In this article, I will criticize evidentialism. My criticism will be maintained in the spirit of William James’ earlier critique of it and I will mention his arguments where necessary. It will become clear from my arguments that, in spite of the critique mounted against it in the century after James wrote his famous article ‘The Will to Believe’ (see below, 1.2), evidentialism is still a ‘live’ doctrine. That is to say, in spite of the philosophical critique of evidentialism it is still in high regard as a ‘popular’ doctrine and, as such, has serious ramifications for our culture at large. Among others, it has serious ramifications for the moral domain.

However, quite like James, my main interest in criticizing evidentialism is not of an ethical but of a religious nature. My critique of the consequences evidentialism harbors when pursued as a moral doctrine serves as a means to criticize evidentialism as a tool to attack religion. In my view, the evidentialist critique of religion is utterly mistaken.

I begin by arguing that, depending on what sort of evidentialism we are talking about, it is either rather uninteresting or, else, leads to absurd or imprudent moral consequences under certain circumstances. Hence, it should not be utilized as a universal doctrine. Next, I bring this point to bear upon religious questions. I argue that evidentialism fails for criticizing religion. It is either rather uninteresting for judging religious claims, or, else, it is imprudent or even inappropriate.

Evidentialism has a long history. In the Continental tradition, the notion of ‘evidence’ is used by e.g. Husserl and Brentano. [1] In the English-speaking world, Locke and Hume are probably the classical defenders of evidentialism. [2] Of the more recent philosophers, Bertrand Russell defends a similar position with his ‘true precept of veracity’ that goes as follows: ‘We ought to give to every proposition which we consider as nearly as possible that degree of credence which is warranted by the probability it acquires from the evidence known to us.’ [3]

Furthermore, evidentialist principles not only play a role in reflective, philosophical activities but are also widespread in the public domain. At least in the Western hemisphere, it is frequently held that one ought to believe only that for which one has sufficient evidence. Put more moderately, in Western culture, it is considered to be preferable to believe something which is warranted by evidentialist means than to believe something which is not.

Additionally, evidentialism is a frequently utilized tool for the purposes of criticizing religion. [4] Upon being asked what he would say if, after death, he would be brought into the presence of God, Russell responded: ‘Not enough evidence God! Not enough evidence!’ [5] Evidentialism’s critical potential with regard to religion becomes particularly perspicuous when we consider that being warranted by evidentialist means mostly implies being warranted by sense perception. Our modern Western culture is dominated by the idea that every reasonable theory should be testable, and visual capabilities suggest themselves for the purposes of testing theories. It is widely believed that these capabilities provide us with hard evidence, whereas cognition-mechanisms which are not based upon the senses provide less reliable evidence or even none at all. [6] Since religious claims are usually not based upon the senses, they cannot meet the standards set by this sort of evidentialism. Thus, evidentialism supports the modern critique of religion.

Much of James’ famous article ‘The Will to Believe’ is devoted to rebuking an evidentialist opponent. This opponent accepts the methodological ideal that one should remain skeptical of a claim until one has sufficient evidence to believe in it. [7] As a consequence of this ideal, the opponent remains skeptical with regard to religion and its claims. For example, W.K. Clifford argues that there is simply not enough evidence for believing in religion and that ‘it is wrong always, everywhere, and for anyone to believe anything upon insufficient evidence.’ [8]

For a proper understanding of the nature of the evidentialist principle as it is used by Cliffordians and like-minded people, it is important to note that evidentialism is a universal ideal. Clifford holds that the maxim that is implied in his statement, viz. not to believe anything upon insufficient evidence, should be applied everywhere, in all domains of inquiry, under all circumstances. In other words, the evidentialism that Clifford recommends consists of two components, a maxim to refrain from believing in cases of insufficient evidence plus the claim that this maxim should be applied universally.

Consequently, there are two strategies available for the purpose of refuting evidentialism, viz. to attack the maxim to refrain from believing implied in it and to attack the claim to universality implied in it. I will pursue the second strategy. And for the purposes of attacking the claim to universality implied in it, one successful counterexample is in principle sufficient.

Rorty, however, uses the first strategy. He attacks the evidentialist maxim frontally on the grounds that it is logically dependent upon realism or foundationalism, both of which are obsolete in his eyes. [9] James is more careful and uses the second strategy. He attacks the universalist pretensions implied in evidentialism. He argues that there are certain cases in which it cannot be applied. For example, there are ‘forced options’, cases in which it would not be wise or it is even impossible to remain indecisive and wait for sufficient evidence. [10] In other words, James leaves the ideal of evidentialism principally intact but claims an exception to its alleged universal applicability. There are cases in which evidentialism should not be applied. Religious claims are such cases, he argues, and consequently, evidentialism should not be applied to judge them.

In particular, James suggests that certain situations exhibit logical features which make it necessary to decide in spite of the fact that there is not enough evidence available—this is a point I will defend extensively below, viz. in sections 2.3 and 2.4. In contrasting reliance on ‘intellectual’ faculties with reliance on non-intellectual ones, such as will or passion, he means to say that, when our intellectual capacities are exhausted, we are entirely justified, entirely within our epistemic rights, to depend on non-intellectual ones in certain circumstances. Since those circumstances apply in the case of certain religious claims, we are entirely within our epistemic rights to rely on will or passion in order to judge them.

I will follow James’ clues when refuting evidentialism. First, I follow the general line of argument he proposes against evidentialism. Second, I accept the particular argumentative strategy he uses to undermine evidentialism, viz. to show that evidentialism leads to absurd consequences if applied to practical problems. Moreover, I follow the method by which James achieves his goal, viz. demonstrating by examples that applying evidentialism in the ‘real world’ (as opposed to using it purely as a philosophical doctrine) leads to imprudent practical results.

Yet, I would like to emphasize that, although I subscribe to many aspects of James’ account, I do not share the doctrine that truth is expediency of belief or utility that is usually attributed to him. [11] A thorough criticism of evidentialism does not require reliance on such a highly contested doctrine, nor does it require reliance on any other full-blown pragmatist doctrine of that sort. The arguments in ‘The Will to Believe’ are logically independent from hard-nosed pragmatist doctrines or can be reconstructed as such.

Neither James nor his evidentialist opponents take great pains to clarify the notion of ‘evidentialism’. Yet, for my purposes, viz. to scrutinize evidentialism’s relevance for criticizing religion, it is crucial to distinguish between two sorts of evidentialism. The sort of criticism of religion that emerges from evidentialism hangs on it. I distinguish between truth-oriented evidentialism and quiescent evidentialism. Let me explain both, beginning with truth-oriented evidentialism.

Truth-oriented evidentialism entails the view that, if there is not enough evidence on a given issue (GI), one should refrain from forming beliefs on GI. This injunction to form beliefs amounts to saying that one should not distribute the truth-values ‘true’ or ‘false’ over GI. Note that the claims of truth-oriented evidentialism are restricted to considerations regarding belief or truth proper. That is to say, the truth-oriented evidentialist may prohibit you – if his arguments succeed – from forming beliefs about GI, but he cannot prohibit you from making educated guesses, hypothetical conjectures or whatever concerning GI. And, closer to James, he cannot prohibit you from forming opinions over GI on non-epistemic grounds. [12] All he can say is that you should not regard claims pertaining to GI to be true in the proper sense of the word, if you have not enough evidence concerning GI. If his claims reach further than that, the truth-oriented evidentialist has abandoned his position in the direction of quiescent evidentialism.

Mentioning it leads to the second version of evidentialism, viz. quiescent evidentialism. This sort of evidentialism implies not only the injunction to refrain from holding GI to be true if you do not have enough evidence, but, rather, it implies that you should not make a decision for or against GI. You should stay away from it. Presumably, you should wait until sufficient evidence has become available. That is to say, you should not only abstain from forming beliefs about GI but also from educated guesses and the like. Moreover, you should not form beliefs or opinions on GI on non-epistemic grounds. For example, you should not form opinions on GI based upon hope, will, or passion.

James is not clear what sort of evidentialism he has in mind in his various attacks on evidentialism. I assume that he directs at least some of his attacks towards quiescent evidentialism. For example, the following quote can be interpreted as being directed towards quiescent evidentialism rather than truth-oriented evidentialism: ‘To preach scepticism to us as a duty until ‘sufficient evidence’ for religion be found, is tantamount therefore to telling us, when in presence of the religious hypothesis, that to yield to our fear of its being error is wiser and better than to yield to our hope that it may be true…’ [13] What James targets here is an evidentialist position that does not only prohibit you from believing religious claims to be true in the proper sense of the word but prohibits you from holding religious claims in whatever sense of the word. In the face of insufficient evidence, you are not only prohibited to distribute the relevant truth-values over religious claims but you are also prohibited from hoping them to be true. The sort of evidentialism James targets here goes beyond truth-oriented evidentialism. He targets a stronger version of evidentialism, viz. what I called above quiescent evidentialism. This type of evidentialism is – rightly or wrongly – attributed to Clifford and like-minded people.

I take it that this sort of evidentialism is still held in our days. It may not be that widespread in the philosophical arena any more – for good reasons, as will become clear below – but it is still used in the public domain. There, it is not only used as a tool to criticize religion but it is also used in the moral discourse. A more or less randomly chosen example is the frequently used argument that, since there is not yet enough evidence available that passive smoking causes cancer, we have no right to impose restrictions on smoking in order to protect passive smokers. Another example would be an argument that was frequently used after the Chernobyl catastrophe in 1986, whereby large amounts of radioactivity were released. In countries that were potentially affected by this release, e.g. Germany, the question came up as to whether one should not impose restrictions on the consumption of certain sorts of food (e.g. dairy products and mushrooms). In this situation, the following argument was introduced: ‘Since we do not have enough evidence on the long-range effects of radioactivity in the food-chain, we should not impose any restrictions on food-consumption.’ [14]

This sort of evidentialism is in a certain sense stronger than truth-oriented evidentialism since it bears not only on the question of the distribution of truth values but also on the theory of action (in a broad sense of the term). It prescribes a certain maxim of action, or, more precisely, a certain maxim of inaction. It prohibits to take a certain course of action, viz. to make a decision for or against GI. In the cases mentioned: We should not (yet) take action against smoking or certain sorts of food because we do not have enough evidence on the issues (passive smoking or the long-range effects of radioactivity in the food-chain). Since this sort of evidentialism entails the injunction to refrain from acting (as long as we do not have sufficient evidence on a given issue), it entails quiescent consequences. Hence, the name ‘quiescent evidentialism’.

Let me summarize the distinction between both sorts of evidentialism because it is central to my argument. Truth-oriented evidentialism implies that if, on a given issue, GI, not enough evidence is available, we should refrain from distributing the relevant truth values over GI. However, it does not prohibit us from making reasonable guesses, conjectures, forming hypotheses or whatever on GI. And it is silent about non-epistemic justifications of GI. Rather, its range of applicability is restricted to the alethic domain, i.e. to the domain in which questions of truth proper are discussed.

Quiescent evidentialism goes deeper. It implies not only alethic consequences but also practical ones, viz. action-relevant consequences (in the broad sense of the word). It implies that, if there is not enough evidence available on GI, we should refrain not only from distributing the relevant truth values over GI but also from taking action concerning GI. If we do not have enough evidence in a disputed case, we should not make a decision on it. Presumably, we should wait until sufficient evidence has become available. In the meantime, however, we should stay away from GI.

How does this distinction between both sorts of evidentialism ramify into the criticism of religion? In what sense differs a criticism of religion based upon truth-oriented evidentialism from a criticism of religion based upon quiescent evidentialism?

The truth-oriented evidentialist critic of religion argues that we should not distribute the relevant truth-values over religious claims, if there is not enough evidence available on them. That is to say, she argues that we should not hold them to be true or false in the proper sense of the word. Furthermore, if she happens to think that ‘believing p’ implies holding p to be true, [15] she will suggest that we have no right to believe in religious claims, if there is not enough evidence available. But in either case, her claims are restricted to the alethic domain in the sense specified above. That is to say, if her arguments succeed, the most she can squeeze out of them is the contention that religious claims should not be held to be true, and therefore we have no right to believe in them.

However, she cannot prohibit us from embracing religious claims on other than alethic grounds proper. She cannot prohibit us from embracing religious claims on the basis of being e.g. educated guesses, hypotheses etc. For example, she cannot prohibit us from holding religious beliefs to be hypothetically true because they provide sufficient motives for action or something of that sort. And, most important for the discussion of James, she cannot prohibit us from holding religious beliefs on non-epistemic grounds, i.e. from having faith if this faith is based upon (reasonable) hope, will, or passion.

The quiescent evidentialist critic of religion, however, goes further. She does not only hold that we should refrain from distributing the relevant truth-values over religious claims but that we have no right to embrace those claims on non-alethic grounds, either. That is to say, we should not hold them because they are educated guesses, hypotheses etc. (in the case of insufficient evidence). We have no right to embrace them on non-epistemic grounds, either. We are not epistemically entitled to hold them because we direct our hope or will towards them or steer our passion in their direction. In short, we have no right to hold religious claims, and therefore should stay away from them altogether.

Since the quiescent evidentialist does not argue that religious claims are false, she is not an atheist. Her concern regarding religion is not ontological, as it is in the atheist’s case but, rather, epistemological. She argues that for epistemological reasons, viz. because there is not enough evidence available, we should stay away from religion. Rather than being an atheist, the quiescent evidentialist is an agnostic. Since his agnosticism goes deeper than that of the truth-oriented evidentialist critic of religion, who restricts his claims to considerations concerning truth proper, we can regard him to be an absolute epistemological agnostic.

Let me proceed to evaluate both sorts of evidentialism. What are their merits and what are their weak points? As indicated above (see introduction), we will follow James in considering them not only as theoretical-philosophical doctrines but also as practical ones, i.e. as doctrines that potentially have consequences for the pursuit of our everyday practices.

Let us begin with truth-oriented evidentialism. Its implication, that we have the right to hold a claim to be true only if there is enough evidence available, may be helpful in cases in which there are reasons to assume that enough evidence will surface in the near future (against Rorty’s all-out rejection of evidentialism). But in cases in which there is not enough evidence available and in which there is no hope to achieve it in the near future – let us call these ‘situations of uncertainty’ – this maxim is not at all helpful.

Let me give an example from everyday practice. Whether the current changes in our climate (‘global warming’) are caused by the Greenhouse effect or not, is yet uncertain. The evidence is insufficient, as far as I know. Thus, telling us that we need to have sufficient evidence before distributing the relevant truth values over the issue, as the evidentialist does, means courting a nice theoretical ideal. However, it does not bring us very far. It does not help us in deciding the issue at stake. If we cannot tell what the relevant truth values are, we have to make our decisions on other than truth-related grounds. If we cannot wait for truth to emerge on the issue, then we have to make our moral policies without it.

The reason that we cannot afford to wait is that what is at stake is of such importance that we cannot afford the luxury of remaining indecisive. If we do, we run the risk that the living conditions on this earth drastically deteriorate (e.g., that the flood rate and intensity on earth will dramatically increase so that entire regions will overflow). In short, when, in spite of a significant amount of uncertainty, we have to make a decision one way or the other, the evidentialist maxim has little to offer.

One may object that this is an extreme sort of example, which occurs rarely. However, more dilemmas of this kind arise in ordinary practice than one may expect. The number of cases in which insufficient evidence is available, and yet we have to make a decision one way or the other, is rather high and, most importantly, will increase steadily. The reason is that current technological developments, e.g. the emergence of new bio-medical possibilities and related developments, have reached a degree of complexity that makes it increasingly difficult to assign clear evidence to issues pertaining to those developments. Yet, often the issues at stake are of such grave importance or have such wide-ranging or long-lasting consequences [16] that we cannot afford to refrain from decision-making on them.

Since truth-oriented evidentialism cannot account for such cases, it is irrelevant to a large number of cases. If it is true that the situations will increase in which uncertainty reigns but in which we still have to make a decision, truth-oriented evidentialism increasingly loses ground. The reason is that all that the truth-oriented evidentialist has to offer in those cases is a prohibition to distribute the relevant truth values over such cases. Yet, if we have to come to a decision on them one way or the other, this will not help a great deal. If we cannot make decisions on truth-related grounds, we have to make them on other grounds, say, on the basis of educated guesses, conjectures, non-epistemic grounds, such as hope, will or whatever. In short, truth-oriented evidentialism is not of great help when being confronted with issues of the sort mentioned. Therefore it is by far not as interesting a doctrine as its proponents suggest.

What about quiescent evidentialism? I think that it is worse off. In the cases mentioned, quiescent evidentialism is not only inapplicable but also morally irresponsible. In those cases, it entails highly questionable or imprudent practical consequences. It can be used for the purposes of promoting the status quo in situations in which it could be vital to consider and implement drastic changes. It can be used to legitimize the failure to make a decision in cases in which decision-making is vital- and the above-mentioned examples of passive smoking and the Greenhouse effect could be such cases. It justifies an extreme conservatism and may easily be abused for political or economic purposes, say, of the tobacco industry and its interest to prevent action being taken against smoking.

Let me give another type of example that reveals the conservative tendencies in quiescent evidentialism. This issue receives much attention in Germany right now and, presumably, in other countries that have a public health system that is supposed to cover almost all medical expenses. I think of discussions over the question whether there is enough evidence that a given alternative medical treatment, ATM, is efficient. In countries which have a public health system of the sort mentioned, answering that question is important since only if there is sufficient evidence for its efficiency, the public health carriers will take over the costs for ATM. Representatives of those carriers, eager to cut expenses, often invoke the argument that there is not yet sufficient evidence for ATM, and accordingly, that its costs cannot be taken over. Given that research into the efficiency of medical treatments lasts usually long and is highly complicated, [17] invoking this argument is a convenient way for the public health carriers to shun all novel treatments. My point is that invoking quiescent evidentialism can lead to an extreme conservatism that rules out innovation on a priori grounds. However, there are cases in which innovation is mandatory. In these cases, quiescent evidentialism to morally doubtful and imprudent results.

Let me conclude my evaluation of evidentialism with yet another example that is admittedly rather personal but has the advantage of being less complex than the above ones. It demonstrates clearly that, in certain circumstances, time is a crucial factor. That is to say, it shows that there are cases in which decisions have to be made now or in the near future. Postponing the decision – the strategy that both sorts of evidentialism suggest – is no option. In this case, therefore, evidentialism leads to dubious results.

Imagine that there are two ways to treat a certain sort of cancer, one being the standard medical paradigm, SMP, the other an alternative paradigm, AP, that holds some promise but for whose success there is as yet insufficient evidence. Imagine also a person who suffers from that sort of cancer and who seriously considers to be treated by AP rather than SMP (say, because she is afraid of SMP’s side effects). How do both sorts of evidentialism cope with such a situation? What practical consequences do they have?

The truth-oriented evidentialist will advise her to refrain from forming beliefs on the issue since there is not enough evidence available. He will suggest that she should not consider claims pertaining to AP to be true in the proper sense of the word. However, she will reply – and she is absolutely right in doing so – that this does not help her at all. She cannot afford to wait until sufficient evidence will be available. She has to make a decision now. And if she cannot make it on the basis of sufficient evidence, then she has to make it on different grounds. If there is no truth of the matter (in the proper sense of the word), then she will have to make her decision on other than truth-related grounds proper. Whatever she does, it is clear that the truth-oriented evidentialist answer is irrelevant for the purposes of solving her problem.

And the quiescent evidentialist solution to her problem? The quiescent evidentialist will advise her to refrain not only from forming beliefs about AP but also from taking action in the sense specified above. She should not even make hypothetical conjectures and the like on the issue. She should wait until enough evidence has become available and, until then, refrain from making a decision on the issue. However, she cannot afford to wait much longer. If she does, chances are that she will either not enjoy medical treatment at all and will have to face the serious consequences of failing to act. The price for refraining from decision-making is simply too high in this case. Or, else, if she happens to live in a country in which everybody is automatically insured by a public health-carrier prescribing one particular treatment in such a case – SMP – refraining from decision-making is tantamount to choosing in favor of SMP. In this case, her refusal to make a decision results in her being treated by the standard means, viz. by SMP.

Obviously, my point is not that SMP is necessarily worse. My point is that relying on a quiescent evidentialist method deprives you of the means to deliberate on the issue. You do not actively and consciously decide in favor of SMP but, rather, have given the decision out of your hands. You leave it over to the contingent circumstances of the situation rather than taking the decision into your own hands. Thus, by supposedly refraining from making a decision on an issue, you may take sides implicitly in certain circumstances.

Or, else, quiescent evidentialism can simply be too costly in certain cases. Refraining from decision-making, as the quiescent evidentialist suggests, results in her not being treated at all. Obviously, it would be absurd to forfeit treatment and risk the grave consequences of that inaction only to follow an abstract quiescent evidentialist rule.

Let me summarize my criticism of both sorts of evidentialism. Quiescent evidentialism leads to questionable results in certain cases. In the above-mentioned case, applying it is either too costly or, else, results in making a decision implicitly without actively reflecting on it. And truth-oriented evidentialism is simply irrelevant in a case like this.

And, as indicated above, cases like this are not exceptional. Rather, if the above provided hints point in the right direction, then we live in conditions in which many decisions have to be made under a significant amount of uncertainty. Given the complexity that our technological developments have reached at this point, given the serious and long-lasting impact they can have on the human race as a whole, and given the new possibilities that have emerged (e.g. alternative medical treatments that cast doubts on what used to be taken for granted); given all that, uncertainty in the sense specified above reigns to a good extent—at least, in the Western hemisphere.

I would even go as far as saying that many of the crucial decisions we, as Westerners, have to make on an individual but also collective basis have to be made under a significant amount of uncertainty. And evidentialism does not provide any help in many of those cases. That being the case, it should not be applied on a universal basis, as its proponents suggest.

How does my criticism of evidentialism relate to the different positions regarding evidentialism named above? I disagree with Rorty by suggesting that it is unnecessarily radical to abrogate evidentialism entirely. Where applicable, it may well be used. Proceeding in such a fashion has the advantage of requiring less strenuous revisions than Rorty suggests. My suggestion does not require abrogating foundationalism and realism. Thus, my arguments against evidentialism have the potential of convincing the foundationalist and the realist as well, whereas Rorty’s do not. That being the case, my stance with regard to evidentialism is prima facie preferable to Rorty’s.

But in accordance with James, I suggest that there are many cases in which evidentialism cannot be applied. Thus, contrary to what the classical defenders of evidentialism propose, I suggest that its range of applicability should be restricted. Clifford is mistaken in assuming that ‘it is wrong always, everywhere, and for anyone to believe anything upon insufficient evidence’. There are circumstances in which it would be irrational to follow this maxim. There are cases in which it would be imprudent or unwise to follow any evidentialist maxim, be they truth-oriented or be they quiescent in nature.

However, if that much is true, why is evidentialism still held in such high esteem? Why do people still think that it is the ‘high road’ to answer questions of belief-acquisition and of justification? Why do philosophers defend it that stubbornly and why does it enjoy such a great respect in the public domain? And why are attempts to challenge it, such as James’, greeted with such horror? I think there are, at least, two reasons for that, one philosophical, the other political in nature. Let me begin with the philosophical reason.

The philosophical reason for defending evidentialism ferociously is that here one of the most precious tenets of the Enlightenment is at stake. What I mean is the Enlightenment optimism concerning the possibilities of acquiring knowledge. Since the Enlightenment, the following idea has been deeply ingrained in us: ‘We, the human race, know everything. And if we happen ignore something at this point, we can be confident to we shall come to know it very soon. Situations of cognitive uncertainty will not prevail, our knowledge increases steadily so that, in due time, we will know everything and be capable of deciding appropriately’. Let us call this the Promethean myth (concerning the possibilities of human cognition).

Admitting that there are situations in which uncertainty prevails and will, in all likelihood, continue to prevail, and yet we still cannot afford to refrain from deciding, challenges this Promethean myth. It reminds us of our cognitive finitude. That is unwelcome to the Enlightenment mind.

Furthermore, the Promethean myth was invented to battle religion: ‘After the death of God, we, the human race, are the true heirs of him in matters cognitive. That is why our cognitive resources are potentially unlimited’. And people like James shake this myth not for philosophical or related reasons but in order to make space for – horribile dictu – religion. That is the reason why representatives of the Enlightenment, such as Russell have attacked James particularly ferocious. [18]

And this myth was created for the purposes of promoting the ‘higher faculties’ of (intellectual) reason over the ‘lower faculties’, such as passion. James, however, points out that those higher faculties lead us into a dead end at times and that, in those situations, it is entirely reasonable to rely on passion. [19]

These are (some of the) philosophical reasons why people keep insisting on the importance of the notion of evidence and why they ferociously attack any criticism of it, such as James’. Daring to challenge the Promethean myth and, worst of all, doing that in the name of both of Enlightenment’s ‘no no’s’, viz. religion and passion, must appear to be frivolous to a ‘steadfast Enlightenment mind’, such as Russell’s.

The political reason why people keep insisting on evidentialism is that it can be used to defend an extreme conservatism that serves particular group-interests. By evidentialist means, you can easily justify the status quo. This comes in handy in certain political constellations. It does if, say, you want to deliberately postpone a ban on smoking because such a ban would hurt the tobacco industry or if you do not wish the public health carriers to take over additional expenses for alternative medical treatments.

After all this criticism of evidentialism, a reconciliatory remark may be due. I acknowledge that its early fathers may well have had good reasons to propose it. In a type of cognitive environment in which we still had reasons to believe in an unbroken optimism with regard to human cognitive capacities and/or in which the possibility of free research was severely restrained (say, by dogmatic ways of thinking), this ideal may well have had a liberating effect. In the situation of the (early) Enlightenment, it may well have been a device to emphasize the individual’s autonomy to believe what seems to be reasonable to her over and against heteronomous ways of thinking (say, guided by church-doctrine). In those circumstances, it may well have warranted intellectual honesty in otherwise intellectually suffocating circumstances. However, our cognitive environment has changed dramatically and what was once liberating is now suffocating. Hanging on to evidentialism as a universal ideal unduly deprives us of moral and intellectual resources, including religious ones (see below, section 3).

However, if its beginnings were innocent enough, is there not anything worth saving from the evidentialist heritage? I think that a general maxim to search for sufficient evidence, if available, is worth salvaging from this heritage and can still be applied in our current intellectual climate. But this maxim needs to be applied in a mild-mannered spirit, not as ‘dogmatically’ and self-righteously as the latter-day evidentialists with their anti-religious zeal do. The world is complex – more so than our Enlightenment fathers dreamt it would be – and we are confronted quite frequently with situations in which uncertainty reigns and we still have to decide one way or the other. In those situations, it would be counterproductive to court evidentialism just because it is a valuable theoretical ideal. Rather, we should always be open to the possibility that it cannot be applied in certain circumstances and should not waste our best energies into defending it no matter what. There are better ways to use those energies, for example to search for procedures that allow the possibility of rational decision-making under a considerable amount of uncertainty.

Let us scrutinize what the upshot of these considerations on evidentialism is for the question how strong the evidentialist challenge to religion is. Where appropriate, I will do that by referring to James’ considerations regarding evidentialism.

If our above contentions hold, then the evidentialist ideal cannot be applied on a universal basis. There are cases which fall outside its range of applicability—let us call them ‘uncertainty situations’, US. Evidentialism should not be applied in US-cases. And if religious claims turn out to be US-cases, evidentialism should not be applied to them. That is to say, if religious claims are comparable in their relevant respects to the US-cases as specified above, evidentialism is inapplicable to them. In that case, evidentialists who nevertheless try to apply their ideal to those religious claims are guilty of ‘intellectual imperialism.’ They unwarrantedly attribute universal scope to their ideal and extend its range of applicability to cases to which it should not be extended. [20]

Let us begin by scrutinizing the force of the truth-oriented evidentialist critique of religion. Let us assume that the truth-oriented evidentialist is right in holding that religious claims lack evidence and that, therefore, we should not hold them to be true. Does that imply a fatal blow to religion, as it may seem to be at first sight?

I do not think so. The reason that it may appear to be a fatal blow to religion is only that the word ‘true’ is used in unqualified ways. Using it thus suggests that, since religious claims cannot be held to be true, they cannot be held to be valid or justified in all possible senses. However, the religious person who is aware of this unqualified use has an easy answer to this truth-oriented evidentialist challenge. She can easily grant that we are not justified in holding religious beliefs to be true (in the proper sense of the word) in the same sense in which we are not justified in holding statements concerning ‘global warming’ or the efficiency of an alternative medical treatment concerning cancer to be true. [21] But, she will insist, not being justified in holding p to be true in the proper sense of the word does not entail not being justified to hold p on other than alethic grounds. In short, she will insist that there may well be other than alethic grounds to justify making religious claims.

Thus, the religious person can yield more or less easily to the truth-oriented evidentialist, if she is aware of the limitations inherent in the evidentialist position. And the limitations are that, whatever the truth-oriented evidentialist proposes, cannot preclude us from having faith in religious claims on grounds different from truth proper. For example, the religious believer can suggest that we should believe in them because we think that they are educated guesses, reasonable conjectures or whatever.

3.1.1 James’ answer to the truth-oriented evidentialist challenge

If the believer is a Jamesian, she may hold that nothing that the truth-oriented evidentialist has said thus far can prevent us from holding that there are sufficient and rational reasons to hope that those religious claims are true (see above, 2.1.). Put more generally, the religious person can point out that the claims brought forward by the truth-oriented evidentialist are limited to epistemic considerations. But this does not touch upon non-epistemic ones. The truth-oriented evidentialist cannot prevent us from believing in religious claims on non-epistemic grounds.

That is the point of what James expresses in the later parts of ‘The Will to Believe’ under the contrast of ‘intellect’ as against ‘will or passion’. His point is that forming an opinion on an issue based upon will or passion is not necessarily inferior to forming it on the basis of intellectual capabilities. In some instances, we have no other choice but to decide with the help of passion. For example, when our intellectual resources are exhausted but we are confronted with a genuine option, we are completely justified in relying on our ‘passional nature’:

Notice that, when James speaks in the part that is italicized above about ‘passion’, he uses it differently from the way he uses it throughout this article. Whereas in this article he is generally concerned with removing prejudices against the use of passion here, he takes over the negative evaluations that his opponents associate with the term ‘passion’ here. So what he means is that the evidentialist advice ‘Do not decide, but leave the question open’ is, contrary to what the evidentialist assumes, ‘passional’ in the sense that it is not better, not more rational, than making a choice.

James’ suggestions can be translated into the nomenclature we used above as being directed against truth-oriented evidentialism. His contention that even if there are no sufficient ‘intellectual’ grounds for having faith in religious claims, there may still be legitimate ‘non-intellectual’ grounds for doing that, means that even if we do not have sufficient reasons to consider religious claims to be true in the proper sense of the word, we may well have other grounds to regard them as justified or legitimate. In short, even if we do not have sufficient epistemic (‘intellectual’) grounds to hold religious claims, we may well have sufficient non-epistemic ones, based upon will or passion.

Thus, it turns out that James provides a non-epistemic justification of faith. Quite unlike Plantinga who relies on notions such as ‘proper functioning’ and ‘design plan’, James does not rebut the evidentialist challenge to religion on epistemic but on non-epistemic grounds. He does not argue that religious claims are true in the proper sense of the word. He does not show that, contrary to what the evidentialist critics of religion assume, there is sufficient evidence for religion. He takes another line of argument by showing that the range of applicability of epistemic justifications is limited. He argues that there are also non-epistemic justifications and in certain occasions it is prudent or even unavoidable to rely on them. And applying them to religious claims cannot entitle us to hold those claims to be true in the proper sense of the word. Yet, applying them can entitle us to have faith in them on other than truth-related grounds. For example, we can be justified in having faith in them based upon will or passion.

3.1.2 A last attempt to make the truth-oriented evidentialist challenge cut ice

The truth-oriented evidentialist may at this point make one final attempt to salvage the importance of his criticism of religion and, by implication, of James, by arguing in the following fashion. If we are not entitled to hold religious beliefs to be true in the proper sense of the word, then we are not entitled to ‘believe’ in them in the proper sense of the word either (since ‘believing p’ implies ‘holding p to be true’). And since James claims to defend belief, as the title of ‘The Will to Believe’ suggests, he ultimately fails.

For the sake of the argument, let us grant the existence of such a link between belief and truth-related concerns. In what sense does this help the truth-oriented evidentialist? It does not make his point much stronger. All that he can gain from invoking this link is a Pyrrhic victory. If he is right, James may have to reword the title of his article and skip the term ‘believe’. [23] However, he could easily substitute ‘having faith’ or something of that sort for it. This is no more than a question of nomenclature. As far as the substance of the argument is concerned, James’ point still stands. The truth-oriented evidentialist critique of religion is much less relevant than usually assumed because it ranges only over epistemic justifications whereas there are non-epistemic justifications to legitimate religious claims.

Let me summarize my criticism of the truth-oriented evidentialist challenge to religion. Its range of applicability is restricted to epistemic justifications. However, as indicated above, in many cases epistemic justifications are not the only justificatory means available. There are instances in which it is inappropriate to rely exclusively on epistemic justifications. Religious claims, for example, can be justified by non-epistemic means, e.g. by reliance on will or passion, as James suggests.

Let us turn now to the quiescent evidentialist critique of religion. As indicated above, the quiescent evidentialist argues that, since we do not have enough evidence on religion, we should stay away from it altogether. This claim is stronger than the truth-oriented critique of religion in that it amounts to a full-fledged agnosticism. The quiescent evidentialist rejects not only the acceptance of religious beliefs as true in the proper sense of the word but also forecloses their acceptance on different grounds. He rejects religion and the possibility of making claims about it altogether.

Above, I have argued that quiescent evidentialism leads to consequences that are doubtful, to say the least. Following the maxim implied in it can lead to morally questionable results. This is presumably the case when it is applied to the issue of global warming. Or, it can lead to results that are, to say the least, imprudent. This is the case when it is applied to the problem of the cancer patient considering to be treated by alternative medical means. As was demonstrated above, refraining from making a decision on the issue is either too costly – if it leads to not being treated at all – or else amounts to making a decision without explicitly reflecting on it—if it leads to the patient being treated by standard medical means.

My point is that the same goes for the quiescent evidentialist critique of religion. It is imprudent to apply it in the case of certain religious claims. It leads to results that are either too costly or, else, amounts to deciding on the issue – viz. against religion – without explicitly reflecting on it. I will flesh out both cases in the following. However, before doing that, I will summarize two features that characterize situations in which applying quiescent evidentialism is unwise.

3.2.1 Two features of a situation in which using quiescent evidentialism would be unwise

What are the logical features of situations in which there is a good chance that using quiescent evidentialism is imprudent or morally questionable? Drawing on our above discussion on the matter, we can identify two features of situations in which applying quiescent evidentialism would be inappropriate. First, there must be genuine uncertainty on the matter in the sense that there is not enough evidence available either way and that it is improbable that it will become available in the near future. That is to say, a presupposition for the possibility of rejecting quiescent evidentialism is a situation in which the choice between two options cannot be guided by conclusive evidence. [24]

The second feature of a situation in which applying quiescent evidentialism is inappropriate is that the choice between both options cannot be avoided (for logical reasons), as is the case with ‘forced options’; or, that it would be imprudent to avoid making a choice because the results of this failure are either too costly or, else, amount to making an implicit choice. This is the case when the cancer patient applies quiescent evidentialism. Another possibility is that the stakes involved are extremely uneven. This is the case if the goods at issue differ significantly, e.g. the consequences of not taking a particular action are drastic whereas the consequences of taking action are not. If one or more of those conditions are met, [25] the second feature of a situation in which applying quiescent evidentialism would be inappropriate is at hand.

In sum, if both features apply, viz. that a choice between both options is unavoidable, would be imprudent to avoid or whatever, and if there is not enough evidence available either way (and, in all likelihood, will not become available in the near future), we are confronted with a situation in which applying quiescent evidentialism is most likely inappropriate. In order to avoid misunderstandings, let me point out here explicitly that only when both features apply simultaneously, applying quiescent evidentialism is imprudent.

3.2.2 Do religious claims exhibit the features of situations in which applying quiescent evidentialism would be unwise?

Let us scrutinize as to whether religious claims are conceivable which form a situation in which both features apply at the same time. But before doing that, let me remind you what is at stake in the discussion right now: The quiescent evidentialist argues that, as a universal rule, we should stay away from issues for which we do not have enough evidence. Since we do not have enough evidence on religious claims, we should stay away from them.

I have attacked the universal claims implied in quiescent evidentialism and argued that there are certain situations in which it would be inappropriate to apply this sort of evidentialism. Those are situations in which both features apply simultaneously, viz. that insufficient evidence is available and will become available in the near future and it would be unwise not to choose. My next step will be to show that certain religious claims are situations in which both features apply, i.e. situations in which not enough evidence is available and in which it would be unwise to refrain from deciding. The upshot of those considerations is the claim that applying quiescent evidentialism to those religious claims is inappropriate. In other words, certain religious claims are situations in which we have to make a decision in spite of the fact that there is not enough evidence available.

3.2.2.1 Does the first feature of a situation in which applying quiescent evidentialism would be unwise apply to religion?

Let us begin our discussion with the first feature of situations in which it would be inappropriate to apply quiescent evidentialism. (Notice that, in the following, I will use the notions ‘feature of a situation in which applying quiescent evidentialism would be unwise’ and ‘conditions that have to be met in order that applying quiescent evidentialism would be unwise’ interchangeably.) Hence, the question is whether there is insufficient evidence available on religion and as to whether it is unrealistic to expect it to become available in the near future.

I think that there neither is conclusive evidence for the falsity of certain religious claims nor for their truth. The religious claims I have in mind are e.g. ones in which a transcendent notion of God and/or other deities plays a crucial role. Traditional Christian claims being based upon a robust notion of God, a Trinitarian concept of the Son etc. are examples of such claims. And the reason that there is not enough evidence available concerning those notions is not that, say, theologians are incompetent researchers or something of that sort. The fact that there is not enough evidence available has nothing to do with deficiencies on the side of the researchers working in the field of theology or anything of that sort but, rather, with the nature of the objects under scrutiny.

Let me explain. Certain research-objects are extremely complex. An example of such an object could be the question of what causes the current changes of our weather patterns. Presumably, the difficulties in tracing those causes do not testify to any inadequacies of meteorologists researching on the weather but, rather, to the fact that the constellations that cause our weather are extremely complex (as is witnessed by the notorious failures of weather forecasts). Certain objects are so complex that capturing them in law-like or nomothetical statements is extremely difficult, if not impossible.

In principle, this holds for the religious domain as well - except, of course, that the degree of complexity is significantly or even infinitely higher. The object under scrutiny there is extremely complex as well and probably incapable of being captured in nomothetical statements or something of that sort as a matter of principle (say, if God is absolutely free in His decisions). If, for example, God is truly totaliter aliter in the sense that the early Karl Barth proposes, the fact that we cannot capture Him in nomothetical statements or something of that sort does not come as a surprise. This does not speak to inadequacies on the side of the theologians pursuing research into the nature of God but rather to the fact that God is truly God (to say things in Barthian parlance). That is, it speaks to the fact that God is the Lord of all there is rather than a human idol, and man is truly man, i.e. a sinful being whose capacities including his cognitive ones are misguided.

Or, take a standard philosophical line of argument. Human cognitive possibilities are limited and the object of religious cognition, a transcendent notion of God as is e.g. suggested in the Christian tradition, lies outside those limits. That is to say, in the cognitive sphere the notion of God has only the function, if it has a function at all, of serving as an ‘idea of reason’ in the sense of Kant’s transcendental dialectics or something like that. If that were true, it would not be a surprise that there is not enough evidence available on the issue. Again, that would not speak to any inadequacies of the researchers delving into the notion of God but, rather, to recognizing the limits of human cognition and asserting that a transcendent notion of God falls outside those limits.

These arguments show that the evidence available on certain religious claims is insufficient. [26] These claims are not even uncommon in religions such as the Christian one. Furthermore, those arguments show that it is unreasonable to assume that sufficient evidence will become available in the near future. If God is truly totaliter aliter in Barth’s sense and/or the Kantian rationale or a similar philosophical one apply to religious claims, then it is unreasonable to expect that sufficient evidence will emerge in the near future. [27] But if there is not enough evidence available on certain religious claims and it is unreasonable to expect it to become available in the near future, the first condition is met for there being a situation in which applying quiescent evidentialism would be unwise. Now, what about the second condition?

3.2.2.2 Does the second feature of a situation in which applying quiescent evidentialism would be unwise apply to religion?

It will be remembered that the second condition for there being a situation in which applying quiescent evidentialism is inappropriate is that it would be unwise to refrain from making a decision because the results of this failure are either too costly or, else, amount to taking sides implicitly (3.2.2.2.2), that the stakes involved are extremely uneven (3.2.2.2.1) etc. Let us begin with the question as to whether religious claims can be reconstructed in such a fashion that the stakes involved are extremely uneven.

3.2.2.2.1 Are religious claims decision-situations in which the stakes are extremely uneven?

James seems to have cases of religious claims [28] in mind in which the latter applies, viz. that there exists a grave imbalance between the goods involved. He suggests that we lose much more in the case of unbelief, if religion is true, than in the case of belief, if religion is untrue. [29] What could examples of religious claims be for which the stakes are extremely uneven? A rather crude but particularly perspicuous one is the idea that the believer is better off than the unbeliever is because a deity has decided that believers are better off (e.g. go to heaven, whereas unbelievers don’t). Or, more refined (and probably more Jamesian): The unbeliever is worse off than the believer not because a deity has decided that this is the case but because of a more or less automatic chain of events, e.g. because the unbeliever is betraying himself of the benefits of faith, such as the possibility of consolation.

Whatever you suggest at this point (yet other construals are conceivable), it will help to add a relevance clause to the effect that the benefits to be reaped from faith are vital. Such a clause helps to show the unreasonableness of quiescent evidentialism. Because if the benefits to be reaped from faith were not vital, it would make it too easy to walk away from the necessity to decide in the absence of sufficient evidence. [30] Let me explain what I mean by comparing the case of religious faith to belief in the existence of extraterrestrial life.

Let us assume that in both cases not enough evidence is available to decide either way. Yet, in the case of the question as to whether extraterrestrial life exists, we can easily abstain from making a decision on the issue. Entertaining as this question may be, it does (usually) not have any serious existential relevance to us. The case of (certain) religious claims is different, however. Here, questions as to whether a deity possesses particular attributes, e.g., is capable of comforting the miserable, of rectifying grave historical injustices or even of overcoming death (in principle), can be questions of great existential relevance. [31] But if they are, it is imprudent to walk away from them as easily as walking away from the question as to whether extraterrestrial beings exists. One would deprive oneself of potential vital benefits that way.

In sum, it is well conceivable that certain religious claims are situations in which the stakes involved are extremely uneven. And since being extremely uneven is one characteristic of the second condition of situations in which applying quiescent evidentialism would be imprudent, those claims meet the second condition. That is to say, those claims are good candidates for falling into the set of situations in which applying quiescent evidentialism would be imprudent. [32]

3.2.2.2.2 Are religious claims decision-situations in which refraining from deciding would be too costly or would amount to taking sides implicitly?

Above, I had argued that applying quiescent evidentialism to cases such as that of the cancer patient considering to be treated by an alternative medical paradigm is not an option. It leads to an outcome that is either too costly or, else, amounts to taking sides implicitly. The same holds in the case of certain religious claims: Applying evidentialism is (a) either too costly or, else, (b) amounts to taking sides implicitly. Let me flesh out that claim.

With regard to (a), (certain) religious claims are equivalent to that of the cancer patient in the following fashion. The cancer patient cannot afford to abstain from deciding because this could result in her not being treated at all (in certain circumstances). And the person pondering (certain) religious questions can also not afford to walk away from deciding on them because this would be too costly—e.g., it might deprive her of her only means to survive in miserable circumstances or even of eternal happiness, or else, it could result in being condemned eternally (see above, 3.2.2.2.1).

The same holds for the second undesirable outcome of applying evidentialism, viz. (b) that of implicitly taking sides. James seems to have something like that in mind. In particular, his considerations on ‘forced options’ (see above, 1.2.) and some of his examples point in that direction. For example, he speaks of a man waiting indefinitely before proposing marriage to a woman because he is not perfectly sure that she will turn out to be an angel. The point James tries to raise here is that this man implicitly makes a decision. His indecisiveness has the same result as believing that she will not turn out to be an angel. That is the point of his rhetorical question: ‘Would he not cut himself off from that particular angel-possibility as decisively as if he went and married someone else?’ [33]

As we have seen, in some circumstances being indecisive on certain religious claims amounts to implicitly making a choice—viz. against religion. An example that is particularly pertinent in the Christian tradition is the following. How does one decide when confronted with Jesus’ claim that whether one believes in Him as being Christ the Son of God will decide one’s future fate? The way one responds to Jesus’ claim decides the action of the Father. This idea is prominent in the Johannine gospel and is frequently used in the existentialist tradition. For example, Bultmann uses it to underpin his ‘present eschatology’, i.e. the idea that the fate of the believer will not be decided in some distant future but, rather, ‘here and now’—viz., in the confrontation with the kerygma. The question whether she believes in it or not will decide the believer’s future fate. [34]

The reason why I bring up these examples here is that they make clear that, in certain cases, one cannot remain neutral. You cannot remain neutral with regard to Jesus’ claim that He is Christ the Son of God. Refraining from making a decision when confronted with this claim is tantamount to rejecting it, according to (Bultmann’s interpretation of) the Gospel of John. In short, in certain cases of religious claims refraining from decision-making amounts to implicitly making a choice—viz. in favor of unbelief .

Or, take another example that abstracts from specifically Christian concerns. Let us reconsider the above provided idea that the believer is (significantly) better off than the unbeliever. Remaining indecisive on religious claims pertaining to this idea implies logically a decision against the truth or probability of this idea. This is the case when holding certain religious beliefs doesn’t cost one very much, say, does not pose serious restrictions on one’s everyday conduct.

Disbelieving a claim the acceptance of which does not cost you very much, but the rejection of which could result in your not obtaining eternal happiness, would be highly imprudent. In order to do that in a prudent fashion, you will have to, at least, tacitly presuppose that those beliefs are definitely false, highly unlikely or something of that sort. [35] As not deciding on the question whether the global warming is caused by the Greenhouse effect [36] implies accepting the belief that this is not the case, not deciding on the religious claim that the believer is (significantly) better off than the unbeliever implies rejecting it. Because if there was any realistic chance that the Greenhouse effect was indeed the cause of the global warming or that the religious beliefs under consideration were true, it would be imprudent or even irrational to refrain from deciding on either issue.

Whatever the details of the argument, however, my main point is that in all those cases (and many more can easily be construed), we are confronted with situations in which refraining from making a decision on religious claims amounts to implicitly rejecting them. Since the condition that refraining from decision-making amounts to making an implicit decision is another ingredient of the second feature of situations in which applying quiescent evidentialism is inappropriate, I conclude that the cases provided are good candidates for being situations in which applying quiescent evidentialism is inappropriate. Again, whether they truly are depends on the question as to whether the first condition is met at the same time.

Concluding this point (3.2.2.2.2), I would like to propose that certain religious claims are decision-situations in which refraining from deciding is either too costly or, else, amounts to implicitly taking sides. Collecting the point I made in 3.2.2.2.1, I would like to propose as well that some religious claims are decision-situations in which the stakes are extremely uneven. As mentioned above, decision-situations in which refraining from deciding is either too costly or, else, amounts to implicitly taking sides and decision-situations in which the stakes are extremely uneven form the above named second condition for cases in which applying quiescent evidentialism is imprudent or inappropriate. Thus, those religious claims are (examples of) cases that make a good chance of being situations in which it is imprudent or inappropriate to apply quiescent evidentialism.

It will be remembered that the first condition for there being a situation in which it is imprudent to apply quiescent evidentialism is that there is not enough evidence to judge a given claim either way and there is no hope to collect sufficient evidence in the near future (see above, 3.2.2.1). If any of the religious claims named under 3.2.2.2, i.e. claims that meet the second condition, meets the first condition as well, then this claim exhibits the logical features of a situation in which applying quiescent evidentialism is imprudent or inappropriate. Since meeting the first criterion is not particularly difficult for religious claims – think e.g. of God being totaliter aliter or the Kantian rationale for placing religious claims outside of the limits of human cognition – it is not difficult to construe the claims that meet the second condition as also meeting the first criterion. And if they do, then they are situations in which applying quiescent evidentialism is imprudent or inappropriate.

In sum, in the case of all those religious claims for which, (a), there is not enough evidence either way and for which there is no hope to collect it in the near future and, (b) which are simultaneously decision-situations in which, (b.1.) refraining from deciding is either too costly or amounts to taking sides implicitly or, (b.2.) in which the stakes involved are extremely uneven, quiescent evidentialism should not applied. It would be imprudent, inappropriate, irrational etc. to apply quiescent evidentialism in those cases.

If the evidentialist challenge to religion is based upon a truth-oriented evidentialism, it is rather uninteresting. If, on the other hand, it is based upon quiescent evidentialism, it is interesting but irrelevant for a large number of cases of religious claims. Moreover, as soon as it assumes universal pretensions, it is plainly mistaken. But being either uninteresting or irrelevant/mistaken, the evidentialist challenge of religion should not be taken as seriously any longer as it usually is. Believers, philosophers of religion and other people engaged with the pursuit of and/or reflection on religion should not spend too much time bothering with the evidentialist challenge. Rather, they should spend their time on more important issues. [37]