Book Search:  


Google full text of our books:


Philosophy as a Humanistic Discipline
Bernard Williams
Selected, edited, and with an introduction by A. W. Moore

Book Description | Reviews | Table of Contents

COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2006, by Princeton University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers. Follow links for Class Use and other Permissions. For more information, send e-mail to

This file is also available in Adobe Acrobat PDF format

Chapter 1


Non pudet, quia pudendum est . . . prorsus credibile est, quia ineptum est . . . certum est, quia impossibile.
     --Tertullian, de carne Christi, v.

(1) This paper does not deal directly either with Tertullian or with his paradox. In considering the most famous and most widely misquoted of Tertullian's paradoxes, I do not try to explain it, still less to explain it away; but take it as the starting-point and end of a discussion of religious language and of its relations to theology and to the kind of philosophical inquiry with which this book* is principally concerned. In particular, I try to bring out a certain tension, a pull between the possible and the impossible, a sort of inherent and necessary incomprehensibility, which seems to be a feature of Christian belief, and to locate this point of tension more exactly within the structure of the belief. This tension Tertullian seems to have felt very strongly, and characteristically proclaimed it with vigour; but it is only by this rather thin string that my remarks are tied to what Tertullian said, the strict interpretation of which would require something quite different.

As the path of this paper is rather circuitous, a rough map may help. After stating the paradox (2), I go on to a short discussion of paradoxes in general, their uses and demands (3). I then leave Tertullian for a while, and attempt to show some features which distinguish religious, or at least Christian, language from other kinds of language (4); this is done by presupposing the existence of God, which may seem a rather peculiar procedure for a sceptic, but which will, I hope, serve for a discussion which tries to show something about religious language as used by believers. The thesis is then proposed that Christian belief must involve at least one statement which is about both God and the world, and that this statement must be partly incomprehensible--which I hold to be suggested by Tertullian's paradox, if given its head (5). Some remarks are then made on theology, and its relations to religious language and to the philosophy of religious language; these raise considerations that stop an incipient discussion of the incarnation, and suggest some rather disheartening conclusions about both the philosophy of religious language and theology (6). I end (7) with some observations about faith and about what one may or may not be said to believe on faith.

Tertullian's paradox I represent as a paradox both about Christian belief and about theology, but it is the former that is the more important point. In both cases I consider it as a paradox about meaning rather than about truth; that is, it is with questions of what is being said in religious language that I am concerned, rather than with questions of whether what is said is true, although the two sorts of question are not (and cannot be) kept clinically apart.

(2) Tertullian, the first Latin father of the Church, started his career as a lawyer and ended it as a heretic. After his conversion from heathenism in 196 he remained for only five or ten years a member of the Orthodox Church; both then and after his lapse into the Montanist heresy, he produced a series of theological works remarkable for vigorous reasoning, an unabashed use of legalistic rhetoric against his opponents, and an intransigent acceptance of paradoxical conclusions. The paradox I want to discuss comes from a work entitled de carne Christi which he wrote in the year 208, 'libris', as the Patrologia (Vit. Tert.) elegantly puts it, 'iam Montanismam redolentibus'--'at a time when his writings were already stinking of Montanism'--but the work is not itself, I believe, heretical. He is attacking Marcion, who believed that Christ was not actually born of the flesh, but was a 'phantasma' of human form. Marcion's refusal to believe in a genuine incarnation, Tertullian argues, could come only from a belief either that it would be impossible, or that it would be unworthy, a shameful degradation of the divine nature. Against the view that it would be impossible he produces the sweeping and general principle 'nihil impossibile Deo nisi quod non vult'--'nothing is impossible for God except what he does not wish to do'. In particular Marcion had argued that the idea of the incarnation of God involved a contradiction, because being born as a human being would involve a change in the divine nature;2 but a change involves ceasing to have some attributes and acquiring others; but the attributes of God are eternal; therefore he cannot change; therefore he could not have been born as a human being. Against this Tertullian says that this is to argue falsely from the nature of temporal objects to the nature of the eternal and infinite. It is certainly true of temporal objects that if they change they lose some attributes and acquire others; but to suppose that the same is true of God is just to neglect the necessary differences between God and temporal objects (de c. C. iii). (I shall in section (6) of this paper say something about this, perhaps not immediately convincing, argument.) Finally, against the view that, even if it were possible, God could not wish to be incarnated, because it would be unworthy of him, Tertullian, summing up his objections to Marcion in a passage of great intensity, accuses him of overthrowing the entire basis of the Christian faith: his argument would destroy the crucifixion and the resurrection as well. 'Take these away, too, Marcion,' he says (ibid., v), 'or rather these: for which is more unworthy of God, more shameful, to be born or to die? . . . Answer me this, you butcher of the truth. Was not God really crucified? And as he was really crucified, did he not really die? And as he really died, did he not really rise from the dead? . . . Is our whole faith false? . . . Spare what is the one hope of the whole world. Why do you destroy an indignity that is necessary to our faith? What is unworthy of God will do for me . . . the Son of God was born; because it is shameful, I am not ashamed: and the Son of God died; just because it is absurd, it is to be believed; and he was buried and rose again; it is certain, because it is impossible.' 'Non pudet, quia pudendum est . . . prorsus credibile est, quia ineptum est . . . certum est, quia impossibile': that is Tertullian's paradox.

(3) People who express themselves in paradoxes are in a strong position; and the more outrageous the paradox, in general the stronger the position. For an objector who insists on pointing out the absurdity of what has been said is uneasily conscious that he is making a fool of himself, for all he is doing is pointing out that the paradox is paradoxical, and this was perfectly obvious already: he is like a man who has missed the point of a joke or an ironical remark or an imaginative comparison, and insists on taking it literally. But ironical remarks and imaginative comparisons can have their point, and so can paradoxes; so it will not do, either, for the objector to dismiss the paradox in the hope that its evident absurdity makes it unworthy of discussion; for this is again to suggest that the person who uttered the paradox had overlooked its absurdity, but on the contrary he knew that it was absurd, and that was one reason why he uttered it. Because people do not in general utter absurdities unless they make a point by doing so, it is felt that the paradoxographer must have been saying something important. He not only prevents the critics answering, but makes them feel that in some mysterious way he is in a better position than they are; he is rather like a normally well-dressed man who appears at a function in a black tie and tails: the others present can't mention it to him, they can't overlook him, and they feel uneasy about their own turn-out. Or, again, he is something like a man who firmly closes a door in one's face: not only preventing one from going on, but making one feel one has no right to.

So far the paradoxographer has everything on his side, but it is not entirely so. For, as the man in the black tie, to make his effect, has usually to be well-dressed, and the man who closes the door has to be someone one respects, so the paradoxographer has to have some other claim on the attention of his audience: for in general a paradox, however suggestive in itself, does not represent solid earnings--it draws a little on yesterday's credit or mortgages a little of tomorrow's. This claim on one's attention can be possessed in various ways: positively, by the utterer being a good and impressive and genuine person whose life commands love and respect, or by other utterances of his being original and profound; and negatively, by other conflicting, or apparently conflicting, claims on our attention being confused and unhelpful, or made by persons whose way of life seems trivial, evil or disastrous. If this is so, we might expect to find the beliefs of a religion, for instance, being put forward with a particularly defiant paradoxicality in two sorts of situation: first, when its believers are intensely bound together by a new and compelling faith, and fighting for survival in a hostile but decaying society whose beliefs they utterly reject; and second when, whatever the divisions and discredit that have fallen on the belief itself, those who reject it, their own hopes perishing, seem to have little to offer in its place except angst, tyranny or imminent thermonuclear annihilation.

This, however, so far as it goes, suggests only why people, and in particular religious believers, should tend at one time rather than at another to express themselves in paradoxes; it says little about why anyone should ever at all choose to speak in paradoxes, or suggests at most that they do this as a striking way of getting people to listen to or consider something else. Often it is not much more: to say, for instance, that the Holy Roman Empire was not Holy, nor Roman, nor an Empire, is, or should be, a brisk way of preparing for a new historical analysis. But there are other paradoxes which seem more important and significant; where to grasp the paradox seems an essential part of understanding what is being said. Here we have the feeling that a paradox, granted that it has to be understood against a background of other beliefs or a way of life, itself tells us something: that it is in a certain way the essence of what is to be believed. This is particularly so in the case of religious beliefs, where the feeling has itself been expressed in many ways: perhaps by saying, that there is an infinity of things that are beyond our comprehension; or that our reason cannot embrace the deepest truths; or that what we say can only be an unsatisfactory (or, perhaps, analogical) account of what we believe on faith. I shall try to show how such a point of tension, of failure of language, must occur in religious belief, and I think, therefore, that we should take Tertullian's paradox seriously; not as just a rhetorical expression of his objections to a particular doctrine, but as a striking formulation of something which I shall suggest is essential to Christian belief.

(4) There has been much discussion in recent years of religious language and its relations to other types of language; a good deal of this discussion has been concentrated on religious statements, and a good deal of this on the one statement 'God exists'. I think it is now time to consider whether such concentration has not been too narrow: for in each respect it has had undesirable results. First, there is an unclarity in the idea of a language--meaning by this, of course, not a national or dictionary language, such as French or Esperanto, but a logically distinguishable language or type of discourse. Second, the concentration on religious statements, as distinct from other types of religious utterance, has produced a string of disruptive effects: it has overemphasized the difference between the apparently unfalsifiable religious statement and the falsifiable statement of the sciences, which is indeed important and will appear later in what I have to say, but which taken by itself leads to an impasse which looks a little like a reduplication in linguistic terms of the barren nineteenth-century dispute between science and religion; and efforts to get out of this impasse have involved, in some cases, attempts to reduce statements of religion to statements of something else, for instance, of mystical experience, and in others attempts to reduce statements of religion to other things that are not statements at all, such as commands or exhortations to a religious way of life--all of which either involve an evident circularity or omit the peculiarly religious character of the statements altogether. Third, there has been the concentration on the logic of the particular statement 'God exists'; this shows a kind of hopeless courage. It shows courage because this statement seems to be the lynch-pin of the whole system: to uncover what is involved in believing this should be to uncover the whole nature of religious language and the essence of religious belief. But it is just the peculiar importance of this statement that makes hopeless an inquiry that starts with it. Its peculiarity is such that it is extremely untypical of religious statements; a peculiarity emphasized by Collingwood, for instance, when he said that it was not a religious statement at all, but rather the presupposition of any religious statement. We might say that the statement of God's existence has indeed great logical power, but that it is the power not so much of a lynch-pin as of a lever: if we knew, from outside the religious system, how to work with it, we might move heaven and earth; but from outside we do not, because we know neither where we may fix a fulcrum nor where we can insert the other end of the lever. So rather than attempt such a direct approach, we must obey the Boyg, and go round.

I cannot hope to go far round, but perhaps something can be said. First, then, I think we must always bear in mind the fact that religious language is not used just for making statements, but that there are many other kinds of religious utterance: commands, for instance ('Thou shalt not take the name of the Lord thy God in vain'), and, very importantly, prayers, and expressions of trust ('Though he destroy me, yet will I wait for him'), and promises, and reprimands, and many others. Furthermore, none of these utterances, including the statements, is made in vacuo: sometimes they are used as part of a religious ceremony or observance, sometimes as part of a religious person's deciding what to do in a practical situation; and generally as part of the activities of life. This as a general point is one constantly emphasized by Wittgenstein; and in considering religious language it is, I think, particularly disastrous to ignore it.

But what is religious language? Is there one thing which is religious language? With what is it being contrasted? One thing, certainly, with which we must be wary of contrasting or comparing it is that nebulous and pervasive substance, 'ordinary language'. For one thing ordinary language should be the language used by most of us in going about our ordinary occasions, and the question of how religious that is, is the question of how religious or professedly religious most of us are; and if some of us all of the time, and most of us most of the time, do not bring talk about God into our affairs, that seems to be at least as much something about us as something about talk about God. This raises the question of dispensing with talk about God, of what is involved in doing without it; and about that I shall later say a little.

So one might ask, 'What are in general the distinguishing marks of a language, of a type of discourse?'; and in attempting an outline of an answer, one can think at once of at least five possible distinguishing marks. For one language might be distinguished from another by the types of logical relation holding within it; by its subject-matter; by its use of technical terms; by its purposes; or, more generally, by the activities with which its use is associated. But it would, of course, be an illusion to suppose that these five, even if they were satisfactorily distinguished one from another, would be competitors for the position of the one and only distinguishing mark of one language from another; it is rather that from the inter-relation of features like these we can, in particular cases, justifiably claim to distinguish one type of discourse from another. Which of these features one would particularly consider is a question partly of at what level the distinctions are being drawn. If we concentrate on distinctions between the sciences, at a low level of generality, we tend to fasten on distinctions of subject-matter, for we all learn at school that mycology is the study of fungi, and geology the study of rocks, and so on. But it is clear that in doing this we presuppose already a distinction between scientific and other discourse, and between one type of science and another: for not all talk about plants, for instance, and not even all scientific talk about them is botanical talk. Nor will the distinction of subject-matter apply at all to any but the most naïve distinctions between subsidiary sciences, the distinction elsewhere--for instance, between physics and physical chemistry--lying rather in the scope and terminology of the laws formulated and employed.

But it is not to the present purpose, even if it were possible, to attempt the high Aristotelian task of characterizing the differences between organized bodies of knowledge. For while it may be possible to characterize the language used by some type of scientist in his professional work, or to characterize a professional scientific activity to distinguish it from some other professional activity, such as that of the historian, this is beside our purpose, which is to characterize some unprofessional uses of language as distinct from others. It does seem clear, however, that when we, as laymen, speculate on the distance of a star, for instance, we are using language differently from when we remark on how beautifully it now shines; and that if we say that the first is a scientific use, part of what we mean is that we are asking a question to which the professional scientist is in the best position to give an answer--it is the sort of question he is asking. So we can at this point reintroduce the idea of a professional use of language, and say at least this much: that some of our utterances ask or involve questions that are properly to be answered by techniques and methods of inquiry professionally employed by some types of specialist, and others do not do this.

This distinction does not apply in any simple way to our investigation of religious language. In the case of religious belief, there is indeed the notion of a person who is a religious authority, but this is something quite different from a scientific authority. For first, the religious authority, if there is one, is at least not just someone who has a good training in the methods of answering certain sorts of question, but someone who has the authority to lay down what is to be believed or done. Second, the question of whether there is a religious authority even in this sense and, if so, who it is, has been the occasion of violent dispute, and many people have been killed in the attempts to settle it. But the dispute was about the settling of admittedly religious questions, so a reference to the authority cannot come into the characterization of a religious question. Third, even if we were to say that a specialist or professional use of religious language was to be found in its theological use (and about theology I shall have something to say later), it is clear that the relation of religious language to the theologian is different from the relation of scientific language to the scientist; one who speaks scientifically is at least an amateur scientist, but one who speaks religiously is not necessarily a theologian, even an amateur one.

How, then, can we attempt a characterization of religious, or at least of Christian, utterances, made in their ordinary occasions by persons other than professional theologians? Could we say, for instance, to take up one of the suggested criteria, that certain language was to be characterized as religious, or more specifically as Christian, by reference to certain practices or observances in the course of which it was used? It is clear that such a reference could not give us enough. For, first, the ceremonies would themselves have to be characterized as religious ceremonies, and if we could do this, we should already have a clearer idea of what religious language was. Second, many religious utterances are made outside such ceremonies; unless everyone speaks religiously only on Sundays. The ceremonies might in the end have to be mentioned in a full characterization of a religious life; but what we are looking for must first be found elsewhere. We have seen that in attempting the characterization of other kinds of language, the distinction of subject-matter, of what the language is used about, did not take us very far; but in the case of religious language perhaps we should after all return to it. For religious language, we might say, is, peculiarly, language about God; and by 'peculiarly' I mean not only that all religious language is language about God, but--and this seems to me an important point--that all language about God is religious language.

But to say that religious language is language about God immediately raises three related difficulties. For first, the word 'about' is misleading. In the most normal linguistic sense of 'about', it is statements that are about things or persons; but, as we have already seen, not all religious utterances are statements--a prayer, for instance, is not about God, but is addressed to him. If we are to say, then, that religious language is language about God, we have to take 'about' in an extremely wide sense. I take it that it would not be disputed by Christians that every religious utterance in some sense comes back to God, perhaps in the sense that if the purpose of the utterance is to be explained, God has in the end to be mentioned. In something like this sense, the word 'about' must be understood. I think we have to say, further, that the mere occurrence of the word 'God' in an utterance does not mean that it is actually about God, and so religious; for the most devout may use the word 'God' in idle phrases and not mean really to speak about God. An utterance which includes the word 'God' must be seriously meant to be about him for it to be actually about him.

Conversely--and this is the second difficulty--it is not the case that the word 'God' has to occur in an utterance for it to be religious. We could put this by saying that the distinction of subject-matter cannot be reduced to another distinction I mentioned, that of technical terminology. For there are many utterances that are religious even in the sense of 'Christian' but do not involve the word 'God'; and, more widely, there is religious language that is not the language of Christianity. To say that this other religious language is language about God, where 'God' is understood in a Christian sense, is at least to prejudge a particular theological issue, concerning the reference and truth of religious beliefs other than Christianity; but the fact that there is an issue here shows that there must be some characterization of a religion, and so of religious language, which is independent of the beliefs of Christianity. Thus we have to say that our characterization is one not of religious language in general, but of that of Christianity; and this will do for the purposes of this discussion.

But is it even this? For--and this is our third difficulty--saying that Christian language is language about God evidently presupposes the truth of Christianity in a far more radical sense, for it presupposes the existence of God. Therefore it looks as if we have to say that, if God exists, the language of Christianity is language about God, and this seems useless as a characterization of such language. For if we start from the statement of God's existence, the characterization is vacuous unless we already know that statement to be true; but if we know that statement to be true, the characterization appears superfluous. If, however, we start from the evident existence of Christian language, in the sense of language used by Christians, we might be tempted to arrive at the statement of God's existence, and so involve ourselves in a kind of ontological proof which might well be considered suspect. This all illustrates the peculiar relation to religious language of the statement of God's existence, which I have already mentioned. If we were seeking an independent characterization of Christian language these difficulties would be damning; but my present aim is not to do this, but to leave on one side the question of God's existence, and to try to show something about Christian language as used by Christians. So perhaps this rather paradoxical approach will not prove entirely useless. I shall therefore continue to speak of Christian language in a way that involves a suspension of disbelief, the suspension being achieved, evidently, by our own bootstraps. I have suggested, then, that all Christian language is language about God. I suggested before that all language about God is religious language, and this must stand, if it stands at all, in its original form: for to say that all language about God is Christian language is to prejudge to the opposite effect the theological issue, which I mentioned before, about the status of other religions.

But here perhaps we have an important point about religious language: for we saw before that, while the language of botany is language about plants, not all language about plants is botanical: for poets, painters, ramblers and so on may also speak of plants, but not botanically. But I want to suggest that all language about God is religious language--one cannot speak non-religiously about God. It does not follow from this that atheists are necessarily speaking religiously: for they are denying the statement 'God exists', and to do this goes behind the presuppositions of the present discussion. It does, however, follow from the present thesis that blasphemy is a kind of religious language, and such it must be--for how else could it give so much offence? Blasphemy is the misuse of religious language: it is to say things about God, or to ask things of him, but the wrong things. Yet there seems to be a sort of paradox here: for the blasphemer says, for instance, that God is wicked, and gives offence by so speaking of the Christian God. But the Christian God is good; so must not the blasphemer be speaking of some other God? But if he does this, he either gives no offence, because it was not of the Christian God that he was speaking offensively, or gives offence only by suggesting that there is another God--a line of argument that might lead to the intriguing conclusion that the only form of heresy is polytheism. Perhaps here we must say something like what we say about disagreements concerning characters in the historical past, that there must be some beliefs, and in the case of religion some practices, in common, between the blasphemer and the orthodox to support the idea that they are both talking about the same God: when Housman referred to 'Whatever fool or blackguard made the world', the description 'He who made the world' provided the place from which the offence was to be taken.

If we say, then, that all language about God is religious language, we have said something about religious language and its subject-matter which distinguishes these from, say, botanical language and its subject-matter. We must next consider one type of utterance which, very importantly, occurs within the body of religious language. I have stressed the fact that religious utterances can be of very various types: statements, commands, prayers, etc.; but it is important also that when we consider only those religious utterances that are statements we find that they as well can be of very various types. Some may be statements directly about the nature of God: 'God is three Persons'; some about historical events: 'God sent the Jews into exile'; some about human nature: 'God has given men free-will'; and so on. That is to say, there are many religious statements that are not just religious--although they are about God, they are about something else as well, something involving the affairs of men.

(5) We must now look more closely at the way in which some religious statements, to confine ourselves to statements, are not purely about God, but about human affairs as well; for by doing this we may become clearer about the range of religious language, its relation to other language and to theology: and we shall return, at long last, to Tertullian's paradox. Because religious statements are so various, many different ones should be discussed, but here I shall mention only one. It raises in itself some well-flogged issues, but these I do not want to discuss: I take this example only to illustrate a more general point.

If a people suffer from occasional failure of their crops and subsequent starvation, a person of rather Old Testament faith might say: 'God makes the crops fail to punish the people for their wickedness.' Such a statement is certainly a religious statement of a sort, but it is also a statement about certain events in human life, and seems in fact to provide an explanation of them; and it seems most clearly to do this, and looks logically like a non-religious explanation of the same events, because it connects with each other two sets of human events--the wickedness of the people and the failure of their crops--with God, as it were, as a middle term. As such, the statement seems also to be in a crude sense falsifiable. For when the agriculturists arrive, the irrigation is improved, the crops never fail, and the people riot in wickedness in the midst of plenty, the man who said that the crops failed because of the people's wickedness notoriously falls into discredit. People will cease to talk of God in explaining the success or failure of the crops: one sort of religious statement will cease to be made. This is the familiar phenomenon of the elimination of religious language from a context; and it has been eliminated here not just because people have come no longer to speak in a certain way--as a people might cease to write some sort of poetry--but because the religious statement, in this particular crude example, was a kind of explanation, and was run over by a rival and better explanation. We mentioned before a distinction of languages in terms of specialists and their techniques, and where a language in the specialist sense can clash with religious language, religious language tends to be driven out; because the specialist techniques give explanations which are recommendations for effective action, and where religious language claimed to do that, it failed: for either it gave an 'explanation' which wasn't an explanation in this sense at all, and provided no recommendation, or it gave an explanation, as in the case of the crops, but a very bad one--for if anyone believes that the best way to prevent natural disasters is to live a better life, he appears to be in error.

It would, of course, be a crude mistake to suppose that these antique considerations could, in some sense, 'disprove religion'. What they do show is that if religious language is used to give certain sorts of explanation, it clashes with a more effective explanation and tends to be eliminated. Such elimination has its effect, too, on the theology of the user of the language. For the religious explanation, as we have seen, was a statement both about certain events and about God; and if these statements are seen to be inapplicable to events, they are seen to be inapplicable also to God. Hence it will come to be seen, perhaps, that certain things cannot be said of God: for instance, that he produces particular disasters as a punishment to men. This in turn leads to new speculation about the nature of God and his relation to the world; so that a change in the possible uses of religious language is connected with a change in the views about the nature of God. This works also the other way; for it would be false to represent the situation as one of the constant retreat of religious language, with consequent trimming of theological doctrine. Undoubtedly this has happened; but there may also be new thoughts about God and new moral views and following from these, changes in what can be said in religious language; as with the coming of Christianity less was said about the anger of God and more about his mercy, and as there comes with a change in religious belief a change in what it is considered proper to ask of him in prayer. An attempt may be made at each stage to co-ordinate the implications of what men think they can and ought to say in religious language, and such co-ordination takes the form of a series of statements about the nature of God: and this is systematic theology.

But although the changes in the range of religious language are not to be described entirely as a retreat, the retreat, as we all know, has its dangers for the religion. The supposed religious explanation that we mentioned was in its rough way one statement about both God and the world; and if all statements that are about both God and the world were to be abandoned, what would be left? Such statements would not need to be, as that one was, explanatory of what goes on in the world, and indeed could not be; but there are connections other than explanations, and some such there must surely be. Wittgenstein said (Tractatus, 6.432): 'How the world is, is completely indifferent for what is higher. God does not reveal himself in the world.' But if all talk about God were talk only about God, and all talk about the world talk only about the world, how could it be that God was the God of the Christian believer, who is a toiler in the world of men? Would not the views about the nature of God retire more and more away from the world of men: his existence would become like that of the gods of Epicurus, 'far remote and cut off from our affairs' (Lucretius, de Rerum Natura, II, 648). And if that happened, it could not be of much concern whether he were there or not.

(6) This is where we return to Tertullian. Tertullian's paradox is relevant to this question both because it is a paradox and because it is about the incarnation. For the incarnation seems to be the point for the Christian faith, where there must essentially be an intersection of religious and non-religious language; it has to be said not only that a certain person was crucified, but that that person was the Son of God. This has to be said, as Tertullian clearly saw, if there is to be a Christian faith; and as he equally clearly saw, it is a paradox. The paradox comes about because, although we must have some statement which says something about both God and the world, when we have it we find that we have something that we cannot properly say. For when God is spoken of in purely religious language, he is said to be a Person eternal and perfect, that is, we do not speak of him in terms appropriate to the temporal and imperfect objects and persons of this life; or if we do, it is notoriously by the analogy of which theologians speak, and therefore imperfectly. For there is no language for God's eternity and perfection beyond the statement of it: it can be said that God is eternal and perfect, not how he is, for God's eternity and perfection must be beyond the reach of our understanding. So when we come to a statement that is about both God and temporal events, it must be unsatisfactory; for if it were not, we should have adequately described the relation of the temporal events to God in terms appropriate only to the temporal events: and this would mean either that we had described only the temporal events, and left God out, or had described God as a temporal being, which he is not.

The difficulty seems to follow not from the eternity of God by itself, but from the conjunction of this with his perfection as a personal being. For some have held, for instance, that the numbers are eternal objects, although mathematical statements about things in the world can satisfactorily be made. One difference of this case from that of God could be marked by saying that, leaving aside the question of application, the nature of the numbers in themselves can be adequately expressed in the language appropriate to this, the language of pure mathematics, but the nature of God cannot be adequately expressed in any human language. But if we say this, it looks as though we were defending now a different thesis about religious language. For this seems to say that any statement about God, whether we say that there is a relation between God and the world or not, will be unsatisfactory, just because it is made in the words of human language; but the thesis was that it is the fact that there must be a relation between God and the world that made religious language unsatisfactory. But it is not really a different thesis; for it is just the fact that there is at some point such a relation, and a statement or set of statements that try to express it, that makes religious language elsewhere also unsatisfactory. The question of the applicability of mathematics to the world does not affect the question of the expression of the nature of the numbers by pure mathematics; but the question of the relation of God to the world does affect the question of the expression of the nature of God in religious language. The actual effect is that God is said to be a perfect personal being; because, for instance, prayers are addressed to him, and because he has a Son who was born into the world. The statement of these relations will be itself unsatisfactory, and will involve others that are so: because the concepts required--of fatherhood, for instance, and of love, and of power--are acquired in a human context; the language of these things is a language that grows and is used for the relations of humans to humans. To say that, while this is so, religious language requires merely an extrapolation from the human context,3 is not to solve the problem but to pose it again. For the extrapolation required is an extrapolation to infinity, and in even trying to give a sense to this we encounter the incomprehensibility. This incomprehensibility Tertullian has brought out in his paradox.

In fact, it is a double paradox: 'because it is shameful', Tertullian says, 'I am not ashamed . . . it is certain, because it is impossible'; that is, there is something that is morally outrageous about it, and there is something intellectually outrageous. The two paradoxes can perhaps be seen by considering the incarnation from two different directions. That God, a perfect being, should be willing to be born and to be crucified, is morally astonishing; that this man on a cross should actually be God is intellectually astonishing. Of these, the moral paradox is perhaps the more readily comprehensible to the unbeliever; for this at least he has a model, in the ideas of humility and sacrifice and the finding of the greatest value not where the worldly are looking for it. So the unbeliever, perhaps impertinently, may feel that he sees a point to the moral paradox--that it has turned upside down the standards of what is to be admired and loved. The feeling is easier, perhaps dangerously so, because we have a Christian tradition: to the educated Roman, for instance, it must have been deeply shocking. In this case, too, we can understand to some extent what is in fact the centre of the paradox, Tertullian's saying not just that it was absurd and he believed it, but that he believed it because it was absurd. It was just the outrageousness of the crucifixion that pointed the new way one had to try to follow in one's life,4 and how can any of this be applied to the second part of the paradox, to the fact of the incarnation? How can it be certain, because it is impossible? How can we come to understand, how can we give any sense to, the statement that this man who was crucified was God?

Here I encounter fully a difficulty that has been gradually making itself felt throughout. For the examination of the meaning of statements about the incarnation is, or certainly has been, a task for the theologian; and not being a theologian I cannot feel competent to undertake it. And yet, by starting out to look at religious language I seem to have reached a point at which it is necessary to turn into a theologian. I think we can see the reason for this if we consider what has already been said about religious language and theology. I suggested before that there was a relation between what can and cannot be said in religious language and systematic theology; that a contraction or extension in the use of religious language leads to changes in the theology; and that the systematization and explanation of the implications of what can be said about God is a task of theology. But this seems to have two consequences. For the theology examines and changes by reference to the logical consequences of speaking in this or that way about God: if we cannot say 'God sent the drought to punish the people', we must say that God does not intervene in the operations of natural law; if we say this, are we to say that God's power is limited or that he himself is willing not to intervene?--and so on: all traditional problems of theology. And if we say that God was incarnated, are we to say that he changed?--Marcion's problem. But if the raising of these questions is a task of theology, then theology seems to include the logical analysis of religious language; for surely the logical analysis of religious language is just this, asking how, and with what implications, utterances are made in religious contexts. So the philosopher who regards his task as the logical analysis of language and who sets out to examine religious language will find himself, I suspect, as I have done, doing theology. This, which is the first consequence, seems to me not too happy a one for the supposedly independent analyst of religious language. For I have a suspicion that as a theologian he will turn out rather poorly; as some indeed have, supposing themselves to be raising for the first time logical difficulties in Christian language which have in fact in one guise or another been the concern of theologians for centuries. If he is not a believer at all, his case will be worse still, for the utterances are not just there, to be pulled to pieces without understanding of the context in which they are used; but can he understand what is the context and importance of a prayer, for instance, unless he understands what it is to pray? Any more than a man can write on the language of aesthetics who cannot see beyond a coloured photograph of 'The Laughing Cavalier'.

The first consequence I suggest, then, of the status of theology is that there is not much hope for an independent logical analysis of religious language; and the second is its converse: that if one task of theology is such an analysis, theology is committed to making itself coherent, and coherent not only with itself, but outside as well. We have already seen how religious language might retreat from human affairs into an Epicurean remoteness, and that this must not be, if it is to be of any use. So it is that theology must show how religious language can gear into other language, and must lay bare the points of intersection. Yet in the end, it seems, it cannot be successful in this; for the points of intersection, as I have tried to say, must contain something incomprehensible. In saying this, I am only saying what theologians and other religious people have nearly always said; and this shows, what in any case follows from the nature of the thing, that while one should be a believer to be a theologian, being a believer does not eliminate the incomprehensibility. For if the belief is true, it is a belief in an eternal but personal God with a concern for the world, and it is from this that the incomprehensibility follows.

Having just disqualified myself from becoming a theologian, I shall not pursue the question of the logic of statements of the incarnation. I shall say, however, that I think it is clear that one cannot deal with the difficulties in the summary way which, in the work under discussion, Tertullian takes. It will be recalled that Marcion had said that if God had been incarnated, he would have changed; but change involves losing some attributes and gaining others; and God cannot do this. Tertullian briskly replied that what Marcion had said was true of temporal objects, but God is not a temporal object, and that therefore what Marcion said did not apply. But this is to counter one's opponent's move by smashing up the chessboard. For Marcion's objection, we might say, is a point about the logic of the word 'change'; we only understand the word 'change' in terms of the losing or gaining of temporal properties: so how can we use it of God? So something else must be said; but then, again, if the beliefs are true, nothing can be said that will really do. Tertullian's paradox is also a paradox of theology: it seems committed to what on its own premisses must be an impossible task.

(7) If it is impossible, what is to be done? Here it may be said that we must have faith; and further that the incomprehensibility I have been discussing is not only a necessary feature of Christian belief but necessary to it, for it is this that provides a place for faith. Tertullian himself I take, on my freewheeling interpretation, to suggest this in the core of his paradox: it is certain, he says, because it is impossible.

We must distinguish here several things that may be meant by having faith. For we may have faith in a person, in the sense that we continue to trust their honesty, good intentions, wisdom, etc., despite perhaps an apparent perversity in their actions. Or we may believe on faith a statement that such-and-such is the case, despite all the evidence being to the contrary. Or we may have faith that such-and-such is what ought to be done, despite the fact that actions and the results of actions involved in carrying out this policy are such as otherwise we should consider wrong. These kinds of faith are, of course, found together: when, for instance, Lenin asked the Bolsheviks before the Revolution to have faith in him, although many of his actions would appear to them inexplicable, he was asking them to believe, among other things, that the aims of the Party would be effected by his policies, although often they seemed to be moving in the opposite direction; and a humanitarian member of the Party had to continue to believe that the Bolshevik state was the right thing to aim at, although murder and misery were involved in doing so. These kinds of faith can be paralleled in the case of religious beliefs; but in the former cases, one thing at least seems to be clear, what it is that is being believed; for if a man had faith in Lenin as leader of the Party, or in the belief that his policies would forward the Revolution, he knew what it was he was believing, although he might be able to give very little in the way of rational grounds for believing it. But it is a stranger request to ask someone by faith to believe something that he does not properly understand; for what is it that he is being asked to believe? Faith might be a way of believing something, as opposed to believing it on evidence; but how could it be a way of stepping from what is understood to what is not understood?

Well, it might be said, faith can be a way of coming to understand something; and here it might be suggested that there is an analogy in the arts. 'You think this stuff is all nonsense,' someone might say about a poem; 'but just believe that the poet is not trying to fool you, take it seriously, and you will come to see what it is about.' The eighteenth-century hymnologist, in slightly more utilitarian terms, made something like this point when he wrote: 'O make but trial of his love; Experience will decide How blest are they, and only they, Who in his strength confide.' But the analogy is not good enough. For here again the initial faith is in a belief that is itself comprehensible: the belief that the poem has a meaning, if one can only find it. But in the case of religious belief it is just the belief itself, and not a prior belief about its comprehensibility, that one has, on the position being discussed, to take on faith, in the hope that afterwards it will become clear what it means. Here again I encounter the same difficulty: for if you do not know what it is you are believing on faith, how can you be sure that you are believing anything? And a fortiori how can such belief be the means to something else, viz. coming to understand?

In any case, this is beside the point; for the original argument was that certain religious beliefs must be inherently mysterious and remain so, and that it is the part of faith to accept them. My difficulty is that, if the belief is incomprehensible and necessarily so, one cannot see what is being accepted, on faith or otherwise.

St. Paul (I Cor. 1.20 f.) writes: 'Where is the wise? Where is the scribe? Where is the disputer of this world? hath not God made foolish the wisdom of this world? For after that in the wisdom of God the world by wisdom knew not God, it pleased God by the foolishness of preaching to save them that believe . . . the foolishness of God is wiser than men'; and in explanation a French commentator, F. Godet, has said: 'l'évangile n'est pas une sagesse, c'est un salut'--'the Gospel is not an intellectual system, but a salvation'. It might be objected that my argument has been treating Christian beliefs too much as a 'sagesse', and that a system of coherent and comprehensible beliefs is not to the point. This might be put differently by saying that in the later part of this paper I have neglected what I emphasized in the earlier, that religious language is not used only to make statements but for many other purposes as well; that the statements of religious belief are to be understood only as part of a way of life, which includes prayer and religious observance and so on; and it might be said in connection with the previous discussion that what one chooses, when one chooses to believe, is to live in a certain way, in which the statements play a part. This is true; but the statements do play a part, and the beliefs must be there, and that is the point. We may consider again the possible contraction of religious language, the lessening of its scope, which I discussed before, God may cease to be mentioned in explanations of particular physical events, for instance, or in moral discourse, and they will continue as forms of discourse on their own. What would not make sense would be for God to cease to be mentioned in the forms of religious observance or in prayer, for then they would no longer exist at all. But religious observance and prayer stand for nothing, so far as I can see, unless there are also behind them some beliefs about God, some statements about him: for this would indeed be the end needle point of faith, to pray just to the unknown God, in complete ignorance of whether such an activity had any sense in relation to him or not--or rather, in such ignorance, one would have to say 'it' rather than 'him'; and could one even say that? Something must be believed, if religious activities are not just to be whistling in the dark without even the knowledge that what one is whistling is a tune; and something that connects God with the world of men. But such a connection must involve saying something about God that is interpreted not just in terms of other statements about God, but in terms of the life of men. If this is said, it seems that it must either be so like some non-religious statement, as in our crude pseudo-scientific example of the failure of the crops, that it can conflict with such a statement, which would make the central religious belief falsifiable and in no way what was required; or it must be sufficiently a statement about God, as it were, for it to be mysterious, as involving an attempt to express the appearance in, or other connection with, a human situation of the infinite perfection of God. If it is inherently mysterious, then it cannot be explained by reason; but to say that it is to be believed on faith, and not by reason, does not face the difficulty: for the question was not how it should be believed, but what was to be believed. If, then, the Christian faith is true, it must be partly incomprehensible; but if it is partly incomprehensible, it is difficult to see what it is for it to be true.

(8) This is only Tertullian's paradox with a converse: credibile est quia ineptum; et quia ineptum, non credibile. It follows further, if this is the case, that it is difficult to characterize the difference between belief and unbelief. We can indeed point out that the believer says certain things which the unbeliever does not say; but we want not just this, but to know what it is that the believer believes and the unbeliever does not believe; but this we cannot properly do. But if we cannot adequately characterize the difference between belief and unbelief, we may not be able to characterize the difference between orthodoxy and heresy: for the difference between persons believing different ineptitudes is as obscure as that between those believing one ineptitude and those not believing it. Tertullian, as I mentioned at the beginning, became a heretic.

Return to Book Description

File created: 8/7/2007

Questions and comments to:
Princeton University Press

New Book E-mails
New In Print
PUP Blog
Princeton APPS
Sample Chapters
Princeton Legacy Library
Exam/Desk Copy
Recent Awards
Princeton Shorts
Freshman Reading
PUP Europe
About Us
Contact Us
PUP Home

Bookmark and Share