Can RDF say anything about anything?
Can RDF say anything about anything? The RDF documents are contradictory on this point. The Primer indicates that RDF can be used to let anyone ``say anything they want about existing resources'' with no exception for the resources used by RDF. Concepts says that ``RDF is an open-world framework that allows anyone to make simple assertions about anything''. However, Concepts also says that ``Certain URIs are reserved for use by RDF, and may not be used for any purpose not sanctioned the RDF specifications.'' What is the situation here?
See: http://lists.w3.org/Archives/Public/www-rdf-comments/2003JanMar/0148.html
There is an apparent contradiction surrounding: "Can RDF say anything about anything"?
The phrase "Say anything about anything" is broken. (I think "Anyone can say something about anything" is closer to the intended goal here.) However, this phase does not appear in the last-call version of the concepts document. The document wording should be revised to avoid such misleading suggestions.See: http://lists.w3.org/Archives/Public/www-rdf-comments/2003JanMar/0175.html
This message offers some examples of the contradiction.
I think the first two examples given are legal, and see no problem with them. I think the third example is not legal, because it uses a reserved URI inappropriately. But I don't see this illegality as being at odds with the idea that "Anyone can say something about anything". I think the final example offered is syntactically legal, but unsatisfiable because the RDF specifications don't define any use of rdfs:Class as a property. So I think the text about what is sanctioned by RDF may require some clarification: some rdf/rdfs URIs are reserved by RDF for specific syntactic purposes, and may not be used to denote resources (e.g. rdf:ID); other rdf/rdfs URIs are reserved to identify specific concepts, for which the corresponding resource is constrained by the RDF specifications, and may not be given an interpretation that is at variance with them.See: http://lists.w3.org/Archives/Public/www-rdf-comments/2003JanMar/0193.html
See: http://lists.w3.org/Archives/Public/www-rdf-comments/2003JanMar/0194.html
See: http://www.w3.org/TR/rdf-syntax-grammar/#section-grammar-summary
More clarifying discussion:
I think the position concerning rdf:ID, etc, in RDF/XML is clear from the syntax spec, section 7.1, which excludes 'coreSyntaxTerms' from 'nodeURIs', etc. I acknowledge it is less clear in the Concepts document and that this should be clarified. Concerning the response to TC2: I think there is a difference between "saying something about anything", and "saying something using any token". The "anything" here is any resource for which a URI can be created. Again, I think the abstract syntax should note that certain URIs are reserved from being used in the subject, predicate or object position.