Existential import revisited (Q2522269)

The author proposes to `amend' quantification theory with identity by basing it on the following axioms : A1. \((y) : (x) \cdot Fx \cdot{} \supset Fy\); A2. \((x) \cdot Fx\supset Gx\cdot {} \supset :(x) \cdot F(x){} \cdot\supset \cdot (x)\cdot Gx\); A3. \((\exists x) \cdot Fx \supset Fx\); A4. \(x=x\); and A5. \(x=y\cdot {} \supset {} \cdot Fx \supset Fy\). He allows for the usual rule of substitution, modus ponens, and the rule which enables him to infer \(A \supset (\alpha) B\) from \(A \supset B\) provided \(\alpha\) is not free in \(A\). A system like this, he points out, is valid for every non-empty domain. It is valid for every domain, including the empty one, if A3 is dropped. The advantage of such a system consists, apparently, in that `\(Fx \supset \cdot (\exists y)\cdot Fy\)', which is deemed to be offending, is not deducible in it. No rule for `\((\exists x)\)' is given. In particular, it is not quite clear to the reviewer whether or not the author admits the equivalence `\((\exists x) \cdot Fx \cdot \equiv \cdot \sim (x) \cdot \sim Fx\)'. It appears that the author's treatment of the problem of existential import rests, among others, on the following presuppositions, whose rationale is taken by him for granted: statements of the form `\((x) \cdot Fx\supset Gx\)' are correct translations of the universal affirmative statements of the traditional logic; the result of substituting a non-referential singular narne for `\(y\)' in `\((\exists x) \cdot x = y\)' is false; the result of substituting such a name for `\(y\)' in `\(y = y\)' is true; the quantifier `\((\exists x)\)' is the fundamental means of expressing existence. Page 291, lines 18 and 32, `\((\exists x)\)' should read `\((\exists y)\)'.