>>I still maintain that *pelh2-
(including Slavic *polje) is a different root, though if *polh1- _had_ survived
in Slavic, a partial semantic overlap of homonymous roots would not be
surprising. I wonder if Polish opole < *o(b)-pol-ije (an old administrative
unit, grouping several villages) might not derive from hypothetical **pol(i)-
'town, fort' rather than *polje 'field. Semantically, it would fit very nicely,
as *ob- means "around".
Cf. Lith. apygarda "district, circuit"
(apy- "around" + gard-) - if it is not neologism of Simonas Daukantas
(1793-1864) - the author, who liked to create new words.
>>The chief difficulty is that Slavic
**pol(i)- is conjectural: it does not seem to be attested in placenames (other
than Opole [= German Oppeln]) and the pan-Slavic term for "town" is
*gordU.
Lith. gardas means "pen, fence, enclosure;
stall". This word failed to replace the word <pilis> in the Baltic
languages. Nevertheless we can trace such tendencies in the Curonian and
Samogitian toponymy. First of all there are three Lith. oikonyms with the root
gard- -- all in West Lithuania:
Gardamas - town in S'ilute.
district;
Gardeliai - village in Kelme.
district;
Gardis'ke. - village in S'ilale.
district.
In the documents of the last quarter of the
16th century much more placenames of this type is mentioned in Samogitia (I
mention them in Ruthenian spelling, as they appear in the
documents):
Gardelis - land in Liauda manor, Veliuona
distr.;
Gardel'kalnas - hill (on which are
"graves") in Padubisys manor, Kraz'iai distr.;
Gardpiavnia - soil in Paju:ris
distr.;
Gardupia - rivulet in Vidukle. distr., near
Little Pale.kiai manor;
Gardy (also Gordy) - manor in Kraz'iai
distr., on the rivulet Krakles;
Gardyny - field in Bielevic'i manor,
Raseiniai distr.;
Gardys'ki - village in Jaunodava manor,
Kars'uva distr.;
Gardelovki - land in Paju:ris distr.,
Garilauki manor;
Gorda - river in Gru:ste. distr., Gordy
manor; in Seda manor, Tels'iai distr.; in Veliuona distr., Vol'montovic'i and
Birs'tiny fields;
Gordeny - manor in Kraz'iai distr., near
the river Dubysa; village in Padubisys manor;
Gordina - river in Vies've.nai distr.,
Kus'lutis'ki manor.;
Gordino - in S'arnele. manor, Vies've.nai
distr.
Gordova and Gorduva - river in Veliuona
distr., Liauda manor or "field";
Gordovo - manor in Gru:ste. distr., on the
river Varduva;
Gordy - manor in Vies've.nai
distr,;
Gordy - manor in Gru:ste. distr., on the
river Gorda;
Gordy - meadow in Seda manor, Tels'iai
distr.
(From: Sprogis I. Geograficheskiy slovar'
drevney Zhomoitskoi zemli XVI stoletiya. Vil'na, 1888, p. 66, 82.)
The earliest evidence comes from the Ceklis
land of South Curonia (now Tels'iai, Plunge., Skuodas and Maz.eikiai districts
of Lithuania) - the castle Garde is mentioned together with 43 other castles of
Ceklis in the document of 1253 (the term "castle" is not used, but all place
names may be identified with certain hillforts). Castle of Garde = manor of
Gordy (in Vies've.nai distr.) = modern town Z'emaic'iu Kalvarija (*Gardai
renamed to Kalvarija in 1639).
Though some of the place names, mentioned
above may be connected with the modern meaning of the word "gardas", some of
them, and particularly the last one (Garde of 1253) are obviously applied to a
castle.
On the other hand we have such place name
as Pylene (Pilene) in the same list of Ceklis castles (1253). In the Duvzare
land of Curonia the castle Empilten is mentioned in the same document of 1253
(=modern I:piltis/Impiltis, Kretinga distr., Lithuania). Cf. Piltene - main
residence of the Bishop of Curonia (though it was not mentioned in
1253).
Thus we can see, that Baltic root pil-
could be replaced by gard- in the meaning of "castle", but pilis survived.
I think a similar process, which was more "successful", erased the memory of
Slavic *pol(i)- "town, castle". The process might be connected with appearance
of a new type of fortifications; perhaps this happened in times, when the Slavs
still lived in a comparatively small area.
Polish <opole>, as you noted, may
bear traces of the extinct pole "castle". "Pole" (Lith. "laukas") as a kind
of territorial organisation is also worth attention.
See: Wojtkowiak Z. "Pole" jako jednostka
osadniczo-terytorialna na Litwie w XVI w. (na przykladzie powiatu
wilkomirskiego) // Zeszyty naukowe uniwersytetu im. Adama Mickiewicza w
Poznaniu. Historia. - Poznan, 1971. - Zesz. 11, p. 157-173.
In this article you'll find a brief review
and references on "pole" structure in the other countries (opole is treated
as a kind of such structure).