[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
Robert LeChevalier scripsit:
> 2. As noted in the ensuing discussion, kau can attach to a definite
> sumti/cmavo value as well as to the question words (and therefore can
> attach to any word that can end a construct equivalent to a question word
> selma'o, so all possible terminators of sumti structures should be able to
> take kau). Thus Cowan's grammar change proposal is incomplete even for the
> current known applications.
Quite so. I'll mess around with allowing it to attach to sumti and selbri.
> 3. Derivative of 2, it seems likely that there could arise usages of kau
> BEFORE there arise question words for a construct. There is no termset
> question word and I can't imagine wanting one, but one can stick kau on the
> end of a termset to indicate that one knows it is the correct termset value.
This is the pernicious "kau" = "knowledge" meme surfacing again <stomp/>
"kau" only has to do with knowledge when the outer bridi is a "djuno" one.
This is a dangerous case to reason about because of factivity issues.
Stick with "wonder" or "doubt" instead.
> 4. If kau is not in UI, I don't think it can go on a UI question
> word. Don't know if it matters, i.e. whether peikau is usable.
I added explicit rules for UI+KAU and CAI+KAU to cover xukau and peikau.
--
John Cowan <jcowan@hidden.email> http://www.ccil.org/~cowan
Raffiniert ist der Herrgott, aber boshaft ist er nicht.
--Albert Einstein