> In Russian, <y> (bI) certainly sounds diphthongal (especially after
> labial consonant: my-, by-, vy-). On the other hand, Russian may be
> the only Slavic language where *Uj (in the nom.sg.masc. adjective *-
U
> + *-jI) did not merge with *y < **u:, since we have the
pronunciation
> -oj when stressed (bol'shój) and a colloquial variant -@... when
> unstressed, in rivalry with "official" -yj/-ij (russkij = [rusk@...]
or
> [rusk^ij]).
>
>
So, if I got you right, you neither support nor reject diphthongial
*y1.
Must admit your note on *Uj puzzled me.
As is commonly known, the merge *Uj with *y is not a quantum of a
process, it's based on another phonetic phenomenon - different
reflexes of 'tense' *U and 'tense' *I (i.e., *U and *I before *j) and
normal *U and *I in a number of Slavic languages. In most Slavic
dialects tense *U was reflected as /y/, in most Russian dialects -
as /o/ when stressed, # (null) or /o/ when unstressed (i.e. merged
with reflex of normal *U). Then, under the influence of the Church
Slavonic orthography, an official norm with [I] established itself
for unstressed positions after velars (as if it was a normal
development like ky>ki etc), and offical and informal norms merged
after other sounds. Thus we have <russkij> ['rusk&J](informal,
from /russkoj/):['rusk^IJ](formal, as if from /russkyj/>/russkij/) <
*rusIskUjI, <malyj> ['mal&J](here official norm, which is as if
from /malyj/, and informal, which is from /maloj/, merged) and
<rodnoj> [r&d'noj] < *ordInUjI (no Church Slavonic influence in
stressed positions - it would be too unnatural).
By the way,<bol's^oj> [b&l^'s^oj] doesn't fit here as <*boljIs^IjI,
with *Ij not *Uj.
Sergei