On unification and admissible rules in Gabbay-de Jongh logics (Q386639)

The paper studies intermediate propositional logics with the disjunction property, that is, such logics \(x\) that \((A \vee B) \in x\) yields \(A \in x\) or \(B \in X\). A substitution \(\sigma\) unifies a formula \(A\) relative to \(x\) if \(\sigma(A) \in x\). A multiple-conclusion rule (m-rule for short) is an ordered pair of finite sets of formulas \(\Gamma/\Delta\). A rule \(\Gamma/\Delta\) is said to be admissible for a given logic \(x\), if every substitution that unifies all formulas from \(\Gamma\), unifies at least one formula from \(\Delta\). A set of m-rules forms a basis (of admissible rules) of a given logic \(x\), if this set induces the greatest multiple-conclusion consequence relation having \(x\) as its set of theorems. The introduced de Jongh m-rules (which are the modifications of the multiple-conclusion version of the Visser rules) play a very important role in the investigation of the admissible m-rules of the Gabbay-de Jongh logics \(\mathbf{D}_n\). Namely, the main result of the paper is that \((n + 1)\)-th de Jongh rule \(\mathsf{J}_{n+1}\) forms a basis for the admissible \(m\)-rules of \(\mathbf{D}_n\). Moreover, if \(\mathsf{J}_{n+1}\) is admissible in an extension \(x\) of \(\mathbf{D}_n\), then \(\mathsf{J}_{n+1}\) forms a basis of the admissible rules of \(x\) too.