THE FOUR GOSPELS - A Study of Origins, The Manuscript Tradition, Sources, Authorship, & Dates. By B.H. Streeter, Hon.D.D.Edin. Fellow of Queen's College, Oxford; Canon of Hereford. First published by MacMillan & Co Limited 1924. - Prepared for katapi by Paul Ingram 2004.




IN the Preface to the Vulgate Gospels, which takes the form of an open letter to Pope Damasus, Jerome defends the principles he has adopted in revising the text of the Old Latin. Of the Latin, he says, no two copies are alike, but we have a standard of authenticity, since for the New Testament, unlike the Old, the Greek was indubitably the original. Hence the discrepancy between the different Latin copies can only be corrected by reference to the Greek original. But the MSS. of the Greek text, which are named after Lucian and Hesychius, both in the Old and the New Testament, have been badly edited and interpolated as may be seen by comparing them with ancient vernacular translations. He has, therefore, used, as a standard by which to correct the Latin, Greek MSS. which are really old.

Si enim Latinis exemplaribus fides est adhibenda, respondeant, quibus; tot enim sunt exemplaria quot pene codices. Sin autem veritas est quaerenda de pluribus, cur non ad Graecam originem revertentes, ea quae vel a vitiosis interpretibus male edita, vel a praesumptoribus imperitis emendata perversius, vel a librariis dormientibus aut addita sunt, aut mutata, corrigimus? Neque ego de Veteri dispute Testamento. ... De Novo nunc loquor Testamento: quod Graecum esse non dubium est. ... Hoc certe quum in nostro sermone discordat, et diversos rivulorum tramites ducit, uno de fonte quaerendum est. Praetermitto eos codices, quos a Luciano et Hesychio nuncupates, paucorum hominum asserit perversa contentio: quibus utique nec in Veteri Instrumento post septuaginta interpretes emendare quid licuit, nec in novo profuit emendasse: quum multarum gentium linguis Scriptura ante translata, doceat falsa esse quae addita sunt. Igitur haec praesens praefatiuncula pollicetur quattuor tantum evangelia, quorum ordo (he means the Greek as opposed to the Old Latin order) est iste, Matthaeus, Marcus, Lucas, Johannes, codicum Graecorum emendata collatione, sed veterum.

Seeing that Jerome is writing a careful and considered Preface to a revised version of the Four Gospels, and that he only mentions the Lucianic and Hesychian versions in order to contrast their inferior text with that of the "ancient codices" he has himself used, I simply cannot understand why some scholars have raised doubts as to whether the Lucianic and Hesychian recensions included the New Testament as well as the Old.

But why, we may ask, was Jerome so contemptuous of the work of Lucian and Hesychius?
I suggest two reasons:
(1) Jerome at this date was quite convinced that the true text of the O.T. was only to be found in Origen's Hexapla [Cf. passages quoted by Swete, op. cit. p. 76 f.]; it was only later in his life that he had recourse to the original Hebrew. But the Hexapla had been published a generation before Lucian and Hesychius began their work. What then but native perversity, "a cantankerous wrongheadedness" (perversa contentio), could have induced these worthies to insist on producing a text of their own?
(2) Jerome had just returned to Rome from Constantinople, where, as he tells us, the recension of Lucian was accepted. Now, if we accept the suggestion (cf. p. 102 ff.) that the MSS. with which Constantine had provided his new capital represented the fam. Θ text, the Lucianic text must have been a recent importation. Inevitably, then, there would have been some conservatives at Constantinople who grumbled at the innovation; and these might have very pertinently appealed to the Old Syriac (multarum gentium linguis scriptura antea translata, doceat falsa esse quae addita sunt), as evidence that the many striking passages absent from both fam. Θ and Syr S. (cf. p. 88) but found in the Lucianic text, were interpolations. If Jerome had heard from someone at Constantinople some general statement to this effect, it would only confirm his a priori suspicion of anyone who dared to think he could improve upon Origen's text of the LXX, and, with his hasty temperament, he would forthwith conclude—without any really careful study—that Lucian's temerity had had equally fatal results on the New Testament. Nor would he wait for further evidence before including Lucian's fellow offender, Hesychius, in the same sweeping condemnation.

The question, however, in regard to which we should have liked Jerome to have made a clear statement is the type of text represented by those ancient MSS. which he himself so much preferred.
Can we identify the Greek text which Jerome preferred?
The materials on which an answer to this question must be based are collected in Wordsworth and White's edition of the Vulgate Gospels and Acts. [Wordsworth and White, Nov. Test. Lat. six. ed. S. Hieronymi (Oxford, 1889-98, 1905).]

(1) In his Commentary on Matthew Jerome seven times discusses various readings in MSS. known to him. Each time he quotes with approval a reading found in א. In two of the seven readings x differs from B, and in two of them it differs from fam. Θ. The fact that Jerome had a MS. agreeing with א seven times out of seven would, even if it stood alone, be a remarkable coincidence. But it does not stand alone.

(2) When in the Acts Jerome departs from the text of the Old Latin version, two-thirds of his alterations are in the direction of agreement with the Alexandrian text which for Acts is represented by B א A C. Where א and B differ he usually goes with the one which is backed by the other MSS. of the family; but six times he agrees with א alone, but only twice (sic) with B alone.

(3) For the Gospels Wordsworth and White give (p. 665) a table showing 39 readings, in which Jerome's Vulgate agrees with extant Greek MSS. against the Old Latin. Of these 25 are found in one or more of the MSS. א B L ΔΜk.; but the significance of this figure is altered when we note that only 11 of the 25 seem to be exclusively Alexandrian readings.

Everything, however, depends on their estimate of Codex f  (Brixianus), the text of which is a sort of half-way house between that of Old Latin MSS. of the type of b and the Byzantine. In f  occur a very large number of readings in which Jerome's text and the Byzantine agree against the other Old Latin MSS. Wordsworth and White hold that the Latin MSS. which Jerome started with before he began his revision must have had a text very similar to f  ; that is to say, that Jerome used a form of the Old Latin version which had already been partially revised by Greek MSS. of the Byzantine type. They conclude that the Greek MSS. that he employed must have been of the א B L type, but most closely related to א.

On this view most of the agreements of the Vulgate and the Byzantine text were due, not to Jerome, but to the previous revision by an unknown person, which is most nearly represented by f  . Burkitt, on the other hand (J.T.S. i. p. 129), thinks that the Old Latin which Jerome attempted to revise was more like b. The agreements between the Vulgate and f  he explains on the hypothesis that f  is a text very largely Vulgate which has been copied from the Latin side of a bilingual Gothic-Latin MS., the Latin, text of which had been sometimes conformed to that of the Gothic. I cannot pretend to the knowledge of the Old Latin version, which would entitle me to express an opinion on this controversy. I believe, however, that the majority of experts incline to Burkitt's view.

But if f  is, not the parent of Jerome's version, but its child, and is therefore not to be reckoned as an Old Latin MS. at all, then, so far as the Vulgate Gospels are concerned, the case for Jerome's use of MSS. of the א B L type collapses. As Wordsworth and White point out (p. 671), from Luke xvii. to the end of the Gospel, and in a large part of John, Jerome's alterations of the Old Latin are almost entirely into conformity with the later Greek MSS.; and they cite 27 instances of the same thing for Matthew, Mark, and the earlier part of Luke. It would look then as if the MSS. used by Jerome had a text something like A mainly Lucianic but with a sprinkling of Alexandrian and other earlier readings. That Jerome, without knowing it, should have used a text almost identical with the Lucianic recension, about which he is so scornful, is really funny. I can only suppose that one of the MSS. he possessed looked so old that he imagined it antedated Lucian.

There still remains, however, the far more impressive evidence quoted above that, in his Commentary on Matthew and in his revision of the Acts, Jerome used a text akin to א. I suggest that this is to be explained by his visit to Alexandria, which took place in 386, that is, between the publication of the Vulgate Gospels and the work on Matthew. [Whether his revision of the Acts and Epistles took place before or after the visit to Alexandria is unknown. His text of the Catholic Epistles is closely allied to that of A.]   At Alexandria he listened with enthusiasm to the lectures of the famous Origenist teacher Didymus.    What more likely, then, than that he should seize the opportunity to acquire a copy of the N.T. in the text, which the school of Origen approved.   
I would relate this surmise to three facts.,   
(a) Our actual MS. א was, perhaps, forty years old when Jerome was in Alexandria;
(b) Jerome died at Bethlehem in 420;
(c) not many years later א was in Palestine in the library at Caesarea. This we learn from a marginal note by a (probably) fifth-century corrector, who says that he collated some parts of the Old Testament with the autograph of Pamphilus which was there preserved.   Historically, therefore, there is no difficulty in supposing that א was brought to Palestine by Jerome.

The hypothesis that our א was actually one of the MSS. used by Jerome would remove a great difficulty. In Jerome's Commentary on Matthew, in addition to the seven variants alluded to above, there is a long discussion of the reading οὐδὲ υἱος, Mt.xxiv.36.
Jerome asserts that the words were absent from the approved copies of Origen and Pierius. But we know that as a matter of fact the words in question did stand in the text used by Origen. For Origen, in his Commentary on Matthew, discusses at considerable length the theological difficulty raised by them. He gives two alternative ways of meeting it; but he never suggests, as he does elsewhere under similar circumstances, that he knew of any MS., which omitted the offending words. Moreover, we have other evidence that they stood in all the texts likely to have been known to Origen; for they occur, not only in B א D Old Lat., but also in 13 &c., 28, Φ Arm., from which we may presume their presence in the text of fam. Θ. But if Jerome used א, and supposed it to represent the text of Origen and Pierius, his statement is explained; for although the words οὐδὲ υἱοςwere written by the original scribe, they were deleted by a very early corrector.

The deletion was made—according to the usual practice— by a row of dots above the word. A subsequent corrector has erased the dots. א was corrected by two scribes, who seem to belong to the fifth century and whose corrections can usually be distinguished by minute differences in handwriting and in the colour of the ink they used. But a row of dots, all but obliterated, cannot by these criteria be assigned to one corrector rather than another. And Tischendorf, in the notes to his four-volume edition of א, in assigning the dots to the corrector אca, and their erasure to his successor אcb, does so with the qualification ut videtur. So far as I can see, there is no reason for identifying the person who deleted οὐδὲ υἱος with the corrector אca, except that אca was the first systematic reviser who worked on the MS. after the original διορθωτής. But this particular reading was one that might well have stimulated the activity of an earlier owner of the MS. who was not concerned to revise it throughout. Origen had already found the words theologically embarrassing; during the Arian controversy the Son's knowledge of the Father was near the centre of the point at issue; the words are not found in Syr. S. nor in the Byzantine text; what is even more significant, they are absent from L 33 and from both the Egyptian versions. That is to say, we have evidence that even in Egypt, the home of the B א text, the words were being discredited as an heretical interpolation. I suggest, therefore, that the deletion in א was made when the Arian controversy was at its height, and, therefore, before the time of Jerome. Indeed, even if the MS. used by Jerome was not א itself but a sister MS., the probability that the words in question had there also been deleted is by no means low.

On the only other occasion on which Jerome refers to the copies of Origen, the reading quoted (Gal. iii.1) as Origen's is one supported by B א against the great majority of Greek MSS. This affords some evidence for the view that Jerome regarded the type of text represented by א as that approved by Origen. But if so, we must ask the question: How exactly did he connect his MSS. with those of Origen? I would hazard the suggestion that Jerome's phrase exemplaria Adamantii et Pierii (Adamantius was a second name of Origen) does not mean two different codices, any more than the text of Westcott and Hort means two different editions. Pierius may well have attempted to popularise the text of the Gospels on which Origen lectured, much in the same way as Pamphilus did for the LXX column of the Hexapla. In favour of this view two considerations may be advanced,
(a) Origen stoutly affirms his belief that the Shepherd of Hermas is an inspired work; Athanasius definitely excludes it from the Canon. Now א contains the Shepherd; whether B also did we do not know, as the end of the MS. is missing. (b) א is written in four columns to the page, B in three, A in two, C in only one which last became the common type, though two is not unusual. The larger number of columns reproduces the format of the papyrus roll which preceded the codex;
is transitional; א represents the most antiquated style of all. Thus, though it is slightly the younger MS., in this respect א reproduces an older tradition than B. This would be accounted for if א was a conservative copy of a MS. of Pierius.

The conjecture that א represents the recension of Pierius is in no way incompatible with the view that B represents that of Hesychius. Hort's arguments,
[W.H. ii. p. 213 ff. It is worth noting that K. Lake, in his Introduction to the facsimile reproduction of א, refutes Tischendorf's suggestion (here alluded to by Hort) that a portion of B and א were written by the same scribe.]

especially if supplemented by a study of the immense list of minute differences between B and א drawn up by Hoskier, [Codex B and its Allies, pt. ii. (Quaritch, 1914).] makes it difficult to accept von Soden's view that their common ancestor was at all recent. It is more probable that they represent, either two independent traditions of the oldest text of Alexandria, or recensions by two scholars each of whom based his text on the oldest MSS. obtainable in Egypt. top