An Essay on Belief and Acceptance.

Bezuidenhout, Anne

Cohen, L. Jonathan. Oxford: Clarendon Press, 1992. 163 pp. Cloth, $35.00; paper, $18.95--As the title suggests, this book is centered around a distinction between belief and acceptance. A parallel distinction is drawn between desire and intention. Cohen argues that acceptance and intention are voluntary states, whereas belief and desire are involuntary dispositions. Acceptance is active, whereas belief is passive. Acceptance is subjectively closed under deducibility (that is, anyone who accepts certain premises and accepts that some proposition follows from those premises, accepts that proposition), whereas belief is not. Acceptance is an all-or-nothing affair, whereas belief comes in degrees, ranging from having an inkling that something is the case at one extreme to feeling certain that something is the case at the other. Similar contrasts can be drawn between desire and intention.

To believe that p is to be disposed to feel that it is true that p, whereas to accept p is to adopt the policy of using p as a premise in one's deliberations. Belief and acceptance often, perhaps even usually, coincide. If one believes that p, this is a prima facie reason for accepting p. And if one accepts that p, it may be that one will come to believe that p. That is, the acceptance will induce the belief. But there are also cases in which belief and acceptance fail to coincide. Similarly, although the desire for something often goes along with the intention to bring that thing about, there are cases in which they come apart. Cohen's interest is in those cases in which these fail to coincide.

For instance, it seems correct to think of a jury as deliberating on the basis of admissible evidence in order to arrive at a verdict in which they either accept that the accused is guilty or accept that he or she is innocent (sec. 20). However, it may be that although the jurors accept that the accused is innocent, they cannot bring themselves to believe in the accused's innocence. Beliefs arise involuntarily, sometimes as the result of irrational factors which are difficult to control (such as racial prejudice). So even though one' accepts that the balance of the admissible evidence does not establish beyond reasonable doubt that the accused is guilty, these irrational factors may continue to operate and produce a feeling, even a strong one, that the accused is guilty.

Cohen's principal aim is to show that certain long-standing puzzles and problems in philosophy can be solved once we recognize that these, too, are cases in which belief and acceptance fail to coincide. Because philosophers have failed to distinguish belief and acceptance, they have tended to cast these problems in terms simply of beliefs (and desires) and hence have been led to see contradictions and conundrums where there are none.

This, Cohen argues, is the case with the paradoxes of self-deceit and weakness of will (sees. 2-28). For instance, it has seemed to some that self-deceit must be described as a case in which a person simultaneously believes and fails to believe a certain proposition. We should instead see self-deception as a situation in which someone believes that not-p, yet accepts that p. Since belief and acceptance are two different cognitive attitudes, we resolve the apparent antinomy. Of course, in order to account for the element of intellectual dishonesty in self-deception, we need to add that the self-deceiver represses the belief that not-p by ignoring (either intentionally or unintentionally) the evidence which runs counter to what he consciously accepts.

There are many other issues which are cast in a new light once one makes the distinction between belief and acceptance (and the analogous distinction between desire and intention). For example, Cohen revisits debates in epistemology concerning whether knowledge consists in justified true belief (sees. 16-18); debates in action theory concerning the nature of action explanations and as to whether reasons are causes (sees. 7-12); and debates in confirmation theory concerning whether subjective probabilities are a measure of strength of belief (sec. 19).

Cohen's distinction is an important one, which has indeed been neglected, to the detriment of discussions in epistemology and action theory. To take just one example at random: there is currently a debate (among Kihyeon Kim and Paul Tidman, as well as others) as to whether we have an epistemic duty to critically evaluate our beliefs in order to root out those which fail to live up to our epistemic standards. Some writers think that the notion of such a duty is impugned by the fact that it entails doxastic voluntarism (namely, the view that we can will to believe something). But if beliefs are indeed involuntary, this casts doubt on the claim that we have such a duty. Nevertheless, if one keeps Cohen's distinction in mind, the various parties in the debate can be seen to be at cross purposes. One can agree that beliefs are involuntary, but this by itself does not refute the deontological conception, for epistemic duties might still have a role to play with regard to what we accept. Perhaps we have an epistemic duty to accept only what survives critical scrutiny, and since we can accept and reject things at will, the notion of such a duty is not problematic.

However, sometimes Cohen's resolution of problems in terms of his distinction is less satisfying. For instance, Cohen says that to explain someone's behavior by appeal to his beliefs and desires is to cite the causes of his behavior. On the other hand, to explain behavior in terms of acceptance and intention is to exhibit the logical relations between the behavior and its reasons, in a way which will make the behavior intelligible to others with a similar capacity for rational action. (Cohen also allows the possibility of mixed explanations; that is, one might explain an action by appeal to a belief and an intention, or by appeal to what someone accepts together with what he desires.)

Cohen stresses that normal adult human beings have cognitive states and dispositions of all four sorts. Hence both causal and logical explanations will be applicable to adult humans, though perhaps not always in the same situations. Cohen concludes that the question "Are reasons causes?" does not admit of a single answer. Sometimes reasons are causes, when the reasons are beliefs and desires. Sometimes reasons are not causes, when the reasons are what the person accepts and intends.

However, it is not clear that the debate about reasons and causes can be settled so easily. If what one accepts and intends are not causally efficacious, it may seem that appeals to acceptance and intention can at best rationalize one's behavior, not explain it. Cohen argues that it is only in those cases in which belief and acceptance or desire and intention come apart that an appeal to acceptance and intention will seem to be a rationalization rather than a genuine explanation. But why should the mere coincidence of acceptance and intention with belief and desire make the difference between a mere rationalization and a genuine explanation? One answer is that when acceptance and belief coincide they are causally integrated with one another. This would indeed give us reasons for supposing that explanations in terms of acceptance and intention are more than mere rationalizations, but only at the expense of making such explanations a species of causal explanation, contrary to what Cohen intends.

However, even if one does not agree with Cohen on every score, his book contains a wealth of insights and anyone with an interest in epistemology and action theory has much to gain by a careful study of Cohen's arguments.