>The truth-statement that all truths can be empirically verified is itself empirically unverifiable.

To be fair, if this is stipulated as a definition rather than a truth-claim, then this issue disappears. This was the problem with the "yeah but can verificationism be verified?" objection to Ayer and the positivists: it only works if verificationism is taken as a proposition rather than a definition or criterion of meaning (which is precisely how it was posited, at least by Ayer).

But ironically enough, I think verificationism and this claim about truth fail empirically; it is a demonstrable matter of empirical fact concerning human linguistic practice that we use language to do things other than assert empirical truths, and there are truths and types of knowledge that are meaningful but cannot be empirically verified (truths about the self obtained via introspection, for instance, all truths that are generalizations, poetic/literary truth, etc).