Wednesday, June 23, 2010

William James, The Will to Be a Self-Deceived

Theists almost invariably proceed from study of arguments for the existence of God—the cosmological, ontological and teleological arguments—to focus on epistemology. Having failed to prove the existence of God given accepted standards of evidence, they change the standards of evidence. Their belief becomes rational by altering what it means for a belief to be rational. Unsurprisingly, when the standards are lowered for religious belief, they must either be lowered for every belief, which leads to obvious cases showing the new standards are too low, or that lowering must be limited to just theism and thus involve special pleading.

One philosopher who offers a lower standard for rational belief is William James. I started writing a long post on faith and reason and got distracted rereading William James’s “The Will to Believe”. This article, or an excerpt of it, appears in almost every introductory anthology on philosophy, and James is widely respected for his work both in philosophy and psychology. So, it is surprising that he would reason so poorly, but the fact is that James’s reasoning has no real merit and is just a rationalization for his preexisting religious belief.

The conclusion of James’s argument is that it is rational to believe in God or religion even though we do not have sufficient evidence that God exists. It is sometimes rational to believe even if you do not have sufficient evidence, and James thinks we do not have sufficient evidence for religious belief. Note that I will be discussing rationality of belief with insufficient evidence rather than rationality of belief when there is sufficient countervailing evidence. The latter is often part of religious belief, and such a view would be more problematic even than James’s, but it is not my target here.

After introducing the issue, James sets up W.K. Clifford as a foil who argues that “it is wrong always and everywhere to believe with insufficient evidence.” James points out that there are two distinct epistemic goals: acquiring true beliefs and avoiding false beliefs, and that Clifford focuses entirely, in this passage, on avoiding false beliefs and never considers the goal of acquiring true beliefs. This is a good point and argues in favor of stating Clifford’s evidentialist (as we call it) claim more positively as well as negatively. For example, one might follow David Hume’s language and say that we should “proportion our beliefs to the evidence”—for example, assenting when there is sufficient evidence and withholding assent otherwise. But that is not James’s strategy.

The irritant in reading James is that he cannot refrain from irrelevant ad hominem attacks on Clifford, James calls Clifford “nervous” and cites his “fear” of making errors and contrasts this fear with his own hopefulness. But the issue is not Clifford’s emotional state; the issue is whether it is responsible for one to believe something when there is insufficient evidence to think it’s true. More importantly, however, James’s remedy for Clifford’s oversight is far too broad: he counts beliefs as rational when it’s obvious that they should not be. Put another way, a serious consideration of Clifford would involve working on the project of what counts as sufficient evidence given the two distinct goals of rational belief. However, James does not of that; James does not offer standards of sufficiency that approach some ideal compromise between acquiring truth and avoiding falsehood. James wants us to believe when we do not have sufficient evidence, but only pragmatic reasons, to think a claim true. Clifford criticizes those who believe with admittedly insufficient evidence, and James defends them. But James’s considerations are so broadly applicable, that all sorts of patently irrational beliefs count as rational on his account. Perhaps that should be the point of the post today, but I’ll put that aside for another time. For now, I want to note that the flaws James claims to find in Clifford’s position are not as great as James believes.

The problem with sufficiency:

James thinks that we do not have sufficient reason to believe any, or most, of the beliefs that are most fundamental and necessary to our interactions with the world and others. We do not have sufficient reason to believe in the value of democracy and progress or the truth of science or the love of others. And so, we must make a guess, one that helps us survive in our society, take a leap of faith and believe—in democracy, in science, in our loved ones—and in so doing we will benefit ourselves and others, and by our faith we may even make our beliefs come true.

Everyone in our society, James argues, takes certain assumptions for granted, and most of us really have no sufficient reason to think these assumptions are true. Indeed, often the critics of these standard views are better justified in their rejection than we are in our acceptance. For example, Sarah believes that humans landed on the moon. She has never given much thought to the belief, but it is simply part of the background set of beliefs that comprises her social inheritance. She has been told it and taught about it so often that she does not consider that it might not be true, and she cannot consciously bring to mind any evidence that might be used to prove that the moon landing indeed happened. Fred, on the other hand, believes that the moon landing was a hoax and is familiar with the evidence put forth in a documentary on Fox asserting that it was faked. Now Fred can debate with Sarah and reduce her to stuttering disbelief and incoherence by demanding explanations for some supposedly anomalous facts about the videos of the moon-landing. Sarah cannot refute Fred, and Fred can rebut any piece of evidence that Sarah manages to assert. So, isn’t Fred the rational one and Sarah the irrational one? Sarah believes without sufficient evidence, and Fred has evidence of a kind. Yet we think Sarah is right to have the belief she does, and Fred is wrong. How can that be unless it is sometimes acceptable to believe with insufficient evidence?

Here’s how: Sarah’s belief is based on the testimony of multiple authorities who are in a better position to know than she is. Her history teachers who taught her about the moon landing and the textbooks she read are based on the authority of people in a position to know whether the moon landing was real or not. And because she believes based on that authority (even though she may not consciously be able to appeal to that authority in her debate with Fred), she is basing her beliefs on sufficient evidence. Even if she did not learn this information from the reliable sources of teachers and textbooks, it’s likely that her belief is better off than the Fred’s since his arguments, while they might seem compelling to the uninitiated, are not in reality convincing to those with some legitimate knowledge. At worst, supposing Sarah’s belief came from others who were not authorities, did not gain their information from authorities and whom she could not reasonably expect to be authorities, Fred’s belief, since it is based on fallacious reasoning, is no better justified than hers.

James overlooks legitimate forms of justification in claiming that our ordinary beliefs are not based on sufficient evidence. He acts as though legitimate appeals to authority do not exist and that, if they do, we have to be able to consciously state our reasoning in those terms. But this is not how all, or probably even most, justification works. We may have little idea how our eyes work, but we still can know beliefs we form on the basis of our first-hand observation. We may not be able to refute the objections of the skeptic—just as we cannot prove that Descartes’ evil genius does not exist—but it does not follow that our beliefs based on authority or the senses are not sufficiently grounded. James is too skeptical here; he rejects legitimate sources of knowledge and justification.

Our belief in the truth (or justification) of science is similar to the moon-landing case. Not all areas have real authorities, as science and history do, so in those areas the justificatory link breaks down. We are then left with nothing more than an appeal to popularity and no sufficient justification. For example, if we believe in God because everyone around us does, this belief is fundamentally different, although perhaps psychologically indistinguishable to Sarah, from the belief in the reality of the moon landing. People observed and participated in the moon landing and the enormous amount of labor and technology necessary to bring it about; there were intersubjectively verifiable first hand observations and accounts connected to the moon-landing. No one has observed God in a way that could be verified by others in the same position to know; no one is in the kind of authoritative position with respect to God that many are in with respect to the moon landing.

Democracy and progress may be a different matter since this involves normativity (facts about what should or ought to be the case) and there are supposed to be no authorities on normative claims. However, there may certainly be sufficient historical evidence for the improvement of the human condition measured as objectively as possible. So, while it is not necessary for my claim, I think a measured respect for democracy as an instrument of human progress is based on sufficient evidence. The important point is that many claims about politics and history are justified by sufficient evidence.

The third case struck me as particularly out of date. James asks us to consider the man who, through his fervent belief that a woman loves him, despite that belief’s lack of foundation in sufficient evidence, convinces the woman that she does in fact loves him. To the modern ear (mine, anyway), this sounds more like stalking and harassment than it sounds like rational behavior. But, being charitable to James, we suppose that there are steps we can take to engage others socially that, he thinks, do not make sense if we are awaiting sufficient evidence that the other person wishes to engage us socially as well. James imagines the rationalist standing aloof from others awaiting their commitment before making overtures of friendship. But that rationalist must be left out in the cold since none would befriend someone so cold and uncaring.

James’s case here is a strawman, however. We all recognize, based on evidence, that people are basically friendly, and that making friends with people often involves politeness and even overt friendliness. There is nothing irrational in thinking that such gestures are more likely to be returned with friendship than is cold aloofness. Must I believe, with insufficient evidence, that someone is my friend or loves me before I proffer my own friendship or even love? Not at all. I only have to believe, based on sufficient evidence, that such overtures are more likely to be received positively than no such overtures, and that the probability of a positive response and the benefits thereof outweigh the meager costs of being friendly myself (or the higher cost of declaring my love).

There is one further type of case, related to the “faith makes it happen” case, and that is when belief would allow for a positive result when the evidence does not render such a belief rational. I may believe irrationally that I will survive my heart attack, and this belief increases my chance of survival, but if I looked rationally at the evidence, I would conclude that I would not survive. Assuming I took all the actions that someone preparing to die should take, and it is by no means certain that I would since doing so might lessen my irrational belief in my own survival, this would make a genuine case in which my being fully rational would have a worse outcome than the alternative.

However, the problem is that, when you take a standard of belief with insufficient evidence, you cannot limit your belief to just these cases. You cannot know that you are in this position if you base your beliefs on insufficient evidence. And if you base your beliefs only on sufficient evidence, it is too late for you to take advantage of the power of positive thinking here; you already have recognized your low probability of survival. The problem is that, for the small increase in probability of survival in a small number of marginal cases, you have to give up rationality in too many other cases. On balance, belief with insufficient evidence would lead to far more harm than good. The possibility of a gain in an unusual case does not justify reasoning in this way in all the other cases in which it would cause harm. And there’s simply no way to compartmentalize the cases in which the belief with insufficient evidence is ultimately beneficial from the much greater number of cases in which the belief based on insufficient evidence is harmful.

In sum James’s examples do not prove that belief with insufficient evidence is necessary. They simply do not support the idea that we cannot survive if we base our judgments only on sufficient evidence.

This leaves me to respond to James for one further criticism of Clifford. James says that he cannot refuse to believe unless he has sufficient evidence because he thinks: “a rule of thinking which would absolutely prevent me from acknowledging certain kinds of truth if those kinds of truth were really there, would be an irrational rule.”

James is trying to make his case more plausible by redescribing Clifford’s advice in a way that overlooks its most salient feature. There really is no problem with following a rule that would “absolutely prevent me from acknowledge certain kinds of truth”. For example, there are certain kinds of truths (although I don't really know what a "kind of truth" is--the issue is the conditions under which one accepts the belief, not the "kind of truth" it is) one could have access to by hitting oneself on the head with a baseball bat (or by using dangerous drugs or exposing oneself to radiation or experimental brain surgery or joining a dangerous cult). Surely, it is possible to come to have true beliefs by hitting oneself on the head with a baseball bat that one would not have otherwise. I might, rationally, come to believe that hitting myself in the head is painful, but I might also believe that aliens are stealing my luggage [Thanks, Steve Martin] or that, honest to God, Lynyrd Skynyrd’s Freebird should be the US national anthem [Thanks, Jeff Foxworthy], or that the third planet in orbit around Betelgeuse is inhabited by intelligent arthropods with a dozen limbs. Some of these beliefs may be true, and at least some of them I cannot get without experimenting on myself in physically dangerous ways. Yet if I adopted the rule: “Do not come to have beliefs by hitting yourself on the head with a baseball bat (or the other methods described),” I would be denying myself the opportunity to come to have these true beliefs. However, it is emphatically not irrational to follow this rule; it’s about as reasonable a thing as I can imagine that I should not hit myself in the head with a baseball bat (or use dangerous, mind-altering drugs or . . . ) even if this results in me missing out on some true beliefs. For one thing, I would never know they were true. And for another, this method is stupidly dangerous. So, James’ argument against Clifford, that his rule irrationally excludes a means of acquiring true beliefs, is not convincing. One often should exclude some ways of acquiring true beliefs if it is harmful to acquire them in that way (as Clifford tries to show believing with insufficient evidence must be) and if there is no reliable means to distinguish the true from false beliefs one acquires by that means.

Here’s another rule for belief: Do not believe if there is overwhelming evidence that the belief is false. Overwhelming evidence is not the same as certainty, so some beliefs which we have overwhelming reason to think are false will be true. So, according to James’s reasoning, it is irrational to close off this avenue of acquiring true belief. Thus, according to James, we should not follow the rule: Do not believe when there is overwhelming evidence that the belief is not true. Unfortunately for James, rejecting this rule is plainly irrational.

This post has gone on far too long, and I didn’t manage to address several of the issues I meant to. Perhaps I can return to them later. But the summary of my argument today is that William James makes rather embarrassingly poor arguments in attempting to undermine the need for sufficient evidence before one can rationally believe. None of the cases he uses to show that we need to believe with insufficient evidence adequately support his point. James’s criticism, then, is not itself based on sufficient evidence.


  1. The following two passages make a curious juxtaposition:

    James is widely respected for his work both in philosophy and psychology. So, it is surprising that he would reason so poorly, but the fact is that James’s reasoning has no real merit and is just a rationalization for his preexisting religious belief.

    The irritant in reading James is that he cannot refrain from irrelevant ad hominem attacks on Clifford, James calls Clifford “nervous” and cites his “fear” of making errors and contrasts this fear with his own hopefulness.

    First, it is my understanding that James was not a religious believer but only what Dennett would call a believer in belief. If this is correct (it is what I recall from my reading about James), then your assessment of James's reasoning as "just a rationalization for his preexisting religious belief" is founded on a factual error. In any case, your claim is certainly not founded on the text itself and is plainly an instance of the kind of "irrelevant ad hominem attacks" for which you fault James.

    Second, I am not persuaded that James engages in any such attacks. Clifford's epistemological position implies an attitude toward the conduct of life. This attitude is not itself derived from any evidence: it is not a conclusion, but a condition on the formation of conclusions. James's point (one of his points, anyway) is that such fundamental epistemological attitudes do not spring from our rational nature; hence his characterization of the Cliffordian attitude as a certain "nervousness" about being deceived. You may not be persuaded by his argument on this point, but it is an argument, not a rhetorical irrelevancy.

  2. I'm glad to read this. I just read the first chapter of 'The Varieties of Religious Experience' (called 'Religion and Neurology') and was quite shocked, really, at his weak reasoning (I was expecting greatness) and straw-men characterizations of the 'materialist' position regarding religious experiences being derived entirely from the brain. The problems seemed greater than could be accounted for by merely considering his (and his opponents') arguments to be scientifically antiquated. It reminded me of similar disappointments with C.S. Lewis ('The Problem of Pain' - truly awful) and the supposedly classic Grand Inquisitor episode in Dostoyevsky.
    I'm enjoying your blog, btw.