Re: [tied] Finnish hevonen "horse"

On Tue, 30 Oct 2001 09:45:03 -0800 (PST), Knut Holt
<aquila_grande@...> wrote:

>There are some striking correspondances between the
>phonenemics of many core elements in uralic and
>indoeuropean. You can easyly see them by comparing
>proto-indo-european elements with the corresponding
>finnish (finnish is so conservative that such a
>comparison can be done meaningfully).
>
>
> IE Ur
>I/me me mi-nu
>you te-we si-nu<ti-nu
>this/that/it so/to se/tuo/tä-
>ablative case suffix es/ed ta
>accusative suffixe sg m m
>Nom pl suffixe s t
>1. p sg suffixe m n<m
>2. p sg suffixe s/tha t
>1. p pl suffixe mo/me me
>2. p pl suffixe te te

This is all true enough.

>In both IE an and FU many elements have variants where
>one element contains an -s- and the other a -t-. Where
>IU has s/t, also FU often have s/s^/t.
>
>The abowe examples illustrate very well this
>correspondance.
>
>In both the groups s/s^ tend to change into h in some
>instances.

In Indo-European s- becomes h- in Greek, Iranian and Brythonic (any
other?). The only group that FU may have been in contact with is
Iranian.

>Therefore the word hevo-nen may very well be genetic
>related to the greek word hippo-, but it could also be
>a early loanworld from IU.

There is no evidence for an s- in PIE *(h1)ek^wos "horse" (Gamqrelidze
and Ivanov make an attempt at *s^ek^wos, with the *s^ they also
reconstruct for *s^okw- "eye" -> Hitt. sakuwa "eyes"; PIE *okw- "eye",
this based on Semitic (Akkadian <sisû>, Hebrew <sûs>, etc.), Hurrian
(<essi>, <issiia->), Egyptian (<ssm.t>) and assorted Caucasian (Geo.
<ac^ua>, Axvax <ic^wa>, Abkhaz <(a)c^y>) parallels, but even if this
were so, it would only show up as s- in Anatolian (and the Anatolian
words we know seem to be based on Indo-Iranian *aswa). Certainly
Iranian has no forms with s- or h- (Avest. <aspa->, OPers. <asa->,
Sogdian <asp->, Wakhan <yas^>, Oss. <jäfs>), and Greek h- (hardly a
source for a Finno-Ugric borrowing) is much more likely to derive from
*y- (*yikkwos) than from *s-. Even if G & I were right, and (pre-)PIE
had *s^ek^wos, then it would still be problematical to relate that to
Finn. *s^epo-, because of *k^w ~ *p (the PIE form does _not_ have a
labiovelar *kW, but a cluster *k^ + *w, which suggests earlier **kVw).