Re: [tied] Syllabic liquids in Slavic

From: Piotr Gasiorowski
Message: 9546
Date: 2001-09-17

----- Original Message -----
From: Miguel Carrasquer Vidal
Sent: Monday, September 17, 2001 12:35 AM
Subject: Re: [tied] Syllabic liquids in Slavic

> I don't see how Czech <srdce> and <c^ervený> could have easily influenced <serce> (<sirzce>) and <czerwony> (<czyrzwiony>), given their rather different phonetical structure.  Surely, <obywatel> may very well be a Czech loan. [I see that Ban'kowski in fact gives it as a "bohemizm", which he of course does not for <serce> or <bez>].
No great problem here. Orthographic <d> was mute in <srdce>; early Polish <rz> [r^] changed into [r] before apical consonants including [ts] (cf. starzec 'old man' : starca []), so the normal reflex of *sIrdIce was <sirce> or <sierce> (the latter dominant after the 13th c.). Polish had no syllabic rhotic, so the most natural adaptation of Czech [srtse] in view of native [s'ertse] was [sertse]. Again, <czyr(z)wiony>, <czer(z)wiony> (with <rz> still standing for a fricative trill) were not very different from <c^ervený>. In effect, *<czerweny> + <czer(z)wiony> --> <czerwony>. As for <bez>, Urban'czyk ("Historia przyimków <bez> i <przez>", 1946) shows the role of translations from Czech in the reintroduction of <bez> in Polish after 1400 (older <biez> was extinct by that time).

> I'm interested.  I would like to point out that in <serce> (< *k^r.d-) [see above] the palatalization has been lost both on the /s/ and the /r/: one might have had <sirzce> ~ <sierzce>
OK, just give me a little time. As I said above, <rz> > <r> is completely regular before <c> (marzec : w marcu, korzec : w korcu, etc.)
>In such cases the surviving yers yielded /e/ (*U) or palatality + /e/ (*I). But closed-syllable developments before liquids are _older_ than the change *& > e ~ zero in open syllables, and need not parallel it. /a/ for *& < *U (or dispalatalised *I) appears only before *r. So what? 

> The development of /r./ in Polish is not what would be expected for /Ur/.  It's possible to suppose an ad-hoc development /Ur/ > /er/? > /ar/, ...
I'd propose *Ur > *&r > ar, *Ir > *j&r > *ir/*'er, where *& represents the phonemically merged qualities of *I and *U. Why "ad hoc"? And how do you know what other development would be "expected" for _preconsonantal_ *Ur (where *U and *r are not separated by a syllable boundary)? These changes are regular in this particular context. You could with equal justice call the proposed developments of syllabic *r "ad hoc".
> ... so I retract my "impossible", but it's certainly not a necessary development.  To paraphrase you: "Of course in [..] individual branches syllabic consonants came to be pronounced asynchronically as [...]*-VR- or *-RV- combinations, but even so divergent *V values show that this development is post-PSlav. E.g. syllabic *r happens to be reflected as such in Czech, Slovenian, Serbo-Croatian and Slovenian, but otherwise gives ORussian Ir/Ur, OCS rI/rU, Bulg. r&/&r, Polish ar/ier, Polabian or/ar, Sorbian ar/or/er, Slovenian dial. &r, ar, ro, etc. It's far less trouble to derive all these reflexes directly from syllabic *r than from anything like *&r."  Even, I might add, if we have to derive them from two distinct flavours of syllabic *r.
Rather than reply to this separately, I'll try to cover these questions in my promised essay on the reflexes of "sylalbic liquids" in Polish.