Re: GROUPLANG: optional features and case

I use to say : 'to free yourself from rules, tame them tight'.
I re-read Herman, Pablo and Carlos' posts regarding genitive, copulative,
predicate, determinant and modifier 'cases'. These topics may be discussed
again unfortunately upon issues that are not be case-related at first glance
like the translation of 'to be', 'to have', of adjectives and participials like
'powerful', 'bitten', etc. So maybe I should SUM all it UP here from
linguistics viewpoint :-) Sorry if it's a few more lines, but I do believe it's
a good tool for those who make their own conlangs and above all ubderstand each
other.
W, X, Y, Z represent specific cases herebelow :

1. Reference = 'the_red_dog I'm talking about'
W-dog pred-the_red_dog = (this) dog is the_red_dog.
=> *mu-the_red_dog-W-o dog = *the dog who is equated as the_red_dog  = THE RED dog.
Case W is an identitive equative.

1. Reference = 'dog'
X-red_thing pred-dog = the red_thing is equated as a dog = the red_thing of the class of dog
=> mu-dog-X-o red_thing = the red_thing that is equated as a dog = the red_thing
that is of the class of dog = the RED dog.
Case X is a classifying equative.

2. Reference = 'red_thing'
X-dog pred-red_thing = the dog is of the class of red_thing => mu-red_thing-X-o dog
= the dog that is a red_thing = the red DOG.
Case X is a classifying equative.

3. Reference = 'red_color'
Y-dog pred-red_color = the dog features red_color => mu-red_color-Y-o dog = the dog
that is a red_color-ed = the red DOG.
Case Y is an attributive : a loose attributive if it's only paint on her or
inherent attributive if it's a feature of her hair.
Same if reference = 'redness' as a state > case is then the 'absolutive'.

Regarding all cases the 'Reference' is actually the predicate. The predicate is the
reference whereby all cases relate to each other. That's why predicate is
usually not a case except in very rare agent/patient systems or in Turkish
attributive : of-room its-door = the door of the room. One conlanger among us
uses a similar construction he calls 'master-servant' system in a Japanese-like
language called Amaterasu or else I can't remember.

'modifier', 'determinant' are only a way to fasten a predicate to a noun more ('modifier')
or less ('determinant') tightly.

abs-me pred-bite(-e) = I bite;
bite-abs-o = biter; the one biting;
bite-pat-o = bitten; the one bitten;
age-me pred-bite-abs-o = I am a biter;
bite-abs-e = biting state;
bite-pat-e = bitten state
mu-bite-abs-e dog = biting dog;
*dog mu-bite-abs-age-o = *a biter-dog;
dog abs-ki pred-bite = dog who is a biter.

Cases W, X, Y above are unaspective, i.e. they don't refer to a process, only to an
identity, a state, or a class outside any process as nouns usually do. To refer
to a process from a noun, you need make the noun aspective, that is, you need
refer to the usual symbolic behaviour or work of that noun. For instance, you
infer that mill grinds grain, that hammer hammers, that river flows, that light
lights, that fruit grows on plants, that boss commands, that eye sees, etc. You
'symbolize' a process through its AGENT. To do so you may either tag on the
noun predicate 'hey, it's a symbol for a specific process', or use another
case, the agentive case. It's very handy a case in English : 'to shelter', 'to
stream', 'to loom', etc.
So you have 4 cases :
Identitive equative
Classifying equative
Attributive
Agentive
In this language we only use the last two ones. The first two ones would be
replaced with roots like 'identity' and 'class' :
att-she age-red_dog pred-identity = she is the red dog I talk about
abs-she age-red_dog pred-class = she is of the class of red dog

Note that attributive and agentive are very different from equatives because they
use the symbolic 'hidden' meaning of the noun to make it a predicate :
brother > to relate as a brother > age-he att-me brother = he is my brother;
disease > to strike as a disease > age-smallpox att-me disease = I suffer smallpox.
Many 'archaic' languages use this double nature of nouns to develop a symbolic
world or construe omens : 'hawk' is the symbol for 'ruthlessness', 'lark' for
'fickleness', 'bucket' for 'scooping' in a river-country and for 'reservoir' in
a desert, etc:-)
Essive (locative and situative cases) and lative cases are like equatives : they
don't infer a specific symbolic hidden meaning in the predicate. That's also
why I suggest we replace them with verbs.
Note also that the number of symbolic meanings 'hidden' in one noun used with
agentive ranges from 0 to infinity, where 0 = infinity = classifying equative :
tree > age-me tree = I am a symbol for tree = I am of the class of trees (0 or
infinity of meanings)
seat > age-me seat = I am (used as) a seat for xxx (1 meaning)
tooth > age-me tooth = I behave like chewing (?) cutting (?) biting (?) tooth (at
least 3 possible meanings : make your choice:-)
Mathias

-----
See the original message at http://www.egroups.com/list/conlang/?start=17513
--
Free e-mail group hosting at http://www.eGroups.com/