Re: Are we elements or animals? (was: Use of fragment identifiers in XML)

>On Thursday, October 31, 2002, at 10:19 PM, pat hayes wrote:
>>>>RDF documents do not DESCRIBE fragments. They USE them.
>>>
>>>I have a triple:
>>>
>>>ex:John rdf:type <http://www.example.org/#Dog> .
>>>
>>>I grab http://www.example.org/, it's an RDF document that says (in part):
>>>
>>><rdf:Description rdf:about="#Dog">
>>>   <dc:description>a dog, an animal with four legs</dc:description>
>>></rdf:Description>
>>>
>>>According to the URI spec (via the links I cited), the #Dog is an 
>>>XML element.
>>
>>Wait a minute. What are you saying here? Are you DESCRIBING that 
>>syntactic thing that starts with a hash sign and has four ascii 
>>characters, or are you USING it to refer to something?
>
>Well, I'm defining it. That's a form of description, I guess.

Right. You are talking about it, not using it; like when I say. 'Dog' 
has three letters, Im talking about the word; but when I say I don't 
own a dog, I'm using that word to talk about the things with four 
legs.

>
>>If the latter, what language do you take it to be in (and whose 
>>semantic rules you will use to help determine what it refers to)? 
>>Myself, I would use RDF, seeing as it occurs in an RDF document. In 
>>which case, the URI spec is irrelevant, since the entire body of 
>>all URI (and XML) specs ever written do not say anything at all 
>>about what it is that fragIDs must be used to refer to. And in that 
>>case, http://www.example.org/#Dog is a class (of dogs). '#Dog' is 
>>an XML element.
>
>Er, they do. That's what I just pointed out. According to the URI 
>spec (and its references I cited) http://www.example.org/#Dog 
>identifies

"Identifies" is ambiguous. You tell me what you think it means, and 
I'll tell you which of the uses goes with which sense. In one sense, 
an identifier in a piece of program text identifes a datastructure; 
in another sense, it identifies whatever that datastructure is 
supposed to denote (which might be a number, or a process, or a fact 
stored in a database, or whatever). URI specs, the last time I 
looked, didn't say anything much about the second sense, the sense 
Ive been calling 'denotes', the sense that is used in model-theoretic 
semantics.

>the XML element (<rdf:Description 
>rdf:about="#Dog">...</rdf:Description>). And according to the RDF 
>spec that URI identifies a class. Which is it?
>
>Perhaps I'm doing a poor job of explaining my argument; it sounds 
>like I haven't gotten through.

Possibly. I have the same feeling, however.

Pat

-- 
---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola              			(850)202 4440   fax
FL 32501           				(850)291 0667    cell
phayes@ai.uwf.edu	          http://www.coginst.uwf.edu/~phayes
s.pam@ai.uwf.edu   for spam

Received on Friday, 1 November 2002 09:50:40 UTC