Re: [tied] Re: Various loose thoughts

----- Original Message -----
From: "Miguel Carrasquer" <mcv@...>
To: <cybalist@yahoogroups.com>
Sent: Sunday, January 16, 2005 11:15 PM
Subject: Re: [tied] Re: Various loose thoughts


>
> On Sun, 16 Jan 2005 20:11:54 +0100, Mate Kapovic
> <mkapovic@...> wrote:
>
>>From: "Sergejus Tarasovas" <s.tarasovas@...>
>>> --- In cybalist@yahoogroups.com, "Mate Kapovic" <mkapovic@...> wrote:
>>>
>>>> Hm, but the fact is that only -u- does *not* disappear from the
>>> last
>>>> syllable in Latvian. Cf. in o-stems Lith. -as : Latv. -s and in u-
>>> stems
>>>> Lith. -us : Latv. -us. Also, *-u is preserved in Auslaut as well in
>>> Latvian.
>>>> Thus, it couldn't have been *-mus in pre-Latvian.
>>>
>>>
>>> By that logic, modern Standard Lith. -ms can continue anything but *-
>>> mus, since -u- otherwise does *not* disappear from the last syllable
>>> in Lithuanian. One should not forget we are speaking of a
>>> *disyllabic* desinence, which well may have developed by slightly
>>> different (morpho)phonological rules.
>>>
>>> I'm not an expert in Latvian historical phonology, but I think
>>> Endzeli:ns had his reasons to not exclude *-mus.
>>
>>I still think you cannot compare (Modern) Lithuanian developments with
>>pre-Latvian ones. Do you have any example of the cases of unregular
>>behaviour of vowels in the last syllable in Latvian?
>
> Not me. But is the loss of -s in Dpl. and Ipl. regular?

Modern Latvian -m in D./I. pl. is really an original dual-ending.

>>In Latvian, the rule is
>>simple, short vowels are dropped in the last syllable, except *-u-. When
>>we
>>see that in Lith. -mus yield -ms and maybe earlier *-mas yielded -mus, we
>>have to assume a special development. But in case of Latvian, we do not.
>>Of
>>course, there is always a possibity of it I think that we have stick with
>>the simplest and regular development if we have no strong proof pointing
>>otherwise.
>
> But *-mas > -mus in Lith. may be simple, it's not regular.
> *-mo:ns > Lith. -mus, OP -mans _is_ regular. And so is
> *-mas > Latv. -ms, OP -mas. Therefore that is the optimal
> solution for Baltic (nominal innovative *-mo:ns, pronominal
> original *-mas), even if it introduces an irregularity at
> the Proto-Baltic level. And in view of Slavic a:-stem Gsg.
> *-a~s > *-a:ns, after Acc.pl. (> Nom.pl) *-a:ns, it's an
> irregularity I can live with.

Well, *-mas > -mus in Lithuanian is more "regular" than the supposed *-mas >
*-mo:ns in pre-Lith. I see no compelling evidence to push the OP innovation
such as -mans in D. pl. so far back into time. Lith. -mus, Latvian -ms and
OP -mas can be derived without any problem from Baltic *-mas and this from
PIE *-mos/bhos. OP -mans is probably just a later innovation. If we can
derive Lith. -mus from *-mas there is no need to posit *-mo:ns for it as
well. Besides, it would be a little bit strange if Lith. had the same
innovative form as OP but different from Latvian. Of course, that is not
impossible, but I would rather not reconstruct it in such a way if it is not
absolutely necessary.

Mate