At 3:26 pm +0100 17/11/98, Kristian Jensen wrote:
>>>I need a (di)graph for the nasalysed velar (/N/) of my TUNU
>>>language. This phoneme occurs only in the end of a word after a
>>>vowel. I write it "q" currently. Other graphs already used are :

>The graph 'g' alone could also work since Tunu does not have it in
>the first place. I don't see why you should bother with a digraph
>consisting of a particular graph that does not exist in the language
>in the first place when you can be using this particular graph
>alone. Some Polynesian and Melanesian orthographies do this. Perhaps
>the best known examples are Samoan and Fiji where /N/ is represented
>by 'g' in both languages.

Yes, indeed they do. And this, I agree, makes it appropriate for Tunu

> Fijian goes further to represent /mb/ by
>'b', /nd/ by 'd' and /Ng/ by 'q'.

Interesting - I've thought of giving {b}, {d} and {g} the values /mb/, /nd/
and /Ng/ in "briefscript"  :)

> though 'g' seems more appropriate because many Pacific
>languages do so in the first place and because it is also a
>Romanized graph originally representing a latin sound much closer to
>/N/ than 'q' ever did.

Yes, indeed - and in Latin, which first used the Roman script, {g} almost
certainly was pronounced as /N/ in words like 'agnus', 'dignus' etc.