comments from G Burgess on phonology paper

[Logical Language Group <lojbab@access.digex.net>]

Gary had only one major comment.  In the first section of sounds, you sometimes
list several IPA values for individual Lojban letters.  He thinks that one
of these should in each case be designated the nominal value, with the others
being "acceptable alternatives".  He noted that you half did this in the
back section, where for most letters you have said something like "this is
the preferred sound for Lojban "c", etc.  But the front table does NOT show
that there are preferences.  Furthermore, even in the back section, you do
NOT express a preference, the given sound is just "one of the eacceptable
forms of r".  he thinks we should choose one in particular as the norm
(or perhaps two if we need a different one for syllabic 'r'.  Among the
Fairfax Lojbanists, it is safe to say that our preference is for a trilled
'r' (as found in Russian of course %^, but actually this was first suggested
way back by Jack Waugh,who referred to German)  However it is selected, a 
"rolled/trilled r of some kind is less likely to be heard as some other sound
especially by those languages that have r/l confusion.

lojbab