1 This point is highlighted at length in F. Watson, Gospel Writing: A Canonical Perspective (Grand Rapids: Eerdmans, 2013), chapters 3–7. See also Jens Schröter, From Jesus to the New Testament: Early Christian Theology and the Origin of the New Testament Canon (trans. Wayne Coppins; Waco, TX: Baylor University Press, 2013), chapter 12.

2 I have argued elsewhere that in fact Tatian never gave his composition the title ‘Diatessaron’, but instead simply called it the ‘Gospel’, in a manner akin to the literary product of Marcion. In this way he sought to erase the memory of his source texts and supplant them with his own new creation, thereby preserving the unity and anonymity of the Jesus tradition. Cf. Crawford, M. R., ‘Diatessaron, a Misnomer? The Evidence From Ephrem's Commentary’, Early Christianity 4 (2013) 362–85CrossRefGoogle Scholar.

3 So C. E. Hill, Who Chose the Gospels? Probing the Great Gospel Conspiracy (Oxford: Oxford University Press, 2010) 104–12. Hill proposes that the Dura fragment of Tatian's work may not have served as the congregation's ‘sacred text’ but instead as ‘a pastor's study tool’ (p. 111), though he himself acknowledges that in many later Syriac churches Tatian's work was indeed treated as sacred scripture. If so, then it seems likely that it was used in a similar way at Dura. Hill does correctly draw attention to the fact that the Dura fragment is written on a roll, whereas all known manuscripts of the fourfold gospel take the form of codices, a feature which he takes as indicative of its non-authoritative status. However, since so much of our early manuscript evidence comes from Egypt, we should not automatically assume that the codex format was universally preferred for liturgical texts in other locales as well. On the relevance of manuscripts for reconstructing the history of early Christiantiy, see especially L. W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006). Hill's interpretation of Tatian's work is largely a response to W. L. Petersen, ‘The Diatessaron and the Fourfold Gospel’, The Earliest Gospels: The Origins and Transmission of the Earliest Christian Gospels – the Contribution of the Chester Beatty Gospel Codex P 45 (ed. C. Horton; Journal for the Study of the New Testament Supplement Series 258; T&T Clark: London, 2004).

4 Though, as is well known, the witnesses for Tatian's work reveal lingering traces of what appears to be the influence of extra-canonical gospels. Nevertheless, these elements are slight in comparison with the prominent position he granted to Matthew, Mark, Luke and John.

5 For an exploration of one such example, see Crawford, M. R., ‘Reading the Diatessaron with Ephrem: The Word and the Light, the Voice and the Star’, Vigiliae Christianae 69 (2015) 70–95 CrossRefGoogle Scholar.

6 See C. McCarthy, Saint Ephrem's Commentary on Tatian's Diatessaron: An English Translation of Chester Beatty Syriac MS 709 (Journal of Semitic Studies Supplement 2; Oxford: Oxford University Press on behalf of the University of Manchester, 1993). On Aphrahat, see T. Baarda, The Gospel Quotations of Aphrahat, the Persian Sage: Aphrahat's Text of the Fourth Gospel (Amsterdam: Vrije Universiteit Amsterdam, 1975).

7 On the Latin tradition, see especially U. Schmid, Unum ex quattuor: Eine Geschichte der lateinischen Tatianüberlieferung (Aus der Geschichte der lateinischen Bibel 37; Freiburg im Breisgau: Herder, 2005).

8 D. C. Parker, D. G. K Taylor, M. S. Goodacre, ‘The Dura-Europos Gospel Harmony’, in Studies in the Early Text of the Gospels and Acts (ed. D. G. K. Taylor; Text-Critical Studies 1; Atlanta: Society of Biblical Literature, 1999); Joosten, J., ‘The Dura Parchment and the Diatessaron’, Vigiliae Christianae 57 (2003) 159–75CrossRefGoogle Scholar.

9 See a survey of this debate in W. L. Petersen, Tatian's Diatessaron: Its Creation, Dissemination, Significance, and History in Scholarship (Supplements to Vigiliae Christianae 25; Leiden: Brill, 1994) 196–203, who summarises the earlier contributions of Baumstark, A., ‘Das Griechische “Diatessaron”-Fragment von Dura-Europos’, Oriens christianus 32 (1935) 244–52Google Scholar; Burkitt, F. C., ‘The Dura Fragment of Tatian’, JTS 36 (1935) 255–9CrossRefGoogle Scholar; C. H. Kraeling, A Greek Fragment of Tatian's Diatessaron from Dura (Studies and Documents 3; London: Christophers, 1935) 15–18; Lagrange, M.-J., ‘Deux nouveaux textes relatifs à l’Évangile’, Revue biblique 46 (1935) 321–7Google Scholar; Plooij, D., ‘A Fragment of Tatian's Diatessaron in Greek’, The Expository Times 46 (1935) 471–6CrossRefGoogle Scholar; Merk, A., ‘Ein griechisches Bruchstück des Diatessaron Tatians’, Biblica 17 (1936) 234–41Google Scholar. Petersen follows Plooij in seeing the Dura fragment as ‘a very early Greek translation of a Syriac Vorlage’. See, however, the recent reassessment in Parker, Taylor and Goodacre, ‘The Dura-Europos Gospel Harmony’, 209–16, who conclude that there is no evidence that the text was translated from Syriac, and suggest instead that it was composed in Greek. Debate over the original language of the Diatessaron has been going on since the contributions of Adolf von Harnack and Theodor Zahn in the late nineteenth century. For recent discussions of this issue, see W. L. Petersen, ‘New Evidence for the Question of the Original Language of the Diatessaron’, Studien zum Text und zur Ethik des Neuen Testaments: Festschrift Heinrich Greeven (ed. H. Greeven and W. Schrage; Beihefte zur Zeitschrift für die neutestamentliche Wissenschaft und die Kunde der älteren Kirche 47; Berlin: de Gruyter, 1986); id., Tatian's Diatessaron, 384–97; T. Baarda, ‘Tatian's Diatessaron and the Greek Text of the Gospels’, The Early Text of the New Testament (ed. Charles E. Hill and M. J. Kruger; Oxford: Oxford University Press, 2012) 337–8; U. Schmid, ‘The Diatessaron of Tatian’, The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis (ed. B. D. Ehrman and M. W. Holmes; New Testament Tools, Studies and Documents 42; Leiden: Brill, 2013) 115 n. 5. Schmid astutely remarks that ‘the bare mechanics of composing a gospel harmony appear to require sources and end product to be in one and the same language. It seems hardly conceivable to perform a close word-by-word harmonization from Greek gospel texts and a Syriac translation simultaneously, without at least one intermediate Greek harmony stage during the compositional process.’ See also F. Millar, Religion, Language and Community in the Roman Near East: Constantine to Muhammad (The Schweich Lectures of the British Academy; Oxford: Oxford University Press, 2013) 114: ‘Tatian will also have been a subject of the Roman Empire, and we have no reason to think that, in the second century, there was anywhere within that Empire where literary composition in Syriac took place. What Ephrem read will have been a Syriac translation of the original Greek.’

10 An image of the fragment may be accessed at http://brbl-legacy.library.yale.edu/papyrus/oneSET.asp?pid=DPg%2024 (accessed on 13 June 2014). Higher resolution images of the recto and verso may be downloaded from my personal page at www.academia.edu, reproduced with the kind permission of the Beinecke Rare Book and Manuscript Library of Yale University.

11 On the Christian house-church and its artwork, see Serra, D. E., ‘The Baptistery At Dura-Europos: The Wall Paintings in the Context of Syrian Baptismal Theology’, Ephemerides Liturgicae 120 (2006) 67–78 Google Scholar; Dirven, L., ‘Paradise Lost, Paradise Regained: The Meaning of Adam and Eve in the Baptistery of Dura Europos’, Eastern Christian Art 5 (2008) 43–57 CrossRefGoogle Scholar; D. Korol and M. Stanke, ‘Gehen die David- und Goliathdarstellungen im “Baptisterium” von Dura-Europos sowie im Vatopedi Psalter “auf den gleichen Archetyp” zurück? Neues zum ursprünglichen Aussehen und zur Deutung der Darstellung im “Baptisterium”’, Syrien und seine Nachbarn von der Spätantike bis in die islamische Zeit (ed. I. Eichner and V. Tsamakda; Spätantike, Frühes Christentum, Byzanz, Reihe B: Studien und Perspektiven 25; Wiesbaden: Reichert, 2009); U. Mell, Christliche Hauskirche und Neues Testament: Die Ikonologie des Baptisteriums von Dura Europos und das Diatessaron Tatians (Göttingen: Vandenhoeck & Ruprecht, 2010); D. Korol and J. Rieckesmann, ‘Neues zu den alt- und neutestamentlichen Darstellungen im Baptisterium von Dura-Europos’, Ablution, Initiation, and Baptism: Late Antiquity, Early Judaism, and Early Christianity (ed. David Hellholm et al.; Beihefte zur Zeitschrift für die neutestamentliche Wissenschaft und die Kunde der älteren Kirche 176; Berlin: de Gruyter, 2011); Peppard, M., ‘Illuminating the Dura-Europos Baptistery: Comparanda for the Female Figures’, Journal of Early Christian Studies 20 (2012) 543–74CrossRefGoogle Scholar. On Dura-Europos more generally, see especially F. Millar, The Roman Near East, 31 bc–ad 337 (Cambridge, MA: Harvard University Press, 1993) 131–3, 445–52, 467–71; T. Kaizer, ‘Religion and Language in Dura-Europos’, From Hellenism to Islam: Cultural and Linguistic Change in the Roman Near East (ed. H. M. Cotton, R. G. Hoyland, J. J. Price, D. J. Wasserstein; Cambridge: Cambridge University Press, 2009) 235–53.

12 See Kraeling, A Greek Fragment of Tatian's Diatessaron From Dura, 5–7. Cf. C. B. Welles, R. O. Fink, J. F. Gilliam, The Excavations at Dura-Europos Conducted by Yale University and the French Academy of Inscriptions and Letters: Final Report 5, Pt. 1: The Parchments and Papyri (New Haven: Yale University Press, 1959) 73–4.

13 Parker et al., ‘The Dura-Europos Gospel Harmony’, 194–9.

14 Parker et al. question the connection with the house-church in light of the cutting apparent on the fragment, which suggests that the original scroll was not simply haphazardly destroyed along with the house-church as a part of the Roman defensive preparations. Moreover, they argue that if the scroll had been cut into scraps for reuse because it had simply ended its useful life, the period between the founding of the church and its destruction would have been an insufficient amount of time for a scroll to wear out (‘The Dura-Europos Gospel Harmony’, 197–8). However, there are other plausible explanations that would not militate against the connection of the parchment with the house-church. For example, the scroll might have been used in some other location before coming to Dura, in which case it might well have ended its useful life as a result of usage at the house-church. It seems to me unduly sceptical not to assume some connection between the only two surviving pieces of material evidence for Christianity in Dura-Europos, especially when they were discovered only two blocks apart.

15 Parker et al., ‘The Dura-Europos Gospel Harmony’, 200–1. I have slightly altered their translation in a few places.

16 Millar, The Roman Near East, 460.

17 Cf. Petersen, Tatian's Diatessaron, 71. So the date is given in Eusebius’ Chronicle. To obtain the same date from Epiphanius one has to substitute the name of one emperor for another.

18 See e.g. Petersen, Tatian's Diatessaron, 127–9; Joosten, ‘The Dura Parchment and the Diatessaron’, 163–5. The term ‘Vulgatisation’ originally was used to describe Codex Fuldensis, whose text largely conforms to the Latin Vulgate, while retaining a harmonised format. However, the term is now widely used to refer to this same tendency as it occurred in the entire history of the reception of Tatian's text, in both the Western and Eastern traditions.

19 For this reason, the study of Tatian's work is encumbered with unusually complex debates over methodology. For one influential proposal, see Petersen, Tatian's Diatessaron, 357–77. However, since Petersen's writing, a methodological shift has occurred, represented especially in the following studies: Schmid, U. B., ‘In Search of Tatian's Diatessaron in the West’, Vigiliae Christianae 57 (2003) 176–99CrossRefGoogle Scholar; id., ‘Genealogy by Chance! On the Significance of Accidental Variation (Parallelism)’, Studies in Stemmatology ii (ed. P. van Reenen, A.A. den Hollander, M. van Mulken; Amsterdam: John Benjamins, 2004); id., ‘The Diatessaron of Tatian’; den Hollander, A. and Schmid, U., ‘The Gospel of Barnabas, the Diatessaron, and Method’, Vigiliae Christianae 61 (2007) 1–20 CrossRefGoogle Scholar.

20 Cf. Crawford, M. R., ‘The Fourfold Gospel in the Writings of Ephrem the Syrian’, Hugoye 18 (2015) 3–46 Google Scholar.

21 As also recognised by Lagrange, ‘Deux nouveaux textes relatifs à l’Évangile’, 325; Merk, ‘Ein griechisches Bruchstück’, 239–40.

22 Crawford, M. R., ‘“Reordering the Confusion”: Tatian, the Second Sophistic, and the so-called Diatessaron’, Zeitschrift für antikes Christentum 19 (2015) 209–36Google Scholar.

23 T. Baarda, ‘Διαφονία-Συμφωνία: Factors in the Harmonization of the Gospels, especially in the Diatessaron of Tatian’, Essays on the Diatessaron (ed. T. Baarda; Contributions to Biblical Exegesis and Theology 11; Kampen: Kok Pharos, 1994).

24 On Lucian's treatise, see G. Avenarius, Lukians Schrift zur Geschichtsschreibung (Meisenheim am Glan: Hain, 1956); H. Homeyer, Lukian: Wie man Geschichte schreiben soll (München: W. Fink, 1965); B. Baldwin, Studies in Lucian (Toronto: Hakkert, 1973) 75–95; Fox, M., ‘Dionysius, Lucian, and the Prejudice against Rhetoric in History’, JRS 91 (2001) 76–93 Google Scholar.

25 Ulrich Mell has argued that Tatian's version must have included as witnesses to the resurrection four named women, plus a larger anonymous crowd of women. Mell then connects this to the baptistery artwork in which one scene is understood to be a depiction of the women at the tomb, showing five female figures (the four named women plus an additional figure representing the anonymous crowd). He therefore thinks that the baptistery preserves images of scenes drawn directly from Tatian's work (Christliche Hauskirche, 253–9). This strikes me as too speculative an enterprise, because it requires one figure to represent a group of persons, and especially because alternate interpretations of the scene are possible, which do not view it as a depiction of Easter morning. Cf. Peppard, ‘Illuminating the Dura-Europos Baptistery’, 556–71.

26 Kraeling, A Greek Fragment of Tatian's Diatessaron, 28–30; Parker et al., ‘The Dura-Europos Gospel Harmony’, 201–5. Cf. the similar use of ἐκ in Luke 24.13: δύο ἐξ αὐτῶν.

27 The participial form is presented as a nomen sacrum (στα), on which see Parker et al., ‘The Dura-Europos Gospel Harmony’, 206–8. Parker et al. point out that the same type of contraction is found twice each in Codex Bezae and P46, and they conclude that the contraction with only στ- is a primitive version that was abandoned in favour of στρ-. Similarly, Merk, ‘Ein griechisches Bruchstück’, 237, called this ‘die einzige wichtigere Abweichung vom herkömmlichen Evangelientext, die im ganzen Fragment zu beobachten ist’.

28 See similar constructions in Paul at 1 Cor 1.23; 2.2; Gal 3.1.

29 Dial. 106.1 (P. Bobichon, Justin Martyr, Dialogue avec le Tryphon: Edition critique (Fribourg: Academic Press Fribourg, 2003) 470).

30 Dial. 110.3 (Bobichon, 478).

31 Dial. 101.3 (Bobichon, 456).

32 1 apol. 32.4 (D. Minns and P. Parvis, Justin, Philosopher and Martyr: Apologies (Oxford Early Christian Texts; Oxford: Oxford University Press, 2009) 168).

33 See 1 apol. 13.3, 14; 21.1; 35.6; 53.2; 61.13; 2 apol. 6.6; Dial. 10.3; 35.2; 11.4, 5; 30.3; 34.8; 38.1; 46.1; 49.8; 67.6; 71.2; 73.2; 76.6; 85.2; 89.3; 91.4; 93.4; 95.2; 96.1; 111.2; 112.2; 116.1; 117.5; 131.5; 132.1; 137.1.

34 Eusebius, HE 5.13.5 (G. Bardy, Eusèbe de Césarée: Histoire Ecclésiastique, Livres v–vii (Source Chrétiennes 41; Paris: Les Éditions du Cerf, 1955) 43). Rhodon is here actually summarising the teaching of Apelles, Marcion's student, but his expression in this passage may reflect his own speech pattern, rather than that of his opponent.

35 In Irenaeus, see Haer. 2.32.4 (= fr. gr. 9) (Ἰησοῦ Χριστοῦ τοῦ σταυρωθέντος ἐπὶ Ποντίου Πιλάτου = in nomine Christi Iesu crucifixi) (A. Rousseau and L. Doutreleau, eds., Irénée de Lyon: Contre les Hérésies, Livre ii (Sources Chrétiennes 294; Paris: Les Éditions du Cerf, 1982) 342–3); 3.12.4 (= fr. gr. 16) (τὸν Ἰησοῦν τὸν ἐσταυρωμένον = Iesum qui crucifixus est) (A. Rousseau and L. Doutreleau, eds., Irénée de Lyon: Contre les Hérésies, Livre iii (Sources Chrétiennes 211; Paris: Les Éditions du Cerf, 1974) 192); 4.23.2 (Jesum crucifixum hunc esse Christum Filium Dei) (A. Rousseau et al., eds., Irénée de Lyon: Contre les Hérésies, Livre iv (Sources Chrétiennes 100ii; Paris: Les Éditions du Cerf, 1965) 698). The only other second-century occurrences I have found are: Gospel of Peter 13.56; letter from the churches of Gaul, apud Eusebius, HE 5.1.41; Hegesippus, apud Eusebius, HE 2.23.12. Mark Goodacre has suggested that the famous walking and talking cross of the Gospel of Peter may have arisen in the text through a scribal misunderstanding of an original abbreviation στα (http://ntweblog.blogspot.co.uk/2010/10/walking-talking-cross-or-walking.html; accessed on 11 November 2014). On this idea, which in one form goes back to Adolf von Harnack, see most recently Foster, P., ‘Do Crosses Walk and Talk? A Reconsideration of Gospel of Peter 10.39–42’, JTS 64 (2013) 89–104 CrossRefGoogle Scholar.

36 The strongest argument against this interpretation of the evidence is that the designation does not occur in Tatian's only other surviving work, the Oratio. This, however, is not really a problem, since the Oratio is mostly an attack on Greek παιδεία and makes only one passing allusion to the crucifixion, calling Jesus τὸν διάκονον τοῦ πεπονθότος θεοῦ (Or. 13.3, M. Whittaker, Tatian: Oratio ad Graecos and Fragments (Oxford Early Christian Texts; Oxford: Oxford University Press, 1982) 28). On this unusual phrase, which differs from Justin's linguistic usage, see Hanig, R., ‘Tatian und Justin: Ein Vergleich’, Vigiliae Christianae 53 (1999) 61–3CrossRefGoogle Scholar.

37 Lucian, How to Write History 51.

38 Lagrange, ‘Deux nouveaux textes relatifs à l’Évangile’, 325 came to a similar conclusion with respect to the στα: ‘il est plus sûr de reconnaitre dans Tatien un trait de génie que de fidélité, puisque nous venons de voir qu'il en prenait à son aise’.

39 On this point Lagrange, ‘Deux nouveaux textes relatifs à l’Évangile’, 324 remarks, ‘Il faut une fois de plus rendre hommage à l'extrême habileté de l'harmoniste.’

40 According to Parker et al., ‘The Dura-Europos Gospel Harmony’, 226, Codex Bezae also has τὸ ὄνομα, against the standard Matthean τοὔνομα.

41 As noted by Parker et al., ‘The Dura-Europos Gospel Harmony’, 227, the καί is also omitted by Codex Vaticanus.

42 A possibility raised by Lagrange, ‘Deux nouveaux textes relatifs à l’Évangile’, 326. On Tatian's encratism, see most recently Koltun-Fromm, N., ‘Re-Imagining Tatian: The Damaging Effects of Polemical Rhetoric’, Journal of Early Christian Studies 16 (2008) 1–30 CrossRefGoogle Scholar and M. R. Crawford, ‘The Problemata of Tatian: Recovering the Fragments of a Second-Century Christian Intellectual’, JTS (forthcoming).

43 Jerome, Comm. Mt. 27.58. In the same paragraph Jerome also notes that another evangelist calls Joseph a βουλευτής, which he equates with a consiliarius. Augustine's gospel text translated βουλευτής with decurio (De consensu evangelistarum 3.22), and he similarly points out that it was Joseph's fiducia dignitatis which granted him access to Pilate. On the reception history of Joseph, see W. J. Lyons, Joseph of Arimathea: A Study in Reception History (Biblical Refigurations; Oxford: Oxford University Press, 2014), who, in keeping with Jerome and Augustine, remarks, ‘Mark's description of Joseph as an esteemed member of the Sanhedrin responsible for Jesus’ death was more probably intended to help illuminate Pilate's decision to grant his request for the body’ (p. 11).

44 Jerome, Comm. Mt. 27.59.

45 On this point, see R. Duncan-Jones, The Economy of the Roman Empire: Quantitative Studies (Cambridge: Cambridge University Press, 1982²) 82–8 (on North Africa), 147–55 (on Italy). As noted by J. E. Lendon, Empire of Honour: The Art of Government in the Roman World (Oxford: Clarendon Press, 1997) 85 n. 269, ‘by the 2nd cent., perhaps the largest source of a city's “public” money was summae honorariae, the required payments made to the city upon election to the city council, a magistracy, or a priesthood’.

46 Cf. Joosten, ‘The Dura Parchment and the Diatessaron’, 172: ‘The harmonist may have judged that the notion of wealth was sufficiently expressed in the description of Joseph as a member of the council (βουλευτής). More importantly, Tatian was an advocate of evangelical poverty who despised riches. He may well have been tempted to underplay this characteristic in his description of the man who gave Jesus a dignified burial-place.’

47 Ibid., 167.

48 So also Lagrange, ‘Deux nouveaux textes relatifs à l’Évangile’, 324.

49 According to Cook, J. G., ‘A Note on Tatian's Diatessaron, Luke, and the Arabic Harmony’, Zeitschrift für antikes Christentum 10 (2007) 471Google Scholar, ‘Tatian's harmony … give[s] priority to Luke and John in the beginning and ending sections – although Matthew dominates most of the rest’.

50 So Lyons, Joseph of Arimathea, 10–11. Cf. Tcherikover, V. A., ‘Was Jerusalem a ‘Polis’?’, Israel Exploration Journal 14 (1964) 61–78, at 72Google Scholar: ‘We thus arrive at the conclusion that, although Josephus uses the Greek noun βουλή to designate the supreme institution of the Jewish people under Roman rule, and its members are referred to as bouleutai not only in Josephus but also in other sources, actually this was not a new municipal institution on a Hellenistic pattern, but the traditional Jewish body that had existed under different names throughout the period of the Second Temple’.

51 The Dura text actually reads Ἐρινμαθαία for Ἁριμαθαία, with the shift in the inital vowel and the addition of an internal nu, a feature that was taken as evidence for a Syriac original by Baumstark, ‘Das griechische “Diatessaron”-Fragment von Dura-Europos’, 249–50. However, Parker et al., ‘The Dura-Europos Gospel Harmony’, 211–13 point out that these changes may be plausibly explained by similar shifts that occur in other Greek manuscripts though these particular changes for this specific word do not appear in the rest of the textual tradition of the New Testament.

52 So Lyons, Joseph of Arimathea, 17.

53 Joosten, ‘The Dura Parchment and the Diatessaron’, 171 points out that this variant is not found in the entire Greek tradition, though it does occur in the Arabic harmony, Codex Fuldensis, the Old Syriac, the Peshitta, the Old Latin and the Vulgate.

54 Though by this point the province of Judea had been renamed Syria Palaestina, following the Jewish revolt of the 130s. Cf. Millar, The Roman Near East, 107–8, 374ff. So also Parker et al., ‘The Dura-Europos Gospel Harmony’, 222: ‘The change of Luke's text to “of Judaea” may be the consequence of the political situation in the time of Tatian. With the dispersal of the Jews, the town is simply in a Roman province called Judaea’.

55 Joosten, ‘The Dura Parchment and the Diatessaron’, 171 n. 33 notes that the phrase ‘the Jews’ is also missing or transformed in the Old Syriac and Old Latin at John 4.22 and 9.22b; and at John 7.13 of the Old Syriac, perhaps under the influence of Tatian's version.

56 LSJ s.v. ἐπί B.ii.1. So also Joosten, ‘The Dura Parchment and the Diatessaron’, 173, who sees this construction as a ‘probable sign of ineptness’ since it does not occur in NT Greek with this sense. He suggests it might therefore be ‘an erroneous transformation of the Markan phrase’.

57 Justin, 1 apol. 13.3 (Ἰησοῦν Χριστόν, τὸν σταυρωθέντα ἐπὶ Ποντίου Πιλάτου, τοῦ γενομένου ἐν Ἰουδαίᾳ ἐπὶ χρόνοις Τιβερίου Καίσαρος ἐπιτρόπου) (Minns and Parvis, 110). See the other examples of ἐπί plus the dative for time in G. W. H. Lampe, A Patristic Greek Lexicon, s.v. ἐπί ii.E.

58 LSJ s.v. ἐπί B.ii.2.

59 Parker et al., ‘The Dura-Europos Gospel Harmony’, 205–6.

60 Joosten, ‘The Dura Parchment and the Diatessaron’, 172–3. He is here following the prior analysis of Parker et al., ‘The Dura-Europos Gospel Harmony’, 220–1, who compared the Dura text with the following supposed witnesses for the Diatessaron: Fuldensis, Tuscan, Venetian, Zacharius, Pepysian, Heiland, Arabic and Persian.

61 See especially the recent summary in Schmid, ‘The Diatessaron of Tatian’.

62 The difficulty of the text was also recognised by Lagrange, ‘Deux nouveaux textes relatifs à l’Évangile’, 326, who nevertheless regarded it as authentically Tatianic. He remarked, ‘Je croirais plutôt qu'il n'a su que faire de cet ἐπιφώσκω qui embarrasse encore les modernes.’ Michael Goulder argued that the phrase σάββατον ἐπέφωσκεν in Luke 23.54 arose from Luke's misreading of the unusual phrase ὀψὲ δὲ σαββάτων τῇ ἐπιφωσκούσῃ εἰς μίαν σαββάτων in Matthew 28.1, taking the participle in the sense of ‘drawing on’ rather than ‘dawning’ (Luke: A New Paradigm (JSNT Supplement Series 20; Sheffield: Sheffield Academic Press, 1989) ii.772–3). I am grateful to Mark Goodacre for pointing me to this reference. Similar phrases occur in the Gospel of Peter: σάββατον ἐπιφώσκει (2.5); Πρωΐας δὲ ἐπιφώσκοντος τοῦ σαββάτου (9.34); Τῇ δὲ νυκτὶ ᾗ ἐπέφωσκεν ἡ κυριακή (9.35). The usage at 2.5 is a statement of Herod made before the crucifixion has even begun, so the author of the text clearly intended it in the sense that the Sabbath was impending, rather than having already begun.

63 E. Ranke, Codex Fuldensis (Marburg and Leipzig: Elwert, 1868) 156–7. A new edition of Codex Fuldensis is in preparation by Nicholas Zola. None has been published since Ranke's version.

64 Interestingly, the sequence of Ephrem's gospel text seems to agree with that of Fuldensis in inserting the Johannine verses between the two pericopae of the Dura fragment. See Ephrem, Commentary on the Diatessaron xxi.8–21. The onlookers at the cross are mentioned at xxi.8, the piercing with the lance is interpreted in xxi.10–13, and the burial is discussed at xxi.20–1. The Arabic agrees with the Dura fragment, against Ephrem and Fuldensis, in that it includes the piercing with the lance at Tat^AR 52.14–20, just before the report of the women watching the crucifixion, which is followed directly by the burial account. This division in the textual tradition is curious and implies that further rewriting must have already occurred between the date of the Dura fragment and the date of Ephrem's commentary, perhaps in conjunction with the translation of the text from Greek into Syriac. If so, the Arabic might represent further editing involving the relocation of the Johannine scene to earlier in the text.

65 Luke 23.54 does not appear at all in the version of these episodes in Codex Fuldensis (cf. Ranke, pp.156–7).

66 Tat^AR 52.21–7 (A.-S. Marmardji, Diatessaron de Tatien (Beyrouth, 1935) 501). On the Arabic version, see further Joosse, P., ‘An Introduction to the Arabic Diatessaron’, Oriens Christianus 83 (1999) 72–129 Google Scholar.

67 So also Joosten, ‘The Dura Parchment and the Diatessaron’, 167–70, who points out that in the Old Syriac the three elements in Luke 23.51 are inverted in precisely the same manner as in the Dura fragment, while the Peshitta partially corrects this inversion, presenting the same order as is found in the Arabic harmony.

68 Luke 23.54 does not show up at all in the Arabic harmony. See the table at Marmardji, Diatessaron de Tatien, cxxxv.

69 Pace Nicholas Perrin, who argues that Tatian's rewriting of the canonical gospels was merely a sign of his esteem for them as authoritative texts (‘Hermeneutical Factors in the Harmonization of the Gospels and the Question of Textual Authority’, The Biblical Canons (ed. J.-M. Auwers and H. J. de Jonge; Leuven: Leuven University Press, 2003) 599–605). Though his choice to use these texts suggests that he regarded them as suitable sources, he also thought himself capable of improving upon and surpassing them! Perrin presents this trend to rewrite previous texts as a Jewish phenomenon. That it surely was, but not exclusively so, and Tatian would have been operating more from a Hellenistic rhetorical background than a specifically Second Temple Jewish one. For further discussion of the process of μετάφρασις (‘paraphrase’), see S. F. Johnson, The Life and Miracles of Thekla: A Literary Study (Hellenic Studies 13; Washington, DC: Center for Hellenic Studies, 2006) 67–112. Johnson cites Tatian's work as a ‘fascinating, if elusive’ indication that early Christian authors were already aware of the ‘metaphrastic relationships among the synoptic gospels’ (p. 110). Further evidence to that point may be found in the Diatessaron-Gospel of Ammonius of Alexandria, on which see Crawford, M. R., ‘Ammonius of Alexandria, Eusebius of Caesarea, and the Origins of Gospels Scholarship’, NTS 61.1 (2015) 1–29 CrossRefGoogle Scholar.

70 Schröter, From Jesus to the New Testament, 270, who follows D. C. Parker, The Living Text of the Gospels (Cambridge: Cambridge University Press, 1997).

71 Bovon, F., ‘Beyond the Canonical and the Apocryphal Books, the Presence of a Third Category: The Books Useful for the Soul’, HTR 105 (2012) 125–37CrossRefGoogle Scholar. Bovon's description of texts in this category is particularly apt to the Diatessaronic tradition: ‘The destiny of books in the third category, by way of contrast, was very different. Their text is so flexible that it is often impossible to publish a single critical edition. Multiple forms of the text – each having different titles and being recorded in a wide variety of manuscripts – orient one's attention to evidence of a situation where each scribe achieved an individual performance’ (p. 134). See also pp. 129–30, where Bovon describes Victor of Capua's attempt to rescue Tatian's work from shipwreck. Cf. what Johnson says about saints’ lives: ‘authors often sought, or felt compelled, to reclassify, reorient, and purify the textual past for the sake of their audiences and readers-to-come’ (The Life and Miracles of Thekla, 106).