Hostname: page-component-74656f9884-6grmk Total loading time: 0 Render date: 2026-11-06T07:42:03.744Z Has data issue: false hasContentIssue false
  • English
  • Français

The Testimony for Marcion's Gospel in NA28: Revisiting the Apparatus to Luke in the Light of Recent Research

Published online by Cambridge University Press:  09 December 2021

Dieter T. Roth Open the ORCID record for Dieter T. Roth [Opens in a new window]
Dieter T. Roth*
Affiliation:
Boston College Theology Department, Stokes 310N, Chestnut Hill, MA 02467, USA Email: dieter.roth@bc.edu
Rights & Permissions [Opens in a new window]

Abstract

Scholarly work on Luke has often noted the significance of Marcion's Gospel for understanding the textual history of the third canonical Gospel. It is not surprising, therefore, that in the past new insights into Marcion's Gospel have led to revisions in the apparatus of the highly influential Nestle-Aland Novum Testamentum Graece, now in its 28th edition. In view of the precedent for continually updating the Nestle-Aland text and apparatus, this article revisits the apparatus to Luke in the light of recent research on Marcion's Gospel in order to highlight problematic references that should be changed or removed in the apparatus of future Nestle-Aland editions.

Keywords

Marcion's GospelLukeNA28textual criticism

Information

Type
Articles
Information
New Testament Studies , Volume 68 , Issue 1 , January 2022 , pp. 52 - 60
Check for updates
Copyright
Copyright © The Author(s), 2021. Published by Cambridge University Press

1. Introduction

The recent resurgence of scholarly interest in Marcion's Gospel has served to underscore anew the importance of readings attested in this Gospel text for questions surrounding the textual history of Luke. Particularly significant in this regard are both Greek reconstructions and translations of Marcion's Gospel that have been published over the past several years.Footnote 1 Though the recognition of the significance of readings in Marcion's Gospel is not novel,Footnote 2 new work on the text of Marcion's Gospel offers an opportunity both to revisit and to reconsider references to the attestation for Marcion's Gospel found in the critical apparatus to Luke printed in the highly influential Nestle-Aland Novum Testamentum Graece. Footnote 3 The following study of Marcion's Gospel in the Lukan apparatus in NA28 offers an overview of the references to Marcion's Gospel in that apparatus and then evaluates critically numerous references in the light of recent work on Marcion's Gospel. As will be seen, even though there are numerous helpful and appropriate references to Marcion's Gospel in the NA28 apparatus, in more than a few instances they are misleading at best and erroneous at worst. Based on this analysis, and in view of revisions in previous editions of the Nestle-Aland apparatus when scholarship revealed problematic references to Marcion's Gospel, it is evident that further changes should be made in the apparatus of future editions.

2. Precedence for Revisiting ‘Mcion’ in the Nestle-Aland Apparatus

Even a brief perusal of previous editions of the Nestle or Nestle-Aland Greek New Testament reveals that both the printed text and its apparatus have been changed and updated over its lengthy history.Footnote 4 For this reason, it is not at all surprising to discover additions or deletions of references to Marcion's Gospel along the way, especially in the light of actual or perceived scholarly progress in the reconstruction of Marcion's Gospel. Of course, one would expect the publication of Adolf von Harnack's magisterial work on Marcion in the 1920s to have been particularly influential in this regard, and this is indeed the case. In fact, a notable example involves the manner in which a change at Luke 24.12 in the second edition of Harnack's monograph led to a significant change in the Nestle-Aland apparatus.Footnote 5 In his original 1921 publication, Harnack did not offer any comments concerning Luke 24.12 in his reconstructed text;Footnote 6 however, in the second edition published in 1924, he stated in his reconstruction, ‘Die Petrus Perikope [v. 12] fehlt’, despite there being no evidence to support this assertion in any source for Marcion's Gospel. Harnack's rationale for his assessment is found in the note to the verse: ‘von M. [Marcion] gestrichen, der Petrus hier nicht wünschte’.Footnote 7 As pointed out subsequently by Frans Neirynck, Harnack's view that Marcion had omitted this verse made its way into the Nestle apparatus in the 16th edition (1936) and it remained there until NA25 (1963).Footnote 8 Since the sources for Marcion's Gospel are all silent concerning this verse,Footnote 9 however, I would contend that Neirynck was correct in arguing that Marcion cannot be marshalled as a witness to the omission of Luke 24.12. The editors of the Nestle-Aland apparently agreed with this position since, starting with the apparatus of NA26 (1979), Marcion no longer appears as a witness for the omission of Luke 24.12.Footnote 10

More recently, a comparison of the apparatus in NA27 and NA28 also reveals five differences between the two that are related to Marcion's Gospel. A problematic change in Luke 4.23 and problems in 5.24 are discussed below; yet, the three other editorial changes are also worth noting. First, in Luke 12.8–9, NA27 indicates that א*; McionT(E) attest the omission of τῶν ἀγγέλων in the phrase τῶν ἀγγέλων τοῦ θɛοῦ found at the conclusion of both verses.Footnote 11 NA28, however, has completely removed the reference. In all likelihood, this change is due to the original reading in Codex Sinaiticus (א) actually being illegible.Footnote 12 Though the reference to Mcion could be retained even without any other manuscript witnesses,Footnote 13 in that case it should read McionT, E since Epiphanius explicitly attests the omission, making the parentheses unwarranted.Footnote 14 Second, in Luke 12.14, concerning the reading κριτὴν ἢ μɛριστήν, the NA27 apparatus lists McionT as attesting κριτην vl δικαστην due to the lack of certainty surrounding the Greek word that Tertullian's Latin text is attesting. NA28 avoids the ‘or’ reference by printing the Latin term iudicem, which is the only instance in which the apparatus presents attestation for a reading in a source for Marcion's Gospel in a language other than Greek. Finally, a change in Luke 16.20 probably reflects the desire to provide ‘more clarity’ in the NA28 apparatus.Footnote 15 Here the NA27 apparatus combines the evidence for the addition of ην and ος to the Nestle-Aland reading and lists McionA a single time, though attesting the txt reading in both instances (i.e. without the additions). NA28 provides the evidence for each reading separately, leading to McionA appearing twice in the apparatus since two variants are in view. Minor updates such as these final two are understandable; however, after a brief overview of the references to Marcion's Gospel in the NA28 apparatus, it will be demonstrated that several additional and, at times, more substantial changes are necessary.

3. Overview of the References to Marcion's Gospel in the NA28 Apparatus

The list of ‘Abbreviations for the Church Fathers’ found in the introduction to NA28 indicates that the apparatus labels readings attested for Marcion as ‘McionT/E/A’, referring to ‘Marcion according to Tertullian, Epiphanius, Adamantius’.Footnote 16 Tertullian, Epiphanius and the so-called Adamantius Dialogue are the most important sources for Marcion's Gospel,Footnote 17 and by my count, there are sixty-five verses in Luke for which the apparatus includes at least one reference to a reading attested for Marcion in one or more of these three sources. It is worth noting, however, that in Luke 11.3 there is also one reference to McionOr for a reading attested in Origen, a siglum easily understood even though it is not listed with the other abbreviations. Consonant with the amount of material attested for Marcion's Gospel in Tertullian, Epiphanius and the Adamantius Dialogue,Footnote 18 it is not surprising that McionT appears most often (in forty-three verses),Footnote 19 followed by McionE (in twenty-one verses)Footnote 20 and McionA (in thirteen verses).Footnote 21 The clear benefit of this manner of citation is that the data found in the sources can be presented without ultimately deciding whether the attested reading is accurately reflecting Marcion's Gospel. The following analysis, therefore, does not focus upon the question of whether an attested reading was actually present in Marcion's Gospel or not, but rather upon the accuracy of the manner in which the attestation in a source for Marcion's Gospel is presented in the apparatus of NA28.Footnote 22

4. Problematic References to Marcion's Gospel in the Apparatus

On the basis of my own work reconstructing Marcion's Gospel, along with the work done by Matthias Klinghardt and Andrea Nicolotti, it has become evident that despite numerous references to the attestation for Marcion's Gospel found in the NA28 apparatus being accurate and helpful, others are problematic. One important issue to bear in mind when examining the data, however, is that despite their valuable contributions, the reconstructions of both Klinghardt and Nicolotti draw on manuscripts of Luke (often D) in order to present wording beyond that which is attested in the heresiological sources.Footnote 23 For this reason, their works are consulted primarily for their interaction with the patristic sources and when their reconstructed Greek text is actually reflected in undisputed sources for Marcion's Gospel.Footnote 24 Bearing this point in mind, the following discussion highlights eight instances where references in the apparatus should be considered candidates for editing or removal in future Nestle-Aland editions.

(1) The first problematic reference to ‘Mcion’ is also the first appearance of an attestation of Marcion's Gospel in the apparatus of NA28. This particular reference in Luke 4.23 is a change from NA27 and seems to have come about as a further instance of desiring to provide ‘more clarity’ in the apparatus.Footnote 25 Unfortunately, in this instance it did not do so. In NA27, for the spelling Καφαρναούμ found in the main text, the apparatus refers the reader to Luke 4.31 by stating ‘v. ad vs 31’. NA28 removes this cross-reference and lists a series of witnesses, including McionT, A for Luke 4.23. The problem, however, is that even though this spelling is attested for Marcion's Gospel in Luke 4.31, no source attests Marcion's text for Luke 4.23. Despite the fact that Ψ 044 is a much later text, it is striking that this manuscript reads Καφαρναουμ in 4.23 but Καπɛρναουμ in 4.31, demonstrating that though it may be unusual, it is not impossible for the same manuscript to attest variant orthography. Given the absence of any evidence for the reading in Marcion's Gospel in Luke 4.23, the orthography here is unknown and McionT, A should not appear in the apparatus as evidence for the reading in the main text.

(2) In Luke 5.24, for the phrase ἔγɛιρɛ καὶ ἄρας τὸ κλινίδιόν σου πορɛύɛυ ɛἰς τὸν οἶκόν σου, the Nestle-Aland draws attention to three variants, each of them involving McionT: the reading ἆρον instead of ἄρας, the reading κράβαττον instead of κλινίδιον and the insertion of καί before πορɛύɛυ. NA27 combined the first and last variant into one entry (⸀1 αρον et ⸆και), which in NA28, once again probably in the pursuit of ‘greater clarity’, is split into two separate entries. By listing all the entries separately, however, attention is drawn to a problem in listing McionT as a witness for all three variants. Though Nicolotti claims that ἆρον τὸν κράβαττόν σου καί is ‘confermato da Tert’,Footnote 26 it is actually not the case that Tertullian attests the καί at the end of the phrase.Footnote 27 The entirety of Tertullian's citation of this section of the verse reads Exurge, et tolle grabattum tuum (Marc. 4.10.1),Footnote 28 and therefore, though McionT can be marshalled as a witness for the readings ἆρον and κράβαττον, it should not be listed as a witness for the reading with καί before πορɛύɛυ.

(3) McionT appears twice in Luke 6.9, with the first reference indicating that Tertullian's attestation of Marcion's text evidences the presence of ɛἰ before ἔξɛστιν. Though Harnack reconstructed the beginning of Jesus’ question in Marcion's Gospel in this manner,Footnote 29 Tertullian's reference to what Jesus asks begins simply with licetne sabbatis benefacere, an non? (Marc. 4.12.11).Footnote 30 Nicolotti does not place ɛἰ in bold in his text and Klinghardt explicitly opines that ɛἰ is not attested by Tertullian.Footnote 31 At the very least, I have argued that on the basis of Tertullian's attestation, ‘it is unclear whether Marcion read ɛἰ’,Footnote 32 and I would therefore contend that McionT should not appear here in the apparatus.

(4) In Luke 6.38, McionT is listed as supporting the reading τω γαρ αυτω μɛτρω ᾧ. Though the reading attested by Tertullian clearly has points of contact with this reading, his reference to the relevant section of text reads eadem, qua mensi eritis, mensura … (Marc. 4.17.9).Footnote 33 It is curious that though the apparatus indicates that P45 Θ f 13 and 700 do not attest γάρ, it does not note the fact that McionT does not attest the conjunction either. In addition, the fact that Tertullian attests an alternative word order reveals that the reference should be placed in parentheses in the apparatus, since this is the manner in which NA28 presents patristic evidence that ‘supports the given reading, but with some slight variation’Footnote 34 and the manner in which slightly variant readings are referenced in other instances.Footnote 35

(5) Luke 10.27 is a further instance in which parentheses should be placed around the reference to McionT. As it is, the apparatus lists ‘D it; McionT’ as attesting the reading ɛν ολη τη ψυχη σου και ɛν ολη τη ισχυι σου. Though Tertullian, like D, does not include the final element in the sequence (καὶ ἐν ὅλῃ τῇ διανοίᾳ σου), the citation of this section of the verse reads et ex tota anima tua et totis viribus tuis (Marc. 4.25.15).Footnote 36 Not only does Tertullian use the preposition ex (d reads in), he also does not repeat the preposition. Furthermore, Tertullian's ex toto corde is not listed in the apparatus as supporting the reading ɛν ολη τη καρδια in D. As in Luke 6.38, the variation in the text attested by Tertullian means that the apparatus should read (McionT).

(6) The next problematic reference is an instance in which the apparatus does use parentheses; however, in Luke 11.2, as I have argued extensively elsewhere, even the ‘qualified’ reference of (McionT) does not alleviate the difficulty regarding how Tertullian is listed as a witness.Footnote 37 First, though the petition for the Holy Spirit brings Marcion's text into contact with other readings attesting such a petition, it is notable that Tertullian's a quo spiritum sanctum postulem (Marc. 4.26.4)Footnote 38 has only minimal points of contact with the reading ɛλθɛτω το πνɛυμα σου το αγιον ɛφ ημας και καθαρισατω ημας of 700 and GrNy printed in the apparatus or with the related reading ɛλθɛτω σου το πνɛυμα το αγιον και καθαρισατω ημας in 162.Footnote 39 Second, and more significantly, locating McionT here obscures the fact that the apparent petition for the Holy Spirit is attested in a different position in Marcion's version of the Lord's Prayer and leads to the apparatus incorrectly implying that it replaces the petition for the kingdom as in the other witnesses.Footnote 40 In fact, the petition for the kingdom is attested by Tertullian for Marcion's Gospel as in Marc. 4.26.4, following the text quoted above, Tertullian's question concerning the ensuing petition in the prayer begins with eius regnum optabo venire ….Footnote 41 Though one can recognise and appreciate the challenge of presenting the testimony to Marcion's text in an apparatus, the manner in which it is currently included in NA28 is more misleading than helpful.

(7) Luke 16.25 is an instance in which the apparatus listing McionA, E as supporting the reading ὅδɛ instead of ὧδɛ is complicated by the reality that considering readings in Marcion's Gospel involves not only the sources for this text but also the sources for the sources. That is to say, the texts of the church fathers found in critical editions and used as sources for Marcion's Gospel are themselves based on manuscript sources. Though Klinghardt presents both Epiphanius and the Adamantius Dialogue as reading ὅδɛ,Footnote 42 Harnack had already noted that the manuscript evidence for these texts is not consistent,Footnote 43 a point most recently reiterated by Nicolotti.Footnote 44 The fact of the matter is that, on the one hand, the Greek manuscripts of the Adamantius Dialogue are not uniform, with some reading ὧδɛ and others ὅδɛ and, on the other hand, no manuscript of Epiphanius’ Panarion reads ὅδɛ.Footnote 45 Concerning the Adamantius Dialogue, Rufinus’ Latin translation reads hic,Footnote 46 leading Harnack to conclude, ‘die schwankende LA ὅδɛ oder ὧδɛ wird für den Dialog durch Rufin (ὅδɛ) entschieden’.Footnote 47 W. H. van de Sande Bakhuyzen's 1901 critical edition does read ὅδɛ,Footnote 48 but more recently Kenji Tsutsui's 2004 work presents a critical text reading ὧδɛ.Footnote 49 Concerning the Panarion, the reading ὧδɛ in Pan. 42.11.6 μɛ (45) // 42.11.17 Σχ. μɛ (45) was emended in the Karl Holl and Jürgen Dummer edition to ὅδɛ with reference to Theodor Zahn.Footnote 50 Zahn had argued, ‘Auch Ep. sch. 45 bestätigt οδɛ, obwohl es dort nicht uberliefert ist; denn wozu sonst hätte Ep. geschrieben ωδɛ (lies οδɛ) παρακαλɛιται, ο αυτος Λαζαρος, als um das aus dem Zusammenhang gerissene οδɛ zu deuten?’Footnote 51 Similarly, Harnack contended that ὅδɛ ‘muß auch für Epiph. nach seinem erklärenden Zusatz hergestellt werden’.Footnote 52 As I have argued elsewhere, however, this argument ‘is problematic as the reading ὅδɛ is not absolutely necessary to explain Epiphanius's remark. The lack of an external subject for the verb could just as easily have led to the clarification.’Footnote 53 At the very least, recent work on Marcion's Gospel has highlighted the problem in ascertaining the reading attested in the sources for Marcion's Gospel at Luke 16.25, thus revealing that the inclusion of McionA, E in the apparatus here should at the very least be revisited.

(8) The final verse to consider is also the only instance in which the apparatus refers to a reading attested by all three primary witnesses for Marcion's Gospel. Unfortunately, however, the attestation is not quite correct. At Luke 24.25, the apparatus provides the following reference for the readings replacing ἐλάλησαν οἱ προφῆται printed in the main text: ɛλαλησɛν (ɛλαλησα McionA, E) προς υμας McionT, A, E. Though Tertullian attests ἐλάλησɛν πρὸς ὑμᾶςFootnote 54 and the Adamantius Dialogue ἐλάλησα πρὸς ὑμᾶς,Footnote 55 Epiphanius actually attests ἐλάλησα ὑμῖν.Footnote 56 The apparatus to this verse should be updated since there is in fact no point at which all three sources agree.

5. Conclusion and Direction for Future Work

Though only time will tell how significant the recent resurgence in scholarship on Marcion's Gospel will ultimately be, revisiting the reconstruction of Marcion's Gospel, at the very least, allows for a critical revisiting of the apparatus in NA28. Though in many instances the references to Marcion's Gospel attested in various sources provide helpful and important insight, in at least eight instances these references are problematic. For this reason, and on the basis of the discussion above, the reference to the attestation for Marcion's Gospel in Luke 4.23, the third reference in 5.24, the first reference in 6.9 and perhaps the reference in 16.25 should all be considered as candidates for removal from the apparatus. In addition, in Luke 6.38; 10.27; 11.2; and 24.25, the references should be revisited and edited in order to avoid a misleading presentation of the evidence for a particular reading.

Beyond revisiting references currently found in the apparatus, renewed work on Marcion's Gospel may also offer, on the one hand, an opportunity to revisit the broader rationale for when to include a reference to ‘Mcion’ as a type of ‘patristic witness’ in the Nestle-Aland apparatus. For instance, it seems that the omission of Luke 17.10b, for which the NA28 apparatus lists no witnesses other than McionE, is the only one of numerous instances in which Epiphanius accuses Marcion of having excised elements from his Gospel that has been included in the Nestle-Aland apparatus.Footnote 57 Should other such attestations of ‘omissions’ be included? Or should this attestation be removed? Are there additional instances in which readings could or should be removed or added? On the other hand, the careful and critical work done on Marcion's Gospel over the past few years may raise the question of whether or not a reconstruction of this Gospel text could appear in the apparatus as a ‘text’, particularly in instances with secure readings. In sum, revisiting both specific references in the current apparatus and the broader issue of how and why the attestation for Marcion's Gospel appears would allow the apparatus in future editions of the Nestle-Aland Novum Testamentum Graece to guide students and scholars more helpfully and accurately with regard to the text attested for Marcion's Gospel.

References

1 Recent works presenting a Greek reconstruction of the text of Marcion's Gospel include M. Klinghardt, Das älteste Evangelium und die Entstehung der kanonischen Evangelien, 2 vols. (TANZ 60; Tübingen, Francke: 2015, 20202); idem, The Oldest Gospel and the Formation of the Canonical Gospels, 2 vols. (BTS 41; Leuven: Peeters, 2021); Roth, D. T., The Text of Marcion's Gospel (NTTSD 40; Leiden: Brill, 2015)CrossRefGoogle Scholar; and Gianotto, C. and Nicolotti, A., Il Vangelo di Marcione (Nuova Universale Einaudi 22; Torino: Giulio Einaudi, 2019)Google Scholar. An English text can be found in BeDuhn, J., The First New Testament: Marcion's Scriptural Canon (Salem: Polebridge, 2013)Google Scholar, and an Italian text in Gramaglia, P. A., Marcione e il Vangelo (di Luca). Un confronto con Matthias Klinghardt (Collana di studi del Centro interdipartimentale di scienze religiose Università di Torino 7; Torino: Accademia University Press, 2017)Google Scholar.

2 J. K. Elliott's observation is indicative of this recognition at the turn of the century: ‘we ought to work more systematically on the writings of Marcion and Irenaeus to learn what they can reveal about the Biblical texts and specifically the New Testament text-types which they were using and quoting’ (‘The New Testament Text in the Second Century: A Challenge for the Twenty-First Century’, NTTRU 8 (2000) 1–14, at 12). For additional references to both NT and Patristic scholars highlighting the importance of work on Marcion's Gospel, see Roth, Marcion's Gospel, 1–3.

3 I initially mentioned this point in the final chapter of my own work on Marcion's Gospel, a chapter in which I offered a few reflections upon avenues for future research (see Roth, Marcion's Gospel, 439).

4 See the forthcoming study by G. S. Paulson on the history of the Nestle-Aland Novum Testamentum Graece, to be published by the German Bible Society.

5 I have discussed this example previously in Roth, D. T., ‘Marcion and the Early Text of the New Testament’, The Early Text of the New Testament (ed. Hill, C. E. and Kruger, M. J.; Oxford: Oxford University Press, 2012) 302–12CrossRefGoogle Scholar, at 310–11.

6 See von Harnack, A., Marcion: Das Evangelium vom fremden Gott. Eine Monographie zur Geschichte der Grundlegung der katholischen Kirche (TUGAL 45; Leipzig: J. C. Hinrich, 1921) 220* (cf. 229*)Google Scholar.

7 Both citations are found in A. von Harnack, Marcion: Das Evangelium vom fremden Gott. Eine Monographie zur Geschichte der Grundlegung der katholischen Kirche (TUGAL 45; Leipzig: J. C. Hinrich, 19242) 238* (cf. 247*). As I have observed elsewhere, ‘Luke 5:39, 22:43, 24:12, and 24:40 all involve instances where the Marcionite text is unattested in the sources, but Harnack believed Marcion excised the passages for dogmatic reasons’ (Roth, Marcion's Gospel, 26).

8 See F. Neirynck, ‘Lc. xxiv 12. Les témoins du texte occidental’, Miscellanea Neotestamentica i (ed. T. Baarda, A. F. J. Klijn and W. C. van Unnik; NovTSup 47; Leiden: Brill, 1978) 45–60, at 52.

9 BeDuhn rightly observes that the verse ‘is unattested’ and adds that it is ‘generally thought to be secondary in Luke’ (First New Testament, 195).

10 See more recently the assessment of B. Ehrman who, though viewing the verse as inauthentic, concluded that Neirynck ‘has convincingly shown that Marcion … cannot be cited in support of the Western text here’ (The Orthodox Corruption of Scripture: The Effect of Early Christological Controversies on the Text of the New Testament (Oxford: Oxford University Press, 20112) 298 n. 129). It seems to me that Ehrman's position remains correct despite recent publications suggesting or even definitively concluding that the verse was not present in Marcion's Gospel (e.g. Klinghardt, Das älteste Evangelium, ii.1130; J. M. Lieu, ‘Marcion and the New Testament’, Method & Meaning: Essays on New Testament Interpretation in Honor of Harold W. Attridge (ed. A. B. McGowan and K. H. Richards; SBLRBS 67; Atlanta: SBL, 2011) 399–416, at 413; eadem, Marcion and the Making of a Heretic: God and Scripture in the Second Century (Cambridge: Cambridge University Press, 2015) 218; and Smith, D. A., ‘Marcion's Gospel and the Resurrected Jesus of Canonical Luke 24’, ZAC 21 (2017) 41–62, at 50)CrossRefGoogle Scholar.

11 The asterisk following the siglum for the manuscript, as noted in the introduction to the Nestle-Aland edition, ‘identifies the original reading when a correction has been made’ (NA27, 54*). The ensuing section of this article briefly explains and discusses the siglum Mcion.

12 See Roth, Marcion's Gospel, 152 n. 312.

13 As pointed out in the conclusion of this article, such a reference occurs in Luke 17.10b.

14 See Roth, Marcion's Gospel, 313–14. The discussion of Luke 6.38 below briefly explains the use of parentheses in the apparatus.

15 Though the major revisions in NA28 involve the Catholic letters, for which the Editio Critica Maior was already available, the ‘Vorwort’/‘Foreword’ also states that a revision of the entire apparatus was undertaken. The website of the Deutsche Bibelgesellschaft notes, concerning the first of two tasks that the NA28 was to accomplish, that ‘the apparatus had to be revised thoroughly to give it more clarity and make it easier to use’ (‘The Novum Testamentum Graece [Nestle-Aland] and its history’, www.academic-bible.com/en/bible-society-and-biblical-studies/scholarly-editions/greek-new-testament/nestle-aland).

16 NA28, 81* (English) 38* (German).

17 For discussion of the sources for Marcion's Gospel, see the overview in Roth, Marcion's Gospel, 46–7 along with the critical analysis in D. T. Roth, ‘The Link between Luke and Marcion's Gospel: Prolegomena and Initial Considerations’, Luke on Jesus, Paul and Christianity: What Did He Really Know? (ed. J. Verheyden and J. Kloppenborg; BTS 29; Leuven: Peeters, 2017) 59–80, at 61–8.

18 Of the 486 verses attested as present in Marcion's Gospel, Tertullian attests 438 (for 328 he is the sole witness). Epiphanius provides data for readings in 114 verses and the Adamantius Dialogue contains seventy-five verses that at least ought to be considered as possibly witnessing Marcion's Gospel. For details of the statistical analysis, see Roth, Marcion's Gospel, 86, 271–2, 355–6. Tables providing an overview of all verses attested as present, verses attested as absent and unattested verses for Marcion's Gospel can be found in Roth, Marcion's Gospel, 49–78.

19 Luke 4.23, 31, 41; 5.14, 24, 34; 6.9, 20, 21, 22, 25, 26, 31, 36, 37, 38; 8.3; 9.35; 10.11, 21, 22, 24, 27; 11.2, 4, 38, 41, 48; 12.1, 5, 14, 27, 39, 51; 16.12, 17, 18; 19.26; 20.36; 21.19; 24.6, 25, 39.

20 Luke 5.14; 6.5, 16; 7.36, 38; 8.46; 9.16, 22, 35; 10.21; 11.11; 12.5, 31; 16.25; 17.10; 22.14; 23.2, 45; 24.6, 25, 39.

21 Luke 4.23, 31; 5.38; 9.1, 2, 6, 22; 12.47; 16.20, 22/3, 25, 31; 24.25.

22 For instance, there are, in my estimation, serious questions concerning the accuracy of many of the attested readings for Marcion's Gospel in the Adamantius Dialogue. Nevertheless, the apparatus for Luke 5.38, for example, correctly presents that which is attested in this source, even if it may be tenuous to view the Adamantius Dialogue as correctly representing the reading in Marcion's Gospel for this verse (see Roth, Marcion's Gospel, 357–8, 359–60). In addition, since Tertullian rarely mentions explicitly that Marcion omitted an element from a verse and, on occasion, abbreviates citations himself, the fact that Tertullian does not include νῦν in his reference to Luke 6.21 or μου in Luke 10.22 does not necessarily indicate that it was not present in Marcion's text (see Roth, Marcion's Gospel, 100, 133). Though one could, therefore, in these instances and others debate whether Tertullian is actually attesting the omission, I have also not pursued such debates here.

23 For critical reflection on Klinghardt's approach along with a variety of issues related to reconstructing Marcion's Gospel, see Roth, D. T., ‘Marcion's Gospel and the History of Early Christianity: The Devil Is in the (Reconstructed) Details’, ZAC 21 (2017) 25–40CrossRefGoogle Scholar; idem, ‘The Link between Luke and Marcion's Gospel’, 59–80; and idem, ‘Marcion's Gospel and the Synoptic Problem in Recent Scholarship’, Gospel Interpretation and the Q-Hypothesis (ed. M. Müller and H. Omerzu; LNTS 573; London: T&T Clark, 2018) 267–83.

24 Klinghardt presents ‘wörtlich genau gesicherte Passagen’ in bold and underlined text (Das älteste Evangelium, ii.450). Nicolotti, who is responsible for the Greek text in the volume published in collaboration with Gianotto, indicates that bold text ‘identifica parti del testo sicuramente o molto probabilmente presenti nel Mcn, in quella forma o in una molto simile, perché citate da qualche autore antico, letteralmente o quasi, come facenti parte di quel Vangelo’ (Il Vangelo di Marcione, cxiv–cxv).

25 See the comments on Luke 16.20 above.

26 Nicolotti, Il Vangelo di Marcione, 16.

27 See Harnack, Marcion, 189*; Klinghardt, Das älteste Evangelium, ii.491; and Roth, Marcion's Gospel, 413.

28 The reference is embedded in an argument involving statements by the prophet Isaiah and is found in Tertullian's observations: quoniam cum redintegratione membrorum virium quoque repraesentationem pollicebatur: Exurge, et tolle grabattum tuum, simul et animi vigorem, ad non timendos qui dicturi errant. The Latin text here and throughout is cited from volume iv of Contre Marcion (critical text, ed. C. Moreschini and trans. R. Braun; SC 456; Paris: Cerf, 2001). In every instance, I have compared the critical text and apparatus with Quinti Septimi Florentis Tertulliani Opera (ed. E. Kroymann; CSEL 47; Vienna: F. Tempsky, 1906) 290–650; Quinti Septimi Florentis Tertulliani Opera (rev. edn of the Kroymann text by E. Dekkers; 2 vols.; CCSL 1, 2; Turnhout: Brepols, 1954) i.441–726; and Tertulliani Adversus Marcionem (ed. C. Moreschini; TDSA 35; Milan: Instituto Editoriale Cisalpino, 1971).

29 See Harnack, Marcion, 190*.

30 Tertullian writes: in hunc ergo sensum legis inducere volens illos per manus arefactae restitutionem interrogat: Licetne sabbatis benefacere, an non? In the CCSL (Dekkers) edition, the question is found at the end of Marc. 4.12.10 (i.571).

31 See Nicolotti, Il Vangelo di Marcione, 24 and Klinghardt, Das älteste Evangelium, ii.520, where he notes ‘ɛι/si: om Tert.’

32 Roth, Marcion's Gospel, 196.

33 The entirety of Tertullian's citation of Luke 6.38 reads: date et dabitur vobis. mensuram bonam, pressam ac fluentem dabunt in sinum vestrum. eadem, qua mensi eritis, mensura remetietur vobis.

34 NA28, 79*.

35 See, among several examples, Luke 9.22, where the apparatus references the attestation for the reading μɛθ ημɛρας τρɛις as Mcion(E) since Epiphanius attests μɛτὰ τρɛῖς ἡμέρας (Pan. 42.11.6, 17), or Luke 11.38, where (McionT) references the ‘Western’ reading ηρξατο διακρινομɛνος ɛν ɛαυτω λɛγɛιν δια τι since Tertullian attests retractabat penes se cur (Marc. 4.27.2).

36 In the CCSL (Dekkers) edition, the citation is found at the end of Marc. 4.25.14 (i.614).

37 See D. T. Roth, ‘The Text of the Lord's Prayer in Marcion's Gospel’, ZNW 103 (2012) 47–63, at 54–9.

38 In the CCSL (Dekkers) edition, the question is found at the end of Marc.4.26.3 (i.615).

39 Because of the slight difference in the manuscript's reading, the NA28 apparatus rightly places 162 in parentheses. The fact that Tertullian references a petition for the Holy Spirit in Marcion's Gospel has often led to the postulate that Marcion's text read very similarly to the other witnesses (see e.g. Harnack, Marcion, 207*, and more recently Klinghardt, Das älteste Evangelium, ii.725 and Nicolotti, Il Vangelo di Marcione, 88). Though this supposition is understandable, I remain convinced that my conclusion concerning Harnack's reconstruction applies in all relevant instances: ‘Harnack's reconstruction rightly recognizes that Marcion's text had some type of request for the Spirit in place of the first petition, but moves far beyond what the evidence allows as to the suggested wording of that request’ (Roth, ‘Lord's Prayer’, 57).

40 Independently of my own work, Judith Lieu also noted that ‘NA27 wrongly gives the impression that it [the petition for the Holy Spirit in Marcion's Gospel] replaces the petition for the coming of God's kingdom as in MS 700 and Gregory of Nyssa’ (‘Marcion and the Synoptic Problem’, New Studies in the Synoptic Problem: Oxford Conference, April 2008. Essays in Honour of Christopher M. Tuckett (ed. P. Foster, A. Gregory, J. S. Kloppenborg and J. Verheyden; BETL 239; Leuven: Peeters, 2011) 731–51, at 738 n. 17).

41 The petition ἁγιασθήτω τὸ ὄνομά σου is not attested for Marcion's text.

42 Klinghardt, Das älteste Evangelium, ii.877.

43 Harnack, Marcion, 221*.

44 Nicolotti, Il Vangelo di Marcione, 134: ‘C’è incertezza fra i mss di Epif e Adam tra il meglio attestato ὧδɛ e il sinonimo ὅδɛ.’

45 See the notes in the W. H. van de Sande Bakhuyzen edition Pseudo-Origen: Der Dialog des Adamantius. ΠΕΡΙ ΤΗΣ ΕΙΣ ΘΕΟΝ ΟΡΘΗΣ ΠΙΣΤΕΩΣ (GCS 4; Leipzig: J. C. Hinrich, 1901) 76 and the Holl, K. and Dummer, J. edition Epiphanius ii: Panarion haer. 34–64 (GCS 31; Berlin: Akademie Verlag, 1980) 113, 143Google Scholar.

46 See the V. Buchheit edition Tyranni Rufini librorum Adamantii Origenis Adversus haereticos interpretation (STA 1; Munich: Wilhelm Fink, 1966) 35.

47 Harnack, Marcion, 221*.

48 Bakhuyzen, Dialog des Adamantius, 76.

49 Tsutsui, K., Die Auseinandersetzung mit den Markioniten im Adamantios-Dialog: Ein Kommentar zu den Büchern i–ii (PTS 55; Berlin: de Gruyter, 2004) 328CrossRefGoogle Scholar. I also drew attention to this fact in Roth, Marcion's Gospel, 379 n. 141.

50 See Holl and Dummer, Epiphanius ii, 113, 143.

51 T. Zahn, Geschichte des neutestamentlichen Kanons (2 vols.; Erlangen: Andreas Deichert, 1888–92) ii.480.

52 Harnack, Marcion, 221*–2*.

53 Roth, Marcion's Gospel, 322 n. 181.

54 I am thankful to Simon Gathercole, who, in email correspondence in 2016, pointed out an inconsistency in my previous discussion of this verse. In the consideration of Tertullian's testimony, I erroneously presented the deponent verb in quae locutus est ad vos (Marc. 4.43.4) as attesting a Greek passive (οἷς ἐλαλήθη πρὸς ὑμᾶς), whereas, in my discussion of Epiphanius’ testimony and the testimony of the Adamantius Dialogue, I correctly presented the reading attested by Tertullian as οἷς ἐλάλησɛν πρὸς ὑμᾶς. Unfortunately, it is the erroneous reading that also appears in my reconstruction, though with the lowest level of confidence (see Roth, Marcion's Gospel, 268, 344, 392, 435). Similar errors appear in Lieu's suggestion of the translation ‘what was spoken to you’ for Tertullian's text (‘Marcion and the New Testament’, 413 n. 19) and in Nicolotti's statement that Tertullian ‘sembra attestare … ἐλαλάθη [sic] πρὸς ὑμᾶς’ (Il Vangelo di Marcione, 199).

55 See Roth, Marcion's Gospel, 392.

56 See Roth, Marcion's Gospel, 343–4. Klinghardt's reconstruction actually follows Epiphanius’ text (Das älteste Evangelium, ii.1131), which is a reading that the apparatus of NA28 does not even indicate as attested in a source for Marcion's Gospel.

57 For the list of verses that Epiphanius explicitly attests as not being present in Marcion's Gospel, see Roth, Marcion's Gospel, 75–6.