(2) As regards t/te non-accenting or the accenting of icords, much latitude must be conceded to the living language and to music, so that it would be very difficult to lay down strict and inviolable rules according to which this or that word is under certain circumst.inces to be non-accented or accented. In this way verse-members which appear to the eye very uneiiual may yet from the rhythmical point of \'iew be counted of equal value.t
(3) We have, moreover, no certain guarantee for the intention to carry through with perfect uni- formity the measure which in general rules in a poem. It is possible that it was considered legiti- mate to admit at times a line with four rises be- side one with three, and conversely to introduce a whole verse with a diilerent length of line, or finally to put a verse of three lines alongside of others with only two. On this whole subject cf. what W. Max Miiller (op. cit. p. 11) has established for Egyptian, and Zimmem (ZA xii. 382) for Baby- lonian poetry.
(4) In general, one receives the impression that in the older poems greater freedom rules than in the later oncs.t An unemngly regular parallelism, exact counting of the rises in verses of uniformly identical construction, all this is, nearly without
• For the former cJ. H. Zimmem, ZA viU. 121fl., x. Iff. ; for the latter W. Jlax Miiller, Vie fAi-bcupopsie d«r alien jfSgi/ptcr, 1899, p. 10 ff. Whether, in this elate of things, the actual relation of the falls to the rises can be reduced to suuunary formula) is another question. This \s'ill depend mainly upon the structure of the particular lan^iia^e. Tlius Ziinmern now (XA xii. 3S2(T.) thinks he can budd the Bab. poetic rhythm practically upon the foundation of the loniciui a mirwri. liut when the result is to obtain in all six dilTcrent feet odniist^ible in the same verse, when from one to three falls are possible bet^^'ecn two rises, when occasionally (cf. Schiipjting, iv. 4, p. 8310 two more falls are elided in accordance with an assumed licence, there is certainly enough of field-room, Zim- mem (p, 38;i) tells ua that Sievers has succeeded in 'provin(f' the existence In Heb. poetry of a pronounced * uniform rhythm,' Since his observations for Bab^vlonian are based upon work carried on in conmion with Sievers, and be several times emphasizes the a^jreement between it and Hebrew, the above remark aa to Ziminem's scheme will probably hohl good also of .Sievers' observations on Hebrew, with which the present writer has not yet made acquaintance.
t Uf. for instance in the Old Oenn, poem Udiand v,22 with T,& or v.*, or the two halves of v.89 or v,»ii' with one another.
t W. Max Miiller (op. <n<. p, 10) says rit'htly: *To me it is a very suHpicious circumstance that the .Sont; of Debondi and the latent Psalms sUll continue to be measured tn mxe and the aaine fashion.'
exception, the mark of later poems. The gap was, no doubt, lilled up by music, which always accom- panied poetry in early times, whereas in later times learned scansion with the pen in the hand and without regard to musical sound appears to have been the rule. But, on the other hand, one is entitled to make stricter demands on lyrical poetry in the narrowest sense, especially on dance- songs such as jierhaps meet us in Canticles, than on longer didactic poems like the Book of Job, which can hardly at any time have been sung.
(5) The more decided and sharply cut any par- ticular measure is, the more conlldently may this be used as a medium for restoring the text. Thus, for instance, one may undertake the work of textual criticism on the /rina/i-measure with surer results than in the case of an eveuly-llowing measure, because the peculiar limping form of the kinah must have demanded closer attention on the part of the poet. In any case, we should do well, in all textual criticism which deals with anything beyond superfluous expletives, to fissure ourselves of strong support on other grounds be- sides metrical, and not repose too much confidence in emendations based on metrical gro\inds alone.
(0) Finally, it must always be Kept stea<lily in view that the quality and the effect of poetry are still in by far the majority of instances secured for the texts by the parallelism, even where regularity in the measure is not carried out. Hence one must guard against assigning too great importance to metrical regularity.
f. Strophes. — We must deal more briefly with the use of strophes, i.e. larger formal units em- bracing several verses. The first to put forward a s])ecial stroi)he-theory was Fr. Koster in hia article, 'Die Strophen oder der Parallel ismus der Verse der heb. Poesie,' in SK, 1831, pp. 40-114. Ilis example was widely followed, and, long before the stricter verse-theories were put forward, the division of the OT poems into strophes of lengths more or less equal or artistically interchanging was prosecuted as nothing short of a pastime. The results correspond exactly to those described above (])p. 6 and 7") in the case of verse-theories. The variety of conclusions and the contradictions between them are perhaps even greater in this instance than in tliat. Here too in varying degrees may be seen mere strophic arrangement of the material received from tradition, alternat- ing with a re-shaping of the text based upon a settled theory ; great irregularity alternating with the strictest attention to rule ; simplicity in the form obtained alternating with the extreme of artificiality ; recognition of the jiarallel verse as the basis of the strophe alternating with accept- ance of the line as the fundamental unit, reach- ing even to the denying and destruction of the parallel verse, etc. At present, in addition to the before-named leading upholders of diilerent verso- theories, who also all put forward a special stroiihe- theory, the most ])rominent place is occupied by D. II. Miiller, with a most ingeniously worked- out stro]ihic system b.ised upon three fundamental priiKi]>les — the rcsputisio, the roncntcnatio, and the iricluaio.* In opposition to the line followed by him, a disposition at jncscnt prevails, following the lead of Bickoll, Duhm, and others, to rest content, wherever possible, with the simplest strophic framework, consisting of four lines, equal to two \'erse3 each of two parallel members.