Provenance as a first-class citizen

There's been some blogging on Planet RDF lately about provenance being a
first-class citizen in RDF (as Seth Ladd put it, at
(http://www.semergence.com/archives/2006/03/17/02/56/22/), or as John
Barstow wrote
(http://www.nzlinux.org.nz/blogs/2006/03/17/thinking-about-rdf-lite/),
to be able to have provenance without reification.

I wanted to point out that this is merely an issue of serialization
convenience, and not an issue with RDF per se, in the bigger picture.
That is, provenance *requires* reification -- if you're going to assert
something arbitrary about a statement, the statement *has* to be denoted
by a Resource.  The question is whether your serialization syntax of
choice lets you do that concisely.

N3, for instance, has formulas.

Maybe that was obvious, but it seemed like there was some conflation
between the RDF/XML syntax and RDF in general.

And, you're always free to interpret the statements in a document not as
global truths but as resources asserted by some other resource.  John
suggests "There’s no reason I can’t have optional who [provenance] and
when parameters that default to “source document” and “now”,
respectively." -- and that seems like a good idea.  That is, we could
formalize that arbitrary RDF documents found on the web should be
treated as
   <> :asserts ( ... ) ; :atTime "time" .
i.e. the document found at a particular time, as a resource, asserts a
set of statements, where 'asserts' means to claim that at that time
those things were true.

-- 
- Joshua Tauberer

http://taubz.for.net

"Unfortunately, we're having this discussion. It's too bad,
because guess who listens to the discussion: the enemy."

Received on Friday, 17 March 2006 14:47:17 UTC