The Peshitta and other important versions
A. The Syriac Versions of the New Testament
These versions are of the highest importance for the critical study of the Gospels, especially for criticism of the text.
1. The Peshitta
That is, the “simple” version is said to be so-called by contrast with the later Havcleian version (see infra), though Moses bar Cephas, d. A.D. 913, so terms it by way of contrast with the Syro-Hexapla version which retained the obeli and asterisks inserted by Origen in his edition of the Septuagint; it must be confessed that this is a far more rational explanation of the term since the Harcleian version is a slavish translation of the Greek, and would itself merit the term “simple.” The Peshitta version is that to be found in the ordinary printed Syriac text, and corresponds to the Latin Vulgate; both the Peshitta and the Vulgate have merited the appellation of “Queen of Versions.”
The date of this version is a subject of keen dispute (see infra); one school would refer it to the second century A.D., another would regard it as a revision dating from the first half of the fifth century. Whatever conclusion may be arrived at on this point it remains that we have more numerous and more ancient MSS. of the Peshitta than of almost any version, some of them even dating from the fifth century.
2. The Curetonian Syriac Version
In the year 1858 Dr. Cureton published from a collection of Syriac MSS. discovered at the monastery of St. Mary Deipara, in Nitria, a MS. of the Gospels, or rather, of large portions of them. It was only natural that the editor should magnify the importance of his “find,” but it was hardly to be expected that he should imagine that he had discovered “the original — to a great extent — of the Hebrew Gospel of St. Matthew”! It was generally agreed, however, that the Curetonian Syriac, C, as it came to be called, represented an older version than the Peshitta; hence it is commonly known as “the Old Syriac,” O.S. Cureton himself assigned the actual MS. to the middle of the latter half of the fifth century, Burkitt would assign it to the opening of that century. The Curetonian differed largely from the Peshitta, and it was natural to suppose that the latter represented a later state of the Syriac version, perhaps even a formal revision of it, while the Curetonian stood for the Old Syriac text which had prevailed previous to that revision. The relation between the two versions would, on this hypothesis, be parallel to that which subsists between the Old Latin and the Vulgate Latin.
3. The Sinaitic Syriac
In 1892 Mrs. Lewis discovered in the Convent of St. Katharine on Mount Sinai a palimpsest MS. of the Syriac Gospels. This is known as the Codex Lewisiana, or the Sinaitic Gospels, S; Burkitt would refer it to the end of the fourth century. This MS. proved to be much more in agreement with the Curetonian than with the Peshitta; it was therefore hailed as another example of the hypothetical pre-Peshitta unrevised text, and took its place with the Curetonian as part of the “Old Syriac.” The curious feature about this MS. is its heretical character, thus in Matt. 1:16 “Joseph, to whom was espoused Mary the Virgin, begot Jesus,” 1:21; “she shall give thee a son,” 1:24; “he took unto him his wife and she gave him a son”; and, as though to accentuate these heretical statements, the first portion of 1:25 “And he knew her not till she brought forth…” is omitted! It is worth recalling that St. Jerome  speaks of the Gospel of the Hebrews as composed in Syriac though presented in Hebrew characters. Moreover, he calls it an Ebionitic Gospel, and these features of the Sinai-Syriac are distinctively Ebionitic. The omissions of the ordinarily disputed passages are equally striking; the last twelve verses of Mark are omitted, likewise the Sweat of Blood, Luke 22:43-44, and the story of the woman taken in adultery, John 7:53 — 8:11. But the most curious omission is that of Luke 23:10-13, the reconciliation of Pilate with Herod, a passage which is elsewhere uniformly attested.
4. Tatian’s Diatessaron
This was a harmony of the Four Gospels which was apparently in almost universal use at one time. Eusebius refers to it as a work which was, apparently, composed by Tatian after he had ceased to be a disciple of St. Justin at Rome and had returned to Assyria: “Tatian,” says Eusebius, “composed a harmony and collation (συνάφειάν τινα καὶ συναγωγήν) of the Gospels, I know not how; to this he affixed the name of Diatessaron, and it is even now in the hands of some.” It is uncertain whether this Harmony was made according to the Greek or the Syriac text; the latter is the more probable view. It has perished long since, but we have (a) St. Ephraim’s Commentary on it in an Armenian version, (b) an Arabic version of it from the eleventh century, and (c) the Latin Codex Fuldensis of the Gospels; this latter was drawn up by Victor of Capua about A.D. 540, but it “is completely assimilated to the Latin Vulgate; and the order of the events, while agreeing in the main with the Arabic Harmony and the Commentary of St. Ephraim, has in many places been altered.”
5. The Philoxenian Syria
cThis is a literal translation of the entire Bible save the Apocalypse; it was made in A.D. 508 by the Chorepiscopus Polycarp for Philoxenus, Bishop of Mabug, A.D. 488-518.
6. The Harcleian Syriac Version
This was made in the year A.D. 616, when Paul of Tella translated the Old Testament into Syriac and Thomas of Harkel did the same for the New Testament; the version was made at Alexandria. The Greek MSS. used for Acts resembled Codex Bezae; for the Epistles recourse was had to MSS. in the library at Caesarea. At one time it was thought that the Apocalypse was not included, but it was discovered in A.D. 1627.
These two translations represent an undoubted revision and are therefore of interest, in view of the controversy touching a possible revision by Rabbula at the opening of the fifth century; the former revisions have left traces, the latter practically none.
7. The Karkaphensian version (so-called)
This is not a version at all, but corresponds to the Massorah on the Hebrew text, i.e. it is a collection of passages given with a view to the preservation of the best traditions touching the orthography and pronunciation.
B. The Syriac N.T. Canon
The question of the completeness of the Canon in use in the Syrian Churches is still undecided. The treatise known as The Doctrine of Addai (Thaddeus) has the following remark: “The Law and the Prophets, which ye read every day before the people, and the Epistles of Paul, which Simon Peter sent us from the city of Rome, and the Acts of the Twelve Apostles, which John the son of Zebedee sent us from Ephesus; these Books read ye in the Churches of Christ, and with these read not any other, as there is not any other in which the truth that ye hold is written.” This would seem to exclude all the “Catholic Epistles.” Indeed, a Codex found at Mount Sinai, and published in the Studia Sinaitica, I. p. 11 ff., gives a list of all the Canonical Books of the N.T. showing the stichi or clauses which each contained. These embrace “the Four Gospels,” the “Acts,” and “Paul the Apostle,” and there follows immediately the total for “the Holy Books which the Holy Church receives.” Hence it is commonly held that the other books found in the printed editions of the Peshitta were incorporated from the later Philoxenian version. Similarly Gwilliam says: “The Peshitto Canon does not include 2 Peter, 2-3 John, Jude, Revelation.” But Gwilliam himself cites a MS. of the Peshitta as containing Jude, while various old translations of the Apocalypse have been discovered of late years, and it is claimed for them that they represent an old version. In addition we have to take into account the witness of a remarkable Syriac-Chinese stone inscription erected at Singan-fu in A.D. 781 and discovered by the Jesuit missionaries in 1625; this speaks of the twenty-seven Books of N.T.
C. These Syriac Versions are an Important Factor in the Textual Problem of New Testament
We have four series of documents: the Peshitta Syriac, P, the Curetonian, C, the Sinaitic, S, and Tatian’s Diatessaron, T. What is their relative order in point of time? To what extent are they dependent on one another?
Can we say, for instance, that C and S represent the Syriac Gospels as they stood in the second century, that T presents a harmony based on them, and that P represents a later revision which, like St. Jerome’s Latin Vulgate, held the field?
This would be an exceedingly simple solution, but there are many difficulties in the way. In the first place, C and S differ considerably from one another. Secondly, we have no real trace of the actual text of T. Thirdly, it is difficult to find any historical allusion to a revision which must have been well known. The hypothesis that P did actually represent such a revision was put forward by Dr. Hort, and consequently P became known as the Syriac Vulgate, as though there was a demonstrated historical resemblance between it and St. Jerome’s Latin Vulgate, both of them, that is, being due to a formal revision. Similarly P has been aptly termed “the Queen of versions,” an honorific title equally due to the Latin Vulgate. But while St. Jerome’s revision is a great historic fact, the same can hardly be said for a revision of the Syriac Gospels. Still we have, it is true, three historical statements which must be weighed. Theodoret, Bishop of Cyrrhus, d. A.D. 457, says:
“He (Tatian) also compiled the Gospel known as the Diatessaron, in which he omitted the genealogies as well as other passages which show that the Lord was born of the seed of David. And not only do those who belong to his following use this, but even those do so who follow the Apostolic teachings; for they do not detect the fraud of this compilation, but use it in guileless fashion as a compendious book. I myself found more than two hundred of these books held in honor in churches of ours, and I collected them all and set them on one side and introduced in their place the Gospels of the Four Evangelists.”
The second statement is that made by the biographer of Rabbula, Bishop of Edessa, A.D. 412-435, viz., that this Bishop ordained that in all Churches the “Gospels of the Separate” should be kept and read. This term “Separate” can only be understood as applying to the “Separate” Gospels as distinct from the Diatessaron. It is of interest, then, to remark that both C and S bear the title of “Separate” Gospels, da Mepharreshe. Unless this title is a later addition we seem to have in it a proof that C and S are later than T. The third statement bears directly on the question of revision: the above-mentioned biographer says that Rabbula “translated the N.T. out of the Greek into Syriac, because of its variations, exactly as it stood.” It has been suggested that this translation is our Peshitta P. An interesting attempt was made by Mr. Woods to overthrow this hypothesis by showing  that St. Ephraim, d. A.D. 373, was well acquainted with P, and that consequently this version was well known some fifty years previous to Rabbula’s work. But Burkitt was able to show that the Roman edition of St. Ephraim’s works contained an immense amount that was spurious, and that whereas Woods had maintained that out of 168 quotations of the Gospels in Ephraim no less than 43 were exactly P, the truth was that after elimination of the spurious works there only remained 8 quotations of P by St. Ephraim. Burkitt further showed that Ephraim in reality used T, and was three times as often in agreement with C and S as with P. These discussions seemed to point to the conclusion that P could not be traced back further than the date of Rabbula, and that there was therefore an antecedent probability that the translation made by the latter really was our present P. Confirmation of this view was sought in the oft-repeated statement that P coincided in text with Codex Alexandriniis, with the later Uncial MSS., with the mass of the Cursive MSS., and with the text found in St. Chrysostom; in other words, with the text dubbed “Syrian” by Westcott and Hort on account of its presumed connection with Antioch and St. Chrysostom.
But Gwilliam  retorted that the dependence of P on Codex Alexandriniis was by no means proved, since out of 167 passages taken from Matt, i-xiv., no less than 65 supported B or Codex Vaticanus, viz. the “Neutral text,” to which Westcott and Hort practically pinned their faith, while 108 (many of the passages contained more than one disputed point, hence the apparent discrepancy in numbers) supported the Traditional text. By the “Traditional text” is not meant the so-called textus receptus of the Elzevirs, but “the text which has been handed down to us by and in the Catholic Church, and which is contained in the mass of copies and is attested by ecclesiastical writers.” Gwilliam further pointed out that in 137 other passages in these same chapters P agreed neither with B nor with the Traditional text, but in the majority of these passages agreed either with C or with S or with both; while in many it agreed with certain Old Latin MSS. or with a few Greek MSS. In view of these facts Gwilliam was justified in maintaining that (a) sufficient proof of a revision of the Syriac N.T. had not been brought forward, due consideration being paid to the power and importance of the Syriac Church and to the effect such a revision must have had on the Church at large; (b) that whereas abundant traces of the pre-Hieronymian Latin text remained in the shape of Old-Latin MSS., no such traces of the pre-Rabbula text were to be found; (c) that while Jerome’s Vulgate only found its way into favor by slow degrees, the supposed revision of the Syriac text, viz. P, exists in an immense number of MSS. of which some at least belong to the fifth century, i.e. to the period assigned to this supposed revision. Gwilliam further pointed out that on the supposition that it could be shown that P represented Rabbula’s revision in the early part of the fifth century, the only result would be a further canonization of P; for it is certain from the account of Rabbula’s translation that he would have made it with care and from the best Greek MSS. available. But in that case P would represent a version practically as old as the Latin Vulgate, and one for our knowledge of which we have MSS. both older and more numerous than those of the Vulgate. It would also represent what Rabbula held to be the best Greek text then available. And that Greek text is by no means simply the Traditional text, nor does it agree with B and the Westcott and Hort groups; as we have seen, it stands alone in thirty-one places in Matt, i-xiv. Therefore Gwilliam draws the further conclusion: either these variants of P are found in the Cursive Greek MSS. or not. If they are not found there, then the Cursives are justly relegated to the background in the textual controversy; but if they are found scattered up and down in the mass of Cursives at our disposal, then we have no right to pin our faith simply to the great Uncials. Thus the value of P for the textual criticism of the Greek New Testament is immense, whether it belong to the second century or to the fourth, whether it be an original text or a revision.
While it must be acknowledged that Burkitt’s argument in disproof of the statement that St. Ephraim was well acquainted with P is exceedingly strong, it must not be forgotten that he was compelled to allow that in the authentic works of that Father there were eight passages which showed a knowledge of P. But whether there are eight passages or only one matters little. One single passage which betrays Ephraim’s indubitable acquaintance with that version is sufficient to prove that P was not the result of a revision by Rabbula. Moveover, on the supposition that P really is due to Rabbula, how are we to explain the amazing gaps in the Canon of the Peshitta? By the opening of the fifth century the present Canon of N.T. was practically universally accepted. Yet P apparently never contained the Catholic Epistles. If this is a fact it is an awkward one, for such an omission at the date of Rabbula’s episcopate is well-nigh incredible. Thus, briefly, while there are certain a priori arguments for assigning P to a late date and regarding it as a revision, yet there are one or two considerations, e.g., the deficiencies in the Canon and the fact that St. Ephraim has eight acknowledged quotations which betray an acquaintance with P, which must make us hesitate before accepting P as a fifth-century revision carried out by Rabbula.
Nor is this discussion one of purely academic interest. For if the early date of P as it now stands could be proved, then the whole critical edifice so laboriously raised by Westcott and Hort would come tumbling down. Since, if the true place of P be that assigned to it in the above scheme, then it will still be of first-rate importance, in view of the statements made by Gwilliam (see  above). But if the Peshitta is not really a late revision its importance is immensely enhanced; for then the unusual underlying Greek text to which it bears witness will not be merely the best Greek text known to Rabbula at the opening of the fifth century, but will represent the Greek text current among the Syriac-speaking Fathers two centuries earlier at least. It has been stated that the early date of the Peshitta is “the sheet-anchor” of the upholders of the Traditional text. This is misleading if it is meant to imply that the disproval of its early date means the downfall of the Traditional text. That text does not depend on any one MS. or family of MSS. It depends on the concurrent testimony of a cloud of witnesses. If one of these proves a frail prop the system or theory does not come tumbling down like a house of cards; it is only one of many props, and the others, the host of Cursive MSS., the later Uncials, the Fathers, etc., are well able to uphold it. The same cannot be said for the neutral type of text. It is dependent on one or two MSS., if these fail the results are disastrous.