The Oxyrhynchus Papyri, Part 15

text with an English translation, Latin and English versions of the Armenian fragment, and the Greek text from Barlaam and Josaphat.

The question then presented itself, how far the Greek of Barlaam and Josaphat could be regarded as representing the ipsissima verba the very words of Aristides. That certain modifications had been introduced by the author of the romance was evident, e.g., a passage near the end in which the Christians were defended from certain charges made against them by early enemies was naturally discarded as out of date. But there remained considerable divergences which could not be easily accounted for. The Syriac has a number of repetitions and details not found in the Greek, the difference in total length approximating to the ratio of 3 to 2. Was this the result of expansion or compression? Had the Syriac translator amplified the original or the redactor of the Greek cut it down? The latter explanation, as Dr. Armitage Robinson observed in discussing this problem (op. cit. pp. 71 sqq.), seemed a priori from the outset the more probable, but careful consideration of the opening passage in which the testimony of the Armenian fragment was also available showed that the faults were by no means all on one side. While in the Greek there could here be traced one serious modification with a consequent displacement, one considerable abbreviation, and an added phrase in a Christological passage, the Syriac was found to be often loose and inaccurate, dropping some phrases and inserting others, sometimes with a distorting effect. Dr. Robinson’s general conclusion was 'that the Greek will, as a rule, give us the actual words of Aristides, except in the very few places in which modification was obviously needed. Where the Syriac presents us with matter which has no counterpart whatever in the Greek, we shall hesitate to pronounce that the Greek is defective, unless we are able to suggest a good reason for the omission, or to authenticate the Syriac from some external source.' Harnack agreed that the Greek was the truer witness, but proposed to account for the variations of the Syriac and Armenian by postulating as the basis of these a later Greek ‘revision’ (Gesch. der altchristlichen Litt. i. 1. 97)—a needlessly complicated hypothesis. Again, R. Raabe, in his commentary in Texte und Untersuchungen, ix. 1, has no high opinion of the accuracy of the Syriac translator. On the other hand, Dr. Rendel Harris in a recent essay seeks to show that Celsus, in replying to Aristides, used a text of the Apology which was in close agreement with the Syriac (Bulletin of the John Rylands Library, vi, pp. 163 sqq.).

With the welcome discovery of what is undoubtedly a fragment of the original text, the problem now reaches a new phase. The relation of the Greek of the fragment (P) to that of Barlaam and Josaphat (BJ) and to the Syriac version is discussed in detail in the notes below on ll. 8 sqq. and 26 sqq. In