On the logic of contingent relevant implication: A conceptual incoherence in the intuitive interpretation of R (Q1119621)

The author offers criticisms, not all of them very convincing, of A. R. Anderson and N. D. Belnap jun.'s system \(\underset \tilde{} R\) considered as a formalizaion of nonlogical conditionals. He then presents his own alternative formalization \(\underset \tilde{} R\underset \tilde{} I\), which is included in \(\underset \tilde{} R\) but weaker. Specifically, many \(\underset \tilde{} R\) theses of the form \(A\to (B\to C)\) are provable in \(\underset \tilde{} R\underset \tilde{} I\) only in the imported form (A\&B)\(\to C\). This includes the transitivity principle (A\(\to B)\&(B\to C)\to (A\to C)\). \(\underset \tilde{} R\underset \tilde{} I\) is given in both an axiomatic and a natural-deduction formulation, the latter without any reiteration rule. \(\underset \tilde{} R\underset \tilde{} I\) is then extended to \(\underset \tilde{} R\underset \tilde{} I^{\square}\), with \(\underset \tilde{} S\underset \tilde{} 4\)-style necessity and a rule for the reiteration of necessities. \(\underset \tilde{} R\underset \tilde{} I^{\square}\) is shown to be equivalent to Anderson and Belnap's system \(\underset \tilde{} E\) of entailment on the obvious translation.