W. B. Henning, BSOAS, (1950)
The study of the Pahlavi poetry, so spiritedly initiated by M. Benveniste twenty years ago, seems to have come to a dead end. That certain Pahlavi texts, as the Ayadgar-i Zareran or the Draxt-i Asurig (the Dispute of the date-palm with the "goat), are poems, is conceded on all sides; but the formal problems, the problems of rhythm, metre, and rhyme, remain in the dark. It seems doubtful whether the material at hand is capable of leading us to definite conclusions.
There are two main obstacles. Firstly, the notorious sloppiness of the copyists leaves too much room for conjecture; the mere addition or omission, at the editors' discretion, of the word for "and" and the harf-i idafet is sufficient to disturb the rhythmical balance. Secondly, as a rule we do not know the dates of composition, and therefore cannot tell how the words were pronounced by the authors; it makes a considerable difference to the metre (whatever it was) whether we put down padak or paig, mazdayasn or mazdesn, rosn or rosan, adak or aig, sikanj or skanj, giyan or gyan, yazat or yazd, awis or os, druyist or drist or drust or durust, haeadar or azer.
One thing is clear: a biased approach will not lead to convincing results. On the strength of the preconceived notion, carried forward from the study of the Avesta (where matters are equally dubious), that the metre is a purely syllabic one, the Pahlavi poems were made to suffer a great deal of emendation; where the usual procedure of omitting inconvenient words produced lines too short to fit into the scheme, either words were added or their pronunciation distorted." The alternative theory, namely that the metre is accentual, seems to offer better prospects. It relieves us of the necessity of changing the texts overmuch; the number of syllables to a line can be left as variable as it is; and the precise pronunciation, rosn or rosan, becomes a matter almost of indifference.
Clear evidence in favour of the accentual verse can be found in the very text that formed the starting-point of M. Benveniste's investigations, the Draxt-i Asurig. The whole of this poem, which is less encumbered with glosses than most other Pahlavi texts, is written in fairly long lines, of twelve syllables on an average, with a caesura in the middle. There is a recurring formula, which fills the first half of lines, x. az man kareul "they make x. out of me". The first word can be one of one, two, or three syllables, so that the first half of a line can have five, six, or seven syllables. Does this not indicate that the metrical value of a word is wholly independent of its number of syllables. The second halves of the lines are not in anyway affected by the greater or lesser length of the first halves.
Please view the attached PDF file to read more!