Re: isolation vs inflection & other features

ebera <ebera@...> wrote:
>>>
Anyway, it doesn't change what I meant. Chinese has the same level of
case-marking than any other language.
<<<
if you refer to the existence and the identification of the semantic core actors
(AGENT, PATIENT, ATTRIBUTE, etc.) rather than to the tagging of grammatical
cases (NOMINATIVE, ACCUSATIVE, GENITIVE, etc.) then i'd say: yes.

>>>
I don't ram open doors. I don't know where open doors are. I've never read
any book on general linguistics. I far prefer the 'feel the truth on your
own' method. Here I just tell you what I 'felt true' from my one-year
experience and wait your comments to make it better (what you've done).
BTW, I'm French too.Don't underestimate my modesty, it's my main quality :)
<<<
i don't think you're vain or wrong--but maybe you're too impatient. you may read
books on general linguistics not necessarily to learn linguistics but to learn
the linguistic vocabulary in order to avoid misunderstanding with other people
on this list when discussing this kind of issue. communication means using a
common vocabulary. people here understand the difference between core and
surface cases very well--yet provided that you speak their own tongue.

>>>
Case can be marked by a preposition, an affixe or a postposition. Sometimes
word order can be used to ellipsis these 'spoken' markers. As for Khmer, if
serial verbs are of the kind 'is the owner' then sure I would consider it a
preposition for the possessive case.
<<<
ok, so i can see you're writing about "semantic" cases, which we often call
"roles" here.

>>>
It was meant to mean 'empirical naming'. You don't tell me much on how
professional linguists dealt with this classification.
<<<
french grammarians and semanticians have set a classification of what you call
"levels" (in french: "niveaux d'énoncé"):
1. concepts pair up into one entity (entité) with one behaviour (comportement)
2. a behaviour is either unaspective ("outside time") or aspective ("within
time")
then there are stuff like attribute vs. role, rheme vs. theme, predicate vs.
argument, etc.
note that an "entité" may be a "comportement" regarding its own "comportement"
and reversely so that there is no possible isolated "comportement". if you get
what i mean, then you may understand that i agree with your criticism of Lojban
in some way and myself am criticizing Rick Morneau's compounding relators.
all this is analysed at the "lexie" level, that is beyond the word, phrase and
clause levels--or rather it transcends them. note as well that i discovered from
chatting on this list that all this seems typically french. what is felt obvious
and unquestionable in french books i've read and even in the national automatic
translation machine lab i've visited is not necessarily considered so elsewhere.
now, from what i read from you, i think you will like this kind of "philosophy"
so i recommend the following books:
"Linguistique générale" de Bernard Pottier (éditions Klincksieck)
"L'Homme de parole" de Claude Hagège
"La Sémantique" de Vincent Nyckees (éditions Belin)--the foreword of the last
one reads "Ce que les Anglo-saxons appellent la sémantique formelle [...] n'a
guère produit à ce jour de résultats généralisables." just to tell you how much
chomskyans are welcome in france :-)

>>>
Cases marking relations between nouns, I see genitive as the undefined
case. Just like the undefined article. When you say 'a dog' you speak about
one specific dog but you don't specify which one, and haven't done
previously. Using genitive, a speaker tells his listener there is a
relation between nouns, but doesn't tell which one. It's supposed to be
obvious. If not, the speaker should use one of the defined cases
(possessive, locative). So for 'a girl in bikini' you could say 'girl
bikini' and for 'the girl's bikini', 'bikini girl'. This, of course, is
pointless if the genitive case is marked in the same way than possessive.
<<<
ok. i think you mean that the "true" relation between two nouns or between a
noun and a subordinate predicate may be omitted and you don't feel like tagging
that relation. i would translate in french that you refer to the general
relation between two entities while omitting their behaviour ("to possess", "to
aim at", etc)

> >>>
To my eyes, you're describing me the fact that carnovorous is an
agglutinated word (carn: meat, vor: eater, ous: adjective case) and the
fact that it has the strict definition of animals that always eat meat.
What I meant was to use unmarked genitive when it makes sense and adjective
case (quality relation) otherwise. With the preceding word order, 'a
carnivorous bird' would be 'bird carnivor', leading to the confusion with
'the carnivor's bird'. So here we should use the adjective case. Except if
carnivor is only a quality and a carnovorous animal in general can't be
refered to as 'a carnivor'.
<<<
making "carnivorous" a noun or a verb is completely beside my point. what i
meant is that saying "an animal is carnivorous" is not the same as saying "an
animal eats meat." in the first case, the behaviour of "eating meat" is
unaspective ("outside time"): "eating meat" is an attribute. while in the second
it is aspective ("inside time"): you eat meat sometime, now, always, etc.. yet
both the verb and the adjective are behaviours (and in this case they are
predicative). some langs lexify given behaviours as "inside time" while others
lexify them as "outside time". however, this is not quite obvious on this list
either, so that i read threads like "adjectives are state verbs" vs. "no,
they're nouns", etc.: of course "nouns" (in french: "substantifs") are typically
"outside time" (in french: "posés hors aspect") while "verbs" are "inside time"
(in french: "posés en aspect"). we'd say that whether adjectives are "noun-" or
"verb-"rooted is a matter of mood. for instance, latin "bonus" is "outside time"
and means "the good one" as well (hence your "carnivorous" as a noun vs. an
adjective). when put "inside time" as a predicate (or commonly said: as a
"verb"), the adjective is said to take or imply a "copula" like "to be" in "to
be good".

OBCONLANG: some conlangers make conlangs based on nouns only (allnoun) and some
others based on verbs only (allverb). maybe this radical choice shows the
conlanger's psychology: allnoun makes everything outside time and shows a desire
to escape the real world's decay--while allverb makes everything inside time and
shows a desire to master the real world's blooming. may this line set fire to
the whole list! mwahahahahahaaaaaaaaaaaaaaaaa!!!!

Mathias
http://takatunu.free.fr/tunugram.htm

Replies