On Sun, 22 Jun 2003 23:58:04 +0000, Richard Wordingham
<
richard.wordingham@...> wrote:
>I gather that 'ablaut now holds few
>secrets'. How do the various words
>for elbow, forearm, e.g. English
>'ell', Latin 'ulna', Greek
>'o:lene:', 'o:le:r' and 'o:llon' all
>relate?
I cannot find a Greek form o:le:r in IEW, EIEC or Boissacq.
The Greek forms given are <o:léne:>; <o:lé:n>, <-énos> and (Hes.) <o:llón>
(IEW <ô:llon>). Curiously, neither IEW, EIEC nor Boissacq make any attempt
to explain the long o:- in Greek (which also appears in Armenian ulu "back,
shoulder", and, as *o:l(e)kt-, in the Baltic words for "elbow").
Latin ulna is reconstructed as *olina:, the Germanic forms as *alina: (=
*alino:), except that Gothic has <ali:na> (= aleina). Irish uilen(n) is
from *oli(:)na:, with o > u because of i, and i > e because of -a:. Welsh
elin derives from *oli:na: ~ *ole:na:. EIEC gives the Tocharian forms A
a:leM (du.), B alyiye, both meaning "palm (of hand)". If I'm not mistaken
about Tocharian phonology, they do not contain *o but rather *a/*&. In
Albanian lërë, the initial vowel/schwa has been completely lost.
In summary, the forms seem to have:
*{o:|o|&}l{e:|e|i:|i}n-
Despite frequent incorporations into the a:-stems, Greek o:lé:n indicates
that the word was originally a consonant stem, animate (unless the form
with -r is real and there was also a neuter variant).
All forms can perhaps be derived from a paradigm:
N *h1éh3le:n(-s) > *ó:le:n
A *h1éh3le(:)n-m. > *ó:lenm.
G *h1&3lén-os > *&3lénos (*alénos, *olénos),
which accounts for almost everything except the *i(:) in the second
syllable.
In my speculative pre-PIE scheme, the above forms would have to be derived
from:
(1)
N **húx.lin-z
A **húx.lin-m
G **hux-lín-as
With svarita lengthening:
(2)
N **húx.li:n-z
A **húx.li:n-m
G **hux-lín-a:s
With labialization/palatalization:
(3)
N **húxW.li:n-z
A **húxW.li:n-m
G **huxW-lín^-a:s
With loss of **/i(:)/ and **/u(:)/:
(4)
N **héxW.le:n-z
A **héxW.le:n-m
G **hexW-lén^-o:s
After zero grade / Szemerény lengthening:
(5)
N *h1éh3.le:n
A *h1éh3.len-m
G *h1&3.léy-os (= Toch. B. alyiye ?)
Forms like *o(:)lin-a: can then be explained by early transfer into the
a:-stems (e.g. at stage (3) above: **húxW.li:n-z -> **húxW.li:n-áh2 >
*&3liná:).
=======================
Miguel Carrasquer Vidal
mcv@...