Greek Genitive Plural

[David L. White]

        The only reason I know of to think that /s/ once existed in the gen
plural of /a/-stems in Greek is that late PIE apparently did not tolerate
hiatus to the extent that /a-oon/ would mandate, at least not in V-stem
nouns.  Note that even in Sanskrit V-stems, where intrusive /s/ was not used
(not counting some pronouns), it was still evidently felt necessary to begin
the final syl with some consonant, in this case /n/.  It seems that
different branches adopted different solutions to the same perceived
problem.  So we have some indirect evidence of a phonotactic constraint.  I
have not bothered to check the rest of posited PIE morphology to see if this
constraint holds up across the board, though right off the top of my head I
cannot think of any evidence that it does not.

Dr. David L. White