Re: [tied] Some accentological thoughts...

On Sat, 3 Mar 2007 12:34:48 +0100 (CET), Mate Kapović
<mkapovic@...> wrote:

>On Sub, ožujak 3, 2007 11:46 am, Miguel Carrasquer Vidal reče:
>> On Fri, 2 Mar 2007 22:45:18 +0100 (CET), Mate Kapović
>> <mkapovic@...> wrote:
>>
>>>On Čet, ožujak 1, 2007 11:07 pm, Miguel Carrasquer Vidal reče:
>>>
>>>> The MAS-approach also fails for verbal roots, as is clear in
>>>> Balto-Slavic itself: be^gnoNti is a.p. a, be^z^e^ti is a.p.
>>>> c. So what is the root *be^g- (*bhegW-): dominant or
>>>> recessive?
>>>
>>>It's recessive ((-)acute) but *-noN- "immobilizes" the (-)acute, whereas
>>>*-ě- does not.
>>
>> Then what does that make *-noN-? It's not simply (+) or
>> (-). And why not an analysis making the root (+), and the
>> suffix -ě- "mobilizing" it?
>
>Because Latvian has be^gt

Well, the infinitive suffix was originally stressed, so
Latvian verbal roots with an acute vowel always have broken
tone in the infinitive, except when the ictus was drawn back
by Hirt's law, which if course doesn't apply here (the acute
is due to Winter's law, not to a laryngeal).

But upon reflection, I'd say verbal roots are always (-)
[usually recessive accent in Greek, usually unstressed in
Vedic]. The verbal suffix, if any, determines where the
stress goes (or the prefix, in the case of the Grk/Ved.
augment).

=======================
Miguel Carrasquer Vidal
miguelc@...