Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

re:RE: SUO: RE: First piece of 4D ontology




>### You and others use KIF to specify axioms, rather than an
>"accepted" syntax such as mathematics ("accepted" = "universally"
>semantically agreed, at least since Hilbert and consorts --and
>pardon the many quotes) in which, incidentally, the 4 axioms could
>be written much more concisely.

The use of KIF syntax is one choice among many, and is of little 
significance in itself. There is no 'standard' syntax for logic or 
mathematics ; there are prefix, postfix, infix and many other 
syntactic variations (not to mention graphical notations), and the 
choice between them is only an aesthetic or pragmatic issue.
Offer a silent prayer of thanks that we arent writing it all in XML, 
which is rapidly becoming the 'standard' for a new generation.

>Clearly this is done with a certain
>form of future computer processing in mind, i.e. interpretation by
>a program for purposes of storage, validation, reasoning, ... by
>computers. This requires a formal definition of the syntax used --in
>the case at hand this likely involves an (the!) ontology itself we
>are trying to define. So, in order to avoid infinite regress, we
>have no alternative but to axiomatically define (= agree on) the
>basic predicates such as "exists ?x ?y".

Exists is not a predicate, it is the existential quantifier, as used 
throughout logic for about a century now. Its meaning is not (cannot) 
be given axiomatically, but is properly expressed by the rules of 
interperation for the formal language. In this case, (exists (?x) 
PHI[?x] ) is true in an interpretation I if there is something in the 
universe of I such that if ?x is interpreted as denoting that thing, 
then PHI[?x] denotes True in that interpretation. (This is a sketch 
of the precise definition.)

>E.g. if this merely involves
>finite sets, one "only" has to agree on the definition of the extension
>of ?x and on the lookup of ?y in this extension --but one assumes there
>is more at play here, otherwise the whole SUO becomes an exercise
>in triviality.

There is no requirement that a universe of interpretation be finite.

>I believe thick volumes of logic have been written on
>this subject of "exists" alone (cfr. the Intuitionists, ...) so a
>proper initial agreement seems required. Of course I don't doubt that
>[the interpreter for] KIF++ has/will "prescribe" this; my point is
>that it will be required at every step and needs to be performed
>explicitly.

If this means that the formalism being used should be given a clear 
semantics, I agree wholeheartedly. I would however point out that 
this requirement seems to be at odds with a Peircian emphasis on 
meaning depending on the circumstances of use of a symbol by an agent.

Pat Hayes

---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes