I have limited myself here to the most salient points of comparison.^70 The
sum of correspondences, not only in the structure of individual lines but also
in their relationship to one another and in the patterns in which they are
combined, is sufficient to show the persistence both in the Rigveda and in
Greek poetry of forms already established at any rate by the Graeco-Aryan
period.
We have, of course, early evidence for another branch of Graeco-Aryan,
namely Iranian. Zarathushtra’sGa ̄tha ̄s are in verse, and we should expect
their metres to show some relationship to the system reflected in Vedic and
Greek. And so they do, inasmuch as they are based on lines with a fixed
number of syllables and a caesura, arranged in strophes of three, four, or five
lines. On the other hand there is no regulation of quantities. In view of the
agreements between Vedic and Greek, we must suppose that such regulation
had existed at an earlier stage but lost its significance. Kuryłowicz has argued
that its role had been taken over by a stress accent on the penultimate syllable
of every word and of the verse.^71 A prosodic difference from Vedic and Greek
is that the meeting of a final with an initial vowel never results in syllabic
loss. This may be connected with the fact that, according to the oral tradition
as recorded in the Sasanian orthography, all final vowels had come to be
pronounced long, as was also the case in Old Persian.
The lines are of the following varieties: 7 | 9 (Y. 28–34, three-line strophes);
4 | 7 (Y. 43–50, five- and four-line strophes); 7 | 7 (Y. 51, three-line strophes),
and in a more complex strophe (Y. 53) 5 | 7 twice, followed by 7 | 7 | 5 twice.
They all involve a seven-syllable colon. The 4 | 7 and 5 | 7 combinations can be
directly compared with the Vedic eleven- and twelve-syllable lines, the four-
line strophe of 4 | 7 corresponding to the Indian tris
̇
t
̇ubh. If Kuryłowicz’s
stress theory is right, the original cadence... ∪−− || has been replaced by...
×× ́×||. In the Younger Avesta, as for example in the Hymn to Mithra (Yt. 10),
the commonest metre is an octosyllable, varied occasionally by lines of ten or
twelve syllables.
There have been several attempts to find metre in some of the Phrygian
inscriptions, especially the Neo-Phrygian epitaphs from the Roman period,
where a number of recurrent and probably traditional formulae appear.
Most recently I have pointed out that several of these appear to show metrical
patterns resembling the Greek glyconic and pherecratean, in some cases
(^70) For fuller surveys of Greek lyric cola and their interpretation in terms of the same system
see Watkins (1963), 195–210= (1994), 350–65; West (1973), 165–70; Gasparov (1996), 54–64.
(^71) J. Kuryłowicz, L’accentuation des langues indo-européennes (2nd edn., Paris 1958), 369–80;
id., BSL 67 (1972), 47–67; id. (1973), 102–38. On Gathic metre see also Christian Bartholomae,
Die Ga ̄θa ̄s und heiligen Gebete des altiranischen Volkes (Halle 1879), 1–19; J. Hertel, Beiträge
zur Metrik des Awestas und des R
̇
gvedas (Leipzig 1927); J. Gippert, Die Sprache 32 (1986),
257–75.
50 1. Poet and Poesy