On 11/5/14 3:10 PM, Karen Coyle wrote: >> If the graph union includes the ontology, and the sparql query >> includes "?S ?P :Work" , you /will/ get bf:workTitle rdfs:domain >> :Work . > > I was speaking of the case in which bf:workTitle rdfs:domain bf:Title. I always need to work these things out looking at actual triples, and thanks to insomnia, this is what I came up with last night: <X> rdf:type bf:Work . <X> bf:workTitle _:aa . _:aa rdf:type bf:Title . _:aa bf:titleValue "Here's my title" . So SPARQL query something like: SELECT ?s ?p ?o Where { ?s a bf:Work } Would you not get: <X> rdf:type bf:Work . <X> bf:workTitle _:aa . If the bf:workTitle were of type bf:Work instead of bf:Title, you would get: <X> rdf:type bf:Work . <X> bf:workTitle _:aa . _:aa rdf:type bf:Work . _:aa bf:titleValue "Here's my title" . Does that clear it up? And I wish to gawd that there was some easier to use software that you could throw these kinds of tests to. I might be able to create a file to test it out in Protege but it's a big pain. However, having such a test environment would make it easy to try things out and see the consequences. kd -- Karen Coyle [log in to unmask] http://kcoyle.net m: +1-510-435-8234 skype: kcoylenet/+1-510-984-3600