I should first like to confess that I'm still stuck in the old heresy of an Odin/Ariovistus invasion from the east (mix of Alans/As and Vani(r)) giving rise to the present spread of the Germanic languages, which isn't fazed by the below information, but I'd like to know if anybody else has an explanation for this bizarre coincidence?
Gamkrelidze and Ivanov
Indoeuropean and the Indo-Europeans, p. 881 / 777
'Proto-Kartvelian prior to its breakup must be placed, on the evidence of archaic lexical and toponymic data, in the mountainous regions of the western and central part of the Little Caucasus (the Transcaucasian foothills). The first wave of Kartvelian migrations to the west and northwest, in the direction of the Colchidian plains, must have begun with one of the western dialects in the third millennium B.C. and led to the formation of Svan, which spread to the western Transcaucasus and was superimposed on local languages, probably of the Northwest Caucasian type, which thus became substratal to Svan. Svan was gradually displaced to the north, to the Great Caucasus range, by the next wave of migrations, which occurred approximately nine centuries later (on glottochronological evidence) and removed the westernmost remaining dialect as far as the Black Sea coast. This western dialect gave rise to the later Colchidian - or Zan, or Mingrelian-Laz -language, one of the languages of ancient Colchis.'
Klimov,
Etymological Dictionary of the Kartvelian Languages
Preface, p. IX
'Using the methods of glottochronology as introduced by Morris Swadesh, the absolute time of separate development can be fixed for these languages at approximately 2,600 and 4,200 years, respectively. In other words, the Georgian-Zan complex should have begun to disintegrate at the beginning of the first millennium B.C., while the differentiation of Georgian-Zan and Svan should date from a period not later than the last centuries of the third millennium B.C.'
[Zan = Megrelian-Laz}
p.326
'GZ *xarg- 'heap, pile (of stones, wood)':
Georg. xerg- 'heap, pile (of stones, wood)';
Megr. xorg- 'heap, pile of stones';
Laz xorg- 'pile'.
Unattested in Old Georgian. The vocalic shape of the Zan form presupposes the former *a (the origin of the parallel Megrelian variant xurg-, which can also mean 'threshold', remains unexplained: possibly it is based on the well-known functional value of the Kartvelian alternation a - u: cf. *bard-). Its Georgian correspondence can be derived from the same protoform taking into account the influence of the ancient umlaut factor. The lexeme underlies the respective verb stems; cf. the action nouns Georg. xergva and Megr. xorgua- 'to heap up with stones, wood'. Comparison with Iranian material proposed in Tedeevi (1988: 38-39) is difficult to accept.
...
Tedeevi, Olg.a 1988. Nark.vebebi osur-kartuli enobrivi urtiertobidan
(Studies of Ossetic-Georgian language interrelations).
Tbilisi: Mecniereba.'
...
It's a fair guess that Tedeevi is including Ossetic material. As for Ossetic connections with Northern Europe, cf.
http://en.wikipedia.org/wiki/Alans
(which goes for the 'sea' word too)
...
de Vries,
Altnordisches etymologisches Wörterbuch
'ho,rg-r m. 'steinhaufe, opferstätte, stein-altar',
nisl. hörgur, nnorw. horg f. 'haufe, schar', dial. 'bergkuppe',
aschw. hargher 'opferstätte', hargh in ON., nschw. dial. harg, horg, horv 'stein-haufe'. >
ne. dial. hurrock 'haufe loser steine' (Thorson 66); >
finn. harju 'hügel, klippe' (Karsten ANF 22, 1906, 191; sehr fraglich, s. Setälä FUF 13, 1913, 363).
ae. hearg 'heidnisches heiligtum', nnl. Hargen ON.,
ahd. harug, haruh 'heiliger stein, opferstätte', harahus 'steinkreis um grab'. Vielleicht zu schw. dial. har 'steiniger boden', mnd. mnl haar, hare 'anhöhe' in ON. (Ekwall MASO 3, 1941, 41).
Man hat verglichen
1. air. carn 'steinhaufe' (WP I, 30),
2. lat. carcer 'umfriedigung, einschluss, schranke' (Meringer WS 10, 1927, 188), und dann weiter zu der unter h a r f r behandelten wurzel *ker (J. Trier, Holz 1952, 77); also auszugehen von 'heiliger hain', oder von 'zaun um einen heiligen ort' (vgl. h r i n g r).
3. ai. s´arkara: 'gries, kies', gr. króke:) 'steiniger strand'.
Der u-stamm könnte für den sakralen gebrauch des wortes zeugen (Specht KZ 64, 1937, 12)'
...
Pokorny
'3. kar-, redupl. karkar- u. dgl. ,hart'.
...
Wegen der Existenz eines weitverbreiteten iber.-mediterr. *carra ,Stein, Steinhaufe' (v. Wartburg FEW. s. v.) ist die Zugehörigkeit folgender Worte zweifelhaft, die teils hierher oder zu (s)ker- ,schneiden' gehören, teils nichtidg. Herkunft sein können:
air. carrac f. (nir. carraig) ,Felsen, Klippe' und crec f. d.s., Gen, craice, creice, mir. auch Nom. craic (nir. craig) f. ds., Gen. creca, stammen alle aus dem Brit., ebenso mir. crach ,rauh' (= schorfig); zu acymr. carrecc, cymr. carreg f. ,Stein, Felsen' (*karrika:), bret. karreg, com. carrek ds., vgl. cymr. carrog ,Bach' (*karra:ka:), eig. ,Steinbach' = mir. carrach ,schorfig' und gallo-rom. *cracos ,Stein' (v. Wartburg FEW. s. v.), acymr. creik, cymr. craig f. (*kraki:) ,Felsen', bret. krag ,Sandstein', cymr. crach ,Schorf = bret. (Vannes) krah ,kleiner Hügel, Spitze' (*krak-, *krakk-, vielleicht durch Metathese aus *kar-k-) usw. Grundformen sind voridg. *karr- oder idg. *skr.-s- (ergäbe kelt. carr-), bzw. *kar-k-;
ebenso zweideutig
ital.-venezian. (ven.) scaranto, caranto ,steiniger Bergbach' und
ven. ON Scarantia > Scharnitz (Tirol), Carantania ,Kärnten', usw.;
desgleichen germ. *har(u)gaz ,Steinhaufen, Opferstätte' in
anord. ho,rgr ,Steinhaufen',
ags. hearg m. ,heidnischer Tempel',
ahd. harg ,Hain, Tempel'
(finn. Lw. karko ,Holzstoß, Stapel', harkko ,Klumpen, Haufen'),
vielleicht auch zu
air. cymr. bret. carn ,Steinhügel, Steingrab', und
ahd. hart ,Bergwald', ags. harað, -eð ds.;
kaum hierher als ,Waldbewohner' der altgerm. VN Harudes, ags. Hæreðas, aisl. Ho,rðar;
eher zu air. caur, cur ,Held' (*karut-s).'
Any connection with 'sharp'?
Why is the Germanic match with Georgian-Zan so much better than anything de Vries and Pokorny can come up with?
(Nothing similar in UEW, BTW, so it's not an 'areal' thing)
Is the out-of-place a/u 'ablaut' in GZ relevant?
Torsten