RE: Expressiveness of RDF as Rule Conclusion Language (was Re: W hat is an RDF Query? )

>  > The classical approach is to think of queries as negated
>>  and think of the entire inference process as finding a contradiction.
>>  That is a very 'logical' viewpoint which many find unnatural,
>
>It is not just unnatural but also not very general, because in
>many formalisms/logics the principle underlying this approach,
>viz proof by refuting the negation (or "reductio ad absurdum"),
>is *not* valid!

I wouldn't say 'many', myself.  Even in weak logics which do not 
support the transformation as an inference, the 'negated' 
interpretation still has some utility. But your general point is well 
taken.

>Although it is quite common to view the inference process
>in logic programs in this unnatural way (apparently, because
>many people like to be 'logical')

We would all get along better if ad hominism were kept in check. The 
reason why I, and I suspect others, find this construction natural is 
not from some social need to impress people, but because it is 
sanctioned by the usual semantics for classical logic, and moreover 
pre-dates logical programming.

>, there is absolutely no
>need to do so. It is conceptually much more adequate (more
>natural and more general) to view the inference process as
>being based on simple rule application (detachment).

This has its own problems and limitations, however. It confuses 
'rule' in the sense of inference rule, with 'rule' in the sense of an 
implication; and,  moreover, it obscures the underlying semantic 
unity between modus ponens and modus tollens inference forms (in 
classical logic, that is.)

Pat
-- 
---------------------------------------------------------------------
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

Received on Wednesday, 26 September 2001 11:04:03 UTC