SKEPTICISM AND THE JUSTIFICATION OF BELIEF¹
William A. Wisdom


ABSTRACT

For some years I have had going through my mind a number of related--some closely and some loosely related--thoughts on logic, science, religion, and what C. S. Peirce calls "the fixation of belief". I have wanted but not found an occasion to draw these thoughts together into a single essay. Skeptic magazine provides the perfect audience for such an essay. Most readers consider themselves skeptics in some sense or other. While most are not professional philosophers, they are intelligent people concerned about centrally philosophical issues: which sorts of beliefs, and methods for arriving at beliefs, are responsible and which irresponsible, and why? Particularly for those who have not read a great deal of professional philosophy, I offer an introduction to some basic theory of knowledge and philosophy of science.

This study has two parts. Part I is on the justification of belief and related matters. I consider some important views on empirical significance, on the distinction between the factual and the logical, and on the notions of explanation, natural law, causation, hypothesis, confirmation, and scientific theory, and I apply some of these thoughts to questions about religious belief and its justification. In Part II, on skepticism, I offer what may be a new way of understanding skepticism, and I present some things that have been said on its behalf. To sharpen our understanding of the skeptical stance, I also present and evaluate some of the things that have been said on behalf of gullibility. I wind up with a close analysis of William James's defense thereof in "The Will to Believe".²

As a philosopher with over forty years of interest in logic and science, I've been startled by the sloppiness of many self-styled skeptics when it comes to talk about knowledge claims, and the justification of belief generally and the justification of religious belief in particular. The systematic study of these--of the nature of knowledge and of the justification of beliefs--is Epistemology, which, with Metaphysics and Ethics, is generally considered one of the three main branches of western Philosophy ever since its origins in the Greek world of the sixth century B.C.E. Today's remarks are an introduction to Epistemology especially for skeptics. Let me lay the groundwork for my talk by briefly citing three examples of the sort of carelessness into which skeptics sometimes lapse when trying to think about Epistemological issues.

For one example, a number of skeptics have written--and I'm sure many more have believed--that David Hume regarded belief in miracles as in principle unjustifiable. In a paper I gave to an international conference in Spain in 1997, I tried to correct this mistake.

For a second example, I refer you to Tom Napier's article entitled "Can Critical Thinking be Overextended?" in the December 2001 issue of Skeptical Briefs, where he lays out a mistaken view common to a number of skeptics. He says that critical thinking--"the use of reason to evaluate data and make deductions from them"--cannot be applied to religion because there are no relevant data. This is apparently meant to be supported by the lines that follow: "One of the central tenets of Protestant theology is that there cannot be proof that God exists. With proof it would be unnecessary to have faith and without faith there is no salvation." This is variously confused or mistaken.

First, I invite everyone to avoid the word "proof" in serious discussion. The word is thoroughly ambiguous. On the one hand, it typically--and, I should propose, correctly--means something like "logically or mathematically compelling demonstration", as when we talk of a proof of the Pythagorean Theorem. If this is what Tom had in mind, then there may well be no proof--no logico-mathematically compelling demonstration--that God exists, but that would hardly be interesting. Very few people have ever thought that "God exists" is true in the same sense that "All red fire-engines are red" and "2+2=4" are true. On the other hand, "proof" is sometimes used to mean something like "substantial justification". I suppose that Tom has some such meaning in mind when he says that "one of the central tenets of Protestant theology is that there cannot be proof that God exists". So let's understand "proof" in this sense.

Now it is simply false that "one of the central tenets of Protestant theology is that there cannot be proof that God exists". (By the way, even if it were true, that would not mean that there cannot be proof that God exists. But it is not true that this is a central tenet of Protestant theology.) Christian theology teaches that God has given us two different sorts of revelation. We have a general revelation of His existence in the whole of the beautifully ordered universe: any rational people scanning the world around them can see or infer the activity of a magnificent Designer and Creator therein. The Psalmist tells us that "the heavens declare the glory of God; and the firmament sheweth his handiwork" (Ps. 19:1). We also have a special revelation, in Scripture and perhaps in the traditions of the Church, which teaches what cannot be inferred by the unaided reason from our observations of the world: namely, that God's nature is triune, that Christ was born of a virgin; that He died for our sins, arose from the grave on the third day, ascended into heaven, and so on. These latter are the sorts of doctrines at least some of which must be believed for salvation…and must be held on faith.

How do I know all this? Because I discussed the matter with theologians at the Presbyterian and Lutheran seminaries in the area and at Episcopal church headquarters; and I read the Roman Catholic doctrine on the matter as laid down by the first Vatican Council in 1870.

Tom seems to reason this way: because on the Protestant view faith is required for salvation, on that view we must believe on faith that there is a God. That just doesn't follow, and is the sort of sloppy thinking among skeptics that distresses me. Belief that there is a God won't take us any distance toward salvation. "Thou believest that there is one God," writes James. "Thou doest well: the devils also believe, and tremble" (2:19). The faith required for salvation is something like faith that Christ suffered and died for our sins--that without Him we are lost sinners doomed to eternal damnation.

One final example of the careless thinking into which skeptics slip. In the February/March 2002 issue of Phactum (Vol. 7 No. 6), Tom writes that a distinction between beliefs and claims supports his notion that "PhACT…[is] not a proper forum for discussing religion". He says: " 'Jesus performed miracles' is a belief. 'I can perform miracles' is a claim." This doesn't go any distance toward establishing the distinction Tom wants, since, though he says that "Jesus performed miracles" is a belief, it can as easily be a claim, if somebody claims it; and though he says that "I can perform miracles" is a claim, it can as easily be a belief, if somebody believes it. Tom goes on: "A belief…is not normally subject to experimental proof." This is just plain silly. Thousands of my beliefs--and of yours, I'm sure-- are "subject to experimental proof". What follows is also silly: "In the sense in which skeptics use it, 'claim' refers to a specific here-and-now ability which, presumably, can be demonstrated." Yesterday I read the claim that the soul lives on the astral plane for exactly ten thousand years between incarnations. There was no presumption that the claim could be demonstrated. I don't know where Tom learned English, but it was not among standard speakers of the language.

Before proceeding, let me add a note on the distinction between claims and beliefs. What I have in mind here involves the notion of what philosophers often call a proposition. A proposition is what is true or false. It is not a bit of language, but is the meaning of a declarative or indicative sentence: for example, it is what is common to "Il pleut", "Es regnet", and "It's raining". And, most importantly for the present purposes, it's the object of propositional attitudes like belief, doubt, assertion, denial, and so on; it's what is believed, doubted, asserted, denied, and so on. It's the "that P" clause in the following:

X believes that P
X doubts that P
X claims that P
X denies that P
X suspects that P
X fears that P
…and on and on.

And we also have:

It is true that P
It is false that P
It is possible that P
It is probable that P
It is demonstrable that P
…and on and on.

These last don't express propositional attitudes, but express what might be called modes of the proposition that P. It is a proposition that is falsifiable, verifiable, mathematically demonstrable, "subject to empirical investigation", or so on. That's why it's so absurd for Tom to say: "A belief…is not normally subject to experimental proof". If it is experimentally provable that P, and I believe that P, then my belief is subject to experimental proof.

But these issues are not central to my paper. I mention them only to establish the need for skeptics to be careful when discussing the nature of knowledge and the justification of belief.


I. THE JUSTIFICATION OF BELIEF

INTRODUCTION

I mean to bring together in Part I a number of interrelated themes clustering around the general topic of the justification of belief. By way of illustration, I shall have occasion to talk about logic, mathematics, natural science, and religion as areas of thought in which questions about beliefs and their justification are at least sometimes central.

When I talk about justifying beliefs, I have in mind what we mean when we talk about supporting or grounding them or showing them to be legitimate: a familiar notion, not very fancy. But a couple of preliminary distinctions are important to sharpen our talk about justification.

First, I suppose that there are at least two different senses in which a belief might be said to be justified. These senses parallel the (at least) two senses in which we say that one should or ought to do something. In the prudential sense, we ought to do an action X for the sake of what it will get us--because it will help satisfy our interests. Being eager to get to town quickly, we ought to take the left fork, since the bridge is out on the road to the right: taking the left fork is, under the circumstances, the prudent thing to do. But there is also a moral sense of ought in which we ought to do an action Y not for what it will get us but simply because it is right. One might say that, whether or not it is prudent to protect the innocent, we ought to do so just because it is right.

Analogously, we might justify a belief Z on prudential grounds: it is appropriate or wise or legitimate or prudent for you to believe Z because you will be better off if you do. In this spirit, Pascal argues that you should believe the tenets of Christianity, not because they are true, but because you risk eternal torment if you do not. Again, people argue on prudential grounds that we should reject the theory of evolution, not because it is false, but because belief in evolution undermines wholesome family values.

Contrasted with prudential grounds for believing something are what I shall call rational grounds. These are any considerations that tend to show that the claim in question is true. This is the sense that I shall henceforth have in mind when I talk about the justification of belief.

One might wonder why we should ever care about the justification of belief in this sense, or whether beliefs need justification. The answer seems simple. To believe something is to commit oneself to its truth. Of course we have strong beliefs and weak beliefs. But this only means that our commitment to the truth of claims is sometimes strong and sometimes weak--which in turn means only that we sometimes have a great resistance to changing our mind on some matter, and sometimes little resistance. The basic point remains: to believe something is to be committed to its truth. So caring that our beliefs should be true rather than false is part of what it means to believe at all. In the absence of justification we can only hope that our beliefs are true, but we have no reason to think that they actually are.

The word "faith" has many different meanings. But in one of its senses, to believe something on faith is to believe it in the absence of rational justification. In this sense, then, those who believe something on faith have no reason to think that their belief is true. Faith in this sense is not a special kind of justification; it is the absence of any justification at all.³

So there is a first distinction between what I've called the prudential and the rational justification of belief. Within the realm of rational justification, there is a second important distinction to be made: the distinction between what I shall call verification and confirmation. (There are parallel notions of falsification and disconfirmation, whose meanings should be clear from the discussion below.) To verify a claim, in the sense that I shall use the term, is to show that it is true, to establish or guarantee or demonstrate its truth. I would call a proof that 7+5=12 a verification of the claim. And I would call pointing to it in normal light a verification of the claim that this piece of litmus paper is blue.

But very few claims--and fewer interesting ones--are subject to verification. So verification is not primarily what I have in mind when I talk about the justification of beliefs. Of course verification is justification--indeed, justification of the strongest sort. But it is not the only sort of justification of belief. Remember the earlier points: "When I talk about justifying beliefs, I have in mind what we mean when we talk about supporting or grounding them or showing them to be legitimate." And "rational grounds [for believing something]…are any considerations that tend to show that the claim in question is true."

To confirm a belief is to enhance its credibility, to show that it is more likely to be true than false, to give good but less than conclusive grounds for thinking that it is true. We are well aware, then, that our beliefs so justified may be false. We nonetheless believe them, not because we are sure that they are true, but because we have better reason to believe them true than false. But we believe them only provisionally--that is, with a readiness to change our mind if better reasons come in on the other side.

Still, none of these remarks tells us how beliefs are confirmed. The point will be addressed in more detail later. But this much can be said here. In establishing beliefs--and in particular, in trying to anticipate future events--people over the ages have followed a large number and wide variety of practices. Some of these practices have turned out to be successful--beliefs adopted on their basis have often turned out to be true; expectations guided by these practices have often been fulfilled--while other belief-forming practices have turned out to be unsuccessful. A belief is confirmed, then, if it is supported by practices that have so far regularly led to the truth. What some of these practices are we shall see later.

In the next section, I mean to present and discuss what I shall call "the classical view" on the justification of belief. The label refers not to what Plato and Aristotle thought on the matter, nor to the current orthodoxy, but to a broad position running from Francis Bacon, John Locke, and David Hume through John Stuart Mill and Auguste Comte to Bertrand Russell, Rudolph Carnap, and A. J. Ayer. It forms the background against which virtually all epistemology and philosophy of science have been conducted for the past fifty or more years in the English-speaking world. I'll sketch "classical" views--with which I am largely sympathetic--on the justification of belief: on empirical significance, on the factual vs. the logical, and on the notions of explanation, natural law, cause, hypothesis, confirmation, and scientific theory.

There have been many criticisms of the classical position over the past three or four decades. Later I will sketch parts of one particularly influential position, that of W. V. O. Quine, certainly America's most important philosopher of the last half century. His view in some ways develops and in other ways criticizes, but in any event enriches, the classical view on the justification of belief.

In the final section of this first half of the paper, I would like to review and summarize the main points of what has preceded, and apply these insights to questions about religious belief and its justification. Along the way, I'll address the question of which if any religious beliefs are susceptible to rational investigation. I'll propose what I think is a new rationale for religious belief. And I'll consider the charge that scientific practice, like religious practice, depends on faith.

"THE CLASSICAL VIEW"

Before trying to justify a claim, we have to know what it means. And before deciding what its meaning is, we have to decide whether it has any meaning at all--whether it is a meaningful claim or a meaningless, vacuous, empty one. While there are no doubt many senses of "meaning", I want to focus on what has been called the empirical significance or factual meaning, or the cognitive content, of a proposition.

Roughly, the question is: is this or is it not a claim about "the world" in the broadest sense, about reality, about the way things are in fact? Here is the test. Imagine that someone asserts a proposition P. You ask: "How would the world be observably different if P were false?" Put otherwise: "What experiences would be incompatible with P, or would count against P? What experiences would incline you to change your mind, and deny rather than assert P?" Of course, you are not asking the asserter to grant that P is false. You are rather asking what the world would be like if it were false. If the claimant can specify observable states of affairs incompatible with the truth of P, then you recognize P as having cognitive content, as being a factual claim--not necessarily a true claim, but a claim about matters of fact. The proposition P then describes (or misdescribes) the world, because it distinguishes the world as the claimant thinks it actually is from other ways that it might be but (presumably) is not.

But suppose that the claimant cannot specify how things would be different if P were false. Put otherwise, suppose that the claimant considers P compatible with every possible state of affairs. Then P is not about the world at all; it does not describe reality, because it does not distinguish things as they are from any other ways they might be. Several examples will help to clarify this important principle.

Somebody solemnly announces: "Time is unreal." This has an air of great profundity about it. But before lapsing into speechless awe, we think to ask the speaker: "What would things be like if time were real? How would they be different than they are now?" Do you see how we can thus zero in on just what the substance of the claim is, on just what it says or commits the believer to? Perhaps the claimant can specify how things would be observably different if time were real than they are now; and if so, we know that the claim is meaningful and we are on our way to understanding what it means. But it is my guess (from years of experience with sophomoric philosophy) that the claimant cannot say how things would be different if the claim were false. In that event, I reject the claim--not because it is false, but because it is literal non-sense; it is meaningless, and hence not even a candidate for my belief.

Other examples abound. Someone thinks that this is a deep question: What if I have the inner color-experience when looking at grass that you have when you are looking at a clear sky, and vice versa? Of course you and I both call the grass and like-colored things "green", and the sky and like-colored things "blue". But what if our experiences were reversed? Before wasting time on this puzzle, ask how things would be observably different if things were one way rather than the other in this respect--what experiences could possibly settle the issue. I propose that there is no objective or empirical difference between the world either way, and hence that the claim that we have different color-experiences when looking at grass (like the claim that we have similar color-experiences) is vacuous and hence not worth thinking about--there being nothing to think about.

Again, someone says: "God loves each of us as a father loves his children." If the claimant can specify possible (though of course not actual) states of affairs with which this claim is incompatible, we can begin to get a handle on what it means. But in many cases, those who assert this can propose no states of affairs ruled out by the belief. I should think that it would be incompatible with the excruciatingly painful deaths of hundreds of infants in a natural disaster. But the believer does not think so. I should think that it would be incompatible with eternal torment for the damned. But the believer does not think so. And if the claim is compatible with every possible state of affairs, then it is not a claim about reality, since it would be assertable regardless of how things really are, and thus is no more about things as they are than it is about any other possible state of affairs.

Now consider the claim that 2+3=5. I ask myself how the world would be observably different if this were false…and I cannot say. On reflection, I realize that I consider this proposition compatible with every possible state of affairs. Or again: "if something is a red fire-engine then it is red." I cannot imagine how things would be different if this were false. I think that I would assert it regardless of what the world was like.

On the "classical" view being put forth, then, these last two claims are as meaningless or empty as the preceding--they are not statements of fact; they are not about the world. They are, however, true. They differ from empirical or factual truths in that mathematical and logical claims not only are true of the actual world, but would be true of any possible world as well. In that sense they are not descriptions of the world. They are true not because of the way the world is, but are true merely by virtue of the meanings of their constituent words. (All true statements are true at least in part because of the meanings of their constituent terms: "Grass is green" would be false if by "green" we meant what we now mean by "red". But the truths of logic and mathematics are true solely by virtue of those meanings; no extra-linguistic facts contribute to their truth.) Statements true by virtue only of the meanings of their constituent terms have been called analytic truths; the remaining truths are called synthetic.

The "metaphysician" and the "theologian" can derive no comfort from this last point, though. People who say "Time is unreal" are not likely to regard the claim as true solely by virtue of what "time" and "is" and "unreal" mean: they think that the sentence is genuinely informative about extra-linguistic reality. And theologians are not likely to say that we can learn, merely by analyzing the meanings of words, that God loves us as a father loves his children: they consider the claim descriptive of the way things actually are, as a matter of extra-linguistic fact.

On the classical view, then, a statement is significant--is a candidate for truth or falsehood, and hence for our belief or disbelief--if and only if either (a) its truth or falsehood can be determined solely by reflection on the meanings of its constituent terms, or (b) its truth would be incompatible with some specifiable experiences: some observations, in the broadest sense, would count against the statement.

Having made a distinction first between the meaningful and the meaningless, and then between the factually or empirically meaningful on the one hand and the logically or mathematically meaningful on the other, the classical view proceeds to an analysis of a cluster of concepts in terms of which we describe and evaluate responsible thinking about matters of fact: explanation, law, cause, hypothesis, confirmation, and theory.

To explain an event or state of affairs is to reduce it from the "phenomenal" or puzzling to the "natural" or familiar, to fit it into a pattern, to show that it is what had to happen, given the attendant facts both specific and general. By specific facts I mean things like the wind's shifting to the west or the gun's going off. By general facts I mean things like the boiling point of water at sea-level or the relation between the pressure, temperature, and volume of an enclosed gas. We call the statements expressing well-established generalities or regularities in nature laws. When the generalization expresses a regularity in the sequences of events--"events of kind A are always followed, nearby and soon after, by events of kind B"--we call the law a causal law, and call the prior event the cause of the subsequent event, the effect.

The basic pattern for the explanation of a fact F, then, is its subsumption under natural law. We show that from a description of attendant circumstances (specific facts) and relevant laws (general facts) we can deduce a description of F. In the causal case we show that the particular prior circumstances are related to the fact F as a special case of a natural law: whenever this happens, that happens. Prior conditions being what they were, and the world in general running the way it does, F is what had to occur. When explanation is successful, we realize that, had we known earlier what we know now, we could correctly have predicted the occurrence of F. That is the sense in which explanation makes what had been puzzling seem natural or ordinary--just what we would have expected had we been more familiar with all the facts.

In this way we explain specific facts. To explain the rupture of the car's radiator last night, we call attention both to attendant circumstances (the radiator was made of copper of such-and-such thickness; it was filled with plain water; the temperature went down to 0∞ F; etc.) and to a number of laws (about the behavior of water below the freezing point; about the strength of copper; etc.). From all these statements together we can logically deduce the breaking of the radiator. We thus explain its breaking by showing that it had to break under the specific and general circumstances obtaining last night.

We explain general facts in the same way. Kepler's Law says that the planets of the solar system move in elliptical orbits, a line from each planet to the sun sweeping out equal areas in equal times. When it was proposed, this was a "brute fact" about the solar system. But we now know why the planets should move thus: Kepler's Law can be deduced from Newton's laws of motion, together with some information about such things as the relative masses of the planets and the sun. So "lower level" or less general laws are also explained by being subsumed under "higher level" or more general laws, together with information about the special circumstances within which those higher laws operate.

We have yet to see how a statement gets to be a law. A law is a statement expressing a well-established regularity in nature. Call a general statement after it has been proposed as a law, but before it has been well-established, an hypothesis. The process by which such a law-like claim becomes a law is called confirmation. We said earlier that "to confirm a belief is to enhance its credibility, to show that it is more likely to be true than false, to give good but less than conclusive grounds for thinking that it is true". But what counts?

Sometimes, confirmation of a general statement can be quite direct: the hypothesis "All swans are white" is confirmed by the observation of individual swans that are white. The greater the number and variety of swans observed to be white, the better our reasons for believing the hypothesis to be a law.

But more often confirmation is indirect, following a somewhat more complicated pattern. Typically, from the hypothesis being tested (together with background information that we already have good reason to believe) we deduce statements whose truth or falsehood can be observed. These observable consequences of the hypothesis are statements that have to be true if the hypothesis is. The fulfillment of these consequences counts as confirmation of the hypothesis--reason to believe that the hypothesis is true. Of course, the observation that things are as they would have to be if the hypothesis were true is not verification--it is not a guarantee that the hypothesis is true. (The observation of a white swan is no guarantee that all swans are white. But it does go some distance toward rendering the general statement credible. How far it goes is the difficult subject matter of inductive logic, probability theory, and statistics.)

It is an interesting question why the verification of a logically necessary consequence of an hypothesis counts as confirmation of the hypothesis. It may be because what follows of logical necessity from a claim expresses part of its content. (For an oversimple example: "Both Tom and Mary are here" logically implies "Mary is here", which is part of what the former statement says. Seeing that Mary is here enhances the believability of the claim that Tom and Mary are here.) On this account, confirmation is partial verification. In the normal case, we cannot observe the truth of a general claim. But we can observe the truth or falsehood of specific claims that are logically entailed or implied by a general claim. And it surely makes sense to think that we raise the credibility of a general claim by showing that part of what it says is true.

Another kind of reason has been given for regarding as confirmation of a general claim the verification of its consequences. Because the observational consequences deduced from it are logically necessary consequences of the hypothesis (together with the background information assumed), the falsehood of one of those consequences falsifies the hypothesis (on the assumption--which is well-grounded but itself subject to revision--that the background beliefs are true). Typically the testing of a hypothesis, then, involves the effort to falsify it--and its confirmation is the failure of those efforts. If our best efforts to show that our hypothesis is false have failed, we have earned the right to regard it as true. (How would you show that "All sodium salts burn yellow" is false, if it is? You would burn a sodium salt and see that the flame is not yellow. So to confirm the hypothesis we burn as large a number and wide a variety of sodium salts as we can, maximizing our chances of finding a falsifier if there is one.) On this account, confirmation is delayed falsification. The longer our aggressive efforts at falsification are unsuccessful, the better reason we have to treat the hypothesis as true.

Finally, we need to say something about what a scientific theory is. This has been and remains the subject of much controversy in the philosophy of science. But a few things can confidently and usefully be said here. First of all, in everyday discourse we sometimes use the terms "theory" and "hypothesis" interchangeably: "I have a theory (or hypothesis) about how the window got broken." But when scientists talk about the heliocentric theory or the theory of evolution or relativity theory or the kinetic theory of gases, they do not mean an educated guess or an unsubstantiated claim. For that they reserve the word hypothesis. A theory is a comprehensive point of view that integrates a number of laws into a coherent body of thought. Frequently a theory does its characteristic work by appeal to postulated entities, properties, events, or processes that are unobservable. So the support for a theory cannot come directly from observation. It comes rather from the power of the theory to organize, to systematize, a large body of otherwise disconnected facts. The virtues of a theory are such things as its simplicity, its scope, its internal consistency, its compatibility with other well-established beliefs, and its ability to suggest and guide further successful research projects.

Although I am quite unsympathetic with psychoanalytic theory on other grounds, it nonetheless can be used to illustrate the main features of a scientific theory. Suppose that people who have had early and severe toilet training grow up to be authoritarian. Suppose that boys who lost their fathers early in life fear bats when they grow up. Suppose that men who must frequently wash their hands also often dream about flagpoles. As such these would be disparate and inexplicable facts--mere facts, brute facts. What the machinery of psychoanalysis--id, ego, superego, libido, cathexis, the unconscious, repression, Oedipus complex--what this apparatus does is propose a whole realm of unobservable entities and processes, with specified features, that variously are affected by our experiences and affect our behavior. It proposes that, and how, apparently unrelated phenomena like toilet training, neurotic behavior, slips of the tongue, and dreams are really related beneath the surface of conscious life. And it purports to explain why certain alleged regularities of human life should be as they are.

A scientific theory tries to make sense of a number of well-established but disconnected beliefs by organizing them into a coherent system. So the justification for adopting theories is not that they correctly describe the world: what correctly describe the world are the particular beliefs organized by the theory. The justification for adopting a theory is rather its power to integrate a large number and wide variety of general facts about the world that would otherwise remain disconnected. In this way, for example, the theory of evolution connects an amazing number and variety of facts from astronomy, geology, biology, paleontology, and other fields. Its appeal, like that of any theory in science, lies primarily in its power and elegance.

In this section we have briefly surveyed a number of related issues bearing on the justification of belief. We have distinguished between meaningful and meaningless claims, and within the former category have distinguished between synthetic (factual, empirical) truths and analytic (logico-mathematical) truths. We have then discussed--again, too briefly--the concepts of explanation, law, cause, hypothesis, confirmation, and theory in science. In the next section we shall refine some of these notions along lines laid out by W. V. O. Quine.

THE WEB OF BELIEF

W. V. O. Quine--in his "Two Dogmas of Empiricism" ⁴ , The Web of Belief ⁵ , and other works--develops a view that is critical of the foregoing distinction between synthetic and analytic truth, and that supports a somewhat different account of the justification of belief.

Picture the body of one's beliefs as a web of interconnected nodes attached to experience only at the periphery. The lines of connection within the web represent the dependency of some beliefs on others: some of our beliefs either evidentially support or are supported by others. Only some of our beliefs directly depend on, or report, experiences: "There's a patch of blue in my visual field", or perhaps "There's a coffee cup on the table before me." The rest of our beliefs depend not immediately on experience but on other beliefs, which themselves depend on still other beliefs, and so on. Ultimately, all of our beliefs are connected to experience--but most of them only indirectly, and some very indirectly, via many links to many other beliefs.

Again imagine the web. The distance of a belief from the periphery--experience--represents a number of things. The farther a belief is from experience, the more other beliefs is it connected to in the network of interdependency, and hence the larger the number and wider the variety of experiences on which it ultimately depends. Because of the interdependency of beliefs, changes in one belief (plus our desire for consistency) will of necessity occasion changes in others. Beliefs at or near the periphery are connected to relatively few others, so a change in a peripheral belief--a single observational datum--seldom requires a change in many other beliefs. Hence our readiness to change beliefs near the edge. I believe that there are no bars on Diamond Street, but would readily abandon that belief in the face of unexpected experiences, because so little depends on it.

But we have a legitimate reluctance to change beliefs more deeply imbedded in the network, since to change one of them will require changes in many others, each of which presumably lies within its own network of justification. Changes in our more general and fundamental beliefs are in this sense more costly than changes in our more specific beliefs. So the distance of a belief from the periphery is also a measure of our willingness or reluctance to change our minds in the face of recalcitrant experience. The failure of only a few experiences to come out as we expected is unlikely to cause us to change our minds about our very general beliefs, since they are connected to a wider variety of experiences than are our more specific beliefs.

Quine says that "our statements about the external world face the tribunal of sense experience not individually but only as a corporate body" ("Two Dogmas", p. 41). Regardless of what our experiences are like, we can consistently hold virtually any belief if we are willing to adjust enough other beliefs to accommodate it. You can believe, consistently with ordinary experiences, that we live not on the outside but on the inside of a big sphere, if you are willing to hold beliefs about gravitation, the nature of light, and so on, that are radically different from most people's. But the resulting belief-system may be so arbitrary and ad hoc as to be thoroughly unacceptable.

This general view holds that all beliefs depend ultimately on experience, some more and others less directly. The beliefs most distantly related to experience--those at the center of the web--are remotely connected to virtually all of our experiences. They are also the beliefs that it would be most costly to change, since their change would require changes throughout the whole network. Indeed, we can understand why people might be psychologically unable to imagine how their experiences would be different if their central beliefs were false; but this is just a failure of imagination.

Quine claims, then, that the "classical" distinction between synthetic and analytic truth, between truths that have empirical content and those that do not (those that are true solely by virtue of the meanings of words) cannot be defended. All beliefs, even the most fundamental and apparently incorrigible, are properly subject to change. It strains the imagination to regard the principles of logic and mathematics as susceptible to revision in the face of radically new experiences; but Quine ably argues that they are revisable.

Which beliefs are at the center of the network would, of course, vary from individual to individual. I think the paradigmatically rational person would have the principles of logic and mathematics there. But someone else might have there their religious convictions, their political principles, or virtually anything else--whatever beliefs are so deeply involved in all their other beliefs that only changes in all of their experiences could justify changes in those central beliefs, which latter changes would have repercussions throughout the entire body of their beliefs.

APPLICATIONS

I want to illustrate some of the foregoing remarks by considering religious beliefs and their justification. It should be clear from the foregoing what sorts of claims--religious or otherwise--are subject to rational, scientific investigation: any claims that purport to be either descriptive of a matter of fact or true by virtue of the meanings of terms.

We can bypass the latter category. It is a relatively easy matter to determine whether or not a claim is true by virtue of the meanings of its constituent terms. And if we doubt it, the proponent bears the responsibility of providing the requisite linguistic analysis or proof. In any event, few characteristically religious claims are presented as true on linguistic grounds alone. Perhaps the only interesting exceptions are ontological arguments for the existence of God (‡ la Anselm, Descartes, Spinoza, Leibniz, Hartshorne, Malcolm, Plantinga), which most people in the mainstream of western thought--including myself--think cannot be successful.

Which if any religious claims, then, would not be subject to scientific investigation? First of all, there are those that purport to be substantive but are not--those that the proponent considers compatible with every possible state of affairs (perhaps "God loves us as a father loves his children"). Evidence one way or the other is unavailable, because every state of affairs would count in favor of such a belief; so it is properly rejected, since it is no more about this world than about any other possible state of affairs.

Then there are those claims that do not even purport to be descriptive. The most conspicuous examples are pure, basic, ultimate ethical principles, if (as seems plausible) these are statements not about how things are but about how they ought to be. Such principles would not be susceptible to scientific investigation. (Jeremy Bentham says that ethical judgments are ultimately factual claims--in his case, claims about what acts will maximize happiness. But his is a minority view, I think.) If there are other religious beliefs that do not purport to be descriptive, then they too would be immune from scientific scrutiny.

What about sheer declarations of faith--"I believe on faith alone that: the universe came into existence about 10,000 years ago; Jesus arose from the dead; Aunt Minnie's cancer is cured; every line of Scripture is literally true; a personal, omnipotent, omniscient, perfectly benevolent Being is one of the many, many things that actually exist"? Are such declarations of personal faith also immune from scientific investigation? Of course not. Each of these is a statement about matters of fact, whose truth or falsehood would have observable consequences and would be compatible or incompatible with other well-established principles. So in each case there is empirical evidence that points one way or the other, or perhaps some evidence that points one way and some that points the other. In any event, such claims are prime candidates for scientific investigation.

Remember that to say "I believe P on faith alone" is not to claim a special kind of justification; it is rather to admit that I have no rational grounds for believing P, but believe it anyway. In such a case, nothing prevents someone else from seeking evidence one way or the other on the matter, if the matter is what we have been calling empirical or factual. After all, what is believed on faith may be false--even demonstrably false.

What has been said in the two preceding sections may provide a basis for regarding religious beliefs as substantive or broadly empirical, and hence meaningful, and yet insusceptible to scientific investigation--at least in the most straightforward sense where investigating means looking for evidence. What we called theories on the classical view, and central beliefs on Quine's, are claims whose characteristic function is not the description of the world but the systematic organization of a number of other beliefs--virtually all other beliefs, in the case of beliefs at the center of the web. It seems appropriate to call these integrating claims empirical. But they are not supported by observations in the same way that, say, reports of general or particular facts are. I suppose that some people might regard their most basic religious beliefs as of this sort--not themselves confirmed by experience, but nonetheless justified by their power effectively to make sense of a large number and wide variety of other less fundamental beliefs, themselves confirmed in the ordinary ways.

I do not know whether either the theologian or the believer-on-the-street would be comfortable with such a defense of, say, belief in the existence of God. This defense is surely not traditional. Traditional (other than ontological) arguments typically offer the God-hypothesis as the best explanation for certain observable features of the world, which features are considered evidence for God's existence--very general features (as that things change), or middle-sized features (as that the earth is so admirably suited to human existence), or very specific features (as that an image of Jesus appeared on the scrotum of a Guatemalan lad). These traditional arguments for the existence of God, then, have roughly the structure of ordinary scientific arguments, and purport to meet the same criteria.

The fact is, however, that appeal to God's activity explains nothing at all. I shall illustrate this on two quite different scales. Many people think that "God created the universe ex nihilo" offers a causal explanation of its origin--perhaps not the correct explanation, but at least what could be an explanation. But it does not. Remember that to explain something means to fit it into a well-established pattern, to subsume it under a law. To explain the occurrence of a particular event Y by saying that event X caused it is at least to say that (a) event X occurred before Y, and (b) our experience shows that whenever an event like X occurs, it is followed by an event like Y. So in order for appeal to the creative power of God's will to count as explaining the origin of the universe, we would have to have lots of experiences of God's wanting a universe, followed by the appearance of one. But of course nothing of the sort is the case. The believer might say that we do have relevant experiences, but of a different kind: we have experiences of semi-potent wills bringing things about; God is omnipotent; and so we know what it means to say that God always gets what He wants--that is the requisite "law". But while we may know what this means, we have no reason to think that it is true. That is, we have absolutely no reason to believe that there is a general principle (or, for that matter, a prior state of affairs) of the sort needed for an explanation in this case. One who says that God created the universe is saying exactly this: I have no idea how the universe came into existence.

Similarly, some people think that "God caused the rain to fall yesterday on the parched fields of Kansas" offers what could be an explanation. But it does not, unless we have empirical evidence that (a) God wanted it to rain yesterday, and (b) it rains whenever God wants it to. Of course we have no such experience. (The story might be elaborated to have thousands of Kansans praying for rain. But this will not help unless it rains whenever lots of people pray for it. Despite the Scriptural promise (Matthew 21:22; Mark 11:24; John 14:13-14, 15:7, 15:16, 16:23)⁶ , faithful petitions are not always granted. And even if they were, we'd have to say that prayer, not God, caused the rain.) One who says that God caused the rain is saying exactly this: I have no idea what caused the rain.

One final issue: does scientific belief and practice, like religious belief and practice, depend on faith, on unjustified assumptions? Some people have claimed that it does, and even that it must. An argument for the latter view runs thus: any belief that is justified at all is justified by appeal to other beliefs, and those to still other beliefs, and so on. So if there are no unjustified beliefs at the bottom, on which all the rest ultimately rest, we are faced with either an infinite regress or a vicious circle of justification; and in either of these last two cases, nothing can really be justified. So if any beliefs are justified, some foundational beliefs must be unjustified.

This line of thought is insensitive to the "bootstrap" nature of scientific practice. Whenever testing a belief, some other beliefs are taken for granted on that occasion. But the assumptions of that occasion are themselves subject to confirmation on another occasion (on which, of course, different beliefs are taken for granted). If this sounds like a circle, at least it is not a simple vicious circle. And it is more like a spiral, in which large bodies of beliefs are given support over time. If any of the temporary assumptions in this on-going process is false, there is every reason to think that its falsehood will eventually be exposed in the ordinary way.

A more common charge is that scientific practice presupposes certain principles, which therefore cannot themselves be scientifically justified and must remain articles of scientific faith. The main candidates for this status are versions of "Every event has a natural cause" or "Nature is orderly" or "The future will be like the past." It is true, of course, that if there were no regularities in nature, science could not operate. But hoping (not assuming) that there are regularities, we look for them, find lots, and expect to find more.

Why should we hope to find regularities (even if there were not very many)? Because it is only by identifying patterns in nature that we can understand what is going on, anticipate the course of events, and exercise some control over what happens. We have discovered that nature has been orderly. We have no guarantee--nor need we assume--that it will continue to be orderly, that the future will be like the past; nor need we believe that every event has a cause. But we will continue to hunt for causes and other regularities in nature, since only by looking for them can we find them if they exist.


II. ON SKEPTICISM

Against the background of what has been said in Part I about the justification of belief generally, I will here offer what I think is a novel account of skepticism, and will present some considerations that have been offered both in its defense and in defense of its antithesis: gullibility.

TYPE I AND TYPE II ERRORS

RenÈ Descartes in his Meditations on First Philosophy, William James in "The Will to Believe", and others have noted that, with respect to our intellectual lives, we have these two distinct though related goals: (i) to acquire true beliefs, and (ii) to avoid false ones. Accordingly, we can go wrong in two different ways. We commit what we shall call a Type I Error when we fail to believe a truth; we commit a Type II Error when we believe a falsehood. We want to avoid errors of both kinds, of course. The ideal intellectual strategy would prevent the commission of error generally. But there is no such strategy. (We cannot guarantee freedom from error by "playing it safe"--by withholding our judgment on some proposition. For if that proposition happens to be true, we will thereby commit a Type I error.)

There is a simple strategy we could adopt to avoid the commission of all Type I errors: believe everything! Believing everything, we would be sure to believe every truth. But of course we would also believe every falsehood--and hence would commit every Type II error possible.

Similarly, there is a strategy we could adopt to avoid the commission of all Type II errors: believe nothing! Believing nothing, we would never believe a falsehood. But of course we would also believe no truths--and hence would commit every Type I error possible.

Unable to eliminate the risk of error entirely, we would like to minimize it. But how? A single strategy to minimize error seems no easier to come by than a single strategy to eliminate it. A suggestion comes from what, in fact, we do when confronted with some specific candidate for our belief--call it proposition P. We are guided by a preliminary decision--often but not always a quick decision--about the cost of being wrong in either of the two ways available. Not knowing whether P is true or false, I ask myself which would be worse on this occasion: failing to believe P if it were true (i.e., committing a Type I error)? or believing P if it were false (i.e., committing a Type II error)? Sometimes one and sometimes the other seems considerably "worse" or in some sense more costly than does the other, and we adjust our policy on that occasion accordingly.

If we feel that it would be worse to commit a Type I than a Type II error--worse to miss the truth--then we allow ourselves to believe P on relatively slight evidence, lest we fail to believe true P (willingly running the risk of believing false P). For example, Pascal says: "I would have far more fear of being mistaken, and of finding that the Christian religion was true, than of being mistaken in believing it true".⁷ [The text here ends with "...than of not being mistaken in believing it true"; but that has to be wrong. Clearly, he is considering the choice between accepting and not accepting Christianity. The point has to be: "I would be more afraid of not accepting it and its turning out to be true than of accepting it and its turning out to be false." In other passages--especially #233--he gives the details of what is known as "Pascal's Wager".] That is, with respect to the proposition "the Christian religion is true", committing a Type I error is far worse--infinitely worse, he says (considering the prospect of eternal torment)--than committing a Type II error (presumably: living a finitely long life of wasted virtue). Accordingly, Pascal recommends that we accept Christianity without any supporting evidence at all.

If, on the other hand, we feel that it would be worse on some particular occasion to commit a Type II than a Type I error--worse to hold a false belief--then we require a lot of evidence before we assent, lest we believe false P (willingly running the risk of not believing true P). For example, we hope that the responsible authorities approach in this spirit the proposition "This new drug [pesticide, infant formula,…] is safe." If they commit a Type I error here--they do not believe it though it is true--then profits may be lost. But if they commit a Type II error--they believe that it is safe while it is not--lives are lost. So we hope that this new product will not get to the market unless there is very strong evidence of its safety.

The issue on any particular occasion, then, is: how much evidence must I have on behalf of P before I am willing to believe it? how ready or reluctant am I to believe P? The greater I take the relative cost of a Type II over a Type I error to be--i.e., the more reluctant I am to believe--the more or stronger evidence I require to overcome that reluctance. And conversely, the more costly I think a Type I error to be, the readier I am to believe P on relatively weak grounds.

This analysis of what happens on a specific occasion provides the basis for a distinction between what might be called intellectual "personality-types". There are some folks who in general have relatively low standards of evidence--who believe "at the drop of a hat". They behave as if they regard Type I errors as worse than Type II...as if in their desire to believe truths they are willing to accumulate a lot of false beliefs as well. These people are called credulous or gullible. And we might generously think of credulity as a policy for minimizing error generally--at least errors of the more serious sort.

On the other hand, there are folks who in general have relatively high standards of evidence--who display a reluctance to believe until overwhelmed by evidence. They behave as if they regard Type II errors as worse than Type I...as if they have, in the colorful phrase of William James, a "preponderant private horror of becoming a dupe" ("Will to Believe", p. 18). These people are called skeptical. And we might think of skepticism as a policy for minimizing error generally--at least errors of the more serious sort.

Skepticism, while not universally held, might seem to be the more respectable stance of the two. People are not ridiculed for being skeptical, as they are for being gullible. Indeed, skepticism has its advocates, while it would seem that nobody recommends gullibility. But this is a feature of our particular age. In other times, it seems, the credulous--at least people credulous with respect to the most important matters--were honored, while the skeptical were tortured and burned. So it might be interesting to survey some of the things that have been said for and against skepticism and gullibility respectively.

Because so much has been said about the importance of keeping your standards of evidence high, I shall address only two "skepticisms" that are specially important for this study. Descartes is important as an early modern exponent of our distinction between two ways of erring. Clifford is important for his influence on James, whose views we shall go on to consider.

DESCARTES AND CLIFFORD

In an effort to establish a firm foundation for his thinking generally, RenÈ Descartes in his Meditations on First Philosophy (1640) sets out to find which if any propositions are indubitable. This he does by raising his standards of evidence as high as possible: he will not believe anything that can be doubted...anything that could be false. In the present terminology, he begins with the determination to commit no Type II errors, even if that leaves him believing nothing.

Certain propositions soon turn out to be indubitable and hence utterly believable on his account: that he exists, that he is a thinking thing, that God exists, and that God is not a deceiver. The non-deceptive character of God then provides Descartes with a guarantee that "whatever I perceive very clearly and distinctly is true" (Descartes, "Meditation Three"). Clear and distinct "perception" is not sensory but intellectual; and we can tell whether or not we have got it. The paradigm is our grasp of our own existence, and--as it turns out--of the elements of mathematics. Some things initially perceived only obscurely and confusedly can by careful attention to the contents of the mind be brought to clarity and distinctness.

Because God is not a deceiver, says Descartes, He has given us no mental faculty the exercise of which could lead us into incorrigible error. But there could be no stronger assurance of the truth of something than our clear and distinct perception of it. So we could not correct our false assent to something clearly and distinctly perceived. So what is clearly and distinctly perceived cannot be false. We should confine our belief to what is clearly and distinctly perceived, and thus avoid all Type II errors.

Of course by keeping our standards so high, we will commit a lot of Type I errors--we will fail to believe a lot of truths. But Descartes provides a recipe for eliminating more and more Type I errors--i.e., for accumulating more and more true beliefs. By critical refinement, we should bring to clarity and distinctness of apprehension more and more of our concepts, and thus bring to indubitability more and more candidates for our belief--thus continually reducing the number of Type I errors, while still preventing the occurrence of Type II errors entirely. [While this material is spread through much of the Meditations, the hard core of it is in the last few paragraphs of Meditation Four and the first paragraph of Meditation Five.]

Descartes's position fits our general picture of a skeptic exactly. He proceeds from the very beginning of his Meditations precisely as if he regards Type II error--believing a falsehood--as far worse than Type I error--missing a truth. He is willing to forgo indefinitely many truths rather than allow into the body of his beliefs the slightest falsehood. As time goes on, he will assent to new propositions, but only after they too have been brought to indubitability--i.e., have met the very highest standards of evidence.

In "The Ethics of Belief" (1876) William K. Clifford argues that "it is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence". Though clearly hostile to religious belief, he nonetheless uses terms like "sinful" and "evil" for "belief...given to unproved and unquestioned statements", even if those statements are true. We have a "duty to mankind...to guard ourselves from such beliefs as from a pestilence". As with skeptics generally, Clifford demands that we keep our standards of evidence very high, recognizing that in doing so we will cut ourselves off from a lot of true beliefs.

As bad as unjustified belief is, "a greater and wider evil arises when the credulous character is maintained and supported, when a habit of believing for unworthy reasons is fostered and made permanent." That character, that habit, is particularly to be found in "those simple souls...who have been brought up from the cradle with a horror of doubt." Precisely in the language of our essay, Clifford identifies credulity as the antithesis of his own position.

PRO INSIPIENTE
(IN DEFENSE OF THE FOOL)

It seems that the dominant Christian position today considers it legitimate or appropriate, though not necessary, for the believer to exercise reason--i.e., to apply logic to the material of experience. This view holds at least that it is rational (not crazy or stupid or silly) to be a Christian, and perhaps even that the exercise of reason can lead one to, or support belief in, the main tenets of Christianity. Such is one familiar reading of Psalm 19.1: "The heavens declare the glory of God; and the firmament sheweth his handywork"--i.e., the observable world provides ample evidence of the existence and nature of God. But however strong or old this attitude toward reason, there is a contrary attitude that goes back to the very earliest Christian documents.

In Paul's first letter to the Corinthians (ca. 55-60 A.D.) is the familiar paean to love--Chapter 13--and a less familiar paean to folly--Chapters 1-3. Here are some representative thoughts. "It is written, I will destroy the wisdom of the wise, and will bring to nothing the understanding of the prudent....Hath not God made foolish the wisdom of this world? For after that in the wisdom of God the world by wisdom knew not God, it pleased God by the foolishness of preaching to save them that believe....The foolishness of God is wiser than men....Not many wise men after the flesh...are called: but God hath chosen the foolish things of the world to confound the wise...." (1:19-27) "The natural man receiveth not the things of the Spirit of God: for they are foolishness unto him: neither can he know them, because they are spiritually discerned." (2:14) "If any man among you seemeth to be wise in this world, let him become a fool, that he may be wise. For the wisdom of this world is foolishness with God....For it is written,...The Lord knoweth the thoughts of the wise, that they are vain." (3:18-20)

The lesson is unavoidable: to the natural mind, the Christian message is foolish, irrational. [This attitude is noteworthy in that the Psalmist says precisely that the atheistic position is the foolish one. (14:1, 53:1)] "Your faith", says Paul, "should not stand in the wisdom of men, but in the power of God". (2:5)

Some forty years later, a similar attitude toward reason was put in the mouth of the risen Jesus himself. Many people have not read or have forgotten the punchline to the story of Doubting Thomas. Told that Jesus has risen from the dead, Thomas says that he will not believe it until he can see and feel Jesus' wounded body. When Jesus presents himself to Thomas and invites empirical investigation, Thomas says, "My Lord and my God." Jesus' rejoinder is interesting: "Because thou hast seen me, thou hast believed: blessed are they that have not seen, and yet have believed." (John 20:28-29; emph. mine) Thomas is not blessed for his belief, which however belated is at least justified; the Lord's blessing is reserved for those with unjustified, groundless belief.

It is with this kind of scriptural backing that Tertullian says, perhaps a hundred years after John's gospel: all you need to do is believe what Jesus taught, however implausible. Once you have accepted Jesus' teaching, close your mind to everything else and stop thinking. "With our faith, we desire no further belief. For this is our palmary [outstanding] faith, that there is nothing which we ought to believe besides." "…Nor have you anything further to do but to keep what you have believed, provided you believe this besides, that nothing else is to be believed…" "[The Christian is not] permitted to give his inquiries a wider range than is compatible with their solution....It must, however, be added that no solution may be found by any man, but such as is learned from God; and that which is learned of God is the sum and substance of the whole thing." "...It is really better for us not to know a thing, because He has not revealed it to us, than to know it according to man's wisdom...." In his History of Philosophy, Wilhelm Windelband says: "...With Tertullian, the content of revelation is not only above reason, but also in a certain sense contrary to reason....The gospel is not only incomprehensible, but is also in necessary contradiction with worldly discernment: credibile est quia ineptum est; certum est, quia impossibile est--credo quia absurdum" (Chapter 2, ß18, Sec. 4)). It is believable because it is foolish, it is certain because it is impossible--I believe it because it is absurd.

Traces of such hostility to reason and praise of credulity run through the history of Christian thought--though for the last many centuries it seems to have been a minority view, represented most prominently by the likes of Pascal and Kierkegaard.

WILLIAM JAMES

No doubt the most elaborate, and I would say the strongest, defense of credulity comes from William James (1842-1910), one of America's most distinguished and influential philosophers and psychologists. His position is laid out most fully in "The Will to Believe" (1896 (WB)). But parts of it appear in "Reflex Action and Theism" (1881), "The Sentiment of Rationality" (1882 (SR)), "Is Life Worth Living?" (1895 (LWL)), and other works.

As we shall see, he clearly argues that, under certain broad circumstances, we are entitled to hold beliefs--and, most importantly, religious beliefs--that are absolutely groundless. He calls "The Will to Believe" a "justification of faith...in religious matters" (p. 1), understanding faith as we defined it earlier--belief held without rational support. Before examining his argument, we must deal with two red herrings.

At a number of points James offers a perfectly screwy definition of faith. Faith, he says, is the adoption of "a believing attitude...in spite of the fact that our merely logical intellect may not have been coerced." (WB 1-2). "Faith means belief in something concerning which doubt is still theoretically possible;...faith is the readiness to act in a cause the prosperous issue of which is not certified to us in advance." (SR 90) "Faith is synonymous with working hypothesis....[He who tests a theory] acts as if it were true, and expects the result to disappoint him if his assumption is false. The longer disappointment is delayed, the stronger grows his faith in his theory." (SR 95)

I call it screwy because on this definition nearly all beliefs about matters of fact, including the best-confirmed laws of nature, would be held on faith. Surely that is the wrong way to use the word faith. It is clear that we all must have "faith" in this sense. But James does not run his argument thus; his defense of credulity does not depend on redefining the word faith in this bizarre way. So we can disregard this definition.

James also alleges that the scientist must hold on faith "the proposition...that the course of nature is uniform. That nature will follow to-morrow the same laws that she follows to-day is...a truth which no man can know; but in the interests of cognition as well as of action we must postulate or assume it." (SR 91) On this he is simply wrong: we need not postulate or assume the uniformity of nature. As pointed out earlier, we certainly hope that nature is orderly. Proceeding on that hope, we would soon discover that we were wrong if the laws of nature changed or ceased. So this "justification of faith" also fails, since the uniformity of nature is not a necessary article of faith. But this is not James's key argument either. So we can disregard it too.

I would urge everyone to read "The Will to Believe". It is only thirty pages long in my edition, and well worth the effort. But for those who have not read the essay (and are willing to take my word for it), I'll summarize his argument. It is based on a number of key definitions.

An hypotheses is "anything that may be proposed to our belief". An hypothesis is live "which appeals as a real possibility to him to whom it is proposed". An option is a "decision between two hypotheses". An option is living "in which both hypotheses are live ones". "If I say,...'Either call my theory true or call it false,' your option is avoidable....You may decline to offer any judgment as to my theory. But if I say, ['Either accept my theory or don't'], I put on you a forced option, for there is no standing place outside of the alternative. Every dilemma based on a complete logical disjunction, with no possibility of not choosing, is an option of this forced kind." An option is momentous when (a) the opportunity it represents is unique, (b) the stake is significant, and (c) the decision is irreversible. "For our purposes we may call an option a genuine option when it is of the forced, living, and momentous kind." (WB 2-4) "When I say 'willing [or passional] nature,'...I mean all such factors of belief as fear and hope, prejudice and passion, imitation and partisanship, the circumpressure of our caste and set." (WB 9)

"The thesis I defend is, briefly stated, this: Our passional nature not only lawfully may, but must, decide an option between propositions, whenever it is a genuine option that cannot by its nature be decided on intellectual grounds...." (WB 11)

The rest of his argument depends crucially on our earlier distinction between errors of Type I and Type II: "Believe truth! Shun error!--these...are two materially different laws; and by choosing between them we may end by coloring differently our whole intellectual life. We may regard the chase for truth as paramount, and the avoidance of error as secondary; or we may, on the other hand, treat the avoidance of error as more imperative, and let truth take its chance....These feelings of our duty about either truth or error are in any case only expressions of our passional life," (WB 18) since rational grounds cannot be given for preferring Type I to Type II errors or vice versa. In our somewhat provocative language, there can be no rational grounds for preferring either skepticism or gullibility.

When evidence one way or the other is unavailable (or evenly balanced), I am obliged to decide a forced option on passional grounds--i.e., on other than intellectual or rational grounds--and hence I am fully entitled to use "the subjective method, the method of belief based on desire". (SR 97) That is, I am entitled to believe something simply because I would like it to be true--or for any other reason. The skeptic would--he notes that Clifford does--recommend the suspension of judgment in such a case. But to follow that advice is to risk a Type I error: losing the truth and its attendant benefits--just as the believer risks a Type II error: believing falsely. "Dupery for dupery, what proof is there [or could there be] that dupery through hope [of being right: the commission of a Type II error] is...worse than dupery through fear [of being wrong: the commission of a Type I error]?" (WB 27)

Finally, as promised, James applies these principles to religious belief. He first explains "the religious hypothesis". "Religion says essentially two things. First, she says that 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. 'Perfection is eternal,'--this phrase…seems a good way of putting this first affirmation of religion, an affirmation which obviously cannot yet be verified scientifically at all. The second affirmation of religion is that we are better off even now if we believe her first affirmation to be true." (WB 25-26) In case you do not yet understand exactly what religion says, James adds: "[The] feeling…that by obstinately believing that there are gods…we are doing the universe the deepest service we can, seems part of the living essence of the religious hypothesis." (WB 28)

First, James takes himself to be addressing only those who regard the option as living: if the religious hypothesis makes no appeal whatsoever to your belief, there is no point in proceeding. Second, he says that the religious option is momentous: "We are supposed to gain, even now, by our belief, and to lose by our non-belief, a certain vital good." (WB 26) So, third, if the option is forced, it is genuine. Further, it surely "cannot by its nature be decided on intellectual grounds". So if the religious option is forced, it legitimately may, because it must, be decided on passional grounds--it would be fully responsible to adopt (or reject) the religious hypothesis on whatever whim might move us.

James guarantees that the option is forced. Forcing an option is easy, mechanical, a matter of simple logic. To force an option, just say, as he does: "Either accept X or don't." This alone will force the option, but it is not very interesting. What James does with it, though, is interesting. He regularly, in this context and others, alleges that the failure to accept the religious hypothesis is tantamount to outright rejection. To simplify the point, suppose that the religious hypothesis were "There is a God." He wants us to see the suspension of judgment as tantamount to rejection--to regard the agnostic position as indistinguishable from atheism. For if he can do that, he may be able to lure into the theist fold some of the many people who find outright atheism distasteful if not unacceptable.

Consider what he says. The skeptic's "command that we shall put a stopper on our heart, instincts, and courage, and wait--acting of course meanwhile more or less as if religion were not true--till doomsday, or till such time as our intellect and senses working together may have raked in evidence enough,--this command, I say, seems to me the queerest idol ever manufactured in the philosophic cave." (WB 29-30) "If I must not believe that the world is divine, I can only express that refusal by declining ever to act distinctively as if it were so, which can only mean acting on certain critical occasions as if it were not so, or in an irreligious way. There are, you see, inevitable occasions in life when inaction is a kind of action, and must count as action, and when not to be for is to be practically against." (LWL 55) "It is often practically impossible to distinguish doubt from dogmatic negation....He who commands himself not to be credulous of God, of duty, of freedom, of immortality, may again and again be indistinguishable from him who dogmatically denies them....Who is not for is against. The universe will have no neutrals in these questions. In theory as in practice, dodge or hedge, or talk as we like about a wise scepticism, we are really doing volunteer military service for one side or the other." (SR 109) The skeptical counsel--withhold judgment in the absence of evidence--is thoroughly discredited.

James's argument is clever, charming, enticing. What is there to say? I believe that there are at least three things to say that together should remove whatever allure James's argument might have.

First, do not be tricked by his effort to exhibit the suspension of judgment as virtual rejection of the religious hypothesis. This sleight-of-hand is accomplished by his particular way of forcing the religious option. But his is not the only way. I could also force the religious option by demanding of you: "Reject God or don't!" Sensing the inadequacy of the evidence, you are unwilling to reject the God-hypothesis, so you don't. But that's tantamount to theism! Who is not against is for. "The universe will have no neutrals in these questions." This of course is preposterous. If by this simple trick of logic I can simultaneously exhibit the suspension of judgment, in the absence of evidence, as both theism and atheism, then we have to reject James's maneuver as a rhetorical ploy with no probative force at all.

Second, James insists that his authorization of groundless belief applies only when an option "cannot by its nature be decided on intellectual grounds". On the other hand, "in our dealings with objective nature we obviously are recorders, not makers, of the truth....Throughout the breadth of physical nature facts are what they are quite independently of us…." (WB 20) "The future movements of the stars or the facts of past history are determined now once for all, whether I like them or not. They are given irrespective of my wishes, and in all that concerns truths like these subjective preference should have no part…." (SR 97) In such cases, James insists that the only responsible thing to do is to seek out the relevant evidence, and suspend judgment until its dictates are clear.

I have already pointed out that on the classical view of justification, with which I am largely sympathetic, an allegedly substantive claim that cannot even in principle be confirmed or disconfirmed is empty non-sense, and hence not a candidate for my belief at all, since it does not say anything. So for me at least--and, I would think, for any rational person--an option "that cannot by its nature be decided on intellectual grounds" cannot be a living option because the hypothesis involved can make no appeal to my belief, there being nothing to believe. So for people like me at least, there can be no "genuine option that cannot by its nature be decided on intellectual grounds"; so there can be nothing for James's permission to work on.

That second point is a general and abstract one. My third and final criticism concerns his understanding of the religious hypothesis. If a claim "that cannot by its nature be decided on intellectual grounds" is literal non-sense, James's religious hypothesis is worse--it's sheer gibberish. I defy anyone to make sense of it:

"First, [religion] says that 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. 'Perfection is eternal,'--this phrase…seems a good way of putting this first affirmation of religion…." "The second affirmation of religion is that we are better off even now if we believe her first affirmation to be true." "[The] feeling…that by obstinately believing that there are gods…we are doing the universe the deepest service we can, seems part of the living essence of the religious hypothesis."

I think I know why he puts "the religious hypothesis" in this ridiculous way. He must make sure that it "obviously cannot…be verified scientifically at all". The more clearly sensible the religious hypothesis is, the more obviously susceptible it will be to rational investigation, and hence the less eligible for groundless adoption.

The following recognizable version of a religious hypothesis is not gibberish: "The natural world was brought into existence about 10,000 years ago by the will of the one eternal, omnipotent, omniscient, benevolent Being, whose son Jesus died and arose from the dead about 2000 years ago." That makes sense: I've got a good idea of what the whole sentence, and most parts of it, mean. Why does James not use this or some similarly intelligible version of the religious hypothesis? Because this and the other intelligible formulations are clearly false. In the choice between what is false and what is unintelligible, the proponent of religious belief will (almost) always opt for the unintelligible. (An aside: It is to the credit of the Biblical Literalists that they regularly opt for the intelligible--they intend the plain meaning of what they say. That it is as plainly false seems no deterrent to Fundamentalism.)

So my final criticism is that James's religious option cannot be genuine because his religious hypothesis cannot be live: it can make no appeal to my belief--I would make bold to say anyone's belief--because it is gibberish. Therefore the religious option cannot be one that we are entitled to settle on non-rational grounds.

Where does this leave us? I have been arguing that there is no successful defense of credulity--not of credulity in general, and not of religious credulity in particular. The argument, in brief, is this: William James's effort to justify credulity in religion and some other areas is far and away the best argument available--and it will not work. So there is no area of our intellectual lives in which we are entitled to lower our standards of evidence, admitting groundless beliefs. All beliefs responsibly held require rational support.

Copyright © 2004, William A. Wisdom

NOTES

¹ Part II appeared, in much the same form as here, as "Skepticism and Credulity" on pp. 96-100 of Skeptic magazine (V, 2). That article is Copyright © 1997, Millennium Press.

² The Will to Believe and Other Essays in Popular Philosophy. Dover Publications, 1956, p. 29.

³ In "The Will to Believe", William James offers the schoolboy's definition: "Faith is when you believe something that you know ain't true." This is groundless belief opposed by a mass of contrary evidence, not just belief without evidence.

⁴ In "The Will to Believe", William James offers the schoolboy's definition: "Faith is when you believe something that you know ain't true." This is groundless belief opposed by a mass of contrary evidence, not just belief without evidence. The Will to Believe and Other Essays in Popular Philosophy. Dover Publications, 1956, p. 29.

⁵ With J. S. Ullian, Random House. 1978.

⁶ JESUS’ TEACHING ON PETITIONARY PRAYER: God gives believers everything they ask for in Jesus’ name--

Matthew 17:20. And Jesus said unto them, … Verily I say unto you, if ye have faith as a grain of mustard seed, ye shall say unto this mountain, Remove hence to yonder place; and it shall remove; and nothing shall be impossible unto you.
Matthew 18:19. Again I say unto you, That if two of you shall agree on earth as touching any thing that they shall ask, it shall be done for them of my Father which is in heaven.
Matthew 21:21-22. Jesus … said unto them, Verily I say unto you, If ye have faith, and doubt not, … if ye shall say unto this mountain, Be thou removed, and be thou cast into the sea; it shall be done. And all things, whatsoever ye shall ask in prayer, believing, ye shall receive.
Mark 9:23. Jesus said unto him, If thou canst believe, all things are possible to him that believeth.
Mark 11:23-24. … Verily I say unto you, That whosoever shall say unto this mountain, Be thou removed, and be thou cast into the sea; and shall not doubt in his heart, but shall believe that those things which he saith shall come to pass; he shall have whatsoever he saith. Therefore I say unto you, What things soever ye desire, when ye pray, believe that ye receive them, and ye shall have them.
John 14:13-14. And whatsoever ye shall ask in my name, that will I do, that the Father may be glorified in the Son. If ye shall ask any thing in my name, I will do it.
John 15:7. If ye abide in me, and my words abide in you, ye shall ask what ye will, and it shall be done unto you.
John 15:16. Ye have not chosen me, but I have chosen you, and ordained you, that ye should go and bring forth fruit, and that your fruit should remain: that whatsoever ye shall ask of the Father in my name, he may give it you.
John 16: 23-24. … Verily, verily, I say unto you, Whatsoever ye shall ask the Father in my name, he will give it you. Hitherto have ye asked nothing in my name: ask, and ye shall receive, that your joy may be full.

⁷ Pensees in Pensees and the Provincial Letters, Modern Library, 85.