Belief

views updated

BELIEF

Beliefs are a species of propositional attitude distinguished by their having the mind-to-world direction of fit.

Propositional attitudes are psychological states characterized by a psychological mode, Ψ, and a propositional content, P, schematically: Ψ(P ). My belief that the earth moves has belief as its psychological mode, and that the earth moves as its propositional content. A desire that the earth move has the same propositional content, but a different psychological mode, desire. Within a psychological mode, propositional attitudes are distinguished by their contents. I could not have two beliefs with the content that the earth moves. Many, though not all, propositional attitudes admit of a bivalent evaluation. Beliefs are true or false. Desires are satisfied or unsatisfied. Intentions are carried out or not carried out. Propositional attitudes with a bivalent evaluation have either the mind-to-world direction of fit or the world-to-mind direction of fit (Searle 1983, chapter 1). Its direction of fit expresses the basic function of a propositional attitude in our mental economy. Beliefs aim to represent how the world is independently. They aim at truth. The belief that Solomon was wise is true if and only if (iff) its content matches the world, that is, iff Solomon was wise. Belief's aim to represent how the world is independently is reflected in its being irrational to retain a belief when one sees that it does not match the world. Thus, beliefs have the mind-to-world direction of fit. A desire, in contrast, seeks not to match how the world is independently, but for the world to come to match its content. Desires seek satisfaction. The desire that a toothache go away is satisfied iff the world comes to match its content, that is, iff the toothache goes away. It is not irrational to retain desires known to be unsatisfied, for seeking satisfaction gives them their point. Desires thus have the world-to-mind direction of fit. Beliefs and desires, in virtue of their opposite directions of fit, have an interlocking role in the production and explanation of action.

Belief, Sensation, Experience, and Concept

Beliefs (and other propositional attitudes) must be distinguished from sensations, sensory images, and experience, on the one hand, and concepts, on the other. The classical British empiricists of the seventeenth and eighteenth centuries—John Locke (1632–1704), George Berkeley (1685–1753), and David Hume (1711–1776)—were unable to provide an adequate account of belief because they assimilated all of these to sensations or sensory images, like the taste of apple or a toothache. But sensations are not an adequate model for belief, or for other propositional attitudes. Sensations are not true or false, or satisfied or unsatisfied. They do not admit of a bivalent evaluation, as propositional attitudes do. They do not have propositional contents. Their differences are differences of qualitative feel. These differences are not variations in psychological function, as are the psychological modes of belief and desire. In particular, propositional contents are required to make sense of the logical relations that obtain between beliefs, and which are crucial to understanding the role of beliefs in reasoning and action. For example, understanding the validity of the inference from the belief that gold is a metal, and the belief that this ring is gold, to the belief that this ring is metal, requires seeing the logical connections between the propositional contents of the first two beliefs and the last, and their shared elements. Similarly, as explained below, the logical relations between the contents of belief and desire are crucial to understanding rational action (see the discussion of the practical syllogism below). Since sensations lack propositional contents, treating beliefs as a subclass of sensations obliterates distinctions needed to understand the role of beliefs in thought and action.

Perceptual experiences, unlike sensations, can be veridical or nonveridical. A hallucination of a rhinoceros on the sofa is a nonveridical perceptual experience. It represents what is not so. Perceptual experiences are like beliefs in admitting of bivalent evaluation and having the mind-to-world direction of fit. But though many beliefs are based on perceptual experience, they are a different coin. Perceptual experience is a fieldlike representational medium and makes use of the qualitative features of modes of consciousness associated with different sensory modalities (visual, auditory, tactile, etc.) to represent how things are around us. Beliefs, in contrast, do not in the same way make use of qualitative features of modes of consciousness to represent. Their mode of representation is purely propositional. Beliefs are to perceptual experiences as statements are to maps. Perceptual beliefs, those based directly on perceptual experience, in effect abstract from the richer representational content of perceptual experience.

Just as beliefs, sensations, and experiences must be distinguished from each other, they must also be distinguished from the shared elements, concepts, in different attitude contents. The concept of gold, for example, is shared between the belief that gold is a metal and the belief that this ring is gold. It is a constituent of both contents. Concepts are neither true nor false, though they apply or fail to apply to things. Sensations neither have such constituents nor are identical with them, for belief contents do not have sensations or images as constituents. Experiences involve concepts, much as beliefs do. A visual experience as of a baseball represents a spherical object as a baseball. The experience is distinct from the concept of a baseball, but represents something visually presented as falling under the concept. Without the concept, there could be no such visual experience.

Someone who believes that gold is a metal possesses the concept of gold and the concept of metal. Possessing a concept requires having beliefs expressing the simplest conceptual connections that the concept enters into. For example, someone who possesses the concept of a gun must believe that guns are weapons, that they are physical objects, that they can be aimed, and the like. Perhaps no precise set of beliefs is required, but if a person lacks a network of beliefs articulating the connections of a concept with other concepts, he does not possess the concept. This shows that we can make sense of attributing a belief to someone only by locating it in a pattern of beliefs, and that the other propositional attitudes and perceptual experience presuppose belief.

Beliefs and the Explanation of Actions

Beliefs, because they aim at truth, play a central role in theoretical reasoning (reasoning about what is so), and hence in practical reasoning (reasoning about what to do).

Theoretical reasoning is central to practical reasoning because we get what we want by doing something that promotes it. We therefore need to know what we can do, and how what we can do is related to what we want. When seeking knowledge of these things, we seek true beliefs about them. Thus, what we do is conditioned by what we believe, whether our aim at truth hits the mark or not. Accordingly, when one explains an action, it is not enough to cite a desire that the action satisfies. One must also cite a belief that the action increased the likelihood of satisfying the desire. If I want my rival to come to grief and an idle comment of mine leads to his downfall, my desire that he come to grief does not help explain my bringing about his downfall if I did not think that my idle comment would have that as a consequence.

An action explanation provides the materials for a practical syllogism that justifies the action from the point of view of the agent. Suppose that I lifted my finger because I wanted to signal you and believed that my lifting my finger, F, would constitute signaling you. One can construct the following argument in favor of this action, drawing the evaluative premise from the desire and the factual premise from the belief:
My signaling you is desirable (has a desirable aspect).
F, being lifting of my finger, is a signaling of you.
F is desirable (has a desirable aspect).

This does not show that the action is desirable all things considered, but only that it has a desirable aspect. Actions have many consequences and properties, some of which one may want and others of which one may not. In deciding all things considered what to do, the agent must rank his desires and take into account his degrees of confidence in desired outcomes attending certain actions, or, in a common phrase, his degrees of belief in outcomes, given the actions. Candidate actions are evaluated on the basis of the value of their results and one's degree of confidence in their having those results. If the chance is low but the value great, the undertaking may still be judged best overall. The notion of degree of belief in a proposition is often treated as a generalization of the ordinary notion of belief. Typically, on buying a ticket, one does not believe that one will win the lottery, though one does have a nonzero degree of belief that one will. However, it may be that degrees of belief in a proposition can be assimilated to beliefs about its probability.

Behaviorist Theories of Belief

Logical behaviorism, a form of materialism, holds that ascribing beliefs and desires and other mental states to an agent is just a compendious way of describing a complex pattern of his actual and potential behavior. On this view, there are no inner mental states or events—no "ghost in the machine," in Gilbert Ryle's memorable phrase. If logical behaviorism is true, action explanation is not causal explanation, but rather functions by locating some particular behavior in a broader pattern of behavior. Logical behaviorism was championed by the logical positivists of the 1930s (see Ayer), and in more subtle forms by Ryle and Ludwig Wittgenstein in the 1940s and 1950s. A central motivation for the logical positivists was their verificationist criterion of meaning, according to which the meaning of a statement should be sought in the conditions that verify or falsify it.

Logical behaviorism fell from fashion after the Second World War (Block). One reason was disenchantment with the verificationist criterion of meaning. A second was the failure to provide necessary or sufficient conditions for attributing psychological states in purely behavioral terms. This failure is connected with the interlocking roles of belief and desire in explaining behavior, and carries an important lesson about the relation of belief (and other propositional attitudes) to behavior.

Let us try to say what behavior is characteristic of the belief that there is an apple pie in the pantry. It is clear that what behavior we expect will depend on what the agent wants and what else she believes. If she is not hungry, we expect no tendency to investigate the pantry. If she likes apple pie but wants to save it for the guests more than to indulge herself, we expect a delayed advance on the pantry. If she believes it is poisoned, we expect her to dispose of it. And so on. The important point is that the behavior we expect from someone who has a particular attitude is conditioned by the other psychological states that we think he has. This makes it impossible to state, in purely behavioral terms, what it is to believe, for example, that there is a pie in the pantry.

Once one understands the role of desire, in particular, in action, one can see that it will play the spoiler for any behaviorist program. For desires can be about behavior usually taken to be a sign of a given sort of psychological state. In particular, one may want to exhibit misleading behavior. One can pretend to believe or want things one does not, and one can suppress behavior that expresses what one wants. It seems plausible, as Hilary Putnam argued in "Brains and Behavior," that the limited deceptions with which we are familiar could take forms that would preclude any behavioral expression of some of an agent's psychological states throughout his life.

Beliefs as Causes of Behavior

The moral is that beliefs issue in behavior only in conjunction with appropriate other propositional attitudes. Desires motivate behavior, but beliefs guide it. Each is impotent without the other. To vary a phrase of Kant's, desire without belief is blind; belief without desire is empty. We cannot read back from behavior to the motivating belief and desire, because any given behavior may issue from different sets of beliefs and desires. Behavior is related to belief and desire as symptoms to a disease. The symptoms may be expressed in the absence of the disease, and the disease may be present without being expressed by any symptoms, or by the usual symptoms. The reason is that the disease is a cause of the symptoms. Its giving rise to the usual symptoms depends on the usual background conditions being present. This analogy suggests that, as with disease and background conditions, beliefs and desires are interlocking causes of behavior. This would explain the failure of the behaviorist programs, since, as Putnam put it, "causes are not logical constructions out of their effects " (p. 27).

This conclusion is bolstered by an argument by Donald Davidson in his seminal paper "Actions, Reasons, and Causes." We sometimes have multiple reasons (belief and desire pairs that potentially explain an action) for doing something, but we act on only one of them. I may believe that if I do not obey the speed limit when driving, I will likely receive a ticket, and I may wish to avoid receiving a ticket. I may believe also that obeying the speed limit is the right thing to do, and wish to do the right thing. I may obey the speed limit for the first reason rather than the second, or for the second rather than the first. In either case, each reason would justify what I do, but only one would explain it. We can make sense of this being so if we hold that the reasons for which I act are those that cause my action.

Voluntarism about Belief

Beliefs play a role in guiding and explaining action, but can they be the products of actions? Can one come to believe something by choosing or deciding to believe it? According to voluntarism about belief, the answer is yes. René Descartes (1596–1650), "the father of modern philosophy," apparently endorsed belief voluntarism. In his explanation in the Meditations (pp. 37–43) of how we fall into error despite God's supreme benevolence, he assumed that when one lacks adequate evidence one can choose to believe something and hence fall into error through the exercise of free will, and thereby absolve God of responsibility for the error. With his famous wager in Pensées (pp. 151–153), Blaise Pascal (1623–1662) argued that the infinite utility that attaches to believing truly that God exists means that no matter how small the probability, one is rationally compelled to belief. This argument likewise seems to assume belief voluntarism. In "The Will to Believe," William James (1842–1910), like Pascal, argued that it is not only possible but sometimes rationally required that we make decisions about belief when evidence underdetermines choice.

One may, of course, bring oneself to believe something indirectly. I may pay another to brainwash me. I may adopt the outer forms of religious faith in the hope that inner conviction will follow. But the issue is not whether I can bring myself to believe something by doing something else that brings it about, but whether I can do this without doing anything else to bring it about.

This seems not to be something that is typically within our power. I cannot just decide now to believe that I do not have hands or a head, or that I am flying, or fabulously wealthy. Religious belief, which is less engaged with the practical, is a more difficult case and has historically been the focus of the debate about belief voluntarism. One can try to inculcate religious belief indirectly, but can one simply decide to believe that God exists, while also believing one lacks adequate evidence? One might answer yes because it can be reasonable to continue to believe that God exists in the face of doubts. But this falls short of what is required. For this is not deciding to believe, but rather deciding not to give up a belief one already has.

Still, we are sometimes faced with an unavoidable practical choice where evidence bearing on a crucial fact leaves the fact, and hence the choice, undetermined. Must we not then make a choice about what to believe despite not having reasons to believe one way or the other? No. We must make the practical choice about what to do. But this does not imply belief. When some action is better than none, we take a chance and hope for the best, without belief.

Is the situation arguably different when a belief itself has a practical value? It is dubious that a belief itself having a practical value makes it easier to conceive of choosing to believe. If someone were to offer me a million dollars to believe that there is life on the Sun, I might wish to do so, but I would be baffled about how to comply.

The purpose of belief is to represent the world accurately. Therefore, belief serves its role only if the formation, retention, and revision of belief are sensitive to what one takes to be one's evidence. This does not mean that we always believe in proportion to our evidence. We make mistakes of assessment and reasoning; we are lazy; we fail to attend to evidence we have. Nor does it mean that what we take to be evidence is evidence, or that all our opinions were entrenched with the spade of reason. Further, it does not mean that belief is never influenced by desire. Otherwise, wishful thinking—believing what one wants to be true—would not be possible. Rather, it means that belief, by its nature, aims at truth, that its function is undermined if one lets belief formation be sensitive to anything other than what one takes to be evidence. Where there is uncertainty, one may be conservative and persist in a belief. But not to give up a belief in the face of strong contrary evidence is irrational. Worse still is to believe where no evidence bears, as in the case of wishful thinking. Even when belief has a practical benefit—as when believing one will win a race increases one's chances, but not enough to warrant the belief—rationality is at best at war with itself. Belief voluntarism thus seems to be something that can occur, at best, only on the fringes of rationality—in the shadow regions of self-deception.

Bibliography

Ayer, Alfred Jules, ed. Logical Positivism. New York: Free Press, 1959.

Block, Ned. "Psychologism and Behaviorism." Philosophical Review 90 (1) (1981): 5–43.

Davidson, Donald. "Actions, Reasons, and Causes" (1963). In his Essays on Actions and Events. New York: Clarendon Press, 2001.

Descartes, René. "Meditations on First Philosophy" (1641). In The Philosophical Writings of Descartes. Cambridge, U.K.: Cambridge University Press, 1984.

James, William. "The Will to Believe" (1897). In his The Will to Believe and Other Essays in Popular Philosophy. Cambridge, MA: Harvard University Press, 1979.

Pascal, Blaise. Pensées (1670). Harmondsworth, U.K.: Penguin, 1966.

Price, H. H. Belief: The Gifford Lectures Delivered at the University of Aberdeen in 1960. London: Allen and Unwin, 1969.

Putnam, Hilary. "Brains and Behavior." In Readings in Philosophy of Psychology, Vol. 1, edited by N. Block. Cambridge, MA: Harvard University Press, 1980.

Ryle, Gilbert. The Concept of Mind. New York: Barnes and Noble, 1969.

Searle, John. Intentionality: An Essay. Cambridge, U.K.: Cambridge University Press, 1983.

James, William. "The Will to Believe" (1896). In his The Will to Believe and Other Essays in Popular Philosophy. Cambridge, MA: Harvard University Press, 1979.

Kirk Ludwig (2005)

Encyclopedia of Philosophy