The coherence theory of empirical knowledge

Abstract

In a paper written for a commemorative symposium on the philosophy of C. I. Lewis, Roderick Firth remarks that Lewis liked to confront his Harvard epistemology students with a fundamental choice between a foundation theory of knowledge based on 'the given', like that advocated so ably in Lewis's own books, and "a coherence theory like that of Bosanquet" .' As Firth notes, there are many different philosophical views which have been called 'coherence theories', including theories of truth and of meaning; but what Lewis seems to have had primarily in mind is a coherence theory of epistemic justification: the view that the epistemic warrant or authority of empirical statements derives entirely from coherence and not at all from any sort of 'foundation'.2 Since Lewis's strong version of foundationism is by now everywhere in eclipse, it seems appropriate to examine the Bosanquetian alternative. The purpose of this paper is to explore, and tentatively defend, a view of the Bosanquetian sort, which I shall call "the coherence theory of empirical knowledge" (hereafter CTEK). As discussed here, the CTEK is not to be identified with any specific historical view, though it has obvious affinities with some. It is intended rather as an idealized reconstruction of a relatively pure coherence theory, one which avoids all versions of foundationism.3 Views like the CTEK, though often employed as dialectical bogeymen, have rarely been treated as serious epistemological alternatives, since they have been thought to be subject to obvious and overwhelming objections. Thus the essential first step in a defense of such a view is to provide a sketch of its overall shape and rationale and show on this basis that these supposedly fatal objections can be answered. Such a preliminary defense of the CTEK, aimed at establishing its epistemological viability, is the goal of this paper.