Re: Tone Romanization: Opinions Sought

John Cowan wrote:

> Philip Newton scripsit:
>> This wouldn't work quite as pictographically in languages such as
>> Cantonese or Sheli which have multiple tones with the same contour
>> (e.g. level) but at different pitches, but I think that marking above
>> the vowel may be the best.

> The To`ngwa" Lo`ma'ji romanization of Cantonese
> (http://tongwalomaji.onlineblast.com) uses a", a^, ah, a`, a', and a
> (or rather their diacritic equivalents) for the six tones of Cantonese
> (high level, high rising, mid level, falling, low rising, and low level),
> and uses digraphs to write things like rounded vowels (Cantonese has 8
> vowel nuclei).  This makes the system completely representable within the
> constraints of Latin-1, except for the syllabic nasals m and ng, which
> have to be represented with trailing punctuation instead of diacritics.

And the problem with the syllabic nasals could be circumvented by placing
a vowel letter (something neutral looking) after the _m_ or _ng_. This
extra vowel can then act as a carrier for the tone-diacritic. This is
what I've been doing in Tiemish(*), using -e, and it looks fine.

This mention of To`ngwa" Lo`ma'ji has made me think about implementing it
for Tiemish. I've always liked the look of tone-spelling, and this at least
has -h, and it's not as ubiquitous as in the Yale system (2 diacritics and
the lack of one encode level, rising and falling, and pretense or absence of
-h indicates whether we're talking about the higher three tones or the lower
three. But I find the result has too many _h_s). The only problem I find then
is that it reminds me of my mystery coda, also denoted _h_. It's a mystery
since I can't decide what it represents. I can use -h in the absence to any
other tone diacritic for "level tone", and -h with a dicritic to mean "the
coda h", and then "-hh" for the combination of the two...
Hmmmm... (or should I say Hmoob?)

In anycase, the diacritics of To`ngwa" Lo`ma'ji look nice (I've always had
a soft spot for the circumflex), they let me use Latin-1, and they remind
me that I'm not so fond of the macron, tilde and caron I'd been using before.
The only pain now is converting pages.

Amusingly, a glance at a transformed version of the Tiemish page reveals to
me my orthographic bias: there are no numeral words, except for "sep" (now
"sehp"), which is "twenty" in mid-level tone, which was the unmarked tone!

* Tiemish is at http://ataltane.net/conlangs/chineselikelang.html

s.
--
Work is of two kinds: first, altering the position of      Stephen Mulraney
matter at or near the earth's surface relative to       http://ataltane.net
other matter; second, telling other people            ataltane@ataltane.net
to do so.         --  Bertrand Russell     http://livejournal.com/~ataltane