On 2007-12-01 20:37, Andrew Jarrette wrote:
> I know cybalist has dealt with this topic before (e.g. message
> #*44128*), but I have not found any satisfactory answer to the question,
> what is the origin of vrddhi in denominal derivatives in Sanskrit? How
> could e.g. /graiva-/ arise from /gri:va:/, or /sauma-/ from /soma/-?
> Piotr has mentioned "e-insertion" into the main syllable, but this would
not explain forms like /graiva-/, and does not explain why the process
> of "e-insertion" should arise in the first place.
> I hope that someone would care to respond to this question once again.
> I want to know how "legitimate" vrddhi in denominals is (as well as
> other forms with vrddhi).
The usual story is that in (late) PIE such derivatives were originally
formed from ablauting nominal stems. Whatever the exact mechanism, the
output was as follows: an accented thematic vowel was added and an *e
was inserted into the nil-grade of the root, not necessarily in the
original vowel slot. Thus, from *diw- (the weak alternant of *djeu-) we
get *deiw-ó-. Then the rule was extended to non-ablauting stems that
already had an *e, hence *swe:k^uró- 'belonging to the household of the
in-laws' (Skt. s'va:s'ura-, OHG swa:gur) from *swék^uro- 'father-in-law'
(this is similar to the case of <soma-> vs.
<sauma->: if //au// = o,
then //a+au// = au). Sanskrit just followed this path ever further,
making the generation of such "reinforced" strong grades an extremely
productive morphological device.