[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
On Tue, Dec 17, 2002 at 03:04:54PM -0500, John Cowan wrote:
> Jordan DeLong writes an admirable post, but then undercuts it with:
>
> > [1] even in fuzzy logics this has nothing to do with anything. My
> > understanding is that a logic with infinite truth values ranging
> > 0-1 considers the value of the expression to be a measure of our
> > certainty of its truth (or whatever).
>
> No, certainty is neither here nor there; it is "truthishness" that's
> at stake. A better way to view it is to map all talk of truth into
> talk of set membership: a car is blue iff it belongs to the set of blue
> things. Now we can understand a fuzzy-logic claim that "the car is blue"
> being 90% true by mapping it to a fuzzy-set-theory claim that the car
> 90% belongs to the set of blue things.
I think this probably depends on which system. McCawley describes
a fuzzy logic in which things are based on certainty. He's also
not very rigorous about things though, and it's all second hand
descriptions---I understand what you're saying, that in a real fuzzy
logic system it's possible to view relations as fuzzy sets (as they
are sets in normal logic) and treat set membership as the more
primative concept. I dunno much of anything about fuzzy logic other
than a few second-hand readings, so ... I dunno. :)
This all gets off the real point, though, which was:
> Certainly the numerical intensity of the blueness is irrelevant, you're
> right about that.
--
Jordan DeLong - fracture@hidden.email
lu zo'o loi censa bakni cu terzba le zaltapla poi xagrai li'u
sei la mark. tuen. cusku
Attachment:
bin8BzXkqZkIY.bin
Description: application/ygp-stripped