From: Miguel Carrasquer
Message: 32658
Date: 2004-05-16
>The nom.sg. ay-(ám) has a different origin, although bothI should add that the accidental merger of sg. absolutive
>happen to have converged on the shape *ey.
>
>The pronoun *is is peculiar in having a stem *i- in the
>singular strong cases, *e- in the weak ones. My hypothesis
>is that the root was *i (giving *i-z, *i-m, *i-d) with
>oblique *i-a. The G, D/L, Ab/I were built on *iá- +
>endings, with accent shift as expected, and absorption of *i
>by a stressed thematic vowel, as elsewhere, giving *esyo,
>*e(sm)i, *e(sm)ot etc. The unextended oblique *ía (> *íya)
>regularly gave *éy (this is completely parallel to the
>i-stem vocative *-ey), and this is seen in Vedic *ay-ám,
>etc.
>
>The plural forms were *i-átu > *és(W) and oblique *i-áti >
>*éy.
>
>So singular *ey (*ía > *íya > *íy > *éy) and plural *ey
>(*iáti > *áti > *ác^ > *é3^ > *éy) are in origin very
>different.
>
>The unextended oblique (I like to call it "absolutive")
>leads a rather marginal existence in PIE. We find it in
>these rare nom.sg. pronouns in *-ey, and as the vocative in
>nominal stems. It also functions as the basis of the
>nominative plural, as is best seen in the i- and u-stems
>(nom.pl. -ey-es, -ew-es).