Theiling Online    Sitemap    Conlang Mailing List HQ   

Re: USAGE: Adapting non-Latin scripts

From:Tristan Alexander McLeay <conlang@...>
Date:Wednesday, May 24, 2006, 23:21
On 25/05/06, John Vertical <johnvertical@...> wrote:
> >As long as the rules were regular with how to spell sounds, it > >doesn't matter if we spelt /æ/ as "a", "ä", "æ" or "ae" and /ei/ as > >"ai", "aa" or "é". If we just picked one and stuck to it, we'd be > >laughing. The Finns are (hopefully not at us ;), and they have more > >vowels than we do. > > >Tristan. > > Digress: > We actually tend to view most of ours as sequences of multiple vowels - only > the 8 monofthongs are seen as independant phonemes. A syllabe-final glide > interpretation works too, but then at least an additional schwa must be > positioned.
Umm... So does that mean a word like /kAt/ you'd consider to rhyme with a word like /pA:t/? (I also thought there's a set of diphthongs /ie/, /y2/, /uo/, which I can't see how you'd interpret them as having syllable-final glides.)
> From *our* POV, an analysis of English as /i i: e ei & A A: Ai Au o o: oi ou > u u: @/ (7 basic vowels + combinations) of course works just fine, but you > might protest...
I'd have nothing against something like that in principle. Of course, the precise symbols you've chosen aren't the ones I do (frex I consider "/Au/" to be "/&O/"), but I'm perfectly happy with length. (I vote for eight basic vowels + length & diphthongs, though two of those vowels are only ever long.) In any case, there are some vowels which no matter if you consider vowel length phonemic or not certainly have a status as bimoraic. It makes plenty of sense to analyse them as long vowels so even if you consider quality the sole distinguishing factor between a vowel /i/ and one /I/, spelling them as "sii" and "sit" still makes sense. (Makes it easier to use the orthography if you say "single vowels can never end a word, except -a" (for schwa).) [Mark J. Reed wonders:
> Well, you can of course use whatever phonemic symbols you want, but > assuming even an approximate phonetic connection, I can't imagine what > phonemic distinction you are capturing via /A/ vs /A:/.
[I assume John was referring to the vowels commonly transcribed as /V/ as in "come" vs /A:/ "calm". That is a perfect length distinction in Australia (albeit with a low central vowel), and I could easily see how a Finn learning English would use it even for American or British sounds.]
> BTW, I dout regularization would suffice to solve all problems of English. > Even with the exceptions aside, there's just too many rules - like the rule > of final <-ost> being "long".
No-no, that's not what I meant. I was suggesting we the Latin alphabet could be used to phonemically spell the English language if we completely redid the orthography. So "host" wouldn't be spelt "host", it'd be spelt "hoast" or "houst" or "hoost" or something (depending on the exact principles of the orthography), same as "toast" is spelt "toast" or "toust" or "toost". We could start with what we've got but radically reduce the ruleset so we've got a few dozen digraphs (and trigraphs) to spell all the vowels and consonants but almost no positional variants. It mightn't look much like English, but that was never my point. -- Tristan.

Reply

Mark J. Reed <markjreed@...>