[YG Conlang Archives] > [ceqli group] > messages [Date Index] [Thread Index] >
--- In ceqli@yahoogroups.com, "Jim Henry" <jimhenry1973@...> wrote: > > On 4/23/06, Rex May <rmay@...> wrote: > > > rules. Currently, the rule is stress the first syllable. That may > > be the optimum way to do it. (And I shouldn't really be saying > > 'words', but 'morphemes'). > > Are morphemes stressed the same independently of whether > they stand alone or are in a compound word? I've often felt that Tceqli could distinguish compounds from discrete words by an application of stress. Perhaps a level or unstressed profile for core words and a non-level profile for compounds. Admittedly, stress is not ironclad, but it can be a useful clue when combined with other clues (presence of little words, knowledge of lexicon). > > >I've also considered penultimate stress, > > which works okay, and having stress fall on the last syllable that > > doesn't end in a vowel. I wonder if my propensities are confusing > > me. A word like 'kanor' seems like it should be stressed on the > > last syllable, no matter the rule. And a word like 'himel', seems to > > need stress on the first syllable. Am I reacting to some intrinsic > > linguistic tendency here, or am I just exporting my own prejudices. > > I don't have any problem with /'ka.nor/; it seems as good as /ka.'nor/ > if not better, so I don't think it can be a strong intrinsic tendency if it > two native speakers of the same language can differ in their > inclination so. Ditto in reverse. Second-syllable stress on both those words seems natural to me. And I even lived for seven years in Colorado, just 90 minutes south of Rex, so it clearly isn't anything intrinsic even to regional accents. > > > If the former, perhaps Tceqli should have slightly more complicated > > stress rules. > > This is probably a bad idea. Auxlangs should have a simple > stress rule; of the ones I've tried, first syllable and penultimate > syllable seem more euphonious than last syllable. Though > maybe better still is to have even stress as in Japanese, > or design the language so that speakers can stress whatever > syllables they like and still be understandable -- I think your > self-segregation rule for Tceqli probably accomplishes that, > but we would need face-to-face conversational testing to be > sure. I generally agree. The one area where I get lost in Tceqli is that I lose track of where I am syntactically, sometimes because I've interpreted a compound as distinct words. --kraunzo