Review of Paul Helm’s Belief Policies

by John M. Frame

[A shorter version of this review was published in Westminster Theological Journal  57:1 (Spring, 1995), 248-251.]

Paul Helm: Belief Policies. Cambridge: Cambridge University Press, 1994. 226 pp.


Paul Helm is Professor of the History and Philosophy of Religion at King’s College of the University of London. He is also a Calvinistic Christian who has served as Associate Editor of The Banner of Truth magazine. Recently in the Journal I reviewed his The Providence of God, published by Inter-Varsity Press. The present volume is very different.

Belief Policies is a philosophy book of a very academic and technical sort. Unlike Providence, there is no attempt here to reach a popular audience. Nor is there any indication in this book of Helm’s personal religious or theological commitment, although the book does discuss issues of religious epistemology among other things. Belief Policies plays the philosophers’ game as it is played today. It analyzes and evaluates a great many concepts and arguments without digression. The organization is very tight; one gets the impression (which one did not get from Providence) that every sentence, every discussion, is precisely where it ought to be. Each sentence advances the argument in some way; this is not an easy book to skim. As much as any other philosophy book I have read, Belief Policies shows a mastery of the literature relevant to its subject.

For all its philosophical sophistication, however, this book is not only of interest to philosophers. I shall focus my attention upon its implications, and the questions it raises, for theology and for Christian thought in general.

In the book, Helm explores the role of the will in the formation of human beliefs. This is an important matter for Christian theology, which has always been concerned with the epistemology of knowing God and his world. Specifically, Scripture commands people to believe in Christ. Opponents of Christianity have often ridiculed the idea that people can be under obligation to believe something. Belief, they argue, is not under our control, so how can God require us to believe something that we do not in fact believe? In one sense, Reformed theology agrees with the objectors: what we believe is ultimately under God’s control rather than our own. But Scripture does assume that we are responsible for beliefs as for actions. Therefore, as secondary causes, we do have control to some extent over what we believe as we have control over what we do. Certainly, Romans 1 presents unbelief as the result of a voluntary decision, to “exchange the truth for a lie.” But how can that be?

At first glance it seems plausible to argue that the will has no role to play in belief. Obviously, I cannot simply will to believe anything at all. Being American, and believing that I am American, I cannot by fiat will to believe that I am Indonesian. So Bernard Williams argues that one cannot decide to believe something “just like that” (p. 46). Helm, however, who draws many parallels in this book between beliefs and actions, replies that “Breaking the smoking habit may involve the will, even though a person cannot break the habit ‘just like that'” (p. 46).

Helm points out that in ways somewhat more subtle than those Williams has in mind, the will does influence belief. People do, after all, sometimes believe without proper justification. In such cases, we are inclined to say that their belief is “willful.” There are, for example, such phenomena as wishful thinking and projection. And often people willfully fail to give proper attention to evidence, or to evaluate it according to proper standards.

Since belief is subject to “standards,” to evaluation, there is something like an “ethics of belief.” Many writers have commented on the subject of what we are obligated or permitted to believe. Helm says that Locke at one point even “makes epistemology a branch of ethics” (p. 123). (Helm himself does not seem to go that far: see pp. 26-28). Formulations of the ethics of belief have differed greatly: from W. K. Clifford’s view that nothing should be believed without sufficient evidence (pp. 90-97) to O. K. Bouwsma’s view that at least in matters of Christian belief we must “resist any attempt to ground… faith in evidence” (p. 207; discussion on pp. 202-207).

Helm gives to such principles the name “belief-policies.” He says, “As I shall use the expression a belief-policy is a strategy or project or programme for accepting, rejecting or suspending judgment as to the truth of propositions in accordance with a set of evidential norms” (p. 58). A belief-policy governs our use of evidence in developing our beliefs. In Helm’s view, belief-policies are the most important area in which the will influences our beliefs. “We choose to believe by choosing, or choosing to retain, belief-policies for acquiring, retaining, or discarding our beliefs” (p. 58).

Not all beliefs are justified by belief-policies, for “there are cases of instinctive and infantile beliefs which are justified” (p. 7). Nor are belief-policies necessarily conscious: they may be “dispositional and tacit, or the result of an overt choice” (p. 58). Belief-policies are of many kinds. In addition to the above examples, Helm discusses the belief-policies of Locke, James, and Plantinga at length and others more briefly. Belief-policies can address questions about the strength of belief, verification, falsification, permission, obligation, the importance or priority of some beliefs over against others, the degree of conservatism in maintaining or revising past beliefs, burden of proof. Some belief-policies seek to maximize the number of true beliefs, others to minimize erroneous ones; others, presumably, seek to balance these concerns in some way.

Helm uses the concept of a belief-policy to illumine the ways in which “weakness of will” and “self-deception” bear upon belief (pp. 142-163). As weakness of will can inhibit our actions, so it can hinder the implementation of a belief-policy. The belief-policy specifies the ends we seek to achieve in believing; the will seeks to reach those ends. Failure to carry through one’s belief-policy can be seen in actions as well as words.

Sometimes, through weakness of will, we fail to believe as our belief-policy dictates, even though we know what it requires. In Helm’s view, this is not self-deception (p. 149). Self-deception enters when “the agent does not recognize the wrongness of his failure to adopt or follow a belief-policy” (p. 150). The “logical form” of self-deception with respect to the belief that proposition p is true “is a person’s believing that p, while not believing that p is believed; or more strongly while it is believed that p is not believed” (p. 153). Helm therefore recognizes that a person is able to hold two contradictory beliefs at the same time.

Helm argues, therefore, that belief is to a large extent  governed by the will, so that people are responsible for their beliefs. Does this view imply the legitimacy of persecuting people for false beliefs? Helm addresses the question of toleration, pointing out that some classical defenses of toleration are compatible with his doctrine of responsibility for beliefs and belief-policies (pp. 164-188).

The last chapter, in which Helm discusses fideism, is the most obviously interesting from a theological point of view. He mentions Popkin’s definition of fideism as the claim that “truth in religion is ultimately based on faith rather than on reasoning or evidence” (p. 189). But he also presents fideism as something which occurs outside religious discourse narrowly defined: “a fideist is someone who holds that one may justifiably form a belief supported by insufficient evidence for the truth of what is believed or even unsupported by the evidence, or even in the teeth of evidence against; or that one may justifiably give a greater degree of strength to a belief than is warranted by the evidence for the proposition that is believed” (p. 189). Thus some forms of epistemological skepticism, which recommend the formation of beliefs on non-evidential grounds, may be called fideistic (pp. 195-200).

On Helm’s view, fideism is a belief-policy, but one which paradoxically sets itself against some uses of evidence and reasoning. Nevertheless, some fideists do argue for their positions, using various rational arguments. In religious discussions, for example, the divine transcendence and our obligation to accept revelation on divine authority are used to warrant fideistic epistemologies. Thus, although these thinkers disavow or de-emphasize the use of evidence in forming beliefs about God, they nevertheless affirm the use of rational argument in defending their belief-policies. That rational argument is a “second-order” defense of a “first-order” fideism (p. 193).

Helm argues that there are many differences within the family of positions called fideisms. Some are “global,” claiming that all knowledge is by faith; others limit fideism to some area(s) of knowledge, such as the knowledge of God. Some defend themselves by “second-order” rational arguments, others do not. Some defend their positions on epistemological grounds, others on moral or religious grounds. Fideisms vary as to the specific role of evidence: some seek to go beyond evidence, others to avoid any evidential defense at all. Helm notes Arvin Vos’s suggestion that Aquinas, of all people, was in one sense a fideist, because “the transcendent subject-matter of much theology necessarily goes beyond human understanding” (p. 194), although for Aquinas there is “some evidence for faith” (p. 195).

Alvin Plantinga claims that belief in God does not require evidence or argument, but may be placed among the “basic propositions” of one’s epistemic system. Nevertheless, he denies that this “Reformed epistemology” is fideistic, because in his view these propositions are “grounded” (e.g. through direct experience) and because, as epistemically basic propositions, they are among the “deliverances of reason.” Helm nevertheless finds some parallels between Plantinga’s view and some forms of fideism, particularly his assertion that the grounds for believing that God exists are person-relative or community-relative (p. 216; discussion on pp. 207-216).

At one point I do question Helm’s account of Plantinga. Helm thinks that when Plantinga says that “one who takes God as properly basic can also know that God exists” he really should have substituted “does” for “can,” in order to state his view clearly. Helm then points out that when one makes that substitution a contradiction ensues: for an unbeliever who made the non-existence of God properly basic would then “know” that God does not exist. Then the believer would know that God exists and the unbeliever would know that God does not exist. But that is impossible. In my view, Helm’s argument here raises questions about the legitimacy of his initial substitution of “does” for “can” in Plantinga’s statement (pp. 212-213).

The importance of this book for theology is that Helm has shown some of the important ways in which human belief is influenced by the will. The notion of a belief-policy is an important one. When Scripture rebukes people for their unbelief in the face of clear revelation (Rom. 1:18-32), it implicitly rebukes their belief-policy, namely their attempt to avoid proper consideration and evaluation of the evidence given by revelation. When people are regenerated by God’s Spirit, they receive a new ability to see that revelation for what it truly is. Their new openness to God’s revelation may be described as a new belief-policy. When Scripture commands repentance and faith, it implicitly commands people to renounce their old belief-policy and to adopt a new one, one which will generate new beliefs (or, on Van Til’s account, formerly repressed beliefs) about God, the world, and themselves.

Helm’s account also brings out some of the complexities of human psychology and epistemology which make self-deception possible. That also illuminates the biblical teaching that unbelief is self-deceptive, a repression of what one knows to be true.

When non-Christians object to the biblical commands to believe on the ground that belief is involuntary, Helm’s book provides us with many useful replies. The consideration of belief-policies is especially useful. For to discuss with non-Christian objectors the role of the will in the formation of belief-policies is, in effect, to question what Dooyeweerd called the “autonomy of theoretical thought.” And such a challenge to human autonomy brings us to the central concerns of Christian witness.

Some questions do remain, of course. Helm’s account of the justification of belief-policies is somewhat unsatisfying. He lists various factors that should play a role in an individual’s determination of what belief-policy to follow (summarized on pp. 140-141). Nevertheless, he insists that there “can be no rationally compelling second-order argument for the superiority of one policy over all others, though there can be fairly compelling arguments for the superiority of one belief-policy over some others” (p. 59; cf. pp. 140-141). A belief-policy governs the use of evidence, and so it cannot itself be wholly based upon evidence (pp. 59, 67). Helm connects the lack of rational compulsion in this matter to the essential role of the will. Since belief-policies are partly voluntary, they cannot be wholly based on rational considerations.

I do not think it is necessary to oppose “will” and “reason” as sharply as Helm seems to suggest. My own inclination would be to say that “will” and “reason” are simply aspects of, perhaps perspectives upon, the integrated human personality. On this basis, the will is always rational and reason always acts voluntarily. To be sure, reason is sometimes constrained, and that might seem to separate it from the will; but recall Helm’s earlier illustration of breaking the smoking habit: the will may be involved even when it operates under relative constraints. And, on the other hand, the “rationality” of the will is sometimes a defective rationality, because human reason itself sometimes functions defectively; but the will does not operate apart from reason.

Helm himself does not define “will” or “reason,” so I am not sure why he finds it necessary to say that the role of the will in belief-policies necessarily excludes a rationally compelling argument for a belief-policy. But I must ask why we cannot say that there are rationally compelling grounds for, say, a Christian belief-policy, and still maintain that people willfully reject that policy? That seems to be the implication of the biblical doctrine of the clarity of revelation in Rom. 1:18-21.

As we have seen, Helm also argues that a belief-policy cannot be based on evidence since it governs the use of evidence. Otherwise, circularity ensues. But, in my Van Tillian mode, I ask, what’s wrong with a little circularity?

Seriously: at this point, we need to define what is meant by “evidence,” as in the above discussion we needed a precise definition of “will.” If “evidence” is limited to what is “self-evident” and “evident to the senses,” then I would say that Helm is right. The grounds for adopting a belief-policy are broader than “evidence” in this narrow sense, and the belief-policy governs the use of such evidence without thereby completely controlling its own justification.

But if “evidence” includes everything that legitimately justifies the adoption of a belief, then there is no escape from the circularity in view. The grounds for adopting belief-policies are necessarily evaluated by the belief-policy in effect.

Now, why should we assume, as Helm does, that belief-policies govern evidence in the narrow sense, but not in the broad sense? Is it not the case that human beings adopt policies for dealing with all sorts of rational considerations, not only with “evidence” narrowly defined? And must not those broad belief-policies be, in the nature of the case, self-reflective and, indeed, self-justifying?

Van Til’s point, of course, was that a Christian belief-policy, taken as a datum of revelation, is self-attesting, self-justifying. It is self-attesting because it is itself supremely authoritative, not dependent on any other authority for its verification. It does not, therefore, depend for its justification on any other belief-policy. In that sense, its justification is “circular.” Taken as a datum of the believer’s consciousness, however, rather than as a datum of revelation, the Christian belief-policy does have an external justification– in divine revelation itself.

I have defended Van Tillian “circularity” in my Doctrine of the Knowledge of God (Phillipsburg, NJ: Presbyterian and Reformed, 1987), pp. 130-133, and in Apologetics to the Glory of God (Phillipsburg, NJ: Presbyterian and Reformed, 1994), pp. 9-14, and in Chapter 22 of Cornelius Van Til (Phillipsburg: Presbyterian and Reformed, 1995). I shall not, therefore, repeat those arguments here. I do think that circularity of a sort is unavoidable when one is seeking to justify what we might call, in terms of the current discussion, an ultimate belief-policy, a policy governing all beliefs.

Nevertheless, Helm’s book has an excellent grasp of the fact that voluntary factors enter into human knowledge, and that they enter precisely at the point where people seek to justify their beliefs. The book is therefore a powerful weapon against claims to “neutrality,” against claims that knowledge is religiously unproblematic. In my view it takes some large steps in the right direction, even if it has not reached its proper destination.


Sign up to receive new posts via e-mail