Re: lightweight reification

At 10:18 AM 6/2/00 +0200, Pierre-Antoine CHAMPIN wrote:
>Graham Klyne wrote:
> > But reification described in M&S involves making a node resource for the
> > statement, AND adding four graph arcs originating from that node.  If we
> > can find a usable identity for the statement resource, then the four
> > additional arcs aren't necessarily needed.
>
>a reified statement has intrinsically these 4 properties (property in the 
>general meaning, not RDF's), and the arcs are the way RDF expresses 
>properties !
>
>This does not mean that implementation must keep those arcs explicit (I 
>agree there are better ways of doing it), but it's true that M&S may be 
>misleading about it.

But, without a (common) way to _name_ a statement, the property values are 
the only way to identify it -- which leads back to full reification.

#g

------------
Graham Klyne
(GK@ACM.ORG)