Luke 22:43–44: An Anti-Docetic Interpolation or an Apologetic Omission?1 Lincoln H. Blumell Brigham Young University Abstract: This article examines the text-critical history of Luke 22:43–44 and argues, primarily on external grounds, that it is more likely that this passage was deliber- ately excised from Luke rather than interpolated. Along these lines it argues that this passage was excised from some early manuscripts of Luke prior to the end of the third century for apologetic reasons. Additionally, this article will question the anti-docetic interpolation theory, which is seemingly held by the majority of inter- preters of this passage, and will argue that this is not the only way to understand the text-critical evidence. Introduction The textual integrity of Luke 22:43–44 has long been a matter of dispute.2 Since the pioneer- ing work of Brooke F. Westcott and Fenton J. A. Hort on the Greek New Testament at the close of the nineteenth century the authenticity of this passage has been hotly disputed.3 In fact, over 1 I would like to sincerely thank the anonymous reviewers of this article for their lucid and forth- right critique of this submission at its various stages; as a result it has been greatly improved. For journal abbreviations and abbreviations of other standard works I have followed P. H. Alexander et al. (eds.), The SBL Handbook of Style: For Ancient Near Eastern, Biblical, and Early Christian Studies (Peabody, MA: Hendrickson, 1999). For Patristic abbreviations not appearing in the SBL Handbook I have followed G.W.H. Lampe (ed.), A Patristic Greek Lexicon (Oxford: Oxford Uni- versity Press, 1961). For Latin Patristic abbreviations not appearing in the SBL Handbook I have followed H. Chirat, Dictionnaire Latin-Français des auteurs chrétiens (Paris: Librairie des Méridi- ens, 1954). English translations are my own unless otherwise noted. 2 Luke 22:43–44 (NA28): ⟦43ὤφθη δὲ αὐτῷ ἄγγελος ἀπ’ οὐρανοῦ ἐνισχύων αὐτόν. 44καὶ γενόμενος ἐν ἀγωνίᾳ, ἐκτενέστερον προσηῦχετο· καὶ ἐγένετο ὁ ἱδρὼς αὐτοῦ ὡσεὶ θρόμβοι αἵματος καταβαίνοντες ἐπὶ τὴν γῆν.⟧ (“43Then an angel from heaven appeared to him and gave him strength. 44In his anguish he prayed more earnestly, and his sweat became like great drops of blood falling down on the ground” [Luke 22:43–44 NRSV]). 3 B.F. Westcott and F.J.A. Hort, The New Testament in the Original Greek: Introduction, Appendix (London: Macmillan, 1896), appendix 64–67, argue that the passage represents an interpolation, although a very early one: “On the other hand it would be impossible to regard these verses [Luke 22:43–44] as a product of the inventiveness of scribes. They can only be a fragment from the traditions, written or oral, which were, for a while at least, locally current beside the canonical Gospels, and which doubtless included matter of every degree or authenticity and intrinsic value. These verses … may be safely called the most precious among the remains of this evangelic tradi- tion which were rescued from oblivion by the scribes of the second century.” Cf. D. B. Weiss, Das 1 2 Luke 22:43–44 the course of the past hundred-plus years this passage has been subject to dozens of studies.4 As a result, different Bible versions produced during this period have reflected the ongoing debate over this passage; some bibles have chosen to omit this passage altogether; others have chosen to place the verses in double brackets to highlight their uncertain nature; and still oth- ers have preferred to leave the passage in without any discriminating apparatus.5 In this schol- Neue Testament: Die Vier Evangelien. Band I (Leipzig, 1905), 434 does not include Luke 22:43–44 in his NT text stating: “v. 43f verurteilt ist, da Jesus weder der stärkung bedarf, noch in heißem kampfe ringt.” For earlier text critical treatments of this passage in the nineteenth century see: J.S. Porter, Principles of Textual Criticism with their Application to the Old and New Testaments (Lon- don: Simms and McIntyre, 1848), 462–64; S.P. Tregelles, The Greek New Testament, Edited from Ancient Authorities, with their Various Readings in Full, and the Latin Version of Jerome. Part II: Luke and John (London: Samuel Bagster and Sons, 1857), 357; C. Tischendorf, Novum Testamen- tum Graece. Editio Octava Critica Maior: Vol. I (Lipsiae: Giesecke & Deverient, 1869), 694–96; C.E. Hammond, Outlines of Textual Criticism Applied to the New Testament (Oxford: Clarendon Press, 1872), 103–4; F.H.A. Scrivener, A Plain Introduction to the Criticism of the New Testament (3rd ed.; Cambridge: Deighton, Bell and Co., 1883), 599–602. 4 For a detailed listing of those supporting and rejecting the authenticity of Luke 22:43–44, see R.E. Brown, The Death of the Messiah, 2 vols. (New York: Doubleday, 1994), 1.180 n. 2. For listings in earlier scholarship, see D. Crump, Jesus the Intercessor (WUNT 2/49; Tübingen: J.C.B. Mohr [Paul Siebeck], 1992), 116–17 n. 25; I. H. Marshall, The Gospel of Luke (Grand Rapids: Eerdmans, 1978), 831–82; and L. Brun, “Engel und Blutschweiss Lc 22.43–44,” ZNW (1933): 265. For the more recent studies on the subject (post Brown [1994]) see: D.C. Parker, The Living Text of the Gospels (Cambridge: Cambridge University Press, 1997), 157–59; J.M.-O’Connor, “What Really Happened at Gethsemane?” BRev 14.2 (1998): 28–39, 52; M. Patella, The Death of Jesus: The Diabolic Force and the Ministering Angel (CahRB 43; Paris: Gabalda et Cie, 1999), 9–15; B.D. Ehrman, “Text and Interpretation: The Exegetical Significance of the ‘Original’ Text,” TC: A Journal of Biblical Textual Criticism 5 (2000): 32–48; A. Fuchs, “Gethsemane: Die deuteromarkinische Bearbeitung von MK 14,32–42 par Mt 26, 36–46 par Lk 22, 39–46,” SNTNU (2000): 23–75; R.G. Regorico, “Agonía de Getsemaní,” Mayéutica 26 (2000): 377–473; C.A. Smith, “A Comparative Study of The Prayer of Gethsemane,” Irish Biblical Studies 22 (2000): 98–122; P. Murray, “The Prayers of Jesus in Luke’s Passion Narrative,” Emmanuel 107.2 (2001): 88–95, 105–6; G. Sterling, “Mors philosophi: The Death of Jesus in Luke,” HTR 94.4 (2001): 383–402; J.J. Pilch, “The Nose and Altered States of Consciousness: Tascodrugites and Ezekiel,” HvTSt 58.2 (2002): 708–20; R. Riesner, “Versuc- hung und Verklärung (Lukas 4,1–13; 9,28–36; 10,17–20; 22,39–53 und Johannes 12,20–36),” TBei 33 (2002): 197–207; C.M. Tuckett, “Luke 22, 43–44: The ‘Agony’ in the Garden and Luke’s Gospel,” in A. Denaux (ed.), New Testament Textual Criticism and Exegesis: Festschrift J. Delobel (BETL 161; Leuven: Peeters, 2002), 131–44; C. Clivaz, “The Angel and the Sweat Like ‘Drops of Blood’ (Lk 22:43–44): 𝔓69 and f13,” HTR 98.4 (2005): 419–40; B.D. Ehrman, “Did Jesus Get Angry or Agonize? A Text Critic Pursues the Original Jesus Story,” BRev 21.5 (2005): 17–26; P.W. Comfort, New Testa- ment Text and Translation Commentary: Commenting on the Variant Readings of the Ancient New Testament Manuscripts and How they Relate to the Major English Translations (Carol Stream, Ill.: Tyndale House Publishers, Inc., 2008), 233–35; T.A. Wayment, “A New Transcription of P.Oxy. 2383 (𝔓69)” NovT 50 (2008): 351–57; C. Clivaz, L’ange et la sueur de sang (Lc 22,43–44): ou comment on pourrait bien écrire l’histoire (BiTS 7; Leuven: Peeters, 2010); T.A. Wayment, “P.Oxy. 2383 (P69) One More Time,” NovT 54 (2012): 288–92; F. Bovon, Luke III: A Commentary on the Gospel of Luke 19:28–24:53 (Hermeneia; Minneapolis, 2012), 201–11. 5 The RSV omits this passage entirely. The ASV, Phillips, and NJB leave the passage as is with no discriminating apparatus. In the GNB, NAB, NASV, NIV, NKJV, NRSV, this passage is either placed in double brackets or has an accompanying footnote explaining that these verses are not found in certain early manuscripts. Both the NA27 and NA28, as well as the UBS3 and UBS4, in- clude the passage but place it in double brackets. See R.L. Omanson, A Textual Guide to the Greek New Testament (Stuttgart: Deutsche Bibelgesellschaft, 2006), 150. An Anti-Docetic Interpolation or an Apologetic Omission? 3 arly quagmire perhaps the most widely cited and influential study of Luke 22:43–44, at least in English, is that of Bart D. Ehrman and Mark A. Plunkett.6 In their examination Ehrman and Plunkett surveyed a wide variety of evidence in an effort to determine whether this passage should be considered authentic. While they admitted that the evidence did not point in one definitive direction, they argued that the passage was likely not original but represented a later interpolation.7 Notwithstanding the tentative nature of their conclusions, their article seems to have largely carried the day. If there is anything approaching a “consensus” in the field, it is that Luke 22:43–44 is not genuine but represents an interpolation.8 Leaving aside for the moment the issue of the manuscript evidence, one of the primary reasons this view has obtained a majority position is because its proponents have been able to marshal a seemingly probable explanation for why these verses might have been added to the Gospel of Luke. Ehrman and Plunkett asserted that sometime before ca. 160 CE the passage was added to Luke as anti-docetic polemic.9 More recently, Ehrman has argued that in the sec- ond century one of the greatest threats to emerging “proto orthodoxy” was Docetism, and that Luke 22:43–44 was added because it reinforced the humanity and corporality of Jesus and thus served as an evangelistic safeguard against docetic views of Jesus.10 On the other hand, those who have argued that the passage is authentic to Luke but was subsequently excised have not generally put forth a detailed explanation of how this occurred; most often this suggestion is merely mooted as a passing remark that is not thoroughly argued and is usually tangential to some other primary argument for the authenticity of these verses.11 The lone exception is the 6 B.D. Ehrman and M.A. Plunkett, “The Angel and the Agony: The Textual Problem of Luke 22:43– 44,” CBQ 45 (1983): 401–16. Potentially the most significant study may turn out to be Clivaz, L’ange et la sueur de sang, given that it is the first book-length treatment of this passage. 7 Ehrman and Plunkett, “The Angel and the Agony,” 416, state: “No one argument yields a defini- tive solution. Rather, the cumulative force of a group of arguments must be assessed, and even then the critic is left with a probability-judgment.” More recently, Ehrman has become more emphatic that this passage is in fact an interpolation. See B.D. Ehrman, The Orthodox Corrup- tion of Scripture: The Effect of Early Christological Controversies on the Text of the New Testament. Updated and with a New Afterword (2nd ed.; Oxford: Oxford University Press, 2011), 220–27. 8 The rating system employed by the UBS Greek NT is perhaps most indicative of this trend. Whereas the UBS3 places this passage in double brackets and gives it a “C” rating, signifying some degree of uncertainty about the decision, in the UBS4 the rating has now been upgraded to an “A,” signifying little or no doubt about the decision. On this point see Tuckett, “Luke 22, 43–44: The ‘Agony’ in the Garden and Luke’s Gospel,” 131 n. 2. Similarly, M.L. Soards, The Passion according to Luke: The Special Material of Luke 22 (JSNTSup 14; Sheffield: Sheffield University Press, 1987), devotes an entire monograph to Luke 22 but can no more than devote a single footnote (144–45 n. 1) to Luke 22:43–44 stating why it is not authentic and therefore dismisses it. Likewise, P.M. Miller, “The Least Orthodox Reading is to be Preferred: A New Canon for New Testament Tex- tual Criticism,” in D.B. Wallace (ed.), Revisiting the Corruption of the New Testament: Manuscript, Patristic, and Apocryphal Evidence (Grand Rapids: Kregel Publications, 2011), 60 n. 16 summarily dismisses the authenticity of Luke 22:43–44 because: “Ehrman and Plunkett have persuasively argued that these verses were added by orthodox to combat against docetic theology.” Many other similar examples could be cited. 9 Ehrman and Plunkett, “The Angel and the Agony,” 416. 10 Ehrman, The Orthodox Corruption of Scripture, 225–27. As evidence (p. 226) Ehrman points out that when Luke 22:43–44 is cited by patristic writers in the first few centuries it is typically used as part of an anti-docetic polemic; cf. Ehrman, “Did Jesus Get Angry or Agonize?,” 17–26. 11 Brown, The Death of the Messiah, 183–84, raises a few possibilities for why it might have been omitted as well as added. Similarly, J. Duplacy, “La préhistoire du texte en Luc 22:43–44,” in E.J. Epp and G.D. Fee (eds.), New Testament Textual Criticism: Its Significance for Exegesis (Oxford: 4 Luke 22:43–44 recent monograph by Claire Clivaz in which she not only argues that Luke 22:43–44 (along with Luke 23:34a) is authentic but also that it was deliberately excised as part of an anti-gnostic polemic: specifically, that non-gnostic Christians omitted the passage(s) from early manu- scripts of Luke in response to a gnostic separationist reading in which Jesus was viewed as an agonist/ἀγωνιστής (“fighter”) who struggled against the Demiurge.12 Not to take away from Clivaz’s wide-ranging study, which has much to offer and certainly presents a historically plau- sible scenario in which these verses could have been excised from certain early copies of Luke, it seems possible that there could still be other contextual possibilities that could account for their excision in some early copies of Luke. As Clivaz convincingly demonstrates that there were different hermeneutical contexts in which Luke’s passion narrative was being read in antiquity (as well as modernity), it seems readily possible that if Luke 22:43–44 was at times de- liberately excised then it could have been done for different reasons at different times.13 In fact, the evidence of Epiphanius of Salamis (treated below) who witnessed the excision of these very verses from select copies of Luke in his own day reveals that the reasons behind their excision were rooted in fourth-century problems directly arising from Arianism and emerging Nicene orthodoxy. Therefore, the present investigation is not merely intended to offer a competing narrative to Clivaz’s study that is mutually exclusive but draws on the momentum generated by her work to pursue an alternative scenario that could have arisen out of the complexity of different reading contexts in antiquity. The present analysis seeks to address this problem by setting forth a plausible contextual explanation for the omission of this passage. It will be argued that Luke 22:43–44 was first omitted from certain copies of Luke sometime after the middle of the second century and before the end of the third century and that its excision was primarily done for apologetic reasons. Certain early Christians felt that these verses were especially challenging, as is shown by early Christian commentary on them, since they seemingly depicted Jesus in a rather feeble light and were the target of a growing anti-Christian polemic. Consequently, some Christians felt it was easier to simply excise this problematic material. While this analysis can only offer a circumstantial case for why this passage might have been omitted, it is no more circumstan- tial than the widely-accepted argument that this passage represents an interpolation that was Clarendon Press, 1981), 86, proposes that perhaps the verses were first removed in Egypt in an attempt to harmonize the gospels but never moves to a fuller explanation of exactly why this may have happened. Likewise, A. von Harnack, Studien zur Geschichte des Neuen Testaments und der alten Kirche: I. Zur neutestamentlichen Textkritik (Berlin/Leipzig: de Gruyter, 1931), 88, believes these verses were omitted because they seemed offensive to some Christians since an angel should comfort Jesus and that as Lord he should agonize and bleed. Along the same lines see K. Jaroš, Das Neue Testament und seine Autoren: Eine Einführung (Cologne: Böhlau, 2008), esp. 85–87; Cf. C.S.C. Williams, Alterations to the Texts of the Synoptic Gospels and Acts (Oxford: Basil Blackwell, 1951), 7; M.-J. Lagrange, Évangile selon Saint Luc (Paris: Lecoffre, 1927), 563. Clivaz, “The Angel and the Sweat,” 439–40, concludes her article with a plea that there needs to be some contextual examination “of the historical and sociological components” that may have led certain Christians to either omit, or possibly add, these verses to select copies of the Gospel of Luke. 12 Clivaz, L’ange et la sueur de sang, 609–18. Clivaz draws upon Theodotus, the Valentinian Gnostic, whose teachings are preserved in Clement of Alexandria (Exc. 3.58.1: ὁ μέγας Ἀγωνιστής, Ἰησοῦς Χριστός) to help make this case and even goes on to argue that the same persons who omitted Luke 22:43–44 and 23:34a also may have interpolated Luke 24:51b in an effort to stop the prolifera- tion of “special revelation” to the disciples following the resurrection. 13 One of the most useful contributions of Clivaz’s study is her lengthy and lucid analysis of the dif- ferent hermeneutical contexts in which Luke’s passion narrative has been read. See Clivaz, L’ange et la sueur de sang, Part I. An Anti-Docetic Interpolation or an Apologetic Omission? 5 added to Luke as part of an anti-docetic polemic. As such, this article hopes to show that the anti-docetic argument is not the only conceivable explanation for this difficult text-critical problem. Luke 22:43–44: The Manuscript & Patristic Evidence Revisited Before attempting to explain why this passage would have been especially susceptible to excision, it is worthwhile to briefly review the manuscript and patristic evidence since it is fundamental to any text-critical assessment and is periodically misrepresented. While those who argue that this passage is not genuine will sometimes assert that the available manuscript evidence favors, or even strongly favors, the view that Luke 22:43–44 represents a later inter- polation, in actuality the current manuscript evidence is fairly even and is in no way lopsided.14 Though it could be admitted that the textual evidence is such that it slightly favors excluding the passage, overall this is not definitive as various early manuscripts both omit and include this passage.15 Most notably, this passage does not appear in 𝔓69vid, 𝔓75, ℵ2a, A, B, N, T, W, but it is attested in 0171, ℵ*, 2b, D, L, Θ, Ψ, 0233.16 Here the evidence of 0171 (=PSI II 124), which is sometimes simply unacknowledged or even misquoted,17 is very significant since this fragment represents a very early and important witness to this passage.18 In fact, in the most recent pa- 14 Comfort, New Testament Text and Translation Commentary, 233, states, “the manuscript evidence for this textual variant is decidedly in favor of the exclusion of [Luke] 22:43–44.” Cf. B.M. Metzger, A Textual Commentary on the Greek New Testament (2nd ed; Stuttgart: Deutsche Bibelgessellschaft, 2002), 151, who notes that the manuscript evidence alone “strongly suggests” that this passage was not originally part of Luke. 15 Ehrman, The Orthodox Corruption of Scripture, 220, concedes this point, “In this particular in- stance [i.e. Luke 22:43–44], the manuscript alignments prove inconclusive for resolving the tex- tual problem”; Tuckett, “Luke 23,43–44,” 132, notes, “It is widely agreed that the manuscript evi- dence alone is inconclusive, though perhaps slightly inclining in favor of omitting the verses”; cf. Brown, The Death of the Messiah, 1.181. 16 For a more expansive listing of texts and manuscripts see Clivaz, L’ange et la sueur de sang, 590. 17 J.A. Fitzmyer, The Gospel according to Luke: Introduction, Translation, and Notes (2 vols.; New York: Doubleday, 1985) 2.1443, who confuses this fragment with the twelfth-century minuscule 1071. Not surprisingly, since Fitzmyer confuses this fragment he judges that the manuscript evi- dence is such that it decidedly favors the omission of this passage. Cf. Tuckett, “The ‘Agony’ in the Garden and Luke’s Gospel,” 131–32, who corrects this mistake. J. Hernández Jr., “The Early Text of Luke,” in C.E. Hill and M.J. Kruger (eds.), The Early Text of the New Testament (Oxford: Oxford University Press, 2012), 121–39, who sets out to treat all pre-fourth century textual witnesses of Luke, completely disregards 0171 although he includes 𝔓7. Similarly, Comfort (New Testament Text and Translation Commentary, 233) cites this fragment as “0171vid” but this is not entirely ac- curate: there is nothing about v. 44 at least that ought to relegate it to the realm of uncertainty as it can clearly be detected on the fragment. See also Clivaz, “The Angel and the Sweat,” 422–24 who appropriately highlights the importance of this fragment as early evidence for Luke 22:43–44. 18 0171 is a parchment fragment that comes from Hermopolis Magna in Upper Egypt and contains portions of Matthew (10:17–23, 25–32) and Luke (22:4–50, 52–56, 61, 63–64). In 1966 K. Treu dis- covered a parchment codex fragment of Matthew 10:17–20 and 21–23 on one side and 10:25–27 and 28–32 on the other side in the Berlin Papyrussammlung. Since there were many similarities (paleographic, codicological, etc.) with the Luke fragments he determined that these fragments came from the very same codex. As a result these two fragments (of Luke and Matthew) have been treated together. See K. Treu, “Neue neutestamentliche Fragmente der Berliner Papyrus- sammlung,” APF 18 (1966): 25–28; cf. NewDocs 2.126–27. For detailed analysis of this fragment see J.N. Birdsall, “A Fresh Examination of the Fragment of the Gospel of St. Luke in ms. 0171 and an Attempted Reconstruction with Special Reference to the Recto,” in R. Gryson (ed.), Philologia Sa- 6 Luke 22:43–44 leographical assessment of this piece by Willy Clarysse and Pasquale Orsini, 0171 is dated to the late second or early third century—one of only a handful of New Testament fragments as- signed to this early period.19 Therefore, in their opinion it predates both 𝔓69 and 𝔓75, which they assign to the third century, and so our earliest extant piece of manuscript evidence for Luke 22 attests vv. 43–44!20 Similarly, the evidence presented by Sinaiticus needs to be highlighted and reevaluated. Though it is periodically minimized because the passage in question was omitted by a later corrector21—albeit some two hundred years after the manuscript was produced22—it cra: Biblische und patristische Studien für Hermann J. Frede und Walter Thiele zu ihrem siebzigsten Geburtstag (Herstellung: Beuroner Kunstverlag, Beuron, 1993), 1.212–27; M. Naldini, Documenti dell’ antichità Cristiana (Florence: Libreria Editrice Fiorentina, 1965), 16 (nos. 11–12); LDAB at: http://www.trismegistos.org/ldab/text.php?tm=61828. 19 P. Orsini and W. Clarysse, “Early New Testament Manuscripts and their Dates: A Critique of Theological Paleography,” ETL 88/4 (2012): 455, 458, esp. 466 where they note: “In only a few cases we propose an earlier date (𝔓18, 𝔓30, 𝔓64, 𝔓67, 𝔓4, 𝔓116, 0171, 0188, 0212, 0308). There are no first century New Testament papyri and only very few can be attributed to the second century (𝔓52 𝔓90 𝔓104, probably all the second half of the century) or somewhere between the late second and early third centuries (𝔓30, 𝔓64+67+4, 0171, 0212). Biblical scholars should realize that some of the dates proposed by some of their colleagues are not acceptable to Greek paleographers and papyrologists.” Cf. R.S Bagnall, Early Christian Books in Egypt (Princeton: Princeton University Press, 2009), 1–49, who earlier noted that dates proposed for certain New Testament papyri are too early and generally need to be re-evaluated. 20 I might also add here that Brent Nongbri is presently making the argument that 𝔓75 is likely a fourth-century composition. If this is the case, then the ms. evidence for the absence of Luke 22:43–44 is pushed back yet further. He currently has a forthcoming article titled “Reconsidering the Place of P.Bodmer XIV–XV (P75) in the Textual Criticism of the New Testament” where he makes a compelling case on paleographic and codicological grounds that 𝔓75 fits very well in the fourth century. He notes that paleographically the closest parallels to the text of 𝔓75 are P.Herm. 4 and 5 that can be objectively dated to the 320s. He also notes that on codicological grounds the Nag Hammadi codices, especially Nag Hammadi Codex II, is remarkably similar. I want to thank Brent Nongbri for providing me with a draft of this forthcoming article and for permission to reference it in this paper. 21 Comfort, New Testament Text and Translation Commentary, 233–34 gives the following errone- ous report about the corrector of Luke 22:43–44 in Sinaiticus: “(The first corrector of ℵ was a contemporary of the scibe who produced the manuscript of Luke; indeed, he was the diorthotes who worked on this manuscript before it left the scriptorium.) Other signs of its doubtfulness [i.e. Luke 22:43–44] appear in manuscripts marking the passage with obeli or crossing out the passage (as was done by the first corrector of ℵ).” 22 The later corrector who removed the passage by placing hooks at the beginning and closing of each line and dots over the letters belongs to group “C” of correctors who are dated roughly to the sixth to eight centuries, almost two hundred years after the manuscript was written. Then, another scribe belonging to group “C” of correctors erased the the dots and hooks, either because he felt the passage was legitimate or because he thought the text looked better without them. On the “C” correctors see J.M. Milne and T.C. Skeat, Scribes and Correctors of the Codex Sinaiticus (London: British Museum, 1938), 46–50, 65 where a seventh-century date for the “C” correctors is given. Cf. A. Myshrall, Codex Sinaiticus, Its Correctors, and The Caesarean Text of the Gospels (Dissertation, University of Birmingham, 2005), 90–92, where she notes that a seventh-century date for the “C” correctors is possible but also suggests that they may be a little earlier; on Luke 22:43–44 and Sinaiticus see pp. 564–65. See also D. Jongkind, Scribal Habits of Codex Sinaiticus (Piscataway: Gorgias Press, 2007), 10–11: “The various C correctors (C, Cb, Cc, Cc*) are all placed around the seventh century.” According to Jongkind (p. 9) it was scribe “D” who served as the διορθωτής. An Anti-Docetic Interpolation or an Apologetic Omission? 7 was nevertheless included by the original scribe (A) of Luke and as such is part of the text of Sinaiticus: thus, Sinaiticus is a witness of this passage. Likewise, the evidence of Codex Al- exandrinus should at least have a caveat when it is cited as a witness against this passage; the scribe placed the Eusebian canon 10.283 (σπγ/ι), which corresponds to Luke 22:43–44, in the margin at the end of Luke 22:42 and otherwise suggests that the scribe was aware of the passage (i.e. Luke 22:43–44) even if it was not included in the manuscript.23 Furthermore, it should be noted that if 𝔓69 does indeed preserve a fragment from a Marcionite recension of Luke, which has recently been argued, then the text-critical weight of this witness must surely be reevaluat- ed.24 Finally, the argument that Luke 22:43–44 is to be regarded as spurious because in certain manuscript families, such as f13, it has been transferred so that it follows Mathew 26:39 and therefore suggests that it originated from a non-Lukan source,25 can no longer be given much credence; Clivaz has convincingly demonstrated that the transfer only establishes the influ- ence of the liturgy on the textual tradition of this manuscript family and not the non-Lukan origin of these verses.26 Turning to the patristic evidence for the first four centuries, which is sometimes down- played or even ignored in certain text-critical assessments, while there is some disparity, Luke 22:43–44 is known by a number of early and important Christian writers with geographic distribution all over the Mediterranean.27 The first to reference these verses is Justin in his Dialogue with Trypho (ca. 155 CE). Here Justin remarks, “For in the memoirs which I say were drawn up by his apostles and those who followed them, [it is written] that ‘His sweat fell down like drops of blood’ while he was praying, and saying, ‘[Father] if it be possible, let this cup pass.’”28 Granted that Justin does not specifically point out that this passage was from Luke, the reference to the “memoirs” that were written by the “apostles” should point in this direction.29 23 The Eusebian canon σπγ/ι appears in the top left margin of the left page of the codex at the end of Luke 22:42. For an image of this page see Facsimile of the Codex Alexandrinus: New Testament and Clementine Epistles (London, 1887), fol. 63. There is also a deliberate space between v. 42 and v. 45. 24 Clivaz, “The Angel and the Sweat Like ‘Drops of Blood’ (Lk 22:43–44): 𝔓69 and f13,” esp. 425–32; Clivaz, L’ange et la sueur de sang, 460–67. 25 Comfort, New Testament Text and Translation Commentary, 234; Omanson, A Textual Guide to the Greek New Testament, 150; Metzger, A Textual Commentary on the Greek New Testament, 151. 26 Clivaz, “The Angel and the Sweat,” 432–38; Clivaz, L’ange et la sueur de sang, 490–99. 27 For useful treatments of Luke 22:43–44 in patristic literature see: J. Duplacy, Études de critique textuelle du Nouveau Testament (BETL 78; Leuven: Leuven University Press, 1987), 349–85; J. Du- placy, “La préhistoire du texte en Luc 22:43–44,” 77–86; Clivaz, L’ange et la sueur de sang, 456–588, who surveys a wide variety of evidence (patristic, Jewish, apocryphal); Bovon, Luke III: A Com- mentary on the Gospel of Luke 19:28–24:53, 204–11, who surveys select patristic and even Byzan- tine and medieval commentary on the passage. Cf. The American and British Committees of the International Greek New Testament Project (eds.), The Gospel according to St Luke Parts 1 and 2 (The New Testament in Greek, 3; Oxford: Clarendon, 1984, 1987), 2.190 is useful but is not com- prehensive. 28 Dial. 103.8: ἐν γὰρ τοῖς ἀπομνημονεύμασιν, ἅ φημι ὑπὸ τῶν ἀποστόλων αὐτοῦ καὶ τῶν ἐκείνοις παρακολουθησάντων συντετάχθαι, 〈γέγραπται〉 ὅτι ἱδρὼς ὡσεὶ θρόμβοι κατεχεῖτο, αὐτοῦ εὐχομένου καὶ λέγοντος· 〈Πάτερ,〉 παρελθέτω, εἰ δυνατόν, τὸ ποτήριον τοῦτο·. Greek text taken from M. Marcovich (ed.), Iustini Martyris Dialogus Cum Tryphone (PTS 47; Berlin: Walter de Gruyter, 1997), 249. 29 Cf. Justin, 1 Apol. 66.3; G.N. Stanton, “The Fourfold Gospel,” NTS 43 (1997): 330, notes that this was Justin’s way of referring to the Gospels; cf. C.E. Hill, Who Chose the Gospels? Probing the Great Gospel Conspiracy (Oxford: Oxford University Press, 2010), 123–50. A. Gregory, The Recep- tion of Luke and Acts in the Period before Irenaeus (WUNT 169; Tübingen: Mohr Siebeck, 2003), 8 Luke 22:43–44 Following Justin there is the evidence from Irenaeus. In a section of his Against Heresies (ca. 180 CE) where he criticizes Christians who denied that Jesus actually assumed flesh and expe- rienced a fully human existence, he remarks among other things that he “sweated great drops of blood.”30 To the second-century evidence of Justin and Irenaeus, Tatian could also be ad- duced, albeit with some caveats, since his Diatessaron is only extant in much later translations that may or may not be accurate renderings of the original composition.31 One other second-century writer who is occasionally cited in this debate is Marcion. In fact, he is sometimes even cited as a witness against the authenticity of these verses since it is alleged that they did not appear in his recension of Luke.32 However, invoking Marcion in such a way goes well beyond his capacity as a witness. Notwithstanding Adolf von Harnack’s eru- dite reconstruction of Marcion’s edition of Luke, it must be remembered that Marcion’s actual text of Luke is no longer extant, and that Harnack’s reconstruction is hypothetical in places.33 Nevertheless, even if one were to concede that Marcion’s version of Luke likely did not contain Luke 22:43–44, how exactly would this constitute evidence for the non-Lukan origin of this passage since it is well known that Marcion’s version of Luke had excised considerable material 280–82 raises the possibility that the phrase “those who followed them” might be taken as a refer- ence to Luke 1:3 so that Justin is referring to Luke’s Gospel. 30 Irenaeus, Haer. 3.22.2 (PG 7.1. 957): sudasset globos sanguinis/ἵδρωσε θρόμβους αἵματος. Interest- ingly, as all of the examples Irenaeus gives for Jesus’ humanity in this section of his treatise are otherwise scriptural proof texts (e.g. John 4:6; Ps 68[69]:27; John 11:35; Matt 26:38; John 19:34), it is therefore probable that when Irenaeus mentions that Jesus “sweated great drops of blood” he was not talking about some oral story but had in mind a scriptural source. 31 The earliest Latin copy of the Diatessaron is contained in Codex Fuldensis and dates to the middle of the sixth century. See (editio princeps) E. Ranke, Codex Fuldensis (Marburg etc.: Elwert, 1868), 146: apparuit autem illi angelus de caelo confortans eum et factus est in agonia et prolxius ora- bat … et factus est sudor eius sicut guttae sanguinis decurrentis in terram (Luke 22:43–44). Cf. ANF2 9.117; see also C. McCarthy, Saint Ephrem’s Commentary on Tatian’s Diatessaron: An English Translation of Chester Beatty Syriac MS 709 with Introduction and Notes (JSS Supplement 2; Ox- ford: Oxford University Press on Behalf of the University of Manchester, 1993), 297 (= Commen- tary on Tatian’s Diatessaron 20.11). 32 Comfort, New Testament Text and Translation, 233; Metzger, A Textual Commentary on the Greek New Testament, 151; Ehrman and Plunkett, “The Angel and the Agony,” 402. 33 A. von Harnack, Marcion: Das Evangelium vom fremden Gott (Leipzig: J.C. Hinrichs, 1924), 177– 255; cf. A. von Harnack, Marcion: The Gospel of the Alien God, trans. J.E. Seely and L.D. Bierma. (Durham, N.C.: The Labyrinth Press, 1990). For a useful critique of Harnack’s reconstruction see D.T. Roth, “Marcion and the Early New Testament Text,” in C.E. Hill and M.J. Kruger (eds.), The Early Text of the New Testament (Oxford: Oxford University Press, 2012), 302–12. See also L.E. Wright, Alterations of the Words of Jesus as Quoted in the Literature of the Second Century (Cam- bridge: Harvard University Press, 1952), 128–34; Williams, Alterations to the Texts of the Synoptic Gospels and Acts, 10–18; E.C. Blackman, Marcion and His Influence (London: SPCK, 1948), 50–54, 57–60, 128–71. What little we know about Marcion’s version of Luke is drawn principally from Epiphanius and also from a few a quotations from Tertullian and Adamantius, and none of these yields any definitive insight about the status of Luke 22:43–44. In Pan. 42.11.1– 11.78 Epiphanius outlines the contours of Marcion’s recension of Luke but does not reproduce the treatise verba- tim. While Epiphanius comments on Marcion’s rendering of Luke 22:41 and then 22:47b (Pan. 42.11.65–66), it cannot be automatically assumed that Marcion’s version did not contain anything from Luke 22:42–47a on this evidence alone. An Anti-Docetic Interpolation or an Apologetic Omission? 9 from Luke?34 Therefore, Marcion can hardly be invoked as a compelling witness against the authenticity of this passage.35 Proceeding to third-century writers, it seems from the writings of a select few authors that they do not seem to be aware of the passage and the inference that could be drawn is that it was not in their copies of Luke. However, it needs to be noted from the outset that just because 34 While Harnack believed that Luke 22:43–44 was not in Marcion’s recension of Luke he also point- ed out that Marcion would have had good reason to omit this material had it been present: Har- nack, Marcion, 234*; cf. Clivaz, “The Angel and the Sweat,” 429–30. Furthermore, Harnack had argued elsewhere for the authenticity of Luke 22:43–44. See A. von Harnack, “Probleme im Texte der Leidensgeschichte Jesu” in Studien zur Geschichte des neuen Testaments und der alten Kirche, vol. 1 (Berlin: De Gruyter, 1931), 86–99. See also Tertullian, Praescr 38.9 (CCSL 1.219): Marcion enim exerte et palam machaera, non stilo usus est, quoniam ad materiam suam caedem scriptu- rarum confecit. (“Marcion expressly and openly used the knife, not the pen, since he made such an excision of the Scriptures as suited his own subject matter”); and P. Lampe, From Paul to Val- entinus: Christians at Rome in the First Two Centuries, trans. Michael Steinhauser (Minneapolis: Fortress Press, 2003), 252–53, notes that Marcion seems to have almost exclusively excised text and that there is very little evidence that he made any additions to Luke. If Luke 22:43–44 was original it is more likely than not that Marcion would have omitted such material in his version of Luke since it would not have adhered to his theology. In Pan. 49.61.2 when Epiphanius was discussing Luke 22:43–44 and criticizing the Arians he noted that this passage opposed Marcion- ite and Manichaean theology since the verses stressed the humanity of Christ (GCS 3.209): καὶ οὐκ ἴσασιν ὅτι ἐὰν μὴ ἔχῃ πάντα ταῦτα καὶ τὸ «μὴ τὸ ἐμὸν βούλημα, ἀλλὰ τὸ σὸν» καὶ ἐὰν μὴ ἀγωνιάσῃ καὶ ἐὰν μὴ ἱδρὼς αὐτῷ γένηται ἐκ σώματος προχεόμενος, ἄρα δόκησις ἦν ἡ ἔνσαρκος Χριστοῦ παρουσία, καὶ εὐλόγως παρὰ Μανιχαίων καὶ Μαρκιωνιστῶν ἡ περὶ φαντασίας [δόκησις] τῆς ἐνσάρκου παρουσίας ὑπόθεσις ᾄδεται, <ὅτι δόκησις ἦν> καὶ οὐκ ἀληθεστάτη (“And they do not know that the human nature of Christ would have been an illusion if he did not have all these things, including ‘Not my will, but yours’; and if Christ had not been in agony and sweat had not poured from his body, there would be some sense to the theory of the unreality of the human nature that Manichaeans and Marcionites yap about, since he would be an apparition and not real at all.” Translation taken from F. Williams, The Panarion of Epiphanius of Salamis: Books II and III, (Sects 47–80, De Fide) (Leiden: E.J. Brill, 1994), 379–80. 35 Although, the evidence provided by him could perhaps lend some contextual weight to the pres- ent proposition that this passage was omitted from Luke. Marcion was not adding material to Luke but deleting material that did not conform to his theology. Therefore, the earliest evidence we have for the deliberate scriptural corruption of Luke is not in the form of interpolation but omission and deletion. An interesting question raised by Marcion’s edition of Luke is whether or not it could have affected non-Marcionite copies of Luke or that perhaps some of the early papyri we possess ema- nated from a Marcionite milieu. Though some have argued that Marcion’s recensions of the scrip- tures indeed affected later non-Marcionite texts and that some of these alterations can be detect- ed in certain papyri (Williams, Alterations to the Texts of the Synoptic Gospels and Acts, 10–18, has in mind 𝔓45; cf. Harnack, Marcion, 206), others have rejected the possibility outright (A.F.J. Klijn, “Matthew 11:25 / Luke 10:21,” in E.J. Epp and G.D. Fee [eds.], New Testament Textual Criticism: Its Significance for Exegesis [Oxford: Clarendon Press, 1981], 14, who notes that “Marcionite influ- ence on early Greek papyri seems impossible.”). Though such a possibility might seem unlikely, Clivaz has made a compelling, albeit circumstantial case, that 𝔓69 may well represent a Marcionite fragment of Luke (“The Angel and the Sweat,” 429–32). Additionally, if the words of the Church Fathers can be trusted when they say that Marcionism had spread everywhere, including Egypt, such a possibility cannot be ruled out (Justin, 1 Apol. 1.26, 58; Tertullian, Praesc. 30; Epiphanius, Pan. 42.1). Lastly, the presence of the “Marcionite Prologues” certainly suggests that Marcionite influence could be had upon later non-Marcionite texts (K.T. Schäfer, “Marius Victorinus und die marcionitischen Prologe zu den Paulusbriefen,” RBén 80 [1970]: 7–16). 10 Luke 22:43–44 a certain author does not cite Luke 22:43–44, it does not necessarily mean that they did not know the passage and should therefore be cited as a witness against it. It is important here not to confuse the absence of evidence with actual evidence and to realize that many authors never cited a number of verses they otherwise knew.36 Thus, very little weight should be put on “negative proof.”37 For example, while Tertullian has sometimes been invoked as a witness against the passage, because he never explicitly cites it, this is a tenuous argument at best.38 Based on the extant literary remains of Tertullian it is difficult to know with a high degree of certainty that he definitely did not know the passage as he never devotes considerable attention to the Gethsemane39 narrative in Luke.40 Turning to Clement of Alexandria, who is regularly cited as a witness against this passage,41 while we may perhaps be a little more certain that the copies of Luke that Clement used did not have this passage there is still a significant degree of uncertainty and it goes well beyond the bounds of the evidence to state that Clement is a witness against the verse just because he never explicitly cites it.42 Like Tertullian, in none of Clement’s extant writings does he ever quote extensively from Luke’s passion narrative so it is difficult to be certain that he did not know the passage; while he may be referenced as an early witness who does not mention the passage, it goes too far to state that he is a witness against it.43 Turning to Origen, who is also frequently cited as a witness against this passage, the same 36 Duplacy, “La préhistoire du texte en Luc 22:43–44,” 78. 37 On the problems of “negative proof,” generally speaking, see D.H. Fischer, Historians’ Fallacies: Toward a Logic of Historical Thought (New York: Harper & Row, 1970), 47–48. 38 Brown, The Death of the Messiah, 180, who cites Tertullian as one of the Church Fathers whose copy of Luke supposedly did not contain 22:43–44. 39 Though Luke never used the term Gethsemane, unlike Matthew and Mark (Matt 26:36, Mark 14:32), and instead speaks rather generically about the location and simply mentions “the place” (Luke 22:40; cf. TDNT 8.195–99) on the “Mount of Olives” (Luke 22:39), for convenience Geth- semane will be used inclusively in this paper to refer to Luke’s account. 40 BiPa 1.373 cites Tertullian, Prax. 27.11 as evidence for Luke 22:44. However, an examination of this reference in Tertullian (anxia usque ad mortem) suggests that Luke 22:44 is not the putative source; a more likely possibility is Matt 26:38 (tristis anima mea usque ad mortem). There is no good reason for necessarily equating Tertullian’s anxia with Luke’s agonia (v. 44). Therefore, this passage should not be marshaled as evidence that Tertullian knew of Luke 22:43–44. On the other hand, Clivaz, L’ange et la sueur de sang, 585 argues that this passage is “probablement allusion à Lc 22,44.” 41 Comfort, New Testament Text and Translation Commentary, 233; Ehrman, The Orthodox Corrup- tion of Scripture, 220; Metzger, A Textual Commentary on the Greek New Testament, 151. 42 Clement cites Luke 22:31 (Strom. 4.74.4). The next verse in Luke explicitly cited by Clement is 22:32 (Strom. 4.74.4). See C.P. Cosaert, The Text of the Gospels in Clement of Alexandria (NTGF 9; Atlanta: SBL, 2008), 180–81. A possible allusion to Luke 22:42 might be found in Paed. 1.6.46 where Clement talks about the “cup” (ποτήριον) but this could just as easily refer to Matt 26:39. In Paed. 2.8.62–63 when Clement talks about the betrayal of Jesus by Judas and how he betrayed him with a “kiss” (φιλήμα) it is likely that he had Luke 22:48 in mind, although given his general description of the betrayal he could have also had Matt 26:48–49 or Mark 14:44–45 in mind. 43 Here it needs to be remembered that we do not have all of Clement’s writings. If Eusebius is correct that Clement wrote commentaries (Hypotyposeis [Ὑποτυπώσεις]) on certain scriptural books of the OT and NT (Hist. eccl. 6.14.1), and if we had his commentary on Luke and he never mentioned the passage, we would certainly be more sure about whether or not Clement knew of the passage. Given the nature of Clement’s extant writings caution and prudence needs to be ex- ercised before automatically rushing to judgments about what he may, or may not, have known.