This article is also available in text-only format.

Kiraz, George Anton. Comparative Edition of the Syriac Gospels: Aligning the Sinaiticus, Curetonianus, Peshitta and Harklean Versions. Vol. 1: Matthew; vol. 2: Mark; vol.3: Luke; vol. 4: John. Leiden: Brill, 1996. ISBN: 90-04-10419-4 (set). Pp. xciv+454+7 plates; ii+257; ii+514; ii+376. US $529.25 (set).

1. The versions of the New Testament are among the least understood, least used tools for studying the text of the NT. They are least understood because, in order to use a version correctly, one must immerse oneself in the peculiarities and particularities of that version: its manuscripts, syntax, vocabulary, and transmission history. They are least used because this immersion means not only learning the language of the version itself but also the languages in which the secondary literature about the version appears. While these obstacles have their positive side (one of the signal identifying marks of the novice is his or her ignorance concerning the versions and patristic material), it is also unfortunate, for the rich textual diversity of the first Christian centuries is made manifest in these sources.

2. Among the versions of the New Testament, the two most important are the Latin and the Syriac. This is not only because of their age (the genesis of both can be dated to between roughly 150 and 180 C.E. [as evidenced by the Acts of the Scillitan Martyrs for the Latin and the Diatessaron for the Syriac]), but also because of their positions as the hyparchetypes from which many other versions descend. For example, the Old High German and all the European vernacular versions (with the exception of the Gothic) are based on the Latin, not the Greek; in the East, the oldest Armenian, Arabic, and Georgian (via the Armenian) versions appear to be dependent upon a Syriac--not a Greek--base. The early date means that these versions (viz., the Latin and the Syriac) may preserve very ancient readings, present in the second-century Greek archetype from which they were translated. The fact that this is so is corroborated by the presence of identical variant readings in other late second- and early third-century sources (viz., the Fathers [e.g., Justin, Clement of Alexandria, Origen, etc.] and the apocrypha [e.g., the Gospel of Peter, Gospel of Thomas, the Protevangelium Iacobi, etc.]). The Syriac is also commended by the fact that it is in a Semitic language--"Christian Aramaic," in the words of F. C. Burkitt--virtually identical with the language presumably spoken by Jesus and his disciples. Hence, the diction, idioms, and syntax found in this version deserve special attention. If, as Papias (as per Eusebius) and Jerome and Epiphanius claim, a "Hebrew" or "Aramaic" Matthew circulated in the early church (Jerome even remarks that this Semitic-language Matthew was regarded as the "original" or "autograph" Matthew "by many": " quod uocatur a plerisque Mathei authenticum..." [Comm. Matt. II, apud 12:13 (CChr.SL 77, 90, 368-369)]), then its language would presumably have been similar to that found in our oldest Syriac version.

3. George Kiraz--formerly a student of Sebastian Brock at the Oriental Institute in Oxford, and now associated with the successor corporation to Bell Labs--has combined his data processing expertise and his love of early Syriac texts to produce a series of volumes which are foundational for critical study of this version. In 1993, he compiled the first concordance to the Syriac NT ever published (Kiraz 1993). In 1994, he published a set of Lexical Tools to the Syriac NT (Kiraz 1994). In 1996, we encounter these four volumes on the gospels.

4. If one ignores the Diatessaron (which is the oldest gospel text in Syriac), then three recensions of the gospels in Syriac exist. (A) The oldest of these three is the vetus syra or "Old Syriac," which exists in two manuscripts: Codex Sinaiticus (Syrs or Syrsin, dated to the mid- or late-fourth cent.) and Codex Curetonianus (Syrc or Syrcur, early fifth cent.). It must be pointed out that these two manuscripts do not appear to be related; rather, each seems to represent a more or less independent translation of a Greek archetype (the Greek archetype apparently differed, as well); that this is the case is demonstrated by the differences in (1) word order, (2) vocabulary choice, (3) handling of passages in the Greek which required circumlocution in the Syriac, etc.

5. (B) The second oldest version is the "Peshitta" (= "common" or "vulgate"; Syrp or Syrpes), extant in over 350 manuscripts (the oldest of which dates from the fifth cent.). Its genesis is placed in the early- to mid-fifth century. Unlike the vetus syra, whose circulation was apparently limited (it was overshadowed by the more ancient Diatessaron), the Peshitta enjoyed the approval of clergy whose allegiance was to the Western "Great Church"; it became the standard NT of the Syrian church. The Diatessaron--which from antiquity had been the standard text of the Syrian church--was swept aside in the 420s by the "Great Church" bishops (e.g., Rabbula of Edessa, Theodoret of Cyrrhus), whose allegiance lay with Rome and Constantinople, not the traditions of Edessa and Jerusalem (cf. the Doctrina Addai and the much-remarked upon and striking disjunction between bishop Aggai [Jerusalem-oriented] and his successor bishop Palut [Rome-oriented]; see, e.g., Bauer 1971: 16-17). The Peshitta represents a careful, quite consistent rendering of its fourth- or fifth-century Greek base.

6. (C) The most recent of the ancient Syriac versions is the Harclean (Syrh), named after Thomas of Harqel, who, in 616, revised the now-lost "Philoxenian" version created by Philoxenus, bishop of Mabbug. Philoxenus' version saw little success, and all manuscripts of it are lost. It survives only in the revised version issued by Thomas (the oldest manuscript of the Harclean dates from the eighth/ninth century; the relationship between the Harclean and the Philoxenian versions was finally stipulated by S. Brock, in his study "The Resolution of the Philoxenian/Harclean Problem" [Brock 1981]).

7. As is well known, modern critical editions of the NT cite their sources inconsistently, generally incorporating only those variant readings deemed significant by the editor(s) (the lone post-World War I exception is Nouum Testamentum Graece/The New Testament in Greek 1935-, prepared under the auspices of the International Greek NT Project, but its text is not an "edition," and its apparatus is--despite its claims to the contrary--incomplete and [perhaps inevitably] error-prone). Hence, when Nestle-Aland/UBS or other editions list "Syrs" in their apparatus, that may be only one of several variants in the same verse; the others were ignored. Such a procedure is inescapable in a pocket edition, but it is obviously an obstacle for a thorough study of the text of the versions. Therefore, when working with the Syriac, the only way to have the full text of the three main traditions (that is, the two manuscripts of the vetus syra and the Peshitta) at one's disposal was to have the three editions named below spread open before oneself. Kiraz's Comparative Edition of the Syriac Gospels changes all that.

8. Each of the four volumes under review is devoted to one of the gospels. The text of each gospel is set out in four horizontal parallel lines, à la Adolf Jülicher's Itala (Jülicher 1938-). The top line is the oldest tradition, namely the Sinaitic Syriac manuscript of the vetus syra (Lewis 1910); the second line is the text of the Curetonian manuscript of the vetus syra (Burkitt 1904); the third line is the Peshitta (Pusey and Gwilliam 1901); the fourth line is--in the absence of a reliable modern edition--the text of a Harclean manuscript (primarily Vat. Syr. 268, but supplemented by Vat. Syr. 267 and Bibl. Laurenz. Plut. 1.40 when Vat. Syr. 268 is defective). The result is a bird's-eye view of roughly half a millennium of textual history. The parallel horizontal line format means that omissions, substitutions, and transpositions are immediately apparent. One can see how the gospel text in Syriac changed between the earliest period (represented by the text of the Sinai and Curetonian mss) and the later period (the Harclean ms[s]). The direction of this change is consistent with what we know from other versions and Patristic citations: just as the sometimes "wild" vetus latina readings are edited out by Jerome in his Vulgate (whose Latin text is close to the Greek text), so here the "wild" vetus syra readings, present in the earliest period (and in some cases clearly drawn from the Greek manuscript[s] upon which this early translation into Syriac was based), are edited out as we move ever closer to the (later) Greek text. This process of assimilation culminates in the Harclean, which is a virtual mirror image of the Greek text current in the seventh century. As Arthur Vööbus remarks:

[in the Harclean] the process of Hellenization celebrated its triumph. It is strange how little respect is shown in [the Harclean] for Syriac idiom and linguistic taste. What [the Harclean] offers is a completely slavish adaptation to the Greek to the extent that even clarity is sacrificed (Metzger 1977: 118).
This imitation of the Greek descended to the level of attempting, where possible, to keep the same number of words in a verse, and in the same order, even to the location of particles. (On the famous "Harclean margin" [Syrhmg], see the definitive analysis by B. Aland [Aland 1986: 111-127].)

9. There is much of value to be gained by the study of the Syriac version of the Gospels, and Kiraz's volumes facilitate these endeavors. Not only do we have a splendid paradigm for the evolution of the gospel text--from the earliest stages (represented by the vetus syra, which was translated from the early Greek gospel tradition), replete with "wild" readings, to the "slavish" conformity to the "sanitized" Greek gospels current in the seventh century (represented by the Harclean)--but we also have numerous interesting readings. Some of these variants are singular to the Syriac; in such instances one cannot know whether the reading is a local variant within the Syriac tradition or a variant present in the Greek base of the third century from which the vetus syra was translated. Two examples of this phenomenon are: (1) Luke 19:44, where the Curetonian Syriac and the Syrian father Aphrahat (first half of the fourth cent.) read, "...the time of your greatness," against the standard "...the time of your visitation"; (2) John 3:18, where the Sinaitic Syriac reads "the chosen son," against all other witnesses, which read "the only-begotten son of God."

10. Other variants in the Syriac are, however, paralleled in the earliest fathers (e.g., Justin, Clement, Origen) and in other early versions (e.g., the vetus latina). In such instances one can be certain that one has recovered a second- or third-century Greek reading. Two examples of this phenomenon are: (1) the omission in the Sinaitic Syriac (Syrs) of Matt 9:34, an omission supported by Greek D (05: Bezae), vetus latina mss a (the oldest extant vetus latina ms) and k (the prime representative of the ancient afra tradition within the vetus latina), as well as the Arabic Diatessaron; (2) the interpolation in Luke 23:48 of "and saying: 'Woe to us! What has befallen us? Woe to us from our sins!'" post "their breasts" in the standard text. The same interpolation (or a recognizably similar one) is found at the same point in Luke in vetus latina ms g1 (Codex Sangermanensis; ninth cent.), Ephrem's Commentary on the Diatessaron, Aphrahat, the Doctrina Addai (Ephr, Aphr, and Addai are early Syrian sources), and the Greek Gospel of Peter.

11. Inasmuch as Kiraz is relying upon established editions, there are few criticisms. No translation is provided, so a knowledge of Syriac is a prerequisite for using the volumes effectively. The novice may be misled by the representation of the Peshitta by a single line of text when, in fact, its tradition is represented by many manuscripts, with many important variant readings. To confine the Peshitta to a single line is as misleading as to represent "the Greek" tradition by the text of Nestle-Aland/UBS. Experts will know that it is always necessary to go to Pusey and Gwilliam's edition and check its apparatus, but others may blithely assume they have "the" text of the Peshitta before them. The same criticism applies to the Harclean line, which reduces an entire tradition to a single line. (In both cases we must wait until the Münster project Das Neue Testament in syrischer Überlieferung reaches the gospels for a full display of the evidence--including Syriac Patristic citations.) But these are minor quibbles when weighed against the brilliance of Kiraz's contribution, which permits one the first "at-a-glance" view of the evolution of the text of the gospels in Syriac. Indeed, the project would have been cluttered if even a dozen Peshitta lines (each representing a manuscript) had been added.

12. In addition to the presentation of the parallel-line text, the first volume (on Matthew) contains a thorough eighty-two page "Introduction," setting out (1) a brief history of the Syriac version, (2) a description of the manuscripts and the editions used, (3) a description of the method of presentation, and (4) a splendid "Introduction to the Harklean Text" (pp. xxxi-lxxxii) by Andreas Juckel of the Münster Institut.

13. As both a tool and as an original contribution to scholarship (viz., the presentation of the text of Vat. Syr. 268), these volumes are a boon to textual studies. Never before has the text of the Syriac version been so readily available; never before has the textual evolution of the Syriac version been so readily discernible.

© TC: A Journal of Biblical Textual Criticism, 1997.


Aland, Barbara 1986. Das Neue Testament in syrischer Überlieferung. Vol. 1: Die grossen katholischen Briefe. ANTT 7. Berlin and New York: de Gruyter.

Bauer, Walter 1971. Orthodoxy and Heresy in Earliest Christianity. Translated and edited by Robert A. Kraft, Gerhard Kroedel, et al. Philadelphia: Fortress. [Electronic edition by Robert A. Kraft, gopher://]

Brock, Sebastian 1981. "The Resolution of the Philoxenian/Harclean Problem." In New Testament Textual Criticism: Its Significance for Exegesis, ed. Eldon J. Epp and Gordon D. Fee, 325-342. Festschrift for Bruce Metzger. Oxford: Clarendon.

Burkitt, F. Crawford 1904. Evangelion da-Mepharreshe: The Curetonian Version of the Four Gospels. Cambridge: Cambridge University Press.

Jüicher, Adolf 1938-. Itala : Das Neue Testament in Altlateinischer Überlieferung. Berlin: de Gruyter.

Kiraz, George Anton 1993. A Computer Generated Concordance to the Syriac New Testament. 6 vols. Leiden: Brill.

Kiraz, George Anton 1994. Lexical tools to the Syriac New Testament. JSOT Manuals, no. 7. Sheffield: JSOT Press.

Lewis, Agnes Smith 1910. The Old Syriac Gospels, or, Evangelion da-Mepharreshe. London : Williams and Norgate.

Metzger, Bruce M. 1977. The Early Versions of the New Testament. Oxford: Clarendon.

Nouum Testamentum Graece/The New Testament in Greek 1935-. Edited by S. C. E. Legg and the International Greek New Testament Project. Oxford: Clarendon.

Pusey, P. E., and Gwilliam, G. H. 1901. Tetraeuangelium sanctum juxta simplicem Syrorum versionem ad fidem codicum, Massorae, editionum denuo recognitum. Oxford: Clarendon.

William L. Petersen
Department of Religious Studies,  Pennsylvania State University