Re: [tied] [j] v. [i]

On Wed, 4 Jun 2003, Miguel Carrasquer wrote:

> >Does it not look like analogy from the "soft" type, just as in many
> >Slavic languages?
>
> I'm not sure what Slavic forms you have in mind. The Lithuanian soft type
> (so-called third declension) adjectives have their plural in -iai (e.g.
> auksìniai). But isn't that just begging the question? How did *-j-oí
> become *-j-oi~?

Ah, come on, spoilsport, this did get us a bit of the way. Now the vowel
of -ai~ and the order of the elements of the diphthong are right. I'm
thinking of South Slavic languages where the nom.-acc. of the fem.
a-declension end in -e (z^ene) from -jeN, only with subtraction of the
-j-. -- If there is a synchronic rule that retained -ai is always
circumflex, then that gets us the rest of the way. I would like to
understand such a rule, however, but even if I don't I have no right to
deny its existence if it works.

Jens