Re: Fuzzy Logic (Was Tarsky)

Warrl kyree Tale'sedrin (warrl@blarg.net)
Tue, 26 May 1998 20:42:49 -0700


From: Daniel Fabulich <daniel.fabulich@yale.edu>

> On Mon, 25 May 1998, Leonardo Gonzalez wrote:
>
> > Multivalent ("Fuzzy") logic does not conflict with absolute truth. It
> > merely breaks it down into cheweable bites instead of forcing you to
> > swallow a whole apple or none at all.
>
> Too true. So long as we're aware that the truth itself is not fuzzy, but
> we are simply using fuzzy logic as a tool to better explain or
> conceptualize the truth.

According to the best current theories, there are indeed times -- at the
subnuclear level at least -- when the truth itself is fuzzy.

> > Believe it or not, there ARE times when one is neither completely right,
> > nor completely wrong.
>
> Perhaps, but not because reality itself is partially objective, but
> because the statement one is making contains some fraction of true
> information and some fraction of false information.

And sometimes because the language contains embedded assumptions, or
because the speaker is making assumptions -- which may or may not have
been declared, or even be readily apparent. (For an example, there are several
embedded assumptions in the word "murder".)

And then there are true statements which seem false -- and vice versa --
because the language encourages dichotomies where reality is a spectrum.

And of course there are also occasions where the truth is not fuzzy, but our
ability to percieve and measure the truth is fuzzy.

> Like I said, it's fuzzy TRUTH that I'm attacking here, not fuzzy logic.

US$500 fee for receipt of unsolicited commercial email. USC 47.5.II.227