Re: Literal as subjects [was: Obsoleting URIs [was: URIs and Unique IDs]]

It may be silly but it helps to restrict the types of RDF you expect to 
find,
and also restricts modelling (which can get very silly if you remove 
restrictions).

rdf:value was designed for this case.
http://www.w3.org/TR/rdf-schema/#ch_value

Imagine removing the restricion - it would cause ontology modelers to 
use it,
and then they need guidelines when to use it or not.

And once people start  using it, the problem of semantic meaning would 
occur,
the triple that includes the literal gives meaning,
annotating a plain literal loses the s-p-Literal context in which the 
literal was used.

Example of meaning-less annotation that people may do:
<bob> <isemployee> "true".
"true" <verifiedby> <HumanResourceDepartment>.
(actually, this would call for reification)

s - p - "true" ... what now?

best
Leo

It was McBride, Brian who said at the right time 10.11.2008 08:34 the 
following words:
>>> Yes, if there were one simple thing I could change about the RDF
>>> specification it would be to remove that silly prohibition against
>>> having a literal as the subject of an assertion.
>>>       
>> I think the RDF semantics allows this, just not the RDF/XML syntax.
>>     
> [bwm]
>
> Its risky jumping into the middle of a thread you haven't been following, but I saw this in passing and ...
>
> My understanding of the intent behind the current recommendations produced by the RDFCore working group was to keep the restriction from the 1999 spec that literals could not be subjects, but prepare the ground to allow a future working group to remove this restriction.
>
> Thus the restriction is imposed in the abstract syntax of RDF, but the semantics document was written to accommodate an extension.
>
> I don't have the specs to hand so can't check the details.
>
> Brian
>
>
>
>   
>> In Section 1.4 of http://www.w3.org/TR/rdf-mt/
>> [[
>> if E is a ground triple s p o. then I(E) = true if
>>
>> s, p and o are in V, I(p) is in IP and <I(s),I(o)> is in IEXT(I(p))
>>
>> otherwise I(E)= false.
>> ]]
>>
>> In section 1.3 IEXT is defined as
>>
>> [[
>> A mapping IEXT from IP into the powerset of IR x IR i.e. the set of
>> sets of pairs <x,y> with x and y in IR .
>> ]]
>>
>> IR is defined in the same paragraph as:
>> [[
>> 1. A non-empty set IR of resources, called the domain or universe of I.
>> ]]
>>
>> Ie a relation can be between anything.
>>
>> In RDF/XML you can always get the same effect in the end by using
>> owl:sameAs in the style of this turtle:
>>
>> [] owl:sameAs "http://example/new-term"^^xsd:anyURI .
>>     :obsoleteUri "http://example/old-term"^^xsd:anyURI .
>>
>> Henry
>>     
>
>
>
>   


-- 
____________________________________________________
DI Leo Sauermann       http://www.dfki.de/~sauermann 

Deutsches Forschungszentrum fuer 
Kuenstliche Intelligenz DFKI GmbH
Trippstadter Strasse 122
P.O. Box 2080           Fon:   +49 631 20575-116
D-67663 Kaiserslautern  Fax:   +49 631 20575-102
Germany                 Mail:  leo.sauermann@dfki.de

Geschaeftsfuehrung:
Prof.Dr.Dr.h.c.mult. Wolfgang Wahlster (Vorsitzender)
Dr. Walter Olthoff
Vorsitzender des Aufsichtsrats:
Prof. Dr. h.c. Hans A. Aukes
Amtsgericht Kaiserslautern, HRB 2313
____________________________________________________