The Ontological argument in neoclassical context: Reply to friedman

Abstract

In his article, 'Necessity and the Ontological Argument',* Professor Fried? man correctly identifies the two major assumptions of Hartshorne's on? tological argument: (PI) It is possible that God exists (perfection is not impossible); (P2) it is necessary that if God exists, God necessarily exists (Anselm's Principle; perfection could not exist contingently). The basic argument of Friedman's paper is that, while Hartsthorne's ontological proof is formally valid in Lewis' S5 system of modal logic, it is not known a priori to be sound. In general, Hartshorne has not been "sufficiently sensitive to the various senses of necessity" (Friedman, p. 307), and hence his argument is based on "acute modal confusion" (p.307). More specifi? cally, for any single notion of de dicto necessity, Friedman contends, either one of the five axioms of S5 or one of the premises of the ontological argument is not known to be true. Thus Friedman concludes that Harts? horne's modal reconstruction of the ontological argument fails, because "there is no single interpretation of necessity which would give us an a priori known sound ontological argument and a priori known sound S5 (or finite fragment thereof)" (p. 324). I am convinced that Friedman's challenge can be met by examining and defending the neoclassical context within which Hartshorne reconstructs Anselm's famous proof. In my opionion, Hartshorne's chief contributions to the contemporary discussion of the ontological argument are two: (1) the substitution of the abstract-concrete (dipolar) understanding of God for the monopolar classical conception of God, and (2) the insistence on a coextensiveness between necessity de re and necessity de dicto in the proof. Both contributions are directly derived from Hartshorne's meta? physics of temporal process. In this reply, I will concentrate on the two sections of Friedman's paper that deal with logical necessity in the Carnapian and in the Kripkean sen? ses. Friedman has shown very well how the axioms of S5 can be known