A contextualist account of sufficient evidence

William James's argument in ‘The will to believe’ is often read in something like the following way: James is defending belief without sufficient evidence, so he must be arguing that in some cases when epistemic reasons are lacking, we can substitute some other kind of reason, such as the potential benefits of believing or even a strong desire that something be true. Stephen Nathanson, for example, says that James is treating belief choices as action choices and arguing, contrary to Clifford, that not only evidence for truth, but ‘the good results of belief’ can count as legitimate reasons for believing.Footnote ¹ Richard Gale says that James is defending belief on insufficient evidence as morally permissible by appealing to ‘the morally desirable consequences of theistic belief’.Footnote ² In philosophy of religion texts James is routinely discussed under headings such as ‘pragmatic justifications of religious belief’,Footnote ³ and his defence of the role of passion is typically read as an alternative to the use of reason and evidence.Footnote ⁴

In contrast to accounts that represent James as substituting non-epistemic reasons for inadequate epistemic reasons, I understand him to be offering a contextualist account of the kind of evidence needed for responsible believing. On my sympathetic and constructive reading of James, the question of whether someone's evidence for a proposition is good enough to permit belief or a belief-like cognitive attitude depends on factors such as how important it is to decide whether or not to act on the basis of the proposition and the extent to which further investigation or deliberation can be expected substantially to improve the individual's evidential situation. In other words practical considerations are relevant to determining what level of evidence is sufficient.Footnote ⁵

The account of James I want to contest misrepresents both the evidential situation he is addressing and the role of practical considerations in his argument. First, the claim that James is defending belief without sufficient evidence is misleading. James undoubtedly wants to argue for the permissibility of belief in the absence of a certain kind of evidence. He takes Clifford's requirement of ‘sufficient evidence’ to extend standards of hypothesis confirmation that are appropriate in a scientific context to believing in general. James thinks that scientific standards are too restrictive when applied to the full range of belief.Footnote ⁶ His primary argumentative response is to shift the focus to practical contexts where refraining from judgement until a fairly demanding evidential standard is met can be in tension with the exercise of practical rationality and also questionable as an epistemic strategy. He argues that under specified conditions it is permissible to pursue an epistemic strategy that is more tolerant of the risk of error than Clifford allows. But it is implausible to view a less risk-aversive strategy as a legitimate way of seeking truth unless the permitted reasons for belief are reasons for thinking that something is true. So we need to conceive of the Jamesian believer as believing on the basis of epistemic reasons, even if those reasons are insufficient to meet Clifford's evidential standard.

Second, while James clearly takes practical considerations to be relevant, it is not that factors such as momentousness constitute reasons for belief. Rather the practical importance of a belief is part of a description of a context in which a potential believer's pursuit of epistemic ends is appropriately constrained by considerations that we can disregard in purely theoretical contexts. James thinks that the level of evidential support we need depends on whether waiting for better evidence is a viable strategy for an agent with both practical and epistemic concerns. Suspending judgement until a particular kind of verification is achieved makes sense in science, he allows, as a strategy for seeking truth and avoiding error in the long run, but he claims that the calculation changes when we move to practical contexts where we do not have the luxury of waiting for the long run.Footnote ⁷ Epistemic grounds that might be insufficient when there is no pressing need to decide can be good enough in a context where we need to commit ourselves to a course of action.

What kind of norms is involved in such a contextualist account of sufficient evidence? Since James is replying to Clifford, who condemns believing without sufficient evidence as immoral, we should surely expect that James is dealing with standards that one might be blamed for failing meet. The most plausible candidate, I think, is in the neighbourhood of our everyday assessments of responsible believing. Richard Foley has contrasted such everyday assessments with those involved in the idealized philosophical concept of epistemic rationality.Footnote ⁸ When we think about what is epistemically rational, we typically imagine an agent whose pursuit of epistemic ends is unconstrained by factors such as whether learning the truth about a particular matter is worth the time and effort or whether continued investigation or deliberation will be likely to lead to a decisive outcome. But in our everyday assessments, when we judge that one should not have believed something (or should not have acted on a belief), we recognize considerations such as our need to prioritize the limited time we have to investigate or deliberate as relevant to assessing blame. Some issues are too trivial to be worth our epistemic efforts. In other cases our grounds may be adequate for the issue at hand, even if we could have tested the matter more thoroughly. For instance, sometimes it is responsible to accept something on testimony that might have been checked, or to operate on the basis of what could be challenged, but seems plausible. In other words, judgements of blame with regard to believing are subject to defence on the grounds that particular activities promoting our cognitive aims are not required for an agent whose pursuit of epistemic ends occurs within a larger practical context.

Norms of responsible believing can be thought of as a species of moral norms, which apply to us in our role as epistemic agents who form beliefs and use them as a basis for action. It has often been noted that there is something odd about speaking of an ethics of belief, since beliefs are not generally voluntarily chosen acts. For the most part, our control over them is indirect. However, believing can be blameworthy when it arises from failures to conduct oneself in ways that are proper for an epistemic agent. For instance, one might show a lack of concern for truth in failing to distinguish reliable and unreliable sources of information. Or one might fail to employ standard ways of checking in cases where they should have been used.

In contrast to thinking about what would be epistemically rational to believe for someone who had no practical concerns, applying norms of responsible believing depends on judgements about how our pursuit of epistemic ends fits with our practical concerns. We surely wouldn't take as an ideal of responsible believing someone who indiscriminately pursued truths without concern for their importance or triviality or who was overly cautious about avoiding error in cases where suspending judgement risked losing out on something important. Our judgements of responsible and irresponsible conduct related to believing presuppose a context within which epistemic efforts can be evaluated.

Clifford proposes a rule for believing that he takes to be independent of context: that we should believe only when we have sufficient evidence. However, such a rule provides little guidance unless we can say what sort of evidence is sufficient for the matter at hand, and that assessment does not seem independent of context. Generally, we can say that when the consequences of being wrong about a proposition are very bad, one needs fairly strong evidence to believe the proposition (or if believing is permissible, to act on the basis of the proposition). Clifford himself appeals to this kind of intuition in his ship-owner case.Footnote ⁹ He imagines a ship-owner as arriving at his belief through a process that disregards relevant evidential considerations. But even if we imagine the ship-owner appropriately reaching the conclusion that the ship is safe for passenger travel on the basis of evidence he possesses, we might judge that he is responsible for acquiring better evidence. Factors relevant to judging that a fairly high evidential standard is called for include both the potentially disastrous results of being wrong and expectations that arise from the social role he occupies.

Even when someone has evidence strong enough to justify believing a proposition, practical considerations can be such that it is blameworthy to take the proposition as a basis for action. When the principal of a high school believes on good evidence that there is no bomb in the school and that the latest bomb threat is another hoax, she may, nevertheless, lack good reason to take the belief as a basis for action, since the consequences of being wrong are so great. James Montmarquet has contrasted what he calls purely epistemic justification where the focus is on whether one has evidence sufficient for knowledge with practical epistemic justification where the issue is whether one has evidence sufficient for acting on a belief.Footnote ¹⁰ There is dispute among contemporary philosophers about whether cases such as the bomb threat show that pragmatic considerations are relevant to whether one has knowledge,Footnote ¹¹ but high-stakes cases such as this one clearly show their relevance for practical epistemic justification. In such cases assessing whether an agent has been epistemically responsible requires us to consider the practical import of being right or being wrong.

One might propose to understand James as claiming that there are cases in which the asymmetrical consequences in situations like the bomb threat are reversed, so that acting on a proposition risks little while refraining from acting risks a significant loss. Just as it makes sense to raise the evidential standards in cases where relying on a belief may have unfortunate consequences, it might be claimed that we should lower the evidential standards in cases where the potential losses from not believing are major, whereas the potential gains from avoiding error are minor.Footnote ¹² However, while James does imagine his potential believer as hoping for a significant good, he does not argue that the costs of being wrong are insignificant. Instead he talks about choosing your own risk in cases where some kind of risk is unavoidable.

So I will understand James's argument to apply to cases where we run a significant risk either in acting on a proposition or in refraining from doing so, yet we have no viable alternative to acting or refraining. I take James's position to be that when (1) there is a pressing need to decide whether or not to act on the basis of some proposition, (2) further epistemic efforts are unlikely to significantly improve one's evidential situation, and (3) one finds the epistemic reasons supporting the proposition convincing, norms of responsible belief do not prohibit believing and acting on the basis of evidence that convinces you. In other words, under specified conditions it is permissible to take a non-sceptical attitude towards evidence that convinces you, even though it would not be unreasonable to adopt a more sceptical attitude and you don't have the kind of evidence that would convince someone who does take a more sceptical attitude.Footnote ¹³

Robert O'Connell has made a persuasive case that the focus of James's concern in ‘The will to believe’ and other essays in popular philosophy from the same era is with world-view beliefs.Footnote ¹⁴ James claims that beliefs on matters such as free will have significant practical import and that there is often no alternative to acting as if a particular world-view belief were true or acting as if it were false. He also holds that we cannot expect to discover the kind of evidence that will be convincing to all reasonable people, but we will often after reflection find a particular view more plausible than available alternatives. The judgement that a particular view is the most plausible one will depend in part on what James calls passionate responses that shape what we take to be evidentially significant.Footnote ¹⁵ While we cannot simply presume that a perspective that is partially constituted by passion-based components results in a correct reading of the evidence, James contends that there is nothing improper in this kind of case about taking the way things seem from our own perspective to be an adequate basis for belief and action.

In the second section I develop an account of the nature of the evidential situation James is considering and apply this account to what he calls the religious hypothesis. In the third section I describe how James argues that the practical context is relevant to the type of evidential support needed. In the final section I consider several objections to the argument I have attributed to James.

Evidence and intellectual undecidability

Near the beginning of ‘The will to believe’ James says that he will be defending religious faith. More specifically, he defends ‘our right to adopt a believing attitude in religious matters, in spite of the fact that our merely logical intellect may not have been coerced’.Footnote ¹⁶ If we take this description as a clue to the evidential situation James has in mind, his use of the terms ‘coerced’ and ‘merely logical intellect’ is suggestive. James imagines approaching the question of what our evidence supports in a way that excludes any involvement of what he elsewhere in the essay refers to as passional responses. The evidence collected and understood in this way is insufficient to compel assent.

Even though James adopts for purposes of argument the passion-free characterization of evidence that he takes to represent Clifford's view, he is suspicious, at least for some kinds of questions, of the idea of excluding all passional elements from the assessment of evidence. In a discussion of philosophical views in another essay he calls the idea of removing any subjective factors an ‘absurd abstraction’.Footnote ¹⁷ His point is not just that the task is difficult, but that the attempt to remove all subjective factors limits our ability to discern truth. Later in the ‘The will to believe’ James suggests that particular passional responses are necessary for recognizing some kinds of truths. If we understand ‘evidence’ to include any truth-conducive reasons, James regards some of what we can become aware of only from a perspective that includes passional elements as evidence. Hence, someone might lack the kind of evidence Clifford requires, but have evidence in this broader sense of the term.

A possible objection to my account is that even if passional elements functioned as external causes of gaining true beliefs, their role in producing true belief would not show that they are internally accessible reasons.Footnote ¹⁸ So James might hold that passion is needed for attaining some truths while still denying that the passional believer had evidence. However, this alternative does not fit well with James's insistence on the importance of instinctive reactions that go deep in human nature in ruling out particular hypotheses. Some views, he claims, are appropriately thought of as irrational when they don't provide a certain kind of subjective satisfaction, and he holds that ‘we are entitled to suppose that the more rational one is the truer of the two.’Footnote ¹⁹ William Wainwright is successful, I think, in showing that James's understanding of what constitutes a good epistemic reason is broader than what contemporary philosophers typically recognize.Footnote ²⁰ James regards some of the intuitions and convictions that are rooted in human emotional reactions as reliable indicators of truth, even if they do not coerce the ‘merely logical intellect’.

Presumably, coercive evidence would leave no logical alternative to accepting a proposition. James's claim that he is defending belief in cases where evidence is not coercive does not precisely specify what sort of evidence one might have. The absence of coercive evidence is compatible with a wide range of evidential situations. However, it will become evident in my subsequent discussion that James is primarily concerned with cases in which what seems apparent to an individual is influenced by features of that individual's perspective whose epistemic value can be questioned, but have contributed to an inclination to believe. The issue is whether to permit or override the inclination.

When stating his thesis, James says that the kind of case in which he defends his permission to believe ‘cannot by its nature be decided on intellectual grounds'.Footnote ²¹ Some interpreters have taken this intellectual undecidability to mean simply that there is not enough evidence to reach a conclusion on the matter.Footnote ²² However, the phrase ‘by its nature’ suggests not merely that a certain kind of evidence is lacking, but that something in the nature of the case precludes resolving the matter in a way that excludes any non-intellectual responses. In ‘The will to believe’ there are two distinguishable types of examples that show different ways something might be intellectually undecidable by its nature.

In one kind of case the existence of a desired state of affairs depends on someone's action. For example, someone's liking you depends on whether you act in a way conducive to producing that kind of response. James posits that the requisite action involves a kind of faith in a particular outcome that at the time of the action exceeds one's available evidence. The matter is intellectually undecidable at a particular point in time because acting on what is not yet supported by evidence is needed to produce the relevant evidence. Late in the essay James suggests the relevance of personal relations situations when he refers to the specific form of the religious hypothesis that represents ultimate reality as personal,Footnote ²³ and some interpreters have taken this sort of example as a key to understanding James's argument about the religious hypothesis.Footnote ²⁴ However, I think that James's fundamental argument for the permissibility of believing the religious hypothesis is stronger if we take the evidential situation he is discussing to parallel his other kind of example. I agree with O'Connell that while James confuses these ‘outcome cases’ with world-view cases in ‘The will to believe’ and other writings, he later gains a clarity about the difference that we can retrospectively apply to interpreting the essay.Footnote ²⁵

The other kind of example of how a purely intellectual resolution is impossible is one in which the significance or weight of the evidence depends on the attitude with which we consider what seems apparent. James's central example (prior to the religious hypothesis) of this type of case is belief in ‘a world of moral reality’.Footnote ²⁶ But earlier in the essay he cites belief in truth and the human capacity to discover it as an example of a belief that depends on a passionate response. James calls this belief a ‘passionate affirmation of desire, in which our social system backs us up’.Footnote ²⁷ That the belief is not grounded in purely intellectual considerations, he says, is shown by our inability to reply to the ‘pyrrhonistic sceptic’. James goes on to say that the dispute with the sceptic is a matter of ‘one volition against another’ and that ‘we are willing to go in for life upon a trust or assumption which he, for his part, does not care to make’.Footnote ²⁸ What is noteworthy here is that from an ordinary non-sceptical point of view we would not say that we lack evidence for belief in truth. The evidential deficiency arises only when we take the non-sceptical viewpoint we are inclined to adopt to require justification over against a sceptical alternative.

In the later discussion of whether to accept the existence of moral truth that is not merely a projection of human preferences or to treat our inclinations to believe in such a thing as ‘only odd biological phenomena’, James says that in the sphere of comparing worths or making judgements of what is good, we rely on what Pascal called the heart. He famously claims, ‘If your heart does not want a world of moral reality, your head will assuredly never make you believe in one’.Footnote ²⁹ In other words our ability to recognize truths in this sphere, if there are such truths, depends on responses rooted in our non-intellectual nature that we can either trust or distrust. James's remarks parallel what he said earlier about belief in truth generally. In fact he explicitly links the two: ‘Moral skepticism can no more be refuted or proved by logic than intellectual skepticism can.’Footnote ³⁰ In both cases James posits that the possibility of a sceptical response precludes arriving at a purely intellectual resolution of the matter. He claims that whether we adopt the sceptical response or the more trusting response to our inclination to believe in a moral reality depends on ‘our whole nature’.Footnote ³¹

Whether the parallel between these other examples and the religious hypothesis holds up depends on how we construe the religious hypothesis. It is often identified with the claim that God exists. However, James makes clear that what he is discussing is something ‘very generic and broad’.Footnote ³² He intends to characterize something that distinguishes religious outlooks from non-religious ones and might be developed in either a theistic or non-theistic direction. James prefaces his characterization of the religious hypothesis by claiming, ‘Science says things are; morality says some things are better than other things; and religion says essentially two things.’Footnote ³³ So if science deals with facts and morality with values, we might expect that religion will have something to do with relating facts and values. James's statement of the religious hypothesis in ‘The will to believe’ is far from clear. However, we can find interpretive help in other writings in which James offers accounts of the essence of religion. One of his best-known accounts is in The Varieties of Religious Experience where he characterizes religion in its ‘broadest and most general terms' as claiming ‘that there is an unseen order, and that our supreme good lies in harmoniously adjusting ourselves thereto’.Footnote ³⁴ James sometimes calls the unseen order a ‘higher universe’ that is friendly to our ideals.Footnote ³⁵ In ‘Faith and the right to believe’ he says that the kind of faith he defends is ‘a greeting of our whole nature to a kind of world conceived as well adapted to that nature’.Footnote ³⁶

James is apparently positing that religious views assert the existence of an ultimate reality that makes possible a harmony between our value aspirations and the nature of things. Different religions offer different accounts of what human fulfilment consists of and how it is achieved, but the various accounts presume that we can make sense of reality, not merely as a set of facts, but in terms congenial to our reflective judgements of value. Is there evidence for the truth of such a presumption? If we take belief in religious reality to be parallel to belief in moral reality, what seems evidentially significant to us will depend on whether we permit the operation of human inclinations to view things morally or religiously, rather than adopting a sceptical attitude towards those inclinations. Someone who does not overrule the inclination to structure reality in moral terms will ordinarily have what seems like an awareness of moral truth and, hence, experiential reasons for confidence in what James calls moral reality. Similarly, someone who follows human inclinations to construe reality in a way that allows for the fulfilment of fundamental value aspirations and finds a particular religious account of how this might occur plausible may also acquire experiential reasons for confidence. But in neither case will these reasons be the kind of evidence that would satisfy someone who approaches the idea of a moral reality or a religious reality with a more sceptical attitude.

It is important to recognize that James is not defending belief in the religious hypothesis for someone who finds it implausible, but only for the individual who finds the hypothesis live. While some have understood James's genuine options as cases in which a hypothesis and its denial are equally live, James nowhere characterizes things this way.Footnote ³⁷ In one of his letters James says that he was defending a right to believe for someone who either already believed or was strongly inclined to believe, but found the belief challenged.Footnote ³⁸ The believer or potential believer, I think, should be imagined as being in an epistemic situation parallel to the moral case where something seems well supported evidentially from a perspective that person is inclined to trust, even though what seems apparent from this point of view can be doubted.

To clarify how one might have evidential reasons to believe, even though the reasons could be doubted, consider one of the analogies that James uses for a different purpose: to illustrate a kind of risk in suspending judgement about the religious hypothesis. James compares waiting for more evidence before deciding about religion to waiting to decide about whether to get married:

The most plausible way of imagining the case is not to think that the man has no evidence for what James quaintly calls the angel-possibility. We can assume I think that the man has enough experience with the woman to give some indication of whether or not she is a good marriage prospect for him. He believes various things about her from personal observation and testimony. On particular occasions he has experienced behaviours that seem to reveal her character and personality traits. He has interpreted times when the relationship has gone well or when tensions have arisen as having various degrees of significance or insignificance. Nevertheless, whatever reasons he has for confidence that the marriage will be a good one, he can view things with a more suspicious eye. Construing the signs as evidence of the desirability of the marriage will depend to some extent on his willingness to trust some of what seems apparent to him, even though in more sceptical moments he may question his own construals.

Similarly, the person for whom James's religious hypothesis is live should not be imagined as having no evidence. Presumably, it is someone who has at least experimented with construing things in accordance with some elaboration of the religious hypothesis and found a resonance with the ways of thinking, valuing, and acting that account provides. Such an individual might be able to point to particular facts or events or experiences that seem illuminating when they are interpreted in religious terms, while finding available non-religious alternatives to be less persuasive. Discovering narratives of others who have followed the same religious path may provide additional assurance. The point of view from which all this seems evidentially significant is subject to doubt,Footnote ⁴⁰ but as in the marriage case, the issue may turn on how sceptical to be about what seems apparent from a point of view one is inclined to trust.

In the earlier discussion of moral truth James had said that our capacity to recognize worths depends on ‘what Pascal calls our heart’.Footnote ⁴¹ The person who permits the operation of this feature of human nature becomes aware of what he or she takes to be moral truth. But that person may also stand back from the perspective in which moral reality seems convincing and question whether human inclinations that lead to this sort of belief are reliable. James thinks that when we do stand back, we have no way to demonstrate that the sceptical possibility is mistaken. We can either trust our capacities for recognizing moral truth or distance ourselves from them. One who does trust the experience of moral reality ‘in the inarticulate heart … clings to it that he is not a dupe’.Footnote ⁴²

It is not obvious, however, that the sceptical attitude in this case is epistemically superior to the more trusting attitude. In discussing the religious hypothesis James says that a rule requiring us to veto the operation of features of our nature that are our only means for recognizing potential truths of some type would be an ‘irrational rule’.Footnote ⁴³ James is not claiming that adopting the sceptical perspective and suspending judgement about the religious hypothesis is irrational, but he is claiming that it is a choice of one kind of epistemic risk over another that cannot be required on purely epistemic grounds. However, he also holds that thinking about these risks within a larger practical context may give us reasons to prefer one epistemic strategy over the other.

Forced and momentous options

James claims that because there is rarely a pressing need to act on the basis of a scientific conclusion, we can adopt an epistemic strategy that puts a priority on avoiding error. He characterizes scientific methodology as expressing what he calls nervousness about falling into error:

James cites Clifford in this context as one who would decline to accept a truth unless it could receive the sort of verification scientific method calls for.

In a different essay James calls the command of science to refrain from believing what has not been verified in the specified way a ‘prudential rule intended to maximize our right thinking and minimize our errors in the long run.’Footnote ⁴⁵ He says that even though we lose some truths by obeying the rule, ‘on the whole we are safer if we follow it consistently’.Footnote ⁴⁶ This long-term epistemic strategy, however, becomes problematic when we come to questions of religious faith because the individual does not have the kind of long run we can posit for scientific investigation. What James calls forced and momentous options put the strategy of suspending judgement until decisive evidence is available in a different light.

James's account of options is imprecise.Footnote ⁴⁷ He initially characterizes an option as a ‘decision between two hypotheses' where ‘hypothesis’ means a potential object of belief. But his first two examples of unforced options include a choice between alternative actions (going out with or without an umbrella) and alternative attitudes towards someone (loving or hating). The third example of an unforced option – Call my theory true or call it false – is on the surface a choice of actions, but the actions could be viewed as reflecting alternative cognitive attitudes. Even so, the case does not involve two hypotheses, unless we consider a hypothesis and its denial two hypotheses. The initial formula for a forced option is ‘Either accept this truth or go without it’.Footnote ⁴⁸ Again, we seem to have alternative cognitive attitudes toward a single hypothesis, and James's subsequent discussion confirms that his focus is on cases where the issue is whether or not to accept a particular hypothesis.

I think that the imprecision about whether we are dealing with alternative beliefs or alternative actions is partially explained by the way James thinks about the relation of beliefs to actions. In another essay he describes the mind as a ‘teleological mechanism’ that ‘functions exclusively for the sake of ends’.Footnote ⁴⁹ Hence, he understands cognitive states in relation to their action outputs. He says in multiple places that actions are the measure of belief, and his concept of liveness is connected to both action and belief. Near the beginning of ‘The will to believe’ James says that deadness and liveness in relation to an individual ‘are measured by his willingness to act. The maximum of liveness in a hypothesis means willingness to act irrevocably. Practically, that means belief; but there is some believing tendency wherever there is willingness to act at all.’Footnote ⁵⁰ Because James thinks of the cognitive states that he calls beliefs and believing tendencies in relation to actions, he can alternatively describe an option between two actions as an option between cognitive states that would make the different actions intelligible.

James's initial characterization of a forced option is that the alternatives offer ‘no possibility of not choosing’.Footnote ⁵¹ In other words even a decision not to choose will be effectively a choice of one alternative over the other. It is easy to see what he means in relation to actions. Consider the action choice James later uses as an illustration: the choice of whether or not to get married to a particular individual. One either gets married or doesn't get married. Not being able to decide is simply one way of not getting married.

If we look at beliefs in isolation from actions, it might seem as if there is no equivalent forced option of belief.Footnote ⁵² You can believe that marrying a particular individual is a good idea or you can disbelieve it, but suspending judgement means neither believing nor disbelieving. However, suppose we think about the question of what to believe in the context of a forced option of what to do, such as whether to get married. In this case one either acquires a cognitive state that makes the action of getting married intelligible or one fails to acquire such a state. Not acquiring the cognitive state needed sensibly to decide to get married by suspending judgement is the practical equivalent of deciding against getting married. So I propose to understand James's forced options of belief as cases in which one has a forced choice to perform or not perform some action and performing the action presupposes either belief or a belief-like state that would make the action intelligible.

Forcedness by itself does not create any urgent need to decide, but when a forced option is also momentous, there is a cost to suspending judgement. James says that because scientific questions are rarely of a momentous nature, ‘Such options are not forced on us.’Footnote ⁵³ We can follow the strategy of waiting in ‘sceptical balance’ until the evidence is sufficient to remove reasonable doubt. But moving to matters of significant practical import, James asks rhetorically whether we can always ‘wait with impunity till the coercive evidence shall have arrived’.Footnote ⁵⁴ Although suspending judgement allows us to avoid error, in cases of forced and momentous options of belief, James holds that it commits us to acting as if we disbelieved, hence, losing the opportunity for what might be a vital good.

James says explicitly that the correlation he presumes between beliefs and actions is crucial to his argument. In a footnote to his discussion of the religious hypothesis he says, ‘Since belief is measured by action, he who forbids us to believe religion to be true, necessarily forbids us to act as we should if we did believe it to be true. The whole defence of religious faith hinges on action.’Footnote ⁵⁵ The relevant forced choice of action is whether or not to live a religious life, and James posits that living a religious life involves accepting the religious hypothesis. He thinks that either disbelief or suspense of judgement will mean not living a religious life. Speaking for those who are inclined to believe, he says that a command to suspend judgement until better evidence is available would mean ‘that we shall put a stopper on our heart, instincts, and courage, and wait – acting of course meanwhile more or less as if religion were not true’.Footnote ⁵⁶ It seems clear that James thinks that it would be an unacceptable restriction on practical agency in the case of an intellectually undecidable genuine option to have to act as if the view that seems most convincing from your own perspective is untrue. Hence, a restriction on belief that has this consequence must also be unacceptable.

In addition to this argument about how a requirement to suspend judgement would be in tension with the exercise of practical rationality, James also questions the idea that suspending judgement is the only viable epistemic strategy. One who permits the inclination to believe to produce belief or a belief-like attitude can be viewed as risking error in the hope of gaining truth, whereas someone who follows the strategy of suspending judgement might be viewed as having a greater concern to avoid error. One reason for not interpreting James as defending belief on non-epistemic grounds when epistemic reasons are inadequate is that it makes his epistemic argument puzzling. It would ordinarily be poor epistemic strategy to believe in the absence of reasons that you take to be truth-conducive, even if doing so might enable you to achieve non-epistemic aims. But James wants to argue that the error-reducing strategy and the truth-seeking strategy are legitimate alternative ways of pursuing epistemic ends. To make sense of this claim, we need to recognize that the factors he calls passional or non-intellectual function not in isolation from epistemic reasons, but in producing a perspective that shapes one's view of what is evidentially significant. In seeking truth about matters that are intellectually undecidable by their nature, such as world-view issues, James thinks that we become convinced only from perspectives that include passion-based components. Suspending judgement when we cannot decide from a perspective that excludes passions means vetoing our tendency to form beliefs about such matters, but in doing so, James suggests that we risk blocking those elements of our nature that are needed to recognize some truths.Footnote ⁵⁷ Whether we trust or distrust non-intellectual aspects of our nature, there is an epistemic risk, and the degree of risk that is permissible cannot be decided by epistemic considerations alone.

The question of how risk aversive to be with regard to error correlates with the question of when evidence is sufficient to permit belief. If avoiding error is much more important than gaining truth, we can set the evidential standard fairly high. If some degree of error is tolerable, we can lower the evidential standard. But what is the proper balance?Footnote ⁵⁸ James's insight, I think, is that the answer depends on the practical context in which epistemic ends are pursued. When there is no pressing need to decide, we can give priority to the concern to avoid error and utilize a strict evidential standard. But when there is a pressing need to decide and no reasonable expectation of improving our epistemic situation, we can treat evidence that convinces us as sufficient for belief.

James's permission to believe does not apply to someone who is not convinced one way or the other. Nor is his term ‘will’ an endorsement of believing through effort or willpower.Footnote ⁵⁹ Rather it is a way of directing attention to the precursive influence of passion in determining what is plausible for us.Footnote ⁶⁰ Furthermore, as I have interpreted James's argument, the permission is limited to cases where some subjective condition may be needed to interpret the evidence correctly and there is no alternative to following or overruling an inclination to take the way things seem to be as a reliable indication of the way things are. James claims that in these kinds of cases individuals are entitled to trust their own lights.Footnote ⁶¹

Objections

One kind of objection that might be raised against the argument I have attributed to James is that he illicitly moves from the justifiability of acting as if some proposition were true to the acceptability of believing the proposition. We might agree with James that it is permissible under certain circumstances to act on the basis of the view you find convincing, but note that action does not require belief. In cases where we lack evidence, but must decide how to act, we can often act on the basis of what we presume, while refraining from belief. Suppose, for example, that a relative of yours who has often been in financial trouble asks you for a loan. Over the years his track record in paying back loans to family members has been spotty, but he claims to have reformed and asks you to trust him. In considering the situation, you decide that you don't have a good basis for believing either that he will pay you back or that he won't. You might, nevertheless, have reason to act on the presumption that he will. Similarly, someone might suspend judgement about whether the religious hypothesis is true, but for various non-epistemic reasons decide to act as if it were true.

If my account of James is correct, this objection distorts the evidential situation he is considering. Admittedly, there are cases in which one lacks any reason to make a judgement about truth or falsity of a proposition, yet has reason to act on the presumption of its truth. In the case of the relative we can imagine various non-epistemic reasons for giving the person the benefit of the doubt. But James's potential believer is faced with a different situation. On my account, she assesses her evidence from a perspective that convinces her of the truth of a proposition and inclines her to believe it, even though lacking the kind of evidence that would convince someone who takes a more sceptical attitude. The issue is whether the more sceptical attitude is required for epistemic integrity, and James argues that under the conditions he specifies, it is not. The forced option he appeals to could be described as whether to treat the inclination to believe as reliable or to distrust it. In the case of someone who seems to be aware of ‘a world of moral reality’, it would make sense to act as if there were such a reality if the apparent apprehension is to be trusted, but suspending judgement would mean treating the inclination to believe as untrustworthy. The forced choice of belief occurs within the context of a forced choice of action, but it is in deciding whether to trust an inclination to believe that one decides how to act.

Furthermore, the idea that we can suspend judgement, but nevertheless act as if something were true, is problematic for some kinds of action. When the relevant action is living a religious life, we are not talking about a simple performance like going to the market, where one's state of mind is only peripherally related to the act. Living a religious life ordinarily involves internalizing ways of thinking and feeling that make possible spontaneous actions in accordance with the way of life and experiences that cohere with it. One could conceivably begin a religious way of life by performing overt actions that imitate what a religious person would do, but at some point, continuing the life will involve either acquiring the kind of internal states that are needed for fully entering into it, or retaining a status akin to being a participant-observer. It is at least unclear whether the potential ‘vital good’ that James posits might be obtained through the kind of religious life available to the person who maintains a critical distance from the ways of thinking and feeling that typically motivate this way of life. Yet an intention to continue to suspend judgement about the beliefs that might make this way of living intelligible would mean suppressing any inclinations to believe that might make such a way of life more than going through the motions.

Another objection to James might call attention to the meagreness of the religious hypothesis.Footnote ⁶² Finding the religious hypothesis credible is a long way from finding a specific religion credible, and most people who are convinced by some actual religion do not even consider the question at the level of abstraction described by James. So what does permission to believe the religious hypothesis have to do with actual religious belief?

There is merit to this objection, and I can only speculate about how James might respond. It seems clear that the order of actual belief is not typically accepting something like the religious hypothesis and then finding a particular religion convincing. So perhaps we can take James to be claiming that someone who is convinced by a particular religion might at a reflective level consider an underlying assumption that is involved in finding the religion plausible. Establishing the permissibility of believing the religious hypothesis would allow the reflective believer to narrow the range of possibilities to religious options.

Conceivably, someone who finds the religious hypothesis convincing might find no particular religion believable. James himself seems to fall into this category. However, James wants to defend belief for people who are convinced by a particular religion. In fact he apparently extends this permission to believe to ‘any hypothesis that is live enough to tempt’ one.Footnote ⁶³ That extension seems overly generous if we are talking about responsible believing. A great many things that people are tempted to believe about religious matters are incredible when considered from the point of view of what we know or should know. James could deal with the problem by emphasizing that while liveness is a subjective category, intellectual undecidability is an objective category that puts limits on the range of responsible belief. He could hold that some sets of religious belief can be ruled out on evidential grounds, but there remain versions of particular religions that are candidates for belief by reasonable and informed people.

To apply James's argument to the acceptance of a particular religion, we need to imagine a case where a religious or functionally equivalent non-religious view, which makes a particular way of life intelligible, cannot be ruled out on evidential grounds and after due reflection seems more plausible to someone than any alternative. In cases where we will inevitably act as if the view we find most plausible were true or act as if it were false and further investigation or deliberation will not improve the evidential situation, James does not prohibit adopting a sceptical attitude towards our own judgement of truth, but he defends taking a non-sceptical attitude in such a case as a permissible epistemic strategy.

The approach I have ascribed to James represents him as responding to an evidentialist challenge to religious belief without denying the fundamental evidentialist requirement of evidence for proper belief. I suspect that any viable form of evidentialism will need to allow evidence as adequate in some cases where it can still be reasonably contested. For example, the evidentialist needs to posit that we have evidence for beliefs even though the evidence is not sufficient to rule out some sceptical hypotheses, such as believing that we are brains in a vat. What the contextualist account I have attributed to James provides is a way of specifying when evidence that doesn't remove the possibility of reasonable doubt can nevertheless be good enough to permit belief. It is a proposal for a form of evidentialism that leaves room for some kinds of religious belief.