Re: Universal Translation Language

Dans un courrier dat=E9 du 02/06/99 23:54:07  , Charles a =E9crit :

> Most experts say 3 core cases, I try to get by with 2.
>  The noun cases are somehow "dual" with respect to verb voices.
>  Oblique cases are more explicitly like a verb + noun.
>  Trying to split noun from verb is like cutting a magnet in half.
> =20
i agree 100% with you. the third core case is a trick to avoid referring to =
a=20
second, unavoidable verb.

>  According to deLancey's argument, the fundamental pattern is:
>  "agent CAUSES theme BE/BECOME state", for a total of 3 case roles
>  and 3 fundamental relations. That's where I'm metaphorically at now.
> =20
i disagree 100% to this (although i wish things were that simple). CAUSE doe=
s=20
not take into account FINALITY :
"the man causes the child to perceive an image" does not equate "the man=20
shows the child an image". even Japanese get that point when deriving both=20
verbs from the same root ;-).

the causative states retrospectively that an unprescribed agent (the man or=20
anybody else) causes a specific state subsequent (to perceive).

the finality analysis states prospectively that a specific state or specific=20
agent ("to show" or "the sign") contributes to an unspecified state=20
subsequent : what does the child do when you show the image to him  ? does h=
e=20
always perceive the image ? No, he is only intended to do so but may not=20
after all.

we are human, we think in terms of causality AND finality : "pr=E9voir, c'es=
t=20
r=E9gner".

(Tunu solves that issue with sheer genius, of course ;-) (JOKE!!!!!)

Mathias