Wittgenstein and Religion:

Philosophy Does Not Leave Everything As It Is



                        In this paper I am trying to make two related points, one about religion and one about Wittgenstein's philosophy.

                        I want to advance a view of religion which is subtle enough to reconcile the apparently incompatible views of Wittgenstein's opponents and followers. The idea is to synthesize Wittgensteinian insights into the non-representational or non-literal aspects of religious language with appreciation for the widespread emphasis on its literal aspects (especially as regards the question of religious belief). The literal is emphasized by both religious fundamentalists and positivistic enemies of religion, both of whom must oppose Wittgenstein; whereas followers of Wittgenstein often leave no room for literal religious beliefs. My middle way will be to portray religion as a family of phenomena arranged in certain typical sequences and phases – a dialectical and narrative structure – and I will characterize the typical progression of these phases as moving from more literal or representational uses of language toward thought and speech which has more to do with gaining perspective and arranging priorities.

                        But doing this will require defending what may be called an interventionist account of Wittgenstein's approach to philosophy and religion, against the view that Wittgenstein's philosophy is "merely descriptive"  – a view whose slogan is: "philosophy leaves everything as it is". I will argue that the intrinsic dynamism and ambivalence of religious phenomena entails that leaving religion "as it is" is precisely what an adequate discourse cannot do, since to discern and rank the different senses of religious locutions is to play the religious language game(s). And I will suggest something similar about the language of metaphysics, and Wittgenstein's relation to it.

                        Let me begin by briefly setting out my view of the role of belief in religion, in contrast to the views of most of Wittgenstein's interpreters (pro and con). Their debate, as I see it, boils down to this: can we take the belief out of religion? Without beliefs about supernatural or metaphysical entities, does religion reduce to a matter of morality and ritual customs? If not, if both the customs and statements of belief are associated with a distinctive point of view or sphere of meaning which goes beyond morality, what is the status of that meaning? Is it merely a matter of attitude or style, or does it have some cognitive content which is accessible (or even of special importance) to philosophy?

                        Consider the following propositions:


(1) There is a cognitive dimension to religion; if the religious person is right, one can come to an increased understanding of one's world through religious practices and reflection on what religion says.


(2) Beliefs about supernatural and metaphysical entities often play a role in the development of religious understanding; religious belief-statements are not always metaphorical, nor do they serve only as expressions of attitude.


(3) The cognitive content of religious experience is specified by, and depends on the truth of, what is believed about transcendent entities.


                        Most critics of Wittgenstein, whether they judge religious beliefs to be true or false, agree that those truth-values are what is important, since they feel that truth-claims are what gives religion its authority. Thus they would accept all three propositions.

                         Most followers of Wittgenstein would accept only the first proposition, and would do so only to the extent that it supports the autonomy of religious language games (not because they want to talk about what it is that religious persons might know). They would explicitly deny the third proposition, for they treat references to transcendent entities as mere "pictures", and they hold that any justificatory procedures suggested by the surface grammar of belief (e.g. proofs of God's existence) are irrelevant to the point of religious belief-statements. And this has led them to deny propositions like (2) as well, for it seems that (3) might follow from (1) and (2).

                        I suggest that this is not the case; for the involvement with a religious belief can increase one's understanding even (or especially) when one's interpretation of its truth-conditions changes (when one does not continue to be interested in something that would make it true), or even when one comes to regard it as false. If we look at religious understanding as a process, we can acknowledge the role of different kinds of language in different phases of that process. The two sides in the debate focus on different characteristic phases, and thus attach different senses to the same terms and formulae.

                        Wittgensteinians have been so concerned with pointing out that belief-statements can be evaluated in a way other than through proof and evidence, that they have developed a dichotomy between a religious sense of 'belief' (not subject to justification) and the normal sense of 'belief', which when applied to religious terms they call superstition. (Sometimes 'faith' is reserved for the first sense, but it seems impossible to define 'faith' without using 'belief'.) We can agree that while most uses of 'belief' (and 'faith') can be said to invite requests for reasons and evidence, there is what might be called a non-justificatory kind of case, in which a person can say "I believe P" while not being willing to count anything as grounds for P (and perhaps holding that his belief would be threatened by the very temptation to consider such grounds). But instead of just two cases with clearly distinguished meanings, I see a series of gradations and interrelations, all impinging upon religious experience as its understanding progresses and fluctuates. The utterance of a belief-sentence can be an event in a transitional episode; it can signify doubt or irony as well as simple conviction. And the same locutions can be used differently even by one person at different times.

                        Wittgensteinians are right, in my view, to deny the primacy of metaphysical conceptions in religion; and I think that their focus on religious language as it is used by those who can be said to be relatively free of superstition does bring out what is best in religion – but at the price of over-simplification. As a result their opponents do not take them to be seriously holding that there is any religious understanding at all (for that, it is assumed, would involve the normal kind of belief), only a "religious attitude". This attitude would be something which one simply has or doesn't have, but for which one does not give reasons. The problem I see here is that giving reasons is understood too narrowly and univocally on both sides, and that what one calls "attitude" is thereby excluded from the domain of understanding. In my view religious understanding pertains to the process of developing certain attitudes and rationally relating them to all aspects of one's world. Specifically religious beliefs may play a part in that process, but they do not articulate what one understands thereby.

                        In urging that we pay attention to the evolutionary character of religious understanding, I am urging that religious language should be viewed as belonging implicitly to a kind of narration, whose component assertions, descriptions, predictions and prescriptions are not to be evaluated apart from their narrative functions in a temporally extended context. This narrative model helps us see how beliefs can play a role in religion without dominating its discourse, and how inattention to the temporality of religion has led to confusion in the debate between Wittgensteinians and their opponents.

                        The latter tend to emphasize what John Cook[1] calls the Traditional View of religious belief, which says that what makes language and rituals religious is that they are connected with beliefs about transcendent beings. He opposes this to the "Wittgensteinian" view (of Winch, Phillips and others), which, in order to secure the autonomy of the religious language game, says that what makes expressions religious is that they "originate" in "primitive religious reactions" (behavior and feelings which are logically and temporally prior to metaphysical representations and statements of belief).

                        In contrast, I suggest that what makes both beliefs and attitudinal "primitives" – and ritual behavior and contemplative experiences and moral precepts and much else – religious is that they are given a place in a special kind of authoritative and ostensibly very comprehensive narrative structure. There are no absolutely essential religious attitudes (although things like awe or reverence are typical) nor absolutely essential beliefs; but being embedded in a religious context gives beliefs and attitudes a special character. What is important, in my view, is to understand the kind of change a phenomenon such as reverence (or joy, or charity, or belief in spirits or an afterlife) undergoes when it is integrated into a religious narrative. Religious discourse purports to be the most comprehensive cognitive framework a person or community has available at a given time: this sense of comprehensiveness can be interpreted rhetorically (as the claim to have left nothing out of the story, which competing versions might include), or politically, as the property of actually helping to tie together (religare) the diverse elements of the life of a community or individual. In both ways coherence stands as a criterion. Religion can make use of beliefs in its story, without (or in spite of) attempts to analyze and justify them; but at the same time narrative coherence requires that its beliefs should not clash with any important strands in the fabric of a culture – including science. To the extent that they do, religion needs to be reformed to be preserved (i.e. to maintain its claims of comprehensiveness and authority).

                        Now you may agree that religious discourse can fulfill these narrative and political functions, without giving any credence to the religious claims to be providing understanding. If beliefs are not of primary importance to religious language, then its function would not seem to be primarily cognitive. How can we speak of a kind of understanding which involves beliefs, but is not limited to what is believed? We might make use of Nelson Goodman's terminology: the world (or worldview) which is presented in a religion might be measured not by the yardstick of truth alone (although it will include some true beliefs – about human character, for example  –in addition to the presumably false supernatural ones), but by the aggregate of features Goodman calls rightness. Rightness


is primarily a matter of fit: fit to what is referred to in one way or another, or to other renderings, or to modes and manners of organization... And knowing or understanding is seen [by Goodman] as ranging beyond the acquiring of true beliefs to the discovering and devising of fit of all sorts.[2]

...knowing cannot be exclusively or even primarily a matter of determining what is true.... Much of knowing aims at something other than true, or any, belief.[3]


Goodman's attention to the dynamics of understanding is due to his probing of the interdependence of creation and discovery in science, in perception, and in art, which is subject to its own determinations of rightness. The rightness of art does not depend primarily on being informative, but rather on its power to exemplify and provoke, to invite (or exhort) useful reflection and exploration – to change our way of seeing, in Wittgenstein's phrase.


...such works induce reorganization of our accustomed world in accordance with these features, thus dividing and combining erstwhile relevant kinds, adding and subtracting, effecting new discriminations and integrations, reordering priorities.[4]


I want to say the same thing about religion: like art, and like philosophy, it changes our way of seeing.

                        And what is more, I think the specific perspective-changing functions of philosophy (as exemplified by Wittgenstein) and religion (in its progression away from literalism) are intrinsically connected. Unlike all those interpreters who take Wittgenstein's approach to religion to be "neutral" or "merely descriptive" at bottom, I think he wanted to change the role of belief in religion, so that statements about the supernatural might no longer be taken so literally by so many people, so much of the time. A writer like Cook cannot see this because he, like his opponents, regards religion as uniform and static – thus as subject only to either pure analysis or what he calls normative definition. He says that "one who lays down such a [normative] definition is merely inventing something, and Wittgensteinians do not think of themselves as doing that."[5] But this is to rely on a false dichotomy, between describing and inventing; whereas to reform an institution or practice is at once to describe it, to distinguish features which will be considered essential (from the new projected standpoint) from those which are to be discarded, and to present ("invent") a plausible story of the new form of life.

                        The conservative interpretation of Wittgenstein as simply providing a uniform analysis of what religious believers mean in general (so that most of them turn out never to have had any literal beliefs at all) would be paralleled by an interpretation of the Philosophical Investigations as claiming that most philosophers had not really been doing metaphysics, only saying things that sounded like it. (They were really just expressing "primitive philosophical reactions.") But clearly Wittgenstein wanted to change much of what philosophers do, and I claim that the same goes for religion. Wittgenstein does, after all, occasionally use 'philosophy' as a term for what he is doing, thus in a "normative" way; just as he occasionally talks about religion in ways that only apply to those of whom he approves. Of course Wittgenstein thought that much of religion – you could even say the best part, or highest phase – is misunderstood when it is taken literally; but that is not the same as denying that literal beliefs have played a role as well.

                        To deny that religious beliefs are often taken literally can be done in two ways, neither of which is very plausible. One way is to disqualify (as being merely superstitious) the great majority of those who call themselves religious. Philosophers unsympathetic to Wittgenstein often say, just look at Mormons, Southern Baptists or Iranian Shiites and you will see how implausible Wittgenstein's view is. It leaves no place for those who literally expect to travel to other solar systems, or who think golden tablets have fallen from the sky; but surely religion encompasses their ideas and activities as well as those of Wittgensteinian heroes like Kierkegaard and Tolstoy. The language and symbolism of religion have been conditioned by the participation of the literal believer in their historical development; the accomodation of erroneous views is part of religious tradition – both explicitly, through the doctrine of levels of meaning and understanding, and implicitly, through institutional accomodation of superstitious elements in ritual. Often (as the most casual glance at the Old Testament shows) these accomodations go too far and reform is required; reform is an essential part of tradition. To declare that the truly religious are perfectly free of superstition is to deny this role of reform, to deny that religion is a process.

                        A second strategy for eliminating all reference to literal beliefs is to posit a great separation between what people say about their beliefs and what the belief-sentences are supposed to "mean" by way of expressing rather than stating. Michael Coughlan[6] defends the view that Wittgenstein's account is "neutral", not "stipulative" with regard to religion, in spite of the fact that it admittedly does not agree with believers' own accounts. It must only agree, he says, with their "performative utterances", not their "conceptual speculations". He even tries to exempt theology from criticism by claiming that it only mentions but does not use metaphysical terms. I think this is clearly false, since many theologians have explicitly claimed to be using the terms they analyze – something which would seem to follow from the claim that faith precedes and enables theology. But in any case I do not accept the boundary between performatives and "speculative" statements of belief which he relies on here. What people say about their beliefs – and the way such discussions bear on situations involving conflict and doubt – is part of their "performance", which does not merely express an attitude or perspective, but also works to achieve, maintain or modify it.

                        In both ways of denying the role of the literal, then, something crucial is lost: the dynamic character of religious belief, its fluctuations with regard to both intensity of affirmation and the degree of literalness with which it is taken. My claim is that this dynamic dimension is also a cognitive one  – that literalism or fundamentalism is not of the essence of religion, but is nonetheless a counter-tendency which religion must perpetually battle within itself. The battle is itself a learning process. And this means that Wittgenstein's intervention in religious language (and our interpretation of it) is itself fundamentally religious.

                        And the same can be said of Wittgenstein's concept of philosophy: the battle against "bewitchment by language" in epistemology and metaphysics shares many of the same strategies, and aims at many of the same underlying motivations, as the battle against the literal interpretation and use of religious language. In both cases what we learn is not whether the beliefs which were initially at issue are true or false, but how to better understand our own investments in the language and practices surrounding them. This is not simply a matter of answering questions or solving problems. Wittgenstein's effort to change our way of seeing aims at changing the way we understand the questions with which both religion and philosophy are concerned. (In his early notebooks Wittgenstein gave this creative definition: "To believe in a God means to understand the question about the meaning of life." He does not say it means answering such a question.)

                        Stanley Cavell has shown[7] the crudity of interpretations of Wittgenstein's writing which portray it as solving (or attempting to solve) the problem of skepticism, as opposed to taking skepticism as a perpetual temptation which the philosopher too must battle in himself. And the temptation to philosophical doubt is bound up with another temptation, to dogmatism and representational thinking. Wittgenstein's discussions of belief – whether in God, other minds, or material objects – also confront a tendency (call it fundamentalism or literal-mindedness) of which none of us is totally free. When he wrote that metaphysics is a kind of magic and said "I must neither speak in favor of magic nor ridicule it"[8], Wittgenstein was not wanting to be "neutral", but to acknowledge how the symbolic forces of magic are still part of us – and the extent to which we still do not know our way around in their domain. "Yes," he said, "the elimination of magic here has the character of magic itself."[9]

                        Consider the way in which Wittgenstein sets himself against the notion of belief which is supposed to be produced, as a response to skepticism, by the "argument from analogy" (for "other minds"): "My attitude towards him is an attitude towards a soul. I am not of the opinion that he has a soul."[10] Here, as in the case of religion, there is a tendency to exaggerate differences and make it look as if a problem had simply been solved (or, if you prefer, ignored). By virtue of sharing a form of life, I can see emotions and attitudes on someone's face (even though I am not always right about what I think I see); what lets me recognize the meaning of the face (or for that matter the face as such, as opposed to the skull's fleshy covering) is prior to any inference I could make about there actually being a person with feelings "behind" it.

                        But from this it does not follow that all questions of belief and analogical reasoning have been eliminated. It remains to consider how "putting myself in the other's place" does function in the perception of persons, and in the sometimes "magical" reactions we have (Wittgenstein:"When someone in ... my company laughs too much, I half-involuntarily press my lips together, as if in that way I believed I could shut his"[11]). It remains for us to consider how our knowledge of each other is conditioned by both trust and (sometimes quite extreme) doubt. Cavell, starting from Wittgenstein, has shown how much the impediment to knowledge of others lies not in them (in the way they are hidden "inside" their bodies), but in us – in our inability or unwillingness to take on the responsibilities and other consequences of adequately situating ourselves to know and be known by each other. Sometimes the temptation to belief is precisely that it affords us a way of evading knowledge. Countless religious writings make the same point.

                        Philosophy and religion are both concerned with organizing our attitudes and beliefs, and both constitute arenas in which conviction and commitment, doubt and denial, are constantly in play as the stories of communities and individuals unfold. I have suggested that we think of religious discourse as a kind of narrative because religion is a phenomenon which is kept alive through participation and creative modification – that is, through retelling. The faithful retelling of religious narratives in our culture should take care to minimize the justificatory and speculative aspects of belief. This caution is not radically different from that which Judaism exercised against idolatry, which Buddha exercised against metaphysics, or which Jesus exercised against the literal interpretation of the Torah. We cannot hope to be entirely successful in overcoming fundamentalism or foundationalism but we should not shirk the task. Wittgenstein's writings always exhibit this attitude of hard labor, never the smug satisfaction of having the "right answer".


                        "We must find the road from error to truth."[12]

[1] John Cook, "Wittgenstein and Religious Belief", Philosophy Oct. 1988

[2] Nelson Goodman, Ways of Worldmaking (Hackett: Indianapolis, 1978) p. 138

[3] ibid. p. 19

[4] ibid. p.105

[5] Cook, op.cit. p.442

[6] Michael Coughlan, "Wittgensteinian Philosophy and Religious Belief" in Metaphilosophy Oct. 1986

[7] Stanley Cavell, The Claim of Reason (Oxford Univ. Press: Oxford, 1979)

[8] Ludwig Wittgenstein, "Remarks on Frazer's The Golden Bough" (hereafter RFGB), ed. Rush Rhees, trans. A.C. Miles, (Brynmill/Humanities: 1979) p.1.

[9] RFGB p. 1

[10] Ludwig Wittgenstein, Philosophical Investigations (Basil Blackwell: Oxford 1953) p. 178.

[11] RFGB p.9

[12] RFGB p.1