I'd like to put on the radar a seeming inconsistent tetrad regarding cco:doctrinal_definition and cco:definition:
- The annotation property cco:definition is intended to have a unique value
- cco:doctrinal_definition is a sub-relation of cco:definition
- The value of cco:doctrinal_definition will be in most, if not all, cases distinct from the value of cco:definition
- For a given IRI and string "XYZ": if cco:doctrinal_definition is a sub-relation of cco:definition and IRI cco:doctrinal_definition "XYZ" then IRI cco:definition "XYZ"
These cannot all be true at the same time. 1 seems motivated by the annotations associated with cco:definition and cco:doctrinal_definition. 2 and 3 don't need defending. 4 is motivated by, among other things, the OWL2 Direct Semantics.
I will note this point is a bit pedantic as - because annotation properties are ignored by reasoners, etc. - I've not encountered a scenario in which such a conflict would generate a problem. So, I am not at this point suggesting we should take steps to avoid the issue. I do think we should at least be clear about this point in documentation.
As for options, the easiest to give up is 2. Alternatively, this may suggest reason to drop 4, perhaps interpreting annotation sub-relations in terms of bookkeeping.
I'd like to put on the radar a seeming inconsistent tetrad regarding cco:doctrinal_definition and cco:definition:
These cannot all be true at the same time. 1 seems motivated by the annotations associated with cco:definition and cco:doctrinal_definition. 2 and 3 don't need defending. 4 is motivated by, among other things, the OWL2 Direct Semantics.
I will note this point is a bit pedantic as - because annotation properties are ignored by reasoners, etc. - I've not encountered a scenario in which such a conflict would generate a problem. So, I am not at this point suggesting we should take steps to avoid the issue. I do think we should at least be clear about this point in documentation.
As for options, the easiest to give up is 2. Alternatively, this may suggest reason to drop 4, perhaps interpreting annotation sub-relations in terms of bookkeeping.