GENGO KENKYU (Journal of the Linguistic Society of Japan)

E. Benveniste is the first to take up the problem of versification in Aßyatkari Zareran. He argues, JA. t. 220, pp.245-293, to the effect that the whole Memoir can be restored to its metrical form: six-syllabic verse, of which five, occasionally four or six, consititute a stanza marked fairly often either withassonance or with right rhyme.
The present writer, however, would draw attention to the fact that in ancientIran an octosyllabic verse was of the type specially favoured in epic literature.It may naturally be supposed that this type of verse should have maintained itspopularity throughout centuries down to the mediaeval times. Was it not exactlythe case with our Aßyatkar? It is throughout the whole text-not exceptingthe initial and the final sections-that text-critical as well as metrical investigationsare likely to reveal eight-syllabic verses, as mentioned above, frequentlycharacterized by either assonance or right rime. Cf. pp.3, 6, 9, 10 where thepresent writer demonstrates how to reconstruct such octosyllabic verses. Thewhole text thus established in its early Middle Persian metrical form, while itabounds in Parthian elements, shows no more than seven instances of the hexametre stanza.

View full abstract

Show abstractHide abstract

The monograph is reviewed with reference to some related articles publishedmore recently by the same author. The notion of grammar as a sentence-generatingdevice is briefly reviewed and its significance in the study of language isemphasized. It is pointed out that the generative grammar with the generalizedtransformation as its part can explain not only the recurrent formal propertieswithin a sentence, but also those outside a sentence. Thus, for example, thechoice of pronouns in reference to the nouns that have appeared in a precedingsentence can be handled formally in terms of a potential applicability of a transformationthat combines the pertinent sentences into one. It is remarked thatthe generalized transformation should be regarded as a grammatical level that isdistinct from the singulary transformation, in view of the substantial differencein the form of rules, the generating capacities, and in particular its recursiveness.
It is also pointed out that the present form of the phrase structure grammarmay not be the most desirable, once it is considered as a part of the transformationalgrammar. Presumably, more formal restrictions and / or sub-structuresshould be introduced into the level. A possible modification may be dividingthe phrase structure into two sub-levels, first a possibly recursive context-freegrammar, and second a level for selecting lexical items in the form of a branchingchoice structure that represent the category analysis. The selectional restrictionsmay be imposed on the choise of categories in the succeeding stage of the‘screening transformation’ that deletes all ineligible strings. It is concluded thatdetails in the grammatical structure including the above-mentioned points have to be determined by further studies of various existing languages.

View full abstract

Download PDF (521K)

Register with J-STAGE for free!