According to an increasingly-influential
interpretation of Gottlob Frege's work, the Fregean conception of
logic is one on which "metatheory" as a whole must be ruled
unintelligible. Because Frege takes logic to be "universal," it is
argued, he cannot make sense of the attempt to evaluate logical
systems themselves, and hence cannot make sense of e.g. standard
questions of the soundness, completeness, and consistency of
logical systems. It has recently been argued that Frege's clear
antipathy towards independence-proofs in geometry (as seen both in
his rejection of David Hilbert's independence-proofs, and in his
rejection of his own tentatively-proposed independence-proof
technique of 1906) is a mark of this rejection of metatheory, and
should be taken to support the "anti-metatheory" interpretation of
Frege.
The central purpose of this paper is to argue that neither
Frege's "universalism" with respect to logic, nor his rejection of
various independence-proof techniques, gives us any reason to view
his conception of logic as anti-metatheoretical. The importance of
this point, as I see it, is that the sense in which Frege takes
logic to be "universal" is a sense in which we all ought to agree
with him, and it is crucial that we understand what is, and what is
not, "ruled out" by this universalist conception.