Re: Two different opposites (again)

Hi!

Carsten Becker <post@...> writes:
> From:    Gary Shannon <fiziwig@...> wrote:
> <<<<<
> know <-> not know
> know <-> unknow (forget)
> do <-> not do
> do <-> undo
> This seems more like the three points of a triangle
> than the two endpoints of a single spectrum.
> Yet in
> other cases these two different opposites really mean
> about the same thing:
> welcome <-> not welcome
> welcome <-> unwelcome
> happy <-> not happy
> happy <-> unhappy
> >>>>
First of all, I see this a bit differently, i.e., no triangle.
E.g. in 'undo' it and 'not do' the first negative reverses the action,
but the section reverses the truth value of the predication, which is
not really triangular, but simply a different scope of what is
negated/reversed.

As to some adjectives, I have two negatives in Tyl-Sjok that work as
follows in a way that 'not good' and 'bad' are different: if you look
at the continuum from -1 to +1, from good to bad, then good is at one
end, and bad is at the other, say, 'good' means x < -0.5 and 'bad'
means x > +0.5.  Now, one of my negatives multiplies the unequation by
-1.  This is 'the opposite'.  The other one switches the direction of
the comparation operator, i.e., 'not good' is x >= -0.5.  This is the
'truth negation'.

In a table:

   good          <  -0.5
   OPP good      >  +0.5
   NEG good      >= -0.5
   NEG OPP good  <= +0.5

Graphically:

   -1...........-0.5...........0...........+0.5..........+1
    <-----------good                      OPP good------->
              NEG good----------------------------------->
    <-----------------------------------NEG OPP good

> This seems to be an interesting feature for a conlang. But is this
> distinction at least a little naturalistic? I'd like to include this in
> Ayeri, but I'm not sure because of my doubts. Ayeri is nearly as regular
> and unambiguous like a loglang, but nevertheless is thought to be a spoken,
> naturally evolved language.
No problem, I think.  It might be a regular feature of a natlang to
have such negations.  If we kindly ask Ray, he will probably show us
that Latin has similar features. :-)))

**Henrik