Open eddieantonio opened 5 years ago
Hey, Aaron! I've started writing this documentation, but I realize it's going to take A LONG time to write fully. Could you please list the preverbs you want documented? I'll update this page to reflect it: https://ualbertaaltlab.github.io/plains-cree-fsts/verbs
I think I'm making progress on this @eddieantonio, for instance my lack of understanding of the FST structure had me thinking that kâ-kî-
was a Preverb when in fact it is implements as the past (Prt
) version of kâ-
, which makes sense.
The way we are approaching realtime lookups on Cree words in our app is by pre-inflecting all possible forms with supplemental information, the goal being a way to link transcribed information back to inflected forms with as much supplemental information as possible tied to that form, including where the inflection was used in transcriptions and other materials (non-functional example here). We already have 'word matching' tied into our work (example here) and are systematically working through the words and forms that aren't "matching" and should be. This is the motivation for many of these questions - how to generate certain forms and possibly if there are discrepancies.
With all that, a good starting place might be some long-vowel forms wî-
and possibly some of these:
ka-kî- (independent) eg: nika-kî-itwân "I could say (thus)"
ta-kî- (conjunct) eg: ta-kî-itweyân "I should say (thus)"
kita- (conjunct) eg: kita-mosiwit "he became a moose".
kâ- (conjunct) kâ-nikamoyân "I who sing"
kâ-kî- (conjunct) kâ-kî-ohkomiyân "when I had a grandmother"
If it is not possible to generate a form maybe it should be classified as a bug?
Originally posted by @aaronfay in https://github.com/UAlbertaALTLab/plains-cree-fsts/issues/6#issuecomment-534255352