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

Re: How to name what you get back? was: Terminology Question concerning Web Architecture and Linked Data

From: Bernard Vatant <bernard.vatant@mondeca.com>
Date: Fri, 27 Jul 2007 10:22:51 +0200
Message-ID: <46A9AB5B.2050004@mondeca.com>
To: Pat Hayes <phayes@ihmc.us>
Cc: www-tag@w3.org, semantic-web@w3.org, 'Linking Open Data' <linking-open-data@simile.mit.edu>

Pat and all
>> Sandro,
>> You should read it well: Chris speaks about "the description.......that
>> identifies....". That is not necessarily a name.
> True, but Sando's point is that this description is returned when you 
> use a particular name (URI) to access the thing identified. If you use 
> a different name of the very same thing, you may well get a different 
> description returned. So the description is OF the thing identified, 
> but is ASSOCIATED with the name, in the sense that only this name may 
> get this particular description returned to you by HTTP.
Is not all this conversation boiling down to the well-known situation of 
"semiotic triangle" [1]. I've published a while ago a representation of 
that at [2] as illustration of this post [3], for those who can read 
French. For those who can't, here it is in a nutshell :

The elements of the triangle are as following
    The "signifiant" (signifier) is the URI
    The "signifié" (signified) is the the 
    The "référent" (referent) is the thing beyond all its descriptions 
(whatever that may be)
Granted, using "signified" in the Web context will not help much ... But 
it's just to show that what is discussed here at length seems not really 
new stuff for linguistics.
> What bothers me is the idea that a description can be said to be 
> 'identifying', which suggests (especially in this context where URIs 
> are called 'identifiers') that it contains enough information to 
> uniquely identify the resource, which is likely to often not be the case.
I agree, but description is often made to identify in a given 
identification context. Identification without context does not exist.
>>  I can give you the
>> description:
>> 1) is a guy
>> 2) sits at the bar
>> 3) wears a hat
>> etc
>> and that identifies a person of which I do not know the name. We call 
>> that
>> descriptive identification.
>> As I said, many things don't even have a name.
> True, but then they will never have such a description returned.
Not using URI and direct http protocol. But such descriptions can help 
to find matching names (URIs) can be returned by a SPARQL endpoint if 
the above description is formalized as a SPARL query. So names can 
retrieve descriptions, and descriptions can resolve to names. Llike in 
natural conversation : the above description is made by John (who does 
not know the guy at the bar) to Peter, who knows the name, and answers : 
"You mean Pat?". No big deal indeed. :-)


[1] http://en.wikipedia.org/wiki/Sign_%28linguistics%29
[2] http://perso.orange.fr/universimmedia/geo/trsemiotic.gif


*Bernard Vatant
*Knowledge Engineering
*3, cité Nollez 75018 Paris France
Web:    www.mondeca.com <http://www.mondeca.com>
Tel:       +33 (0) 871 488 459
Mail:     bernard.vatant@mondeca.com <mailto:bernard.vatant@mondeca.com>
Blog:    Leçons de Choses <http://mondeca.wordpress.com/>
Received on Friday, 27 July 2007 08:22:59 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 07:41:58 UTC