Re: Perfect, Latin 1st conj, Tocharian pret. I

> BTW, it seems strange to me that eg Italian in the perfect has
> parlai
> parlasti
> parlò
>
> where -i and -sti show preserved final vowel but no trace of Latin
> -it in the 3sg (and similarly in Spanish). If late Latin had *-au/-o
> (au -> o) must be Latin, it can't be Italian or Spanish), it must
> have had a parallel history of official -avit. Cf Toch B pret I, and
> Skt 3sg (and 1sg)) dadhau etc. Now if the -ax stative suffix could
> be voiced to *-aG (-> *-aw) we would have a solution for a putative
> 'bad Latin' 3sg perf *-aw and Toch B pret I -a. That also fits in
> with the idea that 3sg pret was the bare stem, except when the stem
> was the root itself, in which case, in order to have some formal
> mark of its status, it was replaced with the bare 'subitive' stem in
> -s, from which later the s-aorist developed.

To be understood so that the roots (possibly the same) of dadháu,
dadáu, tastháu must have ended in a velar which was lost> Torsten


here's another solution:
-ásti tu -> -ásti tu
but
-áwi(t) il(le) -> -áw il -> (in Latin) -ó il

so that the -i in 3sg perf left over after elision of -t merges with
the demonstrative and thus disappears, which can't happen in 2sg.

This can be interpreted either way: that 3sg perf always had a t-less
variant and that the t-form came about through French-style sandhi
(a-t-il), or that the t-less from was generated from the standard one
rather late.


Torsten