> > Would it not be more straight forward to just *define* truth to be a
> > property of a model; a higher truth value corresponding to a "better"
> > representation of "objective reality" (determined by intersubjecive
> > agreement)?
>
> But you're starting assumption is that the "best" model is the one closest
> to objective reality. Take one more step back. Challenge that primary
> assumption. What if the "best" model is defined in some other way?
> Without a necessary regard to its relation with objective reality. What
> new attractor will you cicle around? Is that a "better" (more useful)
> place to go to solve some problems?
Umm. You're right, but that was not my orginal idea. You see, I want
to *define* "truth" to be a property related to the objectivity of the
model.
And so the handle "truth" is what I want to use to describe that *type*
of model. If you want other models (which might be used for other
things), call them by a *different* label.
Like "art" or "model useful for [x]" or "beautiful" or "sham".
ERiC
... heh. I love argument by definition!