402
PARTISAN REVIEW
Unfortunately, however, the fact that a thesis is dispiriting or scary is not
in itself a reason for supposing it false; so it is necessary to articulate why
these misconceptions
are
misconceptions.
Let me begin by distinguishing two projects in which epistemology
may engage: articulating criteria of what constitutes justification of or
good evidence for a belief, and giving guidelines for the conduct of in–
quiry. (Though these are frequently run together in contemporary epis–
temology, they are, though of course related, as different as criteria for
judging roses are from instructions for growing them.)
One source of the first epistemological counter-culturalist thesis, that
epistemic standards are culture- or community-bound, may have been the
perception that neither of the traditionally rival theories of epistemic jus–
tification - foundationalism and coherentism - is acceptable, and the
conviction that the only alternative is to resort to some kind of contex–
tualism, according to which epistemic justification is context- or com–
munity-bound. Another source has been Kuhn's observation that propo–
nents of rival scientific paradigms disagree even about what constitutes
evidence, which has suggested the parallel conclusion, within the philos–
ophy of science, that epistemic standards are paradigm-bound.
I agree that neither foundationalism nor coherentism will do. Foun–
dationalism founders on its failure to identify a class of basic beliefs which
could plausibly be claimed both to be justified exclusively by experience,
and to be sufficient to support all other justified beliefs; coherentism
founders on its failure, consistently with its claim that justification is a
relation exclusively among beliefs, to acknowledge any role to experi–
ence in empirical justification. But contextualism is not the only, or the
most plausible, alternative. A theory intermediate between foundational–
ism and coherentism, which allows the relevance of experience to empir–
ical justification without requiring any beliefs to be justified by experi–
ence alone, and which allows pervasive mutual support among beliefs
without requiring that empirical justification be a matter exclusively of
relations among beliefs, can avoid the difficulties of both foundationalism
and coherentism. In this intermediate theory (I call it, more accurately
than euphoniously, "foundherentist"), the model of evidential support is
- not, as much recent epistemology has supposed, a mathematical proof
- but a crossword puzzle. The clues are the analogue of experiential evi-
dence, and already-completed intersecting entries the analogue of reasons,
of the subject's background beliefs.
Rather than suggesting the context- or paradigm-dependence of
standards of evidence, this model suggests how deep-seated disagreement
in background beliefs will give rise to diagreements about what evidence
