Re: [lojban] gizmu

["Lionel Vidal" <nessus@free.fr>]

And:
> > > Assimilation is natural, yet language-particular.

Yes, but because some others phonetic specific facts ususally
hinder or modify the normal assimilation.

>Contrast russian
> > > _glasnost_ /glasnost/ with English /glaznost/.

Actually the assimilation is indeed a general rule in russian as well
but in a slightly modified form: all unvoiced before voiced, except
 i kratkaie, r, l, m, n and v, become voiced.
e.g.    to do [sdielat] is pronounced zdelat,
          but 3 [tri] is pronounce tri.
This rule explains the pronouncation of glasnost.

John:
> > The latter in Russian means "eyeness" (if it means anything).

[glaz] writen with a z is 'eye', and I am not sure 'glaznost' exists. But
[glasnost] with 's' means advertisement and by extension to make public.
The extension is more apparent with the french word 'publicité'.

(I'll stop here: it is way too painful it to write russian words with
 a latin-font-only-able MUA !)

> Indeed so. But synchronically, word-internal /sn/ is not impossible in
> native vocab.

It is indeed possible, but that will remain stable only if others
phonetic or morphological considerations stronger than
assimilation hold.
That being said, note that 's' is a bit special in indo-european languages
phonetic, because of its constrictive nature.  For instance,
most real 's' have simply disappeared before voiced consonnants
other that 'r' as soon as the transition to ancient latin.  And this
disappearing is explained by a voicing which weakened the consonant.

-- Lionel