The fact that Ben-Sira had before him no Hebrew or national literature which has not been preserved is of great interest ; and, as has been seen, with his grandson the biblical books were classified as they are still. In the book itself certain other names appear. Thus a portion of the projiliecies is called the Remonstrances (48'"), of which we recognize the original in ninDin.T similarly used in Jerus. Peak, ii. 4. Enigmas and Parables is the title taken by Ps 78, to which there is a reference in .W. The Bible as a whole is identi- fied with Wisdom in 24^'' '-, and v.** implies that it had already undergone several generations of ex- pounders. The attitude of Jewish writers to their Bible has so often been dictated by that of their neighbours to their own sacred books that we may be justilied in finding here the traces of the Hellenic estimate of the Homeric poems, to which the Greek translator makes a veiled allusion in the phrase 'narrating epics in writing' (44^'').
3. Poetical form. — To any one who compares the lines of l!en-Sira with the corresponding passages of the OT it is apparent that the latter liave boon subjected to Procrustean treatment ; thus Gn 17*, ' a father of many nations have I made thee,' becomes in Sir 44", ' Abraham was a great father of many people ' ; but for Gn 22'*, ' in thy seed shall all the nations of the earth be blessed,' Sir 45-"' substitutes, ' that nations should be blessed in his seed.' Sometimes the order of the plirases is in- verted ; so Ec 3'^ 'nothing can be added to it nor anything taken from it,' becomes in Sir 18°, 'there may be nothing taken from them, neither may anything be added to them.' Sometimes the verse reproduced undergoes so much inversion and padding that the sense is seriously injured, e.g. Job 8'^ in Sir 40". Since in the first passages cited the author has altered a Divine et.yniolog;^' by the introduction of a monosyllable 3t and seriously reduced a Divine promise by the omission of another mono.syllalile 73, it is evident that single syllables are of imiiort- ance to him, i.e. that his Procrustean metliods are
due to his em])loymcnt of a syllabic metre to which he accommodates the biblical material. That he should do this is very natural, since, as has been seen, he displays considerable aciiuaintance with Greek literature ; and from the nearly contemporary Puitiiilus of Plautus we find that the kindred Plicenician dialect was being accoinmodated to Greek syllabic metres. The metrical scheme is supplied by the correct re-translatiim of any two or three o\ tlie lines, and, where they are taken directly from the OT, this is easy ; and this scheme is a trimeter of the rhythm culled in Greek and Latin Bacchic, in Arabic and Persian miUaqarih, of which the basis is a foot of the form y /_ o. In Persian this rhythm is very popular, the whole of the great classic SJiah-namch\is\ni^ composed in it; the Hebrew variety (except in the substitution of three feet as the linounit for four) resembles the Latin variety used by Plautus, e.g. 'multils res simitu in nie6 corde v6r.so.' Where the lines do not correspond with this scheme, there is some fault either in the tradition or in the re-translation. Thus 27", 0c777T;<rts tucf/ioCs 5cA iravTbi aotfiloLj when re-tr,anslated is one syllable short : but the Latin version which oli'crs ' sicut sol ' indicates that no3n is corrupt for nonj, which gives the ninth syllable requireti, and furnishes a correct antithesis to the changeableness which in clause 6 is compared to that of the moon. Where the lines contain lists, the fact that they are padded in order to obtain a metrical scheme is sometimes very obvious. So in the list 40', divaroi aal alfia xai fpis Kal })Oiuj>ala, eTToryur^al, \lijlI>^, Kal avvrpiixiia Kal tidffTt(, Fritzsche ejected ^iro7u7a(, 'utpote explicationis causa ad- sutum.' It seems unnecessary for the sense, but the two syllables which it represents (nns) are very necessary for the metre.
The re-translation, if ever satisfactorilj' accom- plished, will be of importance for the study of Hebrew grammar, which at present liejiends on a tradition codified some 1000 years later. For it will be found that, when the consonants are re- stored, the metre settles the vocalization (to a certain extent) as in 33", iyKali'i<Toi' a-qnela Kal aWol- waov 6aviM(ri.a, mNSs: nat mn.x vin, where the metre and the sense both require that .isb"! should be rea<l ushnch, not w'shattneh ('and repeat,' not 'and alter').
i. Language. — The language employed by the author was from the nature of the case mainly that of the OT, of which his book is so largely a metrical cento ; but here and there the traces of a later development of Hebrew, such as we find in the Mishna, can be discovered ; and indeed the fragments preserved by the Oral Tradition exhibit a considerable amount of this. No confidence can indeed be placed in the accuracy of these ; it is, however, of some interest that the transmitters of that tr.-idition thought of his language as Uabhinic. One interesting technicality, niaSn, ' rules of con- duct,' wliicli clearly underlies iropeioi in 1*' occurs in an obelized passage ; but comparison of Greek and Syriac seems to reveal Tin.i for 'to beg' in 40-*'', and ]V'2 for 'sbamefast' in 41"; while in 37"" the sentence rendered ' there is a friend, which is only a friend in name' meant 'which is really a friend,' the usage which is here hidden being that of the later Hebrew, where ' to be named ' means 'to be in reality' (e.g. Bab. Gittin, 47a). The use of late or Aramaic words seems at times to have been dictated by metrical reasons ; so in 8'° •isu can be restored with certainty for 'coals,' and it would seem that this word was employed because '^ni c^mtained a syllable too m.any. The Greek word d7W7<is m.ay have been employed in 48'"', but this seems to be isolated.