[DL] Local/global consequence and DL

[Franz Baader]

>   Could anyone tell me whether a TBox is (by default) seen as a set of local
> axioms?

No, TBoxes and the more general GCIs (general concept inclusion axioms)
are gobal. This follows immediately from the definition of the semantics:
In a model I of A = C , you must have A^I = C^I, i.e., for every individual d of
the domain (in modal logic speak: for every world of the Kripke structure),
d in A^I iff s in C^I. This shows that A = C corresponds to a global axiom of
the form a <-> phi.

Best regards,

-Franz Baader