Meta-Epistemology

When doing research for my revised essay paper on the nature of truth and epistemic justification, I read a book by Michael Levine called The Possibility of Religious Knowledge: Causation, Coherentism, and Foundationalism. In it, he made a few interesting points which I have at times thought myself but could not put into words. I will try to do so now:

The first point he made regards epistemic relativism, the possibility of two “equally valid” systems of epistemic justification:

“Relativism would be the view that there are different and equally valid methods of reasoning that could be used to justify epistemic judgments (i.e. that there are multiple structures of epistemic justification)…What may be justified on one foundation may not be justified on another, but then how can we ever know if we are really justified in any of our beliefs? …To choose between foundations we would have to rely on some third foundation…and so on. Ultimately we could never be justified in our choice of a foundation. We would be left with an arbitrary decision…the problem is that knowledge would itself be impossible because knowledge cannot be based on an arbitrary decision. We cannot decide to know something simply by specifying a foundation for our knowledge…” (pgs. pgs. 269-270).

In the same way that two hypothetically equally coherent systems would require an arbitrary choice between them if coherentism is true, if there is more than one “valid” method of epistemic justification, then any choice between them would be arbitrary. If we chose one over against the other on the basis of some supra-criterion – if I could legitimately argue believing one method over against another is reasonable – that itself would become the basis of all epistemic justification and mitigate against the idea there really are two “equally valid” methods of epistemic justification. Two “equal” systems or methods of epistemic justification would, as Levine notes and Gordon Clark noted in A Christian View of Men and Things (pg. 34), lead to arbitrariness and, therefore, skepticism. In other words, these hypotheticals must be just that: hypotheticals. They are useful inasmuch as they incline us to think about and address these epistemological issues, but they cannot be substantial enough to justify skepticism.

The second meta-epistemological point Levine makes is as follows:

“…in epistemology once the meta-epistemologist analyzes what is for him the normative sense (e.g. a coherentist account), then that normative position will be presupposed… Therefore, the meta-epistemic presuppositions or analyses have to be dealt with in considering the consequences of a normative analysis for what one can be said to know…” (pg. 260)

This is especially true for one who holds to a sort of foundationalism or belief in first principles. One’s first principle is his first principle. By definition, a first principle cannot be the conclusion of a prior premise. This is not to say that a first principle cannot be shown to be reasonable by inspection of its epistemic explanatory power, its self-attestation, and the consistency amongst the propositions which follow from it.

What Levine means is that when one is evaluating opposing first principles, definitions of truth, or methods of epistemic justification, he cannot pretend that he is evaluating them abstract from his background as a skeptic, agnostic, coherentist, pragmatist, empiricist, rationalist, Scripturalist, etc. Even the meta-epistemological statements I have made throughout this post have been informed by my own epistemological background. Fortunately, my epistemological background is such that I can apply apagogic argumentation to falsify alternative positions in a non-question-begging fashion.