Re: [tied] PIE 'brow'

On 04-11-09 14:41, whetex_lewx wrote:

> Why *h3kW-bHr-u- is expressed as h3 but not as h2? In Lithuanian h3
> mutates as -uo-, but this reconstructed h3kw- is related to akis
> (eye) (short -a- vowel).

That's as it should be. There's no length if the laryngeal is
prevocalic: *h3o- > (short) *o- > PBS *a- > Baltic a-, Slavic o- .

> Latin acies - of the eye?

Unrelated. This one is from *h2ak^-(ro-) 'sharp, pointed', hence,
figuratively, 'piercing, acute, keen'.

> Sanscrit ak-; Latvian acis, Armenian aku
> and etc... Couldn't be eye in PIE not h3kw-, but akw- (short a
> sound, which became o in Latin

There's no such change in Latin.

> and relicts of short -a- are visible
> only in secondary Latin forms?

But Gk. osse/otte (*hokW-i- + -e), omma (< *opma < *hokW-mn.) and o:ps
(< *ho:kW-s) point unambiguously to *o; so do Toch. ak/ek (< PToch. ëk-
< *hokW-); and of course so does Lat. oculus. Armenian akn (not "aku")
is a difficult word, but the explanation of its shape is an
inner-Armenian problem; the other branches that show the full grade
don't offer decisive evidence either way, since *a and *o have merged
there. To sum up, *h3okW- is the only sensible reconstruction.

> *h3bHruh-; bHruh- may be related to Lithuanian bru:ks^nys (line,
> German strich), brez^ti (to line, German zeichnen, ritzen). Also
> Russian brovka (brov` - eye brush) related to Lithuanian briauna -
> edge.

This is a heterogeneous collection of lookalike words, not a cognate set.

Piotr