[YG Conlang Archives] > [Latejami group] > messages [Date Index] [Thread Index] >
I just finished experimenting with the new system. The results do seem (to me, at least) to be slightly easier to pronounce, but not enough to justify switching to the new system. There are still too many semi-vowels and the words are ugly. Here are a few examples: old: Te Laryase pa jutay sya bweko dwepa libyekyu. new: Dwe Laryaswe bay butwe swa penkoy pebay lijezwa. old: Gote gogyose giku to pa zawtwa dico jutay sya bweko dwepa pa. new: Kodwe kocoswe gigyu tay bay satwa dijoy butwe swa penkoy pebay bay. I'm going to try a different approach - one similar to a much earlier attempt. In this new approach, any modifier can be used as a classifier and vice-versa. The rightmost root morpheme will be the classifier. Any that precede it will be modifiers. However, this means that every word will have to be terminated by at least one suffix in order to maintain self-segregation. [By the way, self-segregation is something that I absolutely insist on because I eventually plan to write speech recognition software for the language.] Since every word will have at least one suffix, words will be longer. This doesn't concern me much because the interlingua is intended for machine translation. Besides, even with the extra suffixes, I suspect that the resulting sentences will still be comparable in length to most natural languages. It should take me a few weeks to try this new approach. If I don't like it, I'll probably stick with the current system, as much as I hate it. Regards, Rick Morneau http://www.eskimo.com/~ram