Path: utzoo!utgpu!watserv1!watmath!att!att!pacbell.com!mips!twg.com!david From: david@twg.com (David S. Herron) Newsgroups: comp.sys.amiga Subject: Re: For all you who want more advertizing Message-ID: <8208@gollum.twg.com> Date: 4 Nov 90 23:38:41 GMT References: <1990Oct24.030917.18454@usenet.ins.cwru.edu> <693@macuni.mqcc.mq.oz> Reply-To: david@twg.com (David S. Herron) Organization: The Wollongong Group, Palo Alto, CA Lines: 45 In article <693@macuni.mqcc.mq.oz> ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar) writes: >In article sutela@polaris.utu.fi (Kari Sutela) writes: >>I'd say that hardcoding a language into a chip wouldn't be too wise. Please, >>keep in mind that there are other languages besides english (and a significant >>portion of Amiga sales comes from non-english-speaking countries). > >Do phonems vary greatly from language to language? Err.. no. Phoneme's are >exactly< units of sound which can be produced by human vocal cords. There is a whole branch of Linguistics which covers this ground: Phonetics. There's a few competing notations for representing Phonemes in text but they all boil down to the same idea; proceed from a knowledge of the vocal tract & what sounds it can possibly make, and come up with a particular glyph (or glyph combination) to denote each sound. Using this one can represent any utterance in any language. Unfortunately for: >>BTW, are there any plans on providing international versions of the >>translator.library in the future? It's not easy/simple to translate from text to phoneme strings. Especially if you wanna do it for multiple languages. Especially since dialect differences is where phoneme differences surface. Especially since Phonemes provide very little control over inflections, something which gives the "humanness" of human speech. However .. half-assed attempts at generating phoneme strings can work pretty well & be understandable. And once the phoneme string is generated you don't need a seperate speech-synthesis library/device for each language. It just turns phoneme's into sound.. So all that would be needed is multiple front ends for the speech device which accepts text in different languages & produces phonemes. The programmer still has to provide the strings in multiple languages. This is because there is no way, that I know of, to translate morpheme's into text. (Morphemes are from yet another branch of Linguistice: Semantics. They are the units of meaning which make up words.) -- <- David Herron, an MMDF & WIN/MHS guy, <- Formerly: David Herron -- NonResident E-Mail Hack <- <- Use the force Wes!