argument is not decisive, Hume is surely getting at something important. We can see
more clearly what it is if we examine the way a similar problem arises for William P.
Alston's epistemology of religious experience.
Alston works within what he describes as a doxastic practice approach to epistemology.
A doxastic practice is a way of forming beliefs and subjecting them to epistemic
evaluation in terms of a background system of beliefs that furnish potential defeaters or
overriders. Alston argues that it is practically rational to engage in socially established
doxastic practices that are not demonstrably unreliable or otherwise disqualified for
rational acceptance. In the religious sphere, he views mystical perception as a kind of
religious experience in which there occurs a presentation or appearance to the subject of
something the subject identifies as the Ultimate. Applying the notion of doxastic
practices to mystical perception, he urges us to suppose that there are different socially
established mystical practices in diverse religions because there are wide divergences in
their overrider systems of background beliefs. Christian mystical practice (CMP) is one
such practice. Alston argues persuasively that it is not demonstrably unreliable. However,
both the outputs of CMP and its overrider system appear to be massively inconsistent
with their counterparts in the mystical practices of other religions. Assuming that the
appearance of conflict cannot be explained away, religious diversity thus gives rise to a
philosophical challenge to the rationality of engaging in CMP or any of its equally well-
established rivals in other religions.
In his book Perceiving God (1991), Alston grants that this challenge is the most difficult
problem for his position that it is rational to engage in CMP. He formulates the problem
in this way. On account of the inconsistency, at most one
end p.396
of the rival mystical practices can be a sufficiently reliable way of forming beliefs about
the Ultimate to be rationally engaged in. But why should one suppose that CMP in
particular is the one that is reliable, if any is? To be sure, CMP can come up with internal
reasons for supposing that it is more reliable than its competitors. But each of them can
do the same. Hence, “if it is to be rational for me to take CMP to be reliable, I will have
to have sufficient independent reasons for supposing that CMP is reliable, or more
reliable or more likely to be reliable, than its alternatives” (269). A cumulative case
argument for the truth of Christianity might provide such independent reasons. Alston
does not try to show, however, that there is a cumulative case for the truth of Christianity
that is decisively superior to the cases that can be made in support of its competitors.
Instead, he elects to proceed in accord with a worst-case scenario in which it is assumed
that there are no independent reasons for perferring CMP to its rivals.
How bad are things for CMP in the worst-case scenario? Alston invites us to look at the
matter in this way. Suppose our sole respectable basis for a positive epistemic evaluation
of CMP were the fact that it is a socially established doxastic practice that has not been
shown to be unreliable. On that assumption, Alston admits, religious diversity would
reduce its epistemic status to an alarming degree. Given the equal social establishment of
several mutually incompatible mystical practices, none of which is demonstrably
unreliable, he concedes that “it is at least arguable that the most reasonable view, even for
a hitherto committed participant of one of the practices, would be that the social