Re: language entailment


I already mentionned this could be a tricky can of worm. I think I just
saw the fist worms crawling out, as this entailment does not seem to
interact very well with other existing entailments.

Assume we have

  :a :p "chat"@en-GB . # (1)

that entails

  :a :p "chat"@en . # (2)

Worm 1: Consider combining it with OWL entailment, and that you also have

  :p a owl:FunctionalProperty . # (3)

So (1,3) entails (1,2,3). As "chat"@en-GB and "chat"@en are two distinct
things, this is inconsistent as it violates the functionality of :p.

Worm 2: if we accept my proposal of making language tags a special kind
of datatype, one could write

  :p rdfs:range rdflang:en-GB . # (4)

Here, (1,4) would entail (1,2,4), which in turn would entail

  "chat"@en a rdflang:en-GB . # (5)

and more generaly would make rdflang:en a subtype of rdflang:en-GB
(while the opposite seemed more intuitive). Furthermore, "chat"@en has
no lexical form in rdflang:en-GB, so that looks plain wrong... :-/


On 05/18/2011 12:11 PM, Nathan wrote:
> Steve Harris wrote:
>> On 2011-05-18, at 10:07, Pierre-Antoine Champin wrote:
>>>  :a :b "chat"@en-GB .
>>> entail
>>>  :a :b "chat"@en .
>>> in any entailment regime defined by the RDF semantics ??
>> No idea.
> Can anybody confirm, if not is it worth us defining this, seems like a 
> useful entailment.

Received on Wednesday, 18 May 2011 12:55:06 UTC