Re: Re : Re: Re : Case, Innateness, Almost Allnoun, NGL.

From Http://Members.Aol.Com/Lassailly/Tunuframe.Html wrote:

> i don't think we can make anything more simplistic
> than the idea that all roles are space & time concepts.
> it is just wrong. as wrong as saying that transitivity is
> merely causation. dream who may. however, if it works
> with a machine, then this should be so for a human being,
> right ? wrong. ever heard of Eurotra ?
Reductionism isn't what we want to do.
I'm just trying to figure out how to SVO without tripping.
Since SVO is child's play, no need for metaphysical roles.

> >  Aspect is more useful for word-building, right?
> word-building is nothing. word-breaking-down
> is *something*. don't you agree ?
Hmm. Maybe so.

> >  So, one should-can systematically derive "a speaker/saying/hearer",
> >  a "donor/gift/recipient", and a "destroyer/destruction/victim".
>
> not quite so. the range of aspects and roles of a process is limited.
> "being spoken to" doesn't mean you actually hear,
> but at least the "to" of "being spoken to" prospectively infers
> that you have the ability to do so ("addressee").
> nothing like a localistic "speak goto you".
> this is not a minor point.
We need 3 simple core cases. Facts may be more complicated,
but we speak SVO or "subject verb patient preposition object".

> "tell about something" is not equivalent to "tell something".
> in "receive from A"  A is either an agent or a source.
> "make something be something else" is a causation,
> not a creation. these are no details either, but different roles
> pertaining to cognition, expression and status mappings.
>
> write in english is a faculty, write on paper is an application, write a
> letter
> is a creation, and all this has nothing to do with space and time.
> recipient is no location, result is no extremal, pattern is no extent.
> ignoring this and other notional roles not pertaining to space & time
> realm is ignoring 90% of language ability. full stop.
True, but secondary. All those finer distinctions are added to
the simple core. Otherwise children would be seen and not heard.

> >  Probably one should also derive "telephone/checkbook/weapon" ...
> >  but I don't know how.
>
> these are "instruments".
Well yeah, but how do I get "weapon" directly from "destroy"?
It doesn't work so well as deriving "destruction/destroyer/etc".

> these are old, basic semantic categories.
> of course, if you consider cognition or expression or
> finis as mere coordinates on a diagram and not as valid
> roles then all this is nonsense.
> there are plenty of instruments : path, access, vehicles,
> containers, clothes, covers, food, etc.
> and very un-locative ones indeed.
>
> but i think you know that by now, Charles, don't you ?
We've been discussing it, but there are 2 sets of derivations
that can be made from verbs: those that are relatively direct
(from core roles, destoy-er -ee -tion etc) versus those that
need some extra input (sword, shield, joke, atom bomb, etc).

I think a simplistic first approximation is necessary.
Then a second approximation adds/corrects the first.
I don't want to be precise, most of the time.
One can always add modifiers when desired.