The last Syriac version of the NT, prepared by Tumo of Ḥarqel (Syr. Orth.) 615/16 in the Enaton near Alexandria (Egypt). This version is a recension (turroṣo) of the NT, with numerous Greek variant readings (translated into Syriac) quoted in the margins or with asterisks in the main text. The recensional method of the Ḥarqlean Version was inherited from Origen’s Hexaplaric Septuagint recension (the Syriac translation of which is the Syro-Hexapla). The Syro-Hexapla therefore can contribute to the proper understanding of the Ḥarqlean version’s features and intention. Historically, both versions are parts of the same translation project the Syr. Miaphysites initiated in the second decade of the 7th cent. According to the subscriptions of the Ḥarqlean Gospels (Kiraz, CESG, vol. 4, 369), Acts, and the Pauline epistles, the translation was executed ‘at the Enaton of (i.e., near) Alexandria, the great city, in the holy Convent of the Antonines … in the year 927 of Alexander, in the fourth indiction (i. e., 615/16)’. The Ḥarqlean Version is mentioned by Bar ʿEbroyo in his ‘Storehouse of Secrets’ (Awṣar roze), proem, in his Ecclesiastical History (I,50 = vol. 1, col. 267), and by Michael Rabo, Chronicle, X,25 = vol. IV 391.

The characteristic features of the Ḥarqlean version are the decisive turn towards the Greek (‘mirror translation’) and the introduction of the philological and thus ‘critical’ dimension into the history of the Syriac NT. The method of recension sketched in the subscriptions remains ambiguous (esp. the role of the Philoxenian); the result, however, is clear: Tumo prepared a NT text that did not exist before; i.e., he produced a recension of the NT. This recension was based on a translation of an existing Greek model and supplemented by textual material drawn from additional Greek mss. In the same way, the Syro-Hexapla is based on an existing Greek model (the Hexaplaric Septuagint recension) and supplemented by textual material drawn from Aquila, Symmachus, and Theodotion (and others). The new philological and ‘critical’ dimension is provided by the distinctive presentation of the additional material. It is located in the margins or in the main text (marked with asterisks) in order not to mix the textual traditions put together in this recension. This distinctive presentation allows the additional material to be read in context as variants of the Graeca veritas. Additions to and omissions from Tumo’s Greek model could be presented in the main text; exchanges of words or passages had to be presented in the margin and connected to the corresponding word in the text by a graphic sign. The obelos is used for the sake of translation technique to mark Syriac words which do not match the Greek Vorlage but are necessary for an intelligible rendering of the Greek. By this way of quoting the additional material and using ‘critical’ signs, Tumo is a pupil of Origen; from a modern point of view, he attached a ‘critical’ apparatus to his version. Of course Tumo was not a modern critic, but he was aware of the non-uniformity of the Greek text. He knew about different textual traditions and that translating a single Greek ms. (like Polykarpos, who prepared the Philoxenian) would provide only an imperfect knowledge of the Greek NT text.

Besides the recensional features, a revisional relation between the Ḥarqlean Version and the Philoxenian is reported in the subscriptions of the Ḥarqlean Version. Although this report is not detailed enough, Tumo could have used the earlier version as the starting point of his own. An attractive feature of the earlier version was its relation to Caesarea in Palestine and to the famous library of Pamphilus, which is reported in the Ḥarqlean subscription to the Pauline letters. A revisional relation to the Philoxenian would have brought Tumo’s own version in line with the Caesarean origin of the Syro-Hexapla and would have supplied the whole Miaphysite translation project with one uniform authority.

As J. Gwynn suggested (‘Paulus [48] Tellensis’, DCB 4 [1887], 266–71, at 267), the instigative historical setting of the laborious translation project could have been the reunion of the Syrian and Coptic Miaphysites after decades of schisma dating from the time of the Syrian Patr. Peter of Kallinikos (581–91) and the Coptic Patriarch Damian (578–606). Although no proof can be given, Gwynn’s suggestion is supported by the common date 616 for the reunion and the completion of the Ḥarqlean Version. This suggestion, however, puts a special linguistic perspective on the Ḥarqlean Version. The reunion had to be performed completely in Greek; and both parties had to determine the authoritative Greek text of the Scriptures for reunion. This text had to be of ‘ecumenical’, not of local currency. For the OT, Origen’s Hexaplaric Septuagint recension (handed down by Pamphilus and Eusebius) could easily be accepted by both parties; for the NT, no such work of Origen existed, and the text had to be determined by the Miaphysites themselves. This text became the Greek model of the Ḥarqlean Version. According to this perspective, the Ḥarqlean Version came into being after the determination of the Greek model as the translation and inner-Syriac promulgation of the approved Greek NT text. The recensional as well as the revisional imprint on the Ḥarqlean version were added during or after translation, when comparison with additional Greek mss. (and with the Philoxenian) was made. Another possibility is that the Ḥarqlean margin was already an integral part of the Greek text established by the Syrian and Coptic Miaphysites in order to quote (i.e., to include into the reunion) different textual traditions of both parties.

After the rise of Islam, the version’s ‘ecumenical’ frame of reference was lost and the version itself was considered as the ‘Greek’ (Bar ʿEbroyo in his Scholia to the NT). The ‘ecumenical’ perspective of the Ḥarqlean Version was replaced by a scientific and philological attitude. Due to the complicated lay-out of the version, not all scribes were capable of transmitting the text properly; some of them omitted the marginal quotations and the asterisks/obeloi completely, thus mixing the readings Tumo intended to keep separate. The philologists were not interested in preserving the original text but in updating it according to the current Greek text of the New Testament, which was the increasingly dominant Byzantine text. This development created ‘corrected’ models (hyparchetypes) and faded out the original non- and pre- Byzantine part of the version.

Due to the large number of mss., the revisional development can be traced best in the Gospel mss. The total is ca. 100 from the 8th to the 19th cent., but only ca. 20 mss. are furnished with the marginal and asterisked additions. Ms. Vat. Syr. 268 (8th/9th cent.) is singled out by an undeveloped stage of text, which preserved a large number of non-Byzantine readings. The majority of the witnesses reflect the version’s development towards the Greek Byzantine text; unfortunately, the oldest dated ms. of the Gospels (ms. Florence, Bibl. Laurenziana Plut. I.40; 757) is of this type. This first revisional stage of the version is well represented in further mss. of the first millennium, e.g., by Vat. Syr. 267 (8th cent.), Harvard Syr.  16 (8th/9th cent.), Brit. Libr. Add. 7163 (9th/10th  cent.), Mingana Syr. 124 (9th/10th cent.), and Chester Beatty Syr. 3 (1177, copied from a model of 841).

In the 12th cent. the revisional development entered a new stage with the activity of Dionysios bar Ṣalibi (d. 1171). Again, Gospel mss. provide the information. The subscriptions, which report his revision (ms. Oxford, New Coll. 334; 12th/13th cent. and London, Brit. Libr. Add. 17,124; 1233/34; see Wright, Catalogue … British Museum, vol. 1, 42) do not mention Greek mss.; it probably relies on several ‘corrected’ models of the version itself. Essential features are the further reduction of the marginal readings and critical signs (though some completely new are added), the inversion of many text/margin readings, and the introduction of the Pericope on the Adulteress (Jn 7:53–8:11).

The extreme rarity of mss. for Acts and the Epistles prevents us from tracing a revisional development in these parts of the NT. There are only two mss. from the first millennium, ms. Jerusalem, St. Mark 37 (Pauline epistles) and ms. London, Brit. Libr. Add. 14,474 (Catholic Epistles), which are of the 9th cent. Two later mss. cover the entire NT with the exception of Revelation: ms. Cambridge, Add. 1700 (1069/70) and ms. Oxford, New Coll. 333 (12th/13th cent.). Fortunately, all except the Cambridge ms. are furnished with marginal and asterisked variants. Revelation is represented by ca. ten late mss., among which Mardin Orth. 36/2 (13th cent., ed. by A. Vööbus as no. 35/2) is furnished with a subscription and with the full range of marginal and asterisked variants. Important steps in the version’s history are: 1.  the creation of a ‘Passiontide Harmony’ composed of literal segments of the single Gospels. It is from the 9th cent. and attributed to Daniel of Beth Baṭin (near Ḥarran); 2. the liturgical use of the version, which entered the lectionaries. This gives color to the version’s appreciation among the Syrians as the ‘correct’ and ‘ecumenical’ text of the NT; 3. the adoption of the version by the Masora (connected with the Qarqaphto monastery near Reshʿayna) in the 8th/9th cent.; 4. the production of bilingual codices in Syriac/Arabic Garshuni (e.g., Vat. Syr. 271 and Paris, Bibl. Nat. Syr. 57); and 5. the attachment of extensive exegetical scholia taken from patristic authorities (e.g., Mingana Syr. 105 and 480; Manchester, J. Rylands Univ. Library, no. 10).

The history of research started with the Editio princeps published by Joseph White 1778–1803; it was based mainly on one single ms. (New Coll. 333, with variants of mss. New Coll. 334 and Bodl. Or. 361). From the very beginning scholars were confronted with the question of the version’s identity (Philoxenian or Ḥarqlean). According to White’s interpretation of the subscriptions it was the Philoxenian version he published, which Tumo of Ḥarqel reissued by only attaching the marginal and asterisked readings. G. H. Bernstein in 1837 considered White’s text to be a separate version, though a thorough revision, of the Philoxenian. The question remained open until the publication of Philoxenos’s late writings, which provided genuine Philoxenian quotations of the NT (especially his commentary on the prologue of St. John’s Gospel, ed. by A. de Halleux [CSCO 380, 1977]). Bernstein’s view turned out to be substantially correct; the Philoxenian version itself is considered lost.

A new approach to the Greek background of the Ḥarqlean Version was launched by B. Aland, who successfully reconstructed the Greek Vorlage of the Ḥarqlean major Catholic Epistles and the writings of Paul. The version’s Greek model can be traced to a group of 10th–15th-cent. minuscule mss. (1611, 1505, 2138, 2495 according to Gregory and Aland), the ancestry of which is not a descendent of the Greek model itself but of a ms. akin to this model having a common 5th/6th-cent. ancestry (B. Aland 1986, introduction). The knowledge of the Greek model opened a new perspective on Tumo’s method of revison, with two principal issues: 1. Tumo is a translator, who faithfully follows his Greek Vorlage; the marginal and asterisked annotations derive from different Greek mss., which supply the version with variant readings of the Graeca veritas; 2. the revisional relation to the Philoxenian (reported in the subscriptions) is rather on the Syriac than on the Greek level. There was a shift from a fairly exact rendering to a ‘mirror translation’; the Greek background of both versions largely remained the same. These results derive from a text-critical study and from a reedition of the Ḥarqlean Version, which is based on a new important ms. (Jerusalem, St. Mark 37; 9th cent.). G. Zuntz’s still dominant approach to the revisional relation between the Philoxenian and Ḥarqlean versions is solely based on an interpretation of the subscriptions and guided too much by the term ‘revision’ instead of ‘translation’.

The complex structure of the Ḥarqlean Version, its liability to distortion, the revisional development, and the scanty attestation of the Praxapostolos confront editors with not a few problems. Not surprisingly, most of the Ḥarqlean editions are one-ms.-publications. A comprehensive critical edition of the Gospels, which sets out the revisional development of the version in some detail, is much to be desired. Ḥarqlean texts published so far are: L. de Dieu, Apocalypsis Sancti Iohannis (Leiden, 1627) (ms. Hebr. Scal. 18); J. White, Sacrorum Evangeliorum Versio Philoxeniana ..., vol. 1, 1–2 (Oxford, 1778) (Gospels); vol. 2, 1–2 (Oxford, 1799/1803) (Acts and Letters) (ms. New Coll. 333, with variants of mss. New Coll. 334 and Bodl. Or. 361 [all Oxford]); G. H.  Bernstein, Das heilige Evangelium des Iohannes. Syrisch in harklensischer Übersetzung  … (Leipzig, 1853) (ms. Vat. Syr. 271); R. L. Bensly, The Harklean version of the Epistle to the Hebrews, chapter XI,28–XIII,25 (Cambridge, 1889) (ms. Cambridge, Add. 1700); M. A. Weigelt, Diatessaric Harmonies on the Passion Narrative in the Harclean Syriac Version (Ph.D. Diss.; 1969) (based on mss. Vat. Syr. 268; London, Brit. Libr. Add. 18,714; Paris, Bibl. Nat. Syr. 31, 51 and 52; Cambridge, Add. 1700); A. Vööbus, The Apocalypse in the Harklean version (CSCO 400; 1978) (ms. Mardin Orth. 35/2 [i.e., 36/2]); B. Aland and A. Juckel, Das Neue Testament in syrischer Überlieferung I (1986) (James, 1 Peter, 1 John; according to mss. London, Brit. Libr. Add. 14,474; Oxford, New Coll. 333; Cambridge, Add. 1700); B. Aland and A. Juckel, Das Neue Testament in syrischer Überlieferung II,1–3 (1991, 1995, 2002) (the corpus of Pauline epistles; according to mss. Jerusalem, St. Mark Syr. 37; Oxford, New Coll. 333; Cambridge, Add. 1700); G. Kiraz, Comparative Edition of the Syriac Gospels (4 vols.; 1996) (ms. Vat. Syr. 268, supplemented by ms. Vat. Syr. 267 and Florence, Bibl. Laurenziana Plut. I,40); P. A. L. Hill, The Harklean Version of St. Luke 1–11: A Critical Introduction and Edition (Ph.D. Diss., Melbourne; 2002) (based on 19 mss.).