Re: [tied] Re: Slavic accentology

From: mkapovic@...
Message: 38563
Date: 2005-06-13

> On Mon, 13 Jun 2005 01:41:08 +0200 (CEST), mkapovic@...
> wrote:
>
>>>>I think Kortlandt's explanation is basically correct: *vo`lja
>>>> (pre-Dybo)
>>>> >
>>>>*vo`l'l'a (*lj > *l'l') > *vo`l'a: (lengthening of final -a: in
>>>>compensation for CC > C, van Wijk)
>>>
>>> The problem I have with that is that Kortlandt's/van Wijk's
>>> explanation suggests a development CCV > CV:, without any
>>> indication that I have seen of parallel developments
>>> anywhere else. I'm not aware of any cases of such
>>> compensatory lengthening. Are there any?
>>
>>Well yes and no.
>
> I meant: are there any examples in any language of a
> geminate consonant (I'll also accept a consonant cluster)
> being simplified with compensatory lengthening of the
> _following_ vowel? I'm aware of thousands of soundlaws, but
> I don't recall ever running into this one. Compensatory
> lengthening of the preceeding vowel, yes. No effect on
> neighbouring vowels, of course. But lengthening of the next
> vowel? I find it very unlikely from a phonetic point of
> view, and it certainly doesn't seem to happen in Slavic
> cases like *attikos > otIcI or *dubno > dUno, *supnos >
> sUnU.

I cannot think of such a change in another language. But this is why I
take it to be so:
There *was* clearly a lengthening (long -a: in Old Polish, Polish dial.
and Slovincian)
All the endings were lengthened (cf. Old Polish), not only N. sg.
The lengthening obviously had something to do with *j.
The length of the ending produced a neo-circumflex on the root.

I think that most will agree on these. Now we just have to think of a way
to make the endings long (which obviously happened) via this *j. I also
thought of an answer like this: -ja was thought of a sort of
quasidiphthong and it was phonetically lengthened. This might even work in
N. sg. but it's hard to imagine a "quasidiphthong" in other cases.
So we need to make a way for this -ja etc. to lengthen. We could just say
"well in -Cja the -a gets lengthened" but that is also not very
satisfying.
Anyway, I don't think that the analogy with your other Slavic examples
apply. *attikos lost it's *tt long before (there were probably no CC in
BSl), and in other two cases we have a plosive and a sonant. One can
imagine that C + j would behave differently.

Mate