- From: Tanel Tammet <tammet@staff.ttu.ee>
- Date: Wed, 13 Jun 2007 08:08:34 +0300
- To: Pat Hayes <phayes@ihmc.us>
- CC: Sandro Hawke <sandro@w3.org>, semantic-web@w3.org, timbl@w3.org
Pat Hayes wrote: > You are begging the question. Suppose an ontology asserts > > ex:Venus rdf:type ex:AstronomicalBody . > > Now, what ties that object URI to the actual concept of being an > astronomical body? And so on for all the other URIs in all the other > OWL/RDF ontologies. The best you can do is to appeal to the power of > model theory to sufficiently constrain the interpretations of the > entire global Web of formalized information. But that argument from > Herbrand's theorem (basically, if it has a model at all then it has > one made entirely of symbols) applies just as well no matter how large > the ontology is. > > The only way out of this is to somewhere appeal to a use of the > symbolic names - in this case, the IRIs or URIrefs - outside the > formalism itself, a use that somehow 'anchors' or 'grounds' them to > the real world they are supposed to refer to. If we all assume that > English words are so grounded (not a bad assumption) then this can be > done in principle by using the URI in English sentences or to other > kinds of representation which are widely accepted as real-world > identifiers, like SS numbers or facial images. I did all three in > > http://www.ihmc.us/users/phayes/PatHayes.html > > If the TAG said this somewhere, and recommended how to do it, that > would be great. Approaching the same issue from an opposite direction: There is not much hope of "guaranteeing" that http://www.ihmc.us/users/phayes/PatHayes.html will always automagically denote the actual Pat Hayes (despite the convincing suggestions on the referred html page). For similar reasons people will use their own URIs to denote Pat Hayes. I could, say, write a similar html page for Pat to, say, http://www.ttu.ee/it/PatHayes.html, and start using that in my formalism and software. After learning about the existence of http://www.ihmc.us/users/phayes/PatHayes.html, I'd probably want to incorporate other formalisms and systems into my own. The only real/proper way to do this would be stating the synonymy as honest first order equality: http://www.ttu.ee/it/PatHayes.html=http://www.ihmc.us/users/phayes/PatHayes.html First, this would be correct _only_ in case when both URI-s above - are not treated as strings (these two URI strings are obviously different) - are not treated as web pages (the referred web pages as strings will - probably - be also different) Additionally, we will need machinery for handling equality. In simple cases like synonymous URIs the decidability of formalisms can be preserved even in the presence of equality, hence there are no strong reasons for dismissing the equality on purely practical grounds. Since the synonyms of the above kind are - obviously - extremely widespread, should not there be a standardised mechanism for equality axioms in the sem web world? For example, extending rdfs to to rdfse by adding the rdfse:equal predicate with a classical semantics? Once we start using the rdfse:equal and state things like http://www.ihmc.us/users/phayes/PatHayes.html rdfse:equal http://www.ttu.ee/it/PatHayes.html we might sometimes want a way to state that the _actual html strings_ referred to by the URI's are equal. For example, if I'd copy the original Pat page to http://www.ttu.ee/it/copyOfOriginalPageOfPatHayes.html, I might want to state something like rdfse:urlpagestring(http://www.ttu.ee/it/PatHayes.html)=rdfse:urlpagestring(http://www.ihmc.us/users/phayes/PatHayes.html) or at least (http://www.ttu.ee/it/PatHayes.html rdfse:urlpagestring ?X) & (http://www.ihmc.us/users/phayes/PatHayes.html rdfse:urlpagestring ?Y) => ?X rdfse:equal ?Y The presence of rdfse:urlpagestring and rdfse:equal in a language would make it easier to avoid the permanent confusion between the actual Pat Hayes and his web page. We could even axiomatize the difference of objects and their web page strings as forall ?X. not(?X rdfse:equal rdfse:urlpagestring(?X)) formally "solving" the question once and for all :) Tanel Tammet
Received on Wednesday, 13 June 2007 05:09:02 UTC