On 2006-05-14 00:46, Richard Wordingham wrote:
> C. (I may be wrong on this one.) Why does _duhita:_ scan as
> long-short-long?
It does. Similarly in <pr.tHiví:> from *pl.t&2wíh2, as well as a number
of other examples (e.g. <rátHa-> with a heavy initial syllable).
The normative pronunciation of the text of the Rigveda was carefully
codified precisely in order to save it from "corruption", i.e. from
being affected by the normal course of changes typical of Old and Middle
Indic. It must have been a very real threat despite the best efforts of
the reciters and their spectacular success in preserving the text. The
discrepancies between the traditionally established sam.hita:pa:t.Ha
(continuously recited text) and the reconstructed metrical text, minor
as they are, show that _some_ changes (especially those caused by the
loss of the vestigial laryngeals) had already begun to distort the metre
beyond easy repair.
Piotr