Re: THEORY: Re : THEORY: French Linguistic Thinking.

Dans un courrier dat=E9 du 07/06/99 22:23:29  , Ed a =E9crit :

> what worries me is that he
>  is=20
>  > now an esperantist convert... ;-)
> =20
>  Really! I hadn't read his "search for the perfect language" book yet.
>   Is that where it all ends?  Esperanto?
>
maybe he found out that you can't find out anything, so he'd better support=20
the most motivated auxlang activists. (;-) you know, i'm always amazed=20
esperantists are so convinced it is the best ial.

>  > (12) to wend
> =20
>  To WEND?
>
yes : "to wend one's way" through the forest, on the field, in life,=20
under-ground :-)
japanese : yama wo aruku.=20

>  > (13) to wend according to the pattern of
> =20
to go round and round, to jump over. (but to go, to be round and to jump are=20
different)
once again, don't take this too seriously. it's ok to have the key but it's=20
difficult to open the door.

>  Not at all!  You've obviously spent a lot of time on this problem.
>
which means i've been spending a lot of time designing a useless=20
construction. like any good conlanger :-)

>  I think the theorists I like are sort of post-Chomskian, trying to
>  scrap that whole Generative Grammar project and start over, as opposed
>  to the French who never took it up to begin with.
> =20
they took it up but understood it quite differently as i've read it : pottie=
r=20
applies chomskian stuff to semantics as well. to make it as short (and wrong=20
and unscientific) as i can i understand that he does like that : a house is =
a=20
building (s=E8me connotatif), a cat an animal, soccer a sport, an armchair a=20
chair with arms (s=E8me d=E9notatif), etc.. an utterance is mapped on a=20
"actance-dependence" diagramme. actance roles tie up the s=E8mes available i=
n=20
utterance according to the actance-dependence diagramme. not reversely :=20
that's where chomskians may shrug it off.
i consider myself that attribution is the "dependence" vecteur so that's why=20
i told you that "gramm=E8mes" are s=E8mes also. pottier doesn't at all of co=
urse=20
and makes the usual  distinction morph=E8me =3D s=E9m=E8me/cat=E9gor=E8me wh=
ere=20
cat=E9gor=E8me =3D gramm=E8me/lex=E8me. so "gramm=E8mes" exist only at a mor=
ph=E8me level.=20
i think differently. it's my way to explain to myself why i have noticed tha=
t=20
"words" put in dependence usually lose core role tags in all languages i kno=
w=20
: "mike wins" > "mike's victory". i agree it's very simplistic. but=20
everything is allowed to a conlanger ;-)

mathias