Bears and why they mostly are called otherwise

[Ante Aikio]

[I wrote:]
>E.g. Finnish has karhu 'bear' < karhea 'coarse' (referring to fur)

[Xavier Delamarre:]

> As for Finnish _karhu_, I have proposed (_Historische Sprachforschung_ 105,
> 1992, pp 151-54) that it could be a loan from a Proto-Aryan form
> _*harkSas_, itself from IE _*h2rtk'os_, with hardening of the laryngeal
> (very much like _kana_ 'chicken' from Germanic _*xana_, or _teke_ 'to do'
> from IE _*dheH1-_) and  regular outcome of Uralic and Aryan -kS- into _h_,
> in Baltic Finnish (O.Ind. _makSa:_ 'fly' = Mordvin _mekS_ = Finn.
> _mehi(lainen)_ 'bee'). Mentioned by Szemerényi in his _Introduction to IE
> Ling._, 53.

I was not aware of this etymology. It sounds phonologically quite
well possible; the only thing one could object to is Aryan *-as > Finnic
*-u (why not Finnish karhas : karhaa- ?). But this is not a big problem, I
guess - e.g., back-formation of non-attested karhas seems possible.

PU *kS > Finnish h is regular, but I don't think it's necessary to
assume *kS here. Aryan *rkS > Pre-Finnish *rS > Finnish rh is equally
possible, since three-conconant clusters are a relatively late development
in Uralic.

> The etymology of _karhu_ by _karhea_ 'coarse' (J. Mägiste, _Estn. Etym.
> Wörtb._, followed by Itkonen & Alii, _Suomen Sanojen Alkuperä_) is
> probably a 'popular etymology'.

I'd still see it as a possibility. But the loan etymology sounds quite
well argumented.

> The existence in Finnish (and Uralic) of very archaic loans from various IE
> strata (late IE, Aryan, Baltic, Germanic etc), which are, so to say,
> 'fossilised' in the language, is in my view the strongest argument for a
> Eurasian Urheimat.

I consider them an argument against e.g. the Anatolian Urheimat of
IE. "Eurasian" is a wide concept - what do you mean, more precisely?

> I am presently dealing with Old Celtic lexicography where, as is well
> known, the word for 'bull' is _taruos_ (Gaulish _taruos_, O.Ir. _tarb_
> etc.) a metathesised form of IE _*tauros_, maybe on the analogy of _caruos_
> 'deer'. The funny thing is that we have a Finnish word _tarvas_ designating
> big cervidae, which shows the same metathesis as Celtic (whereas there is
> none in Aryan, Baltic & Germanic). But it may be pure coincidence or we
> have to postulate that the loan was made somewhere in present Russia,
> before (Proto-)Celts entered Central Europe.

The Finnish metathesis is independent of the Celtic one. Actually,
Pre-Finnish *wr > Finnic / Finnish rv seems to be a regular sound law -
there are no counterexamples, and several Baltic loan words have undergone
the same sound shift (e.g. Finnish torvi 'horn (instrument)', Finnish
karva 'hair (not on head)', cf. Lithuanian taure~, gau~ras). Finnish
tarvas is thus < *tawras < Baltic.

Regards,

Ante Aikio