Re: allnoun revisted

Dans un courrier dat=E9 du 22/04/99 19:16:53  , John Cowan a =E9crit :

<< The Loglans also have this characteristic, though they classify the
 closed-class words into 100 or so subclasses ("lexemes", "selma'o")
 depending on their exact roles: it is guaranteed that any word can be
 replaced by any other word in its subclass without changing the
 grammaticality of the utterance. >>

In French cenemes are called "mots vides" and pleremes are called "mots=20
pleins".
I made it reversely : I have isolated cenemes describing the possible=20
relations between pleremes (and I understand that Ray did the same). I have=20
started with basic unaspective relations (that is : outside any process, lik=
e=20
"to be made of", "to be composed of", "to be of the nature of", "to be=20
relative to", etc.) and then I have tried to figure out what kind of=20
processes relations may result in or from ("to become", "to attribute", etc)=
.=20
These cenemes are used with pleremes according to the roles that the semanti=
c=20
meanings of the pleremes suggest. This, of course, implies that a table is=20
for putting things on it rather than under it like you would understand it=20
;-). So basically the cenemes draw the categories of pleremes. But I could=20
never find more than 46 cenemes. The whole thing is conlang stuff and I do=20
not claim any scientific value for that.

BTW, I had the following problem with some of these cenemes for a few years =
:=20
the address in "I speak to you" or the supplies in "I supply you with food"=20
imply that you are going to do something specific with the message or the=20
food : listen, understand, react, eat, retail, or whatever. In other words,=20
the finitive role of the message or the food (what you are expected to do=20
with it) is clear only in "context". So I did not want to use cenemes=20
equivalent to cases like benefactive, finitive, adversative or so to describ=
e=20
this role (although natlangs do). For a while I thought I should have as man=
y=20
cenemes as there are such implied roles but - can you believe that ? - they=20
are quite a few - and they are all against me, all of them as many as they=20
are, aaargh - and I soon got tired of them ;-) so I ended up having only a=20
few of these cenemes meaning roughly "to be a message addressed to someone"=20
and "to be supplied for some well-known purpose or use that I cannot fully=20
guess by now but wait a minute and it's all clear : ... yep ! he=20
ate/retailed/answered/understood/obeyed/etc. to it".

Only this year I read a whole chapter in a linguistic book explaining what=20
"prospective" v. "retrospective" aspects are. So I know now that these=20
ennoying cenemes are "prospective" ones and I cannot sort them into=20
categories like I can do with the 44 other "retrospective" ones.

Mathias