W3C home > Mailing lists > Public > semantic-web@w3.org > July 2007

Re: literals are resources too

From: Garret Wilson <garret@globalmentor.com>
Date: Fri, 27 Jul 2007 07:26:46 -0700
Message-ID: <46AA00A6.8020502@globalmentor.com>
To: David Powell <djpowell@djpowell.net>
CC: "<>" <_@whats-your.name>, Semantic Web <semantic-web@w3.org>

David Powell wrote:
>   
>>>       not true). I'd like to see something like <xsd:integer
>>>       rdf:literalAbout="123"/>, where "123" plays an analogous role to
>>>       rdf:about. This is the same thing has saying <rdf:Description
>>>       rdf:type="xsd:integer" rdf:literalAbout="123"/>. Suddenly we get
>>>       
>
>   
>> im a bit unclear (perhaps since i don't know rdf/xml). what is the URI for the literal
>> resource. is it a blank node? or is the URI just '123' ?
>>     

RDF isn't really sure about that, which is why I say that currently 
literals have a quasi-resource status. RDF will talk about how, "no, 
really, literals are resources," but they don't have URIs, they can't 
take properties, and they sure as heck don't look like blank nodes on a 
graph. They look like literal nodes, whatever those are.

>
> If we wanted to change the RDF model, I could imagine getting rid of
> literals from RDF triples, and then reintroducing them as an attribute
> (not an RDF property) of resources - like a sort of intrinsic
> value for the resource.
>   

That's sort of what I was saying, which agrees with my point that, for 
the integer 123, the string "123" is *not* the resource. You can also 
think about many programming languages (such as Java) in which all 
objects have a toString() method that generates a string representation 
(although usually this isn't guaranteed to be unique and is for 
informational purposes only).

What we want to do, I think, is have normal resources with normal types 
(such as xsd:integer or whatever) that somehow get initialized with a 
lexical representation, but that lexical representation is then 
subordinate to the actual resource (unlike the current situation, in 
which the lexical representation lives on in some quasi-resource state). 
I wouldn't mind if you could get the lexical representation back, 
through maybe a rdf:lexicalRepresentation property, as David was 
alluding to.

So maybe the thing to do is to define the integer value 123 using 
something like <xsd:integer rdf:literalAbout="123"/>, which creates a 
resource identified by the URI <rdfliteral:123;xsd:integer> with the 
intrinsic rdf:lexicalForm property of "123". (I use rdf:literalAbout, 
not because I like the name, but to show that its purpose is analogous 
to rdf:about.)

Garret
Received on Friday, 27 July 2007 14:26:59 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:47:26 UTC