From: Edgard Bikelis
Message: 61256
Date: 2008-11-02
On Sun, Nov 2, 2008 at 5:11 PM, Andrew Jarrette <anjarrette@...> wrote:--- In cybalist@yahoogroups.com, "bmscotttg" <BMScott@...> wrote:
A bit aside: I've always wondered how English has [wu] as in <womb>
and <woo> ([wum] and [wu]). Is the [w] in this position
extra-rounded, to distinguish it from the almost identical vowel that
follows? It doesn't become [B] or something similar, yet the [w] is
clearly audible before the [u]. We also have [ji] in <year>, <yeast>,
<yield>, <yean>, <ye>. The [j] doesn't become fricativized or
otherwise hardened (cf. Spanish), but nevertheless it is clearly
audible. I know that Mandarin has similar combinations (although in
some dialects /w/ is actually pronounced [v] or the approximant
variety of [v]) -- yet PIE did not seem to allow similar combinations.
Should [w] be considered an allophone of /u/ in PIE, as Arnaud says
it is in French and other Romance languages? (Apparently, as is well
known, this is what the original writers of Latin thought of their /u/
and [w], but it later caused problems in words like <equus> for
earlier *<equos> or <<uulgus> for earlier *<uolgos>).) Or is it
already established doctrine that /u/ and /w/ are allophones of a
single sound in PIE?
Andrew
How could the i- and u- declension work otherwise? Like skt. nom. sing. pát-i-h, pl. pát-ay-ah, or nom. pl. tráy-ah, loc. triSú. And the conjugation... *H1ei-: skt. subjunctive áy-a-ti, imperative i-hí, or *dreu 'to run', dráv-a-ti, part. dru-tá-.
Edgard.