Simplified semantics for basic relevant logics (Q1187984)

Whereas the semantics for the basic (normal) modal system K places no conditions on the binary relation \(R\) -- one adds conditions on \(R\) to model other normal systems such as S4 -- the semantics for the positive, basic (affixing) relevant system \(\text{B}+\) (and its negation extension B) put conditions on the ternary relation \(R\) (and on the unary operation \(*\) in the case of B). One must then add further conditions to model other affixing systems such as T [see \textit{R. Routley} and \textit{R. K. Meyer}, J. Philos. Logic 1, 192-208 (1972; Zbl 0317.02019), and/or \textit{R. Routley}, \textit{V. Plumwood}, \textit{R. K. Meyer} and \textit{R. T. Brady}, Relevant logics and their rivals, Vol. 1 (1982; Zbl 0579.03011)]. This paper declares the intention of balancing the scales, and makes a good beginning by presenting semantics for \(\text{B}+\) (and for some negation extensions) without placing restrictions on \(R\) (and \(*\)). Showing how to model other normal relevant systems by adding appropriate restrictions is left to another time. The new semantics for \(\text{B}+\) simply drops the conditions on \(R\) and the hereditary condition on valuations by changing the interpretation condition for implication formulae at the base world 0 (called `\(g\)' in this article) to: an implicational formula is true at \(g\) iff its consequent is true at every world at which its antecedent is true. This merely stipulates Semantic Entailment (Sement) of \textit{R. Routley} and \textit{R. K. Meyer} [Truth, syntax and modality, Stud. Logic Found. Math. 68, 199-243 (1973; Zbl 0317.02017)]. The extension of the semantics to handle negation yields a pair of interesting results. In the first place, adding the usual Routley-Meyer \(*\) operation without any conditions and with the normal interpretation function for relevant negation does not yield B. Instead, it yields BM which is \(\text{B}+\) together with full DeMorgan equivalences and rule contraposition. One can then get a semantics for B by adding the condition that the starred world of a starred world is simply the original unstarred world. Secondly, handling negation via the American plan [again, see \textit{R. Routley}, \textit{V. Plumwood}, \textit{R. K. Meyer} and \textit{R. T. Brady} (loc. cit.)] with a four-valued valuation function and no star operator yields yet a different system BD, which is \(\text{B}+\) with DeMorgan and double negation. How do model B along these lines remains an open question. Strong consistency and completeness is proved in all of the above cases in more or less the usual fashion for propositional relevant logics.