Den 2017-07-29 kl. 16:38, skrev Raymond Brown:
> I can't pin down a date for General Chinese but, like BPJ, I get 
> the impression it was quite
> a while later.  It does seem a better system in many respects.  If 
> one is going to develop a
> Romanization that goes a long way to distinguishing homophones, 
> one really should look at
> General Chinese if only to avoid re-inventing the wheel.

It's probably hard to improve on by anyone who isn't equal or 
superior to YRC in knowledge of the historical phonology of the 
various 'dialects'.  I do think that he was too anxious about
"the principle of using short transcriptions for common sounds".
The spellings he chose for the dental sibilant series are rather 
misleading for foreign learners; _tz, ts, dz, s, z_ would IMHO be 
preferable: _z_ doesn't look like an affricate beside _ts, dz_ and 
_sz_ suggests a retroflex to those who know Polish and a voiceless 
sibilant to those who know Hungarian. Also I think _dhr_ would 
look better than _jr_, but that is a minor quibble on the part of 
a one-time Sanskritist.

>> The obvious model is Tibetan orthography, which still 
>> essentially depicts the sesquisyllabic[^1] language of over a 
>> millennium ago, which enables those 'dialects' which have 
>> converted complex initials and some finals into tones and those 
>> which preserve more or less of the initials and finals to use 
>> the same orthography -- at a high educational price to be sure.
>> /bpj
>> [^1]: 
>> <> 
> Thanks - an interesting article.  One is always learning new 
> things on this list, even
> at my advance years.   :)

I always wondered why some consonant sequences in Tibetan initals 
are written as stacks and some side by side. Sesquisyllabicity 
would explain that rather well, I think.