> > No, that's not what [] means - (you can't just remove them!) they mean... > You can remove them if you replace them with an existentially > quantified variable, yes? I think so (in so far one can think while having fever ...) > And you can remove an existentially quantified variable in an asserted > context by replacing it with a new term (that is, by skolemization), > right? Well, you could replace that existentially quantified variable with a (Skolem) constant if it is not falling within the scope of universally quantified variables. In that case you can replace it with a (Skolem) function (which has as arguments those universally quantified variables. BTW this thread is actually about function terms ... How could we express these Skolem functions? > In another context, like a negated context or a query, I agree > skolemization would be wrong. I guess I should have been more explicit. Sandro, what do you mean with "negated context"? I could understand that negation changes the "kind" of quantifier (I mean a universally quantified variable becomes an existentally quantified variable and vice versa). I also think that so called "negation-verbs" such as in TimBL example </etc/passwd.rdf> log:doesntSay { :<#timbl> unix:inGroup <#www> }. is a very good thing to use (it was actually the only way we could deal with paradoxes like the Russell set paradox http://www.agfa.com/w3c/euler/russell.axiom.n3 (where :keine is meaning "not an element of")).... -- Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/Received on Wednesday, 7 March 2001 08:57:24 GMT
This archive was generated by hypermail 2.2.0+W3C0.50 : Wednesday, 11 January 2006 15:19:05 GMT