Chapter 4

 

THE WILL TO BELIEVE

            This is rightly considered James's most distinctive and influential doctrine. Had James looked as goofy as Kierkegaard and the street boys in Cambridge been as mean-spirited as those in Copenhagen who followed Kierkegaard around and taunted him with the shouts of "Either or, Either or!" James would have been hounded by shouts of "Will to believe, Will to believe." He wouldn't have minded, because at least he would have known that, unlike his father, someone was reading his books. He expounded it with religious fervor throughout his career whenever there was the slightest pretext for doing so. It seemed to be innate in him, as were so many of his important doctrines. It gets defended in his first major philosophical publication in1878, at the end of each of his final two books, and in every major publication in between. When he was only 24 years of age, he gave expression to a proto-version of it in a letter to Oliver Wendell Holmes, Jr. "But as man's happiness depends on his feeling, I think materialism inconsistent with a high degree thereof, and in this sense maintained that a materialist should not be an optimist, using the latter word to signify one whose philosophy authenticates, by guaranteeing the objective significance of, his most pleasurable feelings." (CWJ, 4, 147) The budding idea is that one can be justified in believing some metaphysical world-hypothesis because of the beneficial consequences of so believing.

            This doctrine, which allows us to believe, or get ourselves to believe, a proposition upon insufficient evidence when doing so will have desirable consequences, works hand and glove with his foregoing accounts of will and belief in Chapter 2 and the conditions supplied in Chapter 3 under which they are free. In what follows, the qualification "or get ourselves to believe" will be dropped for the sake of brevity, and to simplify the discussion we can imagine that each person is able to self-induce a belief in any proposition, p, by ingesting a belief-in-p-inducing pill. This is a sure fire way of self-inducing belief, unlike that of acting as if you believe, which can be messy and chancy. (Since there are an infinite number of propositions, each person's medicine cabinet will be quite crowded and require that each pill, other than the one on the extreme left, be half the size of its immediate left-side neighbor, which is the spatial analogue to a "supertask.") Although belief is not identical with will, because they involve different psychic attitudes in my amended version of James, it resembles the willing state in having behavioral consequences as a result of established neural pathways from the concerned part of the brain to the motor organs. We are supposed to have the power, and the free power at that, to get ourselves, directly or indirectly, into various belief states and thereby bring it about that we overtly act in certain ways. Our free efforts to self-induce belief are part of the crucial x factor we contribute to the cosmic equation and which, like the want or presence of the nail in the parable, can make a crucial difference in the future course of history for good or ill. At a minimum, how we believe can have the most important consequences for our future flourishing as human beings.

            It is at this point that the doctrine of the will to believe enters the story. We have the power to control what we believe, be it directly or, so to speak, through popping a pill, and what we believe can make a decisive difference in whether or not we find the good life of self-realization. Now it often is the case that there is some proposition for which there is little evidence one way or the other but which would be beneficial for us to believe. James's most important example of such a proposition is that we have a contra-causal free will. It seems only reasonable that we should have every right to believe this proposition, in spite of W. K. Clifford's universal admonition that "It is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence." (LE 186) Given that the title of his essay was "The Ethics of Belief," it is clear that by "wrong" he means morally wrong. Were he to have meant epistemically wrong, he would have been uttering the empty tautology that it is epistemically wrong to believe anything upon insufficient evidence, that is, that it is epistemically wrong to believe.

            According to Clifford and his scientistic cohorts there is only one type of justification for a belief, that being an epistemic one based on evidence unearthed through an empirical inquiry or a proof if the belief is a mathematical one. To believe on any other grounds is a violation of our moral duty to measure up to our essence of rationality.[1] James, when he is espousing his will to believe doctrine, challenges Clifford's univocalist account of belief justification. At these times, he claims that there are two different ways to justify  believing a proposition: the epistemic way based on empirical evidence and proofs; and the pragmatic way based on the desirable consequences that accrue to the believer of the proposition. The former is directed toward establishing the truth of the proposition, the latter to establishing the desirability of believing that the proposition is true, quite a different matter.

            I added the qualification "when he is espousing his will to believe doctrine" because I will argue in the next chapter that James, like Clifford, has a univocalist account of belief justification, only it is the reverse of Clifford's, holding that the only justification for believing is pragmatic, based on maximizing desire-satisfaction, with epistemic considerations entering in only in a rule-instrumental manner as useful guiding principles. This should not surprise the reader, since it already has been established that James is firmly committed to both of these propositions.

1. We are always morally obligated to act so as to maximize desire-satisfaction over the other options available to us.

2'. Belief is a free action.

Thus, we are not only morally justified in believing a proposition on pragmatic grounds, we are morally obligated to do so.

            This immediately raises the question, if James was so firmly committed to a univocalist theory of belief justification why, when he was espousing his will to believe doctrine, did he accept the dualistic account of justification, and moreover, as will be seen, in a form that gives pride of place to epistemic over pragmatic reasons? The answer is that James was wisely following the principle of Minimal Ordinance enjoining us to use the weakest, least controversial premises that are required to establish the desired conclusion, since thereby we cut down on the risk of one of our premises being false or rejected by our opponent. James is sweating with conviction that if he only can get his audience to agree that they are morally permitted to believe certain propositions upon insufficient evidence, especially that God and free will exist, great benefits will accrue to them and their society. To win them over he does not need to use the very strong, controversial premise that the only justification for a belief is pragmatic, which almost everyone would reject out of hand on either Cliffordian or mutivocalist grounds. All that he needed was the weaker and therefore less challengeable premise that there are both epistemic and pragmatic modes of justification of belief, with the former taking precedent over the latter, thereby allowing pragmatic reasons to be appealed to only when epistemic ones are of no avail. For the time being, James is willing to work with a hierarchical dualism of justifications for belief that gives a dominant position to epistemic reasons, namely

4. We are always morally obligated to believe in a manner that is epistemically warranted, except when epistemic reasons are not available.

The "except when epistemic reasons are not available" qualification permits believing on the basis of the will-to-believe doctrine so as to maximize desire-satisfaction when epistemic reasons cannot be had. This goes a long way to placate those of Cliffordian leanings in his audience.

            James even argues for the weaker conclusion that we are morally permitted to believe upon insufficient evidence when doing so will have desirable consequences rather than the stronger one that we are morally obligated to do so, as is required by his casuistic rule.[2] He must have realized that many in his audience would not grant him his casuistic rule. At any rate, they shouldn't have because of the many powerful objections that were advanced in Chapter 1. This one problem at a time approach fits the spirit of James's claim that "sufficient to that day will be the evil thereof." (PP 286)

            Before we get down to the details of his account, two different sort of counter-examples to Clifford's universal prohibition will be presented so that they can be set aside as not relevant to James's concerns. First, there is the trust case in which a person is required to believe certain things about another person in virtue of having a special relation to that person, even when they lack adequate evidence for these beliefs. For example, spouses are morally required to believe in the faithfulness of the other person, even when they lack adequate evidence for this belief. Were they even to inquire into the matter, say by hiring a private detective, it would place them outside the trust relation and thus destroy the relationship. They have a duty not even to inquire into the matter. Of course they might get to a certain point at which they can't help but harbor suspicions; as Big Joe Turner used to sing, "You came home in the wee hours of the morning and your clothes didn't fit you right." It is then that the trust relation ends. James did not discuss the trust case, but there is no doubt that he would agree that it is a counter-example to Clifford; however, it is not a very telling one, because Clifford can easily protect his universal prohibition against it by building in an ad hoc restriction that excludes trust cases. Thus, trust cases are not telling counter-examples, because they are too easily localized.

            Much of the recent discussion of the will to believe has concentrated on extreme cases in which there is an overwhelming utilitarian justification for acquiring an epistemically nonwarranted belief, either because doing so prevents an horrendous outcome or brings about some exceedingly beneficial one. For example, an eccentric billionaire might publicly promise to donate a billion dollars to charity if Jones acquires the epistemically nonwarranted belief that Cleopatra weighed 109 pounds when she died, or some overwhelmingly powerful alien invader threatens to destroy the planet unless Jones acquires this belief. Again, James would readily grant that a counter-example has been unearthed but would find it of little interest for his purpose in formulating a will to believe doctrine. In the first place, "extreme cases" are, with very rare exceptions, counter-factual, and James, in general has little concern with merely possible cases in his analyses, being satisfied if his analysis fits the way things actually are. More important is that in the extreme cases the realization of good consequences or prevention of bad ones is completely external to the believer, being connected with his belief only via the intercession of a third person, the billionaire or alien in the examples. James's version of the will to believe is concerned with the personal or existential dimension of belief, the manner in which it changes a person's character and thereby their readiness to act in certain ways. There is a causal theory of value underlying James's will to believe doctrine, according to which the value of an outcome depends, at least in part, on how it is brought about. For this reason, extreme cases can be left out of the discussion of James's doctrine.

            James's intent is not just to produce some counter-examples to Clifford's prohibition. If it was, he could have availed himself of the rather obvious trust and extreme case counter-examples and been done with it. Rather, his aim is to spell out the conditions under which in general we are morally permitted to believe upon insufficient evidence. His dislike of doing philosophy in a formalistic, by the numbers, manner kept him from explicitly listing all these conditions in a neat package of numbered indented sentences, informing us which of them are necessary and which combination of them are sufficient for being morally permitted to believe without epistemic warrant. It will be argued that things go best for James if he is taken as giving a set of sufficient but not necessary conditions, although some individual members of the set are necessary. A lot of work is left to the expositor in extracting these conditions and determining their sufficiency and/or necessity. Thus there is considerable room for alternative interpretations, though there are some conditions that plainly are intended to be necessary.

            James begins his essay, "The Will to Believe," which is his most complete and forceful exposition of the doctrine, by explicitly listing three conditions that together comprise what he calls a "genuine option" to believe. A person's option to believe a proposition at a certain time is a genuine option just in case it is live, momentous, and forced. Because a genuine option is dependent on variable psychological factors, it must be relativized to a person at a time. What is live and momentous can vary across persons as well across different times in a single person's life.

            For a proposition to be live for person at a time, it must then be a real possibility for that person to believe it, as well as to believe its contradictory. His mind must not be made up one way or the other. Thus, the proposition, along with its contradictory, can be seriously entertained even if it cannot win at that time uncontested occupancy of the believer's mind and thereby qualify as a belief for James. To achieve this status will require some work on the part of the believer in the way of making efforts to attend or acting as if he believes (or popping the pill). James parries the capriciousness objection to his account of free will, according to which a person is just as likely to act out of as in character, given that our free choices are undetermined, by pointing out that the range of one's free will is limited to living options. Since a person of benevolent character cannot seriously entertain the thought of doing some sadistic act, it is not a live option for him

            A proposition is momentous for a person at a time when the consequences of his believing or not believing it will have very important consequences, relative to his personal scheme of values, supposedly at the time of his decision to believe rather than at some future time. That an option to believe is unique, pace what James says, does not alone qualify it as momentous. My one and only chance to see Barry Manilow live is his farewell concert tonight, but that alone hardly makes my option to see it momentous. Again, there can be wide spread divergences among persons and a single person at different times in respect to the momentousness of a given option.

            An option to believe is forced when the person will not wind up believing the proposition in question unless he decides to believe it. No one, such as a crazed brain surgeon or mad cyberneticist, is going to compel him to have this belief regardless of what he might do. It is completely up to him whether or not he acquires the belief. We might call the alternative in a forced option that the chooser winds up with if no decision is made the "negative alternative." Dated options, such as to accept a proposal of marriage by midnight tomorrow or never see the man again, are good examples of forced options. A forced option, like a unique one, need not be momentous, which would be the case if this proposal were offered to a lesbian.

            Several supposedly astute and fair commentators have interpreted James as holding that the three conditions for having a genuine option to believe are together sufficient for having a moral right to believe upon insufficient evidence. This will be shown to be a terrible distortion of the text. Once this perverse strawman version is given, they have an easy time denigrating it as the will to gullibility or wishful thinking. It licenses me to believe the propositions that I am the Sultan of Wisconsin, the inventor of the sandwich, and the author of The Critique of Pure Reason, if my psychology is such that they are live and momentous for me and my option to believe them forced. James would be the last person to agree that the pleasure I derive from believing them justifies my doing so.

            The original culprit was Dickinson Miller: "The particular class of cases [for James] in which we have 'a right to believe' was that in which the option before our mind was (1) 'living', (2) 'forced', and (3) 'momentous'."[3] (PA 286)  A. J. Ayer, in a similar vein, wrote, "To claim [as did James] the freedom to believe whatever one chooses may be emotionally satisfying, but I should hardly call it rational."[4] (OP 191) Marcus Peter Ford, who is a very sensitive and imaginative interpreter, wrote: "The 'right to believe' only pertains when a decision is 'forced' and 'momentous' and when it concerns 'live options.'" (WJ 32) What is overlooked by these  interpretations are the further restrictions that James places upon a will to believe option. They not only are stated explicitly but also inform all of his examples, making it hard to understand how any attentive reader could miss them.

            Of these additional conditions, the one that is most important for responding to the wishful-thinking objection is the requirement that the chooser cannot determine at the time of his decision the truth-value of the proposition in question on epistemic grounds. "Our passional nature not only lawfully may [is morally permitted to], but must decide an option between propositions, whenever it is a genuine option that cannot by its nature be decided on intellectual [epistemic] grounds," and, "In concreto, the freedom [moral permission] to believe can only cover living options which the intellect of the individual cannot by itself resolve." (WB 20 and 32) Sometimes James presents the epistemic requirement in a weaker way that would license believing even in the face of significant evidence against the belief. He speaks of "the right to believe in things for the truth of which complete objective proof is yet lacking" and voluntary choice being permitted when "objective proof is not to be had." (MT 138-9 and PP 1177. my italics) Herein the available evidence need not be neutral between the truth or falsity of the proposition. Given that James wants to win over his audience by giving the Cliffordian types as much rope as possible, he ought to go with the stronger version, especially since it is satisfied by his most cherished examples -- belief in free will and the good destiny of the world. For these reasons, it is the one that will be operative in what follows. It should be born in mind, however, that James's real position, the one he keeps in the closet when he is trying to win over the Cliffordians, is that if the option is of overwhelming momentousness to the believer, such as Kierkegaard's option to believe in Christianity, the weaker version is operative: One is permitted to believe in the teeth of quite powerful contrary evidence.      

            When James says that the intellect of the person cannot settle the matter, the "cannot" can be of either the in principle or weaker in practice sort. In practice human neurophysiologists cannot verify that our efforts to attend are not causally determined by brain events due to limits in their powers of mensuration, but it is in principle possible that they do so, for they, along with their instruments, could suddenly make like the incredible shrinking man.[5] It is important to realize that the epistemic undecidability concerns the chooser alone, not other observers. It is conceptually impossible that the chooser, at the time he is deliberating, verify what choice he will make and thereby the occurrence of any event for which his choice is a necessary cause, since he can deliberate only if he is in ignorance of what he shall choose; someone else, however, could verify at that time what his choice will be by appeal to a well founded inductive argument based on his past track record. 

            It must be stressed that James's epistemically-undecidable-by-the-chooser-before-the-choice-is-made requirement is not the weak requirement that at the time of the choice the chooser, as a matter of fact, lacks adequate epistemic grounds or evidence for determining the truth-value of the proposition, which could be realized if the chooser made a point not to investigate the matter, like Clifford's dishonest ship owner who makes a point not to investigate whether his ship is safe to send to sea. Rather, it is the strong requirement that the chooser lacks adequate evidence one way or the other after discharging his epistemic duty to perform all of the relevant inquiries. Laziness, especially self-interested laziness, will not enable the chooser to satisfy the epistemic-undecidability requirement.

            Another question concerning epistemic undecidability that must get resolved concerns whether there is a single method or criterion for determining when a proposition is epistemically warranted that is common to all propositions, regardless of their subject-matter. There are contemporary language-game fideists who would contend that each language-game or doxastic  practice -- normative rule governed human practice for warranting beliefs -- such as a specific organized religion or one of the branches of science, has its own unique ontology, along with its own criteria of rationality and epistemic warrant for a belief. The criteria that inform the different practices are incommensurable in the sense that each practice is a self-contained normatively rule-governed activity that is immune from criticism from without. Whether or not James agrees with these Wittgensteinian language-game fideists does not admit of a simple yes or no answer. Chapter 7 will show that in regard to the ontology part of their thesis he is in agreement, for what is real for James is relative to the assumptions of a given practice or perspective, but in regard to the epistemic part he is not with them. The ontological assumptions of different doxastic practices differ but not their epistemic procedures for warranting belief. Since a choice cannot be made between them by use of their shared epistemic procedures, we are morally permitted to do so in the will-to-believe manner. Not only is there no indication in the text that James believed that criteria for epistemic warrant varied across practices, there is good evidence, based on his attempt to show that existential claims based on mystical experiences are subject to the same checks and tests as are those based on ordinary sense experience, which is a subject for Chapter 10. He even said that "Science, metaphysics and religion...form a single body of wisdom, and lend each other mutual support." (SPP 20) Also, in regard to the seeming opposition between common sense, science, and philosophy, James said that "There is no simple test available for adjudicating offhand between the[se] divers types of thought." (P 93) He did not say that each has its own unique epistemic procedures for warranting belief.

            James not only explicitly states the requirement for epistemic undecidability for a will-to-believe option, all of his many examples satisfy it, thus making it all the more amazing that it could have been missed by some many commentators. The stranded mountaineer who must jump across a chasm to get to safety and can increase his chances of succeeding  if he believes that he has the capacity to do so, is not able, at that time, to epistemically determine whether he has the capacity, though he can do so after he has attempted the leap, though, if he misses, he'll have to do it very quickly. The person who psyches himself up to lead the morally strenuous life by believing the proposition that good will win out over evil in the long run cannot epistemically determine its truth-value, since, in addition to it being beyond our capacity to predict the direction history will take in the long run,[6] the choices that he shall make, which is the crucial x factor that he contributes to the cosmic equation, are not knowable by him in advance of his choices. In contrast to these Jamesian cases, my epistemically nonwarranted beliefs about being the Sultan of Wisconsin and the like are both in principle and in practice epistemically determinable by me now; and, furthermore, I have violated my epistemic duty by not performing adequate inquiries -- to which my response is that since I am the Sultan of Wisconsin, I have lackies to do the drudge work.

            Four conditions for a will-to-believe option have been unearthed so far from the text. A person, A, is morally permitted to believe a proposition, p, at a time T without adequate epistemic warrant if A's option to believe is (1) live, (2) momentous, (3) forced, and (4) A cannot epistemically determine at T the truth-value of p. The text gives some but not decisive reasons to think that James required, in addition to (1)-(4), that A's believing p helps to make p become true: "There are, then, cases where a fact cannot come at all unless a preliminary faith exists in its coming. And where faith in a fact can help create the fact, that would be an insane logic which should say that faith running ahead of scientific evidence is the 'lowest kind of immorality' into which a thinking being can fall," and "In truths dependent on our personal action, the, faith based on desire is certainly a lawful and possibly indispensable thing." (WB 29 and 29) In the 1878 "Some Considerations of the Subjective Method" he wrote that the subjective method, the name he then used for the will to believe doctrine, "can only be harmful, one might even say 'immoral' if applied to cases where the facts to be stated do not include the subjective R as a factor," in which R is the contribution that we make through the actions caused by our belief. (EPH 335)     

            What is not completely clear from these quotations is whether the belief-helping-to-make-true condition is intended to be necessary for any will-to-believe option or is a feature of only some cases of this option. If the latter interpretation is given, as it is by Elizabeth Flower (PA, I, 686), Robert O'Connell (WJ 72) and James Wernham (WB 14), a person would be morally permitted to believe a proposition if (1)-(4) were satisfied, even though his believing made no difference to whether or not it is true, as would be the case with his belief that there are Platonic Forms or that the God of traditional Western theism exists. If his belief gives him pleasure, why deny him this, even if his pleasurable belief state is completely disconnected from activities in the workaday world?

            I favor the former interpretation. First, two of the quotations that were given seem to support it. Furthermore, every example that James gave throughout his career of a will-to-believe option involved a belief that played a causal role in helping to make the believed proposition become true or, as will be shortly seen, some other desirable proposition. The Alpine and the good-will-win-out cases clearly satisfy the causal condition, as do all of his confidence building cases -- the Alpine leaper, the ardent but somewhat unconfident suitor, the person who wants to be liked, et. al. The reason James became so preachy whenever he discussed free will is that he thought that if we believed we had free will we would make the sort of all-out effort to attend to an idea in difficult cases of conflict that would help to bring it about that we do have free will. "If...free acts be possible, a faith in their possibility, by augmenting the moral energy which gives birth them birth, will increase their frequency in a given individual." (WB 84) It is very much in the spirit of James's promethean pragmatism to have a causal requirement, since the significance and value of belief in general are the worldly deeds to which it leads, a thesis for which, as we saw, James gave both physiological and normative reasons. Remember how he railed against sentimentalists and aesthetes, including his own brother. Merely to be in a pleasurable belief or esthetic state is not its own justification. There must be some behaviorally rooted reason for choosing to get oneself into such a state. Thus, it would violate the promethean spirit of James's pragmatism to justify an epistemically nonwarranted belief solely in terms of its being a pleasurable belief state. Furthermore, by having a causal requirement James has yet another way, in addition to having the epistemic undecidability requirement, to protect his doctrine against the wishful thinking objection. My ill-founded pleasurable delusions of grandeur were ones that my believing could not help to make true.

            Assuming that James required that causal condition, there are some questions that need to be answered. First, must P's belief be causally sufficient, necessary, or both for making p true? The answer is none of the above. James's examples, especially the good-will-win-out one, involve only the very weak requirement that P's belief can help to make p true.[7] The Alpine leaper's belief, obviously, does not have to be either causally sufficient or necessary for his leaping successfully; for him to be permitted to acquire the belief; it is enough that his believing increases the probability of success. This is all that common sense, as well as James, requires. Furthermore, this weak interpretation dovetails with James's insistence that a very minor difference in the initial conditions can make the crucial difference in the final outcome, as in the parable of the nail. For these reasons I will include the weak version of the causal condition as a fifth requirement for a will-to-believe option.

            Adopting this requirement requires, in turn, adopting yet another condition, namely, that it is desirable that the believed proposition become true. This requirement is assumed throughout his discussion, for James's will-to-believe justification is a substitution of this argument form:

                        Doing x helps to bring it about that p.

                        It is desirable that p. Therefore,

                        It is morally permissible to do x.

in which "believing p" is substituted for "x" throughout. Notice that the moral permission in the conclusion is not qualified as prima facie, which would make it subject to defeaters or overriders. If we knew only that p's being true would satisfy some desires but not that it would maximize desire-satisfaction, then, if we accept James's casuistic rule, we would have to make this prima facie qualification. Since James's casuistic rule was rejected at the end of Chapter 1 in favor of a vague, mixed bag rule that included deontological factors along with desire-satisfaction, we will interpret "desirable" according to the latter criterion, as we had agreed to do. Thus, even if p's being true were to maximize desire-satisfaction over its being false, this alone would render it only prima facie morally permissible to believe p. Given James's strong commitment to deontological ethical principles in many of his writings, though not in "The Moral Philosopher and the Moral Life," James would agree to this and thus would accept my deontological revision of his casuistic rule.

            To illustrate this, consider the case in which I promise to give Jones a revolver but in the interim he turns into a homicidal maniac and has vowed to kill Smith.

  My giving Jones a revolver helps to bring it about that I keep my promise.           

  That I keep my promise is desirable. Therefore,

  I am morally permitted to give Jones a revolver.

Obviously, the permission must be qualified as only prima facie, for were I to keep my promise and give him a revolver it would result in the death of the innocent person Smith. This deontological defeater is not overruled if it would maximize desire-satisfaction to give Jones the gun: Smith could be a widely disliked person whose death would maximize desire-satisfaction over his continuing to live.    

            At this point it will help the reader to pause for a recap of the six conditions that have so far been unearthed from the text for a will-to-believe option. Person A is morally permitted to believe proposition p without adequate epistemic warrant at a time T if (only a sufficient condition is being given) the option to believe p is: (1) live; (2) momentous; and (3) forced for A at T;  (4) A cannot epistemically determine p's truth-value; (5) A’s believing p helps to bring it about that p; and (6) it is, all things considered, desirable that p become true.

            In fairness to those commentators who have not included the causal requirement for any will-to-believe option it must be pointed out that sometimes James stated his doctrine in a way that didn't justify believing an epistemically undecidable proposition but only adopting it as a working hypothesis, as we do in science when we select some untested hypothesis as a working hypothesis for the purpose of setting up experiments or in everyday life when we simply act as if the proposition were true. It will be shown that the causal requirement is not applicable to some working hypothesis cases.

            The following quotations clearly speak for the working hypothesis version. Some of them simply identify belief or faith with the adoption of a working hypothesis: "Faith is synonymous with working hypothesis," and "To sum up, faith and working hypothesis are here one and the same." (WB 79 and EPH 337) Clifford's prohibition is now interpreted as prohibiting adopting as a working hypothesis (acting as if you believed) an epistemically nonwarranted proposition rather than believing it. "Suppose that, having just read the 'Ethics of Belief,' I feel it would be sinful to act upon an assumption unverified by previous experience..." (WB 80) At the beginning of "The Dilemma of Determinism," James states his intention "to induce some of you to follow my own example in assuming it [the doctrine of Libertarian free will] true, and acting as if it were true." (WB 115) Among the actions to be performed is publicly declaring that it is true, even though you do not believe what you are saying. His presentation of the "faith ladder" in the 1905 "Reason and Faith," which is the final form taken by his will to believe doctrine, says that we are to treat the proposition we desire to be true "as if it were true so far as my advocacy and actions are concerned." (ERM 125) This very same faith ladder gets repeated at the end of each of his final two books. (PU 148 and SPP 113)

            James's examples also reflect his sloshing back and forth between the belief and working hypothesis versions of his doctrine. In "Some Considerations of the Subjective Method" he uses the example of the Alpine climber who must leap a chasm to get to safety. (EPH 332) What is required here to increase his chances of success is good old-fashioned sweating with conviction belief that he has the capacity to succeed, not just the adoption as a working hypothesis that he does. But in the final paragraph of "The Will to Believe" he has an Alpine example of a climber who is confronted with alternative paths and must pick one of them if he is to save himself from freezing to death and has no reason to prefer one of them over the others. Obviously, he must pick one of them and journey along it, for to make no choice assures his death. The chances of his success are in no way increased by his believing that the chosen path is the right one. He only has to adopt it as a working hypothesis and thereby act as if it is the right one by following it.

            The working hypothesis version of the will to believe does not require in general a causal requirement. James himself recognized this when he wrote "And your acting thus [as if you believed] may in certain special cases be a means of making it securely true in the end." (PU 148. my italics) The "in certain special cases" qualification implies that in some cases adoption of a proposition as a working hypothesis does not help to make it true. By adopting a proposition as a working hypothesis a scientist does not help to make it true but only helps to discover that it is true. But "in certain special cases," such as adoption of the hypothesis of Libertarian freedom it does help to make it true, since by acting as if we were free in this sense we help to bring it about that we are. James recognized cases in which believing, in the full-blooded sense, can help in discovering that a proposition is true. By believing that God exists, we increase the chances that we shall have apparent direct nonsensory perceptions of God, such experiences, as Chapter 10 will bring out, counting as evidence for God's existence.[8]

            It is plain that James operated with two quite different versions of the will to believe. They are different because believing isn't the same as acting as if you believe or adopting as a working hypothesis. James recognized this when he gave us the act-as-if-you-believe causal recipe for self-inducing a belief, since a cause cannot be identical with its effect. But why did he keep two sets of books? It cannot be explained in terms of an evolution in his thinking, the belief version being his earlier formulation and the working hypothesis version supplanting it from 1905 on, for he ran them together throughout his career. For example, "The Will to Believe," which gives prominence to the believing version because of its heavy reliance on confidence and courage building cases that require real, sweating-with-conviction type belief, also endorses the working hypothesis version.[9]

            The most likely explanation is in terms of James's attempt to win over his audience by being as accommodating as possible, but his weak, working hypothesis version goes too far in that direction, so far that it trivializes his dispute with Clifford. Clifford, being a man of science, is the last person who would want to issue a prohibition against using working hypotheses, for he must have been aware of their fruitfulness in furthering the progress of science. James is abusing the principle of Minimal Ordinance that enjoins us to use the weakest premises possible to support the desired conclusion, not to replace the desired conclusion with one that trivializes what is at issue. Fortunately, there was no need for James to engage in the trivial exercise of running through open doors, for his exciting, belief version, when strengthened so as to meet certain objections, which shall now be considered, is quite formidable.

 

Objections

            The first objection comes from that tireless critic of the will to believe, Dickinson Miller. The chooser is supposed to get himself to believe a proposition that he himself takes to be evidentially nonwarranted, but this is impossible. "You cannot believe and yet in the heart of that very belief be heroically facing the uncertainty of your whole position. Your state of mind would not be belief, which is regarding something as fact, not as uncertain." (PA 288) Miller is right that to believe a proposition is to believe it is a fact, for a fact is a true proposition and one cannot believe a proposition without believing that it is true; but, pace Miller, that does not require that it is believed to be certain in the sense of supported by overwhelming evidence. That there are so many anti-rationalistic theists of the Kierkegaardian variety shows Miller's inference from take to be true (or a fact) to take to be certain to be bogus. Miller is not alone in his mischaracterization of belief. Many contemporary philosophers, such as Richard Swinburne in Faith and Reason, wrongly claim that to believe a proposition is to believe that its probability is greater than one-half relative to the available evidence.

            There is a close cousin to Miller's objection that might fare better. The point is not that one cannot believe without believing to be certain, but rather that if one believes what he takes to be evidentally unfounded, he will not, pace James, have his confidence and courage boosted so that he can act more effectively in making the believed proposition true. The wrong response to this objection is to find some procedure for making the believer forget that he acquired his belief on the basis of a will to believe option; there could be a second set of pills, neatly arranged in descending order of size from left to right on the second shelf of the medicine cabinet, such that after one has popped a belief-inducing pill he pops the appropriate one from the second shelf that makes him forget the nonrational means by which he acquired this belief and instead implants in his mind the false apparent memory of having acquired it after a successful empirical inquiry. The problem with this way around the objection is that the believer must deceive himself, which is bad enough, but in the process destroys his own integral unity and winds up as a divided, schizophrenic self. James's promethean quest to have it all will be found in Chapter 11 to have this deleterious consequence and ways will be devised to attempt to escape it. The ideal of an integrated, rational self is a powerful one that deserves more respect than is accorded it by this drastic solution.[10]

            A better response is that human psychology is far more variable than this objection envisions. Although it is true that there are some people who are so constituted psychologically that they cannot realize the confidence building benefits from a belief that they take to be evidentially nonwarranted, there are many people whose psychology permits them to do so, such as our nonrationalist theists. It has already been seen that a will-to-believe option is relative to a person at a time because human psychology is variable in regard to which propositions a person takes to be live and momentous belief options. All this objection shows is that there is another psychological reason for relativizing a will to believe option to a person at a time. A  sixth condition could be added requiring that

(7) A's psychology at T is such that he can realize the confidence and courage boosting benefits of a belief that p, even if he takes p to be evidentally nonwarranted.

            James's will to believe is acquiring more and more epicycles, but I will not pause at this point to give an explicit mounting of them, since even more are to come in response to yet other objections. This next objection comes from James himself. Why, he asks, can't someone act so as to help to make some desirable proposition become true without actually believing it? His response:

Since belief is measured by action, he who forbids us to believe religion to be true, necessarily also forbids us to act as we should if we did believe it to be true. The whole defence of religious faith hinges upon action. If the action required or inspired by the religious hypothesis is in no way different from that dictated by the naturalistic hypothesis, then religious faith is a pure superfluity, better pruned away, and controversy about its legitimacy is a piece of idle trifling, unworthy of serious minds. (WB 32)

James even goes so far as to claim that there is no behavioral difference between suspending belief in R and actually disbelieving it. "We cannot escape the issue by remaining sceptical and waiting for more light, because, although we do avoid error in that way if religion be untrue, we lose the good, if it be true, just as certainly as if we positiviely disbelieve." (WB 26) The agnostic must act "meanwhile more or less as if religion were not true." (WB 29-30)

            This is a disastrous response. By the "religious hypothesis" James here means the proposition that

R. Good will win out over evil in the long run.[11]

He claims that because belief is measured by action, a person will act so as to help make R become true by leading the morally strenuous life if and only he first believes that R is true. Such benevolent behavior is "dictated," "required," or "inspired" by R. As an empirical generalization about human psychology, this is false, since we know of many people who do not believe R but nevertheless lead the morally strenuous life.[12]

            Underlying James's response is the false assumption that for every proposition, p, there is a set of actions, B,  such that a person believes p if and only if he performs or is disposed to perform the actions in B. James even went so far as to claim that there is no behavioral difference between suspending belief in R and actually disbelieving it. This assumption fails to do justice to the psychological variability among persons in respect to how their beliefs mesh with their actions. Two persons can believe one and the same proposition but act in radically different ways. Both could believe R but only one of them act so as to help make it true. The person who sits on the sideline might be made overconfident by his belief in R and think that his active participation on the side of the good is not needed or he might have devilist leanings and not want to see R become true. James's assumption of a one-to-one correlation between belief and action is not able to distinguish between believing the factual proposition that R is true and believing the normative proposition that it is good that R is true. The person who acts so as to help make R true could believe the latter but not the former.

            There is an easy way around this difficulty that consists in building in yet another epicycle concerning the way in which a will-to-believe option must be relativized to a person's psychological makeup at the time of the choice, namely

(8) A knows at T that he will act so as to help make p become true only if he first believe that p is true.

Why must A know this fact about his own psychology? The reason is that the conditions for having a will-to-believe option are supposed to justify A's believing or acquiring the belief that p. But what justifies a belief gives the believer a reason for so believing, something that he could give in response to the challenge to justify his belief. This requires that he be aware of this reason or justification. This seemingly innocent point will be the basis for the next objection.

            Imagine that A satisfies conditions (1)-(7) for having a will-to-believe justification for believing that p is true. Among his reasons for this belief is that only by so believing can he act in a way that will help to make this desirable proposition become true. Thus, if we ask A why he is toiling to help make p become true, among the reasons he will give is that p will in fact become true. But that's an irrational reason for trying to make p true. A relevant reason would be that it is good that p become true. Because you believe that Jones will succumb to his cancer of the liver hardly gives you a reason for acting so as to help bring this about.

            Because A has such an irrational reason for acting so as to make p true, he does not do so as a rational, morally responsible agent, and, since rationality is a necessary but not sufficient condition for acting freely, he does not do so freely. This is a very serious matter, especially for the likes of James who prize so highly being a free, morally responsible agent. Just recall the emotional breakdown of 1870 that was occasioned by his doubts that he was such an agent. The problem takes an especially virulent form with a will-to-believe based belief in R, given the very extensive nature of the actions and dispositions that are caused by this belief. Whatever good might be realized by A's irrationally acting so as to make R true is outweighed or defeated by his loss of or diminution in his freedom and moral responsibility. At least these are my deontological intuitions and James's as well, I believe.

            The irrationality objection, devastating though it is, is easily neutralized. All that is required is to separate the proposition that A must first get himself to believe from the one he thereby helps to make true. Thus, (7) must be revised as

(8') A knows at T that he will act so as to help make q become true only if he first believes that p is true, in which q is not identical with p,

By separating p from q, James can give a will-to-believe justification for believing in good old-time theism, not just his pale moral substitute for it, R. A's psychology at T could be such that he will act in the proper good-making fashion so as to help R become true only if he first believes that the God of traditional Western theism exists. Herein he would have a prudential reason for acting benevolently, since he believes that God will reward him for doing so. This may not be an admirable reason but it nevertheless is a rational one. In the case in which a belief that God exists increases the believer's chance of gaining evidence that God exists for his religious experiences, the believer helps to bring about the desirable proposition that there is evidence for the existence of God by believing that God exists.    

            There is some textual evidence that James was on to the need to separate the believed proposition from the one that it is desirable to make true, for he often formulated his will-to-believe option in a way that separated them. Sometimes, though not in "The Will to Believe," he separated them in the confidence building cases. The proposition that the Alpine leaper must believe to increase his chances of leaping successfully across the chasm is not the categorical proposition that he will successfully make the leap but instead that he has the capacity to do so, which is the conditional proposition that if he were to attempt the leap, he would succeed. This is the proposition that he first must believe in order to increase his chances of bringing it about that he successfully makes the leap. When James says that "I wish to make the leap, but I am ignorant from lack of experience whether I have the strength for it" or the "ability for [this] exploit," he is making use of this conditional proposition. (EPH 332. my italics) The leaper's belief that he has the capacity to succeed, unlike the belief that he will in fact succeed, is a rational reason for attempting to make it true that he leaps successfully. The you-will-like-me case admits of the same resolution. A first gets himself to believe the conditional proposition that if he acts in a friendly manner, people will wind up liking him so that he can muster the necessary courage and confidence to act in a friendly manner and thereby help to bring it about that people will wind up liking him. His conditional belief is a rational reason, though not the sole reason, for his acting in a friendly manner. Among the other reasons must be the desirability of being liked by people.

            James flipflops in his manner of stating the common denominator of all religions. In "The Will to Believe" he gives this categorical formulation: "The best things are the more eternal things, the overlapping things, the things in the universe that throw the last stone, so to speak, and say the final word," which I paraphrased as

R. Good will win out over evil in the long run. 

This is the proposition that A first must believe so as to act in the sort of  good-making manner that will help to make R become true. This interpretation is nailed down by his claiming with respect to propositions like R that "There are, then, cases where a fact cannot come at all unless a preliminary faith exists in its coming." (WB 29) Herein the proposition that first must be believed and the one that is to be made true via the belief are identical.

            But in his other writings he gives a conditionalized formulation of religion to the effect that

R'. If we collectively exert our best moral effort, then R (Good will win out over evil in the long run).

For instance, he writes: "Suppose that the world's author put the case to you before creation, saying: 'I am going to make a world not certain to be saved, a world the perfection of which shall be conditional merely, the condition being that each several agent does its own 'level best.'." (P 139. my italics) This conditionalized formulation of the religious hypothesis gets repeated at two places in his lecture notes: "Meanwhile I ask whether a world of hypothetical perfection conditional on each part doing its duty be not as much as can fairly be demanded," and pluralism holds that "the world...may be saved, on condition that its parts shall do their best." (ML 319 and 412) What A does, accordingly, is to get himself to believe R' so that he will act in the morally strenuous that will help to make R become true. Similar considerations apply to a will-to-believe based belief in a metaphysical or world-hypothesis. One believes in theism so that they can get themselves to live in some desirable way.

            Although (8') goes some way to neutralize the irrationality charge, it has to go even further. According to (8'), among A's reason for helping to make the desirable proposition q become true is that p is true. But his psychological makeup at T could be bizarre so that his belief that p is true, although a causal factor in his acting so as to make q true does not constitute a rational reason for so acting.[13] For example, p could be the proposition that Verdi wrote Ernani and q be R, and his psychology be such that he will act so as to help make R become true only if he first believes p. Remember the "Niagara Falls!" routine. Thus, when he is asked why he is living the morally strenuous life so as to make R true, he will respond that it is because Verdi wrote Ernani, thereby rendering his action irrational and thereby not one for which he is morally responsible. Plainly, yet another epicycle is required, namely,

(9) A's belief that p is a rational reason for him to act so as to help make q become true.

            It is now time to pause and give an explicit recap of all the many conditions that together are sufficient for being morally permitted to believe upon insufficient evidence. A is morally permitted at time T to believe an epistemically nonwarranted proposition, p, for the purpose of helping to make true another proposition, q, IF

A's option at T to believe p is:

  (1) live;

  (2) momentous; and

  (3) forced.

And furthermore:

(4)  A cannot epistemically determine at T the truth-value of p;

(5)  A’s believing p can help A to bring it about that q;

  (6) It is, all things considered, desirable that proposition q become true;

  (7) A's psychology at T is such that he can realize the confidence and      

        courage boosting benefits of a belief that p, even if he takes p to be   

        evidentally nonwarranted;

  (8') A knows at T that he will act so as to help make q become true only if  

         he first believes that p; and

  (9) A's belief that p is a rational reason for him to act so as to help make q

        become true.

            I have italicized "sufficient" and put "IF" in block letters to emphasize that conditions (1)-(9) are together taken to be sufficient but not necessary for a will-to-believe option. The reason for not affirming the necessity of (1)-(9) is to avoid the following universalizability objection. Imagine that A has a brother, B, whose psychology exactly resembles his except that A alone satisfies condition (8’) requiring that the believer knows at T that he will act so as to help make q become true only if he first believe that p. Because B is sufficiently strong-willed that he does not need to have the confidence or courage building belief in p in order to do his best to make q become true, he is not morally permitted to believe p upon insufficient evidence whereas the weaker-willed A is.[14] This violates the principle of universalizability -- if A is morally permitted (or forbidden) to perform an action in a certain set of circumstances, then everyone in like circumstances is morally permitted (or forbidden) to do the same. It is implausible to respond that B is not in the same circumstances as A, since he has a stronger will; for someone's being subject to a moral rule should not depend on whether he is weak-willed or cowardly. A similar problem would result if p were live (or momentous) for A but not B.       

            Because conditions (1)-(9) purport to give only a sufficient condition, there is a ready response to the universalizability objection. That B does not satisfy these conditions does not entail that he is not permitted to believe p; this would follow only if the conditions together were necessary. He can be accorded the same moral right to believe p upon insufficient evidence as is A, only he will not have to exercise this right because of his stronger character.

            Given that a major concern of this chapter is to see how James's will-to-believe can justify an epistemically nonwarranted belief in his doctrine of Libertarian free will, it will be instructive to see how conditions (1)-(9) apply to it. There are many people for whom it is a genuine option to believe the proposition that they possess this sort of freedom. Let this be proposition p. Because of limitations in our powers of mensuration -- we're too big and the brain events are too small --  it cannot be epistemically determined whether p is true or p is false, thereby satisfying (4). Let q be the proposition that we exert our best moral effort to attend to the idea of the morally good or right alternative in a case of moral temptation. Certainly, it is desirable that q be true, thus satisfying (6). There are persons, William James for example, whose psychological makeup is such that they can believe that q is epistemically undecidable and yet have their confidence and courage raised by their belief in p, as (7) requires. By raising their confidence level they are able to exert themselves in a way that will help to make q become true, thus satisfying (5), and furthermore know this fact about themselves, in accordance with the demands of (8’). (9)'s conditions are met because their belief that p is a rational reason for their acting so as to make q become true. A good reason for trying to get yourself to attend to a difficult idea in a moral conflict case is that you have the Libertarian sort of free will to do it. Believing that they have free will is both a necessary cause and a rational reason for their attempting to exert themselves to attend to the idea of the morally right alternative in a case of moral conflict .

 

Summation

            I believe that James's reconstructed will-to-believe doctrine in terms of the jointly sufficient set of conditions (1)-(9) for believing upon insufficient evidence is a formidable doctrine that deserves our respect if not our whole-hearted agreement. No doubt, if we put our minds to it, we could think up additional objections, but there is reason to be optimistic that they can be met, maybe by adding on some more epicycles. Some, including William James, might not be happy with so many epicycles. His remark that "The over-technicality and consequent dreariness of the younger disciples at our American universities is appalling" might be directed against my reconstructed version. (PU 13) If I were to be charged with making a crashing bore out of his exciting doctrine, my response is that I am not to be blamed because the universe is so complex. The price that he must pay for having a more defensible version of the will to believe is that he no longer can neatly fit it into his typical one hour lecture for a general audience. No more one night stands on the lecture circuit. He'll have to stick around the visited campus for at least a few days so he can give a whole series of lectures on the will to believe.

            I hope that I have adequately discharged my duty as a commentator to be both unsparingly critical and as sympathetic and constructive as possible. James's doctrine of the will to believe is one of the great contributions to the history of philosophy, and it has been my intent to show its great importance and resiliency. I have attempted to show that a slightly revised version of it is quite plausible, and, moreover marshalable in support of our justification for believing that we have his sort of contra-causal freedom of will.  And, when this is combined with the outcome of Chapter 2 -- that belief is an action -- we are justified in believing that

2'. Belief is a free action.

And when 2' is combined with James's casuistic rule

1. We are always morally obligated to act so as to maximize desire-satisfaction over the other options available to us.

it follows that

3. We are always morally obligated to believe in a way that maximizes desire-satisfaction over the other belief options available to us.

The next two chapters will explore the manner in which James utilized 3 in his analysis of belief-acceptance and truth.