Re: [tied] trzymac'

On Pon, svibanj 29, 2006 9:44 pm, Sergejus Tarasovas reče:

> --- In cybalist@yahoogroups.com, Mate Kapović <mkapovic@...> wrote:
>
>> >Indeed, in that case
>> > Slovincian <tr^îma,> 'I hold' (if <*tri:mň, with -ajo, dialectally
>> > contracted already in Common Slavic?) vs. <tr^ĺ~mac> 'to hold'
> (if <
>> > *trima"ti) would demonstrate shortening before a stressed inner
>> > syllable (*trima"ti) and non-shortening before a stressed final
>> > syllable (*tri:mň,).
>>
>> Actually, the shortening is in the infinitive, but in present
> tense, the
>> length is neo-acute in origin. The 1st person final accent would
> yield
>> exactly the same result in Slovincian since the accent is retracted
> (also
>> a long vowel).
>
> My aim has been more than modest so far -- to read Stang correctly.
> If he didn't mean what I wrote (shortening in <tr^ĺ~mac> because -i-
> goes before a stressed inner syllable in *trima"ti and non-shortening
> in <tr^îma,> because -i- occurs before a stresses non-inner syllable
> in *tri:mň,), what did he mean then?
>
> And then I don't quite understand your comment (no tongue in cheek --
> your formulation is too succinct).
> What do you mean by neo-acute length exactly?

The length of the neo-acute (*~) = the length that has its origin in the
long neo-acute accent. Common Slavic *~ is reflected as length in all West
Slavic languages.

<Definitely not the

> lengthening as a result of retraction from weak jers (not relevant in
> that context), so you must have meant the retraction by Stang-
> Ivs^ic''s Law.

Exactly.

>But does that retraction lengthen short vowels
> aquiring the ictus? I'm not sure it does.

It doesn't. Only long vowels get the long neo-acute (*~), the short ones
get the short neo-acute (*`). But Stang-Ivs^ic''s Law operates before the
two-mora-rule (thus before the shortening of the pretonic syllables when
there are more than two moras after them).

>What it does, actually, is
> that it imparts rising tone to a vowel which was already long before
> the retraction, so it's not the retraction one needs to have length --
> rather ictus on the long (non-acute) vowel before Dybo's Law.

Yes, but the length would be shortened by the 2-mora-rule if there was no
retraction. Thus, the length is preserved only because of the retraction.

>> > Stang didn't know Dybo's Law and considered the place of ictus in
> b-
>> > verbs original, while from contemporary point of view one would
>> > probably expect non-shortening in both cases (*trí:mati >
> *tri:ma"ti
>> > in the same way as *trí:mo, > *tri:mň,),
>>
>> Nope. It is clear that there is a shortening in the first case. I
> have
>> written about it at length. If you are interested, I can send you my
>> article.
>
> Thanks, it would be very helpful, since so far I've been able to
> judge your paper only by Kortlandt's critical article.

Sent.
You can notice that my theory is corroborated by the development in
Polish. Old Polish has, for instance, the expected seNdzic' "judge" -
saNdzisz (a. p. b), which has been transformed analogically to Modern
Polish saNdzic' - saNdzisz. In order to explain that, Kortlandt assumes
some imaginary suffix *-Ij-, which has dissapeared and who knows what.
That is the clear example that his theory just does not work and his
critique is futile.

Mate