This chapter examines the relationship between skeptical themes and debates concerning the rationality of religious belief. This relationship manifests itself not only in terms of critiques of the rationality of religious belief, but also in prominent defenses of the rationality of religious belief, such as Lockean evidentialism, reformed epistemology, and skeptical fideism. The discussion puts particular focus on a specific form of skeptical fideism that employs an epistemic parity argument with regard to the epistemological status of religious and everyday belief. In this regard, it considers quasi-fideism, an account of the rationality of religious belief that applies a Wittgensteinian hinge epistemology to the religious case, and which has a natural affinity with skeptical fideism.

Quasi-Fideism in Outline

In a series of recent papers, I have argued that we should understand the kind of hinge epistemology that arises from Wittgenstein's final notebooks – published as On Certainty (Wittgenstein 1969) – as suggesting a distinctive kind of approach to the rationality of religious belief. In particular, on this interpretation, Wittgenstein is not offering the straight fideistic line in this regard that is usually attributed to him, but rather a very different position, which I call quasi-fideism. On this proposal, while Wittgenstein is claiming that religious conviction, at its root, is to be understood along arational lines (this is the fideistic part), he is not thereby arguing that all religious belief is to be understood arationally; indeed, one's non-fundamental religious beliefs may well on this view enjoy a positive rational standing (which is why the view is not a straightforward version of fideism). Moreover, the idea is that religious belief is no different from ordinary non-religious belief in this regard, in that all belief, religious or otherwise, pre-supposes fundamental arational commitments (this is where the hinge epistemology comes in).

As this last remark indicates, quasi-fideism is effectively the application of hinge epistemology to the problem of the rationality of religious belief, as we shall explain in a moment. But, first, let us briefly remark on the nature of quasi-fideism and its attractions. Part of the appeal of quasi-fideism lies in its relative merits when compared to other accounts of the rationality of religious belief. Straight fideism can seem like a capitulation in this regard, in that rather than offering a positive account of the epistemology of religious belief, it instead contends that such belief is unique (or at least unusual) in not being subjected to rational evaluation. In contrast, the kind of evidentialist treatments of the rationality of religious belief that have been popular historically (at least since Locke), such that (roughly) one should proportion one's religious conviction to the degree of evidential support one has for it, seem to set the bar for religious belief so high as to be unattainable (particularly given the strength of religious conviction and the usual constraints of what counts as applicable ‘evidence’ in this regard).

This chapter explores two influential conceptions of the role of common sense in philosophical theorizing from early analytical philosophy, due to G. E. Moore and Ludwig Wittgenstein. Both approaches set out an important function for our everyday certainties to play in the epistemological enterprise, albeit in very different ways. For Moore, our common-sense certainties serve as a kind of reasonable stopping point in philosophical disputes. In particular, where common sense confronts philosophical theory, we can reasonably side with common sense. While Moore claims that our common-sense certainties have an epistemic weight simply in virtue of being common-sense certainties, for Wittgenstein the certainty that attaches to these commitments entails that they have no rational status at all. Nonetheless, this doesn’t prevent them from having a crucial import to epistemological questions. By setting these two philosophical approaches side by side, we gain an important perspective on how common sense might be appealed to in philosophical theorizing.

Support is canvassed for a new way of approaching some core epistemic issues: anti-risk epistemology. It is explained how anti-risk epistemology differs from anti-luck epistemology by examining some of the subtle (but epistemologically significant) differences between the notions of luck and risk. It is argued that anti-risk epistemology, while essentially an adaption of anti-luck epistemology, can nonetheless resolve some motivational issues that face the latter proposal. In the process, it can provide other important benefits, such as enabling a broader range of epistemic assessments, including capturing the complexity of some important epistemic assessments involving collaborative inquiry. Our ultimate concern, however, is to examine how anti-risk epistemology fits into a wider virtue-theoretic account of knowledge, one that replaces anti-luck virtue epistemology with anti-risk virtue epistemology. As we will see, the latter proposal inherits all the strengths of the former but none of its flaws. It is also better placed to explain why knowledge is never compatible with unsafe belief, and to provide us with a diagnostic handle on the path taken by post-Gettier epistemology.

Although there are some obvious exceptions, for a good part of the first half of the twentieth century the general philosophical attitude toward the problem of radical skepticism was largely dismissive. Whether for broadly antirealist/logical positivist reasons, or for broadly Wittgensteinian reasons, this puzzle was widely thought to be a pseudo-problem, trading on dubious metaphysical claims.1 The second half of the twentieth century, and up to the present day, however, has seen this problem taken seriously once more. Indeed, we have witnessed this philosophical difficulty again not only seriously generating academic discussion, but also being a driver for new proposals within epistemology.

Putnam's BIV argument

A familiar way of arguing for radical skepticism is by appeal to radical skeptical hypotheses, such as the hypothesis that one might be a brain in a vat (BIV) which is being radically, and undetectably, deceived about its environment. Roughly, the skeptical argument goes that since such skeptical hypotheses are by their nature indistinguishable from normal experience, so one cannot know that they are false. Furthermore, if one cannot know that they are false, then it follows that one can't know much of what one believes, most of which is inconsistent with radical skeptical hypotheses.

This last step will almost certainly require some sort of closure-style principle, whereby knowledge is closed under known entailments. Thus, if one does have knowledge of the ‘everyday’ propositions which one takes oneself to know (e.g. that one has hands), and of the fact that these propositions entail the denials of radical skeptical hypotheses (e.g. the BIV hypothesis, because BIVs don't have hands), then one could come to know that one is not a BIV. Conversely, insofar as one grants that it is impossible to know that one is not a BIV, then it follows that one cannot know the everyday propositions which are known to be inconsistent with the BIV hypothesis either.

There are many ways of responding to radical skepticism of this form, which we will refer to as BIV skepticism. One might deny the relevant closure principle, for example, or one might put forward an epistemology according to which one could know the denials of radical skeptical hypotheses, and so on. On the face of it, Hilary Putnam's (1981b) famous argument against BIV skepticism would appear to be a variation on this last anti-skeptical strategy, in that he also seems to be, effectively, claiming that one can know that one is not a BIV.

The parallels between these two anti-skeptical approaches are superficial, however, and the differences significant. In particular, Putnam's overarching goal is not to make the epistemological claim that such anti-skeptical knowledge is possible, but rather to motivate the semantic claim that one cannot truly think the thought that one is a BIV. But these claims are logically distinct, in both directions. That one can know the denial of the BIV hypothesis is obviously consistent with the possibility that one can truly think the thought that one is a BIV.