[tied] Re: Morphology (2/20)

--- In cybalist@..., Miguel Carrasquer <mcv@...> wrote:

> Irregular depalatalization of *-tI to *-tU is a possibilty, but I

would prefer a

> solution that also explains the loss of the ending altogether,

instead of

> introducing that as a second irregularity. What if Proto-Slavic

had both

> primary *-ti and secondary (conjunctive?) *-t in the present, à la

Italo-Celtic.

> That would have given *-tI, and *-0 (after vowel) or *-t (after

consonant, at

> least after -s-). Then regularly *[jes]t > *[jes]tU, and also the

zero-ending

> after vowel is optionally extended with the pronoun *tU. As a

result, we have

> the attested Slavic situation: -tI (Russ. <jest'>, O.Pol.

<jes'c'>), -tU (OCS

> <jestU>, Pol. <jest>) and zero (Ukr., Bul., Cze. etc. <je>) in

apparently random

> distribution.

May be the following fact about Krivichian will be of some interest
to you. According to Zalizniak, in Krivichian -0 is predominant, but -
tI occurs in some cases as well (despite what I wrote in the previous
posting for the sake of simplicity). The distribution seemes to be as
follows:
1. -0 is predominant in conditional clauses, also in adverbial,
purposive and complement subordinate clauses, -0 is not predominant
in other types of sentences.
2. -tI is predominant before pronominal enclitics
(_s'a_,_t'a_,_m'a_,_si_,_ti_,_mi_) and enclitic particle _ti_.
If that distribution reflects something really old, it could indeed
be about subjunctive or other PIE verbal category.

Sergei