Friday, November 14, 2003

A typological consistency dilemma

Last night as I was trying, unsuccessfully, to fall asleep, a thought occurred to me about my current method of nonproductive (mainly) derivation, viz. the application of the stressed prefixes (keli > mákeli). Unfortunately, though I do really like this, I realized that this employs the order Dependent-Head, where everywhere else in the grammar (apart from particles) ordering is Head-Dependent; this includes, of course, compounds, making it all the more strange that a different ordering system would be used word-internally.


As a result, it seems that, if I want to save this system of derivation (which I think is desirable), I need to change this to a suffix rather than a prefix (keli > kélima). I suppose this does have the inherent advantange of drawing attention to the meaning of the root; though I worry that, as such, the suffixes may not be sufficiently stressed for necessary audibility. I seem, however, not to have much choice in the matter.


Incidentally, it appears that this is how numbers need to work, more or less--ten (or so) numerical roots, with monosyllabic suffixes showing decimal place. The alternatives, being (1) particles for numbers, and (2) whole roots for decimal place markers, are clearly unsatisfactory: (1) would use up the entire repertoire of particles to very little effect, and (2) would create enormously long and unpronounceable phrases even for small numbers.


Aside: In addition to strict numerical roots, we also need something for "many," "few," etc. Also develop a way to do the kilkadzesiąt kind of thing.

No comments: