Re: Re : Malat

Garrett wrote :
Wow, we're really going off on this subject :)
>
And that was only a start :-)
snip
> > Garrett :
> > Well, over here, i'm going to make drink and eat the same word. Is there a
> > necessary distinction? The equivalent in english is implied by what is
> > actually consumed (food or liquid).
> > Mathias :
> > Your verb would not be a Malat translation of *to drink* but another concept.
>
> Not an exact translation, but a perfectly usable word in the situation.
No doubt. But I regard the fact you have nice precise verbs like *to cut into
twigs* and *to slurp* also. I'm myself an ex-all-noun-root-ing-language-er ;-)
of an all-noun-root-ifi-ed(/ee?)-ex-language with only 12 operating verbs.
*Let's have a seat and have a few words about the soup we had tonight*. 't was
appalling.
snip
> > Garrett :
> There's an example of some illogical natlang derivation...
>
You dare hint my ancesters could not derive words properly (you b... ! ;-) I
believe otherwise - even after taking into account that the meaning of words
drift at a dramatic speed. I think my ancesters were no less logical than you
(and much more than me). Just not your logics nor mine.
The funny thing is that one famous contemporary standard logical noun-deriving
language is Indonesian, and one famous contemporary standard logical causative
language is Japanese, two peoples I've personally experienced as having very
nice..., er, un-Western logics now and then - aside Rick Morneau's beloved
Swahili.

> Were the derivations illogical or what? why were they hard to remember?
>
I can understand that you remember *logical* derivations quite well. I'm no good
at that beyond derivation from cases. I know I'll have to learn each derived
noun as a new root and with even more difficulty since they all will sound
alike.

Let's take Indonesian. Very easy and very logical noun-deriving language, but within
a narrow range :
patient : -an
quality : ke-an
agent/instrument : pe-
process : pe-an
result : per-an
You would add your *focus* (*-ee*).
Further derivations are skipped because that's where compound words' realm opens. I
personally think that beyond the borferline of cases, compound words should
relay derived words. Concepts are no equations and the meaning of a word is not
tinkered from a lot of this with a bit of that. I've not heard that the
barycenter of a single word was calculated yet. The only way to achieve that
goal is to take a root, reset the scale to zero and stick *derivators* on that
root that have a fixed weigh. Then you pretend that each additional Oz points
the scalehand to another human concept. I imagine that it's what RM does every
single day in his life :-).
And to stray still more off-side (:-) I do believe that wrapping our human, linear
language with mathematics does make it more... mathematically logical, indeed.
But I'm sure you know that a more radical shift would be to switch to
non-linear language like UNL instead.

Did you read Chinese grammarians ? They're now derided because - to make it short
- they believed that in *white horse*, white and horse have the same status.
How couldn't they make a difference between adjectives/verbs and nouns for thousand years ?
Just because they looked at degrees of integration reversely from us : we go from
sentence to compound words, from least to most integrated ; they go reversely
from most to least integrated. So we always tell the predicate from the
argument inside a clause and inside a noun too. It's OK with derived nouns
which are made like an argument incorporating the predicate (that's an image)
and existing therefore. But in these Chinese grammarians' minds, there is no
such syntax within a compound word, so you can't tell which one of *white* and
*horse* is the other's argument or predicate. Actually they are mutually and
simulteanously each other's argument and predicate, each other's actor and
verb.
snip (was interesting but nothing to add to your neat description)
> Oh, the antipassive that you're talking about is what i think Rick morneau
> called 'middle voice'. That's where the subject is de-topicized and made
> unable to be expressed. Another example: "the mouse kills easily". For
> examples one and three above, in malat the first verb would be between agent
> and patient (as the action), and in the third the verb would be after the
> patient (the result of whatever action I performed). Same with 2 and 4.
Right. *Middle voice*, *reflexive (middle) voice* and *deponent verbs* are the
*antipassive* forms of nominative languages. *Antipassive* is used in some
ergative languages where the absolutive already translates as a *passive*.
But in DEL and TUNU antipassive swaps active subject and object :
I buy the car (active) > the car is sold by me (passive) > *the car sells (to)
me (antipassive). I see it > it is seen by me > it appears (to) me. etc. Same
in TOMATO. That's how many Japanese verbs inflect (except that *subject* tag
-ga becomes *indirect object* -ni) and that saves loads of voc-roots IMHO.
What in Malat if the agent misses ? Is there a default agent pronoun ?
Your fellow conlanger
Mathias