The Experiential Basis of Theism

William P. Alston


William P. Alston, principally known for his work in epistemology and philosophy of language, received his Ph.D. from the University of Chicago in 1951. From 1949-71 he taught at the University of Michigan, 1971-76 at Rutgers, 1976-80 at the University of Illinois, and from 1980 at Syracuse University. A frequent panelist and seminar director for the National Endowment for the Humanities, he has also been an active committee member and officer in various divisions of the American Philosophical Association, including President of the Western Division during 1978-79. He also served as President of the Society for Philosophy and Psychology (1976-77) and President of the Society of Christian Philosophers (1978-81), which he helped to co- found. He has served as an editorial consultant to the American Philosophical Quarterly, Nous, Monist, and other professional journals, and is currently Editor of Faith and Philosophy. He has received a number of fellowships over the years, including a fellowship at the Center for Advanced Study in the Behavioral Sciences at Stanford. He has authored or edited four books, including Philosophy of Language in the Foundations of Philosophy series. He has written numerous and often reprinted articles, including eighteen entries in the Encyclopedia of Philosophy, in journals such as the Journal of Philosophy and the American Philosophical Quarterly.
I will try to explain, so far I can in the space allotted, why I think it is rational to be a theist. Rather than being equally sketchy about everything, I will pick one or two issues for special consideration.

First, a word as to how I understand "theism." Rather than tailoring the term to an exact fit with the details of my own beliefs, I will give it a latitudinarian reading. Theism is the view that there is a single, ultimate and supremely perfect source of being, on which all other than itself depends at every moment for its existence, and which appears to us as personal. (The last clause is designed to include those who hold that the ultimate source of being is only "analogically" or "symbolically" describable in personalistic terms.) Theism is only an abstract aspect of elaborate religious systems, but it may be usefully singled out for discussion. However, the involvement of theism in a larger context, comprising not only a richer matrix of belief but a "form of life," is crucial for the ensuing discussion. I will make no attempt to defend the rationality of theism as a philosophical theory, embraced outside of participation in the life of a theistic religion.

I conceive theistic belief to rest on twin pillars, of unequal strength. First, there are the very general considerations of "natural theology," accessible to any sufficiently intelligent and reflective person; they give some support to the belief. Of the traditional arguments for the existence of God, I take most seriously the cosmological and ontological arguments. Without claiming for them anything like coercive proof, it seems to me that the cosmological argument, properly construed, provides a significant support for the view that the physical universe depends for its existence on a necessarily existing source of all being other than itself. And the ontological argument, properly construed, provides significant support for the view that this necessarily existing source of all being is an absolutely perfect being.

The other pillar is the experience of God. It is reasonable to believe in the existence of God because we have experienced His presence and activity in our lives. I consider this to be the more massive pillar. As I see it, the proper role of natural theology is not to bear the whole weight, or even most of it, but to provide basis for reassurance, in moments of doubt, that what we take to be our experience of God is not merely a projection of our needs and fears. I am not concerned to argue that a theistic belief resting wholly on natural theology is a rational belief. Because of this, and because I feel that I have something more distinctive to say about the experiential pillar, I will confine my discussion to that side of the matter.

My first task is to explain how I am thinking of this experiential support. Many discussions of the "argument from religious experience" treat this as one more piece of natural theology. Like the teleological or moral arguments it begins with certain data, in this case religious experiences; and the question is as to whether an adequate explanation of these data requires an appeal to God. That is not the way I am thinking of an experiential support for theism. I am thinking of it as analogous to the experiential support all of us have for the existence of the physical world. Here it is not a matter of gathering data, in the form of sensory experiences, and then claiming that they are best explained by supposing that they are due to the action of the physical world on us. It is rather that we take ourselves, in these experiences, to be aware of the physical world. No explanatory issues are involved. To be sure, a few philosophers have attempted to carry through the explanatory argument just mentioned, but they have been conspicuously unsuccessful; and in any event that is not our normal stance. By analogy, I am claiming that a participant in a theistic religion takes herself to be aware of God at various times, that it is rational for her to do so, and that because of this theistic belief is rational for her.

Note that I have compared the experiential support for theistic belief to our experiential support for belief in the physical world, not to our experiential support for belief in, e.g., trees, much less to my experiential support for the belief that there is now a tree in front of me. With respect to the range of experience I am considering (call it "theistic experience" faute de mieux) God is not one of the items disclosed among others. We don't consult our experience of, e.g., leading the Christian life, to determine whether God is there, rather than someone or something else the way I "consult" my visual experience, i.e., take a look, to see whether that Mercedes sedan is still in front of my house. Similarly, we don't take a closer look or carry out more extensive observations to determine whether there is a physical world. The supposition that there is a physical world (that there are physical things spread out in space, exhibiting various perceivable qualities) is constitutive of the practice of forming particular beliefs about particular physical things on the basis of sense experience in the way we usually do. (Call this "perceptual practice".) To be sure, this very abstract belief in the existence of the "physical world" is not disengaged and explicitly assented to at the earlier stages of mental development. Nevertheless, in learning to form physical-object beliefs on the basis of sense experience we are, at least implicitly and in practice, accepting the proposition that the physical world exists (and that we are aware of it in sense experience). Thus the question of the rationality of this belief is the question of the rationality of perceptual practice. Since, for us at any rate, to accept that belief is to engage in that practice, the belief is rational if the practice is. Analogously, the relatively abstract belief that God exists is constitutive of the doxastic practice of forming particular beliefs about God's presence and activity in our lives on the basis of theistic experience. (Call this "theistic practice".) Here, too, the question of the rationality of that belief, insofar as it rests on experiential support, is the question of the rationality of that practice. In neither case is there any question of using particular experiences to settle the question one way or the other, the way I use particular experiences to settle the question of whether my wife is in the kitchen or of whether God told me that I am handling this problem in the right way.

This is not to say that we can't, or don't, discover God, or the physical world, by experience. We can discover one or the other by coming to participate, reasonably, in one or the other doxastic practice. By learning to engage in the practice we discover various particular facts about God or the physical world; and thereby come to discover God or the physical world. The present point is just that these discoveries are much more global affairs than the discovery of some scientific fact about God or the physical world; they are tied to the practice as a whole rather than to certain particular exercises thereof.

This talk of doxastic practices is reminiscent of talk of language games and the like; and lest I find myself tarred with such dreadful appellations as "Wittgensteinian fideism" let me hasten to make it explicit that I am thinking of these practices in a "realistic" fashion, rather than in an idealist, "pragmatist", culturally relativist, or linguistically solipsistic way. I am not supposing that we create divine or physical "worlds" in Goodmanian fashion by engaging in these practices. I am not supposing that the realities we encounter therein depend for their existence in any way on our conceptual or linguistic activities. Nor am I supposing that the question of their existence can be raised only within the practice in question. (My limited reliance on natural theology is an indication of this.) For better or for worse, I believe that in sensory experience and in theistic experience we encounter and learn about realities that could be just as they are had we not been around to record the fact.

I should say a word about the scope of my term "theistic experience." I mean it to range over all experiences that are taken by the experiencer to be an awareness of God (where God is thought of theistically). I impose no restrictions on its phenomenal quality. It could be a rapturous loss of conscious self-identity in the mystical unity with God; it could involve "visions and voices"; it could be an awareness of God through the experience of nature, the words of the Bible, or the interaction with other persons; it could be a background sense of the presence of God, sustaining one in one's ongoing activities. Thus the category is demarcated by what cognitive significance the subject takes it to have, rather than by any distinctive phenomenal feel.

We have seen that the rationality of theistic belief, in so far as that rests on an experiential base, depends on the rationality of a certain "doxastic" practice, what we have called "theistic practice." But how do we decide that? The evaluation of doxastic practices is a thorny matter, indeed, and I shall only be able to touch on a few points in this paper.

The first point is that where we are dealing with what we might call a basic practice, one that constitutes our basic or most direct cognitive access to a subject matter, it is not to be expected that we can show it to be rational or irrational in a non-circular way, by using only the output of other practices. Very roughly the reason for that judgment is this. The most important consideration in deciding on the rationality of a doxastic practice is its reliability, the extent to which it can be depended on to yield true rather than false beliefs. But if the practice in question, P, is our basic access to its subject matter, we will have no independent way of comparing P's deliverance with the real facts of the matter, so as to determine the accuracy of P's deliverance. I believe that a survey of attempts to establish, from the outside, the reliability, or unreliability, of doxastic practices plausibly taken as basic, like sense perception and introspection, will support this judgment. Thus we should not suppose that unless we can establish the reliability of theistic practice on the basis of premises taken from other practices we will have to deem it irrational. If we imposed that requirement across the board we would be jettisoning sense perception, introspection, rational intuition, and deductive and inductive reasoning; we should be bereft of belief. And how could we justify making this demand of theistic practice alone?

Well, then, what can we do to assess the rational status of a basic doxastic practice? The obvious alternative is to look at its internal "coherence", ways in which it might be self-supporting. This can be illustrated by some of the ways in which perceptual practice is self-supporting. Engaging in this practice and relying on its outputs, we discover (a) that perceptual beliefs formed by a subject at a time can be confirmed by other subjects or by the same subject at other times; and (b) that by relying on its inputs as a basis for reasoning we can discover regularities, thereby putting us in a position to effectively anticipate the course of events and exercise control over them. This does not constitute a non-circular argument for the rationality or reliability of perceptual practice. It is only by relying on the outputs of this practice that we get our reasons for supposing that the practice exhibits these features. (We don't learn this from ESP or from divine revelation.) Nevertheless there are ways in which perceptual practice supports itself from the inside; they are fruits of the practice that encourage us to engage in it and to take it seriously as a source of information. Nor is this self-support a trivial matter. It is conceivable that our perceptual beliefs and what is based on them should not hang together in this way; hence the fact that they do so is a mark in its favor.

Now it is clear that theistic practice does not exhibit these kinds of self-support. Its deliverances do not put us in a position to predict the doings of God, much less control them. Nor is the same sort of intersubjective corroboration possible. This latter point divides into two. First, the community of practitioners is much less extensive. And second, even within that community there are no definite criteria of confirmation. If I suppose that I have seen an airplane flying over my house at a certain moment there are principles that determine what kinds of sensory experiences by observers in what spatio-temporal locations will confirm or disconfirm my belief. But there are no such principles governing the intersubjective confirmation of my supposition that I was keenly aware of God's sustaining love at a certain moment. If no one else was aware of it at that moment, what is that supposed to show?

These disanalogies, and others that might have been mentioned, have been taken by some philosophers to discredit the claims of theistic practice to be a cognitive access to an objective reality. But this is no more than a piece of epistemic imperialism, as much so as the requirement that the reliability of theistic practice be established in other practices, or the requirement that the existence of God be shown by other practices to be the best explanation of theistic experience. It is only that the imperialism is not quite so easy to detect here. In this guise the imperialism consists in taking the mode of self-support exhibited by perceptual practice to be the standard for all other doxastic practices, at least all other experiential doxastic practices, practices of forming beliefs on the basis of experience. If a practice does not hang together in this way, it is to be dismissed; this mode of success is taken to be a necessary as well as a sufficient condition of rationality.

To see what is wrong with this we need to reflect on the point that what features of a doxastic practice betoken reliability is a function of the nature of the subject matter and our relation thereto. If the subject matter is maximally stable, as in mathematics, then always yielding the same beliefs about the same objects is an indication of reliability; but if the subject matter is in constant flux that same feature would rather betoken unreliability. If the subject matter is inorganic nature, the formation of beliefs concerning intelligible communications from the objects would indicate unreliability, but not if the subject matter is persons. What features indicate reliability depends on what features are such that they could be expected to be present if the practice yields mostly truths, given the nature of the subject matter and our relation to it. And how do we tell what the subject matter is like? If the practice in question is a basic one, we have to rely on it to tell us. Thus, a practice sets its own standards for judging its reliability, as well as providing the data for that judgment. It not only grades its own exams; it even provides the criteria for grade assignment.

Let's apply this to the present issue. The picture of the physical world and our perceptual relation thereto that is built up on the basis of sense perception is that of a realm of items that behave and interact (including interactions involving human percipients) in lawful ways that are, to a considerable extent at least, humanly discoverable. This picture provides grounds for supposing that any veridical perceptual experience will be lawfully related to the perceptual experiences of others in specifiable ways that form an effective basis for intersubjective testing. Moreover, the picture suggests that the predictive capacity one gains from a knowledge of general laws is a reasonable expectation, one of the ways in which the by and large veridicality of perceptual beliefs would show themselves. But, by contrast, the picture of God and God's relations to the world that is built up on the basis of theistic practice is quite different. God is too transcendent, too "wholly other" for us to expect to be able to predict, much less control, His behavior, however accurate our particular experiences of His presence. And it is part of that unpredictability that we cannot expect to ascertain lawful regularities in His manifestations to our experience. Even if our particular readings of God's presence were 100% accurate, we could not expect to discover the general conditions under which a human being would experience a certain divine activity if it were really taking place. Insofar as we have any idea of what conditions are required for an individual's being aware of God, they have to do with subtle factors like openness and purity of heart, factors for the presence of which we hardly have an effective test. Therefore, since there is no reason to expect these achievements, however reliable the practice, their absence is no reason to brand it as unreliable, or even to doubt its reliability. It is a further and very important question, into which I will not have time to go into in this paper, how theistic practice, if reliable, could be expected to display that reliability, what fruits of the practice are such that, if they are forthcoming, the practice can be considered self-supporting in a way appropriate to it.

I take these considerations to be a decisive refutation of the claim that theistic practice can be branded as unreliable and/or irrational because it fails to exhibit the kind of self-support characteristic of perceptual practice. Nevertheless, the tendency to take sense perception as our model of a veridical experiential access to reality is so deeply rooted that it will be worthwhile to consider the matter from another angle.

I believe that it will help us to cast a critical eye on this tendency if we try the experiment of looking at perceptual practice from the standpoint of other doxastic practices and consider what judgment on its rationality might be indicated from those other perspectives. Nor do we have to rely on our own imaginations alone for this task. We only have to remember the low esteem in which sense perception has been held by many philosophers from Parmenides to Descartes. Plato, for example, held that sense perception cannot be regarded as a source of knowledge because its objects lack the stability required for true knowledge; moreover they are not fully determinate; there is no definite truth about them. We may, I think, take Plato to be using the standards of mathematics to judge perceptual practice. Since it is incapable of the achievements of mathematics, we cannot take its output as serious candidates for knowledge. In like fashion inductive reasoning has often been denigrated because it cannot meet the requirements of deductive reasoning by displaying a necessary connection between premises and conclusion. I would ask those who reject theistic practice for the reasons we have been considering, why we should take their strictures any more seriously than we do those of Plato against perception. In both cases the critic condemns a doxastic practice for not meeting expectations that are appropriate to a different practice. Of course, in both cases the critic is free to choose not to play the game unless it gives him what his favorite game provides. But why should the rest of us take these preferences seriously?

Here is another way of seeing the essential arbitrariness of my opponent's view. Sense perception, though certainly a useful guide to the physical environment, does not rank as high as it might conceivably rank, even on its own chosen dimensions. There are several respects in which sense perception is less than perfect. First, its deliverances can't all be accurate since there are contradictions between them. Second, though its deliverances have provided a basis for the discovery of regularities that serve as a basis for prediction and control, that achievement has been slow in coming, given the whole stretch of human life on earth, and even now it is spotty. We can certainly envisage a cognitive faculty that would reveal the regularities, along with the particular exemplifications, much more obviously and unerringly. Third, if we are to trust contemporary science, which itself is ultimately based on the deliverances of sense perception, our perceptual beliefs, though useful as a guide to action, rather badly misrepresent the intrinsic character of their objects. Most perceivers take physical things to be intrinsically qualified by the colors they display to our visual awareness: we perceive rocks, tables and leaves as continuously filled in, whereas actually they are mostly empty space with a few particles floating around; and so on. In all these respects perceptual practice is less ideal than it might conceivably be as a source of information about the physical environment. This being the case, why should we take its actual achievement level to be the norm for experiential sources of belief? Why shouldn't we take perceptual practice itself to be irrational for failing to come up to some higher standard? Isn't it just arbitrary to fasten on the level actually achieved by perceptual practice, rather than some higher or lower level? What warrant do we have for supposing that this is the level marked out by the nature of things as the one required for rational acceptance? Once we confront this issue squarely, we will realize that we proceed in this fashion only because we are so thoroughly immersed in perceptual practice that we are irresistibly led to take it as our measuring stick for other experiential sources of belief. But on reflection we can see that we have no rational warrant for doing so. This is just one more form of epistemic imperialism, judging one doxastic practice from the standpoint of another.

Much more needs to be said about these issues. But I hope that I have said enough to show at least that there is an arguable case for regarding what I have called theistic practice as a rational mode of belief formation, and hence that there is an experiential support for the rationality of theistic belief.