302 - Saturday's Child: A New Approach to the Philosophy of Religion

Saturday's Child: A New Approach to the Philosophy of Religion
By James Win. McClendon, Jr. and James M, Smith

"While philosophy of religion in the university was absorbed with something very like natural theology (though often using the newer techniques), in the seminaries it was obliged to struggle for respectability. Meanwhile, both religious studies and philosophy in other areas had taken exciting new turns. Could there emerge a new understanding of philosophy of religion which might restore its role as the critic and illuminator of religion, and as an aid in man's long quest for self-understanding?"

THE philosophy of religion has been, in the view of the generation just passed, an unwanted lovechild. Born of a rude encounter between traditional western Christianity and the Enlightenment of the eighteenth century,¹ bred in the long liberal twilight of the nineteenth and early twentieth centuries, philosophy of religion was hardly the favorite discipline of the from-Brunner-to-Barth theological seminaries of the quarter-century just ended (1940-1965). Meanwhile, in university philosophy departments, the

James Win. McClendon, Jr., educated at the University of Texas, Southwestern Baptist Seminary, Princeton Theological Seminary, and the University of California, taught at Golden Gate Baptist Seminary and has been Visiting Professor at the University of San Francisco, Stanford University, and Temple University. The Jeffery Lecturer in Religion, Goucher College (1970-71), he is the author of Pacemakers of Christian Thought (1962), a contributor to American Philosophy and the Future (1968, ed. Michael Novak), and to various journals.
James M, Smith, educated at the University of Southern California, the University of California at Los Angeles, and Brown University, has taught at the University of Illinois, Princeton University, and the University of Washington. Presently he is Professor and Chairman, Department of Philosophy, Fresno State College, California. A contributor to the forthcoming Ethics, Analysis, and Analytical Philosophy and to literary and political journals, he has just edited Mill's Utilitarianism (1969).

¹ Of course there were philosophical approaches to religion long before the eighteenth century-and in connection with other religions than Christianity. See the elegantly concise history of the more important ones in H. D. Lewis, "Philosophy of Religion, History of," in Paul Edwards, ed., The Encyclopedia of Philosophy, New York, Macmillan and Free Press, 1967. Still there is a point to the claim made above that the discipline as we have received it owes its shape to Hume, Kant, Hegel, Schleiermacher; that is, to the men of the Enlightenment and their successors.

303 - Saturday's Child: A New Approach to the Philosophy of Religion

courses in philosophy of religion were retained often as a hedge against upstart religion departments, and given to someone who could not teach anything more important.

Briefly stated, the trouble was that philosophy of religion in the university had sought to treat religion as a set of doctrines, doctrines which could be examined one by one and given an academic grade: hell was invariably flunked for moral turpitude; God was usually given an incomplete or a withdrawn-failing, depending on the grading system; immortality seemed sometimes to pass, though with difficulty; incarnation rarely took the course for credit. In the seminary, on the other hand, "philosophy of religion" was a clutter of topics not treated elsewhere in the theological curriculum, ranging from comparative religion to church-state relations.

While philosophy of religion thus seemed to slumber over its shovel, other disciplines were breaking exciting new ground. The academic study of religion (still a discipline in search of a name: "Religionswissenschaft," "history of religions," even "religiology" were tried), created excitement by means of techniques which paid meticulous attention to the structure and practice of primitive religions, thus providing, it was believed, clues for religion-in-general. This work seemed to combine impeccable academic objectivity with broad empathy for religion, especially non-Christian, non-Jewish, non-Western religion. In contrast to the old philosophy of religion, the history of religions saw religions as living wholes. Far more than doctrines, far more than competing "systems," religion was humanity itself seen in its intimacy, its variety, its passion, its life.

Meanwhile, contemporary philosophy had passed through a revolution of its own. Earlier in the century, philosophy had found in language a crucial clue to self-understanding in its own work. The positivist phase of philosophical analysis was succeeded by a more catholic phase, in which a variety of techniques were used to bring philosophy's solvents into contact with profound human problems in science, law, ethics, and history. Titles such as "Action and Responsibility," "The Free Will Defense," and "Aesthetic Problems of Modern Philosophy,"² suggest the broadening concerns of recent American philosophers, while in Europe the convergence of existentialism with phenomenology had a similar broadening effect.

² Titles of essays by Joel Feinberg, Alvin Plantinga, and Stanley Cavell in Max Black, ed., Philosophy in America, Ithaca, Cornell, 1965.

304 - Saturday's Child: A New Approach to the Philosophy of Religion

While philosophy of religion in the university was absorbed with something very like natural theology (though often using the newer techniques), in the seminaries it was obliged to struggle for respectability. Meanwhile, both religious studies and philosophy in other areas had taken exciting new turns. Could there emerge a new understanding of philosophy of religion which might restore its role as the critic and illuminator of religion, and as an aid in man's long quest for self-understanding?

I

To find an optimistic answer to this question, we must turn to post-positivist thinkers; that is, thinkers who have faced the challenge posed by the verifiability criterion of meaning. Although that criterion has not survived philosophical scrutiny, the difficulty about the meaning of many ordinary religious utterances remains, and must be faced. Consider two post-positivist philosophers who, while differing greatly from one another, each provide elements of a more promising way of philosophizing about religion. One is the Dutch-American Willem F. Zuurdeeg; the other is the Englishman R. B. Braithwaite.

In 1958, Zuurdeeg published An Analytical Philosophy of Religion,³ which, in spite of its title, had apparently little in common with contemporary Anglo-American philosophical analysis. In it, he focused on convictions, which he defined as "all persuasions concerning good and bad; concerning gods and devils; concerning representations of the ideal man, the ideal state, the ideal society; concerning the meaning of history, of nature, and of the All" (p. 26).

Several things of interest appear in his account. In the first place, the notion of convictions cuts across the lines separating religious, ethical, and a variety of other sorts of convictions. Religion approached in this way is continuous with other sorts of human convictional enterprise. This feature of the approach might commend it in a period when religious studies are attending to non-Christian, non-Western religious expressions in a way that analysis tailored particularly to Christian faith could not serve. It also commends it at a time when our ideas of what religion is (what "religion" means) are in a state of evolution.

Second, Zuurdeeg's notion of convictions was attuned to the power-

³ New York and Nashville, Abingdon, 1958.

305 - Saturday's Child: A New Approach to the Philosophy of Religion

ful role of convictions in human existence: the way in which men not only form, but are formed by, the convictions which they embrace. Zuurdeeg chose "conviction" to express the concept which he was examining because of its etymology: Its root is the Latin convinco, "to conquer," not in the sense of military conquest, but of legal or argumentative triumph. A man "convinced" or "convicted" in this sense is a man conquered, overcome, by testimony which overwhelms him. Thus, a man's convictions go far to make him what he is; "we are," said Zuurdeeg, "our convictions" (p. 58).

Third, this notion comports well with the role of particular human communities (churches, political parties, communes, etc.) in the formation of religious or other convictional outlooks. The convictional community might be defined as those sharing a particular group of convictions, and interacting in their formation and support. Again, Zuurdeeg seemed to imply that communities make convictions, but also that convictions make communities what they are (pp. 32-35). Here, perhaps, is an opening not explored by Zuurdeeg for intercourse between philosophy of religion thus conceived and the sociology of knowledge. It is especially helpful to recognize that the convictions which form a particular person or group might stand in hierarchical relation to one another: A man may be "overcome" by Christian faith conceived thus and thus, but even more "overcome" by patriotism conceived in such and such a way (pp. 27-40), and in the interaction of these might lie interesting analytic possibilities.

Finally, Zuurdeeg, as a post-positivist, was sensitive to the difficulties which lay in the way of either asserting or denying that convictions are "cognitive" in the way in which scientific beliefs are. This dilemma had been vividly expressed by the logical positivists, using the verification criterion of meaningfulness. Zuurdeeg's ultimately unsatisfactory way out of the dilemma was to distinguish "indicative language" (of which the language of science was called a model) from "convictional language" (the language in which men expressed themselves by expressing their convictions).

Had Zuurdeeg taken the line that the only task of "convictional language" was self-expression, he would have had a view not unlike that of Bultmann;⁴ however, with more fidelity to the actual content

⁴ See for example Rudolf Bultmann, "What Does It Means to Speak of God," Faith and Understanding, Vol. I., ed. R. W. Funk, tr. L. P. Smith, New York and Evanston, Harper & Row, 1966.

306 - Saturday's Child: A New Approach to the Philosophy of Religion

of religious utterance than consistency with his own distinction, Zuurdeeg held that when a religious speaker, confessing his conviction, says something like "God led Israel across the Red Sea," he is not only indirectly letting us know about himself, but also purporting directly to tell us something about God, Israel, and the Red Sea crossing.⁵ Only, said Zuurdeeg, this "recital" of biblical history cannot be checked; it is just "repetition"; it is, in a sense of that term which regrettably Zuurdeeg never clarifies, "mythic" (ch. 5). Within the conviction which expresses it, it is simply impregnable; it cannot be impugned, checked out, or discarded as false whatever the evidence might be. We have no objective standpoint from which to judge convictions.

Before we relate Zuurdeeg's work to the present task of philosophy of religion, it is important to note its limitations. It is not at all clear exactly what Zuurdeeg means by "conviction." His definition is a topical one, yet it would be far-fetched to claim that everything of which one is persuaded on certain topics (devils, for example) has convictional status in his life, or that other things omitted by Zuurdeeg (for example, war and peace) do not have such status. Even if the definition were tightened, however, other serious difficulties remain. For example, it is far from clear what Zuurdeeg means, or could mean, by saying that we are our convictions. This cannot be an identity-claim (even if I am blond and am married to a red-head, my convictions are neither blond nor married to a red-head); is it then attributival? or causal? or metaphorical? The truth is, we cannot tell what this claim means in Zuurdeeg's use.

The principal difficulty in Zuurdeeg's account, however, stems from his confusing three different claims: (1) that we are unable to justify our convictions in any way whatever; (2) that we are unable to judge convictional utterances from a standpoint of detached objectivity; and (3) that we are unable to distinguish "indicative elements" in religious utterance. Yet these are distinct, and have quite different claims to our assent. As to the third, we can accept it only if, like Zuurdeeg, we stipulate it. For in actual practice very many convictional utterances as employed by ordinary speakers do have an "indicative" or, as we should say, representative element. To the extent that Zuurdeeg ignores this element, his account of convictions is faulty.

⁵ Zuurdeeg, op. cit., pp. WM, The illustration is ours.

307 - Saturday's Child: A New Approach to the Philosophy of Religion

Concerning the second, we again point out that there is a sense in which in Zuurdeeg's use it is true merely by definition. For if being detached and objective requires one to be convictionless, and if everyone has convictions, it follows that we cannot be detached and objective. Nevertheless, we do not think this claim is trivial. It embodies an important insight which is obscured by Zuurdeeg's broad imprecision about how convictions are related to other beliefs, and about how language expressing convictions is to be understood and related to what he calls indicative language.

Finally, the first claim-that we are unable to justify our convictions in any way-seems different from the other two. It seems not so much trivial, or obscure, as simply unsupported. Of course, if justification were exhausted by scientific justification, or if there were no claims to truth made by convictional utterances, then it would take on a certain plausibility. If, however, those are false suppositions, matters are different. In any case, Zuurdeeg has given us no reason for thinking them true.

If the program of analyzing (religious) convictions is to fulfill its promise, then several further conditions must be met. The concept of "convictions" must be defined with sufficient exactness to enable us to say when a particular belief is indeed a conviction. For example, we might define X's convictions as those of X's beliefs⁶ which he holds persistently (that is, in the face of argument and adversity), and which he cannot change without being significantly changed. Such a definition would demand further refinement (How much "adversity" must X withstand? Which changes in him will count as "significant" ones?), and this refinement would be a part of the new task of philosophy of religion.⁷

Further, the program will require discovering how in general convictions relate to other convictions of their holder, to his other non-convictional beliefs, and to whatever else there is. Means must be discovered for detecting the convictions of an individual or community: It should not be assumed that a particular man knows his

⁶ By defining convictions as a kind of belief, we should not be understood as limiting convictions to merely "cognitive" elements; as will presently appear, beliefs have as rich a range of dimensions (affective, volitional, intentional, etc.) as any other mental states or acts.
⁷ It might occur to someone that in the phrase "religious convictions," the adjective stands as much in need of clarification as the noun. However, "religious" (like "religion") has not shown itself prone to genus/species or other sorts of definition. A part of the promise of the proposal presented by the present article is that the analysis of convictions may proceed before we are in position to define X's "religious" convictions as opposed to his "Political" or "philosophical" or "moral" ones.

308 - Saturday's Child: A New Approach to the Philosophy of Religion

own convictions; he may be mistaken, confused, or both. The same point will apply to the convictions of a community.

II

Rather than develop these possibilities here, let us return to the question of how a conviction is related to other beliefs and to the life of the holder, since this was the point at which Zuurdeeg's analysis (not to mention other examples) failed. How can philosophy of religion overcome the pervasive vagueness which we noted here? At this juncture we turn, not altogether without method, to consider the work of the other post-positivist, R. B. Braithwaite. Since Braithwaite's lecture, An Empiricist's View of the Nature of Religious Belief,⁸ published in 1955, has been widely reported, criticized, and reprinted, we will note only some of its several features. Braithwaite, for example, seeks to give an account of his own highly sophisticated Christian faith, which is, perhaps, not that of the ordinary Christian; but we do not propose to discuss here whether Braithwaite's rather special theological views are, on other grounds, the best or the worst theology going. We will be fully occupied in examining his view of the nature of religious belief as it is determined by his account of what religious utterances necessarily mean, that is, his analysis of religious language.

For the verification principle, Braithwaite substituted the somewhat broader "principle of use," namely, "the meaning of any statement is given by the way in which it is used," which he takes to be Wittgenstein's view of meaning (p. 58). Braithwaite subsumed religious utterances under moral utterances, as being a large and evidently meaningful class of utterances whose use was not to describe anything at all, but to perform a quite different function. And he took that function to be the expression of generalizable intentions, so that he characterized his own theory as "conative" as opposed to the traditional positivist's "emotive" theory of moral language. On this view, to say, "Rioting in the streets is wrong," is nothing other than to declare that, if the situation makes it possible for one to riot in the streets, he intends not to do so. Not all statements of intention, of course, are considered to be moral utterances; what makes such a statement moral is exactly its subsumability under a general

⁸ Cambridge University Press, 1955; reprinted in Ian T. Ramsey, ed., Christian Ethics and Contemporary Philosophy, New York, Macmillan, 1966. Our page references are to the Ramsey edition.

309 - Saturday's Child: A New Approach to the Philosophy of Religion

principle: Thus "I intend to go to the flicks tonight" is not moral unless it is an expression of some principle such as "I intend to go to all the flicks I can," or more conventionally, "Going frequently to the flicks is a good thing to do" (p. 60).

Religious language, further, is a special case of moral language, distinguished from other cases by the association of religious language with "stories." Thus a typical cluster of religious utterances (say the Christian cluster) will consist of expressions of intention (for example, the intention to live a life of agape) associated with stories (for example, stories of Jesus, the apostles, saints, martyrs, and perhaps other sorts of stories). It is particularly important that on Braithwaite's view, while the stories must be capable of being true or false, factual or fictional, it does not, in fact, matter which they are, as far as the force of the main utterances, the declarations of intention, are concerned. This is because Braithwaite maintains that the connection between the stories and the intentions in a given religious system is only "psychological and causal": The stories reinforce the intentions of the one who associates his intentions with them (pp. 66-71).

Braithwaite's theory of how religious language works is a brilliant piece of work, as remarkable for its brevity as for its clarity. The latter inevitably suffers from the former, however, and we must give attention to some matters which Braithwaite has not made clear. Braithwaite gives us only one example of a religious "assertion," namely "God is love," which he says expresses a general principle, the intention to live an "agapeistic" life (p. 63). But Braithwaite does not make it clear whether "God is love" is merely the expression of an intention, or whether it is also the telling of a very brief "story" about God. Unfortunately, there are severe difficulties with either interpretation.

If, as seems more likely to us, Braithwaite intends "God is love" to be taken as a pure expression of intention, one whose exhaustive translation would be, "I intend to live agapeistically," then one sort of difficulty arises. This is that there is no imaginable reason for retaining the sentence "God is love" to express that intention; nor is there any reason to analyze "God is love" as an expression of my intentions. Compare a lawyer's office whose door sign says simply, "Flats fixed"; there seems to be no reason for the lawyer to put out that sign, and no reason, based on the sign, to suppose a lawyer with-

310 - Saturday's Child: A New Approach to the Philosophy of Religion

in. Granting, for argument, that the positivists were quite right, and there is no way to say anything about God, nothing either factual or fictional, true or false, then it is not in the least clear why "God is love," that sentence, expresses or states anyone's intentions.

This may perhaps be seen more clearly by trying to extend the analysis, taking as fresh cases "God is wrathful ," "God is lord of all," "God is omnipotent." Shall we analyze these as: "I intend to be wrathful," "lord of all," "omnipotent"? That does seem counter-Christian; the traditional Christian view might explain that God's wrath makes appropriate my repentance; God's lordship, my obedience; God's omnipotence, my dependence. But those explanations are on this alternative reading of Braithwaite foreclosed.⁹ While there are certainly linguistic problems in references to a transcendent God, there seem to be even more such problems for this type of utterance, if all such reference is on principle excluded.

We can best illustrate this difficulty by imaging an agreed-upon code. In the code whenever a mate wants to express a desire for sex, he will say something about hunger or desire for food. So far so good. But suppose now he badly wants to express a desire for food itself? In this case, he can do it by saying, "Code's off. Let's eat-and I don't mean…. In Braithwaite's analysis, the difference is that there is no way to say "Code's off"-no way to talk about even a fictional God.

But if we say instead that "God is love" is not a pure expression of intention, holding that such utterances can also tell a "story" about God, we are faced both with Braithwaite's apparent rejection of this reading, and with a fresh difficulty. For now "God is love" must be taken to mean not one thing but two things-and two things between which there is no linguistic connection, but only (according to Braithwaite) a psychological and causal connection. There is, for many speakers, a psychological and causal connection between "sugar" and "spice," too, but that is not a linguistic connection; it cannot entitle us to believe that when a speaker says one, he means the other. Think of the confusion that mistake would produce in everyday situations!

⁹ We recognize that "God is wrathful" is not in Christian theology on all fours with "God is love." The love of God is viewed as primary in historic Christianity; wrath is understood as loving wrath. Yet even to give that account seems to presuppose reference to an at least imaginary or delusive God, references which on this alternative reading of Braithwaite are altogether excluded, so that the objection cannot even be made in these terms.

311 - Saturday's Child: A New Approach to the Philosophy of Religion

As a consequence, we have to recognize that Braithwaite's theory cannot adequately account for the very sophisticated religious talk of a believer such as Braithwaite himself; and a fortiori, it cannot account for the talk of more conventional religious speakers. The difficulty arose from the too-exclusive concentration upon one of the dimensions of religious discourse-the declaration of intention-to the neglect of other important dimensions. What we clearly require is an account which is able to include not only the element of representation which the positivists, albeit negatively, emphasized, but also the element of intention and other affective elements, which Braithwaite fastens upon. And perhaps there is something else, of more fundamental importance still, which must be included.

Now, however, we must explain why we introduced this example of the analysis of language into our consideration of a new approach to the philosophy of religion focused upon the role of convictions. The reason is that, whatever its shortcomings, the analysis of language has moved in the direction of that very exactness which we earlier complained was lacking in accounts of convictions such as Zuurdeeg's. What if it were possible to find or construct a theory of religious language which was marked by such exactness and which showed the involvement of a man's speech in his life situation, with its full variety of dimensions and relations, which we have earlier said must be the quality of an analysis of convictions? Then might not the linguistic analysis be directly relatable, indeed directly related, to the convictions of the speaker of this language? If so, still further gains might follow.

III

If this were primarily a historical report, we might introduce at this point a survey of recent contributions to analytical philosophy of religion, showing how a number of workers in the field have provided materials from various directions and labor toward the achievement we have in mind. However, such surveys are already available,¹⁰ and the point of the present paper is a different one. We must now recognize that in the minds of some there is a serious objection to the undertaking we propose. For if convictions are seen as a kind of belief, if talk is only talk, and if, as some philosophers

¹⁰ See, for example, Frederick Ferré, Language, Logic, and God, New York, Harper & Brothers, 1961; supplemented by James A. Martin, Jr., The New Dialogue between Philosophy and Theology, New York, Seabury Press, 1966.

312 - Saturday's Child: A New Approach to the Philosophy of Religion

believe, religious beliefs, attitudes, and responses (among which convictions might be included) are particularly private, inward phenomena, from which realm language is separated by a great uncrossable gulf, then our hopes have been in vain. We cannot here refute this claim itself, much less make evident the actual tight connection between belief and speech which an adequate theory of language would articulate. We can, however, indicate that there are some reasons for skepticism about the uncrossable gulf.

Notice that when beliefs, attitudes, and responses are called private, often nothing is intended beyond calling our attention to the fact that we can keep them secret; we can pretend, dissemble, keep a straight face, lie, or just refrain from speaking. Even here, however, there is need for caution. It is of some importance that the privacy of beliefs is a result of something the believer does. To pretend, even to lie, is a fairly sophisticated activity which we must learn to do. To be an accomplished liar is indeed an achievement. This suggests that it is natural for beliefs to be communicated and that privacy (in this sense) is the exception, not the rule.

Some hold the view that men's beliefs do have expression in their speech and other acts, but also hold that which belief is communicated by which act is a matter of the believer's own choice. Complete privacy of belief is maintained, on this view, by the speaker's arbitrary power of determining how he will express his belief. If, however, Jones, on stepping out of doors, opens his umbrella, and later tells us that he had not believed when he did so that it was still raining, then whether we take him at his later word or his earlier deed depends on a consideration of his acts (including his speech-act) in the surrounding circumstances. These circumstances include Jones's other beliefs, attitudes, desires, and intentions, which are discovered in the same ways as the one in question. Moreover, not just any utterance or other act can (in a given culture, and lacking a prearranged code) mean "It is raining." Wanting or intending to mean something by an expression is not sufficient to make it mean that in the absence of other "public" conditions. Finally, while all sorts of actions may evince our beliefs, there is a class of acts, namely, speaking, or speech-acts, whose point is to do just this."¹¹

Now if we consider utterances in which we describe, identity,

¹¹ The term "speech-act" (or "speech act") was employed or coined by John L. Austin in developing his theory of religious language. See following note.

313 - Saturday's Child: A New Approach to the Philosophy of Religion

grade, or estimate, it is apparent that we issue such utterances happily or appropriately only under certain conditions. We note, however, that these conditions correspond closely to those for happily believing. Beliefs are (hopefully) true and (regrettably) false; but the same could be said for statements, assertions, and some judgments-all of which are speech-acts. There is, however, another spectrum of assessments which even more than the true/false assessment shows the intimate connection of beliefs with speech-acts; namely, being well- (or ill-) founded. Not only beliefs, but assumptions, agreements (with), suspicions, and surmises are unhappy unless there is evidence in their favor. In strictly parallel fashion, statements, assertions, reports, confessions, and warnings commit their utterers to respond to requests for argument or evidence. If I rank (a man first), find (the accused guilty), classify (this bone as neolithic), place, rate, or grade (a performance of Hamlet), or diagnose (an illness), my speech-act will be unhappy in varying degrees if I cannot offer some considerations in its defense when challenged.

In general, wherever one finds a belief, there one should be able to specify a speech-act having conditions for happy utterance which are identical with the conditions for happily having or holding the belief. In setting out the conditions for happy utterance, we would be also setting out the conditions for (happily) believing, and this is no less true if the beliefs in question are religious convictions. Of course, we must reckon with secrecy and hypocrisy-no account of religion, or of human life generally, would be complete without either. Yet there is reason to believe that we will not be frustrated in the task by these sorts of exceptions.

There is, then, the possibility of a new approach to the philosophy of religion, theologically "neutral" with respect to convictions, guided by a consideration for the broad inter-connections between religion and the other modes of human life, informed by an understanding of the ways in which (via our convictions) we become the men we are, and (via our speech) we give remarkably revealing expression to that of which we are convinced. If anyone wishes to link the former of these two paths or techniques with "continental existential philosophy," and the latter with "Anglo-American analysis," we shall not object, although we offer cautions, as our quotation marks indicate, against being misled by those labels.

Is there indeed, or can there be, a form of analysis of religious

314 - Saturday's Child: A New Approach to the Philosophy of Religion

speaking which goes beyond the achievements of Braithwaite, and other analysts of the fifties and sixties, to provide an account of how religious talk works-a form which is adequate to its representative and to its affective dimensions, and to that "something more" which makes speech the unique kind of human activity which it is? While it would be mistake to isolate the work of a single pioneer in the field to the exclusion of many others who have made significant contributions, we are particularly impressed by the prospect of philosophy of language grounded in theoretical work done by John L. Austin.¹² It should be carefully noted, in this connection, that Austin (who himself took little notice of religious language) went beyond his earlier and more widely known "performative" theory by providing the beginnings of a speech-act theory of language. This theory seems to us to provide the groundwork of analytic technique needed to furnish the more exact understanding of convictions which is needed.

No headway is going to be made in philosophy of religion at this stage in its history by mere brilliance or mere insight or mere tour de force. Whether the possibilities we have indicated become actualities, whether, in general, the conjunction of methods suggested here makes possible a philosophy of religion adequate to the new understanding of religion itself (and to the new range of philosophy in the coming generation) is a question whose answer depends upon a great deal of hard work by many workmen-and perhaps as well upon a little bit of Sunday luck, or Providence!

¹² Austin's principal work is available in three volumes; Sense and Sensibilia, reconstructed by G. J. Warnock, Oxford, Clarendon, 1962; How To Do Things with Words, the Williams James Lectures for 1955, J. O. Urmson, ed., Oxford, Clarendon, 1962; and Philosophical Papers, J. O. Urmson and G. J. Warnock, eds., Oxford, Clarendon, 1961. See also James M. Smith and James Wm. McClendon, Jr., "Religious Language after J. L. Austin," forthcoming in Religious Studies.