Re: [tied] Rom. tsarca - Lit. s^árka

From: Piotr Gasiorowski
Message: 35334
Date: 2004-12-06

On 04-12-05 15:49, alexandru_mg3 wrote:

> Some more examples of Hungarian epenthetic preserved vowels:
>
> I think that the best Hungarian example showing the treatement of a
> Slavic methathetic form in Hungarian (indicated also by Piotr) is:
>
> 6. Germ. Karl < Sl. kralj 'king' (Methathetic form) < Hun. Király
>
> (see also: Rom. 'crai' also a Slavic loan showing the Methatetic
> Slavic form and indicating that: lj > j finished in Romanian only
> after this loan)

The normal treatment of CrV.../ClV... sequences in Hungarian loans form
Slavic is epentesis: CVrV.../CVlV... (no matter if in Slavic the initial
cluster was original or had developed as a result of liquid metathesis).
The second vowel may then be syncopated, leaving only CVr.../CVl...
provided that Hungarian syllabification rules allow it. This is what
happened in

szerda 'Wednesday' <-- *sreda/*sre^da < *serda

but also e.g. in

szalma 'straw' <-- *slama/*sloma < *solma
szolga 'servant' <-- *sluga

There is no syncope in words ending in a consonant, or if the syncope
would have led to the creation of a superheavy or otherwise
impermissible cluster, e.g.:

*borzda 'furrow' > *brozda/*brazda --> Hung. barazda
*kolsU 'ear of corn' > *klos/*klas --> Hung. kalász
*korljI 'king' > *krol'/*kral' --> Hung. király

(because *barzda, *kalsz and *kirly are not possible Hungarian words).

The syncope, even where permitted, was not 100% obligatory and some
words were not affected by it, or have retained dialectal variants with
the vowel preserved. Thus, we have

*boltInU > *blotIn/*blatIn > bolatin (AD 1055), modern Balaton
*sle^meN > szelemen 'horizontal beam, ridgepole'

On the other hand, you can find instances of syncope in loanwords that
had the shape *CVlV... in Slavic, e.g.

málna 'raspberry' <-- *malina

Anyway, <szarka> is the expected outcome of either *sroka or *s(v)raka-
in Hungarian. As for the simplification of *sv-, cf. szabad < *svoboda
and szent 'holy, saint' < *sveNtU.

Piotr

PS. You seem to be in favour of the reconstruction *k^orh2kah2 for
'magpie'. However, since it should be clear by now that Alb. sorrë has a
completely different etymology, and since Jens noticed here some time
ago that Skt. s'a:ri:ka: 'myna bird' must also have some other source
despite its superficial similarity, we are left with Balto-Slavic
*c'(w)árka: as the only evidence for *k^orh2kah2. Consequently, the IE
status of the latter is doubtful. You'd like to save the etymology by
adducing Rom. tzarkã, which you would like to be a substratal
"Proto-Albanian" word. Howevere, the absence of any Albanian cognates
renders your claim unprovable. As for Hungarian <szarka>, we have two
competing hypotheses:

(1) It is a loan from Slavic (reflexes of *s(v)órka, such as <soroka>,
<sroka>, <svraka> occur in all the Slavic languages with the meaning
'magpie').

(2) It is a loan from Romanian (where <tzarkã> isn't even the ordinary
word for 'magpie').

Given the distibution of the word and the uncertain status of <tzarkã>,
hypothesis 1 is by far the more likely one.

The ultimate origin of PBSl. c'(w)árka: is an independent question. In
the light of our previous discussions I consider a borrowing from
"Dacian" to be a serious hypothesis, one that provides the word with a
reasonable etymology (< *kWersnah2) and explains its form much better
than any other that I've seen. But even if its etymology is different,
Slavic *s(v)orka remains the most plausible source of Hung. szarka, as I
suppose nearly everyone on this list will agree.

Piotr