Re: *-t-, put

--- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:

>
> On 2007-07-09 00:09, tgpedersen wrote:
> > --- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@> wrote:
>
> >> What's wrong with the orthodox account of the relationship:
> >>
> >> *frawarðiða- ðeðe: > *frawarðiðe: (Goth. frawardida) by haplology
> >> *frawarðiða- ðe:ðun > *frawarðiðe:ðun (Goth. frawardide:dun)
> >>
> >> *wurxta- ðeðe: > *wurxte: (Goth. waúrhta)
> >> *wurxta- ðe:ðun > *wurxte:dun (Goth. waúrhte:dun)
> >>
> >> (where *ðeðe:/*ðe:ðun is the old imperfect of *dHeh1, and the
> >> original meaning of the univerbated phrase was 'made V-ed'
> >
> > The haplololology,
>
> Just try saying "frawarðiða ðeðe:".

I tried and now I'm more convinced than ever that that not even
ProtoGermani would like such a preterite.

Haplology is more or less to be expected in such cases, cf. Lat.
diligenter < *diligentiter.

The 'such cases' are the ones you conjectured.

> > and the fact that in Germanic (and in the lesser Iranian
> > languages, by some odd coincidence) the weak preterite and ppp
> > stems are related;
>
> Given that many of the weak verbs were causatives, a periphrastic
> past tense with an auxiliry verb expressing causativity (e.g. 'he
> made it destroyed') is hardly odd.

I didn't get that. Please spell out.

> > a loss of two syllables should leave a bigger mark on the preterite.

> ??? Please explain. The loss of offending syllables may have been
> gradual -- one by one rather than two at a time. OE has -don for
> Goth -de:dun, which shows that the process didn't stop in PGmc. but
> was continued in the daughter languages.

Second weak conjugation: Gothic salbo:-da < PIE *salbax-dho:; where
-ax is the individuating suffix. Look, Ma, no *-tó-!

> > I think it's
> > *frawarði- ðða: > *frawarðiðe: (Goth. frawardida)
> > *frawarði- ðe:ðun > *frawarðiðe:ðun (Goth. frawardide:dun)
> >
> > *wurx- ðða: > *wurxte: (Goth. waúrhta)
> > *wurx- ðe:ðun > *wurxte:dun (Goth. waúrhte:dun)
> >
> > Note Sabellic 3sg prúfatted "probavit" with -tt-.
>
> Why not a "Iuppiter rule" treatment of *-a:t- > -att-?

Buck has no Iuppiter rule for Oscan afa I could tell and calls its
perfect -tt- 'a mystery'.

> > -i- is a prop vowel
> > also found in Latin -id- 'participles'.
>
> > PIE perf.middle 3sg *-to is from (somehow!) *-dho < *-dhdho.
>
> Curiouser and curiouser ;-)

I'm just hiding behind the broad back of Jens who is a proper linguist
and who wants -t-, -(e)n-, -(e)nt- and -r to be phonetic variants of a
single morpheme. The truth (ie the ancestor) must lie somewhere in
between and -d- is a good place to start, and if one claims /d/ is
actually -Nd- as I do that's an even better place to start. There's
something phonetically wrong with a paradigm that contains -r, -n- and
-nt-, like that of "liver". The -t- must be from -d-, one way or another.


Torsten