In the years 1897-99 considerable fragments were published in Oxford by A. E. Cowley and A. Neubauer, and in Cambridge by S. Scliechter and C. Taylor, of a re-translation made in the 10th or 11th cent., which, doubtless owing to its extreme badness, had been consigned to oblivion in an Egyptian Genizah. This was, indeed, mistaken for the original by the editors and for a time by some otliers, but that it is a re-translation is de- monstrated by all the tests that can be applied, and only a few arguments need be adduced here.
(1) It borrows from the Talmud, and not vice versa. In Bab. Erubin 5ia the following is quoted : ' My son, if thou hast, do good to thyself, for there is not in Sheol luxury, neither is there to death delay.' This comes originally from Sir 14", ' ChUd, according as thou hast, do good to thyself, and offerings to the Lord worthily bring'; v.'^, ' Remember that death will not delay, and a covenant of Hades has not been shown thee ' ; v.'^, ' Give and take and deceive thy soul, for there is not in Hades to seek luxui-y.' It is clear that the reminiscence in the Talmud is of yyUa. 16b. i2»_ fijg Egyptian document for v.^- gives the two clauses v."*'' and v.'^ in the order in which they appear in the Talmud : ' Remember that there is not in Sheol luxury, neitlier will death delay.' The clause which in the Greek is v.'^"" is here transferred to the place before v.'^. But when we come to v.'^ we find the same clause repeated : ' Give to a brother and give and indulge thyself; for there is not in Sheol to seek luxury.' The only explanation of this is that, when re- translating v.'^, the translator recollected the Talmudic citation and inserted it whole, without noticing that the clause about Sheol and luxury came later in the copy before him ; and when he came to v.'"' he translated it afresh. Practically the same thing is done by Wolfsohn, who inserts the Talmudic quotation as v.", but does not repeat the clause about Sheol and luxuiy in v.'* because the Syriac omits it.
(2) In numerous places where the Greek and Syriac versions differ slightly, yet quite clearly represent the same original, the Egyptian docu- ment has two texts, translating or mistranslating both the Greek and the Syriac. So in 30-" where the Greek has ' eunuch ' and the Syriac mhaimna, i.e. 'faithful,' but used ordinarily for 'eunuch,' the Egyptian document has two verses, one with 'eunuch and the other with ' faithful.' Similarly in 30", where the Greek has ' Better is death than a bitter life, and eternal rest than constant sick-
ness,' but the Syriac, ' Better is it to die than an evil life, and go down to Sheol than a sickness svhich is permanent,' the Egyptian document has two verses, one with ' eternal rest ' and the other with 'to go down to Sheol.' Since the (Jreek and the Syriac clearly represent the same original, somewhat differently rendered, it is obvious and certain that the Egyptian document is compiled from the Greek and the Syriac, not vice versa.
(3) The Egyptian document has numerous read- ings which are easily explicable as mistranslations of Syriac or Greek words, e.p. that already cited of mhaimna, 41'-'', 'wisdom' for unyi, which really means 'of injustice,' but would be certainly mis- rendered thus by one acquainted only with the Jewish Aramaic; 35'*, 'a judgment of song' for aiiyKpiixa. ixomiKuiv (!). Others are explicable by the medium of another language ; for it is not probable that the re-translator had access to the Greek directly. This language is identified with certainty as Persian written in the Arabic character by the mistranslation in 43- of ot-ayyiWoiv by ' pouring out heat.' This is obviously due to the Persian skhn, which means both ' speech ' (its Persian sense) and ' heat ' (its Arabic sense). Since the subject is the sun, one who did not remember Ps 19 might not unreasonably think that he poured forth heat rather than speech. Another certain mistranslation from Persian is in 43'^ 'lightning' for 'snow,' since in this language the words are all but indistinguish- able (barq and barf, distinguished by a dot). Be- sides containing mistranslations this document sometimes absolutely fails to understand the author, e.g. the 'decoy-partridge' of IP".
(4) Even if the document were not condemned hopelessly by internal evidence, the external evi- dence would condenm it. As has been seen, the nature of the references in the Oral Tradition makes it certain that the work had been lost before that tradition had been compiled. The only work in which the Egyptian document is quoted is a raedireval squib called Scfer ha-Galuy, composed in mockery of the Gaon Saadyah (t A.D. 941), though ascribed to him ; but even this work is rather against it than in its favour, since it classifies it with a notorious forgery, the Hebrew Hasmonsean Roll, and makes the ludicrous statement that the authors of these works provided them with points and accents — inventions of the 8th cent, at the earliest. The real Saadyah knows Ben-Sira only from the citations in the Talmud. In the pseudo- Josephus, a Hebrew work of the 9th cent., the latter is called Ttsn-Shirach, a form which must come from the Greek ; and in a chronicle of the 11th or 12th cent. (Neubauer, MeclicBval Jewish Chronicles, i. 167) his work is called Maqhil, which is a rendering of the Latin Ecclesiasticus. To- wards the end of the 10th cent, the author of the Arabic Fihrist mentions the work as in the hands of the Christians, but not of the Jews. When the Gaon Hai (t 1038) was asked to account for certain words of Ben-Sira being said in the Talmud to be in the Writings, his reply is ' they were written,' implying that they were so no longer (Teshuboth ha-Gaoniin, Lyck,'5624, p. 12).