimes directly add the annotation but at other times will open a form to allow further details to be selected or entered. For more complex annotations that require further text to be added, such as indicating a correction, a version of the OTE is provided within the form so that the same editor can be used to add annotation to the text of the correction. The correction form is shown in Figure 3.85 The user experience of OTE is described in more detail in Darius Müller, “Zur Elektronischen Transkription von Apokalypsehandschriften: Bericht zum Arbeitsstand,” in Studien zum Text der Apokalypse II, ANTF 50, ed. M. Sigismund and D. Müller (Berlin: De Gruyter, 2017), 19–30. 85

Digital Tools for Editing the New Testament

423

Fig. 3. Form used to add corrections

The OTE only supports the particular customisation of the TEI used by the INTF and the IGNTP for the production of transcriptions for the ECM. While some of the customisation, such as adding corrections, might be applicable to other traditions, some features are specific to biblical texts. For example to establish the structure of the text books, chapters, verses and lections are used. Other supported features such as the annotation of nomina sacra, Eusebian canons, and Ammonian sections are also specifically biblical in their nature. The editor for the Greek New Testament will undoubtedly be lacking support for features common in other traditions. For example, researchers from the Multimedia Yasna project are working with the same developers to produce a similar version of the OTE to support their transcriptions of Avestan Yasna. Both the Avestan language and the text of the Yasna require further customisation of the editor which would be irrelevant for the Greek text of the New Testament in the same way that many of the features in the New Testament editor are not relevant to the Yasna project. Developing the tool was an investment of time and resources, but it would have taken a greater amount of time to train all of our transcribers to become fluent in TEI, and the need to learn this encoding might even have put off some transcribers (as the Transcribe Bentham project found). Using the editor does not mean that our transcribers require no training at all, but it does means that their training can focus on palaeographical skills and the correct identification and annotation of the features of the manuscript. For smaller projects with fewer transcriptions to make, or fewer transcribers to teach, a more generic tool like that in Textual Communities might be more suitable: in that case, investment would instead need to be put into the training and/or learning of TEI and XML. There is a balance that must be found and each editing project will find it in a different place.

424

Catherine Smith

Collation The collation of manuscripts requires the infuriating accuracy of a pedant and the obsessive stamina of an idiot. It is therefore an ideal task for a computer (1989).86

One of the first elements of Collate to be replaced was the core collation algorithm. CollateX was written by Ronald Haentjens Dekker as part of the Interedition project and is specifically intended to be a general tool. The design of the program follows the Gothenburg model (also devised by the Interedition project team) in its separation of the collation process into distinct stages: tokenisation, regularisation, alignment, analysis, and visualisation.87 It is intended that any of the stages can be taken over by another service, should that be necessary or desirable. As a result CollateX can receive input in many formats, has a number of settings that can be used to change its behaviour, and offers several output formats. The flexibility of CollateX is a great asset and is an important part of its appeal to a broad user base. But as a command line tool it is not particularly easy for non-programmers to use. The Collation Editor is a wrapper around CollateX that simplifies some of the interaction. In terms of the Gothenburg model, the Collation Editor uses the alignment functions provided by CollateX, along with the regularisation mechanism provided by the JSON input format. It provides its own visualisation and analysis tools allowing the user to regularise tokens, correct the automatic alignment, set variant lengths and order the readings in each variant unit and control certain CollateX settings, such as selecting the alignment algorithm used and changing the settings for the fuzzy matching of strings. Tokenisation, is left to the user (or the platform in which the collation editor is running), since the JSON input format used requires the data to be pre-tokenised and the preparation of the tokenised data is, to an extent, dependent on the collation editor settings required. Matthew Kirschenbaum suggests that, although interfaces in the Digital Humanities are the primary point of interaction for users, they are often the last thing to be written (normally at a point when the project is running out of both time and money).88 This was certainly not the case with the collation editor. We already had the computationally difficult aspect in CollateX: what was needed was an easy way for editors to interact with the software and the data, making this an interface design project from the start. That is not to say that this was just a project about making things look nice. In order for the collation editor to be fit for 86 Peter Robinson, “The Collation and Textual Criticism of Icelandic Manuscripts (1): Collation,” Literary and Linguistics Computing 4/2 (1989): 99. 87 See the CollateX Documentation available online at https://collatex.net/doc/ §1 [accessed 4 March 2018]. 88 Matthew G. Kirschenbaum, “‘So the Colors Cover the Wires’: Interface, Aesthetics, and Usability,” in A Companion to Digital Humanities, ed. S. Schreibman, R. Siemens, and J. Unsworth (Malden: Blackwell, 2004), 525.

Digital Tools for Editing the New Testament

425

purpose, it had to encapsulate the complex process of preparing a critical edition and that involved understanding and modelling the process alongside building the interface. The role of the computer in scholarly editing was summarised by Thomas Howard-Hill in 1973 and his conclusions are still relevant to the design of editorial tools today. The computer is to be the servant of the editor, and the system is to enable the editor to concentrate where his attention is most required, on problems of textual and critical analysis. No computer-based system which distracts the editor’s attention from the goal of his activity to the means of achieving it is likely to find acceptance.89

This view aligns with both our policy of hiding the XML from our transcribers and with our approach to designing the collation. Deciding on the layout for the collation output was central to the design of the Collation Editor. CollateX itself offers a number of output options including alignment tables with one column (or row) per witness, a JSON based table structure and TEI compliant XML. In addition it provides an SVG serialisation of its internal data structure, a directed acyclic graph.90 In this view, the words of the texts are shown in the nodes in the graph and the lines that connect them (in graph terminology, the edges) are labelled with the witnesses. The text of a witness can be read by following the path using the connecting lines and reading every node the path travels through. This visualisation has also been extended to provide a means of editing the collation output, such as regularising word forms or joining words together into a single variation unit.91 While Stefan Jänicke and his collaborators consider the readability of the CollateX variant graph to be an improvement on previous proposals they suggest that there is still room for improvement.92 They propose five “design rules” to improve the readability of the CollateX variant graph which they have implemented in their javascript library TRAViz.93 The rules proposed are: use font size to give an indication of the number of witnesses attesting a particular reading (the larger the font the more witnesses read the word); avoid “backwards edges” so that all texts read fully from 89 Thomas H. Howard-Hill, “A Practical Scheme for Editing Critical Texts with the Aid of a Computer,” Proof: The Yearbook of American Bibliography and Textual Studies 3 (1973): 336–37. 90 Ronald Haentjens Dekker and Gregor Middell, “Computer-Supported Collation with CollateX,” conference paper at Supporting Digital Humanities 2011: Answering the Unaskable. Copenhagen, Denmark (2011). 91 Tara L. Andrews and Joris van Zundert, “An Interactive Interface for Text Variant Graph Models,” Digital Humanities (2013) http://dh2013.unl.edu/abstracts/ab-379.html [accessed 4 March 2018]. 92 Stefan Jänicke, Annette Geßner, Greta Franzini, Melissa Terras, Simon Mahony, and Gerik Scheuermann, “TRAViz: A Visualization for Variant Graphs,” Digital Scholarship in the Humanities 30 suppl_1 (2015): i83–i99. 93 The code, tutorials and examples of the tool in action can be found on the project website http://www.traviz.vizcovery.org/ [accessed 4 March 2018].

426

Catherine Smith

left to right; use colours rather than labels for edges (connecting lines); bundle the most common edges together into a single thicker line; insert line breaks to avoid horizontal scrolling. As with the CollateX variant graph, TRAViz provides editing features to detach words and realign them as well as joining nodes together. One area of improvement made by TRAViz does not seem to be covered in the rules. This is the vertical alignment of the words and the layout of the edges which combine to make the pattern of shared readings and variants immediately obvious in the outer shape of the graph: it gets wider where there are lots of variants and narrower where more witnesses agree. The examples provided in the paper and on the TRAViz website are certainly visually appealing, but as with the CollateX variant graph, once the number of witnesses starts to rise or witnesses become more divergent, it becomes harder to make sense of the data.94 The use of colour also necessarily limits the number of witnesses TRAViz can meaningfully display. The need to collate upwards of 200 witnesses at a time meant a different approach was required for the New Testament. The layout used can be seen in Figure 4. Internally, CollateX performs baseless collation and both of the visualisations already discussed above have also followed this model. In our layout a base text is used to organise the variants, although we still benefit from the results of the baseless collation. This also has the advantage of ensuring that there are no backward edges. Any of the witnesses in the collation can be used as the base text so long as it contains the verse being collated.95 For the purposes of the ECM the base text used is the text of NA28 (one of the standard hand editions of the New Testament) with any verses not present in the main text added (for example John 5:4). The top line of the display shows the base text with the index numbers used in the ECM (even numbers for words and odd numbers for spaces). Below this the variant readings are presented: each variant unit is in a separate box and each reading is labelled with a letter.96 The a reading is always the reading of the base text. Lacunose and omitted readings are shown at the bottom of each unit. The witnesses that attest each reading can be seen by hovering over the reading text; the reading of a particular witness can be highlighted by selecting the siglum in the menu at the bottom of the page. The layout gives a similar overall outline to the TRAViz display, making it easy to assess the scale of the variation. We have not followed the recommendation of Jänicke et al. to use line breaks rather than horizontal scrolling. This is because we have to be able to handle overlapping variants. These are longer variant units, often ones which have a difference in word order, which are moved into secondary apparatus lines. Having line For an example of such a graph see Jänicke et al., “TRAViz,” i96 Fig. 17. The collation editor can actually display the variants if the base text is lacunose for the verse but the data cannot be edited and manipulated properly, since all of the variants are treated as additions. 96 The use of letters to identify readings follows the convention used in the ECM. 94 95

Digital Tools for Editing the New Testament

427

breaks as well as these extra lines of apparatus would make the display much more difficult to understand.

Fig. 4. Screen shot of the collation editor layout

The most important functional aspect of the Collation Editor is the ability to edit and correct the results from CollateX. Collate was the first program to treat computerised collation as an interactive process between the computer and the editor. When Robinson wrote Collate, he wrote it for the Macintosh computer, as that was the only system at the time that could reliably support what he refers to as “point and click” menus.97 Although cutting edge technology at the time, these menus forced the scholar to interact at a level at least once removed from the data itself. Today there are many more options than were available in the late 1980s. The aim with the Collation Editor has been to keep the data itself central and to keep the interface as invisible as possible.98 To this end the collation editor uses “drag and drop” and the right click context menu for the majority of the interaction, allowing editing to happen through direct manipulation of the text. For example, in the regularisation interface, to create a rule the user drags the word to be regularised onto the target word and this brings up a menu which allows the scope and type of the rule to be specified. When setting variants, one unit can be dragged and dropped onto an adjacent one to join them together; they can be split again by right clicking on the unit and selecting “split words” from the context menu. If words have been aligned incorrectly, this can be corrected in the set variants interface. To move an individual reading to a different location 97 Peter Robinson, “A History of Collate,” Scholarly Digital Editions 29 September 2014: http://scholarlydigitaleditions.blogspot.co.uk/2014/09/the-history-of-collate.html [accessed 4 March 2018]. 98 This follows the advice from Negroponte who states “the secret to interface design [is to] make it go away.” Nicholas Negroponte, Being Digital (London: Hodder & Stoughton, 1995), 93. Originally cited in Kirschenbaum, “So the Colors Cover the Wires,” 524.

428

Catherine Smith

the readings are split using the right click context menu and then each reading can be moved independently. To move a reading to a new unit, it is dropped onto the appropriate unit. To move it to a space without a unit (to be treated as an addition) it is dropped on the number for the space and a new unit is created. While there is no such thing as an intuitive interface,99 this direct manipulation of the data has proved to be a successful model for interaction. This became apparent in one of our regular project meetings during a demonstration of a newly implemented solution to the overlapping variants problem. The example being used for the demonstration had been provided to me as an example of an overlapping variant at a previous meeting. Following the demonstration I was expecting either confirmation that the interface was fine or, more likely, a list of changes or new features that were required. What ensued was a long discussion of whether or not the beginning of John 6:23 should be actually treated as an overlapping variant after all. At the time this was rather frustrating, the issue at hand was not whether John 6:23 was an example of an overlapping variant, but rather whether the interface allowed the editors to record an overlapping variant in the correct way. On reflection, the response was perfect. The editors had just been introduced to a brand new feature of a still relatively unfamiliar interface and they had not noticed the interface, but only data. And that is exactly what Howard-Hill was talking about: the interface should not distract the user from the data being edited. Initially, the Collation Editor was only intended to work within the suite of tools developed as the Workspace for Collaborative Editing. However, it became clear that other editing frameworks also needed a way for their users to interact with CollateX, so the code has been adapted to work in different architectures. There is also a standalone version that works with the local file system and requires no further infrastructure.100 Broadening the user base for the editor in this way also introduced the requirement that it work with a much wider range of traditions, beyond the Greek and Latin New Testament as was the original intention. The fact that the editor was written to work with such a complex tradition and with two languages from the start, meant that some of the features had also been configurable from an early stage: examples of this include the base text and the list of witnesses to be used in each project. Other features, such as the types of regularisation rules, were made configurable early on for practical reasons of project management. For example, the editorial teams did not agree on the types of rules they wanted to use for regularisation so these were set in the configuration files on a project-by-project basis. This pattern has now been extended and 99

J. Raskin, “Intuitive Equals Familiar,” Communications of the ACM, 37/9 (1994): 17–

8. 100 This work was carried out in the summer of 2015 along with Troy Griffitts, the programmer of the NTVMR, one of the other systems which sought access to the collation editor.

Digital Tools for Editing the New Testament

429

all of the Greek, Latin, and New Testament specific settings have been moved into configuration files. It is hoped, having started with such a large and complex tradition, that the collation editor will only require changes to the configuration files in order to handle other traditions. Only time will tell whether that is a realistic hope.101 The result of all of these changes is that, although it will not be suitable for every edition ever made, the Collation Editor is now a more generally applicable tool. It is working in the NTVMR and in the Textual Communities platform and the standalone version was used to make the edition of the Estoria de Espanna.102 There is, of course, a downside to the flexibility introduced into the editor by generalising it in this way. Writing the service files necessary to have it work with different architectures is the first challenge, although most projects will have technical support if they are building their own architecture which may mitigate this issue. More of a problem is the complexity of the configuration process for different traditions and editorial practices. Certain features may require additional code to be written and almost all have an impact on the way the data must be prepared for use. An additional complication is the flexibility of TEI which means that the collation editor cannot currently provide any help with the transformation of data from TEI to the JSON format which it uses. It is simply not possible to predict what features of a text a user will encode, how they will encode them, and what they expect the Collation Editor to be able to do with the result. It would of course be possible to set out rules for what TEI can and cannot be used with a set of automatic import scripts, but that would start to reintroduce use-case restrictions and also go against the principle that has made TEI so widely used. The result is that a lot of time and effort must go into working out what an editor wants to do with their edition and preparing the data and configuration files in such a way that it is all possible. In many cases the technical skills needed to configure the editor are great, perhaps more than might be required to use CollateX on the command line. That is the price to be paid for a tool, particularly one with a graphical user interface, that can be adapted to the specific requirement of multiple textual traditions and editorial approaches.

Users from other traditions have already encountered frustrations around tokenisation differences between manuscripts which makes it difficult for them in the early stages of collation. While I maintain that tokenisation is not a direct responsibility of the collation editor (see the opening of this section), I accept that this might need to be addressed in future developments. 102 See description above and Aengus Ward, “The Estoria de Espanna Digital: collating medieval prose – challenges...and more challenges,” Digital Philology (forthcoming). The edition is available here: http://estoria.bham.ac.uk/edition/ [accessed 4 March 2018]. 101

430

Catherine Smith

Preservation and Sustainability If I have hitherto seemed tedious, and superfluously attentive to detail, my excuse is that I have sought to present a faithful view of every characteristic of my manuscript – to faithfully edit it in short – besides the mere notation of an addition to or an omission from the textus receptus. The advantage of this method, combined with a really accurate collation, is obvious. It presents the whole codex to those at a distance, and does away with that never-ending re-examination of documents, which has been going on – at such a cost of precious time – ever since the days of Henry Stephens (1890).103

Concerns about preservation and sustainability are not new. As the quotation above shows, in 1890 H. C. Hoskier was already thinking about data reuse and was encouraging others to build on his work rather than starting afresh. With the advent of computers this became more of a practical reality: data sharing and reuse was a key motivation for Ellison in his work and he himself reused data created by others; when Fischer was making his concordance of the Vulgate, Busa (via Ott) provided a copy of the lemmatised Latin lexicon that had been created for his concordance project in exchange for information about the words used in the Vulgate, but not by Thomas Aquinas;104 it was possible to purchase the Greek New Testament on punched cards or paper tape from 1969;105 and in 1976 the Oxford Text Archive was established in order to preserve machine readable texts and make them available to the academic community.106 However, the limited storage technology available meant that having access to these texts did not guarantee they could be easily reused for other purposes. There were also already concerns about the quality of the data. For example, an ESTS committee in 1970 called for all existing copies of tapes prepared by Andrew Q. Morton (possibly for his authorship studies) to be proofread and for a system to be established to certify which of the copies had been checked. It also pointed out that the data did not include chapter and verse references, which might be needed, and suggested that diacritical signs and “at least major textual variants” be added.107 Adding additional data was not straightforward in the days of punched cards and paper tape since, as Fischer points out, “the marks can hardly be fitted in afterwards without copying the whole text again.”108 While digital storage, rather than just digital processing, has certainly made data more extensible and therefore easier to reuse, preservation remains a 103 H. C. Hoskier, A full Account and Collation of the Greek Cursive Codex Evangelium 604 (London: David Nutt, 1890), xxviii. 104 Nyhan and Flinn, “The University Was Still Taking Account of Universitas Scientiarum,” 59. 105 The purchase price for the tapes/cards was £12 in 1969 which is the equivalent of about £190 today. Fischer, “The Use of Computers,” 298. 106 Hockey, “The History of Humanities Computing,” 8. 107 Kenneth Grayston, “Computers and the New Testament,” NTS 17/4 (1971): 479–80. 108 Fischer, “The Use of Computers,” 298.

Digital Tools for Editing the New Testament

431

challenge.109 With the amount of digital data being produced it is also a growing challenge. As Pierazzo points out, it is not possible to keep everything. This makes it important to have sensible policies about what to keep and how to keep it.110 One of the key elements to preservation is to preserve the data separately from its display interface and associated tools.111 The data will almost certainly be useful long after the tools that made it have gone. We have already seen this with our own editions. The transcriptions made in plain text using Collate2 were saved independently of the system and were easy to convert to XML so as to be ready for the new tools.112 The editions published with Anastasia, the publishing tool for Collate, were only stored within the Anastasia system and although for the most part we did manage to recover the apparatus data, it took a lot of time and some data was lost. Had the apparatus been stored separately as data and not just as part of the display system this would have been much easier to do. The format chosen for the preservation of data is also important. As Cummings points out, despite its problems for interchange, TEI-XML is still a good format for long-term data preservation.113 Pierazzo suggests that the use of any standard is a positive thing in and of itself, for a number of reasons. If more people use a particular standard there will be more interest in ensuring that that standard is maintained, and if for any reason the standard becomes obsolete and “a critical mass of data exists in one format, tools will be produced to migrate it.”114 Holmes advocates producing different preservation formats to improve the reusability of data, suggesting that we produce a simplified version of our TEI texts using a restricted tag set such as TEI Lite or TEI Simple.115 We also need to think about appropriate licensing.116 In addition to their original purpose of providing early dissemination of academic research, institutional repositories have proved to be a good place to preserve and disseminate XML documents and

See Pierazzo, Digital Scholarly Editing, 169–92; William Kilbride, “Saving the Bits: Digital Humanities Forever?” in A New Companion to Digital Humanities, ed. S. Schreibman, R. Siemens, and J. Unsworth (Oxford: John Wiley, 2016), 408–19. 110 Pierazzo, Digital Scholarly Editing, 170–71. 111 Marilyn Deegan, “Collection and Preservation of an Electronic Edition,” in Electronic Textual Editing, ed. L. Burnard, K. O’Brien O’Keeffe, and J. Unsworth (New York: Modern Language Association of America, 2006), 366. 112 Around 1,000 transcriptions made by the INTF and IGNTP have been converted from plain text to XML. That would have been a lot of data to lose. 113 Cummings, “The Text Encoding Initiative,” 473. 114 Pierazzo, Digital Scholarly Editing, 171. 115 Holmes, “Whatever Happened to Interchange?” i65. 116 See Robinson, “Some Principles for Making Collaborative Scholarly Editions,” §36 on the problems with the Creative Commons non-commercial clause and Houghton, “Electronic Transcriptions,” on the removal of the share-alike condition. In November 2017 the IGNTP took the decision to release its transcriptions under a CC-BY license. 109

432

Catherine Smith

other digital assets.117 A new initiative from the UK research councils aims to make openly available the data collected during the course of publically-funded research; this is the data equivalent to open access in publishing.118 As part of this, institutions are required to provide the infrastructure to preserve this data and make it discoverable. At Birmingham this infrastructure is known as the Research Data Archive and forms part of the Birmingham Environment for Academic Research (BEAR).119 It allows the long-term archiving of data and allows it to be related to published outputs. This is ideal for the ECM which produces a printed volume as alongside the electronic data. While preservation options have improved, sustainability remains a challenge. Pierazzo suggests that bespoke tools are inherently unsustainable, implying that more generic tools are more sustainable.120 It may well be the case that tools of interest to a larger user base, if they are genuinely useful to those users, are more likely to be sustained for longer but it does not mean they are necessarily more sustainable. Sustainability is about more than a large user base: the latter does not necessarily translate into a large development community, particularly with tools in the Humanities. Collate2 had a relatively large user base, but still only one developer; it was also reasonably well documented, yet it became unsustainable because of changes in hardware beyond any developer’s control.121 In ITSEE we stopped using Anastasia partly because of changes to the web server software but also because we discovered other options which we found easier to use. That does not mean that either of those tools was not needed at the time they were written. They were needed, and they have changed the way we edit the New Testament, but they have served their purpose and now we have better tools. The ones that come next will probably be better still. One of the ways that has been suggested for improving the sustainability of tools and, at the same time, to resolve the tension between generalisation and specific requirements is to use a modular approach. This was the model used for the development of CollateX: it focused on microservices that model individual steps in the editing process but can be joined together into larger workflows.122 These smaller elements have many advantages over large scale infrastructures. The focus on a single task means they can be flexible without getting too complex, they are more easily sustained and replaced if necessary, and they can be 117 Alex Ball, Preservation and Curation in Institutional Repositories (Edinburgh: Digital Curation Centre, 2010), 5–6. 118 “Concordat on Open Research Data” (2016) http://www.ukri.org/files/legacy/documents/condordatonopenresearchdata-pdf/ [accessed 25 July 2018]. 119 See https://intranet.birmingham.ac.uk/it/teams/infrastructure/research/bear/index. aspx [accessed 4 March 2018]. 120 Pierazzo, Digital Scholarly Editing, 112. 121 The INTF actually kept a copy of Collate running until 2015 by preserving an old preIntel Macintosh computer as its host. Houghton et al., “The Editio Critica Maior.” 122 van Zundert “If You Build it,” 181.

Digital Tools for Editing the New Testament

433

chained together in customised ways to support a much wider user base. Pierazzo calls for a similar approach, using the term bricks for the component parts, and suggests that these could be joined together with only a small amount of configuration in order to create bespoke workflows for individual editorial projects.123 This approach achieves maximum reuse without compromising individual expression. Of course you still need programmers to pipe the data from one brick to another and an architecture such as the NTVMR, Textual Communities, or the Workspace for Collaborative Editing to control the workflow. These can all share the bricks that work for them and more easily replace the ones that do not by finding or writing an alternative. The smaller the bricks, the better of course, but it depends on the task: some will need to be larger than others because of the tasks they have to perform. The Collation Editor, for example, is a wrapper around CollateX making it bigger than might be ideal but still discrete enough to be used in different workflows. I like to think of the ideal in terms of the mantra for setting variants for the ECM: each tool should be “as small as possible and as large as necessary.” When it comes to sustainability and preservation our efforts should go into the data and the final edition, not into the tools used to create them. This data can be reused in the next generation of tools that will invariably come along. No one can predict how technology will change or how that technology will change editorial practice: in 1950 it was predicted that America might ultimately need a total of 12 computers!124 Just two years before the adoption of the full text transcription workflow by the ECM teams, Robert Kraft hailed Robinson’s Collate as “an excellent program,” but implied that the work involved in making the individual transcriptions is too great a task for the New Testament for it to be worth exploring.125 Because the technology available for building and running tools will always be changing, we should worry less about their long-term survival. This is not to say that we should give up on sustainability altogether. We should follow sustainable software practices and use appropriate data standards; we should ensure that tools are documented, for future developers as well as users; 126 that the code is open source and made available in a repository such as sourceforge or github, preferably with a Digital Object Identifier (DOI); we should publish papers that

Pierazzo, Digital Scholarly Editing, 116–17. Edmund A. Bowles, “Preface,” in Computers in Humanistic Research, Readings and Perspectives, Prentice-Hall Series in Automatic Computation, ed. E. A. Bowles (Englewood Cliffs: Prentice-Hall, 1967), v. 125 Kraft, “The Use of Computers,” 271–72. 126 As Stephen Ramsay, “Hard Constraints: Designing Software,” in A New Companion to Digital Humanities, ed. S. Schreibman, R. Siemens, and J. Unsworth (Oxford: John Wiley, 2016), 456 says, “‘Read the source’ is an absurd substitution for documentation.” 123 124

434

Catherine Smith

explain why the tools were needed, what the tools do and how they do it.127 By following these sustainable development practices we preserve the knowledge that went into the tools and we can build on that knowledge in the future. We will lose tools along the way, which is both inevitable and acceptable: we do not have all of Busa’s punched cards, magnetic tapes, and calculating machines, but his concordance is still available online as part of the Corpus Thomisticum project. We should not stop building bespoke tools if they are needed just because they might not be as quite as sustainable as more general ones. Van Zundert even suggests that it is at the level of specific domain problems that methodological innovation is found.128 In short, we should let concerns about sustainability inform how we develop, but we should not let them control what we develop. New Testament textual criticism has a long history of embracing technology and that trajectory is continuing. The concept of full text transcriptions and machine collation is now fully embedded in our workflow. The second-generation transcription and collation tools have already become central to the production of the ECM. Work is underway in a collaboration between the INTF and the Cologne Center for eHumanities (CCeH) to develop a second generation of tools to support the Coherence-Based Genealogical Method. Again, these tools will not change the method itself, but rather provide new, interactive interfaces: more new wine skins for existing data and techniques. Alongside this, new computation techniques are also being developed. Andrew Edmondson at the University of Birmingham is investigating the use of phylogenetic techniques for establishing manuscript relations using the university’s high performance computing cluster, which like the Harvard Mark IV used by Ellison, is large enough for people to walk around inside. Technology will always be changing, and new technology brings with it new possibilities. If history is anything to go by, New Testament textual critics will continue to make good use of it.

127 A journal has been launched to support the publishing of papers in this area motivated by the problems encountered with getting such papers published elsewhere, see https://open researchsoftware.metajnl.com [accessed 4 March 2018]. Having a platform for publication however does not solve the problem of the low status attached to such papers within University departments. 128 van Zundert “If You Build it,” 167–68.

The Development of the Coherence-Based Genealogical Method (CBGM), its Place in Textual Scholarship, and Digital Editing Klaus Wachtel 1. The Development of the CBGM The CBGM has been developed – and is still being developed – in the context of the Editio Critica Maior (ECM) of the Greek New Testament. A vast degree of contamination1 is a fundamental feature of its manuscript tradition and, therefore, efforts to account for its structure in terms of stemmatology have generally been considered hopeless. The remedy that the CBGM offers against contamination is a methodical analysis of another feature of this tradition, its coherence. The average agreement of all witnesses, compared at variant passages only, is more than 80%. And, as a rule, you can find several relatives for each witness whose agreement rates are much higher. The CBGM provides tools for editing the Greek New Testament in the first place, but it is not restricted to this tradition. It is an appropriate tool for exploring the genealogical structures of all rich manuscript traditions from antiquity. The scholar who invented the method and continues to advance its development is Gerd Mink, one of the ECM editors. The first steps he took were based on the Text und Textwert collations of the Catholic Letters in the 1980s, and he published his first programmatic article on the CBGM in 1993.2 Work on the ECM, on the other hand, was resumed in mid 1990s at the Institut für Neutestamentliche Textforschung (INTF), after initial efforts in the 1970s. At this time, Mink designed a database for collecting, organizing, and processing the evidence, immediately using the available evidence from Greek manuscripts of James to apply the CBGM to the 756 variant passages of that work. It took eight years to finish the first edition of the Catholic Letters of the ECM (from 1997–2005), and in the course of these years the CBGM methodology and 1 I keep using this term instead of “mixture” because it is still most commonly used, referring to the intrusion of variants into a strand of transmission from another strand. 2 Mink, “Eine umfassende Genealogie der neutestamentlichen Überlieferung,” NTS 39 (1993): 481–99.

436

Klaus Wachtel

the display of its results was refined. By 2005, CBGM analyses could be based on an evaluation of all 3,046 variant passages of the Catholic Letters. By 2008, when a conference around the CBGM was held in Münster,3 the CBGM tools appeared online at the “Genealogical Queries” website.4 In preparation for the second edition of the Catholic Letters, the team of editors made direct use of the new tools and the complete dataset from which it draws. The second edition of the Catholic Letters appeared in 2013, featuring 33 textual changes against NA27. These were incorporated into NA28 which came out shortly before the second edition of ECM Catholic Letters. These changes were later also introduced into UBS5. Work on ECM Acts was already well under way when the second edition of the Catholic Letters appeared. For editing Acts, all the Genealogical Queries tools were in place. Three CBGM iterations through the apparatus of Acts were carried out using the old interface, but this interface, as well as its database structures, were no longer up to date. We are now in the process of modernizing both of these components in collaboration with the Cologne Center for eHumanities (CCeH). The new interface has been online for some time and is available together with earlier versions of the software.5 To sum up the history of this stemmatological method, the CBGM was developed in close relationship with the ECM. Through the Nestle-Aland and the UBS Greek New Testament, the CBGM has an impact on the text of the New Testament that is used worldwide, not only in academic teaching and research, but also in the UBS translation programs. The CBGM presupposes a well-formed full apparatus as provided by the ECM, but it is not exclusively applicable to the New Testament. It can be applied to any work with a manuscript tradition that can be referenced word by word.

2. The Place of the CBGM in Textual Scholarship Taking Paolo Trovato’s framework of “genealogical textual criticism in the age of post-structuralism, cladistics, and copy text,” I will now try to determine the place that the CBGM may claim within the broader discourse of textual scholarship.6 Trovato omitted altogether any discussion of the CBGM, which offers me

The proceedings were published three years later in K. Wachtel and M. Holmes, eds., The Textual History of the Greek New Testament: Changing Views in Contemporary Research (Atlanta: SBL, 2011). 4 http://intf.uni-muenster.de/cbgm/en.html [accessed 10 March 2018]. 5 http://intf.uni-muenster.de/cbgm/acts/index.html [accessed 10 March 2018]. 6 P. Trovato, Everything you Always Wanted to Know about Lachmann’s Method: A NonStandard Handbook of Genealogical Textual Criticism in the Age of Post-Structuralism, Cladistics, and Copy-Text, 2nd ed. (Padova: Libreriauniversitaria.it, 2017). 3

The Development of the Coherence-Based Genealogical Method

437

the opportunity to introduce its basics with reference to his otherwise comprehensive presentation. In spite of the title of the monograph, Everything You Always Wanted to Know about Lachmann’s Method, Trovato does not deal much with Lachmann’s method as Lachmann himself explained and applied it, but rather with “the basic rules of the genealogical method inaccurately attributed to Karl Lachmann.”7 The subject of Trovato’s book is “the genealogical-reconstructive method, also known as the common-error or (Neo-) Lachmannian method,” which Trovato himself professes.8 His stance is contrary to scholars who consider the original (authorial) text to be completely beyond our reach, whether they confine themselves to transcribing manuscripts or to reconstructing an archetype of the extant manuscript tradition as an alternative to the search for the original or authorial text. Trovato is one of the “staunch ‘reconstructionists’, who believe that the task of a scientific edition is…to use the archetype as a point of departure, using all available means…to try to come as close as possible to the lost original.”9 Similarly, one of the central goals of the CBGM is the reconstruction of the initial text. Gerd Mink states that “the initial text is not simply a reconstruction on the basis of the surviving variants, which best explains the emergence of the variants and thus represents the archetype of the tradition…The simplest working hypothesis must be that there are no differences between the original and the initial text (except for inevitable scribal slips).”10 Thus the effort of approximating the text of the author is a goal common to the Neo-Lachmannian method à la Trovato and the CBGM. A similarity between Lachmannian methodology in general and the CBGM is the exclusion of descripti (eliminatio).11 Only those witnesses that are not merely copies of the late Byzantine text, which are represented by a relatively small number of straightforward Byzantine witnesses, are included in the ECM apparatus. For the ECM, the eliminatio is carried out on the basis of extensive test passage collations to which all available New Testament manuscripts were subjected.12 The result is a selection of witnesses that includes (1) all manuscripts representing text forms earlier than the largely uniform late Byzantine text; (2) representatives of the late Byzantine text itself; and (3) a large quantity of witnesses which predominantly Trovato, Lachmann’s Method, 23. Trovato, Lachmann’s Method, 15. 9 Trovato, Lachmann’s Method, 15. 10 G. Mink, “Problems of a Highly Contaminated Tradition: the New Testament,” in Studies in Stemmatology II, ed. P. van Reenen, A. den Hollander, M. van Mulken (Amsterdam: Benjamins 2004), 13–85, here 25–6. For a more extensive explanation of the term “initial text,” see also the introduction in Wachtel and Holmes, Textual History, 6–7. 11 Cf. Trovato, Lachmann’s Method, 60. 12 The test passage collations and evaluations were published in the series Arbeiten zur neutestamentlichen Textforschung under the title Text und Textwert der griechischen Handschriften des Neuen Testaments (Berlin: De Gruyter 1986–2017). 7 8

438

Klaus Wachtel

support the Byzantine stream, but which also contain deviating material, either taken over from older traditions or developed in the course of copying. One fundamental difference from Neo-Lachmannian methodology, however, must be mentioned before we delve into more detail. For the CBGM, the distinction between manuscripts and the texts they carry is of paramount importance. Mink states that “the witness of a variant is the text, not the manuscript.”13 Accordingly the final goal of the CBGM is a stemma that displays the relationships between states of text, not between manuscripts. Common Error vs. Coherence The CBGM can certainly also be called a genealogical-reconstructive method. It could not, however, be identified as a common error method. For the CBGM, coherence is the pre-eminent feature of the New Testament manuscript tradition for which it was developed in the first place. In Lachmannian methodology, common errors are used to trace genealogical structures. The starting point is a collation yielding differences which are all seen to be “readings or variants.”14 The next task is to distinguish variants and polygenetic errors from monogenetic or indicative errors. Here the term “variant” refers to morphological and orthographical differences only, while polygenetic errors are likely to have been committed, not copied, by scribes. Trovato states that “only errors…that copyists can reproduce, but that as a rule cannot be made independently by several scribes (monogenetic errors) should be used for the classification of witnesses.”15 The CBGM also starts from variants, leaving open, however, which of the variants should be considered instances of error as compared with the initial text. In the context of the CBGM, it is more important to distinguish variants from clerical errors and orthographica: A reading is the generic term for the wording of a textual unit in which a manusciript is distinguished from one or more or from all other manuscripts. A variant refers to one of at least two reading of the same textual unit which is grammatically correct and logically possible in its context. Errors are readings which do not fulfil these criteria. Alternative and orthographically possible forms of the same variants are classed as orthographica. Examples include certain morphological differences, such as the omission of the present stem element -µ- from the future forms of λαµβάνω, or the occurrence of mixed for strong aorist forms.16

Mink, “Highly Contaminated Tradition,” 29 (emphasis original). Regarding the methodological importance of distinguishing textual from manuscript traditions, particularly in computer aided textual research, see B. Bordalejo, “The Genealogy of Texts: Manuscript Traditions and Textual Traditions,” Digital Scholarship in the Humanities 31 (2016): 563– 77. 14 Trovato, Lachmann’s Method, 52–4. 15 Trovato, Lachmann’s Method, 56. 16 ECM Acts 1.1, 24*. 13

The Development of the Coherence-Based Genealogical Method

439

In the ECM apparatus, clerical errors and orthographica are assigned to the superordinate variants because such readings are largely irrelevant for the genealogy of variants and states of text. Obvious errors were corrected when a scribe copied a manuscript; orthographical differences reflect the history of the language rather than that of the text. In the context of the CBGM, all grammatically sound, or at least tolerable textual differences, which are not merely orthographical, are considered variants. An indicative error would have to be a variant in this sense to be genealogically useful because, as a rule, clerical errors were corrected, not copied, by the scribes. The CBGM abstains from identifying variants as errors, a principle that offers two advantages over against the common-error method: (1) we do not have to know at the outset, relying only on our philological acumen, which variants are errors and which are true renderings of the text in a pristine exemplar; and (2) we are not immediately confronted with the problem of contamination which admittedly, as will be seen below, is the biggest problem for Lachmannian genealogy. Instead, we can make use of quantitative data regarding similarities and differences between witnesses, i.e. pre-genealogical coherence, to get an impression of the consistency of attestations. Such an accounting of readings presupposes a well-formed critical apparatus in which every witness subjected to comparison with any other witness is cited at every passage of variation, including deficiencies such as lacunae. Then, the comparison incorporates all the passages where the compared witnesses are extant.17 The most important innovation brought about by the CBGM compared to Lachmannian methodology is the perception and description of genealogical relations. According to Trovato, the relationship between any two manuscripts A and B can be assigned to one of three types, A>B, B>A, or AB.18 For the textual tradition of the Greek New Testament, it would not be a reasonable goal to describe the relationship between any two manuscripts following this pattern. For any two New Testament witnesses A and B, i.e. states of text preserved in manuscripts, there is textual flow of the type A>B, which stands side by side with flow in the opposite direction (B>A), as well as AB, due to contamination. The contribution that the CBGM claims to make to stemmatological textual scholarship is a cure against contamination, or at least a way to do stemmatological analysis in a contaminated tradition.

For the use of pre-genealogical coherence as indicator of relatedness see G. Mink, “Contamination, Coherence and Coincidence in Textual Transmission,” in The Textual History of the Greek New Testament: Changing Views in Contemporary Research, ed. K. Wachtel and M. Holmes (Atlanta: SBL, 2011), 141–216, here 157–58. 18 Trovato, Lachmann’s Method, 57. 17

440

Klaus Wachtel

3. The Stemmatological Approach of the CBGM19 Paul Maas is one of the first scholars to use the term “Stemmatik,” rendered in English as “stemmatics” or “stemmatology,” albeit usually with inverted commas or scare quotes. In a 1937 article, “Leitfehler and stemmatische Typen,” he apologises for using this neologism and defines it en passant as “Lehre von den Abhängigkeitsverhältnissen der Handschriften” (“doctrine of dependencies between manuscripts”).20 The concluding statement of the article sets a high standard for an acceptable stemma codicum: Just as the chemical formula settles the arrangement of the atoms in each molecule of a compound in unambiguous and unalterable fashion, so the stemma settles the relationship of witnesses for every passage in the text – if we have a virgin tradition. No specific has yet been discovered against contamination.21

Three features of Maas’ comment must be emphasized in order to correctly understand this doctrine and its relevance for understanding the textual history of the New Testament: (1) the usefulness of Maas’ stemma codicum is restricted to “virgin traditions,” i.e. non-contaminated traditions; (2) Maas’ approach regards the relationship of manuscripts, not the texts that they carry; and (3) Maasian stemmatics are guided by indicative errors. As I mentioned above, the CBGM differs on each of these accounts, since it is designed to examine contaminated traditions, focuses on examining states of texts preserved in the physical manuscripts, and evaluates textual relationships based on coherence. Maas first published his Textkritik in 1927 and Giorgio Pasquali wrote a lengthy review about this work in 1929, from which his Storia della tradizione e critica dell testo (1934) developed.22 One overwhelmingly well-documented conclusion of Pasquali’s review was that there is no immaculate manuscript tradition. Contamination is the rule, not the exception in rich manuscript traditions, of which the New Testament is a prime example. The aim of conventional stemmatology is the reconstruction of the archetype, the first manuscript of a tradition. Where gaps in the tradition exist (i.e.

19 This chapter is based on a paper presented under the title “A Stemmatological Approach in Editing the Greek New Testament” at Digital Scholarly Editing: Theory, Practice, Methods – DiXiT Convention, Antwerp 2016. An extended abstract was published in Peter Boot et al., eds., Advances in Digital Scholarly Editing (Leiden: Sidestone, 2017), 223–27. 20 Byzantinische Zeitschrift 37 (1937): 289–94, here 289; English Translation: “Indicative Errors and Stemmatic Types” (appendix 1) in Paul Maas, Textual Criticism (Oxford: Clarendon, 1958). 21 Maas, Textual Criticism, 49. 22 Giorgio Pasquali, “Review of Maas, Textkritik,” Gnomon 5 (1929): 417–35, 498–521; Pasquali, Storia della tradizione e critica del testo, 2nd ed. (Florence: Felice le Monnier, 1971) (1 st ed. 1934).

The Development of the Coherence-Based Genealogical Method

441

exemplars are lost) it is necessary to instead attempt to reconstruct intermediary hyparchetypes. The transmission of the Greek New Testament produced several features with which conventional stemmatology could not cope. There is an overwhelming mass of medieval manuscripts produced from the ninth century onward which, for the most part, contain a standardized Byzantine text. This text, although clearly distinguished from older text forms in many places, developed in different varieties, and these varieties often contain or, to use the pejorative term, are contaminated with variants from older text forms. Additionally, there are a number of witnesses centuries older, the most comprehensive of which date from the fourth and fifth centuries. Our earliest manuscripts, which are more or less fragmentary papyri, date from the second to fourth centuries. The early witnesses often share variants which distinguish them from the Byzantine text, but, when compared with each other, they preserve many more differences than the Byzantine witnesses among each other. In other words, the texts preserved in older manuscripts are heterogeneous to a higher degree than the texts preserved in later manuscripts. This situation is a consequence of the fact that most manuscripts from the first millennium are lost. The documents produced before the ninth century are single survivors of larger traditions and support different text forms in constantly varying combinations. Most links from before the ninth century are missing. This is the reason why the New Testament tradition appears to be so highly, and some say hopelessly, contaminated. An attempt to construct a stemma codicum for such a tradition would be a hopeless enterprise in Maasian terms. The CBGM takes up this challenge and starts by assessing the genealogy of variants assembled in a full critical apparatus. On this basis, coherent structures are traced in the manuscript tradition. The assessments of variant passages are expressed by local stemmata of variants, meaning a stemma for each queried variant unit is constructed. A simple example may look like this:

a b

c

Fig. 1. Sample Local Stemma

A statement about the relationship between variants implies a statement about the relationship between the texts containing these variants at the same time. If variant a is prior to b and c, it follows that this passage is an instance of variation where the states of text containing a are prior to those containing b and c. As a result, the following principle is crucial for the CBGM: A hypothesis about genealogical relationships between the states of a text as preserved in the manuscripts has to rest upon the genealogical relationships between the variants they exhibit.

442

Klaus Wachtel

Therefore a systematic assessment of the genealogy of these variants (displayed as local stemmata) is a necessary requirement for examining the genealogy of textual witnesses.23

If local stemmata for each variant passage have been constructed, we can calculate the number of instances where witness X has the prior variant as compared with witness Y at the places where they differ. Due to contamination, there will also be a number of instances where Y has the prior variant. Moreover, there will be a number of unclear cases where the variants of X and Y are not directly related or cases where it has to be left open as to which is the prior one. These numbers are tabulated in tables of potential ancestors and descendants. Table 1. Potential Ancestors of witness 35 (W1)24 W2

NR

PERC1

EQ

PASS

617 424 468 A 025 323 1739 03 04 P74

1 2 3 4 5 0 6 7 8 0

95.995 95.988 95.588 92.263 91.160 89.638 87.853 87.272 87.262 82.493

2924 2919 2903 2695 2444 2725 2676 2633 1836 278

3046 3041 3037 2921 2681 3040 3046 3017 2104 337

W1< W2 56 51 57 212 99 111 158 201 103 27

W1> W2 46 45 49 0 84 111 115 78 93 22

UN CL 15 23 27 6 46 76 77 90 60 6

NOREL 5 3 1 8 8 17 20 15 12 4

Version 1.0 W2: Manuscript numbers of potential ancestors – NR: Ranking numbers according to degrees of agreement – PERC1: Percentage of agreement with 35 (=W1) – EQ: Number of agreements with 35 – PASS: Passages shared by 35 and W2 – W1W2: Number of priority variants in 35 (=W1) – UNCL: Unclear relationship between W1 and W2 – NOREL: No relationship between W1 and W2.

This table shows the result of a comparison of GA 35 (i.e. the state of text of the Catholic Letters preserved in the minuscule manuscript 35) with all other witnesses included in the ECM apparatus. The view presented in the table is focused on the potential ancestors of 35. The most closely related, and, therefore, first

23 G. Gäbel, A. Hüffmeier, G. Mink, H. Strutwolf, K. Wachtel, “The CBGM Applied to Variants from Acts – Methodological Background,” TC: A Journal of Biblical Textual Criticism 20 (2015) http://rosetta.reltech.org/TC/v20/TC-2015-CBGM-background.pdf [accessed 10 March 2018]. The basis for this comprehensive analysis is a critical apparatus comprising all variants of every Greek textual witness selected for the edition. 24 For the sake of consistency I am referring to a table produced by Genealogical Queries 1.0 (http://intf.uni-muenster.de/cbgm/en.html [accessed 10 March 2018]) which Mink used in his Introductory Presentation. Version 2.0, based on the second edition of the Catholic Letters in the ECM, is available on the same website. Data Source: Cath. Letters (excl. small fragments and extracts).

The Development of the Coherence-Based Genealogical Method

443

potential ancestor of 35 is 617. Both of these witnesses are extant at all 3046 variant passages in the ECM apparatus. They agree at 2924 or 96% of these passages. In 56 instances of the remaining passage where they disagree, GA 617 witnesses the priority variant, in 46 instances it witnesses the posterior variant, and in 20 instances the relationship is unclear or there is no direct relationship. As a result, the predominant textual flow goes from the state of the text in 617 to the state of the text in 35. So the statement, “617 is the first potential ancestor of 35” rests on an objective fact (617 and 35 are equal at 96% of the variant passages they share) and a summary of editorial assessments at the passages where these witnesses differ (617 contains quantitatively more variants that are judged as prior to those preserved in 35).25 Based on such analysis for all included witnesses, textual flow diagrams can be plotted for the attestations of each variant. The corresponding module of Genealogical Queries, “Coherence in Attestations,” was extensively used in editing the ECM volumes that have already appeared.26 Here I would like to focus on a long-term objective of the CBGM which, although conceptualized from the beginning, has not yet been realised completely: the global stemma. For the construction of a global stemma of witnesses, a preliminary methodological step is necessary which defines an optimal substemma for each witness. An optimal substemma comprises only the ancestors that are necessary to account for the individual text of a witness. Mink’s Introductory Presentation

25 It has been shown elsewhere that such evaluations of the relationships between individual witnesses and the emerging overall picture of the structure of the transmission is now taking the place of the old text type theories; see, e.g. David Parker, An Introduction to The New Testament Manuscripts and Their Texts (Cambridge: Cambridge University Press, 2008), 171–74. For recent descriptions of text forms without recourse to text type terminology, see the contributions of Strutwolf, Gäbel, and Wachtel in ECM Acts part 3. 26 For the purpose of the present essay it suffices to point to the following publications that explain the composition of textual flow diagrams and their use in editing, especially with regard to the reconstruction of the initial text: G. Mink, “The Coherence-Based Genealogical Method (CBGM) – Introductory Presentation” (2009), online at egora.uni-muenster.de/intf/service/downloads.shtml, 204–303; Mink, “Contamination, Coherence, and Coincidence,” 162–89; G. Gäbel et al., “The CBGM Applied to Variants from Acts”; K. Wachtel, “Constructing Local Stemmata for the ECM of Acts: Examples,” TC: A Journal of Biblical Textual Criticism 20 (2015) at http://rosetta.reltech.org/TC/v20/TC-2015-CBGMexamples.pdf [accessed 10 March 2018]. Furthermore two recent publications have to be mentioned here, Peter Gurry, A Critical Examination of the Coherence-Based Genealogical Method in New Testament Textual Criticism, NTTSD 55 (Leiden: Brill, 2017) and Tommy Wasserman and Peter Gurry, A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method (Atlanta: SBL, 2017). Both provide useful introductions to the CBGM. A brief and precise description of the methodology is provided by Peter Gurry: “How your Greek NT is Changing: A Simple Introduction to the CoherenceBased Genealogical Method (CBGM),” JETS 59 (2016): 675–89.

444

Klaus Wachtel

contains a very clear explanation of the procedure leading to the production of an optimal substemma (pp. 164–77). In his Introductory Presentation, Mink demonstrates the procedure in detail for 35.27 As a result, four witnesses emerge, 617 468 025 and 1739. A global stemma consisting of a network of optimal substemmata meets the condition that Maas formulated for a stemma based on an immaculate manuscript tradition, which has to be true at each variant passage. This means that the stemmatic ancestors of each witness agree entirely or support a variant prior to the one in the descendant. 617

468

025

1739

35 Fig. 2. Optimal Substemma for 3528

So far only the top of the global stemma for the witnesses included in the ECM of the Catholic Letters has been constructed.29 Note how the relationships displayed in the optimal substemma of 35 are included in the global stemma. Necessary edges from all four ancestors picked for the optimal substemma of 35 point to this witness. In addition there is one connection that may be superfluous, or help to optimize the substemma in a subsequent iteration of the analysis. At the top of the stemma there is A, the reconstructed initial text that by definition has no ancestors. For the witnesses preserved in manuscripts the potential ancestors that were picked for the optimal substemmata have become stemmatic ancestors, i.e. ancestors necessary for explaining the text of each of the witnesses. Note that for every optimal substemma comprised in this graph the closest potential ancestor (the one with the highest agreement rate) is part of the selection. For 03, A is the only stemmatic ancestor, which reflects the importance of 03 for the reconstruction of the initial text. This confirms the traditionally high estimation of 03. Likewise, the high positions of 1739 and 04 will not come as a surprise. It would require another essay, however, to explain why well-known “Alexandrian” witnesses like 01 02 33 and 1175 are not at the top of the global stemma for the Catholic Letters, but instead 35 424 468 and 617. These are all witnesses with high percentages of agreement with the majority text, but also with our reconstruction of the initial text. This is because they differ from the Mink, “Introductory Presentation,” 475–560. Mink, “Introductory Presentation,” 560. 29 Cf. Mink, “Introductory Presentation,” 561–74. The original figure which had to be reproduced in black and white for this publication is on p. 562. 27 28

The Development of the Coherence-Based Genealogical Method

445

initial text almost only where the majority of witnesses differ from it as well. For now it may suffice to state that the aforementioned four well-copied representatives of the majority text are closer to the reconstruction of the initial text (and by the same token to 03) than 01 02 33 and 1175.

necessary edge with the highest agreement

of ancestor and descendant

A

necessary edge with lower agreement

of ancestor and descendant

03

possibly superfluous edge

undirected edge

1739

04

468

025 617

424

35

Fig. 3. Top of the Global Stemma for the Catholic Epistles

Conclusion: An Approach for Coping with Contamination The possibilities of the CBGM illustrate that stemmatic analysis of complex traditions is not a remnant of nineteenth century positivism, but a mode of analysis that is essential for understanding the structure of the tradition. There remains much work to be done in this vein of research. The following four statements are key points that need to be considered in carrying out this research, especially as the CBGM begins to be practiced by those outside the small group of editors who designed it. (1) Do not try to reconstruct the manuscript tradition, but focus on the states of text preserved in the manuscripts. (2) Do not try to use common errors to define strands of transmission because diorthosis is an integrated part in the manual reproduction of texts. Instead, base your research on a well-constructed apparatus of grammatically valid variants and use tabulated rates of agreement (pre-genealogical coherence) to determine the proximity of each state of text. (3) Do not try to reconstruct hyparchetypes because, for all richly transmitted texts from antiquity, you would need several generations of them, piling up sub-

446

Klaus Wachtel

hypotheses about stages in the development of the text. Instead, trace the structures in the preserved states of text and order them according to rates of agreement (pre-genealogical coherence) and according to the genealogy of the variants contained in them (genealogical coherence). (4) To explain the development of the text as documented in extant witnesses (i.e. states of text), determine for each of them their stemmatic ancestors, i.e. the ancestors needed to explain the respective state of text by agreement and descent. Read from the bottom up, the resulting stemma will account for the reconstruction of the initial text at its top.

4. Outlook: The CBGM and Digital Editing Access to “Genealogical Queries” is an integrated part of the new online representation of ECM Acts which is available here: http://ntvmr.uni-muenster.de/nttranscripts. Enter “Acts” plus chapter and verse, and the text of the selected verse will appear. The Greek manuscript witnesses are linked to the transcription and an image of the page containing the verse. In the upper margin of the apparatus frame, there are links to the database of patristic citations, a line-by-line transcription of Latin witnesses, and the Amsterdam conjectures database. Hovering the mouse over the left margin of the apparatus, two blue icons will appear, a balloon and a downward arrow in a circle. Clicking the latter will take you to the local stemma and textual flow diagrams for the respective variation unit in “Genealogical Queries – Acts (Phase 4).” Clicking on the balloon icon will open a textual commentary if available.30 Both of these features are the starting points for an interactive digital edition.31 Users will have the option to open their own editorial workspace in the future where they can edit the local stemmata and publish their views in an online textual commentary. Hitherto, the use of the Genealogical Queries modules is the online part of the documentation of the ECM editors’ discussions and decisions. In the future it will be an integrated part of the open digital edition of the New Testament.

Thus far the online commentaries are reproductions of the commentary printed in ECM Acts part 3. They reflect the views of the editors of ECM Acts on passages with a text differing from NA28, passages with a split guiding line, and on some text-critically difficult passages where the text remained the same. We are now planning on inviting a group of experts to publish new commentaries in this section of the NTVMR forum. All registered users can respond to existing comments. 31 This description reflects the state reached at the time of writing this study (February 2018) and it will probably have developed further when this essay is in print. 30

Apparatus Construction: Philological Methodology and Technical Realization Annette Hüffmeier What a dry title! Perhaps I might distract you by asking a more interesting question that evokes some stimulating answers: what do we expect from a modern critical edition of the New Testament? Herman Charles Hoskier would have answered, somewhat optimistically, “the Truth,” as he wrote in the preface to his Full Account and Collation of the Greek Cursive Codex Evangelium 604 in 1890. Recounting his famous encounter with Dean Burgon on the stairway at midnight, Hoskier quotes Burgon as saying: ‘As surely as it is dark now, and as certainly as the sun will rise to-morrow morning, so surely will the traditional text be vindicated and the views I have striven to express be accepted. I may not live to see it. Most likely I shall not. But it will come.’ The way is not clear yet, and the sun has not yet risen, but I believe those words to-day much more than I did then…And as the Truth must ever triumph, soon or late, in every exact science, so shall it be in this department of Biblical learning, even though it be by means of instruments so poor, so inconsistent, so weak as we are.1

Twenty years later Hoskier described “the Truth” a bit less emotionally and in more scientific detail in the preface to his book Concerning the Genesis of the Versions of the N.T. (1910): I wrote in 1890 that we had laid no certain foundation on which to build up a scientific textual theory, and felt that we must very likely retire in order to advance satisfactorily. The keynote to the proper position was struck by the Abbé Martin and others – (indeed Burgon’s whole contention amounts to no more than this!) – and was that the truth is to be recovered from the side-testimony of documents, irrespective of their age, by deductions as to their parentage and stems, taking into account local and other influences at work on the actual scribes or editors…In other words, work like that of Scholz or Lachmann is comparatively worthless, and it is with minutiae we have to deal at the outset.2

A few pages later, Hoskier continues: H. C. Hoskier, A Full Account and Collation of the Greek Cursive Codex Evangelium 604 (with two facsimiles) (London: David Nutt, 1890), v. 2 Hoskier, Concerning the Genesis of the Versions of the N.T. Remarks suggested by the Study of JP and the Allied Questions as Regards the Gospels (London: Bernard Quaritch, 1910), vii. 1

448

Annette Hüffmeier

Partial conclusions one might have drawn, but the evidence was only fragmentary, and while throwing decided light on certain phases of the subject, it must take its proper place with other material, and be weighed in the future, and in the light of still further investigation. To deduce from fragmentary and incomplete evidence is a crime. We have wasted too much time since Stephen’s and Beza’s days by imperfect collation and examination of documents, so that we are still far behind our objective point of reducing a mass of observations from chaos to a scientific working basis…The whole question is more or less academic, and the study of minutiae first is the only key which will unlock the remaining mysteries, and that is so irksome it is generally shirked. Exhaustive methods are the ones worth using, and accurate transcription of photographic copies the only ways of presenting the primary evidence of important documents.3

Fear not. In this essay I will not evaluate Hoskier’s remarks or mind-set in any detail, but I will discuss a promising “road to Burgon’s sunrise” that meets Hoskier’s standard for exhaustive study. At the same time, this “road” offers a substantial answer for my opening question: what do we expect from a modern critical edition of the New Testament now one hundred years after Hoskier? I think – and this may not come as a surprise – that it is the same claim that we are trying to deliver every single day by advancing the Editio Critica Maior (ECM), whether we are in Münster, Birmingham, Wuppertal, the United States, Amsterdam, Vienna, Örebro,4 Dublin, Göttingen, or wherever else we have partners across the world collaborating on this huge project. The goals of the edition as articulated by the editors of ECM Acts, which shape what users should expect, are as follows: The Editio Critica Maior provides the full range of resources necessary for scholarly research in establishing the text and reconstructing the history of the New Testament text during its first thousand years. These include: – all the variants found in the selected Greek manuscripts and in citations by the Greek Fathers; – the evidence of the three [plus one] most important early versions (Latin, Coptic, Syriac and Ethiopic) where they witness to variants in the Greek text; – the evidence of other versions (Armenian, Georgian and Old Church Slavonic) where they witness to variants in the Greek text, and to the extent they are available in editions.5

Hoskier, Concerning the Genesis, x–xi. I would like to particularly thank our colleague in Sweden, Tommy Wasserman, for his chapter entitled “The Editio Critica Maior: A Milestone in the History of Textual Criticism” in his book The Epistle of Jude: Its Text and Transmission (Stockholm: Almqvist & Wiksell, 2006), 22–5. On p. 23 he quotes William L. Petersen regarding the challenges involved in constructing and presenting an apparatus criticus: “An apparatus is obliged to be a butterfly and an elephant at the same time: an elephantine amount of information must be conveyed in an exquisite minimum of space.” 5 Holger Strutwolf, Georg Gäbel, Annette Hüffmeier, Gerd Mink, and Klaus Wachtel, eds., Novum Testamentum Graecum. Editio Critica Maior. III: Die Apostelgeschichte, 1.1 (Stuttgart: Deutsche Bibelgesellschaft, 2017), 18* (hereafter ECM Acts). 3 4

Apparatus Construction

449

All of this, then, leads to a “text…established afresh on the basis of these resources.”6 Despite the greatness of this achievement as such, a collection of data of which Hoskier could only dream, I do not believe what we have accomplished thus far is sufficient for the discipline, given the necessities of textual scholarship for present day research. Thus it seems relevant that we, the editors of the ECM (as well as those of the Nestle-Aland edition) must not only be aware of the history of the Greek tradition, but we must also consciously present a text as a hypothesis about the initial text, which differs from earlier editions. Truth as Hoskier conceived it is not the business of our editorial endeavours, but we strive to uphold the scientific principle of hypothesis based on the best available data. The quality of this hypothesis may be put to the test by the reader who asks whether the variant accepted into the reconstructed text of the edition has the capacity to explain the emergence of the other variants at the same passage. In many cases,7 the ECM editors even express textual doubt by splitting the guiding line and using a diamond siglum to mark a variant as equally suitable for the reconstruction, thus asking the reader to discuss the passage in question. The aim is no longer to simply state the truth once and for all, as Hoskier might have desired, but to engage in an infinite process of approximation to the initial text in a scholarly discourse. In other words: we, as editors, act critically towards tradition and ask readers to act critically towards our decisions; we even hope to enable readers to make their own choices. This process presupposes a very thorough, transparent documentation of each step of our work. In the long run it requires a global stemma, as well as a fully digitized edition and interactive tools. Regarding the last two points we are on our way, supported by the Cologne Center for eHumanities (CCeH),8 and there are a number of databases and online-tools9 that are already available.

6 For ECM Acts this led to 52 changes in guiding lines. See Klaus Wachtel, “Notes on the text of the Acts of the Apostles,” in ECM Acts 1.1, 28*, for the complete list see pp. 34*– 5*). 7 155 cases for ECM Acts. Cf. Wachtel, “Notes,” 28*, for the complete list see pp. 35*– 7*). 8 Additionally, a Centre for Digital Humanities is currently being founded at the University of Münster. 9 E.g. for all Greek patristic quotations up to the tenth century (soon to be open to the public), for the SMR-list of Coptic manuscripts (http://intf.uni-muenster.de/smr/, accessed 22 February 2018), or for palaeography (http://intf.uni-muenster.de/NT_PALAEO/, accessed 22 February 2018). For more see our “online utilities“ (http://egora.uni-muenster.de/intf/, accessed 22 February 2018). For digital tools, cf. “Genealogical Queries 2.0” for ECM Catholic Letters (with the five components: Potential Ancestors and Descendants, Comparison of Witnesses, Coherence in Attestations, Coherence at Variant Passages and Local Stemmata; http://intf.uni-muenster.de/ cbgm2/GenQ.html, accessed 22 February

450

Annette Hüffmeier

In what remains of this essay, I concentrate on our foundation for establishing a hypothesis about the initial text, i.e. a well-constructed, machine-parsable critical apparatus. This apparatus can then be used for applying statistical methods like the Coherence-Based Genealogical Method (CBGM) which lies at the heart of textual decisions that comprise the ECM. Since many of our steps would remain unintelligible without taking into account a certain degree of subjectivity, interpretation, and the definition of our own parameters, I start at the very beginning. The process of producing an apparatus draws upon the profound framework of the Text und Textwert der griechischen Handschriften des Neuen Testaments series for the process of selecting Greek manuscripts for the inclusion in the editions. This allows transcribers to focus on a set of relevant manuscripts, which are then transcribed and collated electronically by generations of eager scholars, students, and volunteers.10 Each witness is transcribed twice in order to reduce the number of errors and these two transcriptions are carefully compared and, where necessary, corrected by a third party. A final transcription is then published in the New Testament Virtual Manuscript Room (NTVMR 2.0, developed by Troy Griffitts and Ulrich Schmid), side by side with photographs of the original manuscript. I am sure that Hoskier would have been proud of this protocol.11 The transcription process proved taxing and technically difficult. Apart from the expected instances of difficult handwriting and poor quality of the material, other hurdles included reconstructing lacunae, assessing damaged or otherwise illegible passages, and identifying different scribal hands. All these problems required revisiting during the electronic collation of the Greek manuscripts, especially in light of the complex outputs that these transcriptions created. These same issues were tackled also by us, the editors, who are responsible for making sense of the raw transcriptions. We were assisted by software, like CollateX, an achievement of software design by the Interedition Development Group,12 in combination with the Collation Editor, a program designed by Catherine Smith of the Institute for Textual Scholarship and Electronic Editing (ITSEE) in Birmingham. The Collation Editor provides an easy means for regularising orthographical issues, delineating variation units, and reordering these variants. As a result of 2018) and now for “Acts – Phase 4” (with two components: Coherence and Stemmata and Comparison of Witnesses; ntg.cceh.uni-koeln.de, accessed 22 February 2018). 10 In the case of ECM Acts 183 Greek manuscript witnesses, including twenty-eight parchment and papyrus fragments, nine lectionaries and twelve supplements. In the case of ECM Mark, we will have 200 Greek manuscript witnesses, including ten lectionaries and eight supplements. 11 http://ntvmr.uni-muenster.de/ [accessed 22 February 2018]. The NTVMR comprises numerous tools: the transcription editor, NTTranscripts, Imageviewer and NTConjectures, the Amsterdam Database of New Testament Conjectural Emendations. 12 collatex.net [accessed 22 February 2018].

Apparatus Construction

451

this nonetheless complex procedure – to which I will return in more detail – all Greek witnesses are recorded once at each variant passage13 and can thus be evaluated with the assistance of this data organised through the use of the software. Once this data is processed, our web designer, Volker Krüger, feeds the revised output of the Collation Editor into several databases of his own creation: one for the Greek critical apparatus itself, one for each version (mainly the Latin, Coptic, Syriac and Ethiopic), and one for the patristic specialists. At this point, with data accurately fed into their various databases, we, the editors of the ECM in Münster, are finally able to choose those variant passages for which we require versional and patristic data from our colleagues dealing with the relevant evidence whenever it “can possibly contribute to establishing the Greek text or to its history.”14 The criteria for identifying these important passages where versional data are necessary for ECM Acts in particular were defined by following criteria:15 (1) The passage features variants which have significant manuscript support and exhibit translatable linguistic differences;16 (2) In the versional or Greek patristic traditions there are variants which very probably go back to the Greek but are not preserved in the extant Greek manuscript tradition.17

In order to assess the second point of these criteria, we developed a fuller set of principles for evaluating versional, and, for ECM Mark, patristic evidence, that likely derives from now-lost Greek readings: 13 In the 28 chapters of Acts (1,000 verses with 18,500 words and more than 26,000 readings) we have nearly 7,500 variant passages; in Mark’s 16 chapters (nearly 680 verses with 11,300 words and 15,800 readings) we will have more than 4,000 variant passages. 14 ECM Acts 1.1, 20*. This is in contrast to ECM Catholic Letters and ECM John (the latter to be published in Birmingham). For ECM Acts we did not restrict the patristic evidence, but based on this experience, we changed course for ECM Mark. 15 In more than 3,100 places in ECM Acts, “a distinction is made between passages where recording the full versional evidence was mandatory and those passages where the versions were entered only if they could be clearly attributed to a Greek variant” (ECM Acts 2, 127) and “if the linguistic difference could be expressed” (ECM Acts 1.1, 20*). 16 “For selecting the passages at which versions should be cited [in ECM Acts] an attestation was regarded as significant, if it included at least one variant supported by at least two of the following witnesses: P29, P38, P45, P48, P74, P127, 01, 02, 03, 04, 05, 08, 33, 81, 94, 180, 181, 307, 453, 610, 614, 623, 1175, 1292, 1409, 1505, 1611, 1642, 1678, 1739, 1875, 1890, 2138, 2344, 2818, L1188s1. This list contains witnesses featuring a particularly close relationship to the reconstructed initial text, further several manuscripts commonly regarded as forms of the ‘Western’ text (P29, P38, P48, P127, 05, 08, 1884), and additionally representatives of early text forms transmitting special material (P45 and the Harklensis group with 614, 1292, 1505, 1611, 1890, 2138)” (ECM Acts 1.1, 20* n.7). 17 “In these [several hundred] cases the overview of variants provides a Greek retranslation, while the full apparatus shows the variant in the original language on which the retranslation is based” (ECM Acts 1.1, 20* n.8; see ECM Acts 2, 127–29 for the exact description and full list).

452

Annette Hüffmeier

Variants with exclusively versional [in ECM Mark also patristic] attestation were included, if they were assessed as probably going back to a Greek exemplar and intraversional emergence appeared unlikely. Such probability is based on the following criteria: – The variant is supported by at least two versions. This criterion is particularly decisive if the versions are in different languages. Two different versions, however, can also be written in the same language, as, for example, the Sahidic and Bohairic versions in Coptic. – Many variants preserved only in the Harclensis (S:H) are recorded, because the translator is known to conform very accurately to his Greek exemplar. – Other variants, which were exclusively transmitted in single versions were included if they were considered in realm of ‘Western’ paraphrasing or extension of the text. This applies particularly to the Middle-Egyptian and to the Latin text in 05. – If, in addition to a single versional witness of a variant, there is a Greek patristic citation [or the other way round], this constellation is regarded as criterial for listing the variant.18

I offer one example only which fulfils many of these criteria in a single reading: Acts 9:40/3919 where Peter is said to have brought back to life Tabitha through prayer and the command Ταβιθά, ἀνάστηθι. Following Peter’s words, multiple patristic and versional witnesses provide a longer formulaic statement with some internal variation, including two Latin church fathers, five more Latin manuscript witnesses (out of eleven here altogether), the Sahidic, the MiddleEgyptian, the asterisked Harclensis, the Armenian, and the Georgian. We retroverted the Greek as follows: 39 b (εν τω ονοµατι του κυριου ηµων ιησου χριστου) 39 a om. …Chrys. L:V. 50. 56. 58. 70. 189. K:B. S:PH. Ä. Sl b in nomine domini nostri iesu christi L:AM. CY. 51. 54. 57. 6 6 61. 67>. K:S>M. S:HA>. A. G - 014. 88 9,40/39b) L 51 54 57 AM in nomine domini nostri iesu christi. 61 in nomine domini I iesu. 67 in nomine domini iesu christi. CY in nomine iesu christi. 9,40/39b) K:S om κυριου ηµων [K:Smms (sa 17, sa 391)] 9,40/39b) S:HA Der ganze Passus sub aster. / The whole reading sub aster. Fig. 1. Versional Support for an Addition at Acts 9:40/39

In addition to the complex standards for gathering and evaluating data, we simultaneously began applying the CBGM to the material. This happened in iterative phases, using pre-genealogical and genealogical coherence, the classic philological tools of textual criticism, and the versional and patristic evidence. Within this stage pre-genealogical coherence is objective, just counting (dis)agreements 18 19

ECM Acts 2, 127; for the full list see pp. 128–30. GA 05 has no text here, but all other “Western” witnesses preserve variant a.

453

Apparatus Construction

between states of text, whereas genealogical coherence implies a subjective element in assessing the relationships between variants. Our main tools here are the construction of local stemmata as a newly derived external criterion, and the creation of textual flow diagrams displaying relationships between witnesses within attestations based on lists of potential ancestors and descendants for each of the witnesses. The construction of local stemmata is an iterative process as shown by the following diagram:

internal criteria explanations - reasons

external criteria prejudices pregenealogical coherence

local stemmata

revision of local stemmata

genealogical coherence

growing knowledge of  genealogical coherence

revision of prejudices

Fig. 2. The Process of Constructing a Local Stemma

After weighing all arguments and arriving at all necessary decisions within our editorial committee, the main responsibility for establishing the master copy of the apparatus moved on to Volker Krüger. In short, after many complex controlling procedures he went on to extract the information from all our different databases for two purposes: (1) the material from these databases served as the basis for the digital publication of the ECM in our New Testament Virtual Manuscript Room (http://ntvmr.uni-muenster.de/nt-transcripts); (2) Volker transformed this same data into word files and then, with the help of a desktop publishing software, produced the .pdf-Vorlage for the final print edition of the ECM. ***

454

Annette Hüffmeier

With an overview of the process in places, it is now helpful to examine this progression in light of some concrete examples. I have chosen some examples from the Gospel of Mark that will better illuminate the process of constructing the critical apparatus for this next fascicle of the ECM. The first issue that is central to apparatus construction is the question of basic definitions. In the process of regularisation we have to distinguish carefully between the concepts of reading, variant, error, and orthographicum20 because we want to identify all readings that are less valuable for textual criticism. This process is where interpretation starts. In some cases a reading that at first sight appears to be perfectly normal has to be regularised since it makes no sense in its specific context. In Mark 5:2–20 we meet the man “from the tombs” (ἐκ τῶν µνηµείων). He is described as someone ἐν πνεύµατι ἀκαθάρτῳ, who, after numerous fruitless attempts, no one was able to bind anymore, δῆσαι (5:3/30). Apart from the proper variants δαµασαι (“to tame”) and κρατησαι (“to get under control”) we find also the non-sense reading δισαι, alongside δεισαι (“to become frightened”), δυσαι/δυσε (“to plunge”), and λυσαι (“to free, loosen”, ie. the opposite of “bind”). Of course we all know how the erroneous readings came into being: ΔΗΣΑΙ < (because of vowel interchange) ΔΙΣΑΙ or ΔΕΙΣΑΙ or ΔΥΣΕ; or ΔΥΣΑΙ < (because of graphical similarity) ΛΥΣΑΙ. These mechanical changes in the tradition were regularised in the collation process according to the reading from which they derived. I think there can be no doubt that we regularised all these mistakes correctly. In another example, possible errors, like dittography at a page change, might lead to a proper – even if singular – new variant. In Mark 3:12 “he,” meaning Jesus, whose name had been mentioned earlier in 3:7, forbids everybody to make him φανερόν after having healed numerous people. At this point GA 1071, which has a tendency to add singular readings from time to time,21 closes one page with

20 “A reading is the generic term for the wording of a textual unit in which a manuscript is distinguished from one or more or from all other manuscripts. A variant refers to one of at least two readings of the same textual unit which is grammatically correct and logically possible in its context. Errors are readings which do not fulfil these criteria…Exceptions are made only for the most frequent kinds of vowel interchange (αι-ε, ε-η, ει-η-υ-ι-οι, ο-ω), and errors caused by dropping one of a double consonant or doubling a single consonant. When deciding whether or not a reading should be considered an error, the principle is strictly observed that even the remotest possibility of making sense will qualify its inclusion as a variant.” And preceding this statement on the same page: “Alternative and orthographically possible forms of the same variants are classed as orthographica. Examples include certain morphological differences, such as the omission of the present stem element -µ- from the future forms of λαµβάνω, or the occurrence of mixed for strong aorist forms…All orthographica are recorded except for the movable ς and the movable ν῾ (ECM Acts 1.1, 24*). 21 E.g. in Mark 3:28/37 or 34/33.

Apparatus Construction

455

και πολλα επετιµα αυτοις. The case ending of αυτ-οις (dative plural masculine) was abbreviated by a raised ς:

Fig. 3. Mark 3:12 in GA 1071

The next page, however, opens with οις̅ ινα µη φανερον αυτον ποιησωσιν·… οι̅ς̅ (word 9), the dative masculine plural ending that was abbreviated on the previous page is repeated, and is written like a nomen sacrum for ιησους with its article ο.

Fig. 4. Repeated Ending at Mark 3:12 in GA 1071

I think it is obvious that the explicit beginning of 3:12, “and Jesus forbade,” instead of the implicit, but nonetheless understandable “and he forbade,” goes back to αυτ(οι)ςοις, and we do not have to regularize it as a mistake of αυτοις without ο ιησους, but should accept it as a legitimate variant in its own right. Another question that arises from the construction of the apparatus is the difference between errors and orthographica. In Mark 1:19 when Jesus calls his first disciples he moves on with Simon and Andrew and immediately spots the brothers Jacob and John. Jesus’ act of “seeing” is described in word 8 by ειδεν. More than 50 out of our 200 witnesses use the frequent orthographicum ιδεν and merely one, GA 792, reads οιδεν. Can “he knows Jacob and John” here really be meant? Indeed not, but perhaps “he gets to know” might be possible, at least before we finish the sentence which describes how Jacob and John mend their fishing nets in their common boat. I think that “he saw” is the only reasonable option, an option that regards οιδεν as a mistake, due to vowel interchange with either ειδεν or its orthographicum ιδεν. Therefore, οιδεν would be regularised as a sub-reading of ειδεν. The question then arises: how do we distinguish orthographica from variant readings? As a rule we accept, and include in the apparatus, all alternative readings – even orthographica – that we can find in dictionaries (like LSJ, Bauer &

456

Annette Hüffmeier

Aland,22 sometimes the Thesaurus Linguae Graecae23) and grammar books (like BDR24). In a few instances we build on analogous sets of data. We tend not to accept poetic and very late forms. As a result of our work on ECM Acts and ECM Mark we have long lists of vocabulary, but I will not bother with details here, since quite often clear differentiation between forms is hard work. The most difficult problems in this regard are with the treatment of proper names. Another issue that arises is the question of what to do with nomina sacra. Should they be replaced by their unabbreviated forms? In Mark 3:5 Jesus talks to the man whose withered hand or arm he is about to heal. In words 28–30 many manuscripts use the abbreviated form τω ανω(ι) for the dative singular masculine τω ανθρωπω. Regularisation is not a problem in this instance; the abbreviated form is subordinated to the unabbreviated form. A problem occurs, however, in GA 979, which writes τω κω, the nomen sacrum of τω κυριω. Since we cannot seriously assume that Jesus is talking to God the Lord here, we must find another solution. Could κυριος refer to the “owner”25 of the withered upper limb? Or maybe the scribe of GA 979 misread the nomen sacrum τω ανω, which he had used two verses earlier, as the proper word τω ανω “to the one above,” and replaced it by substituting it for the clearer (at least in his mind) τω κω? This reconstruction at least makes it possible to understand it as a mistake from τω ανω(ι). I think the latter option is the more persuasive one. Therefore, we could have called it a pure mistake of variant a, but in order to make the reading visible we decided to introduce a variant of its own, called bf. We also tend to follow the rule that a scribe’s self-corrections are more relevant for the apparatus than the original, uncorrected reading. This principle usually works very well, although sometimes specific readings induce hesitation. Before Mark 5:31 the woman who has been suffering from bleedings for years had touched Jesus’ clothes in the not unfounded hope of healing. In response to this action, Jesus asks, “who touched my clothes?” When his disciples answer they do so by repeating his question in an abbreviated form, “and you say: who touched me?” They omit the last genitive construction των ιµατιων. There is one manuscript, GA 038, which repeats the complete question, adding again the last genitive των ιµατιων in word 33. A second witness, GA 153*, also has των ιµα̣τι̣ in its initial reading. The same scribe, GA 153C*, adds an abbreviated και in front of the new line and overwrites των ιµα̣τι̣ with the following περιεβλεπετο.

22 Walter Bauer, Griechisch-deutsches Wörterbuch zu den Schriften des Neuen Testaments und der frühchristlichen Literatur, ed. K. Aland and B. Aland, 6 th ed. (Berlin: De Gruyter, 1988). 23 http://stephanus.tlg.uci.edu [accessed 22 February 2018]. 24 Friedrich Blass and Albert Debrunner, Grammatik des neutestamentlichen Griechisch, ed. F. Rehkopf, 18 th ed. (Göttingen: Vandenhoeck & Ruprecht, 2001). 25 Bauer & Aland, s.v. κύριος 1.

Apparatus Construction

457

Fig. 5. Mark 5:32 in GA 153

If we follow the above rule we must assume that the scribe of GA 153 realised immediately that in 5:31 he had unthinkingly repeated the end of 5:30 (τις µου ηψατο των ιµατιων) even though his Vorlage did not contain των ιµατιων. The scribe wanted to correct this careless mistake and did so immediately. Alternatively, we might think that the scribe knew of a variant (and perhaps even had it in his Vorlage) that read των ιµατιων and consciously chose not to add it, since he himself judged it to be a careless, misguided echo of 5:30. In constructing an apparatus, we also have to tend with very long lacunae or lacunae at the beginning or at the end of a line which are very frequent. We often end up denoting them with the siglum ↔ (“double arrow-cases”), which indicates “that it cannot be determined which of the variant readings noted is supported.”26 (Partly) illegible words transcribed with underdots or lacunae within a word are similarly problematic, but as a rule easier to handle. Our list of lacunae not only comprises lacunae in the usual narrow sense, but also those rather long and really extensive omissions – often because of parablepsis – which do not make sense in their current forms.27 In Mark 3:17 at the beginning of the list of the twelve apostles, GA 117 omits και ιακωβον τον του ζεβεδαιου between πετρον and και ιωαννην τον αδελφον του ιακωβου, presumably due to the very common parablepsis based on the repetition of και. Perhaps in this special case the twofold use of ιακωβον/ιακωβου and the homioteleuton ζεβεδαι-ου/ιακωβ-ου may have contributed to the omission as well. The syntax of the phrase as it stands in GA 117 is still acceptable, but the number twelve, which GA 117 mentions in 3:14 rather than 3:16, can no longer be correct because the number of names only adds to ten. So GA 117 goes to the list of lacunae with the label “H,” referring to homoioteleuton or homoiarcton, as do GA 713 and 2206 for similar reasons one verse later. The division of variation unit lengths also raises problems for the segmentation of the apparatus. For applying the CBGM and for entering the patristic and versional evidence we prefer to create shorter variation units rather than longer, ECM Acts 1.1, 25*. In our apparatus they received the label “H” (homoioteleuta/homoiarkta) or “A” (an abbreviation for the German Auslassung, referring to pointless and inexplicable omissions). 26 27

458

Annette Hüffmeier

following the general rule that variation units should be “as short as possible, as long as required.” But this leads to some conflicts. After having calmed the storm Jesus asks those with him in the boat whether their fear has to do with their lack of belief: τί δειλοί ἐστε; οὔπω ἔχετε πίστιν; (Mark 4:40/8–18). Following regularisation of readings, everything seems to be clear with words 8 (τι), 10 (δειλοι) and 16 (εχετε). Word 12 (εστε) is also fairly straightforward because it is only omitted once (in GA 28). And of course the copula can easily be omitted. The addition of the pejorative ολιγοπιστοι in word 13 by three manuscripts (GA 579 766 1424), whose scribes apparently could not await the final question of the same verse, is of minor importance. But the devil is in the last two words. There is only one alternative reading to πιστιν (word 18), πιστον. This reading is found only in GA 2174*, and even there it is corrected by another hand to πιστιν (GA 2174C). Do we judge the initial reading to be a mistake, and therefore regularise it as πιστιν? Or is the reading instead a legitimate alternative if one understands it as, “do you not have a faithful (person)?” or “do you not have (something) reliable?” We opted for the latter here. And we have not yet mentioned word 14: ουπω with its variants πως ουκ, ουτω(ς) πως ουκ, ουτω(ς) ουκ and ουτω(ς). But now we discover that the graphically very similar word ουτω(ς)28 cannot only be a variant reading of word 14 but may be an additional reading in word 9 or 11 as well. Thus we have to assess the boundaries of the variation unit anew, even though the data allows for recording all readings word by word, while maintaining small variation units: Mark 4:40/9 a om. b ουτως

4:40/11 a om. b ουτως

Mark 4:40/14 a ουπω b πως ουκ c ουκ d ουτως

4:40/18 a πιστιν b πιστον

4:40/12 a εστε b om.

4:40/13 a om. b ουτως c ολιγοπιστοι d ολιγοπιστοι ουτως

But this presentation of the material would deprive us of the possibility to show probable connections between variants that I have emphasised with the use of bold and underline typeface. A longer variation unit avoids mixing up alternative readings and additions and leaves open the exact position of the omitted εστε in a variant reading which clearly adds ουτως instead of using it as alternative to ουπω. So we prefer the following form of the apparatus in this instance:

28

ΟΥΠΩ < ΟΥΤΩ/ΟΥΤΩΣ.

Apparatus Construction Mark 4:40/10–14 a δειλοι εστε ουπω b δειλοι εστε ουτως c

ουτως δειλοι εστε ουπω

d e f

δειλοι εστε ουτως πως ουκ δειλοι ουτως εστε πως ουκ ουτως δειλοι εστε πως ουκ

g

δειλοι εστε ουτως ουκ

h i j

δειλοι εστε ολιγοπιστοι ουτως πως ουκ δειλοι εστε ολιγοπιστοι πως ουκ δειλοι εστε ολιγοπιστοι ουπω29

k

ουτως δειλοι ουπω

459

4:40/18 a πιστιν b πιστον

Having given you some examples I hope I have been able to make dry food a bit more tasty so that you were able to glimpse a bit of Hoskier’s “sunrise”! The main intention of my article is to present the philological method and the technical aspects of constructing the critical apparatus for the ECM. I want to illustrate the complexity of creating the apparatus and our diligence during that process. Therefore, after a few introductory remarks about the goals of the ECM edition in general, I described the working process step by step: – the selection of Greek manuscripts on the basis of Text und Textwert der griechischen Handschriften des Neuen Testaments; – the transcription of the selected Greek manuscripts by two hands and their critical comparison by another (the results are to be seen in the NTVMR 2.0); – the electronic collation; – the examination – and where necessary – regularisation of each single reading among thousands, setting and ordering all variants, using readings from Mark 1–5 as examples; – feeding the newly arranged data into a number of databases for each tradition that supplies information to the apparatus; – the application of the CBGM in iterative phases (pre-genealogical, genealogical) verse by verse and arriving at editorial decisions; – and the production of the master copy. All in all the critical apparatus of the ECM is constructed to give as much information as possible that contributes to understanding the construction of our hypothesis about the initial text and the work’s (mainly Greek) textual history.

29 The very few manuscripts that add ολιγοπιστοι (word 13) could have an extra entry outside this greater variation unit. I would not think this is so, however, in the case of εστε (word 12). Although in all other witnesses it follows directly after δειλοι, three manuscripts (4 273 2206) preserve it one word later as δειλοι ουτως εστε.

Contributors Garrick V. Allen is Lecturer in New Testament Studies at Dublin City University and Research Associate of the Department of Ancient and Modern Languages and Cultures, University of Pretoria. J. K. Elliott is Emeritus Professor of New Testament Textual Criticism, University of Leeds. Gregory Peter Fewster is a PhD Candidate at the University of Toronto in the Department for the Study of Religion along with the Book History and Print Culture Collaborative Specialization. Peter J. Gurry is Assistant Professor of New Testament at Phoenix Seminary. Juan Hernández Jr. is Professor of Biblical Studies at Bethel University. H. A. G. Houghton is Professor in New Testament Textual Scholarship at the University of Birmingham and Director of the Institute for Textual Scholarship and Electronic Editing. Annette Hüffmeier is Research Fellow of the Institute for New Testament Textual Research and teacher of Ancient Greek at Westfälische Wilhelms-Universität, Münster. Dirk Jongkind is a Fellow of St Edmund’s College and Academic Vice Principal at Tyndale House, Cambridge. Martin Karrer is Professor New Testament and its World at the Kirchliche Hochschule Wuppertal/Bethel and Director of the Institute for Septuagint and Biblical Textual Research (New Testament). Jennifer Wright Knust is Professor of Religious Studies at Duke University. Jan Krans is Lecturer of New Testament Studies at Protestantse Theologische Universiteit and Vrije Universiteit Amsterdam.

462

Contributors

Thomas J. Kraus is a Lecturer in New Testament at the University of Zürich and research associate of the Department of Greek, Latin and Classical Studies at the University of the Free State, Bloemfontein, South Africa. Christina M. Kreinecker is Assistant Professor in New Testament Studies at the University of Salzburg. Curt Niccum is Professor in the College of Biblical Studies and Associate Director of the Center for the Study of Ancient Religious Texts at Abilene Christian University. D. C. Parker is Professor of Digital Philology at the University of Birmingham. Jacob W. Peterson is a PhD Candidate in New Testament language, literature, and theology at the University of Edinburgh and a Research Associate at the Center for the Study of New Testament Manuscripts. Stanley E. Porter is President, Dean, Professor of New Testament, and Roy A. Hope Chair in Christian Worldview at McMaster Divinity College, Hamilton, Ontario, Canada. Catherine Smith is Research Fellow and Technical Office at the Institute for Textual Scholarship and Electronic Editing, University of Birmingham. Jill Unkel is Curator of Western Collections at the Chester Beatty, Dublin. Klaus Wachtel is Research Fellow of the Institut für Neutestamentliche Textforschung, Westfälische Wilhelms-Universität, Münster. Tommy Wasserman is Professor of Biblical Studies at Ansgar Teologiske Høgskole. An-Ting Yi is a doctoral student at Vrije Universiteit Amsterdam.

Bibliography Abbas, Hyder. “‘We want quality and condition’: the Formation of Chester Beatty’s South Asian Manuscripts and Miniatures Collection.” In Original Intentions: Essays on Production, Reproduction, and Interpretation in the Arts of China. Edited by N. Pearce and J. Steuber. Gainesville: University Press of Florida, 2012. Abberley, Will. “Race and Species Essentialism in Nineteenth-Century Philology.” Critical Quarterly 36/1 (2012): 45–60. Abbott, Edwin A. Johannine Grammar. Diatessarica 6. London: A. and C. Black, 1906. Abbott, T. K. Evangeliorum versio Antehieronymiana ex codice Ussheriano. Dublin: Hodges and Figgis, 1884. Abraha, Tedros. La lettera ai Romani: Testo e commentari della versione Etiopica. Äthiopistische Forschungen 57. Wiesbaden: Harrassowitz, 2001. Abraha, Tedros. The Ethiopic Version of the Letter to the Hebrews. Studi e Testi 419. Vatican City: Biblioteca Apostolica Vaticana, 2004. Abraha, Tedros. The Ethiopic Versions of 1 and 2 Corinthians. Rome: n.p., 2014. Achelis, H. and N. Bonwetscheds. Hippolytus Werke. Exegetische und homiletische Schriften. Leipzig: Hinrichs’sche, 1897. Ahmed, Sarah. The Promise of Happiness. Durham: Duke University Press, 2010. Aland, Barbara. “Entstehung, Charakter und Herkunft des sog. westlichen Textes untersucht an der Apostelgeschichte.” ETL 62/1 (1986): 5–65. Aland, Barbara and Kurt Aland. Der Text des Neuen Testaments. Eine Einführung in die wissenschaftlichen Ausgaben sowie in Theorie und Praxis der modernen Textkritik. 2nd edition. Stuttgart: Deutsche Bibelgesellschaft, 1982. Aland, Kurt. “Zur Liste der griechischen neutestamentlichen Handschriften.” Theologische Literaturzeitung 8/9 (1953): 465–96. Aland, Kurt. “Zur Liste der griechischen Neutestamentlichen Handschriften V.” ZNW 45 (1954): 179–217. Aland, Kurt. “Zur Liste der Neutestamentlichen Handschriften VI.” ZNW 48 (1957): 141–91. Aland, Kurt. “Die griechischen Handschriften des Neuen Testaments, Ergänzungen zur «Kurzgefassten Liste» (Fortsetzungsliste VII).” Pages 1–53 in Materialien zur neutestamentlichen Handschriftenkunde 1. Edited by K. Aland. Berlin: De Gruyter, 1969. Aland, Kurt. “Novi Testamenti Graeci Editio Maior Critica.” NTS 16/2 (1970): 163–77. Aland, Kurt. “The Twentieth-Century Interlude in New Testament Textual Criticism.” Pages 1–14 in Text and Interpretation: Studies in the New Testament Presented to Matthew Black. Edited by E. Best and R. McL. Wilson. Cambridge: Cambridge University Press, 1979. Aland, Kurt. Neutestamentliche Entwürfe. Theologische Bücherei 63. Munich: Kaiser, 1979. Aland, Kurt. “Vorwort.” Pages V–VIII in Text und Textwert der Grieschischen Handschriften des Neuen Testaments, I. Die Katholischen Briefe, Band 1: Das Material. Edited by K. Aland. Berlin: De Gruyter, 1987. Aland, Kurt et al. Text und Textwert der Grieschischen Handschriften des Neuen Testaments. Berlin: De Gruyter, 1987–2017.

464

Bibliography

Aland, Kurt et al. Kurzgefasste Liste der griechischen Handschriften des Neuen Testaments. 2nd revised and enlarged edition. ANTF 1. Berlin: De Gruyter, 1994. Aland, Kurt and Barbara Aland. The Text of the New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism. Translated by E. F. Rhodes. 2nd edition. Grand Rapids: Eerdmans, 1989. Albin, C. A. Judasbrevet: Traditionen, Texten, Tolkningen. Stockholm: Natur och Kultur, 1962. Allen, Garrick V. “The Reception of Scripture and Exegetical Resources in the Scholia in Apocalypsin (GA 2351).” Pages 141–63 in Commentaries, Catenae and Biblical Traditions. Edited by H. A. G. Houghton. Piscataway: Gorgias, 2016. Allen, Garrick V. “‘There is no glory and no money in the work’: H. C. Hoskier and New Testament Textual Criticism,” TC: A Journal of Biblical Textual Criticism (2018): 1–19. Allen, Garrick V. “Paratexts and the Reception History of the Apocalypse.” JTS (2019). Allen, Graham. Intertextuality. 2nd edition. London: Routledge, 2011. Allen, P. S. et al., eds. Opvs epistolarvm Des. Erasmi Roterodami Erasmus. Oxford: Clarendon, 1910. American Editorial Committee of the IGNTP. “Proposed Publication of the Manuscript Evidence for the Text of the Greek New Testament.” The Bible Translator 1 (1950): 169–71. Amphoux, Christian-B. “Une édition ‘plurielle’ de Marc.” Pages 69–80 in The New Testament Text in Early Christianity: Proceedings of the Lille colloquium, July 2000. Edited by C.-B. Amphoux and J. K. Elliott. Lausanne: Zèbre, 2003. Andrews, Tara L. “The Third Way.” Variants 10 (2013): 61–76. Andrews, Tara L. and Joris van Zundert. “An Interactive Interface for Text Variant Graph Models.” Digital Humanities (2013). Anonymous. Literary Fragments of Ronald Wood Hoskier 1896–1917. Boston: McKenzie, 191?. Anonymous. “According to Mark IV.” Time (9 August 1954): 68–9. Anonymous. “Bible Index Computed.” The New York Times (17 February 1956): 27. Anonymous. “Bible Labor of Years is done in 400 Hours.” Life 42/7 (1957): 92. Arendt, Hannah. Elemente und Ursprünge totaler Herrschaft. Frankfurt am Main: Europäische Verlagsanstalt, 1955. Argyriou, A. Les exégèses grecques de l’Apocalypse à l’époque turque (1453–1821). Esquisse d’une histoire des courants idéologiques au sein du people grec asservi. Εταιρεια Μακεδονικων Σπουδων. Thessaloniki: Kronoz, 1982. Armoni, Charikleia, James M. S. Cowey and Dieter Hagedorn, eds. Die griechischen Ostraka der Heidelberger Papyrus-Sammlung. Veröffentlichungen aus der Heidelberger PapyrusSammlung 11. Heidelberg: Universitätsverlag Winter, 2005. Armstrong, Elizabeth. Robert Estienne, Royal Printer. An Historical Study of the Elder Stephanus. Cambridge: Cambridge University Press, 1954. Revised edition 1986. Arzt-Grabner, Peter. Philemon. Papyrologische Kommentare zum Neuen Testament 1. Göttingen: Vandenhoeck & Ruprecht, 2003. Arzt-Grabner, Peter. “Onesimus erro: Zur Vorgeschichte des Philemonbriefes.” ZNW 95 (2004): 131–43. Arzt-Grabner, Peter. “‘Neither a Truant nor a Fugitive’: Some Remarks on the Sale of Slaves in Roman Egypt and Other Provinces.” Pages 21–32 in Proceedings of the Twenty-Fifth International Congress of Papyrology, Ann Arbor, July 29–August 4, 2007. Edited by T. Gagos and A. Hyatt. Ann Arbor: Scholarly Publishing Office The University of Michigan Library, 2010.

Bibliography

465

Arzt-Grabner, Peter. “Neues zu Paulus aus den Papyri des römischen Alltags.” Early Christianity 1 (2010): 131–57. Arzt-Grabner, Peter. “Paul’s Letter Thanksgiving.” Pages 129–58 in Paul and the Ancient Letter Form. Edited by S. E. Porter and S. A. Adams. Leiden: Brill, 2010. Arzt-Grabner, Peter. “Neuigkeiten aus der Papyrologie.” Early Christianity 3 (2012): 111–15. Arzt-Grabner, Peter. 2. Korinther. Unter Mitarbeit von Ruth E. Kritzer. Papyrologische Kommentare zum Neuen Testament 4. Göttingen: Vandenhoeck & Ruprecht, 2014. Arzt-Grabner, Peter. “Neuigkeiten aus der Papyrologie (2).” Early Christianity 6 (2015): 561– 69. Arzt-Grabner, Peter and Kristin De Troyer. “Ancient Jewish and Christian Amulets and How Magical They Are.” Biblische Notizen 176 (2018): 5–46. Arzt-Grabner, Peter, Ruth E. Kritzer, Amphilochios Papathomas, and Franz Winter. 1. Korinther. Papyrologische Kommentare zum Neuen Testament 2. Göttingen: Vandenhoeck & Ruprecht, 2006. Askeland, Christian. “The Coptic Versions of the New Testament.” Pages 201–29 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Askeland, Christian. “The Sahidic Apocalypse in Early Islamic Egypt.” Pages 271–87 in Studien zum Text der Apokalypse. ANTF 47. Edited by M. Sigismund, M. Karrer, and U. Schmid. Berlin: De Gruyter, 2015. Askeland, Christian. “An Eclectic Edition of the Sahidic Apocalypse of John.” Pages 33–79 in Studien zum Text der Apokalypse II. ANTF 50. Edited by M. Sigismund and D. Müller. Berlin: De Gruyter, 2017. Aune, David E. Revelation. WBC. 3 volumes. Dallas: Word, 1997–1998. Austin, Linda. “The Nostalgic Moment and the Sense of History.” Postmedieval: A Journal of Medieval Cultural Studies 2/2 (2011): 127–40. Bagnall, Roger S. and Bruce W. Frier. The Demography of Roman Egypt. Cambridge: Cambridge University Press, 1994. Bagnall, Roger S, and Raffaella Cribiore. Women’s Letters from Ancient Egypt, 300 BC–AD 800. With contributions by Evie Ahtaridis. Ann Arbor: University of Michigan Press, 2006. Bagnall, Roger S. “Review of Malcolm Choat, Belief and Cult in Fourth-Century Papyri.” BASP 43 (2006): 205–209. Bagnall, Roger S. Early Christian Books in Egypt. Princeton: Princeton University Press, 2009. Bagnall, Roger S. “The Amicitia Papyrologorum in a Globalized World of Learning.” Pages 1–5 in Actes du 26e Congrès international de papyrologie: Genève, 16–21 août 2010. Edited by P. Schubert. Genève: Droz, 2012. Ball, Alex. Preservation and Curation in Institutional Repositories version 1.3. Edinburgh: Digital Curation Centre, 2010. Bardy, G. Eusèbe de Césarée. Histoire Ecclésiastique. Volume 2. SC 41. Paris: Éditions du Cerf, 1955. Barnes, Timothy David. Constantine and Eusebius. Cambridge, MA: Harvard University Press, 1981. Barthes, Roland. Essais critiques. Paris: Seuil, 1964. Barthes, Roland. The Pleasure of the Text. Translated by R. Miller. New York: Hill and Wang, 1975. Barthes, Roland. Image – Music – Text. Translated by S. Heath. New York: Hill and Wang, 1977. Barthes, Roland. The Rustle of Language. Translated by R. Howard. New York: Hill and Wang, 1986.

466

Bibliography

Bauer, Walter. Griechisch-deutsches Wörterbuch zu den Schriften des Neuen Testaments und der frühchristlichen Literatur. Edited by K. Aland and B. Aland. 6th edition. Berlin: De Gruyter, 1988. Beard, Mary. “Ciceronian Correspondences: Making a Book Out of Letters.” Pages 103–44 in Classics in Progress: Essays on Ancient Greece and Rome. Edited by T. P. Wiseman. Oxford: Oxford University Press, 2002. Bell, Lonnie. The Early Textual Transmission of John: Stability and Fluidity in its Second and Third Century Greek Manuscripts. NTTSD 54. Leiden: Brill, 2018. Bengel, J. A. Apparatus Criticus ad Novum Testamentum. 2nd edition. Tübingen: J. G. Cotta, 1763 (1734). Benjamin, Walter. “L’ouvre d’art à l’époque de sa reproduction mécanisée.” Zeitschrift für Sozialforschung 5/1 (1936): 40–68. Bentley, Richard. Proposals for Printing a new edition of the Greek Testament and St. Hierom’s Latin Version, with a full answer to all the remarks of a late pamphleteer, By a Member of Trinity College in Cambridge (1721). In The Works of Richard Bentley, D.D. Edited by A. Dyce. London: Francis Macpherson, 1838. Berlant, Lauren. Cruel Optimism. Durham: Duke University Press, 2011. Betz, Hans Dieter. “Zum Problem der Auferstehung Jesu im Lichte der griechischen magischen Papyri.” Pages 230–61 in Hellenismus und Urchristentum: Gesammelte Aufsätze. Volume 1. Edited by H. D. Betz. Tübingen: Mohr Siebeck, 1990. Beza, Theodorus. Novum D. N. Iesu Christi testamentum. Geneva: Robertus Stephanus, 1556. Beza, Theodoruss. Novum D. N. Iesu Christi testamentum, sive Novum Foedus. 2nd edition. Geneva: Henricus Stephanus, 1565. Bianchini, Giuseppe. Evangeliarium quadruplex latinae versionis antiquae seu veteris italicae. Rome: de Rubeis, 1749. Birdsall, J. N. “Irenaeus and the Number of the Beast: Revelation 13,18.” Pages 349–59 in New Testament Textual Criticism and Exegesis. BETL 161. Edited by A. Denaux. Leuven: Leuven University Press, 2002. Blake-Hill, Philip V. “The Macmillan Archive.” The British Museum Quarterly 36 (1972): 74– 80. Blass, Friedrich and Albert Debrunner. Grammatik des neutestamentlichen Griechisch. Edited by F. Rehkopf. 18th edition. Göttingen: Vandenhoeck & Ruprecht, 2001. Blomberg, Craig L. and Jennifer Foutz Markley. A Handbook of New Testament Exegesis. Grand Rapids: Baker, 2010. Blomkvist, Vemund. Euthalian Traditions: Text, Translation and Commentary. Berlin: De Gruyter, 2012. Bloom, Harold et al. Deconstruction and Criticism. New York: Continuum, 1979. Blumell, Lincoln H. Lettered Christians: Christians, Letters, and Late Antique Oxyrhynchus. NTTSD 39. Leiden: Brill, 2012. Boismard, M.-É. and A. Lamouille. Le Texte occidental des Actes des Apôtres: Reconstitution et Réhabilitation. Synthèse 17. Paris: Éditions Recherche sur les Civilisations, 1984. Boismard, M.-É. Le Texte occidental des Actes des Apôtres. Revised edition. Études Bibliques 40. Paris: Gabalda, 2000. Böttrich, Christfried. “Codex Sinaiticus and the Use of Manuscripts in the Early Church.” ExpTim 128/10 (2017): 469–78. Bordalejo, Barbara. “The Genealogy of Texts: Manuscript Traditions and Textual Traditions.” Digital Scholarship in the Humanities 31 (2016): 563–77. Bourdieu, Pierre. Homo Academicus. Translated by Peter Collier. Stanford: Stanford University Press, 1988.

Bibliography

467

Bourdieu, Pierre. The Field of Cultural Production: Essays on Art and Literature. Edited and translated by R. Johnson. New York: Columbia University Press, 1993. Bousset, Wilhelm. Textkritische Studien zum Neuen Testament. TUGAL 9/4. Leipzig: Hinrichs’sche, 1894. Bousset, Wilhelm. Die Offenbarung Johannis. KEK 16. 6 th edition. Göttingen: Vandenhoeck & Ruprecht, 1906. Bousset, Wilhelm. “Textkritik II.” Theologische Rundschau 17 (1914): 187–206. Bowers, Fredson. Textual and Literary Criticism. Cambridge: Cambridge University Press, 1959. Bowles, Edmund A. “Preface.” Pages v–viii in Computers in Humanistic Research, Readings and Perspectives. Edited by E. A. Bowles. Englewood Cliffs: Prentice-Hall, 1967. Bowman, J. H. “Macmillan Greek.” Journal of the Printing Historical Society 19/20 (1990): 103–24. Bowman, J. H. Greek Printing Types in Britain in the Nineteenth Century: A Catalogue. Oxford: Oxford Bibliographical Society, 1992. Boym, Svetlana. The Future of Nostalgia. New York: Basic Books, 2001. Brightman, F. E. “On the Italian Origin of Codex Bezae, II. The Marginal Notes of Lections.” JTS 1 (1900): 446–54. Brock, Sebastian P. “The Syriac Euthalian Material and the Philoxenian Version of the NT.” ZNW 70/1–2 (1979): 120–30. Brooks Jr., Frederick P. The Mythical Man-Month: Essays on Software Engineering. Anniversary edition. Boston: Addison-Wesley, 1975. Repr. with four new chapters. Boston: Addison-Wesley, 1995. Brown, A. J. ed. Novum Testamentum ab Erasmo recognitum. Opera Omnia Desiderii Erasmi Roterodami 6. Revelation: part IV. Leiden: Brill, 2013. Bruce, F. F. “Westcott’s Significance.” Pages x–xi in The Epistles of St. John: The Greek Text, with Notes and Addenda. Grand Rapids: Eerdmans, 1966. Buchanan, E. S. The Four Gospels from Codex Veronensis (b), being the First Complete Edition of the Evangeliarum Purpureum in the Cathedral Library at Verona, with two facsimiles. Oxford: Clarendon, 1911. Bucking, Scott. “Christian Educational Texts from Egypt: A Preliminary Inventory.” Pages 132–38 in Akten des 21. Internationalen Papyrologenkongresses. Edited by B. Krämer, W. Luppe, and H. Maehler. Leipzig: Teubner, 1997. Burgon, John W. Letter to the Guardian, 26 July 1882, inserted as a frontispiece to the microfilm copy of GA 700. British Museum Egerton 2610. Burgon, John W. The Revision Revised: Three Articles Reprinted from the Quarterly Review. London: John Murray, 1883. Burkitt, F. C. “Additional Note.” JTS 12 (1911): 457–59. Burkitt, F. C. “Euangelium Gatianum.” JTS 11 (1910): 607–11. Burkitt, F. C. “A Gothic-Latin Fragment from Antinoe.” JTS 11 (1910): 611–13. Burnard, Lou D. “At Home with the Hardware.” The Times Literary Supplement 4024 (1980): 533. Burton, P. H., H. A. G. Houghton, R. F. MacLachlan ,and D. C. Parker, eds. Vetus Latina. Die Reste der altlateinischen Bibel. 19. Evangelium Secundum Iohannem. Freiburg-im-Breisgau: Herder, 2011–. Burton, P. H. The Old Latin Gospels. Oxford: Oxford University Press, 2000. Burton, P. H. “The Latin Version of the New Testament.” Pages 167–200 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013.

468

Bibliography

Busa, Roberto. “The Use of Punched Cards in Linguistic Analysis.” Pages 357–73 in Punched Cards: Their Applications to Science and Industry. Edited by R. S. Casey et al. 2nd edition. New York: Reinhold Publishing Corporation, 1958. Busa, Roberto. “The Annals of Humanities Computing: The Index Thomisticus.” Computers and the Humanities 14/2 (1980): 83–90. Busa, Roberto. “Concluding a life’s safari from punched cards to World Wide Web.” Pages 3– 11 in The Digital Demotic: A Selection of Papers from Digital Resources in the Humanities 1997. Edited by L. Burnard, M. Deegan, and H. Short. London: Office for Humanities Communication. 1998. Cadwallader, A. H. “The Politics of Translation of The Revised Version: Evidence from the Newly Discovered Notebooks of Brooke Foss Westcott.” JTS 58/2 (2007): 415–39. Cadwallader, A. H. The Politics of the Revised Version. Scriptural Traces: Critical Perspectives on the Reception and Influence of the Bible. New York: Bloomsbury, 2018. Cady, Linelle E. and Tracy Fessenden. “Gendering the Divide: Religion, the Secular, and the Politics of Sexual Difference.” Pages 3–24 in Religion, The Secular, and the Politics of Sexual Difference. Edited by L. E. Cady and T. Fessenden. New York: Columbia University Press, 2013. Canfora, Luciano. Il copista come autore. Palermo: Sellerio, 2002. Caragounis, Chrys C. The Development of Greek and the New Testament: Morphology, Syntax, Phonology, and Textual Transmission. WUNT 167. Tübingen: Mohr Siebeck, 2004. Carlson, Stephen C. The Text of Galatians and Its History. WUNT 2/385. Tübingen: Mohr Siebeck, 2015. Caroli, Menico. Il titolo iniziale nel rotolo librario greco-egizio: con un catalogo delle testimonianze iconografiche greche e di area vesuviana. Bari: Levante, 2007. Causer, Tim and Melissa Terras. “Crowdsourcing Bentham: Beyond the Traditional Boundaries of Academic History.” International Journal of Humanities and Arts Computing 8/1 (2014): 46–64. Causer, Tim, Justin Tonra, and Valerie Wallace. “Transcription Maximized; Expense Minimized? Crowdsourcing and Editing The Collected Works of Jeremy Bentham.” Literary and Linguistic Computing 27/2 (2012): 119–37. Causer, Tim and Valerie Wallace. “Building a Volunteer Community: Results and Findings from Transcribe Bentham.” Digital Humanities Quarterly 6/2 (2012). Caygill, Howard. Walter Benjamin: The Colour of Experience. London: Routledge, 1998. Cerulli, Enrico. I manoscritti etiopici della Chester Beatty Library in Dublino. Roma: Accademia Nazionale dei Licei, 1965. Chadwick, Owen. The Victorian Church, Part II: 1860–1901. London: SCM, 1972. Chakrabarty, Dipesh. Provincializing Europe: Postcolonial Thought and Historical Difference. Princeton: Princeton University Press, 2000. Charles, R. H. A Critical and Exegetical Commentary on the Revelation of St. John. ICC. 2 volumes. New York: Scribner’s Sons, 1920. Childers, Jeff W. “The Georgian Version of the New Testament.” Pages 293–327 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Church, C. L. “Westcott, B(rooke) F(oss) (1825–1901), and F(enton) J(ohn) A(nthony) Hort (1828–1892).” Pages 1038–43 in Dictionary of Major Biblical Interpreters. Edited by D. McKim. Downers Grove: InterVarsity, 2007. Ciotti, Giovanni and Hang Lin, eds. Tracing Manuscripts in Time and Space through Paratexts. Studies in Manuscript Cultures 7. Berlin: De Gruyter, 2016.

Bibliography

469

Clark, Kenneth W. Eight American Praxapostoloi. Chicago: University of Chicago Press, 1941. Clark, Kenneth W. “The Effect of Recent Textual Criticism upon New Testament Studies.” Pages 27–51 in The Background of the New Testament and Its Eschatology. Edited by W. D. Davies and D. Daube. Cambridge: Cambridge University Press, 1954. Cleaver, Laura. “The Western Manuscript Collection of Alfred Chester Beatty (ca. 1915– 1930).” Manuscript Studies 2/2 (2017): 445–82. Cleaver, Laura. “The Present.” Pages 104–109 in Miniature Masterpiece: The Coëtivy Hours. Edited by J. Unkel. Dublin: Chester Beatty Library, 2018. Clough, Cecil H. “The Albani Library and Pope Clement XI.” Librarium: Zeitschrift der Schweizerischen Bibliophilen Gesellschaft / Revue de la Sociéte Suisse des Bibliophiles 12/1 (1969): 13–21. Cole, Zachary J. Numerals in Early Greek New Testament Manuscripts: Text-Critical, Scribal, and Theological Studies. NTTSD 53. Leiden: Brill, 2017. Colwell, E. C. and Ernest W. Tune. “Method in Classifying and Evaluating Variant Readings.” Pages 96–105 in Studies in Methodology in Textual Criticism of the New Testament. NTTS 9. Leiden: Brill, 1969. Colwell E. C. and Ernest W. Tune. “The Quantitative Relationships Between MS Text-Types.” Pages 25–32 in Biblical and Patristic Studies in Memory of Robert Pierce Casey. Edited by J. N. Birdsall and R. W. Thomson. Freiburg im Breisgau: Herder, 1963. Repr. in Studies in Methodology in Textual Criticism of the New Testament, 56–62. Colwell, Ernest C. “Genealogical Method: Its Achievements and Its Limitations.” JBL 66 (1947): 114–16. Colwell, Ernest C. “Method in Locating a Newly-Discovered Manuscript within the Manuscript Tradition of the Greek New Testament.” Pages 757–77 in Studia Evangelica. Volume 1. Edited by K. Aland et al. Berlin: Akademie-Verlag, 1959. Repr. in Studies in Methodology in Textual Criticism of the New Testament, 26–44. Colwell, E. C. Studies in Methodology in Textual Criticism of the New Testament. NTTS 9. Leiden: Brill, 1969. Constantinidou, Natasha. “Printers of the Greek Classics and Market Distribution in the Sixteenth Century: The Case of France and the Low Countries.” Pages 275–93 in Specialist Markets in the Early Modern Book World. Edited by R. Kirwan and S. Mullins. Leiden: Brill, 2015. Conybeare, F. C. “On the Codex Pamphili and the Date of Euthalius.” Journal of Philology 23/46 (1895): 241–59. Conybeare, F. C. “The Date of Euthalius.” ZNW 5/1 (1904): 39–52. Coogan, Jeremiah. “Mapping the Fourfold Gospel: Textual Geography in the Eusebian Apparatus.” JECS 25/3 (2017): 337–57. Coogan, Jeremiah. “Rewriting the Temple Incident with Eusebius of Caesarea.” JBL (forthcoming). Coogan, R. Erasmus, Lee and the Correction of the Vulgate: The Shaking of the Foundations. Genf: Droz, 1992. Cook, Clair M. “Automation Comes to the Bible.” Computers and Automation (1958). Reprint of the same article in Christian Century 54.30 (1957). Cook, F. C. The Revised Version of the First Three Gospels. London: John Murray, 1882. Corduan, W. Handmaid to Theology: An Essay in Philosophical Prolegomena. Grand Rapids: Baker, 1981. Reprint Eugene: Wipf & Stock, 2009. Coroleu, A. “On the Reception of Erasmus’s Latin Version of the New Testament in SixteenthCentury Spain.” The Bible Translator 67 (2016): 56–68.

470

Bibliography

Coron, Antoine. “The First Libraries: Blois, Fontainebleau, Paris.” Pages 151–66 in Creating French Culture. Treasures from the Bibliothèque nationale de France. Edited by M.-H. Tesnière and P. Gifford. New Haven: Yale University Press, 1995. Corsten, Thomas, Markus Öhler, and Joseph Verheyden, eds. Epigraphik und Neues Testament. WUNT 365. Tübingen: Mohr Siebeck, 2016. Couchoud, Paul-Louis. “L’Évangile de Marc a-t-il été écrit en Latin?” RHR 94 (1926): 161– 92. Cowe, S. Peter. “The Armenian Version of the New Testament.” Pages 253–92 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Crawford, Matthew R. “Ammonius of Alexandria, Eusebius of Caesarea and the Origins of Gospel Scholarship.” NTS 61/1 (2015): 1–29. Croke, Fionnuala, ed. Chester Beatty’s A–Z: From Amulet to Zodiac. Dublin: Chester Beatty Library, 2014. Cummings, James. “The Text Encoding Initiative and the Study of Literature.” Pages 451–76 in Companion to Digital Literary Studies. Edited by S. Schreibman and R. Siemens. Oxford: Blackwell, 2008. Cutler, Anthony. “From Loot to Scholarship: Changing Modes in the Italian Response to Byzantine Artifacts, ca. 1200-1750.” Dumbarton Oaks Papers 49 (1995): 237–67. Cuvigny, Hélène. “The Finds of Papyri: The Archaeology of Papyrology.” Translated by A. Bülow-Jacobsen. Pages 59–78 in The Oxford Handbook of Papyrology. Edited by R. S. Bagnall. Oxford: Oxford University Press, 2009. Dahl, Nils A. “The ‘Euthalian Apparatus’ and Affiliated Argumenta.” Pages 231–75 in Studies in Ephesians: Introductory Questions, Text- & Edition-Critical Issues, Interpretation of Texts and Themes. Edited by D. Hellholm, V. Blomkvist, and T. Fornberg. Tübingen: Mohr Siebeck, 2000. d’Alveydre, Saint-Yves. Les Mission des Juifs. Paris: Calmann Lévy, 1884. Davies, Rachel Bryant. Troy, Carthage, and the Victorians: The Drama of Classical Ruins in the Nineteenth Century Imagination. Cambridge: Cambridge University Press, 2018. de Bruyn, Theodore. “Appeals to Jesus and the One ‘Who Heals Every Illness and Every Infirmity’ (Matt 4:23, 9:35) in Amulets in Late Antiquity.” Pages 65–82 in The Reception and Interpretation of the Bible in Late Antiquity: Proceedings of the Montréal Colloquium in Honour of Charles Kannengiesser, 11–13 October 2006. Edited by L. DiTommaso and L. Turcescu. Leiden: Brill, 2008. de Bruyn, Theodore. “Papyri, Parchments, Ostraca, and Tablets Written with Biblical Texts in Greek and Used as Amulets: A Preliminary List.” Pages 145–90 in Early Christian Manuscripts: Examples of Applied Method and Approach. TENTS 5. Edited by T. J. Kraus and T. Nicklas. Leiden: Brill, 2010. de Bruyn, Theodore. Making Amulets Christian: Artefacts, Scribes, and Contexts. Oxford: Oxford University Press, 2017. de Bruyn, Theodore and Jitse H. F. Dijkstra. “Greek Amulets and Formularies from Egypt Containing Christian Elements.” BASP 48 (2011): 163–216. Deegan, Marilyn. “Collection and Preservation of an Electronic Edition.” Pages 358–70 in Electronic Textual Editing. Edited by L. Burnard, K. O’Brien O’Keeffe, and J. Unsworth. New York: Modern Language Association of America, 2006. de Guldenstubbé, Louis. La réalité des esprits et le phénomène merveilleux de leur écriture directe demontrées par…Paris: Franck, 1857. de Groote, Marc, ed. Oecumenii Commentarius in Apocalypsin. Leuven: Peeters, 1999. de Grotte, Marc, ed. Index Oecumenianus. Gent: Olms-Weidemann, 2001.

Bibliography

471

de Hamel, Christopher. Meetings with Remarkable Manuscripts. London: Allen Lane, 2016. Deissmann, Adolf. Bible Studies: Contributions Chiefly from Papyri and Inscriptions to the History of the Language, the Literature, and the Religion of Hellenistic Judaism and Primitive Christianity. Translated by A. Grieve. 2nd edition. Edinburgh: T&T Clark, 1909. Deissmann, Adolf. Licht vom Osten. Das Neue Testament und die neuentdeckten Texte der hellenistisch-römischen Welt. 4th edition. Tübingen: J. C. B. Mohr, 1923. Deissmann, Adolf. Light from the Ancient East: The New Testament Illustrated by Recently Discovered Texts of the Graeco-Roman World. New and completely revised edition with eighty-five illustrations from the latest German edition. Translated by L. R. M. Strachan. London: Hodder and Stoughton, 1927 (repr. Peabody: Hendrickson, 1995). de Jonge, H. J. “Jeremias Hoelzlin: Editor of the ‘Textus Receptus’ Printed by the Elzeviers Leiden 1633.” Pages 105–28 in Miscellanea Neotestamentica. Studia ad Novum Testamentum Praesertim Pertinentia a Sociis Sodalicii Batavi c.n. Studiosorum Novi Testamenti Conventus Anno MCMLXXVI Quintum Lustrum Feliciter Complentis Suscepta. NTSup 47. Edited by T. Baarda et al. Leiden: Brill, 1978. de Jonge, H. J. “Novum Testamentum a nobis versum: The Essence of Erasmus’ Edition of the New Testament.” JTS 35 (1984): 394–413. de Jonge, H. J. “Erasmus’ Translation of the New Testament, Aim and Method.” The Bible Translator 67 (2016): 29–41. Delitzsch, Franz. Handschriftliche Funde 1. Die Erasmischen Entstellungen des Textes der Apokalypse. Nachgewiesen aus dem verloren geglaubten Codex Reuchlin. Leipzig: Dörffling & Franke, 1861. der Nersessian, Shiarpie. The Chester Beatty Library: A Catalogue of the Armenian Manuscripts, with an Introduction on the History of Armenian Art. 2 volumes. Dublin: Hodges and Figgis, 1958. Derrida, Jacques. “Structure, Sign, and Play in the Discourse of the Human Sciences.” Pages 247–65 in The Structuralist Controversy: The Languages of Criticism and the Sciences of Man. Edited by R. Macksey and E. Donato. Baltimore: Johns Hopkins University Press, 1972. Derrida, Jacques. Of Grammatology. Translated by G. C. Spivak. Baltimore: Johns Hopkins University Press, 1976. Derrida, Jacques. Writing and Difference. Translated by A. Bass. Chicago: University of Chicago Press, 1978. Derrida, Jacques. Positions. Translated by A. Bass. Chicago: University of Chicago Press, 1981. Deveney, John Patrick. “Spiritualism.” Pages 1074b–82a in Dictionary of Gnosis and Western Esotericism. Edited by W. J. Hanegraaff. Leiden: Brill, 2006. Diekamp, Franz. “Mitteilungen über den neu aufgefundenen Kommentar des Oekumenius zur Apokalypse.” Sitzungsbericht der Kgl. Preuß. Akademie der Wissenschaft 13–14 (1901): 1046–1056. Diekamp, Franz. “Neue über die Handschriften des Oekumenius-Kommentares zur Apokalypse.” Biblica 10/1 (1929): 81–4. Dill, U. and P. Schierl, eds. Das bessere Bild Christi, Das Neue Testament in der Ausgabe des Erasmus von Rotterdam. Basel: Schwabe, 2016. Dill, U. “Kontroversen: Erasmus verteidigt seine Ausgabe.” Pages 167–79 in Das bessere Bild Christi. Das Neue Testament in der Ausgabe des Erasmus von Rotterdam. Edited by U. Dill and P. Shierl. Basel: Schwabe, 2016. Dinshaw, Carolyn. How Soon is Now? Medieval Texts, Amateur Readers, and the Queerness of Time. Durham: Duke University Press, 2012.

472

Bibliography

Diobouniotis, C. and A. Harnack. Der Scholien-Kommentar des Origenes zu Apokalypse Johannis. TU 3/8. Leipzig: Hinrichs’sche, 1911. Dionisotti, Anna Carlotta. “From Stephanus to Du Cange: Glossary Stories.” RHT 14 (1984): 303–36. d’Olivet, Fabre. La Langue hébraïque restituée et les veritable sens des mots hébreux rétabil et prouvé par leur analyse radicale. Paris: Eberhart, 1815. d’Olivet, Fabre. Histoire philosophique du genre humain, ou L’homme considéré sous ses rapports religieux et politiques dan l’état social, à toutes les époques et chez les différents peuples de la terre, précédée d’une dissertation introductive sur les motifs et l’objet de cet ouvrage. 2 volumes. Paris: Hubert, 1824. Dorfbauer, Lukas, ed. Fortunatianus Aquileiensis. Commentarii in evangelia. CSEL 103. Berlin: De Gruyter, 2017. Dorival, Gilles. “Un groupe judéo-chrétien méconnu: les Hébreux.” Apocrypha 11 (2001): 7– 36. Driscoll, Matthew J. “The Words on the Page: Thoughts on Philology, Old and New.” Pages 87–104 in Creating the Medieval Saga: Versions, Variability and Editorial Interpretations of Old Norse Saga Literature. Edited by J. Quinn and E. Lethbridge. Odense: University Press of Southern Denmark, 2010. Drum, Walter. “Ecclesiastical Library Table: Text.” The Ecclesiastical Review 47 (1912): 739– 48. Duchesne, L. “Fragments des Epîtres de S. Paul.” Archives des Missions scientifiques et littéraires 3/3 (1876): 420–29. Dumville, D. N. A Palaeographer’s Review: the Insular System of Scripts in the Middle Ages I. Osaka: Kansai University Press, 1999. Dungan, David L. A History of the Synoptic Problem: The Canon, the Text, the Composition, and the Interpretation of the Gospels. New York: Doubleday, 1999. Duxfield, Polly. “Transcribing the Estoria de Espanna using Crowdsourcing: Strategies and Aspirations.” Magnificat Cultura i Literatura Medievals 2 (2015): 129–48. Ebojo, Edgar B. “A Scribe and His Manuscript: an Investigation into the Scribal Habits of Papyrus 46 (P. Chester Beatty II––P. Mich. Inv. 6238).” PhD diss., University of Birmingham, 2014. Edmondson, Andrew C. “Discovering New Family Trees of Biblical Manuscripts.” Poster presented at the University of Birmingham Research Poster Conference. 16 June 2015. Edwards, David L. Leaders of the Church of England: 1828–1944. Oxford: Oxford University Press, 1971. Edwards, Mark. “Ammonius, Teacher of Origen.” The Journal of Ecclesiastical History 44/2 (1993): 169–81. Ehrhard, Albert. “Der Codex H ad Epistulas Pauli und ‘Euthalios Diaconos’: Eine palaeographisch-patrologische Untersuchung.” Zentralblatt für Bibliothekswesen 8 (1891): 385–411. Ehrman, Bart D. “Methodological Developments in the Analysis and Classification of New Testament Documentary Evidence.” NovT 29/1 (1987): 22–45. Ehrman, Bart D. “The Use of Group Profiles for the Classification of New Testament Documentary Evidence.” JBL 106/3 (1987): 465–86. Ehrman, Bart D. “A Problem of Textual Circularity: The Alands on the Classification of New Testament Manuscripts.” Biblica 70 (1989): 377–88. Ehrman, Bart D. The Orthodox Corruption of Scripture. Oxford: Oxford University Press, 1993. Ehrman, Bart D. “Heracleon and the ‘Western’ Textual Tradition.” NTS 40/2 (1994): 161–79.

Bibliography

473

Ehrman, Bart D. “The Text as Window: New Testament Manuscripts and the Social History of Early Christianity.” Pages 361–79 in The Text of the New Testament in Contemporary Research. Essays on the Status Quaestionis. Edited B. D. Ehrman and M. W. Holmes. Grand Rapids: Eerdmans, 1995. Ehrman, Bart D. Studies in the Textual Criticism of the New Testament. NTTSD 42. Leiden: Brill, 2006. Ehrman, Bart D. and Michael W. Holmes, eds. The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. NTTSD 42. Leiden: Brill, 2013. El-Haj, Nadia Abu. The Genealogical Science: The Search for Jewish Origins and the Politics of Epistemology. Chicago: University of Chicago Press, 2012. Ellicott, C. J. and E. Palmer. The Revisers and the Greek Text of the New Testament by Two Members of the New Testament Committee. London: Macmillan, 1882. Elliott, J. K. “Manuscripts of the Book of Revelation Collated by H.C. Hoskier.” JTS 40 (1989): 100–11. Elliott, J. K. “The Distinctiveness of Greek Manuscripts of the Book of Revelation.” JTS 48 (1997): 116–24. Elliott, J. K. “Manuscripts Cited by Stephanus.” NTS 55 (2009): 390–95. Elliott, J. K. “A Short Textual Commentary on the Book of Revelation and the New Nestle.” NovT 56 (2014): 68–100. Elliott, J. K. “Recent Work on the Greek Manuscripts of Revelation and the Consequences for the Kurzgefasste Liste.” JTS 66 (2015): 574–84. Elliott, J. K. “‘Novum Testamentum editum est’: The Five-Hundredth Anniversary of Erasmus’s New Testament.” The Bible Translator 67 (2016): 9–28. Ellis, John M. Against Deconstruction. Princeton: Princeton University Press, 1989. Ellison, John W. “The Use of Electronic Computers in the Study of the Greek New Testament Text.” PhD diss., Harvard Divinity School, 1957. Ellison, John W. Nelson’s Complete Concordance of the Revised Standard Version Bible. New York: Thomas Nelson, 1957. Ellison, John W. “Computers and the Testaments.” Pages 160–69 in Computers in Humanistic Research, Readings and Perspectives. Edited by E. A. Bowles. Englewood Cliffs: PrenticeHall, 1967. Epp, Eldon J. The Theological Tendency of Codex Bezae Cantabrigiensis in Acts. Cambridge: Cambridge University Press, 1966. Epp, Eldon J. “The Twentieth Century Interlude in New Testament Textual Criticism.” JBL 93/3 (1974): 386–414. Epp, Eldon J. “New Testament Textual Criticism in America: Requiem for a Discipline.” JBL 98 (1979): 94–8. Epp, Eldon J. “A Continuing Interlude in New Testament Textual Criticism.” HTR 73 (1980): 131–51. Epp, Eldon J. “Textual Criticism.” Pages 75–126 in The New Testament and Its Modern Interpreters. Edited by E. J. Epp and G. W. MacRae. Atlanta: Scholars, 1989. Epp, Eldon J. “Decision Points in Past, Present, and Future New Testament Textual Criticism.” Pages 17–44 in Studies in the Theory and Method of New Testament Textual Criticism. Edited by E. J. Epp and G. D. Fee. Grand Rapids: Eerdmans, 1993. Epp, Eldon J. “Toward the Clarification of the Term ‘Textual Variant.” Pages 47–61 in Studies in the Theory and Method of New Testament Textual Criticism. Edited by E. J. Epp and G. D. Fee. Grand Rapids: Eerdmans, 1993.

474

Bibliography

Epp, Eldon J. “The Claremont Profile Method for Grouping New Testament Minuscule Manuscripts.” Pages 211–20 in Studies in the Theory and Method of New Testament Textual Criticism. Edited by E. J. Epp and G. D. Fee. Grand Rapids: Eerdmans, 1993. Epp, Eldon J. “The Multivalence of the Term ‘Original Text’ in New Testament Textual Criticism.” HTR 92/3 (1999): 245–81. Epp, Eldon J. “Anti-Judaic Tendencies in the D-Text of Acts: Forty Years of Conversation.” Pages 111–46 in The Book of Acts as Church History: Text, Textual Traditions and Ancient Interpretations / Apostelgeschichte als Kirchengeschichte: Text, Texttraditionen und Antike Auslegungen. BZNW 120. Edited by T. Nicklas and M. Tilly. Berlin: De Gruyter, 2003. Epp, Eldon J. Perspectives on New Testament Textual Criticism: Collected Essays, 1962-2004. NTSup 116. Leiden: Brill, 2005. Epp, Eldon J. “Textual Clusters: Their Past and Future in New Testament Textual Criticism.” Pages 519–77 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. NTTSD 42. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Epp, Eldon J. “Critical Editions and the Development of Text-critical Methods, Part 2: From Lachmann (1831) to the Present.” Pages 13–48 in The New Cambridge History of the Bible: From 1750 to the Present. Edited by J. Riches. Cambridge: Cambridge University Press, 2015. Epp, Eldon J. “Codex Sinaiticus: Its Entrance into the Mid-Nineteenth Century Text-Critical Environment and Its Impact on the New Testament Text.” Pages 53–90 in Codex Sinaiticus: New Perspectives On The Ancient Biblical Manuscript. Edited by S. McKendrick et al. London: British Library, 2015. Epp, Eldon J. “The Late Constantin Tischendorf and Codex Sinaiticus.” Pages 13–54 in Studies on the Text of the New Testament and Early Christianity: Essays in Honor of Michael W. Holmes. NTTSD 50. Edited by D. M. Gurtner, J. Hernández Jr., and P. Foster. Leiden: Brill, 2015. Epp, Eldon J. “Text-Critical Witnesses and Methodology for Isolating a Distinctive D-Text in Acts.” NovT 59/3 (2017): 225–96. Erasmus, Desiderius. Novum Testamentum iam quintum accuratissima cura recognitum a Des. Erasmo Roter. cum Annotationibus eiusdem ita locupletatis, ut propemodum opus novum videri possit. 5th edition. Basel: Froben, 1535. Evans, Richard J. In Defense of History. London: Granta Books, 1997. Evelyn White, H. G. “Greek Ostraca and Papyri.” Pages 299–330 in The Monastery of Epiphanius at Thebes, Part II. Edited by W. E. Crum. New York: Publications of the Metropolitan Museum of Art Egyptian Expedition, 1926. Falluomini, Carla. “The Gothic Version of the New Testament.” Pages 329–50 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Fee, Gordon D. Papyrus Bodmer II (P66): Its Textual Relationships and Scribal Characteristics. SD 34. Salt Lake City: University of Utah Press, 1968. Fee, Gordon D. “Codex Sinaiticus in the Gospel of John.” NTS 15 (1968–1969): 23–44. Fee, Gordon D. “The Use of the Definite Article with Personal Names in the Gospel of John.” NTS 17 (1970): 168–83. Fee, Gordon D. “The Text of John in Origen and Cyril of Alexandria: A Contribution to Method in the Recovery and Analysis of Patristic Citations.” Biblica 52 (1971): 357–94. Fee, Gordon D. “The Text of John and Mark in the Writings of Chrysostom.” NTS 26 (1979– 1980): 525–47. Fee, Gordon D. “Origen’s New Testament and the Text of Egypt.” NTS 28 (1982)L 348–64.

Bibliography

475

Fee, Gordon D. “P75, P66, and Origen: The Myth of Early Textual Recension in Alexandria.” Pages 247–73 in Studies in the Theory and Method of New Testament Textual Criticism. Edited by E. J. Epp and G. D. Fee. Grand Rapids: Eerdmans, 1993. Fee, Gordon D. “On the Types, Classification, and Presentation of Textual Variation.” Pages 62–79 in Studies in the Theory and Method of New Testament Textual Criticism. Edited by E. J. Epp and G. D. Fee. Grand Rapids: Eerdmans, 1993. Fee, Gordon D. “Codex Sinaiticus in the Gospel of John: A Contribution to Methodology in Establishing Textual Relationships.” Pages 221–43 in Studies in the Theory and Method of New Testament Textual Criticism. Edited by E. J. Epp and G. D. Fee. Grand Rapids: Eerdmans, 1993. ffrench [sic], Patrick and Roland-François Lack, eds. The Tel Quel Reader. London: Routledge, 1998. Firestein, S. Ignorance: How it Drives Science. Oxford: Oxford University Press, 2012. Fischer, Bonifatius. Verzeichnis der Sigel für Handschriften und Kirchenschriftsteller. Vetus Latina 1/1. Herder: Freiburg, 1949. Fischer, Bonifatius. “The Use of Computers in the New Testament Studies, with Special Reference to Textual Criticism.” JTS 21/2 (1970): 297–308. Fischer, Bonifatius. Die lateinischen Evangelien bis zum 10. Jahrhundert. AGLB 13, 15, 17, 18. Freiburg: Herder, 1988–1991. Fischer, Bonifatius†. “Die lateinischen Evangelien bis zum 10. Jahrhundert. Zwei Untersuchungen zum Text.” ZNW 101 (2010): 119–44. Foucault, Michel. Language, Counter-Memory, Practice: Selected Essays and Interviews. Edited by D. F. Bouchard. Translated by D. F. Bouchard and S. Simon. Ithaca: Cornell University Press, 1977. Funghi, Maria Serena, Gabriela Messeri, and Cornelia Eva Römer, eds. Ostraca greci e bilingui del Petrie Museum of Egyptian Archaeology (O.Petr.Mus.). Florence: Edizioni Gonnelli, 2012. Gäbel, G., A. Hüffmeier, G. Mink, H. Strutwolf, and K. Wachtel. “The CBGM Applied to Variants from Acts – Methodological Background.” TC: A Journal of Biblical Textual Criticism 20 (2015) http://rosetta.reltech.org/TC/v20/TC-2015-CBGM-background.pdf. Gäbel, Georg. “The Text of P127 (P.Oxy. 4968) and Its Relationship with the Text of Codex Bezae.” NovT 53/2 (2011): 107–52. Gäbel, Georg. “‘Western Text,’ ‘D-Text Cluster,’ ‘Bezan Trajectory,’ Or What Else? – A Preliminary Study,” Pages 83–136 in Novum Testamentum Graecum: Editio Critica Maior III: The Acts of the Apostles; Part 3: Studies. Edited by H. Strutwolf et al. Stuttgart: Deutsche Bibelgesellschaft, 2017. Gallagher, J. Tim. “A Study of von Soden’s H-Text in the Catholic Epistles.” AUSS 8 (1970): 97–119. Gaselee, S. “Christian Egypt.” Archaeological Report (Egypt Exploration Fund) (1911–1912): 54–79. Gathercole, Simon J. “The Titles of the Gospels in the Earliest New Testament Manuscripts.” ZNW 104/1 (2013): 33–76. Geer Jr., Thomas C. and Jean-François Racine, “Analyzing and Categorizing New Testament Greek Manuscripts.” Pages 497–518 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. NTTSD 42. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Genette, Gérard. Introduction à l’architexte. Paris: Seuil, 1979. Genette, Gérard. Palimpsestes: la littérature au second degré. Paris: Seuil, 1982. Genette, Gérard. Seuils. Paris: Seuil, 1987.

476

Bibliography

Genette, Gérard. The Architext: An Introduction. Translated by J. E. Lewin. Berkeley: University of California Press, 1992. Genette, Gérard. Paratexts: Thresholds of Interpretation. Translated by J. E. Lewin. Cambridge: Cambridge University Press, 1997. Genette, Gérard. Palimpsests: Literature in the Second Degree. Translated by C. Newman and C. Doubinsky. Lincoln: University of Nebraska Press, 1997. Gentry, Peter J. The Asterisked Materials in the Greek Job. Atlanta: Scholars, 1995. Gentry, Peter J. “The Aristarchan Signs in the Textual Tradition of LXX Ecclesiastes.” Pages 463–78 in In the Footsteps of Sherlock Holmes: Studies in the Biblical Text in Honour of Anneli Aejmelaeus. Edited by K. De Troyer, T. M. Law, and M. Liljeström. Leuven: Peeters, 2014. Gerber, Albrecht. Deissmann the Philologist. BZNW 171. Berlin: De Gruyter, 2010. Gibson, Roy. “On the Nature of Ancient Letter Collections.” JRS 102 (2012): 56–78. Gibson, Roy. “Starting with the Index in Pliny.” Pages 33–55 in The Roman Paratext: Frame, Texts, Readers. Edited by L. Jansen. Cambridge: Cambridge University Press, 2014. Gignac, F. A Grammar of the Greek Papyri of the Roman and Byzantine Periods: Volume I Phonology. Milan: Istituto Editoriale Cisalpino-La Goliardica, 1976. Glaue, Paul and Karl Helm. Das gotisch-lateinische Bibelfragment der Universiätsbibliothek der Gießen. Gießen: Töppelmann, 1910. Goff, Philip, Arthur E. Farnsley II, and Peter J. Thuesen. “The Bible in American Life Today.” Pages 5–32 in The Bible in American Life. Edited by P. Goff, A. E. Farnsley II, and P. J. Thuesen. Oxford: Oxford University Press, 2017. Goodspeed, Edgar J. “The Golden Latin Gospels.” The Biblical World 38 (1911): 67–70. Goodspeed, Edgar J. “Review: Hoskier’s Study of the New Testament Versions.” AJT 16 (1912): 652–54. Goswell, Greg. “Early Readers of the Gospels: The Kephalaia and Titloi of Codex Alexandrinus.” JGRChJ 6 (2009): 134–74. Goswell, Greg. “Ancient Patterns of Reading: The Subdivisions of the Acts of the Apostles in Codex Sinaiticus.” JGRChJ 7 (2010): 68–97. Goswell, Greg. “An Early Commentary on the Pauline Corpus: The Capitulation of Codex Vaticanus.” JGRChJ 8 (2011): 51–82. Goulburn, Edward Meyrick. John William Burgon, Late Dean of Chichester: A Biography with Extracts from his Letters and Early Journals. 2 volumes. London: John Murray, 1892. Gilbert, Penny. “Automatic Collation: A Technique for Medieval Texts.” Computers and the Humanities 7/3 (1973): 139–47. Gracia, Jorge J. E. A Theory of Textuality: The Logic and Epistemology. Albany: State University of New York Press, 1995. Grafton, Anthony T. and Megan Williams. Christianity and the Transformation of the Book: Origen, Eusebius, and the Library of Caesarea. Cambridge, MA: Harvard University Press, 2006. Gratfon, Anthony T. Joseph Scaliger. A Study in the History of Classical Scholarship. I: Textual Criticism and Exegesis. Oxford: Clarendon, 1983. Graves, Charles L. Life and Letters of Alexander Macmillan. London: Macmillan, 1910. Grayston, Kenneth. “Computers and the New Testament.” NTS 17/4 (1971): 477–80. Greetham, D. C. Textual Scholarship: An Introduction. New York: Garland, 1994. Gregory, Caspar René. Textkritik des Neuen Testaments. Leipzig: Hinrichs’sche, 1900. Gregory, Caspar René. Die griechischen Handschriften des Neuen Testaments. Leipzig: Hinrichs’sche, 1908.

Bibliography

477

Grenfell, Bernard P. and Arthur S. Hunt. The Amherst Papyri: Being an Account of the Greek Papyri in the Collection of Lord Amherst of Hackney. Part II Classical Fragments and Documents of the Ptolemaic, Roman and Byzantine Periods. London: H. Frowde, 1901. Griesbach, Johann Jakob. Symbolae criticae ad supplendas et corrigendas variarum N. T. lectionum collectiones. Accedit multorum N. T. codicum Graecorum descriptio et examen. Tomus prior. Halle: Curtius, 1785. Griesbach, Johann Jakob. Novum Testamentum Graece. Textum ad fidem codicum versionum et patrum recensuit et lectionis varietatem adiecit…Volumen I. IV Evangelia complectens. Editio secunda emendatior multoque locupletior. Halle: Curtius, 1796. Gryson, Roger, ed. Apocalypsis Johannis. Vetus Latina 26/2. Freiburg: Herder, 2000–2003. Guarducci, Margherita. Epigrafia Greca IV: Epigrafi sacre pagane e cristiane. Rome: Arbor sapientiae, 1978. Guidi, I. Il testo copto del testament di Abramo. Rome, 1900. Gurry, Peter J. “How your Greek NT is Changing: A Simple Introduction to the CoherenceBased Genealogical Method (CBGM).” JETS 59 (2016): 675–89. Gurry, Peter J. “The Number of Variants in the Greek New Testament: A Proposed Estimate.” NTS 62/1 (2016): 77–121. Gurry, Peter J. A Critical Examination of the Coherence-Based Genealogical Method in New Testament Textual Criticism. NTTSD 55. Leiden: Brill 2017. Gurry, Peter J. “The Harklean Syriac and the Development of the Byzantine Text: A Historical Test for the Coherence-Based Genealogical Method (CBGM).” NovT 60/2 (2018): 183– 200. Gwynn, J. The Apocalypse of St. John: In a Syriac Version Hitherto Unknown. Dublin: Hodges, Figgis, and Co., 1897. Haelewyck, J.-C. “Un nouveau témoin vieux latin de Marc. Le ms. Durham, Cathedral Library A.II.10 + C.III.13 + C.III.20.” RBen 123 (2013): 5–12. Haelewyck, J.-C., ed. Vetus Latina. Die Reste der altlateinischen Bibel. 17. Evangelium Secundum Marcum. Freiburg-im-Breisgau: Herder, 2013–. Haentjens Dekker, Ronald and Gregor Middell. “Computer-Supported Collation with CollateX.” Supporting Digital Humanities 2011: Answering the Unaskable. Copenhagen, Denmark (2011). Haines-Eitzen, Kim. Guardians of Letters: Literacy, Power, and the Transmitters of Early Christian Literature. Oxford: Oxford University Press, 2000. Hall, H. R. “Other Donations to the Egyptian and Assyrian Department.” The British Museum Quarterly 5/2 (1930): 48–9. Hall, Isaac H. “The Title-Page of the Elzevir Greek New Testament of 1624.” JBL 10 (1891): 147–50. Halliday, Michael A. K. Halliday’s Introduction to Functional Grammar. 4th edition, revised by C. M. I. M. Matthiessen. London: Routledge, 2014. Hamilton, Alastair. “In Search of the Most Perfect Text: The Early Modern Printed Polyglot Bibles from Alcalá (1510–1520) to Brian Walton (1654–1658).” Pages 138–56 in The New Cambridge History of the Bible. Volume 3: From 1450 to 1750. Edited by E. K. Cameron. Cambridge: Cambridge University Press, 2016. Hamilton, Paul. Historicism. 2nd edition. London: Routledge, 2003. Harari, Y. N. Sapiens. A Brief History of Humankind. London: Penguin, 2014. Haraway, Donna. “Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective.” Feminist Studies 14/3 (1988): 575–99. Haraway, Donna. Simians, Cyborgs, and Women: The Reinvention of Nature. New York: Routledge, 1991.

478

Bibliography

Harris, J. Rendel. Codex Bezae: A Study of the So-Called Western Text of the New Testament. Cambridge: Cambridge University Press, 1891. Harris, J. Rendel. The Annotators of the Codex Bezae [with Some Notes on Sortes Sanctorum]. London: C. J. Clay and Sons, 1901. Head, Peter M. “Additional Greek Witnesses to the New Testament (Ostraca, Amulets, Inscriptions, and Other Sources).” Pages 429–60 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. NTTSD 42. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2014. Heide, Martin. Der einzig wahre Bibeltext? Erasmus von Rotterdam und die Frage nach dem Urtext. 5th edition. Nürnberg: VTR, 2006. Heide, Martin. “Die syrische Johannes-Apokalypse. Zum gegenwärtigen Stand der Forschung.” Pages 71–82 in Die Johannesoffenbarung: ihr Text und ihre Auslegung. ABG 38. Edited by M. Karrer and M. Labahn. Leipzig: Evangelische Verlagsanstalt, 2012. Heide, Martin. “Zur Vorlage und Bedeutung der äthiopischen Bibelübersetzung.” Pages 289– 313 in Studien zum Text der Apokalypse. ANTF 47. Edited by M. Sigismund, M. Karrer, and U. Schmid. Berlin: De Gruyter, 2015. Heide, Martin. “Die syrische Apokalypse oder Offenbarung an Johannes. Kritische Edition der harklensischen Textzeugen.” Pages 81–187 in Studien zum Text der Apokalypse II. ANTF 50. Edited by M. Sigismund and D. Müller. Berlin: De Gruyter, 2017. Heine, Ronald E. Origen: Scholarship in the Service of the Church. Oxford: Oxford University Press, 2010. Hellholm, David and Vemund Blomkvist. “Parainesis as an Ancient Genre-Designation: The Case of the “‘Euthalian Apparatus’ and the ‘Affiliated Argumenta.’” Pages 465–519 in Early Christian Paraenesis in Context. BZNW 125. Edited by J. Starr and T. EngbergPedersen. Berlin: De Gruyter, 2004. Helm, Rudolf. Eusebius Werke, siebenter Band. Die Chronik des Hieronymus. Leipzig: Hinrichs’sche, 1926. Helm, Rudolf. Hieronymus’ Zusätze in Eusebius’ Chronik und ihr Wert für die Literaturgeschichte. Liepzig: Dieterich, 1929. Herder, Johann Gottfried. “Essay on the Origin of Language” (1772). In On the Origin of Language: Two Essays by Johann Gottfried Herder and Jean-Jacques Rousseau. Translated by J. H. Morann and A. Gode. New York: Frederick Ungar, 1966. Hernández Jr., Juan. Scribal Habits and Theological Influences in the Apocalypse: The Singular Readings of Sinaiticus, Alexandrinus, and Ephraemi. WUNT 2/218. Tübingen: Mohr Siebeck, 2006. Hernández Jr., Juan. “Andrew of Caesarea and his Reading of Revelation: Catechesis and Paranesis.” Pages 755–74 in Die Johannesapokalypse. WUNT 287. Edited by J. Frey, J. A. Kelhoffer, and F. Tóth. Tübingen: Mohr Siebeck, 2012. Hernández Jr., Juan. “The Creation of a Fourth-Century Witness for the Andreas Text Type: A Misreading in the Apocalypse’s Textual History.” NTS 60 (2013): 106–20. Hernández, Raquel Martín and Sofía Torallas Tovar. “The Use of the Ostracon in Magical Practice in Late antique Egypt: Magical handbooks vs. Material Evidence.” SMSR 80 (2014): 780–800. Hindley, Meredith. “The Rise of the Machines. NEH and the Digital Humanities: The Early Years.” Humanities 34/4 (2013). Hockey, Susan. “The History of Humanities Computing.” Pages 3–19 in A Companion to Digital Humanities. Edited by S. Schreibman, R. Siemens, and J. Unsworth. Malden: Blackwell, 2004.

Bibliography

479

Hodges, Zane C. and Arthur L. Farstad. The Greek New Testament according to the Majority Text. 2nd edition. Nashville: Nelson, 1985. Hofmann, Josef and Siegbert Uhlig. Novum Testamentum Aethiopice: Die Katholischen Briefe. Äthiopistische Forschungen 29. Stuttgart: Franz Steiner, 1993. Hofmann, Josef. Die äthiopische Johannes-Apokalypse kritisch untersucht. Leuven: CSCO, 1969. Hofmann, Josef. “Das Neue Testament in äthiopischer Sprache.” Pages 345–73 in Die alten Übersetzungen des Neuen Testaments, die Kirchenväterzitate und Lektionare. ANTF 5. Edited by K. Aland. Berlin: De Gruyter, 1972. Holmes, Martin. “Whatever Happened to Interchange?” Digital Scholarship in the Humanities 32 suppl. 1 (2017): i63–i68. Holmes, Michael W. “The Text of P46: Evidence of the Earliest ‘Commentary’ on Romans?” Pages 189–203 in New Testament Manuscripts: Their Texts and Their World. TENT 2. Edited by T. J. Kraus and T. Nicklas. Leiden: Brill, 2006. Holmes, Michael W. “Text and Transmission in the Second Century.” Pages 61–79 in The Reliability of the New Testament: Bart D. Ehrman and Daniel B. Wallace in Dialogue. Edited by R. B. Stewart. Minneapolis: Fortress, 2011. Holmes, Michael W. “What Text Is Being Edited? The Editing of the New Testament.” Pages 91–122 in Editing the Bible: Assessing the Task Past and Present. Edited by J. S. Kloppenborg and J. H. Newman. Atlanta: SBL, 2012. Holmes, Michael W. “From ‘Original Text’ to ‘Initial Text’: The Traditional Goal of New Testament Textual Criticism in Contemporary Discussion.” Pages 637–88 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. NTTSD 42. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Holmes, Michael W. “Reasoned Eclecticism in New Testament Textual Criticism.” Pages 771– 802 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. NTTSD 42. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Holquist, Michael. Dialogism: Bakhtin and his World. London: Routledge, 1990. Holub, Robert C. Reception Theory: A Critical Introduction. London: Routledge, 1984. Hoogendijk, F. A. J. and A. Jördens, eds. Berichtigungsliste der Griechischen Papyrusurkunden aus Ägypten. Volume 13. Collected by J. M. S. Cowey und F. A. J. Hoogendijk. Leiden: Brill, 2017. Horne, Thomas Hartwell and Samuel Prideaux Tregelles. An Introduction to the Critical Study and Knowledge of the Holy Scriptures. Volume IV. An Introduction to the Textual Criticism of the New Testament; with Analyses, etc., of the Respective Books, and a Bibliographical List of Editions of the Scriptures in the Original Texts and the Ancient Versions. 10th edition. London: Longman, Brown, Green, Longmans, & Roberts, 1856. Horsley, Gregory H. R., and John A. L. Lee. “A Lexicon of the New Testament with Documentary Parallels: Some Interim Entries, 1.” Filologia Neotestamentaria 10 (1997): 55–84. Hort, Arthur F. Life and Letters of Fenton John Anthony Hort. London: Macmillan, 1896. Hort, F. J. A. “Reviews.” The Journal of Classical and Sacred Philology 4 (1859): 373–84. Hort, F. J. A. “On the End of the Epistle to the Romans.” Journal of Philology 3/5 (1871): 51– 80. Horton, Charles. “The Chester Beatty Biblical Papyri: A Find of the Greatest Importance.” Pages 149–60 in The Earliest Gospels: The Origins and Transmission of the Earliest Christian Gospels; The Contribution of the Chester Beatty Gospel Codex P45. Edited by C. Horton. London: T&T Clark, 2004. Hoskier, H. C. A Full Account and Collation of the Greek Cursive Evangelium 604 (with two facsimiles). London: David Nutt, 1890.

480

Bibliography

Hoskier, H. C. Table of Arbitrage Parities between New York and London. London: Richard Clay and Sons, 1892. Hoskier, H. C. The Golden Latin Gospels: JP in the Library of J. Pierpont Morgan (formerly known as the “Hamilton Gospels” and sometimes as King Henry the VIIIth’s Gospels) now edited for the first time, with critical introduction and notes, and accompanied by four full-page facsimiles. New York: Private Printing, 1910. Hoskier, H. C. Concerning the Genesis of the Versions of the N.T. Remarks Suggested by the Study of JP and the Allied Questions as Regards the Gospels. 2 volumes. London: Bernard Quaritch, 1910. Hoskier, H. C. “The Elzevir New Testaments of 1624 and 1633.” JTS 12 (1911): 454–57. Hoskier, H. C. “The Authorized Version of 1611.” Bibliothecha Sacra 68 (1911): 693–704. Hoskier, H. C. “New Edition of the Codex Veronensis (b).” The American Journal of Philology 32 (1911): 220–21. Hoskier, H. C. Concerning the Date of the Bohairic Version: Covering a Detailed Examination of the Text of the Apocalypse and a Review of the Some of the Writings of the Egyptian Monks. London: Bernard Quaritch, 1911. Hoskier, H. C. “Evan. 157 (Rome. Vat. Urb. 2).” JTS 14 (1912–1913): 78–116; 242–93; 359–84. Hoskier, H. C. “The Lost Commentary of Oecumenius on the Apocalypse.” The American Journal of Philology 34 (1913): 300–14. Hoskier, H. C. “The New Codex ‘W.’” Expositor 5/5 (May 1913): 467–80. Hoskier, H. C. “Von Soden’s Text of the New Testament.” JTS 15/3 (1914): 307–26. Hoskier, H. C. Codex B and its Allies: A Study and an Indictment. 2 volumes. London: Bernard Quaritch, 1914. Hoskier, H. C. The Text of Codex Usserianus 2., or r2 (“Garland of Howth”) with Critical Notes to Supplement and Correct the Collation of the Late T. K. Abbott. London: Bernard Quaritch, 1919. Hoskier, H. C. “Manuscripts of the Apocalypse – Recent Investigations, part I.” Bulletin of the John Rylands Library 7/1 (1922): 118–37. Hoskier, H. C. “Manuscripts of the Apocalypse – Recent Investigations, part II.” Bulletin of the John Rylands Library 7/2 (1923): 256–68. Hoskier, H. C. “Manuscripts of the Apocalypse – Recent Investigations, part III.” Bulletin of the John Rylands Library 7/3 (1923): 507–25. Hoskier, H. C. “Manuscripts of the Apocalypse – Recent Investigations, part IV.” Bulletin of the John Rylands Library 8/1 (1924): 236–75. Hoskier, H. C. “Manuscripts of the Apocalypse – Recent Investigations, part V.” Bulletin of the John Rylands Library 8/2 (1924): 412–43. Hoskier, H. C. Immortality. Boston: The Stratford Company, 1925. Hoskier, H. C. The Complete Commentary of Oecumenius on the Apocalypse: Now Printed for the First Time from Manuscripts at Messina, Rome, Salonika, and Athos. Ann Arbor: University of Michigan Press, 1928. (repr. Wipf & Stock, 2008). Hoskier, H. C. Concerning the Text of the Apocalypse: Collation of All Existing Available Greek Documents with the Standard Text of Stephen’s Third Edition Together with the Testimony of Versions, Commentaries and Fathers. 2 volumes. London: Bernard Quaritch, 1929. Hoskier, H. C. De Contemptu Mundi: A Bitter Satirical Poem of 3000 Lines upon the Morals of the XIIth Century by Bernard of Morval Monk of Cluny (fl. 1150) Re-edited, with Introduction and Copious Variants from all the Know MSS. London: Bernard Quaritch, 1929.

Bibliography

481

Hoskier, H. C. The Bronze Horses: A Comment on the Prose-poem of Amy Lowell. Portland, ME: Mosher Press, 1930. Hoskier, H. C. What is Nirvana? Portland, ME: Mosher Press, 1930. Hoskier, H. C. (Signpost). In Tune with the Universe. London: Rider & Co., 1932. Hoskier, H. C. “Concerning ‫ את‬and its Very Special Use in the Old Testament.” Pages 96– 117 in Amiticiæ Corolla: A Volume of Essays Presented to James Rendel Harris, C.Litt. on the Occasion of His Eightieth Birthday. Edited by H. G. Wood. London: University of London Press, 1933. Hoskier, H. C. (Signpost). The Back of Beyond. London: The Daniel Company, 1934. Hoskier, H. C. “A Study of the Chester-Beatty Codex of the Pauline Epistles.” JTS 38 (1937): 148–63. Hoskier, H. C. Appendix to an Article on the Chester-Beatty Papyrus of the Pauline Epistles known as P46 in the Journal of Theological Studies no. 150 Setting Forth Here in Supplementary Detail the Shorter Text of that Important Document. Oxford: Oxford University Press, 1937. Hoskier, H. C. A Commentary on the Various Readings in the Text of the Epistle to the Hebrews in the Chester-Beatty Papyrus P46 (circa 200 A.D.). London: Bernard Quaritch, 1938. Houghton, H. A. G. and Catherine J. Smith. “Digital Editing and the Greek New Testament.” Pages 110–27 in Ancient Worlds in Digital Culture. Edited by C. Clivaz, P. Dilley and D. Hamidović. Leiden: Brill, 2016. Houghton, H. A. G., D. C. Parker, Peter Robinson, and Klaus Wachtel. “The Editio Critica Maior of the Greek New Testament: Twenty Years of Digital Collaboration.” Digital Philology 6/2 (forthcoming). Houghton, H. A. G. “A Newly Identified Old Latin Gospel Manuscript: Würzburg Universitätsbibliothek M.p.th.f.67.” JTS 60 (2009): 1–21. Houghton, H. A. G. “The St Petersburg Insular Gospels: Another Old Latin Witness.” JTS 61 (2010): 110–27. Houghton, H. A. G. “Recent Developments in New Testament Textual Criticism.” Early Christianity 2 (2011): 245–68. Houghton, H. A. G. “Chapter Divisions, Capitula Lists, and the Old Latin Versions of John.” Revue Bénédictine 121/2 (2011): 316–56. Houghton, H. A. G. “The Use of the Latin Fathers for New Testament Textual Criticism.” Pages 375–405 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. NTTSD 42. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2014. Houghton, H. A. G. “The Electronic Scriptorium: Markup for New Testament Manuscripts.” Pages 31–60 in Digital Humanities in Biblical, Early Jewish and Early Christian Studies. Edited by C. Clivaz, A. Gregory, and D. Hamidović. Leiden: Brill, 2014. Houghton, H. A. G. The Latin New Testament: A Guide to Its Early History, Texts, and Manuscripts. Oxford: Oxford University Press, 2016. Houghton, H. A. G. “Electronic Transcriptions of New Testament Manuscripts and their Accuracy, Documentation and Publication.” In Ancient Manuscripts in Digital Culture: Visualisation, Data, Communication. Edited by D. Hamidović, C. Clivaz and S. Savant. Leiden: Brill (forthcoming). Howard-Hill, Thomas H. “A Practical Scheme for Editing Critical Texts with the Aid of a Computer.” Proof: The Yearbook of American Bibliography and Textual Studies 3 (1973): 335–56.

482

Bibliography

Huffer, Lynne. Maternal Pasts, Feminist Futures: Nostalgia, Ethics, and the Question of Difference. Stanford: Stanford University Press, 1998. Hull, Robert F. The Story of the New Testament Text: Movers, Materials, Motives, Methods, and Models. Atlanta: SBL, 2010. Hurtado, Larry. Text-Critical Methodology and the Pre-Caesarean Text. Grand Rapids: Eerdmans, 1981. Hurtado, Larry. The Earliest Christian Artifacts: Manuscripts and Christian Origins. Grand Rapids: Eerdmans, 2006. Hutton, Edward A. An Atlas of Textual Criticism Being an Attempt to Show the Mutual Relationship of the Authorities for the Text of the New Testament Up to about 1000 A.D. Cambridge: Cambridge University Press, 1911. Huyssen, Andreas. “Nostalgia for Ruins.” Grey Room 23 (2006): 6–21. Huyssen, Andreas. “Authentic Ruins: Products of Modernity.” Pages 17–28 in Ruins of Modernity. Edited by J. Hell and A. Schönle. Durham: Duke University Press, 2010. Ide, Nancy M. and C. M. Sperberg-McQueen. “The TEI: History, Goals, and Future.” Computers and the Humanities 29/1 (1995): 5–15. International Greek New Testament Project. The Gospel according to Luke, The New Testament in Greek 3. Oxford: Clarendon, 1983, 1987. Jänicke, Stefan, Annette Geßner, Greta Franzini, Melissa Terras, Simon Mahony, and Gerik Scheuermann. “TRAViz: A Visualization for Variant Graphs.” Digital Scholarship in the Humanities 30 suppl. 1 (2015): i83–i99. Jansen, Laura, ed. The Roman Paratext: Frame, Texts, Readers. Cambridge: Cambridge University Press, 2014. Jenkins, R. G. “The First Column of the Hexapla: The Evidence of the Milan Codex (Ralfs 1098) and the Cairo Genizah Fragment (Ralfs 2005).” Pages 88–102 in Origen’s Hexapla and Fragments: Papers Presented at the Rich Seminar on the Hexapla, Oxford Centre for Hebrew and Jewish Studies, 25th–3rd August 1994. TSAJ 58. Edited by A. Salvesen. Tübingen: Mohr Siebeck, 1998. Jongkind, Dirk. Scribal Habits of Codex Sinaiticus. Piscataway: Gorgias, 2007. Jongkind, Dirk. “The Text and Lexicography of the New Testament in the Eighteenth and Nineteenth Centuries.” Pages 274–99 in A History of Biblical Interpretation: Volume 3 The Enlightenment through the Nineteenth Century. Edited by D. F. Watson and A. J. Hauser. Grand Rapids: Eerdmans, 2017. Jongkind, Dirk, Peter J. Williams, Peter M. Head, and Patrick James, eds. The Greek New Testament Produced at Tyndale House, Cambridge. Wheaton: Crossway, 2017. Jones, Brice C. “New Testament Texts on Greek Amulets from Late Antiquity and Their Relevance for Textual Criticism.” PhD diss., Concordia University, 2015. Jones, Brice C. New Testament Texts on Greek Amulets from Late Antiquity. LNTS 554. London: Bloomsbury, 2016. Jones, Steven E. Roberto Busa, S.J., and the Emergence of Humanities Computing: The Priest and the Punched Cards. New York: Routledge, 2016. Juckel, Andrea. “Der ‘harklensische Apparat’ der Acta Apostolorum.” Pages 228–45 in Novum Testamentum Graecum: Editio Critica Maior III: The Acts of the Apostles; Part 3: Studies. Edited by H. Strutwolf et al. Stuttgart: Deutsche Bibelgesellschaft, 2017. Jülicher, Adolf, ed. Itala. Das Neue Testament in altlateinische Überlieferung. I. MatthäusEvangelium. Berlin: De Gruyter, 1938. Jülicher, Adolf, Walter Matzkow, and Kurt Aland, eds. Itala. Das Neue Testament in altlateinische Überlieferung. I. Matthäus-Evangelium. 2nd revised edition. Berlin: De Gruyter, 1972.

Bibliography

483

Kamphuis, Bart L. F. et al. “Sleepy Scribes and Clever Critics. A Classification of Conjectures on the Text of the New Testament.” NovT 57 (2015): 72–90. Kang, Taran. “The Problem of History in Hannah Arendt.” Journal of the History of Ideas 74/1 (2013): 139–60. Karrer, Martin. “Der Text der Johannesapokalypse.” Pages 43–78 in Die Johannesapokalypse. WUNT 287. Edited by J. Frey, J. A. Kelhoffer and F. Tóth. Tübingen: Mohr Siebeck, 2012. Karrer, Martin. “Das Neue Testament des Erasmus und Luthers.” Theologische Zeitschrift 73 (2017): 299–324. Karrer, T. F. “Antiquarisches.” Zeitschrift für die gesammte luterische Theologie und Kirche 11 (1850): 121–24. Kemman, Max and Martijn Kleppe. “User Required? On the Value of User Research in the Digital Humanities.” Pages 63–74 in Selected Papers from the CLARIN 2014 Conference, October 24–25, 2014, Soesterberg, The Netherlands. Linköping University Electronic Press, 2015. Kenney, E. J. “Books and Readers in the Roman World.” Pages 1–32 in The Cambridge History of Classical Literature: Volume 2: Latin Literature. Edited by E. J. Kenney and W. V. Clausen. Cambridge: Cambridge University Press, 1982. Kenyon, Frederic G. Recent Developments in the Textual Criticism of the Greek Bible. London: British Academy, 1933. Kenyon, Frederic G. The Chester Beatty Biblical Papyri: Descriptions of Twelve Manuscripts on Papyrus of the Greek Bible. London: Emery Walker, 1933–1937. Kenyon, Frederic G. The Text of the Greek Bible: A Student’s Handbook. London: Duckworth, 1937 (new edition, 1949). Kenyon, Frederic G. “Hesychius and the Text of the New Testament.” Pages 245–50 in Mémorial Lagrange. Edited by L.-H. Vincent. Paris: J. Gabalda, 1940. Kern, Otto. Inscriptiones Graecae. Bonn: Marcus & Weber, 1913. Keysler, J. G. Travels through Germany, Hungary, Bohemia, Switzerland, and Lorrain. Giving a True and Just Description of the Present State of those Countries; their Natural, Literary, and Political History; Manners, Laws, Commerce, Manufactures, Painting, Sculpture, Architecture, Coins, Antiquities, Curiosities of Art and Nature, Etc. 2nd edition. 4 volumes. London: G. Keith, A. Linde, S. Crowder, P. Davey, B. Law, T. Field, and E. Dilly, 1756. Kieffer, René. Au delà des recensions? L’évolution de la tradition textuelle dans Jean VI, 52– 71. ConBNT 3. Lund: C.W.K. Gleerup, 1968. Kießling, Emil. Wörterbuch der griechischen Papyrusurkunden, mit Einschluß der griechischen Inschriften, Aufschriften, Ostraka, Mumienschilder usw. aus Ägypten. Berlin: Selbstverlag der Erben and Wiesbaden: Harrassowitz, 1944–1993. Kilbride, William. “Saving the Bits: Digital Humanities Forever?” Pages 408–19 in A New Companion to Digital Humanities. Edited by S. Schreibman, R. Siemens, and J. Unsworth. Oxford: John Wiley, 2016. Kilpatrick, G. D. “Atticism and the Text of the Greek New Testament.” Pages 15–32 in The Principles and Practice of New Testament Textual Criticism: Collected Essays of G. D. Kilpatrick. BETL 96. Edited by J. K. Elliott. Leuven: Peeters, 1990. King, Karen L. “‘What is an Author?’ Ancient Author-Function in the Apocryphon of John and the Apocalypse of John.” Pages 15–42 in Scribal Practices and Social Structures among Jesus Adherents: Essays in Honor of John S. Kloppenborg. Edited by W. E. Arnal et al. Leuven: Peeters, 2016. Kirschenbaum, Matthew G. “‘So the Colors Cover the Wires’: Interface, Aesthetics, and Usability.” Pages 523–42 in A Companion to Digital Humanities. Edited by S. Schreibman, R. Siemens, and J. Unsworth. Malden: Blackwell, 2004.

484

Bibliography

Klijn, A. F. J. and G. J. Reinink. Patristic Evidence for Jewish-Christian Sects. NTSup 36. Leiden: Brill, 1973 Klijn, A. F. J. “Jewish Christianity in Egypt.” Pages 161–75 in The Roots of Egyptian Christianity. Edited by B. A. Pearson and J. E. Goehring. Philadelphia: Fortress, 1986. Kloppenborg, John S. “Disciplined Exaggeration.” NovT 59/4 (2017): 390–414. Knapp, Georg Christian. Η καινη διαθηκη. Novum Testamentum Graece. Recognovit atque insignioris lectionum varietatis et argumentorum notationes subiunxit…Editio quarta. London: Priestley, 1824. Knibb, Michael. Translating the Bible: The Ethiopic Version of the Old Testament. Oxford: Oxford University Press, 1999. Knopf, Rudolf. “Eine Thonscherbe mit dem Texte des Vaterunser.” MDAIA 25 (1900): 313– 24. Knopf, Rudolf. “Eine Thonscherbe mit dem Texte des Vaterunser.” ZNW 2 (1901): 228–33. Knust, Jennifer W. and Tommy Wasserman. To Cast the First Stone: The Transmission of a Gospel Story. Princeton: Princeton University Press, 2018. Koester, Helmut. Introduction to the New Testament, Vol. 2: History and Literature of Early Christianity. 2nd edition. Berlin: De Gruyter, 2000. Komoszewski, J. Ed, M. James Sawyer and Daniel B. Wallace. Reinventing Jesus: How Contemporary Skeptics Miss the Real Jesus and Mislead Popular Culture. Grand Rapids: Kregel, 2006. Kraft, Robert A. “The Use of Computers in New Testament Textual Criticism.” Pages 268–82 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. Edited by B. D. Ehrman and M. W. Holmes. Grand Rapids: Eerdmans, 1995. Krans, Jan. Beyond What Is Written. Erasmus and Beza as Conjectural Critics of the New Testament. NTTS 35. Leiden: Brill, 2006. Krans, Jan. “Erasmus and the Text of Revelation 22:19: A Critique of Thomas Holland’s Crowned With Glory.” TC: A Journal of Biblical Textual Criticism 16 (2011): 1–19. Kraus, Thomas J. “Manuscripts with the Lord’s Prayer – They are More than Simply Witnesses to that Text Itself.” Pages 227–66 in New Testament Manuscripts: Their Texts and Their World. TENTS 2. Edited by T. J. Kraus and T. Nicklas. Leiden: Brill, 2006. Kraus, Thomas J. “Die Welt der Miniaturbücher in der Antike und Spätantike – Prolegomena und eine erste methodische Annäherung für eine Datensammlung.” SNTU 35 (2010): 79– 110. Kraus, Thomas J. “Miniature Codices in Late Antiquity – Preliminary Remarks and Tendencies about a Specific Book Format.” Early Christianity 7 (2016): 134–52. Kreinecker, Christina M. 2. Thessaloniker. Mit einem Beitrag von Günther Schwab. Papyrologische Kommentare zum Neuen Testament 3. Göttingen: Vandenhoeck & Ruprecht, 2010. Kreinecker, Christina M. “The Earliest Commentaries on Paul as Sources for the Biblical Text: A New Research Project at the Institute for Textual Scholarship and Electronic Editing at the University of Birmingham (COMPAUL).” Early Christianity 3 (2012): 411–15. Kreinecker, Christina M. “Emotions in Documentary Papyri: Joy and Sorrow in Everyday Life.” Pages 451–72 in Yearbook 2011: Emotions from Ben Sira to Paul. Edited by R. Egger-Wenzel and J. Corley. Berlin: De Gruyter, 2012. Kreinecker, Christina M. “Papyrologie (NT).” Das wissenschaftliche Bibellexikon im Internet (www.wibilex.de), 2018. Kreuzsaler, Claudia. “Der Rechtsalltag von Frauen im Spiegel der Wiener Papyri.” Pages 1– 17 in Emanzipation am Nil: Frauenleben und Frauenrecht in den Papyri. Edited by H. Froschauer and H. Harrauer. Vienna: Phoibos Verlag, 2005. Kristeva, Julia. Σηµειωτική: Recherches pour und sémanalyse. Paris: Seuil, 1969.

Bibliography

485

Kristeva, Julia. Revolution in Poetic Language. Translated by M. Waller. New York: Columbia University Press, 1984. Krüger, Friedrich. Schlüssel zu von Soden’s die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt hergestellt. Göttingen: Vandenhoeck & Ruprecht, 1927. Kubo Sakae. “Textual relationships in Jude.” Pages 276–82 in Studies in New Testament Language and Text. NTSup 44. Edited by J. K. Elliott. Leiden: Brill, 1976. Kümmel, Werner G. Introduction to the New Testament. 14th revised edition of P. Feine and J. Behm. Nashville: Abingdon, 1966. Küster, Ludolf. Novum Testamentum Graecum, cum lectionibus variantibus MSS. exemplarium, versionum, editionum, SS. patrum et scriptorum ecclesiasticorum; et in easdem notis. Accedunt loca scripturae parallela, aliaque exegetica. Praemittitur dissertatio de libris N.T. et canonis constitutione, et s. textus N. Foederis ad nostra usque tempora historia. Studio et labore Joannis Millii S.T.P. Collectionem Millianam recensuit, meliori ordine disposuit, novisque accessionibus locupletavit Ludolphus Kusterus. Rotterdam: Fritsch & Böhm, 1710. Kutzner, Edgar. Untersuchungen zur Stellung der Frau im römischen Oxyrhynchos. Europäische Hochschulschriften 392. Frankfurt: Peter Lang, 1989. Lachmann, Karl. “Rechenschaft über seine Ausgabe des Neuen Testaments.” Theologische Studien und Kritiken 3 (1830): 817–45. Lachmann, Karl. Novum Testamentum Graece. Berlin: Reimeri, 1831. La’da, Csaba A. “Die wirtschaftliche Stellung der Frau im hellenistischen, römischen und byzantinischen Ägypten (332 v. Chr.–642 n. Chr.).” Pages 35–49 in Emanzipation am Nil: Frauenleben und Frauenrecht in den Papyri. Edited by H. Froschauer and H. Harrauer. Vienna: Phoibos Verlag, 2005. Lake, Kirsopp and R. P. Blake. “The Text of the Gospels and the Koridethi Codex.” HTR 16 (1923): 267–86. Lake, Kirsopp and Silva Lake. Family 13 (The Ferrar Group): The Text according to Mark. London: Christophers, 1941. Lake, Kirsopp. “The Text of the Gospels in Alexandria.” AJT 6 (1902): 79–89. Lake, Kirsopp. Codex 1 of the Gospels and its Allies. Cambridge: Cambridge University Press, 1902. Lake, Kirsopp. Facsimiles of the Athos Fragments of Codex H of the Pauline Epistles. Oxford: Clarendon, 1905. Lake, Kirsopp. The Text of the New Testament. 6th edition. London: Rivingtons, 1943. Lang, T. J. and Matthew R. Crawford. “The Origins of Pauline Theology: Paratexts and Priscillian of Avila’s Canons on the Letters of the Apostle Paul.” NTS 63/1 (2017): 125–45. Lange, Armin et al., eds. Textual History of the Bible. Volumes 1–2. Leiden: Brill, 2016–2018. Latour, Bruno. We Have Never Been Modern. Translated by C. Porter. Cambridge, MA: Harvard University Press, 1993. Lavagnino, John. “When Not to Use TEI.” Pages 334–38 in Electronic Textual Editing. Edited by L. Burnard, K. O’Brien O’Keeffe, and J. Unsworth. New York: Modern Language Association of America, 2006. Law, T. M. “Origen’s Parallel Bible: Textual Criticism, Apologetics, or Exegesis?” JTS 59/1 (2008): 1–21. Lawlor, H. J. Chapters on the Book of Mulling. Edinburgh: David Douglas, 1897. Lefebvre, G. “Fragments grecs des Évangiles sur Ostraka.” BIFAO 4 (1904): 1–15. Leipoldt, J. Sinuthii Archimandritae Vita et Opera Omnia. Volume 1. Leipzig, 1906. Leitch, Vincent B. Cultural Criticism, Literary Theory, Poststructuralism. New York: Columbia University Press, 1992.

486

Bibliography

Lorusso, Vito. “Locating Greek Manuscripts through Paratexts: Examples from the Library of Cardinal Bessarion and other Manuscript Collections.” Pages 223–68 in Tracing Manuscripts in Time and Space through Paratexts. Edited by G. Ciotti and H. Lin. Berlin: De Gruyter, 2016. Lowe, E. A. and E. K. Rand. A Sixth-Century Fragment of the Letters of Pliny the Younger: A Study of Six Leaves of an Uncial Manuscript Preserved in the Pierpont Morgan Library, New York. Washington D.C.: Carnegie Institution of Washington, 1922. Lowe, E. A. Codices Latini Antiquiores II. Great Britain and Ireland. 2nd edition. Oxford: Clarendon, 1972. Lee, John A. L. and Gregory H. R. Horsley. “A Lexicon of the New Testament with Documentary Parallels: Some Interim Entries, 2.” Filología Neotestamentaria 11 (1998): 57–84. Lembke, M. et al., eds. Text und Textwert der griechischen Handschriften des Neuen Testaments. VI. Die Apokalypse. Teststellenkollation und Auswertungen. ANTF 49. Berlin: De Gruyter, 2017. Lembke, M. “Beobachtungen zu den Handschriften der Apokalypse des Johannes.” Pages 19– 70 in Die Johannesoffenbarung. Ihr Text und ihre Auslegung. ABG 38. Edited by M. Labahn and M. Karrer. Leipzig: Evangelische Verlagsanstalt, 2012. Lembke, M. “Der Apokalypsetext der Complutensischen Polyglotte und sein Verhältnis zur handschriftlichen Überlieferung.” Pages 33–134 in Studien zum Text der Apokalypse. ANTF 47. Edited by M. Sigismund, M. Karrer, and U. Schmid. Berlin: De Gruyter, 2015. Levinsohn, Stephen H. Discourse Features of New Testament Greek: A Coursebook on the Information Structure of New Testament Greek. 2nd edition. Dallas: SIL International, 2000. Lin, Yii-Jan. The Erotic Life of Manuscripts: New Testament Textual Criticism and the Biological Sciences. Oxford: Oxford University Press, 2016. Lipphardt, Veronika. “Isolates and Crosses in Human Population Genetics; or, A Contextualization of German Race Science.” Current Anthropology 53/5 (2010): S69–S82. Llewelyn, Stephen R. and Jim R. Harrison. Greek and Other Inscriptions and Papyri Published 1988–92. New Documents Illustrating Early Christianity 10. Grand Rapids: Eerdmans, 2012. Lorimer, Douglas A. “Science and Secularization of Victorian Images of Race.” Pages 212–35 in Victorian Science in Context. Edited by B. Lightman. Chicago: University of Chicago Press, 2012. Lüdeke, Roger. “Materialität und Varianz.” Pages 454–85 in Regeln der Bedeutung: Zur Theorie der Bedeutung literarischer Texte. Edited by F. Jannidis et al. Berlin: De Gruyter, 2003. Luijendijk, AnneMarie. “A Gospel Amulet for Joannia (P.Oxy. VIII 1151): Rhetoric and Religion on the Ground.” Pages 418–43 in Daughters of Hecate: Women and Magic in the Ancient World. Edited by K. B. Stratton and D. S. Kalleres. Oxford: Oxford University Press, 2014. MacGregor, Geddes. The Making of the Bible. Philadelphia: J. B. Lippincott, 1959. Mahmood, Saba. “Sexuality and Secularism.” Pages 47–58 in Religion, The Secular, and the Politics of Sexual Difference. Edited by L. E. Cady and T. Fessenden. New York: Columbia University Press, 2013. Mass, P. “Leitfehler und stemmatische Typen.” Byzantinische Zeitschrift 37 (1937): 289–94. Maas, P. Textkritik. 3rd edition. Teubner: Leipzig, 1957. Macksey, Richard and Eugenio Donato, eds. The Structuralist Controversy: The Languages of Criticism and the Sciences of Man. Baltimore: Johns Hopkins University Press, 1972.

Bibliography

487

Macksey, Richard and Eugenio Donato. “The Space Between – 1971.” Pages ix–xiii in The Structuralist Controversy: The Languages of Criticism and the Sciences of Man. Edited by R. Macksey and E. Donato. Baltimore: Johns Hopkins University Press, 1972. Macksey, Richard. “Concluding Remarks.” Pages 319–22 in The Structuralist Controversy: The Languages of Criticism and the Sciences of Man. Edited by R. Macksey and E. Donato. Baltimore: Johns Hopkins University Press, 1972. Maldfeld, Georg. “Die griechischen Handschriftenbruchstücke des Neuen Testaments auf Papyrus.” ZNW 42 (1949): 228–53. Maldfeld, Georg. “Berichtigungen und Ergänzungen zur Liste der griechischen Papyrusfragmente (P) des Neuen Testaments.” ZNW 43 (1950/51): 260–61. Malik, Peter, P.Beatty III (P47): The Codex, Its Scribe, and Its Text. NTTSD 52. Leiden: Brill, 2017. Marchand, James W. “The Gothic Evidence for the ‘Euthalian Matter.’” HTR 49/3 (1956): 159–67. Marcos, Natalio Fernández. The Septuagint in Context: Introduction to the Greek Version of the Bible. Leiden: Brill, 2000. Martens, Peter W. Origen and Scripture: The Contours of the Exegetical Life. Oxford: Oxford University Press, 2012. Martin, Victor and Rodolphe Kasser. Papyrus Bodmer XIV: Évangile de Luc chap. 3–24. Geneva: Bibliotheca Bodmeriana, 1961. Martin, Victor and Rodolphe Kasser. Papyrus Bodmer XV: Évangile de Jean chap. 1–15. Geneva: Bibliotheca Bodmeriana, 1961. Martini, Carlo M. Il problema della recensionalità del codice B alla luca del papiro Bodmer XIV. Rome: Biblical Institute Press, 1966. Martini, Carlo M. “Is There a Late Alexandrian Text of the Gospels?” NTS 24 (1977–1978): 285–96. McCulley, William R. “Univac Compiles a Complete Bible Concordance.” Systems Magazine (March–April 1956): 22–3. McCutcheon, R. W. “A Revisionist History of Cicero’s Letters.” Mouseion: Journal of the Classics Association of Canada, III, 13/1 (2016): 35–63. McGann, Jerome J. A Critique of Modern Textual Criticism. Chicago: University of Chicago Press, 1983. McGurk, Patrick. Latin Gospel Books From A.D. 400 to A.D. 800. Paris: Érasme, 1961. McKenzie, Judith S. and Francis Watson, eds. The Garima Gospels: Early Illuminated Gospel Books from Ethiopia. Oxford: Manar al-Athar, 2016. McKitterick, David. A History of Cambridge University Press, Volume 3: New Worlds for Learning 1873–1972. Cambridge: Cambridge University Press, 2004. McNamara, Martin. “Irish Gospel Texts Publication Project.” Proceedings of the Irish Biblical Association 38 (2015): 85–98. Mercati, Iohannis. Psalterii Hexapli Reliquiae. Pars Prima. Codex Rescriptus Bybliothecae Ambrosianae 0 39 SVP. Phototypice Expressus et Transcriptus. Rome: Bybliotheca Vaticana, 1983. Metzger, Bruce M. “The Caesarean Text of the Gospels.” JBL 64/4 (1945): 457–89. Metzger, Bruce M. “The Influence of Codex Bezae upon the Geneva Bible of 1560.” NTS 8 (1961): 72–7. Metzger, Bruce M. The Text of the New Testament: Its Transmission, Corruption and Restoration. Oxford: Clarendon, 1964. Metzger, Bruce. The Early Versions of the New Testament: Their Origin, Transmission and Limitations. Oxford: Clarendon, 1977.

488

Bibliography

Metzger, Bruce M. “The Westcott and Hort Greek New Testament – Yesterday and Today.” The Cambridge Review 103 (1981): 71–6. Metzger, Bruce M. “Bilingualism and Polylingualism in Antiquity; with a Checklist of New Testament Manuscripts Written in More Than One Language.” Pages 2.327–34 in The New Testament Age: Essays in Honor of Bo Reicke. Edited by W. C. Weinrich. Macon: Mercer University Press, 1984. Metzger, Bruce M. The Text of the New Testament: Its Transmission, Corruption and Restoration. 3rd edition. Oxford: Oxford University Press, 1992. Metzger, Bruce M. A Textual Commentary on the Greek New Testament. 2nd edition. New York: United Bible Societies, 1994. Metzger, Bruce M. The Bible in Translation: Ancient and English Versions. Grand Rapids: Baker, 2001. Metzger, Bruce M. “The Future of New Testament Textual Studies.” Pages 201–208 in The Bible as Book: The Transmission of the Greek Text. Edited by S. McKendrick and O. O’Sullivan. London: The British Library, 2003. Metzger, Bruce M. and Bart D. Ehrman. The Text of the New Testament: Its Transmission, Corruption, and Restoration. 4th edition. Oxford: Oxford University Press, 2005. Middleton, T. F. The Doctrine of the Greek Article: Applied to the Criticism and Illustration of the New Testament. Cambridge: J. & J. J. Deighton, 1833. Mill, John. Novum Testamentum. Cum lectionibus variantibus MSS. exemplarium, versionum, editionum, SS. Patrum et Scriptorum ecclesiasticorum; et in easdem notis. Accedunt loca Scripturae parallela, aliaque ἐξηγητικά et appendix ad variantes lectiones. Praemittitur dissertatio, in qua de libris N.T. et Canonis constitutione agitur, historia S. Textus N. Foederis ad nostra usque tempora deducitur, et quid in hac editione praestitum sit, explicatur. 1st edition. Oxford: Theatrum Sheldoniacum, 1707. Miller, Peter N. “The ‘Antiquarianization’ of Biblical Scholarship and the London Polyglot Bible (1653–57).” JHI 62 (2001): 463–82. Mills, Sara. Discourse. London: Routledge, 1997. Milne, H. J. M. and T. C. Skeat. Scribes and Correctors of the Codex Sinaiticus. London: British Museum, 1938. Mink, Gerd. “Zur Stemmatisierung neutestamentlicher Handschriften.” Pages 100–14 in Bericht der Hermann Kunst-Stiftung zur Förderung der neutestamentlichen Textforschung für die Jahre 1979–1981. Münster: Institut für neutestamentliche Textforschung, 1982. Mink, Gerd. “Problems of a Highly Contaminated Tradition: The New Testament.” Pages 13 –85 in Studies in Stemmatology II. Edited by P. van Reenen, A. den Hollander, and M. van Mulken. Amsterdam: Benjamins 2004. Mink, Gerd. “The Coherence-Based Genealogical Method (CBGM) – Introductory Presentation” (2009). Online at http://egora.uni-muenster.de/intf/service/downloads.shtml. Mink, Gerd. “Contamination, Coherence, and Coincidence in Textual Transmission: The Coherence-Based Genealogical Method (CBGM) as a Complement and Corrective to Existing Approaches.” Pages 141–216 in The Textual History of the Greek New Testament: Changing Views in Contemporary Research. Edited by K. Wachtel and M. W. Holmes. Atlanta: SBL, 2011. Mitthof, Fritz. “Junge Heirat, viele Geburten, frühe Verwitwung oder Scheidung: Demographische Aspekte des Frauenlebens im römischen Ägypten.” Pages 65–73 in Emanzipation am Nil: Frauenleben und Frauenrecht in den Papyri. Edited by H. Froschauer and H. Harrauer. Vienna: Phoibos Verlag, 2005. Moi, Toril ed. The Kristeva Reader. New York: Columbia University Press, 1986.

Bibliography

489

Mowitt, John. Text: The Genealogy of an Antidisciplinary Object. Durham: Duke University Press, 1992. Mosshammer, Alden A. The Chronicle of Eusebius and Greek Chronographic Tradition. Lewisburg: Bucknell University Press, 1979. Morgan, Charles. The House of Macmillan (1843–1943). London: Macmillan, 1944. Moulton, W. F. William F. Moulton: A Memoir. London: Isbister, 1899. Moulton, James H. and George Milligan. The Vocabulary of the Greek Testament Illustrated from the Papyri and Other Non-Literary Sources. London: Hodder and Stoughton, 1929. Mugridge, Alan, Copying Early Christian Texts. WUNT 362. Tübingen: Mohr Siebeck, 2016. Müller, Darius. “Abschriften des Erasmischen Textes im Handschriftenmaterial der Johannesapokalypse.” Pages 165–268 in Studien zum Text der Apokalypse. ANTF 47. Edited by M. Sigismund, M. Karrer, and U. Schmid. Berlin: De Gruyter, 2015. Müller, Darius. “Zur Elektronischen Transkription von Apokalypsehandschriften: Bericht zum Arbeitsstand.” Pages 19–30 in Studien Zum Text Der Apokalypse II. Edited by M. Sigismund and D. Müller. Berlin: De Gruyter, 2017. Müller, Darius. “Die Apokalypse-Handschriften GA 2329 und 2351: Textkritische und textgeschichtliche Beobachtungen zu zwei ‘ständigen Zeugen für die Apokcalypse’ in NestleAland28.” Pages 369–410 in Studien zum Text der Apokalypse II. ANTF 50. Edited by M. Sigismund and D. Müller. Berlin: De Gruyter, 2017. Müller-Sievers, Helmut. “Reading without Interpreting: German Textual Criticism and the Case of Georg Büchner.” Modern Philology 103/4 (2006): 498–515. Murphy, Harold S. “On the Text of Codices H and 93.” JBL 78/3 (1959): 228–37. Nagel, Peter. “Koptischen Bibelhandschriften des Alten Testaments aus frühislamischer Zeit.” Pages 131–55 in Die Koptische Kirche in den ersten drei islamischen Jahrhunderten: Beiträge zum gleichnamigen Leucoren-Kolloquium 2002. Hallesche Beiträge zur Orientwissenschaft 36. Edited by W. Beltz and H.-M. Schenke. Halle: Martin-Luther-Unversität, 2003. Negroponte, Nicholas. Being Digital. London: Hodder & Stoughton, 1995. Neill, Stephen. The Interpretation of the New Testament: 1861–1961. London: Oxford University Press, 1964. Nestle, Eberhard. “Zum Vaterunser von Megara.” ZNW 2 (1901): 347–49. Nestle, Eberhard. “Die eusebianische Evangeliensynopse.” Nueue Kirchliche Zeitschrift 19 (1908): 40–51, 93–114, 219–32. Nestle, Eberhard. Einführung in das griechische Neue Testament. 3rd edition. Göttingen: Vandenhoeck & Ruprecht, 1909. Nestle, Eberhard. “Some Points in the History of the Textus Receptus of the New Testament.” JTS 9 (1910): 564–68. Neuschäfer, Bernhard. Origenes als Philologe. 2 volumes. Basel: F. Reinhardt, 1987. Nevius, Richard C. “The Use of the Definite Article with ‘Jesus’ in the Fourth Gospel.” NTS 12 (1965): 81–5. Niccum, Curt. “The Ethiopic Version and the ‘Western’ Text of Acts in Le Texte Occidental des Actes des Apôtres.” Pages 69–88 in Transmission and Reception: New Testament Text-Critical and Exegetical Studies. Edited by J. W. Childers and D. C. Parker. Piscataway: Gorgias, 2006. Niccum, Curt. The Bible in Ethiopia: The Book of Acts. Ethiopic Manuscripts, Texts, and Studies 19. Eugene: Pickwick, 2014. Niccum, Curt. “Apokalypse Now: The Ethiopic Version of Revelation Fifty Years after Hofmann.” Pages 211–30 in Studien zum Text der Apokalypse II. ANTF 50. Edited by M. Sigismund and D. Müller. Berlin: De Gruyter, 2017.

490

Bibliography

Nichols, Stephen G. “Introduction: Philology in a Manuscript Culture.” Speculum 65 (1990): 1–10. Nichols, Stephen G. “Why Material Philology?” Zeitschrift für deutsche Philologie 116 (1997): 1–21. Nicklas, Tobias. “The Early Text of Revelation.” Pages 225–38 in The Early Text of the New Testament. Edited by C. E. Hill and M. J. Kruger. Oxford: Oxford University Press, 2012. Nongbri, Brent. “The Acquisition of the University of Michigan’s Portion of the Chester Beatty Biblical Papyri and a New Suggested Provenance.” APF 60/1 (2014): 93–116. Nongbri, Brent. “Reconsidering the Place of Papyrus Bodmer XIV–XV (𝔓75) in the Textual Criticism of the New Testament.” JBL 135/2 (2016): 405–37. Nordenfalk, Carl. Die spätantiken Kanontafeln: Kunstgeschichtliche Studien über die eusebianische Evangelien-konkordanz in den vier ersten Jahrhunderten ihrer Geschichte. Tafeln. Göteborg: O. Isacsons, 1938. Nordenfalk, Carl. “Canon Tables on Papyrus.” Dumbarton Oaks Papers 36 (1982): 29–38. Nordenfalk, Carl. “The Eusebian Canon-Tables: Some Textual Problems.” JTS 35/1 (1984): 96–104. Nus, Jean-Baptiste Eugène. Les Grands mystères: vie universelle, vie individuelle, vie sociale. Paris: Noirot, 1866. Nyhan, Julianne and Andrew Flinn. “The University Was Still Taking Account of Universitas Scientiarum: Wilhelm Ott and Julianne Nyhan.” Pages 55–73 in Computation and the Humanities: Towards an Oral History of Digital Humanities. Edited by J. Nyhan and A. Flinn. Berlin: Springer, 2016. Oakman, Robert L. “The Present State of Computerized Collation: A Review Article.” Proof: The Yearbook of American Bibliography and Textual Studies 2 (1972): 333–48. Oakman, Robert L. Computer Methods for Literary Research. 2nd edition. Athens: University of Georgia Press, 1984. Olender, Maurice. The Languages of Paradise: Race, Religion, and Philosophy in the Nineteenth Century. Translated by A. Goldhammer. Cambridge, MA: Harvard University Press, 1992. Oliver, Harold H. “The Epistle of Eusebius to Carpianus: Textual Tradition and Translation.” NovT 3 (1959): 138–45. Oliver, Kelly ed. The Portable Kristeva. Revised edition. New York: Columbia University Press, 2002. O’Loughlin, Thomas. “Harmonizing the Truth: Eusebius and the Problem of the Four Gospels.” Traditio 65 (2010): 1–29. Olsson, Bror. Papyrusbriefe aus der frühesten Römerzeit. Uppsala: Almquist & Wiksells, 1925. Omont, M. H. Notice sur un très ancien Manuscrit Grec en Onciales des Épitres de Saint Paul, conserve à la Bibliothèque Nationale (H ad Epistulas Pauli). Paris: Imprimerie Nationale, 1889. O’Neill, J. C. “The Rules Followed by the Editors of the Text Found in the Codex Vaticanus.” NTS 35 (1989): 219–28. Orbell, John. “Hoskier, Herman.” Oxford Dictionary of National Biography (2004). Orsini, Pasquale and Willy Clarysse. “Early New Testament Manuscripts and Their Dates: A Critique of Theological Paleography.” ETL 88/4 (2012): 443–74. Orsini, Pasquale. Manoscritti in maiuscola biblica, Materiali per un aggiornamento. Cassino: Università di Cassino, 2005. Osburn, Carroll D. “The Text of the Pauline Epistles in Hippolytus of Rome.” Second Century 2 (1982): 97–124.

Bibliography

491

Ott, Wilhelm. “Transcription and Correction of Texts on Paper Tape. Experiences in Preparing the Latin Bible Text for the Computer.” Revue Informatique et Statistique dans les Sciences Humaines 2 (1970): 51–67. Ott, Wilhelm. “Computers and Applications in Textual Criticism.” Pages 199–223 in The Computer and Literary Studies. Edited by A. J. Aitken, R. W. Bailey, and N. HamiltonSmith. Edinburgh: Edinburgh University Press, 1973. Ott, Wilhelm. “Software Requirements for Computer-aided Editing.” Pages 81–103 in Editing, Publishing, and Computer Technology: Papers given at the Twentieth Annual Conference on Editorial Problems, University of Toronto, 2–3 November 1984. Edited by S. Butler and W. P. Stoneman. New York: AMS Press, 1988. Palmer, E., ed. Η Καινη Διαθηκη: The Greek Text with the Readings Adopted by the Revisers of the Authorised Version. Oxford: Clarendon Press, 1881. Papaioannou, Stratis. “Byzantine historia.” Pages 297–313 in Thinking, Recording, and Writing History in the Ancient World. Edited by K. A. Raaflaub. London: Wiley Blackwell, 2014. Parker, D. C., Bruce Morrill, Klaus Wachtel, and Kenneth W. Clark. The Principio Project: Procedures for Transcribing Witnesses (2001) http://epapers.bham.ac.uk/2948. Parker, D. C. “The Development of Textual Criticism Since B.H. Streeter.” NTS 24 (1977): 149–62. Parker, D. C. Codex Bezae: An Early Christian Manuscript and its Text. Cambridge: Cambridge University Press, 1992. Parker, D. C. The Living Text of the Gospels. Cambridge: Cambridge University Press, 1997. Parker, D. C. “A New Oxyrhynchus Papyrus of Revelation: P115 (P.Oxy. 4499).” NTS 46 (2000): 159–74. Parker, D. C. “Through a Screen Darkly: Digital Texts and the New Testament.” JSNT 25/4 (2003): 395–411. Parker, D. C. “Electronic Religious Texts: The Gospel of John.” Pages 197–209 in Electronic Textual Editing. Edited by L. Burnard, K. O’Brien O’Keeffe, and J. Unsworth. New York: Modern Language Association of America, 2006. Parker, D. C. An Introduction to the New Testament Manuscripts and Their Texts. Cambridge: Cambridge University Press, 2008. Parker, D. C. Codex Sinaiticus: The Story of the World’s Oldest Bible. London: British Library, 2010. Parker, D. C. “What Is the Text of the New Testament?” Pages 95–104 in The Reliability of the New Testament: Bart D. Ehrman and Daniel B. Wallace in Dialogue. Edited by R. B. Stewart. Minneapolis: Fortress, 2011. Parker, D. C. Textual Scholarship and the Making of the New Testament: The Lyell Lectures. Oxford: Oxford University Press, 2012. Parker, D. C. “New Testament Textual Traditions in Byzantium.” Pages 2–32 in The New Testament in Byzantium. Edited by D. Krueger and R. S. Nelson. Washington, D.C.: Dumbarton Oaks, 2016. Parsons, Ernest W. “Review: A Third-Century Papyrus Codex of the Epistles of Paul by Henry A. Sanders.” Classical Journal 32 (1936): 113–15. Pasquali, Giorgio. “Review of Maas, Textkritik.” Gnomon 5 (1929): 417–35, 498–521. Pasquali, Giorgio. Storia della tradizione e critica del testo. 2nd edition. Florence: Felice le Monnier, 1971. Patrick, Graham A. F. J. A. Hort: Eminent Victorian. Decatur: Almond, 1987. Patrick, Graham A. The Miners’ Bishop: Brooke Foss Westcott. 2nd edition. Werrington: Epworth, 2004.

492

Bibliography

Parvis, Merril M. “The International Project to Establish a New Critical Apparatus of the Greek New Testament.” Crozer Quarterly 27 (1950): 301–308. Paul, Carol. “Benedict XIV’s Enlightened Patronage of the Capitoline Museum.” Pages 341– 66 in Benedict XIV and the Enlightenment: Art, Science, and Spirituality. Edited by R. Messbarger, C. M. S. Johns, and P. Gavitt. Toronto: University of Toronto Press, 2016. Pavel, Thomas G. The Spell of Language: Poststructuralism and Speculation. Translated by L. Jordan and T. G. Pavel. Chicago: University of Chicago Press, 1989. Payne, Philip B. “Vaticanus Distigme-Obelos Symbols Marking Added Text, Including 1 Corinthians 14.34–5.” NTS 63/4 (2017): 604–25. Peebles B. M. “The Bibliotheca Albana Urbinas as Represented in the Library of the Catholic University of America.” Pages 327–51 in Didascaliae: Studies in Honor of Anselm M. Albareda. Edited by S. Prete. New York: B. M. Rosenthal, 1961. Peikola, Matti. “Manuscript Paratexts in the Making: British Library MS Harley 6333 as a Liturgical Compilation.” Pages 44–67 in Discovering the Riches of the Word: Religious Reading in Late Medieval and Early Modern Europe. Edited by S. Corbellini, M. Hoogvliet, and B. Ramakers Leiden: Brill, 2015. Perosa, A. ed. Lorenzo Valla. Collatio Novi Testamenti. Firenze: Sansoni 1970. Peters, Ronald D. The Greek Article: A Functional Grammar of ὁ-items in the Greek New Testament with Special Emphasis on the Greek Article. Leiden: Brill, 2014. Peterson, Jacob W. “An Updated Correction List for Chester Beatty BP II—P.Mich. Inv. 6238 (Gregory–Aland Papyrus 46 [𝔓46]).” BASP 56 (2019). Petersen, William L. “What Text Can New Testament Textual Criticism Ultimately Reach?” Pages 136–52 in New Testament Textual Criticism, Exegesis and Church History: A Discussion of Methods. Edited by B. Aland and J. Delobel. Kampen: Kok Pharos, 1994. Petzer, Jacobus H. “The History of the New Testament – Its Reconstruction, Significance and Use in New Testament Textual Criticism.” Pages 11–36 in New Testament Textual Criticism, Exegesis and Church History: A Discussion of Methods. Edited by B. Aland and J. Delobel. Kampen: Kok Pharos, 1994. Pfeiffer, Rudolf. History of Classical Scholarship: From the Beginnings to the End of the Hellenistic Age. Oxford: Clarendon Press, 1968. Pickering, Stuart R. “The Significance of Non-Continuous New Testament Textual Materials in Papyri.” Pages 121–41 in Studies in the Early Text of the Gospels and Acts. The Papers of the First Birmingham Colloquium on the Textual Criticism of the New Testament. Edited by D. G. K. Taylor. Birmingham: University of Birmingham Press, 1999. Pickering, Wilbur. The Identity of the New Testament Text. Revised edition. Nashville: Thomas Nelson, 1980. Pickering, Wilbur N. The Identity of the New Testament Text III. Eugene: Wipf & Stock, 2012. Pierazzo, Elena. Digital Scholarly Editing: Theories, Models and Methods. Farnham: Ashgate, 2015. Pinney, T. ed. The Letters of Rudyard Kipling. Volume 3 1900–10. London: Palgrave Macmillan, 1996. Pintaudi, R. “Una nota a PSI I 65.” ZPE 56 (1984): 137–38. Platt, Thomas Pell, ed. Ethiopic New Testament. London: British Foreign Bible Society, 1830. Porter, Calvin L. “A Textual Analysis of the Earliest Manuscripts of the Gospel of John.” PhD diss., Duke University, 1961. Porter, Calvin L. “Papyrus Bodmer XV (P75) and the Text of Codex Vaticanus.” JBL 81/4 (1962): 363–76. Porter, Stanley E. and Jason C. Robinson. Hermeneutics: An Introduction to Interpretive Theory. Grand Rapids: Eerdmans, 2011.

Bibliography

493

Porter, Stanley E. and Andrew W. Pitts. Fundamentals of New Testament Textual Criticism. Grand Rapids: Eerdmans, 2015. Porter, Stanley E. “Why So Many Holes in the Papyrological Evidence for the Greek New Testament?” Pages 167–86 in The Bible as Book: The Transmission of the Greek Text. Edited by O. O’Sullivan and S. McKendrick. New Castle: Oak Knoll, 2003. Porter, Stanley E. “Textual Criticism in the Light of Diverse Textual Evidence for the Greek New Testament: An Expanded Proposal.” Pages 305–37 in New Testament Manuscripts: Their Text and Their World. TENTS 2. Edited by T. J. Kraus and T. Nicklas. Leiden: Brill, 2006. Porter, Stanley E. How We Got the New Testament: Text, Transmission, Translation. Grand Rapids: Baker, 2013. Porter, Stanley E. “What Do We Know and How Do We Know It? Reconstructing Early Christianity from Its Manuscripts.” Pages 41–70 in Christian Origins and Greco-Roman Culture: Social and Literary Contexts for the New Testament. Edited by S. E. Porter and A. W. Pitts. Leiden: Brill, 2013. Porter, Stanley E. The Letter to the Romans: A Linguistic and Literary Commentary. NTM 37. Sheffield: Sheffield Phoenix, 2015. Powerscourt, Sheila. Sun Too Fast. London: Geoffrey Bles, 1974. Preisigke, Friedrich. Wörterbuch der griechischen Papyrusurkunden, mit Einschluß der griechischen Inschriften, Aufschriften, Ostraka, Mumienschilder usw. aus Ägypten. 2 volumes. Berlin: Selbstverlag der Erben, 1925 and 1927. Preisigke, Friedrich. Besondere Wörterliste. Wörterbuch der griechischen Papyrusurkunden, mit Einschluß der griechischen Inschriften, Aufschriften, Ostraka, Mumienschilder usw. aus Ägypten 3. Edited by E. Kießling. Berlin: Selbstverlag der Erben, 1931. Prigent, P. and R. Stehly, eds. “Les Fragments du De Apocalypsi d’Hippolyte.” Theologische Zeitschrift 29 (1973): 313–33. Puar, Jaspir K. The Right to Maim: Debility, Capacity, Disability. Durham: Duke University Press, 2017. Rahlfs, A. Septuaginta, Vol. X. Psalmi cum Odis. Göttingen: Vandenhoeck & Ruprecht, 1931. Rabinow, Paul, ed. The Foucault Reader. New York: Pantheon, 1984. Ramsay, Stephen. “Hard Constraints: Designing Software.” Pages 449–57 in A New Companion to Digital Humanities. Edited by S. Schreibman, R. Siemens, and J. Unsworth. Oxford: John Wiley, 2016. Raskin, J. “Intuitive Equals Familiar.” Communications of the ACM 37/9 (1994): 17–8. Reeve, Michael D. “Lucretius from the 1460s to the 17th Century: Seven Questions of Attribution.” Aev 80 (2006): 165–84. Reuss, Eduard. Bibliotheca Novi Testamenti Graeci cuius editiones ab initio typographiae ad nostram aetatem impressas quotquot reperiri potuerunt. Braunschweig: Schwetschke, 1872. Reynolds, L. D. and N. G. Wilson. Scribes and Scholars: A Guide to the Transmission of Greek and Latin Literature. 4th edition. Oxford: Oxford University Press, 2013. Richard, M. ed. Hippolytus: Commentarius in Danielem. GCS NF 7. Berlin: De Gruyter, 2000. Richards, W. Larry. “A Critique of a New Testament Text-Critical Methodology: The Claremont Profile Method.” JBL 96 (1977): 555–66. Richards, W. Larry. The Classification of the Greek Manuscripts of the Johannine Epistles. Missoula: Scholars, 1977. Richards, W. Larry. “An Analysis of Aland’s Teststellen in 1 John.” NTS 44 (1998): 26–44.

494

Bibliography

Richter, Siegfried G. “Der Wegweiser zum Mittelägyptischen in der ECM.” Pages 221–27 in Novum Testamentum Graecum: Editio Critica Maior III: The Acts of the Apostles; Part 3: Studies. Edited by H. Strutwolf et al. Stuttgart: Deutsche Bibelgesellschaft, 2017. Roberts, C. H. Manuscript, Society and Belief in Early Christian Egypt. London: Oxford University Press, 1979. Robinson, J. Armitage. Euthaliana: Studies of Euthalius Codex H of the Pauline Epistles and the Armenian Version, with an Appendix Containing a Collation of the Eton MS of the Pseudo-Athanasian Synopsis. Cambridge: Cambridge University Press, 1895. Robinson, Maurice A. “Von Soden’s Folly.” Unpublished Ph.D. Seminar Paper, 1977. Robinson, Maurice A. and William G. Pierpont. The New Testament in the Original Greek: Byzantine Text-Form. Southborough: Chilton, 2005. Robinson, Peter. “The Collation and Textual Criticism of Icelandic Manuscripts (1): Collation.” Literary and Linguistics Computing 4/2 (1989): 99–105. Robinson, Peter. Collate: Interactive Collation of Large Textual Traditions, Version 2. Computer Program distributed by the Oxford University Centre for Humanities Computing, 1994. Robinson, Peter. “A History of Collate.” Scholarly Digital Editions 29 September 2014, http://scholarlydigitaleditions.blogspot.co.uk/2014/09/the-history-of-collate.html. Robinson. Peter. “The Background to the Textual Communities Project.” http://www.textualcommunities.usask.ca/web/textual-community/wiki/-/wiki/Main/The+background+to+the +Textual+Communities+Project. Robinson, Peter. “Some Principles for Making Collaborative Scholarly Editions in Digital Form.” Digital Humanities Quarterly 11/2 (2017). Rochlitz, Rainer. The Disenchantment of Art: The Philosophy of Walter Benjamin. Translated by J. M. Todd. New York: Guilford Press, 1992. Rogers, Molly. Delia’s Tears: Race, Science and Photography in Nineteenth-Century America. New Haven: Yale University Press, 2010. Römer, Cornelia Eva. “Ostraka mit christlichen Texten aus der Sammlung Flinders Petrie.” ZPE 145 (2003): 183–201. Römer, Cornelia Eva. “Das zweisprachige Archiv aus der Sammlung Flinders Petrie.” ZPE 164 (2008): 53–62. Ross, J. M. “The Ending of the Apocalypse.” Pages 338–44 in Studies in New Testament Language and Text. NTSup 44. Edited by J. K. Elliott. Leiden: Brill, 1976. Rowlandson, Jane. Women and Society in Greek and Roman Egypt: A Sourcebook. Cambridge: Cambridge University Press, 1998. Royse, James R. “Von Soden’s Accuracy.” JTS 30 (1979): 166–71. Royce, James R. “Scribal Habits in Early Greek New Testament Papyri.” ThD diss., Graduate Theological Union, 1981. Royce, James R. Scribal Habits in Early Greek New Testament Papyri. Leiden: Brill, 2008. Rupp, E. G. Hort and the Cambridge Tradition: An Inaugural Lecture. Cambridge: Cambridge University Press, 1970. Ruwet, Nicolas. “Linguistics and Poetics.” Pages 296–313 in The Structuralist Controversy: The Languages of Criticism and the Sciences of Man. Edited by R. Macksey and E. Donato. Baltimore: Johns Hopkins University Press, 1972. Salmon, George. Some Thoughts on the Textual Criticism of the New Testament. London: Murray, 1897. Sanday, William. “Baron Hermann von Soden.” JTS 15 (1914): 305–306. Sanders, Henry A. “Hoskier’s Genesis of the Versions.” American Journal of Philology 33 (1912): 30–42.

Bibliography

495

Sanders, Henry A. The New Testament Manuscripts in the Freer Collection. The Washington Manuscript of the Four Gospels. New York: Macmillan, 1912 Sanders, Henry A. Facsimile of the Washington Manuscript of the Gospels in the Freer Collection. 2 volumes. New York: Macmillan, 1912. Sanders, Henry A. “The Egyptian Text of the Four Gospels and Acts.” HTR 27/2 (1933): 77– 98. Sanders, Henry A. “The Beatty Papyrus of Revelation and Hoskier’s Edition.” JBL 53/4 (1934): 371–80. Sanders, Henry A. A Third-Century Papyrus Codex of the Epistles of Paul. University of Michigan Studies Humanistic Series 38. Ann Arbor: University of Michigan Press, 1935. Sanzo, Joseph E. Scriptural Incipits on Amulets from Late Antique Egypt. STAC 84. Tübingen: Mohr Siebeck, 2014. Sanzo, Joseph E. “Wrapped up in the Bible: The Multifaceted Ritual on a Late Antique Amulet (P. Oxy. VIII 1077).” JECS 24 (2016): 569–97. Schaefer, Donavan. “The Promise of Affect: The Politics of the Event in Ahmed’s The Promise of Happiness and Berlant’s Cruel Optimism.” Theory & Event 16 (2013). Schaefer, Donavan. Religious Affects: Animality, Evolution, and Power. Durham: Duke University Press, 2015. Schaper, Joachim. “The Origin and Purpose of the Fifth Column of the Hexapla.” Pages 3–15 in Origen’s Hexapla and Fragments: Papers Presented at the Rich Seminar on the Hexapla, Oxford Centre for Hebrew and Jewish Studies, 25th–3rd August 1994. TSAJ 58. Edited by A. Salvesen. Tübingen: Mohr Siebeck, 1998. Scherbenske, Eric W. Canonizing Paul: Ancient Editorial Practice and the Corpus Paulinum. Oxford: Oxford University Press, 2013. Schironi, Francesca. “Plato at Alexandria: Aristophanes, Aristarchus, and the ‘Philological Tradition’ of a Philosopher.” The Classical Quarterly 55/2 (2005): 423–34. Schironi, Francesca. ΤΟ ΜΕΓΑ ΒΙΒΛΙΟΝ: Book-ends, End-titles, and Coronides in Papyri with Hexametric Poetry. Durham: American Society of Papyrologists, 2010. Schironi, Francesca. “The Ambiguity of Signs: Critical Σηµεια from Zenodotus to Origen.” Pages 87–112 in Homer and the Bible in the Eyes of Ancient Interpreters. Edited by M. R. Niehoff. Leiden: Brill, 2012. Schironi, Francesca. “P.Grenf. 1.5, Origen, and the Scriptorium of Caesarea.” BASP 52 (2015): 181–223. Schmid, Josef. “Untersuchungen zur Geschichte des griechischen Apokalypsetextes. Der KText.” Biblica 17 (1936): 11–44, 167–201, 273–93, 429–60. Schmid, Josef. Studien zur Geschichte des griechischen Apokalypse-Textes. Munich: Karl Zink, 1955–1956. Schmid, Josef. Studies in the History of the Greek Text of the Apocalypse: The Ancient Stems. Edited and translated by J. Hernández Jr., G. V. Allen, and D. Müller. Atlanta: SBL, 2018. Schmid, Ulrich. “Scribes and Variants – Sociology and Typology.” Pages 1–23 in Textual Variation: Theological and Social Tendencies? Edited by H. A. G. Houghton and D. C. Parker. Piscataway: Gorgias, 2008. Schmid, Ulrich. “Transmitting the New Testament Online.” Pages 189–206 in Text Comparison and Digital Creativity: The Production of Presence and Meaning in Digital Text Scholarship. Edited by W. Th. van Peursen, E. Thoutenhoofd, and A. van der Weel. Leiden: Brill, 2010. Schmid, Ulrich. “Die Apokalypse, überliefert mit anderen neutestamentlichen Schriften – eaprHandschriften.” Pages 421–41 in Studien zum Text der Apokalypse. ANTF 47. Edited by M. Sigismund, M. Karrer, and U. Schmid. Berlin: De Gruyter, 2015.

496

Bibliography

Schmid, Ulrich. “Editing the Apocalypse in the 21st Century.” Pages 231–40 in Book of Seven Seals: The Peculiarity of Revelation, its Manuscripts, Attestation, and Transmission. WUNT 363. Edited by T. Kraus and M. Sommer. Tübingen: Mohr Siebeck, 2016. Schmidt, Carl. “Christliche Texte.” APF 2 (1903): 381–85. Schmidt, Desmond. “Towards an Interoperable Digital Scholarly Edition.” Journal of the Text Encoding Initiative 7 (2014). Schoene, Alfred. Die Weltchronik des Eusebius in ihrer Bearbeitung durch Heironymus. Berlin: Weidmann, 1900. Schröder, Katharina D. “Hinweise zur Verzeichnung der koptischen Versionen.” Pages 72–80 in Novum Testamentum Graecum: Editio Critica Maior III: The Acts of the Apostles; Part 3: Studies. Edited by H. Strutwolf et al. Stuttgart: Deutsche Bibelgesellschaft, 2017. Scott, David. Conscripts of Modernity: The Tragedy of Colonial Enlightenment. Durham: Duke University Press, 2004. Scott, Joan. “Fantasy Echo: History and the Construction of Identity.” Critical Inquiry 27/2 (2001): 284–304. Scrivener, F. H. A. A Plain Introduction to the Criticism of the New Testament for the Use of Biblical Students. 1st edition. Cambridge: Deighton, Bell, and co., 1861. Scrivener, F. H. A. Bezae Codex Cantabrigiensis, Being an Exact Copy, in Ordinary Type, of the Celebrated Uncial Graeco-Latin Manuscript of the Four Gospels and Acts of the Apostles, Written Early in the Sixth Century, and Presented to the University of Cambridge by Theodore Beza, A.D. 1581. Edited With a Critical Introduction, Annotations, and Facsimiles. Cambridge: Deighton, Bell, and co., 1864. Scrivener, F. H. A. Six Lectures on the Text of the New Testament and the Ancient Manuscripts Which Contain It, Chiefly Addressed to Those Who Do Not Read Greek. London: George Bell and Sons, 1875. Scrivener, F. H. A. The Parallel New Testament: Greek and English. Cambridge: Cambridge University Press, 1882. Scrivener, F. H. A. Novum Testamentum Textûs Stephanici...London: George Bell, 1887. Scrivener, F. H. A. A Plain Introduction to the Criticism of the New Testament. 4th edition. London: George Bell and Sons, 1894. Seaward, Louise. “The Crowd, the Volunteers and the Supertranscribers. Building and Supporting an Online User Community for the Bentham Edition.” Presentation given at the co:op READ Convention Marburg 2016. Available online at https://coop.hypotheses.org/224 [accessed 20 August 2018]. Selden, Raman et al. A Reader’s Guide to Contemporary Literary Theory. 4th edition. London: Prentice Hall, 1997. Shackelton Bailey, D. R. “Introduction.” Pages 3–26 in Epistulae ad Familiares. Vol. 1. Edited by D. R. Shackelton Bailey. Cambridge: Cambridge University Press, 1977. Shillingsburg, Peter L. “Text as Matter, Concept and Action.” Studies in Bibliography 44 (1991): 31–82. Shillingsburg, Peter L. Scholarly Editing in the Computer Age: Theory and Practice. 3rd edition. Ann Arbor: University of Michigan Press, 1996. Sigismund, Marcus. “Die griechisch-lateinischen Apk-Bilinguen (und andere mehrsprachige Manuskripte der Apk).” Pages 315–64 in Studien zum Text der Apokalypse. ANTF 47. Edited by M. Sigismund, M. Karrer, and U. Schmid. Berlin: De Gruyter, 2015. Sigismund, Marcus. “Die neue Edition der Johannespokalypse: Stand der Arbeit.” Pages 3–17 in Studien zum Text der Apokalypse II. ANTF 50. Edited by M. Sigismund and D. Müller. Berlin: De Gruyter, 2017.

Bibliography

497

Sigismund, Marcus. “Quaestiones Aretae I.” Pages 411–32 in Studien zum Text der Apokalypse II. ANTF 50. Edited by M. Sigismund and D. Müller. Berlin: De Gruyter, 2017. Sinaites, Justin. “Newly Recovered Manuscripts of the Scriptures from Saint Catherine’s Monastery, Sinai.” Restoration Quarterly 59 (2017): 5–24. Sion, Tasfa, ed. Novum Testamentum cum epistola Pauli ad Hebreos tantum... 2 vols. Rome: V. and L. Dorici, 1548. Skeat, T. C. “The Codex Sinaiticus and the Codex Vaticanus and Constantine [1999].” Pages 193–237 in The Collected Biblical Writings of T.C. Skeat. NTSup 113. Edited by J. K. Elliott. Leiden: Brill, 2004. Smith, Courtney Weiss. Empiricist Devotions: Science, Religion, and Poetry in Early Eighteenth-Century England. Charlottesville: University of Virginia Press, 2016. Smith, Helen and Louise Wilson, eds. Renaissance Paratexts. Cambridge: Cambridge University Press, 2011. Smith, Jonathan Z. Drudgery Divine: On the Comparison of Early Christianities and the Religions of Late Antiquity. Chicago: University of Chicago Press, 1990. Smith, Kimberly K. “Mere Nostalgia: Notes on a Progressive Paratheory.” Rhetoric and Public Affairs 3/4 (2000): 505–27. Smith, Paul. Discerning the Subject. Minneapolis: University of Minnesota Press, 1988. Smith, W. Andrew. A Study of the Gospels in Codex Alexandrinus: Codicology, Palaeography, and Scribal Hands. NTTSD 48. Leiden: Brill, 2014. Soule, Gardner. “The Machine that Indexed the Bible.” Popular Science 169/5 (1956): 173– 75, 242, 246. Spielmann, Kent. “Participant Reference and Definite Article in John.” Journal of Translation and Textlinguistics 7 (1995): 45–85. Stephanus, Robert. Τῆς καινῆς διαθήκης ἅπαντα. Novum Testamentum. Ex bibliotheca regia. 1st edition. Paris: Robertus Stephanus, 1546. Stephanus, Robert. Τῆς καινῆς διαθήκης ἅπαντα. Novum Testamentum. Ex bibliotheca regia. 2nd edition. Paris: Robertus Stephanus, 1549. Stephanus, Robert. Τῆς καινῆς διαθήκης ἅπαντα. Novum Iesu Christi D. N. Testamentum. Ex bibliotheca regia. 3rd edition. Paris: Robertus Stephanus, 1550. Streeter, B. H. The Four Gospels: A Study of Origins. 4 th edition. London: Macmillan, 1930. Streeter, B. H. “Codices 157, 1071 and the Caesarean Text.” Pages 149–50 in Quantulacumque: Studies presented to Kirsopp Lake by Pupils, Colleagues, and Friends. Edited by R. P. Casey, S. Lake, and A. K. Lake. London: Christophers, 1937. Suggit, John N. Oecumenius: Commentary on the Apocalypse. The Fathers of the Church 112. Washington, D.C.: The Catholic University of America Press, 2006. Suleiman, Susan R. and Inge Crosman, eds. The Reader in the Text: Essays on Audience and Interpretation. Princeton: Princeton University Press, 1980. Tanselle, G. Thomas. A Rationale of Textual Criticism. Philadelphia: University of Pennsylvania Press, 1989. Tasker, R. V. G. “The Text of the ‘Corpus Paulinum.’” NTS 1 (1954–1955): 180–91. Taylor, C. Hebrew-Greek Cairo Genizah Palimpsests from the Taylor-Schechter Collection: Including a Fragment of the Twenty-Second Psalm according to Origen’s Hexapla. Cambridge: Cambridge University Press, 1900. Thackeray, Henry St. John. “Review of Codex B and its Allies: A Study and an Indictment (by Herman C. Hoskier).” The Journal of Hellenic Studies 35 (1915): 275–76. Thomsen, Mads Rosendahl et al., eds. Literature: An Introduction to Theory and Analysis. London: Bloomsbury, 2017. Thorpe, James. Principles of Textual Criticism. San Marino: Huntington Library, 1972.

498

Bibliography

Timpanaro, Sebastiano. The Genesis of Lachmann’s Method. Translated by G. W. Most. London: The University of Chicago Press, 2005. Tischendorf, Constantin. Monumenta sacra inedita sive reliquiae antiquissimae textus Novi Testamenti graeci ex novem plus mille annorum codibus per Europam dispersis eruit atque edidit… Leipzig: Tauchnitz, 1846. Tischendorf, Constantin. Novum Testamentum Graece. Ad antiquissimos testes denuo recensuit, apparatum criticum omni studio perfectum apposuit, commentationem isagogicam praetexuit…Editio octava critica maior. Leipzig: Giesecke & Devrient, 1869. Toda, Satoshi. “The Eusebian Canons: Their Implications and Potential.” Pages 27–43 in Early Readers, Scholars and Editors of the New Testament: Papers from the Eighth Birmingham Colloquium on the Textual Criticism of the New Testament. Edited by H. A. G. Houghton. Piscataway: Gorgias, 2013. Tompkins, Jane, ed. Reader-Response Criticism: From Formalism to Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980. Tov, Emanuel. “Paratextual Elements in the Masoretic Manuscripts of the Bible Compared with the Qumran Evidence.” Pages 73–83 in Antikes Judentum und Frühes Christentum. Festschrift für Hartmut Stegemann zum 65. Geburtstag. BZNW 97. Edited by B. Kollmann, W. Reinbold, and A. Steudel. Berlin: De Gruyter, 1999. Tregelles, S. P. An Account of the Printed Text of the Greek New Testament; with Remarks on its Revision upon Critical Principles. Together with a Collation of the Critical Texts of Griesbach, Scholz, Lachmann, and Tischendorf, with that in Common Use. London: Samuel Bagster and Sons, 1854. Tregelles, S. P. The Greek New Testament. Edited from Ancient Authorities, with Their Various Readings in Full, and the Latin Version of Jerome. Part II. Gospels of St. Luke and St. John. London: Bagster, 1861. Tregelles, S. P. “A Few Notes on Codex Reuchlini of the Apocalypse, together with a collation of it’s [sic] text with the common editions.” Pages 1–16 in Handschriftliche Funde 2. Neue Studien über den Codex Reuchlins und neue Textgeschichtliche Aufschlüsse über die Apokalypse aus den Bibliotheken in München, Wien und Rom. Edited by F. Delitzsch. Leipzig: Dörffling & Franke, 1862. Treu, K. “Griechisch-Koptisch Bilinguen des Neuen Testaments.” Pages 95–123 in Koptologische Studien in der DDR. Wissenschaftliche Zeitschrift der Martin-Luther-Universität. Edited by J. Irmscher. Halle: Martin-Luther-Universität, 1965. Trovato, P. Everything you Always Wanted to Know about Lachmann’s Method: A Non-Standard Handbook of Genealogical Textual Criticism in the Age of Post-Structuralism, Cladistics, and Copy-Text. 2nd edition. Padova: libreriauniversitaria.it, 2017. Tuckett, Christopher. “The Early Text of Acts.” Pages 157–74 in The Early Text of the New Testament. Edited by C. E. Hill and M. J. Kruger. Oxford: Oxford University Press, 2012. Turner, Eric G. Greek Papyri: An Introduction. Oxford: Oxford University Press, 1980. Turner, James. Philology: The Forgotten Origins of the Modern Humanities. Princeton: Princeton University Press, 2014. Tyson, Lois. Critical Theory Today: A User-Friendly Guide. 2nd edition. London: Routledge, 2006. Tzamalikos, P. An Ancient Commentary on the Book of Revelation: A Critical Edition of the Scholia in Apocalypsin. Cambridge: Cambridge University Press, 2013. Uhlig, Siegbert and Helge Maehlum. Novum Testamentum Aethiopice: die Gefangenschaftsbriefe. Äthiopistische Forschungen 33. Stuttgart: Franz Steiner, 1993.

Bibliography

499

Vaganay, Léon and Christian-Bernard Amphoux. An Introduction to New Testament Textual Criticism. Translated bu J. Heimerdinger. 2nd edition. Cambridge: Cambridge University Press, 1991. Vanhoutte, Edward. “The Gates of Hell: History and Definition of Digital | Humanities | Computing.” Pages 119–56 in Defining Digital Humanities: A Reader. Edited by M. Terras, J. Nyhan, and E. Vanhoutte. Farnham: Ashgate, 2013. Vansittart, A. A. “On Two Triple Readings in the New Testament.” Journal of Philology 3/6 (1871): 357–59. van Zundert, Joris. “If You Build It, Will We Come? Large Scale Digital Infrastructures as a Dead End for Digital Humanities.” Historical Social Research/Historische Sozialforschung 37/3 (2012): 165–86. Vervliet, Hendrik D. L. The Palaeotypography of the French Renaissance. Selected Papers on Sixteenth-Century Typefaces. Leiden: Brill, 2008. von Dobschütz, Ernst. Eberhard Nestle’s Einführung in das Griechische Neue Testament. 4th edition. Göttingen: Vandenhoeck & Ruprecht, 1923. von Dobschütz, Ernst. “Zur Liste der Neutestamentlichen Handschriften.” ZNW 23 (1924): 248–65. von Dobschütz, Ernst. “Zur Liste der Neutestamentlichen Handschriften II.” ZNW 25 (1926): 299–306. von Dobschütz, Ernst. “Zur Liste der NTlichen Handschriften III.” ZNW 27 (1928): 216–22. von Dobschütz, Ernst. “Zur Liste der NTlichen Handschriften IV,” ZNW 32 (1933): 185–206. von Soden, Hermann. Die Schriften des Neuen Testaments in ihrer ältesten erreichbaren Textgestalt hergestellt auf Grund ihrer Textgeschichte. Two parts in 4 volumes. 2nd edition. Göttingen: Vandenhoeck & Ruprecht, 1911–13. Vööbus, Arthus. Early Versions of the New Testament. Stockholm: Estonian Theological Society in Exile, 1954. Wachtel, Klaus and David C. Parker. “The Joint IGNTP/INTF Editio Critica Maior of the Gospel of John: Its Goals and Their Significance for New Testament Scholarship.” Paper presented at the SNTS meeting, Halle, 2005. http://epapers.bham.ac.uk/ 754/1/2005_ SNTS_WachtelParker.pdf [accessed 20 August 2018]. Wachtel, Klaus and Michael W. Holmes, eds. The Textual History of the Greek New Testament: Changing Views in Contemporary Research. Atlanta: SBL, 2011. Wachtel, Klaus and Klaus Witte. Das Neue Testament auf Papyrus II: Die Paulinischen Briefe, Teil 2: Gal, Eph, Phil, Kol, 1 u. 2 Thess, 1 u. 2 Tim, Tit, Phlm, Hebr. ANTF 22. Berlin: De Gruyter, 1994. Wachtel, Klaus. Der byzantinische Text der katholischen Briefe. ANTF 24. Berlin: De Gruyter, 1995. Wachtel, Klaus. “Editing the Greek New Testament on the Threshold of the Twenty-First Century.” Literary and Linguistic Computing 15/1 (2000): 43–50. Wachtel, Klaus. “Colwell Revisited: Grouping New Testament Manuscripts.” Pages 31–43 in The New Testament in Early Christianity: Proceedings of the Lille Colloquium, July 2000. Edited by C-.B. Amphoux and J. K. Elliott. Lausanne: Zèbre, 2003. Wachtel, Klaus. “Constructing Local Stemmata for the ECM of Acts: Examples.” TC: A Journal of Biblical Textual Criticism 20 (2015) http://rosetta.reltech.org/TC/ v20/TC-2015CBGM-examples.pdf [accessed 10 August 2018]. Wachtel, Klaus. “A Plea Against the Text-Type Concept.” Pages 137–148 in Novum Testamentum Graecum: Editio Critica Maior III: The Acts of the Apostles; Part 3: Studies. Edited by H. Strutwolf et al. Stuttgart: Deutsche Bibelgesellschaft, 2017.

500

Bibliography

Wachtel, Klaus. “Notes on the Text of the Acts of the Apostles.” Pages 28*–33* in Novum Testamentum Graecum: Editio Critica Maior III: The Acts of the Apostles; Part 1: Text; Chapter 1–14. Edited by H. Strutwolf et al. Stuttgart: Deutsche Bibelgesellschaft, 2017. Waite, Stephen V. F. “Computers and the Classics.” Computers and the Humanities 3/1 (1968): 25–9. Wallace, Daniel B. “Historical Revisionism and the Majority Text: The Cases of F. H. A. Scrivener and Herman C. Hoskier.” NTS 41 (1995): 280–85. Wallace, Daniel B. “The Majority Text Theory: History, Methods, and Critique.” Pages 711– 44 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Wallraff, M., S. S. Menchi, and K. von Greyerz, eds. Basel 1516: Erasmus’ Edition of the New Testament. SMHR 91. Tübingen: Mohr Siebeck, 2016. Wallraff, Martin. Kodex und Kanon: Das Buch im frühen Christentum. Berlin: De Gruyter, 2013. Wallraff, Martin. “The Canon Tables of the Psalms: An Unknown Work of Eusebius of Caesarea.” Dumbarton Oaks Papers 67 (2013): 1–14. Wallraff, Martin and Patrick Andrist, “Paratexts of the Bible.” Early Christianity 6 (2015): 237–43. Walton, Brian Bibliorum sacrorum tomus quintus: sive Novum D. N. Jesu Christi Testamentum. London: Thomas Roycroft, 1657. Walton, Brian. Ad Biblia sacra polyglotta appendix. In quo varii Tractatus, Annotationes, Lectiones variae, Hebr. Graec. Lat. Samarat. Chald. Syr. Arab. Aethiop. Pers. cum Indicibus, etc. quae tomum sextum constituent. London: Roycroft, 1657. Ward, Aengus. “The Estoria de Espanna Digital: collating medieval prose – challenges…and more challenges.” Digital Philology (forthcoming). Wasserman, Tommy. “The Patmos Family of New Testament MSS and Its Allies in the Pericope of the Adulteress and Beyond.” TC: A Journal of Biblical Textual Criticism 7 (2002): pars. 1–59. Wasserman, Tommy. The Epistle of Jude: Its Text and Transmission. Stockholm: Almqvist and Wiksell, 2006. Wasserman, Tommy. “Theological Creativity and Scribal Solutions in Jude.” Pages 75–84 in Textual Variation: Theological and Social Tendencies? Papers from the Fifth Birmingham Colloquium on the Textual Criticism of the New Testament. Edited by H. A. G. Houghton and D. C. Parker. Piscataway: Gorgias, 2008. Wasserman, Tommy. “Historical and Philological Correlations and the CBGM as Applied to Mark 1:1.” TC: A Journal of Biblical Textual Criticism 20 (2015): 1–11. Wasserman, Tommy. “The Coherence Based Genealogical Method as a Tool for Explaining Textual Changes in the Greek New Testament.” NovT 57/2 (2015): 206–18. Wasserman, Tommy. “Was There an Alexandrian Recension of the Living Text of the Gospels?” Pages 1–22 in Liturgy and the Living Text of the New Testament. Edited by H. A. G. Houghton. Piscataway: Gorgias, 2018. Wasserman, Tommy, and Peter J. Gurry. A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method. Atlanta: SBL, 2017. Watson, Francis. The Fourfold Gospel: A Theological Reading of the New Testament Portraits of Jesus. Grand Rapids: Baker, 2016. Weber, R. et al., eds. Biblia Sacra iuxta Vulgatam versionem. 5th edition. Stuttgart: Deutsche Bibelgesellschaft, 2007. Westcott, Arthur. Life and Letters of Brooke Foss Westcott. London: Macmillan, 1903. Westcott, B. F. Lessons from Work. London: Macmillan, 1901.

Bibliography

501

Westcott, B. F. and F. J. A. Hort. The Gospels. Cambridge: Cambridge University Press, 1871. Westcott, B. F. and F. J. A. Hort. The New Testament in the Original Greek. London: Macmillan, 1881. Westcott, B. F. and F. J. A. Hort. The New Testament in the Original Greek: Introduction, Appendix. 2nd edition. London: Macmillan, 1896. Westcott, B. F. and F. J. A. Hort. The Greek New Testament: With Dictionary. Peabody: Hendrickson, 2007. Wettstein, Johann Jakob. Prolegomena ad Novi Testamenti Graeci editionem accuratissimam, e vetustissimis codd. MSS. denuo procurandam; in quibus agitur de codd. MSS. N. Testamenti, Scriptoribus Graecis qui N. Testamento usi sunt, versionibus veteribus, editionibus prioribus, et claris interpretibus; et proponuntur animadversiones et cautiones ad examen variarum lectionum N. T. necessariae. Amsterdam: Wettstein & Smith, 1730. Wettstein, Johann Jakob. Ἡ Καινὴ Διαθήκη. Novum Testamentum Graecum editionis receptae cum lectionibus variantibus Codicum MSS., Editionum aliarum, Versionum et Patrum nec non commentario pleniore ex Scriptoribus veteribus Hebraeis, Graecis et Latinis historiam et vim verborum illustrante. Tomus I. Continens quatuor Evangelia. Amsterdam: Dommeriana, 1751. White, H. J. Novum Testamentum Domini Nostri Iesu Christi Latine. 1. Evangelia. Oxford: Clarendon, 1889–1898. White, John L. The Form and Function of the Body of the Greek Letter: A Study of the LetterBody in the Non-Literary Papyri and in Paul the Apostle. SBLDS 2. Missoula: SBL, 1972. Wilcken, Ulrich. “Bibliographische Notizen und Mitteilungen.” APF 2 (1903): 161–80. Willard, Louis Charles. A Critical Study of the Euthalian Apparatus. Berlin: De Gruyter, 2009. Williams, James. Understanding Poststructuralism. Stocksfield: Acumen, 2005. Williams, P. J. “The Syriac Versions of the New Testament.” Pages 143–66 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013. Williams, P. J. “Ehrman’s Equivocation and the Inerrancy of the Original Text.” Pages 389– 406 in The Enduring Authority of the Christian Scriptures. Edited by D. A. Carson. Grand Rapids: Eerdmans, 2016. Williams, P. J. “The Greek New Testament, Produced at Tyndale House.” Early Christianity 8/2 (2017): 277–81. Wilson, A. J. The Life and Times of Sir Alfred Chester Beatty. London: Cadogan, 1985. Winter, Thomas Nelson. “Roberto Busa, S.J., and the Invention of the Machine-Generated Concordance.” The Classical Bulletin 75/1 (1999): 3–20. Wisse, Frederik. The Profile Method for the Classification and Evaluation of Manuscript Evidence as Applied to the Continuous Greek Text of the Gospel of Luke. Grand Rapids: Eerdmans, 1982. Wolfreys, Julian, ed. The Continuum Encyclopedia of Modern Criticism and Theory. New York: Continuum, 2002. Wordsworth, John. Old-Latin Biblical Texts: No. I. The Gospels according to St Matthew. From the St Germain MS. (g1), now numbered Lat. 11553 in the National Library at Paris. Oxford: Clarendon, 1883. Zacagnius, Laurentius Alexander. Collectanea Monumentorum Veterum Ecclesiae Graecae ac Latinae Quae Hactenus in Vaticana Bibliotheca Delituerunt. Tomus Primus. Rome: Typis Sacra Congreg. de Propag. Fide, 1698. Zuntz, Günther. The Ancestry of the Harklean New Testament. The British Academy Supplemental Papers VII. London: British Academy, 1945. Zuntz, Günther. “Euthalius = Euzoius?” VC 7/1 (1953): 16–22.

502

Bibliography

Zuntz, Günther. The Text of the Epistles: A Disquisition upon the Corpus Paulinum. The Schweich Lectures of the British Academy, 1946. Oxford: Oxford University Press, 1953. Zuurmond, Rochus. Novum Testamentum Aethiopice: The Synoptic Gospels. Part 1: General Introduction and Edition of the Gospel of Mark. Äthiopistische Forschungen 27. Stuttgart: Franz Steiner, 1989. Zuurmond, Rochus, revised by Curt Niccum. “The Ethiopic Version of the New Testament.” Pages 231–52 in The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis. Edited by B. D. Ehrman and M. W. Holmes. Leiden: Brill, 2013.

Ancient Sources Index Old Testament Genesis 1:1 1:4

35 206

Exodus 3:14

54

Psalms 68 90

300 293–295, 300

90:1 90:1–2 90:13 117:26 117:27

294 295 295 291 291

1 Esdras 8:38 9:29

236 236

New Testament Matthew 1:1 1:18 1:18–2:7 1:19 1:22 1:24 2:1 2:4 2:19–4:17 4:23 4:23–24 4:24 4:24–5:29 5:7 5:16 5:19 5:21 5:22 5:28 5:29 6:9 6:9–13

294 249–250, 253 253, 255 257 257 257 257 255 251 298 297 2866:9 253, 255 257–258 257 254 257 81 257 257 294 298

6:11–13 6:12 6:13 7:1–2 8:2–9:8 8:24 9:10 9:35 10:29 13:7–14:1 13:8 13:16 13:17 13:18 13:19 13:23 13:24 13:30 13:34 13:35 13:36 13:38 13:43

296 296 81 242 251 259 111 297 259 253, 255 257 253 253 257 257 257 257 257 257 257 257 257–258 257

504 13:44 13:47 13:51 13:55 13:57 13:58 14:1 14:24 14:35 16:9–17:17 16:13 16:13–18:31 16:17 16:18 16:19 16:21 16:22 16:26 16:27 16:28 17:5 17:12b 17:13 17:14 17:15 17:16 17:19 17:23 17:24 17:26 18:2–5 18:3 18:5 18:7 18:10 18:12–13 18:15 18:17 18:18 18:19 18:21 18:23 18:28 19:26–26:18 19:27 19:28 19:29 20:17 20:19

Ancient Sources Index 253 257 257 257 254 257 257 174 259 251 257 253, 255 316 261 254 258, 260 256 258 255, 262 258 256 259 259 259 259 254 256 255, 258 254 261 258 260 258 258 258, 260 254 260 260–261 254 258 254 258 258–259 253, 255 254 254, 260 261 258 259–260

20:23 20:25 20:26–27 20:28 20:30 20:34 21:2 21:3 21:4 21:15 21:29–30 21:33–41 21:37 21:44 21:46 22:6 22:10 22:16 22:19 22:29 22:30 22:31 22:34 22:36 22:42 23:8 23:10 23:14 23:18 23:25 23:27 23:28 23:39 24:6 24:10 24:14 24:21 24:26 24:29 24:30 24:31 24:33 24:34 24:38 24:39 24:40 24:41 24:42 24:48

258 254 254 259–260 249, 254, 261 258 258 258 261 259 260, 262 258 259 254 259 259 23 254 255, 262 254 254, 261 261 259 258 255 258, 261 258 259 261 260 258 258 258 257 261 253 258 254 259 258 258–259 258 258 258 258 258 259 259, 262 261

505

Ancient Sources Index 25:2–4 25:4 25:5 25:10 25:13 25:15 25:22–23 25:27 25:35 25:38 25:41 25:43 25:44 25:46 26:2 26:3 26:4 26:9 26:13 26:14 26:15 26:39–58 26:45–27:58 26:48 26:49 26:51 26:53 26:54 26:55 26:57 26:60 26:61 26:62 26:63 26:74 27:4 27:5 27:6 27:20 27:24 27:26 27:27 27:28–29 27:29 27:29–46 27:30 27:31 27:31–32 27:32

258 258 254 260–261 254 259 254 262 258 258 253, 259 258, 260 261 259, 261 261 249 259 260–261 258 261 259 251 253, 255 259 258 258 258–259 254 258 259 259 259 254 259 259 261 254 254, 259 259, 261 254 254 258 261 259, 261 251 254, 259 254, 258–259 290 259

27:35 27:38 27:41 27:45 27:46 27:47 27:48 27:49 27:49b 27:57

259 261 259 259 260 254 259, 262 259–260, 262 115 253

Mark 1:1 1:1–8 1:1–3:23 1:19 2:12–3:21 3:5 3:7 3:12 3:14 3:16 3:17 4:19–5:36 4:40 5:2–20 5:3 5:30 5:31 5:32 5:40–41 6:20 6:36–16:20 6:47 6:56 7:18 7:19 7:32–8:35 9:17–22 10:17–52 10:21 14:22–62 15:21 16:9–20

294 300 253 455 251 456 454 454–455 457 457 457 253 458–459 454 454 457 456 457 290 81 254 174 254 253 81 251 290 251 80 251 290 144

Luke 1:1 1:1–7 1:13

294 300 236

506

Ancient Sources Index

1:13–2:5 1:28 1:42 1:60 1:63 3:8–6:39 4:6 6:5 6:17–49 7:11–11:54 8:3 8:12–43 9:37 9:55 9:56 9:62 10:1–15 10:18 10:40–11:32 10:41 11:1b–2 11:2 12:13–16 12:45–14:18 14:3 14:5 14:10 15:25–16:15 17:7–19:10 17:8 19:8 19:38–22:35 22:40–71 22:43 22:43–44 22:44 22:59–23:14 23:12 23:34 23:35–44 23:38 24:8–13, 24–49 24:45

253 291 290 236 236 253 254 253 251 253 194 251 295 81 81 89 410 254 251 81 295 81 290 253 254 254 260 253 253 254 254 253 290 81 290 81 253 255 19, 81 251 81 251 112

John 1:1 1:1–9 1:1–17 1:3–4

294, 318 290 300 299

1:14–17 1:17 1:18 1:23 1:27 1:28 1:38 1:39 1:42 1:45 1:47 2:1–2 2:14 2:15 2:17 2:18–3:31 2:22 3:27 4:2 4:25 4:35 5:3–4 5:4 5:5 5:10 5:12–6:25 5:16 5:30 6:11 6:15 6:17 6:22 6:23 6:24 6:26 6:28 6:40 6:45 6:55 6:56 6:63 6:69 7:1 7:8 7:26 7:28–8:16 7:31

290 239 81, 318 294 318 236, 317–318, 320, 322, 325, 330 318 314, 317–318, 324, 330 317–318, 323, 330 239 239 295 16 330 330 251 330 236 327 330 330 81 426 330 330 253 330 330 330 315, 330 331 331 327, 428 239, 331 239 331 314, 331 331 331 314 331 331 147 331 331 251 331

507

Ancient Sources Index 7:33 7:39 7:40 7:53–8:11 8:7–10:3 8:9 8:29 8:31 8:38 8:49 8:53 9:8 9:9 9:17 10:7 10:14 10:16 10:30 10:34 10:40 10:41 11:19 11:21 11:31 11:41 11:54 12:1 12:3 12:7 12:17 12:17–13:6 12:19 12:21 12:31 12:35 13:24 13:27 13:30–31 14:7 14:22 14:30 15:1 15:7 15:8 16:3 16:16 17:3 17:11

239, 331 331 331 115, 144 253 314–315, 331 331 239 314, 331 315, 326, 327, 331 330 331 331 331 239 331 19 36 331 236 236 331 239 331 332 332 332 239 332 332 251 332 325 315, 326–327, 332 332 314, 332 332 315, 325, 332 327 315, 324, 332 332 315, 323, 325–327, 332 332 315, 332 315, 332 332 239 332

17:12 17:30 18:5 18:7 18:10 18:19–26 18:29 19:2 19:3 19:13 19:15–17 19:16–19 19:34 19:35 19:38 20:1–21:4 20:11 20:18 20:24 20:29 20:30–31 20:31 21:1–23 21:5 21:6 21:15 21:15–17 21:16 21:17 21:24 21:24–25 21:25

315, 332 332 239 239 258 290 315–316, 332 261 332 315–316, 326–327, 332 290 321 17, 259 116 239 250 315, 332 239 332 239 116 315, 325, 332 116 325 332 236, 315, 333 316–317 315, 317, 333 315, 317, 333 117 115–116 116

Acts 3:4 4:6 4:13 4:19 9:40 12:25 13:5 13:25 15:37 24:7 25:14 27:1 28:11

236 236 236 236 452 236 236 236 236 312 312–313 312–313 312–313

508 Romans 1:1 1:7 1:16–11:36 1:17 1:18–4:25 2:16 2:24 3:4 3:10 3:22 3:24 4:6 4:17 5:1 5:1–21 5:15 5:17 5:17–6:3 6:3 6:4 6:5–14 6:11 6:23 8:1 8:2 8:15–25 8:27–9:9 8:31 8:34 8:36 8:39 9:9–22 9:13 9:22–32 9:25 9:33 10:1–11 10:12–11:2 10:15 11:3–12 11:8 11:13–22 11:24–33 11:26 11:35–12:9 12:1–2 12:4 12:10–13:1

Ancient Sources Index

237 237 149 235 149 237 235 235 235 237 237 234 235 148–149 149 237 237 217 237 18 216 237 237 237 237 217 217 291 237 235 237 217 235 217 226 235 217 217 235 217 235 217 217 235 217 295 235 217

12:16 13:2–11 13:12–14:8 13:14 14:9–21 14:9–15:10 14:22–15:10 15:3 15:5 15:9 15:11–19 15:16 15:17 15:20–29 15:21 15:59–16:3 16:3 16:4–13 16:14–23 16:25 16:27

222, 226 217 217 237 217 226 217 235 237 235 217 237 237 217 235 217 237 217 217 237 237

1 Corinthians 1:4 1:4–14 1:14–23 1:22 1:24–2:2 1:27 2:3–11 2:6 2:11–3:5 2:15 3:2 3:6–15 3:16–4:3 4:4 4:4–10 4:11–19 4:20–5:7 5:5 5:8–6:3 5:12–6:7 6:2 6:4 6:4–12 6:13–7:3 6:14 7:4–12

274 218 218 275, 279 218 273 218 224 218 223 273 218 218 273 218 218 218 222, 224 218 242 224 275 218 218 224 218

509

Ancient Sources Index 7:7 7:12–19 7:20–29 7:22 7:30–37 7:37–8:7 7:39 7:40 8:2 8:4 8:7–9:1 8:9 9:4–12 9:5 9:8 9:12–20 9:20–10:1 9:22 9:25 10:1–10 10:3–4 10:4 10:10 10:11–20 10:21–30 10:25 10:29 10:31 10:31–11:6 10:33 11:1–2 11:3 11:5 11:7–17 11:12 11:14 11:18–25 11:25 11:26–12:2 11:29 12:3 12:3–12 12:5 12:13–24 12:20 12:24 12:24–13:1 13:2 13:2–11

276 218 218 271, 279 218 218 278 277, 279 278–279 275 218 227 218 276 278–279 219 219 224, 227 275 219 269, 277, 279 275 234 219 219 224–225 273 275 219, 225 273 225 225 225 219 276 276, 279 219 276 219 278 275 219 275 219 224–225 273 219 279 219

13:5 13:11 13:11–14:6 13:12 13:13 14:6–15 14:10 14:11 14:13 14:16–24 14:18 14:23 14:24–34 14:29 14:34–15:5 14:36 15:6–15 15:9 15:11 15:12 15:17–28 15:22 15:27 15:28–39 15:39 15:39–50 15:45 15:47 15:51–16:2 15:54 15:55 16:2–12 16:6 16:7 16:10 16:11 16:12–23

225 225 219 276 271, 279 219 226 274, 279 275 219 273, 279 275 219 226, 275, 279 219 271, 279 219 73 275 272, 279 219 276 272, 279 219 270, 279 219 272, 279 269, 279 219 275 73 219 272, 279 216, 226, 276, 279 276, 279 275 219, 225

2 Corinthians 1:1–8 1:5 1:8 1:8–15 1:14 1:16–2:1 1:19 1:20 2:3–12 2:8

219 270, 279 225 219 276 219 223 275 219 275

510 2:13–3:3 2:14 3:1 3:5–13 3:5–6 3:6 3:14–4:3 3:18 4:2 4:3 4:4–12 4:13 4:13–5:4 4:16 5:5–13 5:9 5:14–6:2 5:18 5:19 5:20 6:3–13 6:14–7:4 6:16 6:17 7:5–11 7:12–8:3 7:13 8:4–12 8:5 8:11 8:13–24 9:1–7 9:2 9:7–10:1 9:12 9:14 10:1–11 10:2 10:7 10:10 10:12 10:11–11:2 11:3–10 11:12–22 11:23–32 11:33–12:9 12:6 12:7 12:10

Ancient Sources Index 219 275, 279 269, 279 219 279 273 219 235 224 275, 279 219 275 219 275, 279 219 275, 279 219 269 269, 279 275, 279 219 219 272, 279 275 219 219 222, 225, 275, 279 219 273, 279 276 219 219 269, 279 219 224 224 219 227 274, 279 270, 279 278 219 219 219 219 219 276, 279 275 275

12:10–18 12:14 12:18–13:5 12:19 13:2 13:5–13

219 278 220 227 277 220

Galatians 1:6–7 1:8 1:10–22 1:23–2:10 2:12–21 3:2–15 3:16–29 4:2–18 4:20–5:1 5:2 5:2–17 5:20–6:8 5:24 6:10

191 220 220 220 220 220 220 220 220 172 220 220 237 220

Ephesians 1:1–11 1:12–20 1:21–2:8 2:10–20 2:21–3:10 3:11–4:1 4:2–14 4:15–25 4:26–5:6 5:8–25 5:26–6:6 6:2 6:8–18 6:20

220 220 220 220 220 220 220 220 220 220 220 222 220 220, 222

Philippians 1:1 1:5–15 1:5–28 1:17–28 1:20 1:30–2:12 2:2 2:14–27 2:29–3:8

220 220 225 220 222, 226 220 277 220 220

511

Ancient Sources Index 3:10–21 3:15 4:2–12 4:14

220 227 220 220

Colossians 1:2 1:5–13 1:7 1:16–25 1:27–2:7 2:8–19 2:22–3:11 3:3 3:13–25 4:3–12 4:16

220 220 224 220 220 220 220 222 220 220 220

1 Thessalonians 1:2 1:8–2:3 3:12 5:5–9 5:23–28

220 220 235 220 220

1 Timothy 1:12 2:11 3:15

172 172 14

2 Timothy 2:3 2:4

172 172

Titus 2:1 2:2

172 172

Hebrews 1:1 1:7–2:3 1:9 2:3–11 2:11–3:3 3:3–13 3:14–4:4 4:4–14 4:14–5:7 5:8–6:4

222, 226 217 226 217 218 218 218 218 218 218

5:9 6:1 6:2 6:4–13 6:6 6:13–7:1 7:1 7:2 7:2–10 7:11–20 7:20–28 7:25 7:28–8:8 8:9–9:2 9:3 9:10–16 9:12 9:18–26 9:26–10:8 10:5 10:8–20 10:22–30 10:32–11:3 11:4–9 11:9–17 11:18–26 11:26–34 11:35–12:1 12:2–11 12:11–21 12:12 12:21–13:2 13:3–11 13:12–20 13:25

226 226 226 218 226 218 226 226 218 218 218 226 218 218 38 218 18 218 218 173 218 218 218 218 218 218 218 218 218 218 172 218 218 218 222

James 1:20 2:4 2:15 4:10

353 353 353 353

1 Peter 5:1

353

2 Peter 2:18 2:20

353 353

512

Ancient Sources Index

1 John 2:12–14 2:13–16 2:18–22 2:19–22 2:23 3:17–22 4:1–3 4:10–14 4:18–21 5:7 5:8

291 291 291 291 291 291 291 291 291 312 312

2 John 5 12

353 353

3 John 4

353

Jude 23a

340

Revelation 1:11 3:7

66 66

5:1 5:9 8:9 9:13 12:3 12:8 12:10 13:8 13:10 13:18 14:8 15:3 16:2 16:5 18:3 18:10 18:22 19:14 20:11 21:4 21:11 21:16 21:27 22:3 22:16–21 22:21

373 373 31, 51 274 373 20 373 373 373 30, 76, 373, 376 373 55 55 54, 373 374 55 374 373 373 74–75, 77 77 55 373 54 53, 371 373

Early Christian Literature Augustine Serm. 137.15 295.2

262 261

Caesarius of Arles Serm. 28.3 261

Jerome Praef. in Parala.

169

Praef. in Pent.

169

Vir. ill. 55

165

Origen Comm. Matt. 15.14

169

Modern Author Index Abbas, Hyder 379, 385, 389 Abberley, Will 88 Abbott, Edwin A. 238 Abbott, T. K. 24–25, 248 Abraha, Tedros 267, 270, 275 Achelis, H. 57 Ahmed, Sarah 98 Aland, Barbara 82, 122–123, 222, 287– 288, 301, 318, 338, 341, 348, 351, 357, 374–375, 377, 456 Aland, Kurt 82, 122–123, 127, 222, 250, 283, 287–289, 294–295, 299, 301, 318, 338, 348, 351, 365, 374–375, 412, 456 Albin, C. A. 337 Alford, Henry 106 Allen, Garrick V. 26, 51, 64, 68–70, 88, 93, 100, 107, 206, 247 Allen, Graham 135 Allen, P. S. 59 Amphoux, C.-B. 232, 288 Anderson, Bradford XI Andrews, Tara L. 417, 425 Andrist, Patrick 178 Argyriou, A.370 Armoni, Charikleia 292 Armstrong, Elizabeth 307–310, 321–322 Arzt-Grabner, Peter 187–190, 196, 295– 297, 303 Askeland, Christian 49, 58–59 Aune, David E. 65, 365, 367–368, 370– 372 Austin, Linda 85 Baarda, Tjitze 305 Bagnall, Roger 184, 192–194, 197–199, 228 Bailey, D. R. S. 163 Bakhtin, Mikhail 133 Ball, Alex 431 Bardy, G. 169

Barnes, Timothy David 155, 165–166, 170 Barthes, Roland 134–135 Bauer, Walter 456 Beard, Mary 163 Beduhn-Mertz, A. 412 Behm, J. 29 Bell, Harold Idris 383 Bell, Lonnie 335 Bengel, J. A. 334, 340, 354, 359, 398 Benjamin, Walter 79, 84 Bentley, Richard 396–398, 400 Benz, Ernst 29 Berlent, Lauren 98 Betz, Hans Dieter 187 Bianchini, Giuseppe 19 Birdsall, J. N. 344, 376 Blake, R. P. 83 Blake-Hill, Philip V. 105 Blass, Friedrich 456 Blomberg, Craig L. 133 Blomkvist, Vemund 159, 161–162, 167– 168, 172, 177 Bloom, Harold 136 Blumell, Lincoln 184 Boismard, M.-É. 279 Bonwetsch, N. 57 Bordalejo, B. 438 Böttrich, Christfried 177 Bourdieu, Pierre 92, 95 Bousset, Wilhelm 13, 16, 56, 59, 231 Bowers, Fredson 133 Bowles, Edmund A. 410, 433 Bowman, J. H. 111 Boym, Svetlana 96, 99 Brightman, F. E. 97 Brock, Sebastian P. 163 Brooks Jr., F. P. 408 Brown, A. J. 53 Bruce, F. F. 126 Buchanan, E. S. 19

514

Modern Author Index

Bucking, Scott 291 Budge, Ernest Wallis 181 Burgon, John William 6–7, 13, 81–83, 93, 95, 97–99, 119, 341, 398, 447– 448 Burkitt, F. C. 16, 18–19, 23 Burnard, Lou 415, 431 Burton, E. de Witt 39 Burton, Philip H. 49, 252, 261 Busa, Roberto 407–410, 430, 434 Butler, H. E. 32–33 Cadwallader, Alan 104, 114–115, 119, 122 Canfora, Luciano 208 Caragounis, Chrys C. 241 Carlson, Stephen C. 237 Caroli, Menico 205, 208 Cate, Jeff 69 Causer, Tim 419 Caws, Peter 131 Caygill, Howard 79 Cerulli, Enrico 391 Chadwick, Owen 110, 126 Chakrabarty, Dipesh 93 Charles, R. H. 57 Childers, Jeff W. 50 Church, C. L. 104 Ciotti, Giovanni 205 Clark, A. C. 374 Clark, Kenneth W. 341, 355, 399 Clarysse, Willy 227, 342 Cleaver, Laura 380, 384 Clough, Cecil H. 100–101 Cole, Zachary J. 245, 376 Colwell, E. C. 123, 127, 144–145, 342– 344, 346, 360, 399 Constantinidou, N. 310 Conybeare, F. C. 161 Coogan, Jeremiah 155, 159, 174 Coogan, R. 53 Cook, Clair M. 408 Cook, F. C. 104 Coroleu, A. 53 Coron, Antoine 310, 313 Corsten, Thomas 189 Couchoud, Paul-Louis 76 Cowe, S. Peter 50 Cowey, James M. S. 292 Crawford, Matthew 158, 165, 206 Cribiore, Raffaella 192–193

Croke, Fionnuala 380 Crosman, Inge 135 Cross, Samuel H. 32 Culter, Anthony 100 Cummings, James 418, 431 Cuvigny, Hélène 182 Dahl, Nils 158, 162, 174 d’Alveydre, J. A. S.-Y. 28 Davies, Rachel Bryant 88 Debrunner, Albert 456 de Bruyn, Theodore S. 294–300, 303– 304 Deegan, Marilyn 431 de Groote, Marc 29–30, 67 de Guldenstubbé, Louis 69–74, 76 de Hamel, Christopher 123 Deißmann, Adolf 185–188, 191, 195, 197, 199, 295 de Jonge, Henk Jan 55–56 Dekker, Ronald H. 424–425 Delitzsch, Franz 59–60 Delmarter, Steve 280 der Nersessian, S. 391 Derrida, Jacques 131–132, 134 De Troyer, Kristin 295, 297–298, 303 Deveney, John Patrick 72 Diekamp, Franz 21 Dijkstra, Jitse H. F. 294–300, 303–304 Dill, U. 53, 61 Dinshaw, Carolyn 91 Diobouniotis, C. 26 Dionisotti, Anna C. 310 d’Olivet, Fabre 35 Donato, Eugenio 131 Dorfbauer, Lukas 198 Dorival, Gilles 228 Driscoll, Matthew J. 136–137, 140, 142– 143 Drum, Walter 39 Duchesne, L. 172 Dumville, D. N. 262 Dungan, David L. 123 Duxfield, Polly 420 Ebojo, Edgar B. 206, 214–217, 221–223 Edmondson, Andrew 412, 434 Edmunds, Albert 16 Edwards, David L. 109, 113 Edwards, Mark 165 Ehrhard, Albert 158

Modern Author Index

515

Ehrman, Bart D. 104, 122, 124, 138, 208, 307, 329, 336, 343, 347–348, 397, 400, 409 El-Haj, Nadia Abu 98 Ellerton, John 106–107 Ellicott, C. J. 119 Elliott, J. K. 59, 64–65, 312, 364, 372 Ellis, John M. 141 Ellison, John W. 407–413, 430, 434 Epp, Eldon J. 49, 83, 88, 96, 103, 124– 125, 127, 138, 144–145, 213, 231, 334–335, 337, 342, 346, 352, 354– 355, 399–401 Evans, Richard J. 361

Greetham, D. C. 136, 139–140, 142–143, 145 Gregory, C. R. 5, 285–287, 289–290 Grenfell, Bernard S. 175, 182 Griesbach, J. J. 326–327, 334–335, 343, 359 Griffitts, Troy 428, 450 Gryson, Roger 58, 64, 255, 369 Guarducci, Margherita 295 Guidi, I. 42 Gurry, Peter J. 52, 88, 98, 125, 139, 198, 281, 340, 348, 351, 354–355, 357, 375, 411, 443 Gwynn, J. 41

Falluomini, Carla 50 Farnes, Alan Taylor 252 Farnsley II, Arthur E. 104 Farstad, Arthur L. 372 Fee, Gordon D. 127, 145, 147, 222, 232– 233, 238–241, 244, 344–346, 350 Fernández Marcos, N. 169 Firestein, S. 403 Fischer Bonifatius 247–252, 258, 261, 412–413, 430 Flinn, Andrew 413, 430 Foucault, Michel 134–135 Frier, Bruce 192

Haelewyck, J.-C. 248, 252 Hagedorn, Dieter 292 Haines-Eitzen, Kim 222 Hall, H. R. 5 Hall, Isaac H. 13 Halliday, Michael A. K. 146 Hamilton, Alastair 324 Hamilton, Paul 137 Harari, Y. N. 403 Haraway, Donna 98 Harris, J. Rendel 5–7, 10, 18, 25, 29, 31, 33–35, 76, 93, 97, 199 Harrison, James R. 187 Head, Peter 123, 196, 290–291, 301–303 Heide, Martin 50, 53, 58–59 Heine, Ronald E. 155 Hellholm, David 162 Helm, Karl 19 Helm, Rudolph 166 Hendel, Ronald 92 Herder, J. G. 88 Hernández, Raquel M. 296 Hernández Jr., Juan 51, 58, 64–65, 67, 69, 107, 206, 363, 369 Hindley, Meredith 408 Hockey, Susan 407, 430 Hodges, Zane C. 372 Hofmann, Josef 266–268, 271, 276–277, 279 Holmes, Martin 417, 431 Holmes, Michael W. 96, 127, 138, 156, 208 Holquist, Michael 133 Holub, Robert C. 135 Hoogendijk, F. A. J. 183 Horne, Thomas H. 319

Gäbel, Georg 49, 94, 139, 355–356, 375, 442–443, 448 Gallagher, Tim 336 Gaslee, S. 20 Gathercole, Simon J. 205 Geer Jr., Thomas C. 338, 344 Genette, Gérard 156, 202–205, 207, 209 Gentry, Peter J. 170 Gerber, Albrecht 185, 187, 195 Gibson, Roy 163–164 Gignac, F. 242 Gilbert, Penny 413 Glaue, Paul 19 Goff, Philip 104 Goodspeed, Edgar J. 8, 14, 20–21, 39, 42, 58 Goswell, Greg 171–172, 206 Goulburn, Edward M. 340 Gracia, Jorge J. E. 132, 150–151 Grafton, Anthony T. 155–156, 164–166, 169, 307 Grayston, Kenneth 430

516

Modern Author Index

Horsley, Greg H. R. 187 Hort, Arthur F. 105 Hort, F. J. A. 6–8, 16–17, 19, 21, 23–24, 30, 41, 49, 52, 86–89, 91–92, 94–96, 103–127, 231, 281, 306, 323, 334– 335, 338, 341, 343, 359, 371–372, 398–404, 406 Horton, Charles 384 Hoskier, H. C. 3–71, 74–77, 79–83, 86– 88, 91–95, 97–99, 100, 107, 181–182, 185, 195, 197, 199, 227, 231, 248– 249, 252–254, 261–263, 265–281, 293, 305, 329, 337–340, 343, 360, 364, 367, 376, 400, 430, 447–450, 459 Houghton, H. A. G. 25, 163, 176, 196, 199, 247, 249–252, 254, 259, 415, 421, 431–432 Howard-Hill, Thomas 424–425, 428 Huffer, Lynne 84 Hüffmeier, Annette 94, 99, 377, 442, 448 Hull, Robert F. 104, 112, 124 Hunt, Arthur S. 175, 182, 286, 298 Hurtado, Larry W. 151–152, 156, 345 Hutton, Edward A. 125 Huyssen, Andreas 85 Ide, Nancy M. 418 James, Patrick 242 Jänicke, Stefan 425–426 Jenkins, R. G. 164 Jones, Brice C. 196, 283–284, 288, 290– 291, 294–298, 301–303 Jones, Steven E. 407, 409 Jongkind, Dirk 103, 113, 156, 171–174, 176, 372 Jördens, A. 183 Juckel, Andreas 50 Jülicher, Adolf 250 Junack, Klaus 336 Kamphuis, Bart L. F. 306 Karavidopoulos, J. 372 Karrer, Martin 9, 53–54, 364–365, 377 Karrer, T. F. 60 Kasser, Rodolphe 341 Kemman, Max 414 Kenney, E. J. 209 Kenyon, F. G. 38, 181–182, 216, 220– 221, 232–233, 244–245, 265, 343 Kern, Otto 295

Keysler, J. G. 84 Kießling, Emil 183 Kilbride, William 431 Kilpatrick, G. D. 235 King, Karen L. 90 Kirschenbaum, M. 424, 422 Kleppe, Martijn 415 Klijn, A. F. J. 228 Kloppenborg, John S. 160 Knapp, Georg Christian 75–76 Knibb, Michael 268 Knopf, Rudolf 295–296 Knust, Jennifer Wright 7, 14, 89, 92, 155, 159, 170–171, 403 Koester, Helmut 224 Köster, Beate 336 Komoszeksi, J. Ed 122 Kraft, Robert A. 407, 411, 433 Krans, Jan 10, 36, 53–55, 79, 93, 107, 127, 309, 312–313, 316–317, 322– 324 Kraus, Thomas J. 293–295, 299–300, 304 Kreinecker, C. M. 183, 188, 190–192, 195 Kristeva, Julia 133–135 Kritzer, Ruth E. 188 Krüger, Friedrich 336 Krüger, Volker 451, 453 Kubo, Sakae 336, 344 Kümmel, Werner G. 341 Kutzer, Edgar 192 Lachmann, Karl 86, 90–92, 95, 97, 103, 119, 124, 343, 346, 398, 437–438 Lack, R.-F. 134 La’da, Csaba A. 193 Lake, Kirsopp 18, 83, 104, 172, 231, 345, 397 Lake, Silva 345 Lamouille, A. 279 Lang, T. J. 206 Lange, Armin 271 Latour, Bruno 89, 94 Lavagnino, John 417 Law, T. Michael 164 Lawlor, H. J. 248–249 Lee, A. L. 187 Lefebvre, G. 290 Leipoldt, J. 42 Leitch, Vincent B. 136

Modern Author Index Lembke, Markus 26, 31, 50, 60–61, 63– 65, 364–366, 368 Lévi-Strauss, Claude 132 Levinson, Stephen H. 238 Lied, Liv Ingeborg 98 Lightfoot, J. B. 106, 109 Lin, Hang 205 Lin, Yii-Jan 88, 91, 95 Lipphardt, Veronika 88 Llewelyn, Stephen R. 187 Lorimer, Douglas A. 88 Lorusso, Vito 208–209 Lowe, E. A. 164, 252 Lowell, Amy 33 Lüdeke, Roger 91 Luijendijk, AnneMarie 298 Maas, Paul 440 MacGregor, Geddes 104, 120 Macksey, Richard 131–132 MacLachlan, R. F. 252 Maehlum, Helge 271 Maldfeld, Georg 287, 293 Malik, Peter 66, 68, 156 Marchand, James W. 163 Markley, Jennifer Foutz 133 Martens, Peter 155 Martin, Dale 90 Martin, Victor 341 Martini, Carlo 233, 341, 358, 372 Matzkow, Walter 250 McCully, William R. 409 McCutcheon, R. W. 163 McGann, Jerome 159 McGurk, Patrick 249, 252 McKenzie, Judith S. 176 McKitterick, David 118, 121 McNamara, Martin 247, 252 McReynolds, Paul 347 Menchi, S. 61 Mercati, Iohannis 164 Metzger, Bruce M. IX, 50, 103–104, 106, 232, 266, 268, 277, 288, 305, 307, 312, 316, 318–319, 323, 341–343, 354, 359, 371, 373, 397, 411 Middell, Gregor 425 Middleton, T. F. 238 Migne, J. P. 159 Millar, Eric G. 382, 384 Milligan, George 187 Mills, Sarah 136

517

Milne, H. J. M. 171–172, 176 Mink, Gerd 94, 138–139, 198, 348–350, 352, 360, 375, 413, 435, 437–439, 442–444, 448 Mitteis, Ludwig 182 Mitthof, Fritz 192–193 Moi, Toril 135 Morgan, Charles 107 Morton, Andrew Q. 430 Mosshammer, Alden A. 166 Moulton, James Hope 187 Moulton, W. F. 114 Mowitt, John 132–134 Mugridge, Alan 156 Müller, Darius 9, 25, 64, 66, 364, 368 Müller-Sievers, Helmut 91 Murdoch, John 379 Murphy, Harold S. 170, 179 Nagel, Peter 50 Negroponte, Nicholas 427 Neill, Stephen 104–105 Nestle, Eberhard 13, 18, 53, 55, 103, 125–127, 159–160, 283–285, 295 Neuschäfer, Bernhard 155 Nevius, Richard C. 238 Niccum, Curt 37, 50, 58–59, 269, 271, 279 Nichols, Stephen 137 Nicklas, Tobias 369 Nongbri, Brent 227–228, 342 Nordenfalk, Carl 158–160, 163, 168, 176–177 Nus, J.-B. E. 28 Nyhan, Julianne 413, 431 Oakman, Robert L. 408, 413 Öhler, Markus 189 Olender, Maurice 87 Oliver, Harold H. 159, 166 Oliver, Kelly 135 Olsson, Bror 190 Omont, M. H. 172–173 O’Neill, J. C. 232 Osburn, Carroll D. 345 Orbell, John 4 Orsini, Pasquale 227, 342 Ott, Wilhelm 412–413, 414–416 Owen, Robert Dale 71 Palmer, E. 119

518

Modern Author Index

Papaioannou, Stratis 100 Parker, D. C. 10, 65–66, 88, 125, 137– 138, 142, 156, 170, 196, 198, 201, 236, 252, 284, 288–290, 300, 305– 306, 312, 352, 357–358, 369–370, 395, 397, 399–400, 402, 415, 418, 420, 443 Parsons, Ernest W. 231–232 Parvis, Merrill M. 347, 410–411 Pasquali, Giorgio 440 Patrick, Graham A. 105, 109–110, 119– 120, 126 Paul, Carol 100 Pavel, Thomas G. 146 Payne, Philip B. 171 Peebles, B. M. 100 Peikola, Matti 205 Perosa, A. 56 Perowne, J. J. S. 118 Peters, Ronald D. 238 Peterson, Jacob W. 215 Petersen, William L. 137–138, 448 Petzer, Jacobus 137–138 Pfeiffer, Rudolf 169, 175 Pickering, Stuart R. 301–302 Pickering, Wilbur N. 14, 341 Pierazzo, Elena 414, 418, 430–433 Pierpont, William G. 94, 372 Pinney, T. 14 Pintaudi, R. 294 Pitts, Andrew W. 123, 143, 148 Platt, Thomas Pell 267, 269–270 Plummer, Alfred 13, 39, 82 Porter, Calvin L. 341–342 Porter, Stanley E. 123, 131, 133, 139, 143, 148, 152, 301 Powerscourt, Sheila 381, 386, 388, 390 Preisigke, Friedrich 182–183 Prigent, Pierre 57 Puar, Jaspir K. 98 Puech, Aimé 29 Puech, H.-C. 29 Rabinow, Paul 135 Racine, J.-F. 338, 344 Rahlfs, A. 164, 294–295 Ramsay, Stephen 433 Rand, E. K. 164 Raskin, J. 428 Reeve, Michael D. 312 Reuss, Eduard 115

Reynolds, L. D. 169, 175 Richard, M. 57 Richards, W. L. 336, 344, 347, 412 Richter, Siegfried 50 Roberts, C. H. 227 Robinson, Jason C. 131 Robinson, J. Armitage 158–159, 161, 173 Robinson, Maurice A. 40, 94, 372 Robinson, Peter 414–415, 419–420, 423, 427, 431, 433 Rochlitz, Rainer 84 Rogers, Molly 88 Römer, Cornelia Eva 290–292, 303 Ross, J. M. 53 Rowlandson, Jane 193 Royse, James R. 40, 65, 207, 214, 221– 222, 267, 269–279 Rupp, E. G. 109 Ruwet, Nicolas 131 Salmon, George 82, 335, 339 Sanday, W. 23, 337 Sanders, Henry A. 5–6, 15–16, 22, 31, 36, 215, 221, 265 Sanzo, Joseph E. 294, 296 Sawyer, M. James 122 Schäfer, Donovan 98 Schaper, Joachim 164, 169 Scherbenske, Eric 158–159, 162, 206 Schierl, P. 61 Schironi, Francesca 168–170, 175, 205 Schmid, Josef 31, 43–48, 50, 57–58, 61, 62–63, 67, 370 Schmid, Ulrich B. 61, 65, 208, 213, 364– 366, 416, 450 Schmidt, Carl 295–296 Schmidt, Desmond 417 Schoene, Alfred 166 Schröder, Katharina D. 50 Schwab, Günther 188, 191 Scott, David 79–81 Scott, Joan W. 80 Scrivener, F. H. A. 52, 55, 57, 61, 94, 97, 103–104, 118–119, 304–306, 310– 314, 319–320, 324–325, 371, 396, 398, 400 Selden, Raman 136 Semler, J. S. 334 Shillingsburg, Peter 96, 413–414 Sievers, Martin 421

Modern Author Index Sigismund, Marcus 9, 50, 370 Sinaites, Justin 268 Skeat, T. C. 171–172, 176 Smith, Catherine J. 199, 255, 415 Smith, Courtney Weiss 90 Smith, Helen 205 Smith, Jonathan Z. 92, 160 Smith, Kimberly K. 84–85 Smith, Paul 134 Smith, W. Andrew 156, 172–174 Snapp Jr., James 296 Soule, Gardner 408 Souter, Alexander 8, 20, 23, 29, 91 Sperberg-McQueen, C. 417–418 Spielmann, Kent 239 Staab, K. 29 Stehly, R. 57 Streeter, B. H. 83, 232, 397 Strutwolf, Holger 94, 98–99, 352, 375, 442–443, 448 Suggit, John N. 29 Suleiman, Susan R. 135 Tanselle, G. Thomas 133, 145 Tasker, R. V. G. 201 Taylor, C. 164 Terras, Melissa 419, 425 Thackeray, H. St. John 59, 339 Thomsen, Mads R. 136 Thorpe, James 141, 145 Thuesen, Peter J. 104 Timpanaro, Sebastiano 12 Tischendorf, Constantin 16–17, 48, 66, 86, 95, 103, 119, 122, 327, 371, 398 Toda, Satoshi 158 Tompkins, Jane 135 Tonra, Justin 419 Tov, Emanuel 206 Tovar, Sofia T. 296 Treier, Daniel J. 92 Tregelles, S. P. 61–62, 103, 111, 119, 303–306, 312, 318–319, 321, 324, 371–372, 396, 398, 404 Treu, K. 50 Trovato, Paolo 436–439 Tuckett, Christopher 355 Tune, Ernest W. 123, 144–145, 342–344, 346, 360 Turner, C. H. 24 Turner, Eric T. 181–182, 293 Turner, James 127

519

Tyson, Lois 136 Tzamalikos, P. 26 Ubaldi, P. 29 Uhlig, Siegbert 271 Vagany, León 30, 288 Vanhoutte, Edward 408 Vansittart, A. A. 114–115 van Zundert, Joris 414, 425, 432, 434 Verheyden, Joseph 189 Vervliet, Hendrik D. L. 307–309, 313 Voltmann, Niklas 68 von Dobschütz, Ernst 285–290, 292–293, 296, 300–302 von Greyerz, K. 61 von Harnack, Adolf 226 von Soden, Hermann 16, 22, 30, 40–41, 48, 52, 62, 159, 285, 335–337, 343, 347, 360, 401 Vööbus, Arthur 266 Wachtel, Klaus 49, 94, 99, 125, 139, 221, 318, 346, 350, 354–356, 358–359, 375, 377, 413, 418, 436, 442–443, 448–449 Waite, Stephen V. F. 416 Wallace, Daniel B. 7, 60, 87, 93–94, 122, 340–341 Wallace, Valerie 419 Wallraff, Martin 61, 158, 166, 175, 178 Ward, Aengus 420, 429 Wasserman, Tommy 7, 88–89, 92, 98, 139, 155–156, 159, 170–171, 198, 281, 283, 340, 342, 344, 348, 351– 352, 355–359, 375, 403, 411, 443, 448 Watson, Francis 159, 176 Welte, Michael 336 Wenger, Leopold 197 Wessely, Carl 181 Westcott, Arthur 105 Westcott, B. F. 6–8, 16–17, 19, 23–24, 30, 49, 52, 86–89, 91–92, 94–97, 103–127, 231, 281, 305, 315, 334– 335, 338, 341, 343, 359, 371–372, 396–402, 404 Weston, Arthur H. 15 Wettstein, J. J. 77, 322, 326–327, 371 White, H. G. Evelyn 176 White, H. J. 248

520

Modern Author Index

White, John L. 187 Wilcken, Ulrich 181–182, 295 Willard, Charles 158–159, 161, 167, 173, 206 Williams, James 134 Williams, Michelle 155–156, 164–166, 169 Williams, Peter J. 49, 113, 124, 242 Wilson, A. J. 380 Wilson, Louise 205 Wilson, N. G. 169, 175 Winer, Georg 109 Winter, J. G. 190 Winter, Thomas N. 409

Wiseman, T. P. 165 Wisse, Frederik 83, 347–348 Witte, Klaus 221 Wolfreys, Julian 133 Wordsworth, John 248, 308 Youtie, H. C. 190 Yu, Gan 421 Zacagnius, L. A. 159–162, 167–168, 177 Zellman-Rohrer, M. 298–299 Zuntz, Günther 158, 163, 201, 214, 216, 220–222, 227, 341, 354 Zuurmond, Rochus 49, 267–268

Subject Index Alexandrian recension 231–233, 244– 245 Ammonius 158, 160, 164–166, 169, 175, 206 Andrew of Caesarea 21, 29, 46–48, 53, 60, 67, 363, 369 Aramaic 42–43, 341 Back of Beyond 10–11, 28, 35–36, 70, 195 Caesarean text 82–83, 342–344 Carpianus, Letter to 158–160, 165, 175– 178 Chester Beatty, Alfred X, 379–391 Chicago School 399–400 Claremont Profile Method 336–337, 346–348 Codex Alexandrinus (A02) 159, 167, 171–174, 177, 236–237, 239, 241, 324, 343, 353, 398 Codex Bezae (05) 42–43, 97, 156, 251, 259, 311, 321–324, 348, 398 Codex Sinaiticus (‫א‬01) 15–16, 20, 27, 43–45, 47, 67, 159, 167, 170–177, 201, 236, 271, 285, 334–336, 339– 340, 345, 353, 358, 360, 372–374, 386–387, 398–400, 402 Codex Vaticanus (B03) 7–10, 23–24, 29, 41, 52, 59, 61, 67, 103, 171, 175, 231–245, 260, 271, 318, 327, 334, 338–342, 353, 360, 372, 398–401 Coherence-Based Genealogical Method (CBGM) X, 98–99, 138–139, 142, 150, 198, 333, 346, 348–361, 370– 372, 374–377, 411, 435–446, 450, 452, 457, 459 Cologne Center for eHumanities (CCeH) 359, 434, 436, 449–450 Commentary manuscript(s) 8, 21–22, 26, 29–30, 33, 53, 66, 367, 369–370

Complutensian Polyglot 61, 92, 311–313, 323 Contamination (textual) 27, 47, 248–249, 256, 343, 346, 347–349, 351, 357, 360, 436, 439–446 Correction(s) 98, 151, 179, 182, 201– 229, 233–234, 236, 244, 253, 270, 299, 313, 327, 357–358, 374, 422– 423, 456–457 Critical Apparatus 11, 22, 30, 32, 40, 54– 56, 66, 99, 108, 122, 150, 153, 248, 302, 305–333, 336–337, 347, 354, 360, 364, 373, 376, 405, 413–414, 426, 431, 436–437, 439, 441–459 Digital Humanities 49, 195, 199, 407– 434, 449 Editio Critica Maior (ECM) X, 8–9, 49– 50, 56, 58–59, 62–66, 94, 98–99, 125–126, 139, 156, 339, 348, 350– 355, 357–358, 360, 363–364, 367, 371–372, 375–377, 406, 412, 415, 418, 421, 423, 426, 432–437, 439, 442–444, 446–459 Eusebius of Caesarea (Eusebian apparatus) 155–179, 206, 212, 228, 320, 327, 334, 423 Eschatology (eschaton) 31, 51–52 Euthalius (Euthalian apparatus) 157–164, 166–168, 170–175, 177–179, 206, 334 Extensible Markup Language (XML) 418–421, 423, 425, 431 Garland of Howth X, 8, 24–25, 247–263 Guinness Brewery 4 Historic Printed Editions – Bengel (1734) 334, 340, 354, 359, 398

522

Subject Index

– Beza (1624, 1633) 54, 56, 307, 316, 320–322, 326 – Erasmus (1516–1535) 25, 53–56, 59– 61, 308–309, 316–317, 320, 325, 367, 371, 398 – Griesbach (1777, 1785, 1796) 307, 326–328, 334–335, 359 – Knapp (1797) 75–76 – Mill (1707) 111, 307, 312, 321, 324– 325, 328, 333–334 – Stephanus (1550, 1551) X, 9, 14, 21, 32–34, 54–56, 305–333 – Wettstein (1751–1752) 77, 307, 312, 326–328, 371 Hoskier, H. C. (biography) 3–12, 51–53 International Greek New Testament Project (IGNTP) 99, 125, 337, 347, 406, 410–412, 415, 418, 420, 423, 431 Institut für Neutestamentliche Textforschung (INTF) 64–65, 99, 125, 348, 359–360, 364, 370, 375, 412, 416, 418, 420, 423, 431–432, 434– 435 Institute for Textual Scholarship and Electronic Editing (ITSEE) 415, 432, 450 King James (Authorized) Version (1611) 19, 31, 54, 103–104 Kurzgefaßte Liste 65, 283–304, 336, 363–369 Majority Text 60, 93–94, 98, 299, 335, 340, 353, 372, 444–445 Morgan, J. P. 4, 15–16 Morphology 235–236, 241–243 Narrative textual criticism 137–143 New Historicism 136–137 New Testament Virtual Manuscript Room (NTVMR) 199, 280–281, 283, 292, 304, 428–429, 433, 446, 450, 453, 459 Nostalgia 84–90, 96–101 Oecumenius 8, 21–22, 29–30, 30–33, 52, 66–67, 369

Origen 26, 155–156, 164–166, 169–171, 179, 228, 319, 334, 367 Ostraca 186, 189, 289–292, 300–304 Papyri 181–199, 341–342 Chester Beatty 36–38, 214–228, 273, 279, 381–384, 390 Paratext(s) 149–153, 155–179, 201–229 Pericope Adulterae 98, 115, 144, 171, 314, 336, 398, 404–405 Philology X, 38, 127, 136–137, 143, 175, 195, 245, 405 Polyglot(s) IX, 15–17, 20–22, 24, 27, 30, 36–38, 40–50, 53, 59–61, 249, 323 Revised Version (1881) 7, 23, 37, 86, 104, 112, 114, 117–121, 123, 409 Septuagint (LXX) 38, 92, 169, 185, 189, 289, 302, 390 Spiritism/spiritualism 5, 10–11, 28–30, 48, 69–77, 79, 195 Structuralism 131–137, 436 Synoptic Problem 398, 405 Systemic Functional Linguistics (SFL) 146–149 Talismans 292–304 Text Encoding Initiative (TEI) 417–420, 423, 425, 429, 431 Text types (families/groups) 45–48, 83, 89, 139, 232, 238–239, 241, 316, 333–361, 370–373, 443 Textual history 17, 39–50, 52, 56–61, 64–67, 141, 179, 281, 352, 360, 382, 440, 459 Textual variant(s) 20, 38, 54, 62, 69, 143–149, 153, 234, 237, 256–257, 260–261, 270, 272, 290, 308, 311– 317, 319–333, 339–340, 344–346, 349, 351, 353, 358, 373–377, 399, 404, 410–412, 424–426, 428, 435– 436, 438–439, 441–446, 449, 451– 452, 454–459 Text und Textwert (TuT) 26, 31, 50, 62– 65, 279, 318, 363–365, 368, 370, 376–377, 435, 437, 450, 459 Textus Receptus IX, 6–7, 9, 13, 17, 21, 23, 31–32, 37, 45, 52–57, 60, 80, 82,

Subject Index 86, 95, 98–99, 103–104, 106–107, 119, 316, 318, 320, 334, 338, 343, 360, 367–368, 371, 398–399, 401, 410, 430 Theosophy 6, 10–11, 28–29 Thoroughgoing eclecticism 363–377 Translation technique 268, 271–281 Versions – Arabic 32, 269, 300, 323, 379, 381 – Armenian 32, 49–50, 57, 158, 195, 379, 382, 388, 391, 448, 452 – Coptic (Sahidic and Bohairic) 16–17, 20–22, 26, 32, 36–39, 42, 44–47, 49, 56, 58–59, 67, 195, 265, 291, 293, 295, 299, 338, 379, 381–382, 388, 390–391, 448–449, 451–452 – Ethiopic (Ge‘ez) 32, 37, 49–50, 56, 58–59, 163, 265–281, 323, 379, 388– 389, 448, 451

523

– Georgian 49–50, 448, 452 – Gothic 18–19, 49–40, 163, 176, 195 – Latin 8, 14–25, 32–34, 37, 42–45, 47, 49, 53, 56, 58–59, 64, 76, 97, 108, 124, 159, 163, 166, 168–169, 176, 182, 195–196, 247–263, 267, 269, 271, 305–308, 310, 319, 322–323, 325, 334, 336–338, 358, 369, 371, 376, 379, 384, 388, 396, 398, 412, 416, 428, 430, 446, 448, 451–452 – Old Church Slavonic 379, 388, 391, 448 – Persian 323, 379 – Syriac 16–17, 19–21, 27, 32, 37, 41– 46, 49, 56, 58, 77, 110, 124, 163, 170, 195, 281, 323, 336, 355, 379, 382, 388, 391, 448, 451 Walton Polyglot 267, 323–324 World War I 4–5, 11, 24, 31, 34, 51, 65, 100, 182, 195