Stumbled across this Kalaallisut (West Greenlandic) text-to-speech site the other day. It’s not very sophisticated – the creator didn’t include all the required phonemes, for one thing – but it’s not every day you see speech synthesis that isn’t Indo-European, Japanese, or Chinese. Does anyone know of any other speech synthesis projects involving Eskimo-Aleut languages?
Having tried speech synthesis in an advanced lab phon course, I know it’s much more difficult than it seems. I’m impressed when anyone can get results that sound even vaguely like they should. It strikes me that for a language like Kalaallisut, too, phrasing and intonation becomes more important in synthesized speech – less overlook-able, if you will – due to the tendency for very long words. Perhaps that’s just a symptom of my not being a speaker of the language, though.