William James's ‘Will to believe’ essay originated as lectures delivered to philosophy clubs of Yale and Brown. First published in 1896, James's essay received much early criticism, but is now a staple of introductory textbooks in philosophy. John Bishop's important book is a rigorous retooling and defence of James's contention that one can ‘adopt a believing attitude in religious matters, in spite of the fact that our merely logical intellect may not have been coerced’ (William James ‘The will to believe’, in The Will to Believe and Other Essays in Popular Philosophy (New York NY: Dover, 1956), 1–2). Bishop's concern is with what he calls moderate fideism, the thesis that it is permissible, morally and intellectually, in certain circumstances to take as true a proposition in one's practical reasoning, even while acknowledging that the proposition lacks adequate evidential support.

While moderate fideism goes beyond the evidence, it does not ignore the evidence. A more extreme fideism would advocate accepting a proposition (or in Bishop's terminology, taking a proposition as true) even when there's a preponderance of evidence arrayed against that proposition. But when the evidence speaks the moderate fideist listens. When the evidence is silent, however, the moderate fideist is not content to suspend all epistemic action. With this understanding James was a moderate fideist, as too, Bishop argues, are the proponents of Reformed epistemology (97). Indeed, as we will see, if Bishop is correct, we are all fideists of one kind or another.

James had W. K. Clifford for his foil; Bishop aims at ‘hard-line moral evidentialism’. The hard-line moral evidentialist holds that accepting (or taking as true) any theistic proposition lacking adequate evidential support is epistemically and morally impermissible. Clifford is properly classified as a hard-line evidentialist, with his declaration that ‘it is wrong always, everywhere, and for any one, to believe anything upon insufficient evidence’. Bishop demurs by championing a permissibility principle, which he calls (J+):

Principle (J+) is the heart of the book. Clauses (i) and (ii) of (J+) are the Jamesian legacy inherited by Bishop, with (iii) and (iv) added by Bishop, seeking to avoid the charge that moderate fideism enables the acceptance of immoral positions. Along with James, Bishop understands a genuine option as involving a decision to take as true (or in James's case, believe) a proposition the truth of which is left open by the evidence, where the decision is forced, living, and momentous.

Whether James in fact intended (ii) with its essential undecidability is debatable. The cost to pay for insisting on essentiality may be high – how would one know that a particular question is in principle forever open, with no chance of the evidence closing it? One answer Bishop supplies here is an inductive appeal, ‘if one is prepared to agree that human history until now has not managed to resolve the question of God's existence, then surely one should also agree that it is more likely than not that this question necessarily cannot be resolved in the evidence?’ (140–141). But thinking about recent innovations in the philosophy of religion including the fine-tuning version of the design argument, and the argument from divine silence, may make one hesitant to agree with Bishop that it is surely more likely than not that the question of God is in principle unanswerable by any evidence. And in any case, it is doubtful that an inductive appeal provides strong support for showing necessity.

Another answer Bishop provides is that ‘it seems clear that the truth of a framing principle would be essentially evidentially undecidable’ (141). A framing principle is any proposition the truth of which is presupposed by all other propositions within a specified domain or context (or framework). The proposition that God exists would be a framing principle for theism. Does justificatory digging within a framework (what some call a ‘worldview’) require that a presuppositional bedrock always limits the depth of the dig? Many thinkers have held so, most prominently Wittgenstein in his On Certainty. If this line of thought is correct and if no credible epistemic coherentist project is available, all worldviews at their foundations would require a non-evidential commitment for allegiance. Fideism of a kind would be unavoidable.

Clauses (iii) and (iv) share the task of shielding the moderate fideist from a charge that moving away from strict adherence to evidentialism opens the door to morally problematic alternatives. The alleged threat here involves faith commitments, Bishops says, to deities like ‘Nazi gods’ (165). But is (iv) really necessary? Why wouldn't morally problematic alternatives be considered false on that score? If we have a sufficient grasp of correct morality, as required by (iv), wouldn't that provides us with good reason to reject any thing found incompatible as false?

Although (J+) is designed as a moral permissibility condition, Bishop seeks to defend not just the moral permissibility of moderate fideism, but that it is epistemically permissible as well. An externalist himself, Bishop surveys several strategies that might produce this result, but concludes that the standard arguments provide neither the fideist nor the evidentialist with a non-question-begging argument for or against the epistemic permissibility of moderate fideism (or, for that matter, evidentialism). But since it is far from clear that moderate fideism flows out of an improper exercise of one's epistemic capacities, and since there are some moral considerations that favour moderate fideism, Bishop suggests that, all in all, it deserves an endorsement over hard-line evidentialism.

The first of these moral considerations is that a modest Jamesian fideism is a more balanced stance than is hard-line evidentialism, with regard to harmonizing the two inclinations to adhere to evidence, and to accept a worldview the truth of which cannot be evidentially established (220). Modest fideism acknowledges and makes room for both inclinations, while hard-line evidentialism sees only the first. The second moral consideration is that hard-line evidentialism is doctrinaire and dogmatic in its affiliation with naturalism. As no framework (or worldview) can be evidentially supported all the way down to its foundational premises – there are always presuppositions beyond the reach of the evidence – even the hard-line evidentialist makes a passional commitment in aligning with naturalism (and does so permissibly, Bishops argues, only when in conformity with (J+)). But then to deny another the same right to align with theism via (J+), is doctrinaire and dogmatic, a moral failing, Bishop argues (220). The third of the moral considerations Bishops surveys is, I think, the most promising – certain theistic propositions make sense of the hope that acting morally, even when doing so is imprudent, makes sense. As Bishop puts it ‘seeking to live in accordance with morality is not a pointless goal’ (226). The hard-line evidentialist, while allowing a commitment to basic moral claims, would prevent any acceptance or commitment to those theistic propositions that support the practical hope that our sacrifice and effort in the cause of morality makes a real difference.

If Bishop is correct that the three considerations break the impasse between moderate fideism and hard-line evidentialism in favour of the former, then a significant gap in Bishop's argument is closed, as there is no discussion of the argument from divine silence found in Believing by Faith. The argument from divine silence contends that evidential ambiguity is not what we would expect if a perfectly loving being were to exist, and so evidential ambiguity supports naturalism over theism. The argument from divine silence proceeds only with a significant dose of hard-line evidentialism assumed. However, if there is good reason to hold that moderate fideism is morally preferable over hard-line evidentialism, then the fact of evidential ambiguity (if it is a fact) would not support naturalism, and the argument from divine silence is denied a foothold.

The first chapter includes a helpful glossary defining the stipulated terms that figure significantly in Bishop's discussion, and an overview laying out the direction and landmarks of the eight chapters that follow. The justification of religious belief is also briefly discussed in chapter 1. The second chapter takes a look at justification and specifies the kind most relevant for reflective theists – whether their theistic beliefs are morally permissible. The usual philosophical focus is on epistemic justification, and Bishop returns to epistemic justification in chapters 7 and 8, but his main focus is with the moral status of moderate fideism. In the second chapter Bishop also takes on the issue of doxastic voluntarism, and introduces the idea of taking a proposition to be true – investing an evidentially unsupported proposition with full weight in one's practical deliberations.

The ambiguity thesis is introduced in chapter 3. According to the ambiguity thesis, neither theism nor naturalism enjoys the clear support of a preponderance of the evidence. moreover, the ambiguity thesis asserts that both a theistic interpretation and a naturalistic one of the totality of the evidence are viable. Neither stands beyond the other basking in the glow of evidence if the ambiguity thesis is true. The fourth chapter examines two alternative responses to evidential ambiguity. One is what Bishop calls isolationist epistemology. Isolationism asserts that theistic propositions can be evaluated only by principles internal to theism itself. There are no neutral, external standards which can relevantly evaluate theism. Bishop contends that the internal principles appealed to by isolationists themselves are in need of something like (J+). So isolationist epistemologies do not present a rival to moderate fideism. Likewise, Bishop contends, with Reformed epistemology, since it at best provides the theist with a justification contingent on God existing. This conditional result will not suffice for the reflective theist, Bishop believes, so Reformed epistemology likewise needs (J+).

Chapter 5 introduces the idea of a doxastic venture. One makes a doxastic venture when one takes to be true a proposition that one believes, while recognizing that one lacks sufficient supportive evidence. The nature of Christian faith is also discussed here. Chapter 6 is primarily a discussion of James's idea of a genuine option, and the understanding Bishop puts on it. Restriction and limitation are dominant goals of chapter 7, as Bishop tightens the scope of (J+) with the introduction of clauses (iii) and (iv) to the Jamesian (i) and (ii). Much of the argument in support of (J+) is defensive as various objections are examined and defused in the chapters to this point. With chapter 8 Bishop seeks a positive case in support of (J+). He examines various possible supports of (J+), ranging from the claim that hard-line evidentialism is self-undermining to a possible employment of Pascal's wager. None of the seven or so possible supports clearly and successfully favours moderate fideism over hard-line evidentialism, and Bishop suggests that something of an impasse exists in which neither is favoured. In the last chapter Bishop introduces the three moral considerations mentioned above that he contends sway support toward moderate fideism and away from hard-line evidentialism.

Believing by Faith with its original and interesting argument is a significant entry in the debate over the ethics of belief. Anyone interested in James's argument for the right to believe, the debate over evidentialism, or the topic of fideism will find Bishop's book a profitable read. The book would also be appropriate for advanced undergraduates, and for graduate students.