Re: [tied] trzymac'

On Sri, svibanj 31, 2006 1:40 am, Miguel Carrasquer reče:

> On Tue, 30 May 2006 10:58:15 +0200 (CEST), Mate Kapović
> <mkapovic@...> wrote:
>
>>On Uto, svibanj 30, 2006 12:39 am, Miguel Carrasquer reče:
>>
>>> I don't think I can accept Holzer's chronology, in any case
>>> not where it relates to absolute dates. I don't believe
>>> that Slavic <kórljI> has anything to do with Charles Martel
>>> or even with Charlemagne (Karl would have given *kórlU).
>>
>>That is hardly certain. The borrowings are not always subject to strict
>>rules. Also, Slavic had /l/ and /l^/ in that period
>
> Which period? That is the question. Slavic didn't have a
> phonemic contrast between /l/ and /l^/ until the loss of the
> yers. Whether and when there was a phonetic contrast is
> anybody's guess. The word *karljI has /lj/.

Why do you suppose there was no *lj before the loss of the years? There
are sequences *lj in Slavic, which are exactly *lj and not *lIj. Cf. for
instance suffixes with *-j- etc.

>>, and it is a common
>>phenomenon that languages with a "hard" and "soft" l's borrow foreign
>>"hard" l as their "soft" l, thus in Turkish and Albanian for example.
>>There are also other examples when Romance /l/ > Slavic /l^/, as I seem
>> to
>>remember from Holzer's articles.
>
> In his Zagreb article I find one example (Pola > Pűlj), and
> several with l = l. A quick scan doesn't reveal any with -l
>> -o, but that may be coincidental.

There is also a toponym Poljud from something like palu:des, as I remember
it from the top of my head.

>>Anyway, his theory is not at all relying on this word only. It is
>>established by numerous attestations of Slavic words in old documents
>>(German, Latin, Greek etc.), other loanwords into and from Slavic,
>>toponyms etc.
>
> The Charles Martel thing is one of the few links with
> absolute chronology. Latin loans in Croatian are another
> one,

Are you talking about Holzer's IWoBA article? Coz I am not. I'm talking
about his general theory on Proto- and Common Slavic (cf. a number of his
articles in Wiener slavistisches Jahrbuch), which is very well
substantiated and does not depened just on a couple of words. The reason I
mentioned Holzer's view was just to explain my use of Proto- and Common
Slavic as terms. Holzer's IWoBA article is another thing...

>but in view of the data presented in Zagreb by Ranko
> Matasovic', I have my doubts about Holzer's conclusions
> regarding the dating of Dybo's law.

Please not that his Proto- and Common Slavic theory do not depend on that.
Originally, he did not even deal with the accents. His IWoBA article was
actually the first time that he tried to involve accentuation into his
relative chronology.

> I don't have Matasovic''s paper handy right now, but as I
> recall the gist of it was that Latin feminines were adopted
> into a.p. a, while masculines/neuters were adopted into a.p.
> b.

I don't have it here either, but yes, it was something like that.

> Since Latin feminines are never stressed on the -a, a.p. c
> is out of the question, and so is a.p. b if Dybo's law had
> already worked. The fact that Latin feminines became a.p. a
> in Croatian, in spite of their not having anything like
> acute accentuation in the source language, shows that Dybo's
> law predates the influx of Latin words.

Hardly. Romance length could have just been percepted as rising and thus
interpreted as the old acute in Slavic. That's very common in language
contact. For instance, in Croatian the German accent is perceived as
rising, so almost all the German words get rising accents.

Anyway, I don't think it such a good idea to put to much weight to
loanwords. As a native speaker of a tonal language I see very well what
happens with new loanwords (for instance English) in Croatian. And I can
tell you - it's a mess, a general chaos. One can hardly pick up much logic
there. So I am very sceptical on what we can find out from the word loaned
almost 1500 years ago...
Just to illustrate the picture, in my idiolect bo``ks (< English box) is
a. p. a, dz^e``m (< jam "marmalade") is a. p. b and ba^r (< bar) is a. p.
c. All relatively recent English loanwords (20th century)...

> Why masculines/neuters became a.p. b is an interesting
> question. The implication is that not only Dybo's law, but
> also Ivs^ic''s law (retraction c.q. advancement of stress
> from weak yers) predates the contact with the Romans. After
> Ivs^ic''s law, all three a.p.'s supply a more or less
> correct placing of the ictus for Latin masculines in the
> nominative/accusative singular. Apparently, the prosody then
> prevails, and the neo-acute is considered to be closer to
> the Latin model than either the acute of a.p. a or the
> circumflex c.q. stresslessness of a.p. c. Even if that
> means un-Latin stress in the oblique forms (the presence of
> limited mobility in some Latin categories such as rátio,
> ratióne may have helped alleviate that).

Currently, I think that all the a. p. b loanwords were loaned pre-Dybo. I
cannot see another reason for stem-stressed Latin masculines to get a
final accent in Slavic (although I wouldn't say it is absolutely
impossible).

> There are other reasons, one of them outlined above. In
> general, I feel a bit uncomfortable with the implication
> that almost nothing happened between Proto-Balto-Slavic (2nd
> millennium BC?) and 600 AD, and almost everything happened
> between 600 and, say, 1200 AD.

Have you read Holzer's articles? Because he's not just making it up. For
instance, if you look at early Slavic loanwords in Greek, there are
toponyms like Karouta /karu:ta/ ~ Slavic *koryto and Gardiki ~ Slavic
*gordIcI. Get the picture? Slavic *did* indeed change a lot in that
period, that is quite clear. I would urge you not to dismiss Holzer's
claims so easily.

>>> I have a chronological problem with the fact that old Polish
>>> still had uncontracted forms in the XIV ~ XV centuries, at a
>>> time when the accent was already fixed on the initial in
>>> _all_ words. If so, that must mean that the accent
>>> retraction and the contraction are unrelated phenomena, at
>>> least in Lekhitic.
>>
>>Old Polish has uncontracted endings in *pytajes^I?
>
> As I said, my Polish historical grammar seems to imply that,
> without giving much details. I didn't find *pytaje or
> anything similar in the Kazania S'wie,tokrzyskie.

Strange... Anyway, I don't believe it changes much. If it were true (and I
seriously suspect it is, maybe I'm wrong), that would only mean that the
contraction did not occur so far up North. In Czech and South Slavic, it
could have worked easily.

Mate