Sets as singularities in the intensional universe (Q911577)

This paper sets out to reconstruct set theory within an attribute theory based on a relevant logic in much the same way that arithmetic is reconstructed out of set theory. The underlying logic is RQ with the rule gamma (material detachment) as well as machinery for a version of relativized quantifiers. The language includes identity and a special predicate to represent an attribute as being an instance of an attribute, and special postulates are given to govern these. Within this scheme sets are to be identified with ``extensional'' attributes. An attribute x is extensional (Ext(x)) just in case for any z, z's being an instance of x is relevantly equivalent to \(z=y\) for some y an instance of x. With this in mind, axioms are added to specify the attributes in question as being well-founded, and to postulate a (unique) attribute V of ``being a set''. (Roughly, x is a set just in case it is a collection of sets.) It is then shown that V is not itself extensional. Finally, a detailed proof is given of the major result of this paper, namely, that relevant implication collapses to material implication when quantifiers are restricted to sets. As the author puts it, this shows that the ``classical universe'' is contained within the ``intensional universe''. But, as he admits, the addition of the rule gamma gave the ``relevant'' system a decidedly classical slant from the start. He then suggests that set theories with the unrestricted comprehension axiom but based on weak relevant logics will yield a more satisfying solution to the program set out in this paper. Such set theories have been studied by \textit{R. T. Brady} [Notre Dame J. Formal Logic 24, 431-449 (1983; Zbl 0488.03026); Paraconsistent logic, Essays on the inconsistent, 437-471 (1989; Zbl 0691.03038)].