[YG Conlang Archives] > [Latejami group] > messages [Date Index] [Thread Index] >
MorphemeAddict@hidden.email wrote:
>
> Is there any connection in the new scheme between classifier meanings and
> modifier meanings?
> ...
> ci- cipu ci
> faw- kavo ka
> fe- fepi fe
> ...
Yes.
>
> But the following don't following this pattern:
> ca- cala za
> ki- kiva di
> se- begi tay
These were my mistakes, now corrected. Thanks for pointing them out!
cal -> ca
kiv -> ki
beg -> be
> fa- kigi kwi
ki -> kwi still preserved the C avd V.
> kyo- bave ! kwa
"Bav" is the classifier/root meaning 'now', a temporal deictic. It has no
mnemonic link to "kwa" which means 'pull/receive/request/collect/enter'.
> twe- tesye xe
I could have used "xes" for the classifier, but I felt that it might be
confused with other similar-sounding classifiers. Besides, the semantic
link is not that strong. [Classifier "tes" is for true conjunctions,
while "xe" means 'two/divided/opposition'.] Conjunctions can link more
than two constituents, and the most common conjunction, "and", rarely
implies 'division/opposition'. In fact, it almost always implies the
opposite.
>
> Is there any pattern or rule behind the division of consonants into
> C1, C2/C3?
>
I choose them primarily for distinctiveness and variety. However, "l"
was a freebie because, when it terminates a classifier, it still can't
be confused with a prefix. Also, I couldn't use 'c', 'j', or 'x' for
C2/C3 because there are now two part-of-speech markers that start with
'y' ("ye" for true conjunctions and "yu" for open adjectives).
>
> It seems like it'll be hard to separate modifiers from classifiers.
>
I disagree. You just have to learn the new rules. Besides, only the
word-designer needs to be able to parse words, and I have no problem
doing so. The translation software writer does not need to be aware at
all of the self-segregation system.
Regards,
Rick Morneau
http://www.eskimo.com/~ram