Anda di halaman 1dari 204

Wittgenstein and

Philosophy of Religion
Wittgenstein and
Philosophy of Religion

Edited by Robert L.Arrington

and Mark Addis

London and New York

First published 2001
by Routledge
11 New Fetter Lane, London EC4P 4EE

Simultaneously published in the USA and Canada

by Routledge
29 West 35th Street, New York, NY 10001

Routledge is an imprint of the Taylor & Francis Group

This edition published in the Taylor & Francis e-Library, 2003.

© 2001 Edited by Robert L.Arrington and Mark Addis

All rights reserved. No part of this book may be reprinted or

reproduced or utilised in any form or by any electronic, mechanical, or
other means, now known or hereafter invented, including photocopying
and recording, or in any information storage or retrieval system,
without permission in writing from the publishers.

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication Data

Wittgenstein and philosophy of religion/edited by Robert L.
Arrington & Mark Addis
p. cm.
Includes index
1. Wittgenstein, Ludwig, 1889–1951—Religion. 2. Religion—
Philosophy. I. Arrington, Robert L., 1938– II. Addis, Mark.

B3376.W564 W546 2001

210´.92–dc21 00–055323

ISBN 0-203-46258-0 Master e-book ISBN

ISBN 0-203-77082-X (Adobe eReader Format)

ISBN 0-415-21780-6 (Print Edition)

Contributors vii
Editors’ introduction ix
Acknowledgements xiii
Abbreviations xv

1 The gospel according to Wittgenstein

2 Wittgenstein and magic
3 Wittgenstein, religious belief, and On Certainty
4 Creation, causality, and freedom of the will
5 Faith: themes from Wittgenstein, Kierkegaard and Nietzsche
6 D.Z.Phillips’ fideism in Wittgenstein’s mirror
7 Wittgensteinian religion and ‘reformed’ epistemology
8 Wittgenstein and the interpretation of religious discourse
9 Wittgenstein and Wittgensteinians on religion
10 ‘Theology as grammar’: Wittgenstein and some critics

Index 185


Mark Addis is a Lecturer at the University of Central England and the author
of Wittgenstein: Making Sense of Other Minds.
Robert L.Arrington is Professor of Philosophy at Georgia State University. He
is the author of Rationalism, Realism, and Relativism and Western Ethics,
the editor of the Blackwell Companion to the Philosophers, and co-editor of
two other collections of essays on Wittgenstein published by Routledge.
Alan Bailey is College Lecturer in Philosophy at Pembroke College,
University of Oxford.
William H.Brenner is Professor of Philosophy at Old Dominion University
and the author of Logic and Philosophy and Wittgenstein’s Philosophical
Brian R.Clack is Tutor in Philosophy at St Clare’s International College,
Oxford. He is the author of An Introduction to Wittgenstein’s Philosophy
of Religion and Wittgenstein, Frazer and Religion, as well as the co-author
of The Philosophy of Religion; A Critical Introduction.
Paul Helm is Professor of the History and Philosophy of Religion at King’s
College, University of London and the author and editor of numerous
works in the philosophy of religion, including his recent Faith and
Understanding and Faith With Reason.
Michael P.Hodges is Professor of Philosophy and Chair of the Department of
Philosophy at Vanderbilt University and the author of Transcendence and
Wittgenstein’s Tractatus.
John Hyman is a Fellow of Queen’s College, Oxford, and the author of The
Imitation of Nature.
Kai Nielsen is Professor Emeritus of Philosophy at the University of Calgary
and Adjunct Professor at Concordia University. He is the author of
numerous books and articles in the philosophy of religion and other areas
of philosophy, including the recent Naturalism Without Foundations.
Iakovos Vasiliou is Assistant Professor of Philosophy at Georgia State
University and the author of articles relating Wittgenstein to Greek

Editors’ introduction

Wittgenstein’s remarks on religious belief have had an influence quite

disproportionate to their number. He wrote very little on the subject, and
much that we have from him on the topic comes from brief collections of
remarks, notes others made of his lectures, and records of snippets of thought.
In his later period, there are primarily the ‘Remarks on Frazer’s Golden
Bough,’ the ‘Lectures on Religious Belief,’ and occasional remarks in Culture
and Value. Nevertheless, most anthologies in the philosophy of religion and
many collections of essays designed for introductory philosophy courses will
have sections on the Wittgensteinian approach to religion (usually referred
to as a form of fideism). His thought in this area has also had an impact in
cognate areas such as religious studies and theology.
In this volume our hope is to convey some of the excitement about
Wittgenstein’s later thought on religion. We want to show how stimulating
and suggestive Wittgenstein’s remarks can be—how they can lead to a
totally new perspective on religious belief, to new ways of understanding
specific topics such as creation and freedom of the will, and to a new focus
for debating the issue of faith and reason. We also want to demonstrate
how very controversial these remarks are. Wittgenstein scholars are not of
a single mind regarding the significance of what Wittgenstein had to say on
the subject, as will be readily apparent on reading several of the following
essays. Moreover, some Wittgenstein scholars reject what appears to be the
central philosophical message found in the few remarks on magic and
religious belief—even while they accept what Wittgenstein has to say about
language in other areas of discourse. And there are, of course, non-
Wittgensteinians who forcefully repudiate the implications of his approach
to religion.
John Hyman gets us off to a good start with a brief introduction to
Wittgenstein’s overall philosophy—both his early thought in the Tractatus
Logico-Philosophicus and the later thought as found in Philosophical
Investigations. After this survey and a brief treatment of the main themes in
Wittgenstein’s philosophy of religion, Hyman raises some questions about

x Editors’ introduction

the acceptability of Wittgenstein’s remarks on religion. The doubts expressed

in his questions will resonate with many philosophers.
Brian Clack’s essay consists of an interpretation of Wittgenstein’s thoughts
on magic found in his ‘Remarks on Frazer’s Golden Bough’. Clack’s
interpretation is at odds with the prevailing ‘expressivist’ interpretation of
what Wittgenstein has to say on this topic. By extension, Clack can be read
as challenging expressivism as a proper way of understanding Wittgenstein
on religious belief in general.
The next essay—by Iakovos Vasiliou—also gives us a distinctive reading
of Wittgenstein on religion. Vasiliou leads us to see the remarks on religious
belief through the lenses of Wittgenstein’s On Certainty. This approach has
the virtue of demonstrating how Wittgenstein’s philosophy of religion—brief
and scattered as it is—is consistent with themes he developed at length in
some of his last writings.
William Brenner turns to two of the topics that are standard in the
philosophy of religion: creation and freedom of the will. He shows that
although Wittgenstein explicitly rejected a cosmological conception of God
as First Cause, his thoughts on causation and related topics allow us to
develop a new understanding of what many religious believers mean when
they speak of God as creator of the world and when they attribute free will to
themselves. Brenner’s essay demonstrates how Wittgenstein’s often cryptic
remarks can lead a thinker to new and imaginative ways of viewing the
religious life.
The central notion of ‘fideism’—the concept of faith—is given an extended
discussion by Michael Hodges. He examines Kierkegaard’s revolutionary
thoughts on faith and the influence they exerted on Wittgenstein. But Hodges
is also impressed with Nietzsche’s genealogical approach to religion and the
critical perspective on the religious life that this approach assumes. Thus
Hodges is led to raise the question whether Wittgenstein’s infamous
quietism—his insistence that philosophy ‘leaves everything as it is’ and cannot
serve as a higher epistemological authority—can be challenged. Hodges then
envisages several ways in which one might try to gain a critical distance and
grip on religious discourse and the religious life. He wants to know whether
this can be done without violating some of Wittgenstein’s central ideas.
Probably the most influential commentator on Wittgenstein’s philosophy
of religion—and an important philosopher in his own right within this area
of philosophy—is D.Z.Phillips. Phillips is the person one normally associates
with the fideistic interpretation of Wittgenstein. But his readings of
Wittgenstein are controversial, and Mark Addis discusses many of the topics
on which some Wittgenstein scholars would take issue with him and their
reasons for doing so. Addis addresses some of the key notions operating in
most commentaries on Wittgenstein’s remarks on religion—the notions of
language games and forms of life—and attempts to bring clarity to their
meaning and application.
One of the most important positions in recent philosophy of religion is the
Editors’ introduction xi

approach of what is called ‘reformed epistemology’—a point of view closely

associated with Alvin Plantinga. What is the relationship between Plantinga’s
ideas and those of Wittgenstein—and those of Wittgensteinians such as Phillips
and Anthony Kenny? Paul Helm provides a guide to the similarities and the
differences between these two influential interpretations of religious belief. He
points to ways in which the one side has unfairly criticised the other, and he
identifies in both approaches areas where clarity and persuasiveness are less
than what one would hope for. And he tries to see how both sides line up with
regard to today’s realism/anti-realism debate in philosophy.
Alan Bailey begins his essay by pointing to some features of Wittgenstein’s
method, and he then proceeds to identify key elements of Wittgenstein’s
philosophy of religion. Bailey is a critic of Wittgenstein’s thought in this area.
He gives numerous reasons for thinking that Wittgenstein has
mischaracterised the nature and meaning of religious discourse. Bailey’s essay
draws on the work of contemporary philosophers who have studied the idea
of attributing beliefs to others—Dennett, Davidson, and Stitch—and Bailey
uses these studies in developing his own attack on Wittgenstein.
One of the best-known critics of Wittgenstein on religion is Kai Nielsen,
whose 1970s article ‘Wittgensteinian Fideism’ contained a forceful rejection of
much that Wittgenstein had to say on the topic. In his new essay for this
volume, Nielsen expresses an appreciation of many aspects of the later
Wittgenstein’s thought, but he continues to argue against what seem to him to
be the central messages coming from Wittgenstein with regard to religion. As
he develops his interpretation of Wittgenstein on religion, Nielsen cites the
work of two major Wittgenstein commentators, Norman Malcolm and Peter
Winch. He utilizes some of Winch’s thoughts to initiate his criticism of
Wittgenstein, but he goes on to develop his own distinctive reasons for thinking
Wittgenstein wrong, especially about Wittgenstein’s quietism—his insistence
that philosophy cannot provide a critical assessment of religious practices.
The book concludes with an essay by Robert Arrington in which he
attempts to respond to some of the criticisms that are leveled against
Wittgenstein on religion by some of the other contributors to the book.
Arrington focuses on Wittgenstein’s characterization of theology as grammar.
He argues that this notion, developed and extended, reveals the weaknesses
of many of the reasons given for thinking that Wittgenstein has
mischaracterized religious discourse and for believing that Wittgenstein has
unconvincingly insulated religious belief from rational criticism.
Reading Wittgenstein is a philosophical experience to be relished. It leads
many readers to energetic counter-argument; it leads others to new ways of
seeing things that bring intellectual satisfaction of the highest order. It is
hoped that the essays in this book will prompt their readers to go to the
Wittgenstein texts on religion themselves, again or for the first time, and to
participate in the intellectual excitement that Wittgenstein generates in this
area of his thought. And if the essays succeed in casting some light on these
texts, they will completely fulfill their authors’ present aims and ambitions.

The editors gratefully acknowledge the assistance of Ms Ellen Logan, Office

Manager in the Department of Philosophy at Georgia State University, Ms
Andrea Howard, Administrative Assistant in the department, and Mr Robert
Vinten, a graduate research assistant in the department. They also express
appreciation to Blackwell Publishers for permission to print a revised and
expanded version of John Hyman’s essay ‘Wittgensteinianism’ from Philip L.
Quinn and Charles Taliaferro (eds), A Companion to Philosophy of Religion
(Oxford and Cambridge, MA: Blackwell, 1997), pp. 150–7.

Abbreviations of titles of
Wittgenstein’s works cited
in this volume

AWL Wittgenstein’s Lectures, Cambridge 1932–35 (from the Notes of

Alice Ambrose and Margaret McDonald), ed. Alice Ambrose
(Oxford: Blackwell, 1979).
BB Blue and Brown Books (Oxford: Blackwell, 1958).
BT The ‘Big Typescript’ (Ts 213): a rearrangement, with
modifications, written additions and deletions, of Ts 211, 1933.
CV Culture and Value, ed. G.H.von Wright in collaboration with
H.Nyman, trans. P.Winch (Oxford: Blackwell, 1980); rev. edn ed.
Alois Pichler (Oxford: Blackwell, 1998).
LC Wittgenstein: Lectures and Conversations on Aesthetics,
Psychology and Religious Belief, ed. C.Barnett (Oxford:
Blackwell, 1966).
‘LE’ ‘A Lecture on Ethics’, in Philosophical Occasions 1912–1951,
ed.James Klagge and Alfred Nordman (Indianapolis and
Cambridge, MA: Hackett, 1993), pp. 36–44.
LFM Wittgenstein’s Lectures on the Foundations of Mathematics,
Cambridge 1939, ed. C.Diamond (Sussex: Harvester Press, 1976).
NB Notebooks 1914–16, ed. G.H.von Wright and G.E.M. Anscombe,
trans. G.E.M.Anscombe (Oxford: Blackwell, 1961).
OC On Certainty, ed. G.E.M.Anscombe and G.H.von Wright, trans.
D.Paul and G.E.M.Anscombe (Oxford: Blackwell, 1969).
PG Philosophical Grammar, ed. R.Rhees, trans. A.J.P.Kenny (Oxford:
Blackwell, 1974).
PI Philosophical Investigations, 3rd edn ed. and trans. G.E.M.
Anscombe (Oxford: Blackwell, 1958).
PO Philosophical Occasions 1912–1951, ed. James C.Klagge and Alfred
Nordman (Indianapolis and Cambridge, MA: Hackett, 1993).
PR Philosophical Remarks, ed. R.Rhees, trans. R.Hargreaves and
R.White (Oxford: Blackwell, 1975).
RF Remarks on Frazer’s Golden Bough, ed. R.Rhees (Doncaster:
Brynmill Press, 1979).

xvi Abbreviations

‘RF’ ‘Remarks on Frazer’s Golden Bough’, in Wittgenstein: Sources

and Perspectives, ed. C.G.Luckhardt, trans. John Beversluis
(Ithaca: Cornell University Press, 1979; Harvester Press, Hassocks,
1979; Bristol: Thoemmes Press, 1996); also in Philosophical
Occasions, ed. James Klagge and Alfred Nordman, trans. John
Beversluis (rev.) (Indianapolis and Cambridge, MA: Hackett,
RFM Remarks on the Foundations of Mathematics, ed. G.H.von
Wright, R.Rhees and G.E.M.Anscombe, trans. G.E.M. scombe,
rev. edn (Oxford: Blackwell, 1978).
ROC Remarks on Colour, ed. G.E.M.Anscombe, trans. Linda cAlister
and Margarete Schättle (Oxford and Berkeley: Blackwell and
University of California Press, 1977, 1978, 1979).
RPPI Remarks on the Philosophy of Psychology, vol. I, ed. G.E.M.
Anscombe and G.H.von Wright, trans. G.E.M Anscombe (Oxford:
Blackwell, 1980).
TLP Tractatus Logico-Philosophicus, trans. D.F.Pears and B.F.
McGuinness (London: Routledge & Kegan Paul, 1961).
WWK Ludwig Wittgenstein und der Wiener Kreis, shorthand notes
recorded by F.Waismann, ed. B.F.McGuinness (Oxford: Blackwell,
1967). The English translation, Wittgenstein and the Vienna Circle
(Oxford: Blackwell, 1979), matches the pagination of the original
Z Zettel, ed. G.E.M.Anscombe and G.H.von Wright, trans.
G.E.M.Anscombe (Oxford: Blackwell, 1967).
Nachlass All references to unpublished material follow von Wright’s
catalogue Wittgenstein, 35ff. They are by Ms or Ts number
followed by a page number.
1 The gospel according to
John Hyman

Wittgenstein’s early philosophy was worked out in the six years or so
following his arrival in Cambridge in 1911, and published in the Tractatus
Logico-Philosophicus in 1922. After a long hiatus, Wittgenstein took up
philosophy again in 1929, and soon began to develop the ideas which were
published after his death—first in the Philosophical Investigations, the
masterpiece of his mature philosophy, and then in editions of various
notebooks, drafts and collections of philosophical remarks. Both of these
philosophies include highly original and influential views about the nature of
religion. I shall discuss them in turn.

Wittgenstein said that the fundamental idea of the Tractatus is ‘that the “logical
constants” are not representatives; that there can be no representatives of the
logic of facts’ (TLP 4.0312). Perhaps a simpler way of expressing this thought
is to say that the propositions of logic are not descriptions. Frege had thought
that the propositions of logic describe timeless relations between abstract
objects; Russell had thought that they describe the most general features of the
world. We arrive at the propositions of logic, according to Russell, by
abstracting from the content of empirical propositions, and so the propositions
of logic themselves describe the world we encounter in experience, but they do
so in the most abstract and general terms.
Wittgenstein argued that Frege and Russell underestimated the difference
between the propositions of logic and empirical propositions, because they
agreed in thinking, or assuming, that however different these kinds of pro-
positions may be, however different the kinds of things they say are, they still
have this much in common, that they say something. Wittgenstein’s own view
was that the propositions of logic say nothing; they contain no information
whatsoever; they are simply tautologies: ‘For example, I know nothing about
the weather when I know that it is either raining or not raining’; ‘all the
propositions of logic say the same thing, to wit nothing’ (TLP 4.461, 5.43).

2 John Hyman

If logical propositions say nothing, what is it for a proposition to say

something? The answer Wittgenstein gave in the Tractatus is one he later
summarised as follows: ‘The individual words of language name objects—
sentences are combinations of such names’ (PI§1). Accordingly, the sense of
a sentence will depend on the meanings of the words which are combined in
it, and the way in which they are combined. Just as the objects to which the
individual words correspond can be combined or arranged in different ways,
so can the words in a sentence; and the sense of the sentence will depend on
what arrangement of objects it presents to us. Hence, if a proposition says
anything at all, it says that such-and-such objects are arranged in such-and-
such a way. The only thing we can do with words is to describe, or
misdescribe, the facts.
Thus, according to the Tractatus, ‘One name stands for one thing, another
for another thing, and they are combined with one another. In this way the
whole group—like a tableau vivant—presents a state of affairs’ (TLP
4.0311). This is known as the picture theory of meaning. Words are combined
in a sentence to form a picture or model of a possible state of affairs in the
world. If the way that things are arranged corresponds to the way the words
are combined, then the sentence is true; and if not, then it is false.
In his own Preface to the Tractatus, Wittgenstein said that ‘the whole sense of
the book might be summed up in the following words: what can be said at all can
be said clearly, and what we cannot talk about we must pass over in silence’
(TLP, p. 3). So far, I have commented on ‘what can be said’, as Wittgenstein
himself did in the larger part of the Tractatus. But by doing so, I have broken the
very rules which fix the limits of what can be said. For as soon as I try to explain
how a sentence must be related to the state of affairs it represents, I try to do
more with words than merely describe the facts (TLP 4.12).
This implication of Wittgenstein’s doctrine, that philosophical
propositions are themselves nonsensical, did not escape him. He states it
explicitly at the end of the Tractatus:

The correct method in philosophy would really be the following: to say

nothing except what can be said…and then, whenever someone else
wanted to say something metaphysical, to demonstrate to him that he
had failed to give a meaning to certain signs in his propositions…
My propositions serve as elucidations in the following way: anyone
who understands me eventually recognises them as nonsensical, when he
has used them—as steps—to climb up beyond them. (He must, so to
speak, throw away the ladder after he has climbed up it.)
(TLP 6.53–4)

But it is not just philosophy that lies beyond the reach of language. Ethics,
aesthetics, and whatever thoughts we might aspire to have about the meaning
of life, or about God, all belong to what Wittgenstein calls ‘the mystical’; and
they are alike incapable of being put into words. Nothing which touches on
The gospel according to Wittgenstein 3

matters of value can be captured in words. Every human effort to address or

even to articulate what Wittgenstein called ‘the problems of life’ must be in
vain: ‘When the answer cannot be put into words,’ he says, ‘neither can the
question be put into words’ (TLP 6.5).
If we accept Wittgenstein’s austere conception of language, and its
consequence that the ethical, aesthetic and religious aspects of human life—
which he calls ‘the mystical’—cannot be put into words, we may feel tempted
to conclude that the importance we attach to aesthetic experience, to ethics
and to religion is the result of an illusion. Alternatively, we may conclude
that what can be put into words is paltry by comparison with what cannot.
There is no doubt that Wittgenstein intends us to draw the latter conclusion.
In fact, in a letter written in 1919 to a prospective publisher, Wittgenstein
says that the Tractatus ‘consists of two parts: of that which is under
consideration here and of all that I have not written. And it is precisely this
second part that is the important one.’ 1 This cannot have been an
encouragement; but it is a telling remark.
Still, it is difficult to know what place, if any, God and faith have in the
system of the Tractatus. God is mentioned four times, but only the last of
these comments has anything to do with religion. It is this: ‘How things are
in the world is of complete indifference for what is higher. God does not reveal
himself in the world’ (TLP 6.432). The emphasis on the word ‘in’ is
Wittgenstein’s; and perhaps its significance is explained in the next remark
but one: ‘It is not how things are in the world that is mystical, but that it
exists’ (TLP 6.44). Thus, Wittgenstein may have wanted to intimate that God
reveals himself in the fact that the world exists, the fact that ‘there is what
there is’ (NB, p. 86)—though we must not forget that this is not strictly
speaking a fact at all, and is therefore impossible to state.
We should not imagine that this is meant to be an argument for God’s
existence. It would be a strange argument indeed, if this had been what
Wittgenstein intended—one with a nonsensical premise and a nonsensical
conclusion. What may be intended, however, is that a religious attitude is an
attitude towards the world as a whole, an attitude in which it isn’t how things
happen to be in the world that absorbs our attention, but that it exists. And
a religious attitude can also be described as, in some sense, an
acknowledgement of God, although of course it is an attitude which we must
never attempt to articulate by saying that God exists: ‘What we cannot speak
about we must pass over in silence’ (TLP 7).
This is what the Tractatus hints at; but a few remarks in Wittgenstein’s
notebooks, which he wrote in July 1916, when the Austrian Army in which
he was serving was retreating across the Carpathian mountains and his life
was in constant danger, are more explicit. They identify God with ‘the
meaning of life, i.e. the meaning of the world’, with fate and with the world
itself (NB, pp. 73f.). However, the impression they convey most forcibly is
that faith consists in the ability to see that life has a meaning; that this in turn
consists in living in such a way ‘that life stops being problematic’, for ‘the
4 John Hyman

solution to the problem of life is to be seen in the disappearance of this

problem’ (NB, p. 74; cf. TLP 6.521); and that living thus will enable one to
achieve a sort of happiness—something perhaps akin to a Stoic calm—by
detaching oneself from the uncontrollable contingencies of the world, and
accepting it without fear. Wittgenstein incorporated some of these remarks
into the Tractatus; and it seems that the rest continued to exert an influence
on his thought, and remain, albeit with an altered emphasis, in the
background of the Tractatus.
The Tractatus presents an austere view of human language, even a
repressive one, for it denies the intelligibility of much of what we say,
including everything which mattered most to Wittgenstein himself. This
doctrine was the result of a brilliant, profound and subversive critique of
Frege’s and Russell’s philosophy of logic. But the doctrine that religious truths
are ineffable has an important place in the history of religious thought, and it
is likely to appear plausible if one thinks that language cannot capture our
profoundest feelings. Wittgenstein’s upbringing led him to revere musical
creativity, and it is possible that his love of music made him receptive to this
thought. At any rate, the achievement of his early philosophy, so far as the
philosophy of religion is concerned, was to have formulated the doctrine that
religious truths are ineffable in terms which are clear and explicit precisely
because they are founded on a theory of language.

In the 1930s, Wittgenstein’s philosophy of language was dramatically
transformed, and his earlier view of religion could not survive the
transformation. He abandoned the doctrines that a proposition is a logical
picture compounded out of names whose meanings are the things they stand
for, and that the intelligible use of language always serves a single purpose—
to describe the facts. He came to believe—on the contrary—that the meaning
of a word is its use in the language; that words can be used for an indefinitely
broad and heterogeneous range of purposes; and hence that the task of
philosophy is not logical, but hermeneutical. Philosophy, he now contends,
does not consist in logical analysis, but in the description of our various
The term ‘language-game’ has excited controversy and caused some
puzzlement; but a language-game is simply a human activity involving speech
or writing, in which a distinctive range of concepts is employed. The word
‘game’ is there to remind us of three things: first, that these activities are
guided and constrained by the rules we enunciate when we explain the
meanings of words; second, that they are extremely varied and are not
usefully seen as elaborations of a single theme, such as communicating
information or producing beliefs; and third, that they take place and have
significance only in the context of human forms of life and culture.
So, when the later Wittgenstein writes about religious belief, he continues
The gospel according to Wittgenstein 5

to argue that the use of language to express religious beliefs is quite unlike
the use of language to state facts; but he no longer infers that it must therefore
be a misuse of language. His principal aims are to explain how concepts such
as sin, redemption, judgement, grace and atonement can have an
indispensable place in an individual’s or a community’s way of life, and to
show how we can resist assimilating the use of these concepts to hypotheses,
predictions and theoretical explanations.
For example, he talks at length about the belief that there will be a last
judgement. His stated intention is to show that ‘in religious discourse we use
such expressions as: “I believe that so and so will happen,”…differently to
the way in which we use them in science’ (LC, p. 57). But he argues in favour
of a far more surprising and radical doctrine, namely that believing in a Last
Judgement does not mean thinking it certain or probable that a certain kind
of event will occur sometime in the future. This does not merely distance the
use of an expression like ‘I believe that so and so will happen’ in religious
discourse from its use in science; it distinguishes it from any kind of prediction
at all.
So, if the expression ‘I believe that there will be a Last Judgement’ is not
used to make a prediction, how is it used? This is Wittgenstein’s answer:

Here believing obviously plays much more this role: suppose we said that
a certain picture might play the role of constantly admonishing me, or I
always think of it. Here there would be an enormous difference between
those people for whom the picture is constantly in the foreground, and
others who just didn’t use it at all.
(LC, p. 56)

This example is not atypical. Wittgenstein equates having religious beliefs

with using religious concepts and having the attitudes and emotions that their
use implies. He says this most explicitly as follows: ‘It strikes me that a
religious belief could only be something like a passionate commitment to a
system of reference’ (CV, p. 64). And he says this about coming to believe in

Life can educate one to a belief in God. And experiences too are what
bring this about; but I don’t mean visions and other forms of sense
experience which show us ‘the existence of this being’, but, e.g., sufferings
of various sorts. These neither show us God in the way a sense-
impression shows us an object, nor do they give rise to conjectures about
him. Experiences, thoughts,—life can force this concept on us.
(CV, p. 86)

Now, it was one of the fundamental themes of Wittgenstein’s later philosophy

that the concepts we use cannot be justified by reference to reality. Some
philosophers have argued that our concepts, or at any rate the concepts of
6 John Hyman

science, are correct, that they conform to the nature of the things we use
them to describe. For example, our language prevents us from saying that
colours have a pitch or that musical notes are coloured because red cannot be
a semitone higher than blue and a musical note cannot be visibly orange.
Wittgenstein argues, to the contrary, that our network of concepts, which he
calls ‘grammar’, cannot either conflict or accord with the facts. For what we
say conflicts with the facts if it is false and accords with the facts if it is true.
And although the concepts we use may be well- or ill-suited to the business of
stating particular sorts of truths, the truths and falsehoods that we state are
not themselves concepts, any more than the sentences we utter are predicates.
Grammar itself is therefore arbitrary, i.e. not accountable to any reality. (This
does not mean unimportant, capricious or readily alterable.) A system of
measurement, for example, is not correct or incorrect in the way that a
statement of length is; although of course some systems are more useful,
convenient and easy to understand and apply than others. (This doctrine is,
in fact, a constant feature of Wittgenstein’s philosophy, early and late. The
first entry in his Notebooks (p. 1) is: ‘Logic must take care of itself’.)
So, Wittgenstein held that ‘a religious belief could only be something like
a passionate commitment to a system of reference’—that is, a passionate
commitment to the use of certain concepts. And he also held that a ‘system of
reference’ cannot be verified or justified by comparing it with the facts. For
example, the metric system cannot be verified or falsified, although of course
my belief that the earth moves at 30 metres per second can. But if a religious
belief is something like a passionate commitment to a system of reference—
as opposed to a passionate commitment to the truth of an empirical
proposition—then a religious belief cannot be true or false. And Wittgenstein
held that religious beliefs cannot be reasonable or unreasonable either, if that
means that they can or cannot be justified:

I would say, they are certainly not reasonable, that’s obvious.

‘Unreasonable’ implies, with everyone, rebuke. I want to say: they don’t
treat this as a matter of reasonability. Anyone who reads the Epistles will
find it said: not only that it is not reasonable, but that it is folly. Not only
is it not reasonable, but it doesn’t pretend to be.
(LC, p. 58)

The people Wittgenstein regards as unreasonable are apologists for, or

against, religion who make the assumption—Wittgenstein calls it ‘ludicrous’
(LC, p. 58)—that religious beliefs can be corroborated or falsified by
evidence. But unless religion is confused in this way with something quite
different, it is not unreasonable. ‘Why shouldn’t one form of life culminate in
an utterance of belief in a Last Judgement?’ Wittgenstein asks rhetorically
(LC, p. 58). A passionate commitment to a system of reference, as Jacques
Bouveresse put it, ‘bears a closer resemblance to an affair of the heart than to
a rational matter’.2 This should not be interpreted as meaning that faith is
The gospel according to Wittgenstein 7

something which people tend to be irrational about; but rather, that it is

something which reason has no bearing on. And this immunity from rational
criticism extends even to Christian beliefs about Jesus: ‘The historical
accounts in the Gospels might, historically speaking, be demonstrably false
and yet belief would lose nothing by this…because historical proof (the
historical proof-game) is irrelevant to belief’. (CV, p. 32).
In sum, Wittgenstein defends two principal doctrines: first, a doctrine
about the meaning of religious discourse, and second, a doctrine about the
epistemology of religious beliefs. The first doctrine is that the expression of a
religious belief in words is not a prediction or a hypothesis, but instead
expresses ‘something like a passionate commitment to a system of reference’.
And the second is that religious beliefs are therefore immune from
falsification and from verification. Compare a superstition, for example, that
an amulet will protect the person who wears it from disease. This is a
hypothesis, which adequate statistical data will verify or falsify. Critics and
apologists who mistake religious beliefs for hypotheses and muster evidence
in their favour or against them confuse religious faith and superstition,
Wittgenstein says (CV, p. 72).

Is Wittgenstein’s account of religious belief convincing? I doubt it. First, is he
right to argue that a religious belief is ‘something like a passionate
commitment to a system of reference’? Surely not. Admittedly, believing that
God exists is very unlike believing a hypothesis in history or in science. And
the differences have to do with the ways in which we may be led to believe
that God exists, and the ways in which this belief will influence our other
beliefs and our feelings, commitments and actions. But it does not follow
that believing that God exists is nothing but ‘a passionate commitment to a
system of reference’—i.e. a commitment to leading a life in which questions
will be asked, obligations will be acknowledged, decisions taken and actions
performed, which can only be explained or understood by the use of religious
concepts. And surely, if I have and retain that commitment, my belief that
God exists will typically be among my reasons for doing so.
Perhaps one can defend Wittgenstein on this point. In his Remarks on
Frazer’s Golden Bough, Wittgenstein comments on the extraordinary custom
of killing the priest-king in his prime, so that his soul will be kept fresh:
‘Where that practice and these views go together,’ Wittgenstein says, ‘the
practice does not spring from the view, but they are both just there’ (RF, p. 2).
But even if this is often true, it does not follow that the ‘view’ is the
commitment to the ‘practice’. No doubt the expression of a religious belief
can often convey a passionate commitment to a way of life—one which
involves a variety of customs, habits, rituals (in short, practices) in which a
range of religious concepts, amounting even to a ‘system of reference’, are
essentially involved. But we should not infer that the passionate commitment
8 John Hyman

is all that the belief signifies, any more than we should infer from the fact
that a moral judgement may convey passionate admiration or intense disgust
that this is all it signifies. This reductive view of moral judgements seemed
appealing to some during the heyday of verificationism; but it does not carry
Is his account of the difference between the proposition that God exists
and existential propositions in science and history convincing? Again, surely
not. He is right, I think, to insist that there are many different kinds of
existential proposition. Belief in God’s existence is not the same sort of
existential belief as the belief that there are infinitely many primes, or the
belief that there is a planet more distant than Uranus. But one good way of
seeing the various differences between these various propositions is to
consider the different ways in which they are proved or supported. For
example, the existence of infinitely many primes is a necessary truth, which
can be proved by a reductio of its contradictory; and the existence of a planet
more distant than Uranus is a contingent truth, which was initially postulated
to explain the orbit of Uranus, and subsequently confirmed by observation.
Furthermore, since evidence and argument are not the exclusive property
of science, Wittgenstein cannot be right to insist that if we try to prove or
support the proposition that God exists we are already trapped in confusion,
because we are treating religion as if it were science. It would, I think, be
foolish to maintain that Anselm and Aquinas were peddling superstitions (cf.
LC, p. 59), or that apostasy cannot be based on reasons. And it is surely a
mistake to claim, as Wittgenstein does, that when we consider God’s
existence, ‘what is here at issue is not the existence of something (dass es sich
hier um eine Existenz nicht handelt)’3 (CV, p. 82). Kierkegaard may have had
something similar in mind when he wrote: ‘God does not exist; he is eternal.’
But the differences between ‘God exists’ and existential propositions in
science or history does nothing to support this idea; and neither does the
doctrine that God cannot begin to exist or cease to exist. If Democritus
believed that atoms cannot begin or cease to exist, it does not follow that he
did not believe that an atom is eine Existenz—an entity, or something which
Is Wittgenstein right to insulate religious beliefs from ‘the historical
proofgame’? I doubt it. It is certainly impossible to insulate religion entirely
from rational criticism: ‘If Christ be not risen, our faith is vain’ implies ‘Either
Christ is risen or our faith is vain’ for exactly the same reason as ‘If the
weather is not fine, our picnic is ruined’ implies ‘Either the weather is fine or
our picnic is ruined’. But if religious beliefs and systems of religious beliefs
are not invulnerable to logic, why should they be cocooned from other sorts
of rational scrutiny?
Finally, is Wittgenstein right to insist that religious faith is not so much a
matter of assenting to a series of doctrines as cleaving to a form of life? I
think he is. But nothing in his later philosophy of language, and in particular
no part of his doctrine about the relation between language and forms of life,
The gospel according to Wittgenstein 9

implies that a form of life cannot involve historical or metaphysical beliefs

(such as that Jesus rose from the dead or that the soul is immortal) as well as
concepts and attitudes: all of them—beliefs, concepts and attitudes—in a
mutually supporting relation. Nor does it imply that the beliefs which form
part of the core of a form of life cannot include false or incoherent ones.
Christians have traditionally believed a number of well-defined historical
propositions and rather ill-defined metaphysical ones, which Christian
philosophers have sought to formulate precisely. This is not to say that these
formulations—such as Aquinas’ teaching on the soul or the Eucharist—tell
us exactly what Christians have traditionally believed: on the contrary, a
precise formulation cannot exactly capture something imprecise, precisely
because of its precision. However, it can help us to settle whether the ill-
defined belief is paradoxical or implicitly contradictory. How can it do this?
Wittgenstein himself suggests a plausible answer to this question in the
Philosophical Grammar:

But if we wish to draw boundaries in the use of a word, in order to clear

up philosophical paradoxes, then alongside the actual picture of the use
(in which as it were the different colours flow into one another without
sharp boundaries) we may put another picture which is in certain ways
like the first but is built up of colours with clear boundaries between
(PG, p. 76)

Wittgenstein does not maintain explicitly that religious beliefs, unlike the
formulations of them offered by philosophers, cannot be paradoxical. But
several of his remarks point in this direction;4 and some of his followers have
not been afraid to claim this directly. For example, D.Z.Phillips writes: ‘If the
notion of an inner substance called ‘the soul’ is the philosophical chimera we
have suggested it is, whatever is meant by the immortality of the soul cannot
be the continued existence of such a substance’.5 Whether or not Wittgenstein
would have agreed with this inference, it is surely quite unconvincing, because
there is no reason why it should be impossible to espouse, seriously and
sincerely, doctrines which are demonstrably incoherent. In fact, since every
philosopher has managed it at one time or another, and since not every
philosopher is dishonest or insincere, it must be possible. The trick,
presumably, is to avoid explicit contradictions; but whilst it may sometimes
take a philosopher’s ingenuity to do this, when challenged or cross-examined,
a like-minded community with a reassuring intellectual elite seems likely to
make it as easy as falling off a log, even for the man Wittgenstein calls ‘an
honest religious thinker’ (CV, p. 73). If the immortality of the soul is a
contradictory doctrine, it does not follow that one cannot believe in it.
One important source of support for incoherent beliefs is the use of
figurative language. It is a familiar fact that analogies can explain; but
sometimes the explanations are bogus. There is a nice example which crops
10 John Hyman

up from time to time in the philosophical literature: an explanation of

wireless. It begins with an explanation of wire. Wire is like a dog that is so
long that when you pull its tail in Boston it barks in New York. And now for
wireless. Wireless is just the same, but without the dog. Figurative language,
like analogy, has its authentic and its bogus uses; and its principal bogus use
is to make us imagine that we understand something. That is why Aristotle,
who argued famously that metaphor is the main part of excellence in diction,
also insisted, in the Posterior Analytics, that one should never attempt to
define by means of metaphor.6
Are history and philosophy therefore capable of demonstrating that
traditional Christianity is worthless, or that ‘people do all of this out of sheer
stupidity’? (RF, p. 1). Not at all; for if the Gospels are demonstrably false and
no coherent formulation of the central Christian doctrines is possible, it does
not follow that Christianity is either stupid or worthless. The Stoic doctrine
of preferred indifferents may be incoherent; the doctrine that forms are
material certainly is. But it would be a gross mistake to infer that Stoicism
was a stupid or worthless system. Incoherence is a defect in a system of
religious beliefs, because if one can see that a doctrine is incoherent, that is a
compelling reason to disbelieve it.7 But philosophers, who have a professional
interest in coherence and consistency, are prone to exaggerate their

Wittgenstein’s influence in the philosophy of religion is due to scattered
remarks, marginalia and students’ notes. He never intended to publish any
material on the subject, and never wrote about it systematically. Nevertheless,
it is possible to glean a moderately clear picture of his views about the nature
and justification of religious belief.
His semantic doctrine is an ingenious application of a powerful strategy,
which consists in making ‘a radical break with the idea [which lay at the
heart of the Tractatus] that language always functions in one way, always
serves the same purpose: to convey thoughts—which may be about houses,
pains, good and evil, or anything else you please’ (PI§304). This strategy
produced radical and fruitful ideas in the philosophy of mind and the
philosophy of mathematics. Its application in the philosophy of religion has
a kinship with both, but is less successful than either.
The epistemological corollary, that religious beliefs are immune from
rational criticism and incapable of receiving rational support, has the
interesting consequence that, as Wittgenstein said, ‘if Christianity is the truth
then all the philosophy that is written about it is false’ (CV, p. 83), but it has
little else to recommend it. Wittgenstein once described his work as ‘one of
the heirs of what used to be called philosophy’ (BB, p. 28); and it may be
possible to regard a ‘passionate commitment to a system of reference’ not as
religious belief—certainly not Christian religious belief—but as one of the
The gospel according to Wittgenstein 11

heirs of what used to be called Christian religious belief. Whether it could

enjoy as vigorous a life as its ancestor is a sociological and perhaps a
psychological question, which I shall not address.

1 Wittgenstein: Sources and Perspectives, ed. C.G.Luckhardt (Hassocks: Harvester,
1979), p. 94.
2 J.Bouveresse, Wittgenstein Reads Freud: The Myth of the Unconscious
(Princeton, NJ: Princeton University Press, 1995), p. 19.
3 I have followed Winch’s translation; but ‘what is here at issue is not an entity’
may be preferable.
4 See, for example, CV, pp. 22, 32, 73, 85; but cf. CV, p. 1.
5 D.Z.Phillips, ‘Dislocating the Soul’, in D.Z.Phillips (ed.), Can Religion Be
Explained Away? (London: St. Martin’s Press, 1996), p. 237.
6 Poetics 1459a5; Posterior Analytics 97b37f.
7 There is a strand in Christian philosophy which denies this. Indeed, Tertullian
sometimes held that the absurdity of a doctrine is a reason for believing it. See
J.N.D.Kelly, Early Christian Doctrines, 5th edn (London: A. & C.Black, 1977),
p. 152.
2 Wittgenstein and magic
Brian R.Clack

The story of Wittgenstein’s presence in contemporary philosophy of religion

is a peculiar and in many respects a tragic one. This is in part due to the late
appearance of his own writings on religion. While none of these appeared
until 1966, the perceived character of the ‘Wittgensteinian interpretation of
religion’ had by then already been established by virtue of the work of certain
of his followers, who carefully applied leitmotifs of the Philosophical
Investigations so as to illuminate problems of religious language. The result
was that Wittgenstein’s view of religion appeared to be known even before
many had heard what he himself had written, and (here’s the tragedy) that
view was habitually understood in terms of ‘fideism’: religion was a
‘language-game’, a ‘form of life’ neither requiring justification nor susceptible
to criticism or explanation. Absence of such notions in Wittgenstein’s own
considerations of religion seemed to make little difference to these
characterisations. The die had been cast. Whenever Wittgenstein’s name was
mentioned with regard to religion, the talk was always to be of how religious
discourse has its own criteria of meaningfulness, truth and falsity, and it was
hard not to think that his approach was specifically designed to protect
religion from positivist censure.1 I wish to make no comment on the validity
of fideistic devices, but mention this for the sole purpose of highlighting a
regrettable feature of the way Wittgenstein’s thoughts on religion have been
addressed and appropriated, namely that readers find in Wittgenstein’s
writings what they expect to find there rather than what is actually there.
And what is actually there is in fact infinitely more interesting, complex and
challenging than the occasionally banal glosses which are now so familiar.
A similar fate awaited the Remarks on Frazer’s Golden Bough,
Wittgenstein’s most intriguingly pregnant of religious musings. Prejudices
about what Wittgenstein ‘must’ be saying, compounded by only the most
cursory of readings of the Remarks, led to premature and misleading
conclusions, conclusions which have nonetheless assumed the status of
orthodoxy. The orthodox view is this: Wittgenstein, appalled by James Frazer’s
depiction of magico-religious beliefs as mistaken hypotheses, and of rituals as

Wittgenstein and magic 13

futile attempts to achieve empirical ends, sought to overturn this conception of

magic/religion as mistaken science by emphasising the fundamentally
‘expressive’ nature of ritual and belief. Rather than the attempt to describe
(and subsequently manipulate) supernatural forces, magic should be seen as a
way of expressing feelings and attitudes, and is in no manner mistaken or
erroneous. This much is well known. And yet it is a striking fact that the
Remarks on Frazer itself does not yield these conclusions, lending only the most
slender of supports to the expressive interpretation of ritual. As with all of
Wittgenstein’s thoughts on religion, the picture is much less obvious. In this
essay, therefore, I want to explore the roots of the misunderstanding of
Wittgenstein’s view of magic and primitive religion. A good place to start is
with the nature of the expressivist thesis itself.


The expressivist interpretation of religious belief and practice springs from a
violent rejection of the intellectualist anthropology of the Victorian period.
Frazer’s variety of intellectualism, as elaborated in The Golden Bough and
elsewhere, consists of two central contentions: that magic and religion emerged
as explanatory theories of the natural world and its workings; and that these
theoretical systems gave rise to rituals, conceived as instrumental actions,
attempts to influence the course of natural events. Because the theories on
which the rituals rest are erroneous, the rituals themselves are abortive. This
negative conception of ritual is set into a broadly Hegelian account of history,
whereby human history is the story of a battle between three competing
philosophies of life: magic, religion and science; a battle which witnesses the
progressive improvement of the human mind, as less worthy theories (magic,
religion) are discarded in favour of something far superior (science).
As Frazer presents it, magic (the earliest of humanity’s philosophies of life)
emerges when primitive human beings attempt to come to terms with the world
around them. Entirely ignorant of the true workings of nature, and therefore at
the mercy of nature’s capricious power, the primitive tries to help himself by
theorising. The primitive, that is, becomes philosopher, and the result of his
philosophising is the production of what Frazer terms ‘theoretical magic’, a
system of belief based upon the positing of two laws: the law of homoeopathy
and the law of contiguity, laws which can be manipulated to produce desired
results. The first of these laws (or principles of thought) states that ‘like
produces like, or that an effect resembles its cause’, while the contention of the
second is that ‘things which have once been in contact with each other continue
to act on each other at a distance after the physical contact has been severed’.2
When applied, these laws produce practical magic, and for our purposes here
we can focus upon the practical form of the first law: homoeopathic magic.
The belief that like produces like is applied by the primitive in his attempts to
bend the will of nature to his own advantage. If, for example, a great drought
afflicts the land, and rain is desperately desired, magic can be employed to
14 Brian R.Clack

effect a downpour: by sprinkling a small amount of water onto the ground, by

roaring like thunder, by striking two flints together to create the impression of
lightning, or the like. In more nefarious manner, the death of an enemy can be
effected by damaging or destroying a small likeness of the intended victim.
Such is the character of homoeopathic magic as Frazer presents it: desired
events are to be produced by acts of imitation. Though such desires are perfectly
reasonable, the methods employed to achieve them suffer from one fatal flaw:
they are completely futile. No amount of effigy-burning can result in the death
of one’s enemy, and the skies will not respond to the pitiful mimicry of rain.
Magic is thus one huge mistake.
It is this dismissive claim to which Wittgenstein objects so strongly, finding
Frazer’s attitude ‘unspiritual’, ‘savage’, and symptomatic of a sensibility
dulled by immersion in science.3 It is precisely Frazer’s (typically modern)
obsession with the aims and methods of the natural sciences which leads him
to see magic rituals as ineffective actions, blunders. But, as Wittgenstein is
quick to remind us, magic will appear to be such only if one insists on treating
it as science. If magic is to be something other than an error, therefore, it
must be something other than science, something other than a reasoned yet
hopeless attempt to influence the course of nature. Several of Wittgenstein’s
remarks against Frazer are thus aimed at undermining the notion that
primitives engage in magical action because they lack adequate technical
expertise. For instance: The same savage who, apparently in order to kill his
enemy, sticks his knife through a picture of him, really does build his hut of
wood and cuts his arrow with skill and not in effigy’ (RF, p. 4).
Wittgenstein’s point is a pertinent one. If Frazer is correct, then surely we
should expect to see the primitive employing homoeopathic techniques in all
activities. We should, for example, see the primitive building in effigy: he
would construct a miniature hut and then look about to see whether a lifesize
one has magically appeared. So much easier than all the hard physical effort
of real construction. But of course we see no such homoeopathic construction
work, and this suggests to Wittgenstein that magic is not a (poor) substitute
for technical proficiency, but something entirely discontinuous with practical,
empirical techniques. And it is when Wittgenstein considers what the
distinctive character of magic might be that his apparently expressivist thesis
makes itself known.
Consider, to begin with, his remarks on the nature of effigy destruction. In
direct contrast to Frazer’s contention that such actions are attempts to kill a
hated person, Wittgenstein writes:

Burning in effigy. Kissing the picture of a loved one. This is obviously not
based on a belief that it will have a definite effect on the object which the
picture represents. It aims at some satisfaction and it achieves it. Or
rather, it does not aim at anything; we act in this way and then feel
(RF, p. 4)
Wittgenstein and magic 15

By placing the exotic rite (effigy burning) alongside a familiar action which
we ourselves might perform (kissing a photograph), Wittgenstein attempts to
undermine the plausibility of an instrumental rationale. If I am away from
the woman I love, I may carry a photograph of her with me. And when I feel
the pain of her absence, I may kiss that photograph, that image of her. But in
kissing her image, I do not believe that she will really feel the touch of my lips
on her face. No, that kiss is simply an expression of my love, or of my desire
that she were here with me. And if my love affair turns sour, if the woman
hurts me, then I may tear her photograph to pieces. But that would not be an
attempt to kill her. In tearing the photograph I would be expressing my anger,
despair and frustration.
The implication of such a reflection on an instance of homoeopathic magic
seems clear. If our own personal ritual actions can be interpreted as expressive
performances, might this not also be true of magic in primitive society?
Primitive magic is not a crude form of science, a hopeless attempt to effect
practical goals, and it is emphatically not a mistake. It is somehow expressive.
The magician who crafts and mutilates an image of his foe is engaged in an
act, not of homicide, but of catharsis; an act whereby deeply felt emotions
are manifested. This cathartic account of ritual thus lays its emphasis on how
we as human beings have within us certain passions, hopes and feelings which
need occasionally to be let out. Ritual functions as just such an emotional
safety-valve, enabling us to bring those feelings out in the open, to bring to
expression our desires and wishes, both individual and collective. This theme
emerges in Wittgenstein’s consideration of rain-making ceremonies:

I read, amongst many similar examples, of a rain-king in Africa to whom

the people appeal for rain when the rainy season comes. But surely this
means that they do not actually think he can make rain, otherwise they
would do it in the dry periods in which the land is a ‘parched and arid
desert’. For if we do assume that it was stupidity that once led the people
to institute this office of Rain King, still they obviously knew from
experience that the rains begin in March, and it would have been the
Rain King’s duty to perform in other periods of the year.
(RF, p. 12)

If it is the case that a rain-maker is consulted only when the rains are due to
begin anyway, only when the storm clouds are amassing on the horizon, and
not in times of drought, then the plausibility of the expressivist case becomes
clear. Instead of an attempt to bring rain, the ceremony serves as a celebration
of the coming of rain; a celebration of the fact that the crops will not wither
in the earth and die, that the streams will not run dry. Continuing the theme,
Wittgenstein remarks, ‘towards morning, when the sun is about to rise,
people celebrate rites of the coming of day, but not at night, for then they
simply burn lamps’ (ibid.). The idea here is the same as in the rain-making
example: magic serves to mark the cycles of nature, rather than to manipulate
16 Brian R.Clack

them. And even were we to point to examples of rain-making ceremonies

carried out in times of extreme drought, the expressivist thesis would not as
a result be scuppered, for such rites should be understood as the expression
of a community’s anxiety-ridden hope for rain.4
Reflection on such instances illustrates the great appeal of the expressivist
interpretation. Faced with ritualised appeals for rain, faced with the
ceremonious defacing of images, it may seem that to lay stress on the release
of (potentially parlous) emotion is a far more economical explanation than
the gratuitous nature of intellectualist elucidations, such as Frazer’s view that
magical and religious rites spring in part ‘from a mistaken theory of the solar
system’.5 And, moreover, expressivism is perceived to avoid the uncharitable
and perhaps even imperialistic tone of Frazerian intellectualism in which the
primitive is described to all intents and purposes as a laughable clown, a
simpleton lost in a world he can neither understand nor control.
Wittgenstein’s remarks on ‘our kinship to those savages’ (RF, p. 10) seem
intended to make that point: must we be branded as fools because we kiss
and deface images? No, so neither should the ‘savage’.
Laudable though such intentions may be, they are not sufficient to
necessitate the embracing of expressivism, for certain unavoidable problems
make this is a deeply implausible interpretation of ritual. Even its most
compelling elements break apart upon greater examination. Although
Wittgenstein’s carefully chosen example of the African rain-king certainly
suggests a celebratory rationale, as a general account of rain-making
ceremonies such an expressivist interpretation is inadequate. Consider just
one example. Frazer documents the case of an Abyssinian king, the Alfai,
who is held to be able to cast down rain, and ‘if he disappoints the people’s
expectation and a great drought arises in the land, the Alfai is stoned to
death’.6 This must suggest that those petitioning the king really do believe
that he has it within his powers to make rain, and that they are therefore not
merely expressing hopes and fears through their appeal. Otherwise, why
would the Alfai’s fate be determined by the persistence of the drought? An
instance such as this provides greater support for an instrumentalist
conception of ritual than for the expressivist alternative, which, though
perhaps applicable to an explication of personal ritual acts (kissing pictures
and the like), fails fully to engage with the seriousness with which ritual is
undertaken in primitive societies.7
Indeed, all arguments for expressivism tend towards weakness. Consider
just two more. First, the expressivist is keen to stress the dichotomy between
emotive and instrumental motivations. The absence of emotional motivations
in certain instrumental actions (constructing a canoe, preparing a meal) might
lead one to conclude that where we find one such motivation we never find
the other. So if, as is undeniably the case, there are emotional motivations in
the destruction of an effigy, in the ritualised appeal for rain, and so on, one
might be tempted to conclude that such actions are entirely non-instrumental,
that such actions do not ‘aim at anything’. But this would be a bizarre line of
Wittgenstein and magic 17

reasoning, and there are really no grounds for concluding that instrumental
and emotive motivations may not be intertwined together in a specific action.
Think, for example, about the use of love charms, something which appears
particularly appropriate in this context: after all, there is nothing more heart-
rendingly emotional than the wish to have a desired person fall in love with
one. For the expressivist, the love charm offers little interpretative difficulty,
for all we need to understand are the wishes and hopes of the person using
the charm or reciting the spell. And it would be erroneous to conclude that
such a person thinks that his spell will have ‘a definite effect’: ‘The description
of a wish is, eo ipso, the description of its fulfilment. And magic does give
representation to a wish; it expresses a wish’ (RF, p. 4). It is undeniably the
case that the use of a love charm expresses the desire of the ritualist (what
else could it express?). But this expressive element does not preclude the
simultaneous presence of an instrumental motive: the expression of a wish is
of course there in the act, but surely there might also be the feeling that by
this performance, by the manipulation of the beloved’s hair, clothing or
picture, she will actually fall under the spell of the ritualist. Is the expressivist
really denying this? Is the expressivist seriously saying that such an action
has never been carried out for reasons other than emotional catharsis?8
Perhaps Wittgenstein is indeed saying this. But his reasons for so doing are
bizarre. It seems to be his view that no one in his or her right mind could
believe that a magical action would have a concrete impact on the world, in
which case magic must be something other than how it is envisaged by Frazer.
This idea emerges most explicitly when Wittgenstein considers this comment
from The Golden Bough:

At a certain stage of early society the king or priest is often thought to be

endowed with supernatural powers or to be an incarnation of a deity,
and consistently with this belief the course of nature is supposed to be
more or less under his control, and he is held responsible for bad weather,
failure of the crops, and similar calamities.9

Wittgenstein responds: ‘It is, of course, not so that the people believe that the
ruler has these powers, and the ruler knows very well that he doesn’t have
them, or can only fail to know it if he is an imbecile or a fool’ (‘RF’, p. 73).
It is remarkable that Wittgenstein does not think for one moment that
people could believe that a ruler has supernatural powers. Within the British
Isles it was, in former times, firmly believed that the king had the power to
cure scrofula (the ‘king’s evil’). Keith Thomas, who records how popular
‘touching for the evil’ was (between May 1682 and April 1683, 8,577
sufferers were touched by Charles II), notes how ‘Charles I’s sacred touch
made Royalist propaganda during the aftermath of the Civil War’.10 One
might indeed want to say of such a case that the king ‘knows very well’ that
he does not possess supernatural powers, and that his actions are cynical and
politically motivated. But it seems implausible to suggest that the people
18 Brian R.Clack

themselves (those who seek to be touched by the king) are performing some
sort of expressive pantomime. And this goes for many of the cases for which
Wittgenstein wants to deny instrumental motivations. If I, living in an age of
enormous technological advance and with a reasonable understanding of the
workings of nature, may nonetheless, out of the desperation of unhappy
infatuation, strangely feel that a magical charm might secure for me the love
of a woman whose attention as yet eludes me, then is it entirely impossible to
imagine that members of less scientifically advanced societies believe in the
effective (and not merely expressive) power of magic?
A further doubt about the plausibility of expressivism might arise when
we consider another of Wittgenstein’s reasons for rejecting the instrumentalist
conception. It is Frazer’s contention that a magical action is essentially a
mistake which seems to raise Wittgenstein’s ire. ‘Frazer’s account of the
magical and religious notions of men is unsatisfactory: it makes these notions
appear as mistakes. Was Augustine mistaken, then, when he called on God
on every page of the Confessions?’ (RF, p. 1). There is something peculiar
about this remark. Wittgenstein seems to be saying that Frazer is wrong
because his theory entails that magical actions and beliefs are mistakes; and
because magical actions and beliefs are not mistakes, they must be other than
instrumental. But is Wittgenstein really suggesting that a belief cannot be
mistaken, that an action cannot be a blunder? If so, we might as well say that
the geocentric view of the universe was not erroneous at all, but was rather
an expressive belief, encapsulating perhaps the deep feeling of the importance
of human life in the scheme of things, but in no way intended as a genuine
theory. The redescription of mistaken beliefs as purely emotive utterances
may reveal an exemplary sense of charity, but it cannot compel us to accept
the expressivist thesis.
Do these worries necessitate a rejection of Wittgenstein’s thinking on
magico-religious practice? Well, perhaps not.


If Wittgenstein were straightforwardly expressivist with regard to religious

belief, one would have grounds for rejecting his claims. But the picture is
much more complex than that. True, one occasionally stumbles across
remarks which are explicitly expressivist. The rain-making example is one,
as is the following, from Culture and Value:

If someone who believes in God looks round and asks ‘Where does
everything I see come from?’, ‘Where does all this come from?’, he is not
craving for a (causal) explanation; and his question gets its point from
being the expression of a certain craving. He is, namely, expressing an
attitude to all explanations.
(CV, p. 85)
Wittgenstein and magic 19

Such a remark does seem to contain within it the seeds of a full-blown expressivist
thesis. But it would be crucially wrong to extrapolate that conclusion from this
comment and those similar to it. The reason is that there are aspects of
Wittgenstein’s thinking which militate strongly against the expressivist thesis.
The first is external to the Remarks on Frazer, while the second arises from his
explicit comments on magical actions. I will address these one at a time.
As I have argued elsewhere,11 it would be bizarre for Wittgenstein to
embrace a theory of religion which is incompatible with his philosophical
project as a whole. For the expressivist thesis flows from a particular (and
widespread) view of the nature, function and cognitivity of language; a view
which can be found in Wittgenstein’s own Tractatus Logico-Philosophicus,
but which he firmly rejected in his later philosophy. One central contention
of the Tractatus is that language has no function other than to depict possible
or actual facts or states of affairs. Wittgenstein consequently claims that ‘the
totality of true propositions is the whole of natural science’ (TLP 4.11);
meaningful language, that is, is restricted to the description of empirical
phenomena. Any use of language other than that which attempts to describe
the facts of the world is seen as senseless, a consequence of this, of course,
being that the respective discourses of aesthetics, ethics, metaphysics and
religion are banished to silence. What then is to be said of such familiar uses
of language? What is happening when someone uses ethical or religious
language? While Wittgenstein’s response to such questions is notoriously
obscure,12 the response given by the positivist philosophers who followed in
the wake of the Tractatus was unambiguous, at least with regard to the
language of morals. In the hands of such philosophers as A.J.Ayer and Charles
Stevenson, moral utterance, devoid of cognitive significance due to its failure
to possess verification conditions, becomes a domain of emotion. For
instance, instead of functioning as a description of the nature of a particular
act, a statement such as ‘murder is wrong’ serves rather to express (on the
part of the speaker) and to evoke (on the part of the listener) a particular
emotion or attitude toward murder (here: an attitude of disapproval).13
What is requisite to note here is how this emotive theory of ethics provided
a model for post-positivist analyses of religion. Though Ayer was happy to
dismiss religion as entirely meaningless, many other writers sought to provide
emotivist analyses of religious belief. The most (in)famous of such analyses
was that offered by R.B.Braithwaite in his paper ‘An Empiricist’s View of the
Nature of Religious Belief’.14 Accepting that the principle of verification rules
out the possibility of religious utterances having the same descriptive
character as the hypotheses and statements of scientific discourse, Braithwaite
instead maintains that religion is to do with emotion, and as a consequence
of this: ‘If religion is essentially concerned with emotion, it is natural to
explain the use of religious assertions on the lines of the original emotive
theory of ethics and to regard them as primarily evincing feelings or
emotions’.15 We need not concern ourselves with the details of Braithwaite’s
(wholly familiar) theory. The point is that it is precisely this analysis of
20 Brian R.Clack

religion, with its dichotomous understanding of language, which underlies

the orthodox expressivist interpretation of the Remarks on Frazer’s Golden
Bough. So when, in The Coherence of Theism, Richard Swinburne
characterises the Wittgensteinian account of religion as an ‘attitude theory’,
he says that in such an approach, ‘religious assertions, including credal
sentences, express intentions to live in certain ways, or express attitudes of
approval for certain patterns of life or do something else other than stating
how things are’.16 Likewise, consider how Michael Banner summarizes the
essence of the controversy between Frazer and Wittgenstein. Condensing it
into one question—a question about the meaning of religious language—
Banner asks: ‘Is religious discourse concerned primarily with describing and
explaining (hence with making statements to be judged true or false), or is it
primarily concerned with expressing and commending a particular attitude
towards the world?’17 While Frazer occupies the former (i.e. intellectualist)
position, Wittgenstein’s position, Banner is sure, has all the marks of an
expressivist understanding: religious discourse does not (does not even try
to) describe and explain, but rather expresses attitudes, emotions and feelings.
The scandal, thus, of unthinkingly attributing such a thesis to Wittgenstein
is that it entirely ignores the way in which his later philosophy (of which the
Remarks on Frazer form a crucial, pivotal, part18) serves to undermine the
Tractarian and positivist assumptions about language which give rise to
expressivism in the first place. Expressivism appears to rest on the notion
that religious discourse does not match up to the essentially descriptive
function of language, and that therefore if the language of religion is to be
anything other than downright nonsensical, its meaning must be ‘emotive’.
So here we are presented with a precise view of the nature of language: besides
tautologies, language is either descriptive (its primary function) or emotive
(a secondary function). Religious discourse is neither descriptive nor
tautologous, so must be emotive, expressive.
There are (at least) two fronts on which Wittgenstein fights this picture.
One of the most memorable achievements of the Philosophical Investigations
is Wittgenstein’s portrayal of the motley character of language: the
Augustinian/Tractarian picture of language is seen, not as inaccurate, but
rather as overly narrow, so that we might say that it is ‘“appropriate, but
only for this narrowly cirmcumscribed region, not for the whole of what you
were claiming to describe”’ (PI§3). Thus, countering the contention that the
purpose of language is to describe states of affairs, he writes: ‘Language is
not defined for us as an arrangement fulfilling a definite purpose. Rather
“language” is for us a name for a collection’ (Z§322). The lesson we take
from this is clear and straightforward: even were we to say that magico-
religious discourse does not appear to be ‘descriptive’, the automatic labelling
of it as ‘expressive’ would be over-hasty. And, on the second front,
Wittgenstein seeks to undermine the homogeneity of the concept of the
‘descriptive’ itself, which is not one thing, and has no general form: ‘Think
how many different kinds of thing are called “description”: description of a
Wittgenstein and magic 21

body’s position by means of co-ordinates; description of a facial expression;

description of a sensation of touch; of a mood’ (PI §24). ‘Description’ thus
denotes ‘a great variety of thing’ (RPP I §981). The importance of this should
not be lost on us. If the expressivist thesis relies on a particular view of
linguistic meaning which is rejected by Wittgenstein; and if it relies on a sharp
distinction between the descriptive and the non-descriptive which is likewise
rejected by him, then it would be bizarre to find such a picture of magic and
religion emerging from the Remarks on Frazer. For Wittgenstein, this plethora
of dichotomous pairs—the descriptive-non-descriptive, cognitive-non-
cognitive, belief-attitude, explanatory-expressive—will be redundant.
One might feel there is something pedantic about this strategy. Regardless
of such thoughts about linguistic meaning, it might be said, Wittgenstein is
patently articulating an expressivist understanding when he writes against
Frazer: rain-making ceremonies are celebratory actions; image burning and
image kissing are respectively expressions of hate and love; and so on. But this
is only apparently the case. One’s feeling that Wittgenstein’s thesis is an
expressivist one can be overturned by reinterpreting those passages that suggest
expressivism. This will be attempted in the following section. But we can also
point to other evidence. For instance, if Wittgenstein is so adamant that magico-
religious actions aim at nothing other than the satisfaction of the celebrations,
why does he so frequently contradict this understanding? Why, that is, are there
so many passages which suggest an instrumental rationale for rituals? Consider
just three remarks which blatantly run counter to expressivism:

People at one time thought it useful to kill a man, sacrifice him to the god
of fertility, in order to produce good crops.
(AWL, p. 33)

Eating and drinking have their dangers, not only for savages but also for
us; nothing more natural than wanting to protect oneself against these.
(RF, p. 6)

When a man laughs too much in our company (or at least in mine), I
half-involuntarily compress my lips, as if I believed I could thereby keep
his closed.
(‘RF’, p. 73)

In each of these cases, ritual acts are explained in terms of the goals they
seek: people sacrifice a man because they seek thereby to secure a good
harvest; protective magic is employed to neutralise culinary danger; and
homoeopathic techniques are used in an attempt to silence an irritating
person. All of this seems pure Frazer. What can we say about this
simultaneous presence of expressivism and instrumentalism within the
Remarks on Frazer? Is Wittgenstein simply confused? Or does he have
another agenda altogether?
22 Brian R.Clack


The key to understanding what Wittgenstein is doing when he writes against

Frazer is perhaps this remark:

When I am furious about something, I sometimes beat the ground or a

tree with my walking stick. But I certainly do not believe that the ground
is to blame or that my beating can help anything. ‘I am venting my anger.’
And all rites are of this kind. Such actions may be called Instinct-actions.
(‘RF’, p. 72)

It would be easy to construe this as still another example of Wittgenstein’s

expressivism: ‘all rites’ serve to vent emotions. And yet that would be to fall
into the trap warned of at the beginning of this essay, namely that of finding
in Wittgenstein’s writings what we expect to find there. Instead of resting
content with presuppositions about what he must be saying, we should look
and see what this remark is actually contending. If we do this, the expressivist
element falls away, and what becomes striking is that term, ‘Instinct-actions’.
Presumably, an instinct-action is one which is somehow spontaneous, which
is the product neither of thought nor of ratiocination. What in the Remarks
on Frazer Wittgenstein calls an ‘instinct-action’, he elsewhere terms a
‘primitive reaction’, the word ‘primitive’ suggesting that such an action is
‘pre-linguistic: that a language-game is based on it, that it is the prototype of
a way of thinking and not the result of thought’ (Z §541). Armed with this
idea, we can find a unity in Wittgenstein’s reflections on magical practice, a
unity which initially seemed compromised by his apparent wavering between
expressivist and instrumentalist understandings.
In the context of primitive reactions and instinct-actions, consider again
the effigy-burning example. This apparently paradigmatic case of an
expressive action turns out to be suggestive of something else altogether.
True, if we read the remark in one way we do indeed reach the orthodox
conclusion. Witness: This is obviously not based on a belief that it will have
a definite effect on the object which the picture represents.’ If we lay the
emphasis thus, we do appear to arrive at emotivism: I kiss the picture of my
beloved with no external aim in view; my action achieves some personal
satisfaction. But now consider a variant emphasis: ‘This is obviously not
based on a belief that it will have a definite effect on the object which the
picture represents.’ This now suggests something entirely different. The
action does not spring from any reasoning about the connection between
the picture and the person pictured; indeed, it springs from no reasoning
whatsoever. The action is spontaneous, instinctive: I naturally want to kiss
the photograph of the woman I adore. I neither anticipate that by so doing
she will feel my kiss, nor that I will achieve satisfaction from the act. I act.
And note that my instinctive kissing of the photograph may well be
consistent with some strange feeling that it has some kind of effect (‘I forgot
Wittgenstein and magic 23

to kiss her picture last night: how she will resent me!’). Wittgenstein’s point
is simply that the action is not grounded in such a thought. It has no
grounds: ‘we act in this way’. Likewise, Wittgenstein’s example of
compressing one’s lips in an attempt to stop someone from laughing is
subject to precisely the same analysis. It is not about the greater validity of,
in this instance, an instrumental rationale, but is, rather, an example
employed to illustrate the non-ratiocinative nature of ritual acts. Recall
what he says: ‘I half-involuntarily compress my lips, as if I believed I could
thereby keep his closed.’
Viewed in this light, the Remarks on Frazer constitute a far more
interesting argument than if we present Wittgenstein’s thoughts in the
orthodox way, that is, as an expressivist attempt to overturn instrumentalist
conceptions of ritual belief. For Wittgenstein is not denying that people often
think that their ritual actions have an effect. And yet this should not be
construed as a full capitulation to Frazer. Far from it. Wittgenstein is
implacably opposed to intellectualism—not because it contains an
instrumentalist account of ritual, but because it is overly rationalistic, because
it contends that all rituals are the product of reasoning. To see the kind of
approach that Wittgenstein rejects, note this striking passage from Mircea
Eliade’s Patterns in Comparative Religion:

When a sorceress burns a wax doll containing a lock of her victim’s hair
she does not have in mind the entire theory underlying that bit of magic—
but this does not affect our understanding of sympathetic magic. What
does matter to our understanding is to know that such an action could
only have happened after people had satisfied themselves by experiment,
or established theoretically, that nails, hair, or anything a person has
worn preserve an intimate relation with their owner even when separated
from him.19

Wittgenstein certainly opposes the overly intellectualistic nature of this view,

for what did anyone establish theoretically before human beings began
kissing images of their loved ones? Nothing, obviously. Such actions are
performed naturally, instinctively; they are emphatically not the product of a
sustained and elaborate process of ratiocination.
That this is a theme uppermost in the Remarks on Frazer can be seen in
two passages which seek to stress just how much ritual actions are not
grounded in theory:

When he [Frazer] explains to us, for example, that the king must be killed
in his prime because, according to the notions of the savages, his soul
would not be kept fresh otherwise, we can only say: where that practice
and these views go together, the practice does not spring from the view,
but both of them are there.
(RF, pp. 1–2)
24 Brian R.Clack

The characteristic feature of primitive man, I believe, is that he does not

act from opinions he holds about things (as Frazer thinks).
(RF, p. 12)

We should not be misled by Wittgenstein’s reference to ‘primitive’ man here

(as though he were intent on effecting some distinction between the ‘savage’
and the ‘civilized’ mind). For ‘primitive man’ we should rather read: ‘man in
his ritual mode’. In which case, Wittgenstein’s argument in the Remarks on
Frazer is one intended to suggest that religious and magical practice are not
the outcome of some theoretical and explanatory endeavour on the part of
human beings, but are instead actions which spring from our very nature. He
is, perhaps, making a claim about the natural religiousness of human beings:
‘One could almost say that man is a ceremonial animal’ (‘RF’, p. 67).
It would, however, be unsatisfactory to think that Wittgenstein is
contending that such a priority of instinct over thought is peculiar only to
ritual belief and practice. On the contrary, it is definitive of his later
philosophy as a whole that he seeks to undermine an overly rationalistic
conception of human agency, and indeed of the human person per se.
Wittgenstein is keen to stress the animality of the human person and the
central role of instinct in human life and in concept formation. Hence: ‘I want
to regard man here as an animal; as a primitive being to which one grants
instinct but not ratiocination’. (OC§475).
This idea is most famously applied by Wittgenstein in his analysis of pain-
behaviour, when he considers the nature of pain-language, particularly first-
person expressions of pain (such as ‘I have a headache’, ‘I am in pain’), and
the relation of this language to the behaviour typical of being in pain
(groaning, wincing, screaming). One might be tempted to think of this
relation as one of description, so that the sentence ‘I am in pain’ serves to
describe an inner state: I observe that I am experiencing a certain sensation,
and identify this sensation as one of pain; I then use language to inform others
that I am experiencing this sensation. Wittgenstein wants to say that the
language of pain neither originates nor primarily functions in such a fashion.

Words are connected with the primitive, the natural, expressions of the
sensation and used in their place. A child has hurt himself and he cries;
and then adults talk to him and teach him exclamations and, later,
sentences. They teach the child new pain-behaviour.
(PI §244)

The word ‘pain’ is learned, then, not as a description of certain sensations

and behaviour, but, rather, as a replacement for the non-verbal expression of
pain. The language of pain refines the behaviour. In similar fashion, the
sentence ‘He is in pain’ is connected with behaviour, in this instance with
pitying, helping, worrying. And this helping of which the language is a
Wittgenstein and magic 25

refinement is itself instinctual, is itself a primitive reaction. We do not (at

least usually) infer from someone’s behaviour that that person is in pain (‘I
wince like that when my body hurts, so she too must be experiencing a similar
pain’). No, reason does not come into play on such occasions: ‘It is a help
here to remember that it is a primitive reaction to tend, to treat, the part that
hurts when someone else is in pain; and not merely when oneself is’ (Z §540).
This digression into Wittgenstein’s account of pain-behaviour is intended
simply to indicate that there is a structural similarity between this aspect of
our lives and the nature of magico-religious belief and practice, a similarity
which can be summed up by his slogan: ‘In the beginning was the deed’ (CV,
p. 31). In both cases, Wittgenstein is keen to highlight the natural
responsiveness of human beings: in the first case, to the suffering of others, in
the second, to features of the natural world, features which provoke the
‘ceremonial’ aspect of the human being. Witness:

That a man’s shadow, which looks like a man, or that his mirror image,
or that rain, thunderstorms, the phases of the moon, the change of the
seasons, the likenesses and differences of animals to one another and to
human beings, the phenomena of death, of birth and of sexual life, in
short everything a man perceives year in, year out around him, connected
together in any variety of ways—that all this should play a part in his
thinking (his philosophy) and his practices, is obvious, or in other words
this is what we really know and find interesting.
(RF, p. 6)

We are that kind of creature which responds to the world in a ritual way.
Magic and religion thus emerge when the world hits us, when its dramatic
elements draw out beliefs and practices from our ceremonial nature. But to
repeat: this is not because human beings once were struck by certain
phenomena which they then sought (pathetically) to explain. Veneration of
objects is not based on a fearful desire to appease or control, but arises
because human life is intimately tied up with certain phenomena which
become the existential parameters of our lives: birth, sex, love, death, the
natural environment, and so on. As Wittgenstein remarks, once again
emphasising the unratiocinated nature of religion:

It was not a trivial reason, for really there can have been no reason, that
prompted certain races of mankind to venerate the oak tree, but only the
fact that they and the oak were united in a community of life, and
therefore it was not by choice that they arose together, but rather like the
flea and the dog. (If fleas developed a rite, it would be based on the dog.)
(‘RF’, pp. 72–3)

In other words, if we wish to know why it is that we engage in certain ritual

actions, or why it is that we have certain beliefs, Wittgenstein’s response is to
26 Brian R.Clack

point to the conditions of our existence in this world, and to say, quite simply:
‘Human life is like that’ (RF, p. 3).
A short essay such as this is not sufficient to do justice to the multi-layered
and evocative nature of the Remarks on Frazer’s Golden Bough. All I have
tried here to indicate is that it distorts the character of Wittgenstein’s thinking
on magic to see his thoughts simply as a contribution to a standard
philosophical-theological debate about whether religious language is concerned
either with describing and explaining supernatural realities, or else with
expressing attitudes and feelings towards the world. Rather, Wittgenstein is
attempting to say something about human beings, something mediated through
the contention that ritual life is not the product of (mistaken) theories about
the world and its happenings, but is, rather, definitive of us as a species, a
natural manifestation of our character. As such, with its emphasis on deed over
deliberation, the Remarks on Frazer form an early and crucial part of
Wittgenstein’s late attack on the calculating, rational, Cartesian self. Just as it
is ‘humiliating to have to appear like an empty tube which is simply inflated by
a mind’ (CV, p. 11), so it is debasing and implausible for human beings to be
presented as merely reasoning entities, whose actions are consistently the result
of deliberation and calculation. Human life is more intense and colourful than
that: it is ruled by passion, by instinct, by motivations we can barely grasp. As
a result, our life here is strange and disconcerting. Hence why Wittgenstein’s
reflections on magical practice ultimately reach their bedrock with haunting
thoughts about ‘man and his past,…the strangeness of what I see in myself and
in others, what I have seen and have heard’ (RF, p. 18).
Part of the reason why neither expressivism nor instrumentalism will do is
simply because each makes magic appear to be something which can, in the
last analysis, be understood, either as an attempt to manipulate the world, or
as the expression of perfectly understandable feelings. And yet the deepest of
feelings about magic is surely that we cannot understand it. What can we
really make of the fact that we habitually venerate certain images (and
destroy others), that we hold certain days more important than others, that
we decorate our houses with greenery at festive times? Such things are not
obviously transparent, and when we reflect upon the more sinister elements
of our ritual history (human sacrifice, ritual mutilation, cannibalism), who is
to say why such things have occurred? That is to say, our intimacy with ritual,
our appreciation of its naturalness, lead to darker and disquieting thoughts
about our nature as a species, thoughts which will not be stifled by comforting
words about the expression of emotion or the mistaken hypotheses of early
humanity before scientific liberation. It is this sense of the strangeness of
human life which results from a reading of Wittgenstein’s notes on The
Golden Bough. An example from his ‘Lectures on Aesthetics’ perhaps sums
up this sense best: ‘Something hidden, uncanny. Cf.Keller’s two children
putting a live fly in the head of a doll, burying the doll and then running
away. (Why do we do this sort of thing? This is the sort of thing we do do.)’
(LC, p. 25).
Wittgenstein and magic 27

1 See, for example, the passage where John Hick speaks of Wittgenstein’s
‘language-game theory’ of religion: ‘The internal transactions constituting a given
language-game are thus invulnerable to criticism from outside that particular
complex of life and language—from which it follows that religious utterances
are immune to scientific and other nonreligious comment’ (J.Hick, Philosophy
of Religion, Englewood Cliffs, NJ: Prentice-Hall, 1990, p. 97).
2 J.G.Frazer, The Golden Bough (London: Macmillan, 1922), p. 11.
3 See Wittgenstein’s personal attack on Frazer’s character (RF, p. 8), and compare
with the following remark from Culture and Value: ‘Man has to awaken to
wonder—and so perhaps do peoples. Science is a way of sending him to sleep
again’ (CV, p. 5).
4 Compare with John Beattie’s analysis of such customs: ‘the man who consults a
rainmaker, and the rainmaker who carries out a rain-making ceremony, are
stating something; they are asserting symbolically the importance they attach to
rain and their earnest desire that it shall fall when it is required’ (J.H.M.Beattie,
Other Cultures, London: Routledge & Kegan Paul, 1966, p. 203).
5 Frazer, op. cit., p. 79.
6 Frazer, ibid., p. 107.
7 One strategy for protecting the plausibility of expressivism against such a
criticism would be to adopt Beattie’s device of saying that ‘[a]lthough magic is
magic because it is essentially expressive and symbolic, the people who use it
think of it as instrumental’ (Beattie, op. cit., p. 212). Whatever the merits of such
a strategy, it runs counter to a Wittgensteinian analysis and therefore cannot
ameliorate the deficiencies of the latter. For, remember, Wittgenstein appears,
contra Beattie, to be saying that rituals do not ‘aim at anything’; while
Wittgensteinian perspectives on social life notoriously deny any strategy (such as
that suggested by Beattie) of distinguishing between ‘what believers think they
are doing’ and ‘what they are really doing but are unaware of. (On this matter,
see D.Z.Phillips, Religion Without Explanation (Oxford: Blackwell, 1976), esp.
pp. 56–83, and Peter Winch, The Idea of a Social Science (London: Routledge &
Kegan Paul, 1958).)
8 Wittgensteinians may concede that some ritual actions have been performed out
of a desire to influence the course of events, but that such actions are superstitious
aberrations of the genuine article. D.Z.Phillips, for example, frequently draws a
distinction between ‘superstitious’ acts and ‘deep’ religious practices. A boxer who
crosses himself before a fight is superstitious if he believes that by so doing he will
come to no harm, but not so if the action is intended as a dedication of his
performance. Likewise, a prayer of petition is superstitious if intended to alter the
course of events, but deep if employed as a means of reflecting upon one’s selfish
desires, desires which threaten to lead the believer away from the will of God. The
application of this distinction to the material discussed by Wittgenstein in the
Remarks on Frazer is straightforward: magic is deep if expressive; superstitious if
instrumental in intent. This is certainly one way of salvaging the plausibility of the
expressivist thesis, but it has the unhappy consequence that magical rituals are
either condemned or else presented as something other than they naturally appear
to us, not as the attempts of a threatened community to protect itself against the
powers of nature, but as celebratory actions, cathartic releases, or the like. It is
hard here not to suspect that some sort of revisionary exercise is being undertaken
(rather than the purely descriptive work avowedly performed), and one feels bound
to accept John Skorupski’s verdict that in Phillips’ hands the beliefs of mankind
‘make up a mountain of superstition supporting a tiny cairn of Wittgensteinian
wisdom’ (J.Skorupski, Review of Religion Without Explanation, by D.Z.Phillips,
28 Brian R.Clack

Mind, vol. 88 (1979):155). For further consideration of the religion-superstition

distinction, see B.R.Clack, ‘D. Z.Phillips, Wittgenstein and Religion’, Religious
Studies, vol. 31, no. 1 (1995):111–20; and D.Z.Phillips, Wittgenstein and Religion
(London and Basingstoke: Macmillan, 1993) esp. pp. 56–77 and 245–7).
9 Frazer, op. cit., p. 168.
10 K.Thomas, Religion and the Decline of Magic (Harmondsworth: Penguin,
1978), p. 231.
11 See B.R.Clack, ‘Wittgenstein and Expressive Theories of Religion’, International
Journal for Philosophy of Religion, vol. 40, no. 1 (1996):pp. 47–61; and
B.R.Clack, Wittgenstein, Frazer and Religion (London and Basingstoke:
Macmillan, 1999), pp. 37–50.
12 See B.R.Clack, An Introduction to Wittgenstein’s Philosophy of Religion
(Edinburgh: Edinburgh University Press, 1999), pp. 27–50.
13 See A.J.Ayer, Language, Truth and Logic (Harmondsworth: Penguin, 1971), pp.
141–4; and C.L.Stevenson, Ethics and Language (New Haven, CT: Yale
University Press, 1944).
14 R.B.Braithwaite, ‘An Empiricist’s View of the Nature of Religious Belief’, in
B.Mitchell (ed.), The Philosophy of Religion (Oxford: Oxford University Press,
1971), pp. 72–91.
15 Ibid., p. 79.
16 R.G.Swinburne, The Coherence of Theism (Oxford: Clarendon Press, 1977), p.
17 M.C.Banner, The Justification of Science and the Rationality of Religious Belief
(Oxford: Oxford University Press, 1990), p. 69.
18 See Clack, op cit., pp. 45–6.
19 M.Eliade, Patterns in Comparative Religion (London: Sheed & Ward, 1987),
p. 9.
3 Wittgenstein, religious belief,
and On Certainty
Iakovos Vasiliou

Although there is little explicit mention of religious belief in On Certainty, I

shall show that Wittgenstein’s remarks elsewhere1 about the nature of
religious belief are most intelligible when read in light of that work.2 Certain
concepts, particularly world-picture (Weltbild), system of reference
(Bezugssystem), conversion (Bekehrung), and persuasion (Überredung) are
fundamental to both the remarks in On Certainty and to the numerous
remarks on religious belief in Culture and Value.

Religious belief appears to take a stand about a fact under dispute in making
a claim about how things were, are, or will be. We ordinarily suppose that
the truth of a belief is determined by whether, in fact, things are the way the
belief says they are: ‘The Earth goes around the Sun’ is true if and only if the
Earth does indeed go around the sun; ‘Christ was resurrected from the dead’
is true if and only if Christ was resurrected from the dead. If the Earth does
not in fact revolve around the Sun, or if Christ was not in fact resurrected
from the dead, then the belief is false.
Consider an example Wittgenstein uses: the occurrence of a Last
Judgment. The believer says there will occur a particular event, the Last
Judgment, a non-believer says there will not.3 Here the believer and non-
believer take opposing stands on a disputed fact. If it is shown that there will
not be a Last Judgment, or that we have good reason for believing there will
not, then the basis for the belief has been undermined, and a rational person
ought to give up the belief. Correspondingly, if it is proven that there will be
a Last Judgment, or if it is shown that we have good reason to think that
there will be a Last Judgment, then the belief is justified, and a rational person
ought to believe it.
What I have said so far can appear to be simply obvious and to explain
some common attitudes towards religious belief. One familiar line of thought
maintains that for many, if not all, religious beliefs we have overwhelming

30 Iakovos Vasiliou

evidence that the beliefs are false. Therefore, anyone who nevertheless holds
them does so not based on (good) reasons at all, but as a result of irrational
(or, at least, non-rational) faith. The fact that it can seem compulsory to say
that religious beliefs are held on faith results from the acceptance of the views
expressed in the previous two paragraphs. According to these views, religious
beliefs make disputed claims about what has been, what is, or what will be
the case. Since, however, these claims are very badly supported we ought to
describe these beliefs—beliefs that are, to quote Kierkegaard, held ‘on the
strength of the absurd’—as beliefs based not on reason, but on faith. I shall
refer to this as the ‘faith view’ of religious belief.
Another view about religious belief, which has been quite influential
although it is not perhaps in the ascendant at the moment, also maintains
that religious beliefs make claims about what has been, what is, or what will
be the case, but holds that there is good evidence for the truth of these beliefs:
artifacts, signs, testimony, etc. As with any empirical belief, the evidence is
what provides good reasons for thinking that the religious beliefs are true.
Let’s call this the ‘evidence view.’4
Both the evidence and the faith view of religious belief have in common
the idea that religious beliefs make claims about what has been, what is, or
what will be the case. I shall refer to this feature of religious belief as the
‘factuality of religious belief.’ By ‘factuality’ I am referring to the idea that
religious belief makes a claim about the occurrence of a state of affairs. Part
of the factuality of religious belief includes the idea that the belief is either
true or false, the state of affairs either obtains or it does not. Just as either the
Earth goes around the Sun or it does not, either Christ was resurrected from
the dead or he was not. Both the evidence and the faith view have this in
common. A third attitude towards religious belief might maintain that ‘Christ
was resurrected from the dead’ is a ‘metaphor’ or a ‘symbol’ for, say, the idea
that all human beings can in some sense renew their lives. Christ’s
resurrection stands as a symbol of hope. Such a view would say that although,
in reality, it is false that Christ was resurrected from the dead—that is, it did
not really happen—the narrative is still somehow important for human
beings and can function as a sort of allegory or metaphor for the human
condition. I can find no evidence that Wittgenstein ever discusses this sort of
view, or that he ever says that religious belief is simply false. The texts show
that he is interested in religious belief as an example of a belief that makes an
empirical claim, that purports to make a claim about what has been, what is,
or what will be the case. Whether true or false, religious belief concerns
matters of fact. This is what I mean by ‘factuality of religious belief.’
It is the factuality of religious belief that makes it an interestingly different
case from, say, ethical or aesthetic belief. It is a reasonable and
comprehensible (if not ultimately satisfactory) view to maintain that ethical
or aesthetic beliefs are not statements of fact, but rather expressions of
approval or disapproval—a set of positions held by non-cognitivists of
different stripes. The same cannot be said for many religious beliefs: ‘Christ
Wittgenstein, religious belief, and On Certainty 31

was resurrected from the dead’ is surely more naturally understood as making
a claim about the occurrence of a particular event rather than as expressing
an attitude. The factuality of religious belief makes religious belief appear to
be a species of ordinary empirical belief.
Wittgenstein agrees with the factuality of religious belief, but he disagrees
with both the faith and the evidence view. Paradoxically, Wittgenstein’s own
remarks appear to commit him both to the factuality of religious belief (which
it seems difficult to deny) and to the claim that the believer and non-believer
do not in fact contradict each other. There can appear to be very little room
to maneuver here. The factuality of religious belief by itself necessitates that
the state of affairs either obtains or does not; indeed, we might think this is
part of what it means for religious beliefs to be factual. On the evidence view,
the believer and the non-believer disagree about the quality of the evidence
for either side. On the faith view, the believer acknowledges that there may
be no good evidence that justifies his belief, but he believes anyway, on faith.
Both of these views, however, clearly maintain that there is a dispute between
the believer and the non-believer: they hold opposing and mutually
contradicting positions about the obtaining of a state of affairs.
Wittgenstein makes an apparently contradictory claim: the believer and
the non-believer hold contrary beliefs about the occurrence of a state of
affairs, but the believer and non-believer do not dispute with or contradict
one another.

Suppose that someone believed in the Last Judgement, and I don’t, does
this mean that I believe the opposite to him [sic], just that there won’t be
such a thing? I would say: ‘Not at all, or not always.’
If some[one?] said: ‘Wittgenstein, do you believe in this?’ I’d say: ‘No.’
‘Do you contradict the man?’ I’d say: ‘No.’
Would you say: ‘I believe the opposite’, or ‘There is no reason to suppose
such a thing’? I’d say neither.
(LC, p. 53)

Why does Wittgenstein say that the believer and non-believer do not
contradict each other? Doesn’t each take a stand contrary to the other about
the occurrence of some event? Perhaps we can answer these questions if we
consider Wittgenstein’s remarks about the relationship between religious
belief and evidence.

The point is that if there were evidence, this would in fact destroy the
whole business.
Anything that I normally call evidence would not in the slightest
influence me.
32 Iakovos Vasiliou

Suppose, for instance, we knew people who foresaw the future; make
forecasts for years and years ahead; and they described some sort of
Judgement Day. Queerly enough, even if there were such a thing, and
even if it were more convincing than I have described but [sic], belief in
this happening wouldn’t be at all a religious belief.
Suppose that I would have to forgo all pleasures because of such a
forecast. If I do so and so, someone will put me in fires in a thousand
years, etc. I wouldn’t budge. The best scientific evidence is just nothing.
A religious belief might in fact fly in the face of such a forecast, and
say, ‘No. There it will break down.’
(LC, p. 56)

These passages make clear that Wittgenstein does not believe that a proof
that a particular event has occurred or will occur by itself gives rise to a
religious belief. Religious belief does not rest on proof or evidence. In at least
one crucial respect, then, Wittgenstein thinks that religious belief differs from
ordinary empirical belief. We need to understand why a proof would fail to
make a believer out of a non-believer, and why a disproof would fail to make
a non-believer out of a believer. Part of the distinctive nature of religious
belief, according to Wittgenstein, is that it can ‘fly in the face of the ‘best
scientific evidence.’ So he rejects not only the evidence view, the idea that
there is in fact good evidence for religious belief, but he also holds a stronger,
counterfactual thesis: even if there were conclusive proof one way or the
other, that would be somehow irrelevant to the appropriate assessment of
religious belief. How can evidence be irrelevant to the assessment of a factual
Wittgenstein’s claim is not simply a psychological one. If it were, there
would be a straightforward response: the reason that the disproof of a
religious belief may fail to make a believer cease believing is because the
believer is plainly irrational—in the grip of irrational faith. The tenor of
Wittgenstein’s remarks clearly suggests a stronger claim. It is not simply that
a disproof does not work on the believer, but that it ought not to work; it
fails after all to be relevant.5 Something similar is true for the non-believer: a
cogent proof would not cause him to have a religious belief, nor ought it to.
It is true that if the non-believer is rational, then a truly cogent proof would
make him believe, say, that there will be a Last Judgment. But Wittgenstein
says that even so this belief would not thereby be a religious belief.
Does he also reject the faith view? Wittgenstein thinks that the faith view is
better in the sense that it does not fall into the trap of attempting to provide
evidence for religious belief. And he himself occasionally speaks of ‘faith’ or
‘belief’ (Glauben6 ) in Culture and Value (e.g. pp. 33, 53). Many adherents of
the faith view, however, might not share Wittgenstein’s wholesale rejection of
the evidence view. They may hold simply that there is no sufficient evidence for
their religious belief now, but that there could in theory be. God might provide
sufficient evidence for some person or people at some time, say. But in any
Wittgenstein, religious belief, and On Certainty 33

case, the faith view still holds that believer and non-believer occupy opposing
sides of a dispute about a state of affairs, and this is what Wittgenstein seems
to deny, while yet maintaining the factuality of religious belief:

What inclines even me to believe in Christ’s Resurrection? It is as though

I play with the thought.—If he did not rise from the dead, then he
decomposed in the grave like any other man. He is dead and
(CV [1937], p. 33)

Here Wittgenstein shows that he believes in what I have called the factuality
of religious belief: believing in the Resurrection entails believing that a certain
state of affairs either occurred or it did not. And yet he also says:

Queer as it sounds: The historical accounts in the Gospels might,

historically speaking, be demonstrably (erweislich) false and yet belief
would lose nothing by this: not, however, because it concerns ‘universal
truths of reason’! Rather, because historical proof (the historical proof-
game) is irrelevant to belief. This message (the Gospels) is seized on by
men believingly (i.e. lovingly).8 That is the certainty characterizing this
particular acceptance-as-true, not something else.
A believer’s relation to these narratives is neither the relation to
historical truth (probability), nor yet that to a theory consisting of ‘truths
of reason’. There is such a thing.
(CV [1937], p. 32)

Notice that Wittgenstein does not say that the belief is false, but that it is
‘demonstrably false’: that is, shown to be false by ‘the historical proof-game.’
He suggests, rather cryptically, that there is a sort of belief that is neither a
‘truth of reason’ (i.e. an a priori) truth, nor an ordinary empirical claim about
the past (an historical truth). It is a sort of belief that makes a claim about
how things are, were, or will be, and yet does not depend on evidence. As we
shall see in the next section, the nature of such beliefs is arguably the central
topic of On Certainty. For the moment, however, let us work on getting a
clearer sense of Wittgenstein’s description of religious belief.
Wittgenstein describes a person who is religious as fundamentally a person
who lives a certain sort of life.

One of the things Christianity says, I believe, is that sound doctrines

(guten Lehren) are all useless. One must change one’s life. (Or the
direction of one’s life.)
It says that wisdom is all cold; and that you can no more use it for
setting your life to rights than you can forge iron when it is cold.
The point is that a sound doctrine need not take hold (ergreifen) of
you; you can follow it as you would a doctor’s prescription.—But here
34 Iakovos Vasiliou

you need something to move you and turn you in a new direction.—(I.e.
this is how I understand it.) Once you have been turned around, you
must stay turned around.
Wisdom is passionless. But faith by contrast is what Kierkegaard calls
a passion.
(CV, p. 53 [1946])

The best scientific evidence for the occurrence of a Last Judgment makes it a
matter of empirical fact like any other: an event with no necessary connection
to leading any particular kind of life. To tell someone to believe that ‘God
knows all’ is to attempt to effect a change in how she lives her life, to change
what she currently sees as reasons for acting and living the way she does. The
simple belief in the occurrence of any empirical event by itself does not carry
with it any necessary consequences about how a person should live.

Suppose someone were a believer and said: ‘I believe in a Last

Judgement’, and I said: ‘Well, I’m not so sure. Possibly.’ You would say
that there is an enormous gulf between us. If he said There is a German
aeroplane overhead’, and I said ‘Possibly I’m not so sure’, you’d say we
were fairly near.
It isn’t a question of my being anywhere near him [the believer], but on
an entirely different plane, which you could express by saying: ‘You mean
something altogether different, Wittgenstein.’
Why is it that in this case I seem to be missing the entire point?
(LC, p. 53)

Wittgenstein denies that there is a genuine dispute here between the believer
and himself because what is at issue is not simply the belief or disbelief in
some particular state of affairs, such as whether there is a particular kind of
plane flying by, but about how one should lead a life. One could doubt that
there really is a certain kind of plane overhead, without that affecting one’s
life as a whole. But one cannot do that with belief in the Last Judgment.
Religious belief is a guidance for life:

Suppose somebody made this guidance for his life: believing in the Last
Judgement. Whenever he does anything, this is before his mind.
It will show, not by reasoning or by appeal to ordinary grounds for belief,
but by regulating for in all his life.
(LC, pp. 53–4)

This is a critical aspect of religious belief that distinguishes it, on

Wittgenstein’s account, from ordinary empirical belief. A religious belief is
Wittgenstein, religious belief, and On Certainty 35

something the believer has before his mind whenever he acts, something that
is regulating his conduct. Expressing that religious belief is expressing his
way of life. With an ordinary empirical belief, that such and such a plane is
overhead, or that so and so will be president, believing it to be true or false
does not provide any ‘guidance for life.’ This is why people on both sides of
a dispute about who will be president are, after all, quite close. Despite the
apparent similarity in cases, a dispute about whether there will be a Last
Judgment is not something on which one can take a stand without that belief
affecting the believer’s life as a whole. Even a rigorous proof that a particular
event has occurred or will occur will not, thereby, lead to a religious belief
because it will not necessarily—not by virtue of the proof alone—lead to a
change of life.
Now by itself having an overall effect on a person’s life is not enough to
explain the difference between a religious belief and an empirical belief. If a
person learns that he has six months left to live, then it is reasonable to think
that he thereby acquires an empirical belief that affects his life as a whole.
What count for him as worthwhile projects and what he sees himself as
having reason to do might well radically change. So it is not unique to
religious belief that it has an effect on the believer’s actions as a whole.
Let us look at other descriptions of what Wittgenstein thinks a religious
belief is and how it differs from ordinary empirical belief.

It strikes me that a religious belief could only be something like a

passionate commitment to a system of reference (Bezugssystem).
Hence, although it is belief, it is after all a way of living (Art des
Lebens), or a way of assessing life. It is passionately seizing hold of this
(CV [1947], p. 64)

Christianity is not based on a historical truth; rather, it offers us a

(historical) narrative and says: now believe! But not, believe this narrative
with the belief appropriate to a historical narrative, rather: believe,
through thick and thin, which you can do only as the result of a life.
Here you have a narrative, don’t take the same attitude to it as you take
to other historical narratives! Make a quite different place in your life
for it.—There is nothing paradoxical about that!
(CV [1937], p. 32)

Believing in the point of a narrative ‘through thick and thin’ can only be the
result of a particular kind of commitment, a commitment that makes no sense
with respect to ordinary empirical belief. Rationally held ordinary empirical
beliefs are beliefs that need to be revised in light of new evidence. Religious
belief is something you can have only ‘as the result of a life’ and requires a
commitment to a Bezugssystem. The nature of the commitment is not a rational
one, in the specific sense that it has not been arrived at on the basis of the most
36 Iakovos Vasiliou

plausible evidence. It is rather belief that is held passionately (leidenschaftlich).

Although the meaning of this is obscure, I think it suggests that there is a
commitment to a ‘way of assessing life’ that is not based on evidence; rather,
what is the case will turn out to be based on the ‘way of assessing life’ or the
‘system of reference.’ Such a passionate commitment perhaps explains why
even a life-changing empirical belief would not count as religious belief. If
someone changes his life, sees his life differently, as a result of coming to have
a new empirical belief (say, that he has six months to live), his commitment to
a new way of living is not something held passionately in Wittgenstein’s sense,
and it is not a commitment to a way of life. If this person finds out, in three
months, that the diagnosis was incorrect and his illness is in fact something
entirely trivial, then in so far as his new way of looking at the world was based
entirely on the alleged empirical fact of his impending death his attachment to
that way of life may well end. It is certainly also imaginable, however, that as
a result of a sort of conversion, he finds himself committed to living differently,
despite the fact that his death is no longer imminent. Such a person would
resemble the religious person now in so far as his commitment to the new way
of life was not based on any evidence or any facts; he believed in this new
guidance for his life ‘through thick and thin.’
Religious belief does not express simply the obtaining or not obtaining of
some state of affairs (e.g. the Resurrection)—although it does express this
too. Wittgenstein writes above, ‘Although it is belief…’ (his emphasis): he
does not give up on the idea that religious belief takes a stand about the
obtaining of some state of affairs. But it is more than simply a belief, it is also
a ‘way of living’ or a ‘way of assessing life.’ Greek moral philosophy is
famously concerned with a very particular kind of knowledge, which, when
possessed, would enable the possessor to get living well, or ‘happiness’
(eudaimonia), right. The Stoics called the wise man’s grasp of happiness
(eudaimonia) an art or skill of living (technê tou biou). The wise man has
knowledge but that knowledge is particularly knowledge of what is the best
life, a knowledge that implies that he gets living well, or happiness, right.
When Wittgenstein says that religious belief is a passionate commitment to a
‘system of reference’ and a ‘way of living,’ it sounds quite similar to the Greek
idea of the virtuous or wise person’s commitment to a certain conception of
eudaimonia. The comparison with Greek moral philosophy may be quite apt:
the point of the commitment to a certain narrative, a certain conception of
the facts, is not its scientific validity, but its claim to capture correctly the
best way to live. Understanding and believing in a particular conception of
eudaimonia appears to be just this sort of passionate commitment to a way
of life—think of Socrates’ commitment to virtue.
Commitment to religious beliefs for Wittgenstein is more than simply a
commitment to certain claims about how things stand: it is a commitment to
living a certain way. Religious beliefs are still factual, in the sense that they
do make claims about how things stand. But since religious beliefs make such
claims without basing them on evidence, and even in such a way as to fly in
Wittgenstein, religious belief, and On Certainty 37

the face of evidence, they are shown to be held on the basis not of evidence,
but of the believer’s commitment to a particular way of life. This is part of
the reason why there is no dispute in any ordinary sense between the believer
and non-believer according to Wittgenstein. An ordinary dispute arises over
a disagreement about the quality or quantity of evidence justifying a certain
claim. When religious belief flies in the face of what ordinarily counts as good
evidence, when the believer believes without evidence, it shows that he is
part of a different Bezugssystem. We still need to get clearer on what a
Bezugssystem is, and how it conditions our beliefs about the world: for this
the critical text is On Certainty. But first let us look at what Wittgenstein
says about how religious beliefs arise differently from empirical beliefs:

A religious belief might in fact fly in the face of such a forecast, and say
‘No. There it [the best scientific evidence] will break down.’
As it were, the belief as formulated on the evidence can only be the
last result—in which a number of ways of thinking and acting crystallize
and come together.
(LC, p. 56)

In this passage Wittgenstein is arguing that the believer does not believe on
the basis of evidence. When he speaks of ‘the belief as formulated on the
evidence’ I take him to mean the expression of a religious belief in the ‘form’
of an ordinary empirical belief.9 As we have seen, this is the central difficulty
concerning religious belief. It has, one might say, the grammar of statements
expressing empirical beliefs, and yet religious belief can fly in the face of
empirical evidence. Religious belief ‘flies in the face’ of empirical evidence
because what one is trying to express in saying ‘there will be a Last Judgment’
is not simply the belief in the obtaining of a state of affairs, but a willingness
to commit to an entire way of regulating one’s life.
Wittgenstein’s last sentence in the above passage is difficult, but important.
In the case of an ordinary empirical belief, the belief arises on the basis of
evidence; it is the evidence that gives rise to the belief. With religious belief,
despite having the grammar of an empirical belief, the expression of the belief
in the form of an ordinary empirical belief is ‘the last result.’10 What does
Wittgenstein mean by this? He offers a brief gloss ‘—in which a number of
ways of thinking and acting crystallize and come together.’ I think he is saying
that in the case of religious belief, what gives rise to the belief is not any
empirical evidence—we have seen why empirical evidence alone cannot make
a believer—but ‘ways of thinking and acting’ that constitute a kind of life,
say, a Christian life or a Buddhist life. These beliefs get expressed, ‘crystallize
and come together,’ in statements that have the form of ordinary empirical
propositions; that is, as though they arose from evidence, when, in fact, they
are the ‘last result,’ the expressions of adherence to a way of life not founded
on any empirical evidence, but nevertheless making claims about the
occurrence of states of affairs.
38 Iakovos Vasiliou

Isn’t Wittgenstein’s position, then, simply a version of the faith view? At

the moment we should note that Wittgenstein does have in common with the
faith view the rejection of evidence as the basis for belief. In this respect, as I
said earlier, the faith view is clearly closer to Wittgenstein’s than the evidence
view. As I have been characterizing it, however, the faith view still holds that
there is a conflict over the truth of empirical facts between the believer and
the non-believer. I think that when Wittgenstein speaks about the believer
being ‘on a whole different plane’ and about there being ‘an enormous gulf
between the believer and the non-believer, he is rejecting this idea.
Considering On Certainty should offer us a better understanding of why he
does this.
In the ‘Lectures on Religious Belief,’ after some discussion of how
Christian beliefs ought not to be understood as depending on historical
evidence, Wittgenstein says:

Why shouldn’t one form of life culminate in an utterance of belief in a

Last Judgement? But I couldn’t either say ‘Yes’ or ‘No’ to the statement
that there will be such a thing. Nor ‘Perhaps’, nor ‘I am not sure’.
(LC, p. 58)

The expression of religious belief in the form of an empirical statement, say

‘there will be a Last Judgment,’ is part of the ‘culmination’ of a ‘form of life.’
The expression of religious belief in the form of an empirical belief is what
comes last; it is what is said by someone leading a certain kind of life. This is
a recurrent theme throughout Wittgenstein’s writings on religious belief. Part
of what Wittgenstein means by saying that religious belief is a ‘culmination’
of a way of life is that religious belief, as opposed to belief in ordinary
empirical events, often arises as a result of a particular kind of upbringing or
education (Erziehung). It is the result of ‘life,’ and so one can be led to such
beliefs only in the course of being brought up, persuaded, or converted to
follow a particular kind of life: the religious one.

A proof of God’s existence ought really to be something by means of

which one could convince oneself that God exists. But I think that what
believers who have furnished such proofs have wanted to do is give their
‘belief an intellectual analysis and foundations, although they themselves
would never have come to believe as a result of such proofs. Perhaps one
could ‘convince someone that God exists’ by means of a certain kind of
upbringing (eine Art Erziehung), by shaping his life in such and such a

Life can educate (erziehen) one to a belief in God. And experiences too
are what bring this about; but I don’t mean visions and other forms of
sense experience which show us the ‘existence of this being’, but, e.g.,
sufferings of various sorts. These neither show us God in the way a
Wittgenstein, religious belief, and On Certainty 39

sense impression shows us an object, nor do they give rise to conjectures

about him. Experiences, thoughts,—life can force this concept on us.
(CV, p. 85–6 [1950]).

Religious belief is the upshot of a kind of life, a kind of upbringing, which

culminates in a certain sort of belief. Further, when a belief is generated in
this way, it is not the result of ‘evidence,’ such as visions, etc., and so it does
not result in a belief that is simply possible—one that may be true, or, as
Wittgenstein says, ‘a conjecture.’ A person is brought up, or initiated into a
particular tradition, a particular way of life, and then finds himself acting in
certain ways—and then expressing his belief in certain things.
I now turn to On Certainty.


‘Moore propositions’
One of the primary themes in On Certainty is that not all empirical
propositions have the same logical status. Some propositions that have the form
of empirical propositions, and indeed make claims about what has been, what
is, or what will be the case, nevertheless do not function in the same way as
other ‘ordinary’ empirical propositions. Wittgenstein writes: ‘Our “empirical
propositions” do not form a homogeneous mass’. (OC, §213).11 Again: ‘It is
clear that our empirical propositions do not all have the same status, since one
can lay down such a proposition and turn it from an empirical proposition into
a norm of description’ (§167; see also §§ 308, 401).
Examples of these sorts of propositions are plentiful in On Certainty. The
Earth has existed for more than 150 years’; ‘I have two hands’; ‘I have never
been on the moon’; ‘All human beings have two parents.’ Some of these
examples come originally from G.E.Moore,12 and so I shall call them ‘Moore-
propositions,’ or ‘M-propositions’ for short. M-propositions ‘have the form
of empirical propositions’ in so far as they make claims about the obtaining
of some state of affairs, and there is no logical contradiction in imagining
their denial; that is, they are not in the usual sense a priori propositions.
Wittgenstein denies being able to provide any common characteristic to all
M-propositions (§674), but he does say that they are propositions about
which one could not be simply mistaken. As Wittgenstein puts it, if we are
wrong about a M-proposition it would not be considered a mistake, but
rather perhaps some sort of mental disturbance: ‘If Moore were to pronounce
the opposite of those propositions which he declares certain, we should not
just not share his opinion: we should regard him as demented’ (§155; see also
§§71–2, 194, 425, 674).
Although we may agree that Wittgenstein has shown us something
important about the language-game of mistakes, and its relation to M-
40 Iakovos Vasiliou

propositions, it is clearly not by itself an adequate response to the radical

skeptic. There is a lot of debate about how successfully or unsuccessfully
Wittgenstein can respond to the skeptic, but this is not my concern here.13 I
am interested in how Wittgenstein’s careful descriptions of M-propositions
and the roles they play in our thought and action might be related to and
shed light on his discussion of the nature of religious belief.
It is important to understand that what counts as a M-proposition is going
to change in different circumstances and in different historical situations. The
line between M-propositions and ordinary empirical propositions is not a
sharp one. This is part of what makes understanding their role in our
language and thought difficult. Philosophers typically distinguish empirical
claims from necessary, or logical, truths. It is central to the latter that they
have a sort of timeless, true-in-all-possible-worlds, status. M-propositions are
emphatically not like these, but they are not ordinary empirical propositions
either.14 Wittgenstein employs the image of a river:

It might be imagined that some propositions, of the form of empirical

propositions, were hardened and functioned as channels for such
empirical propositions as were not hardened but fluid; and that this
relation altered with time, in that fluid propositions hardened, and hard
ones became fluid.
But I distinguish between the movement of the waters on the river-bed
and the shift of the bed itself; though there is not a sharp division of the
one from the other.
The same proposition may get treated at one time as something to test by
experience, at another as a rule of testing.
And the bank of that river consists partly of hard rock, subject to no
alteration or only to an imperceptible one, partly of sand, which now in
one place now in another gets washed away, or deposited.
(§§ 96–9)

Empirical propositions do not, as their grouping under the label ‘empirical

propositions’ misleadingly suggests, ‘form a homogeneous mass’; the
category ‘empirical proposition’ is not fine-grained enough. Together they
constitute a whole, but that whole is like a river, with heterogeneous parts
functioning in distinct ways. Some propositions, those I have called M-
propositions, form, as it were, the river-bed and are the hardened backdrop
against which the others flow. Part of the river-bed is hard rock and
susceptible only to an ‘imperceptible’ alteration over time, but some of it is
sand, which shifts much more slowly than the flow of the water, but more
quickly than the stone, changing the appearance of the river. The line between
river-bed, sand, and flowing water is not sharp, although one can choose
Wittgenstein, religious belief, and On Certainty 41

examples of propositions that clearly belong in one category or another.

Keeping this image in mind should help us to understand what Wittgenstein
is doing as he considers particular examples of M-propositions and their role
in our thinking and acting.
Wittgenstein claims that M-propositions are treated as ‘rule[s] of testing’
experience. Empirical propositions become M-propositions by becoming
part of the river-bed, that is, by becoming ‘norms of description.’ With an
ordinary empirical belief it is the evidence of experience that justifies the
belief; we test the belief against experience. But an M-proposition is not
(or, perhaps, is no longer) tested ‘by experience’ but has itself become a rule
or a measure of testing (§309). It is what is held fast, and other propositions
are tested against it (§151). M-propositions, then, are not supported by
evidence. As we shall see, they form the background against which other
empirical claims are established and tested. The fact that M-propositions
have ‘the form of empirical propositions,’ the fact that they make claims
about how things stand, and yet are not held on the basis of evidence, marks
an important similarity with what we have seen Wittgenstein say about
religious belief.
Consider the following chain of reasoning. Scientists believe that the Earth
is 4.5 billion years old. They don’t know it, insofar as that number is not
precise; it may be off by millions of years. But they are certain that it is older
than 3.5 billion years, and the certainty increases the smaller the amount of
time that is specified. Everyone is absolutely certain that the Earth is more
than 50 years old.
What is odd about the proposition The Earth is more than 50 years old’ is
that its function and place among empirical propositions are different from
those of a proposition like ‘The Earth is 4.5 billion years old,’ despite the
similarity in grammatical form, and even in content—after all, both
propositions are about the age of the Earth. ‘The Earth is more than 50 years
old,’ which obviously makes a claim about what is the case, is not established
by experience or on the basis of evidence, like other ordinary empirical claims.
Rather it is a rule, it is part of the background against which our entire
practice of determining the age of the Earth operates (cf. §167 and §98,
above). But, one might still object, surely we can test by experience whether
or not the Earth is more than 50 years old?

What we call historical evidence points to the existence of the earth a

long time before my birth;—the opposite hypothesis has nothing on its
Well, if everything speaks for an hypothesis and nothing against it—is
it then certainly true? One may designate it as such.—But does it certainly
agree with reality, with the facts?—With this question you are already
going around in a circle.
42 Iakovos Vasiliou

For something to count as evidence for a claim, there needs to be a conception

of what would count for the claim, and what would count against it. With a
proposition like ‘The Earth is more than 50 years old’, everything counts for
it. We have no clear idea of what sort of evidence could count against it.
Wittgenstein says that we can ‘designate’ this as certain. Although this sounds
like it is slighting the truth of the claim, he thinks that there is something
wrong with simply saying that it agrees with the facts, with reality. We are
going around in a circle here because there are no facts that we are more
certain of against which we can check whether or not the Earth is more than
50 years old. In order to see whether we know something, that is, to consider
its justification and evidence, we must be able to consider the evidence as
more secure than the claim being examined. In the case at hand, we have no
such evidence. We can ‘designate’ the claim as certain, but we are not really
determining the truth of an empirical proposition, in the way we determine
the truth about, say, the number of people in a room—something determined
by perceptual evidence and counting. Rather, in calling an M-proposition
certainly true we are describing part of what Wittgenstein will call our ‘world-
picture’ (Weltbild), part of the river-bed against which we make ordinary
judgments about reality. This ‘designating’ is a way of saying that the claim
functions for us as a rule or norm for testing: it is part of what determines
what it is for us to test in the first place.

Weltbild and Bezugssystem

Everything that I have seen or heard gives me the conviction that no man has
ever been far from the Earth. Nothing in my picture of the world (Weltbild)
speaks in favour of the opposite.
But I did not get my picture of the world (Weltbild) by satisfying myself of
its correctness; nor do I have it because I am satisfied of its correctness. No: it
is the inherited background against which I distinguish between true and false.
[How do I know ordinary historical facts are true?] I have a world-picture
(Weltbild). Is it true or false? Above all it is the substratum of all my enquiring
and asserting. The propositions describing it are not all equally subject to
testing. (§§93–4, 162)

‘The Earth has existed for more than fifty years’ is not simply an ‘ordinary
empirical proposition’ insofar as it is part of our Weltbild, part of the
inherited background against which we distinguish between what is true and
what is false. To dispute it is not the same as disputing, say, whether
Washington crossed the Delaware. In this sort of ‘ordinary’ case we have a
clear idea of what sorts of things count as evidence for and against this type
of belief. Such a belief is not part of the ‘substratum’ of all of our inquiry and
so exempt from testing. M-propositions, by contrast, form the background
or ‘scaffolding’ (§211) of our thinking and describe our picture of the world.
They describe the backdrop against which ordinary empirical beliefs are
Wittgenstein, religious belief, and On Certainty 43

tested: ‘The truth of certain empirical propositions belongs to our frame of

reference (Bezugssystem)’(§83).
Our world-picture, which is described by the propositions that are M-
propositions, constitutes a ‘system of reference’ (Bezugssystem) or system of
evidence (System der Evidenz).15 What count as empirical facts do so within
the system. Our system of reference determines not only the conditions of
truth and falsity for all propositions (as I have already discussed) but also the
truth of certain empirical propositions, M-propositions.16 The world-picture
itself is not grounded or justified by any evidence, and therefore the entire
system cannot, in the same sense at least, be called true or false: ‘It is so
difficult to find the beginning. Or, better: it is difficult to begin at the
beginning. And not try to go further back’ (§471). ‘The difficulty is to realize
the groundlessness of our believing’ (§166).
We cannot call propositions that describe our Weltbild reasonable or
unreasonable, justified or unjustified, because there is nothing against which
to make such determinations. I think this is what Wittgenstein means when
he says we can ‘designate as certainly true’ that ‘the Earth has existed for
more than fifty years.’ To say that such a proposition is justified is inaccurate,
but so is to say that it is unjustified. If we insist on this question of
justification, we are trying to go ‘too far back’ because we are trying to step
outside of the structure that makes justification and the giving of evidence
possible. This is why saying that we ‘know’ these propositions is odd.17 Our
‘believing’ in our Weltbild is not itself based on any grounds; it is what
determines what shall count as grounds, without itself being grounded.
In a sense, however, our world-picture is grounded, but it is grounded in
our actions, in the activities in which we engage in our lives.

As if giving grounds did not come to an end sometime. But the end is not
an ungrounded presupposition; it is an ungrounded way of acting.
Sure evidence is what we accept as sure, it is evidence that we go by in
acting surely, acting without any doubt.
Giving grounds, however, justifying the evidence, comes to an end;—but
it is not a kind of seeing on our part; it is our acting, which lies at the
bottom of the language-game.
(§§110, 196, 204)

We are taught to engage in various activities. We learn to behave, to

investigate, to inquire in certain ways. It is our actions and practices that
provide a sort of ‘foundation’ or ‘justification’ for our Weltbild. Our system
of evidence, characterized by M-propositions, describes not what we believe,
so much as our form of life, and the sorts of activities we engage in. These
activities themselves are not grounded in anything; they are simply what we
do. The language-game itself is not based on grounds, nor is it reasonable or
unreasonable, but ‘it is there—like our life’ (§559).
44 Iakovos Vasiliou

This points to a larger problem with the entire expression ‘form of empirical
propositions’, an expression Wittgenstein himself has been using (e.g. §96).

I want to say: propositions of the form of empirical propositions, and

not only propositions of logic, form the foundation of all operating with
thoughts (with language).
In this remark the expression ‘proposition of the form of empirical
propositions’ is itself thoroughly bad; the statements in question are
statements about material objects. And they do not serve as foundations
in the same way as hypotheses which, if they turn out to be false, are
replaced by others.

…und schrieb getrost

‘Im Anfang war die Tat.’
[…and write with confidence
‘In the beginning was the deed.’ (Goethe, Faust I)].


Wittgenstein expresses his discomfort with trying to categorize the

propositions he is interested in as falling under a particular form. I think that
the problem with ‘proposition of the form of empirical propositions’ is that it
suggests that such propositions are only apparently but not really empirical
propositions. Wittgenstein emphasizes that such propositions ‘are statements
about material objects’, but also that they do not serve as hypotheses, which
could be replaced by others, e.g.: ‘Let’s suppose the Earth existed for more
than fifty years…’; ‘Let’s suppose Washington knew of British troop
movements before…’. Recall the description of the believer’s relation to the
message of the Gospels as neither that of historical truth nor that of ‘truths of
reason’ (CV, p. 32).
It is difficult to be sure how Wittgenstein intends Goethe’s words to be
understood. But it would make good sense if he was making a point similar to
his idea that religious beliefs present the culmination of a way of life. In the
same way, ‘the Earth has existed for more than fifty years’ could never be the
first belief one had about the existence of the earth. Rather, there has already
been a certain form of life, a certain world-picture, that was being lived and
acted upon, which then threw up an expression such as ‘the earth has existed
for more than fifty years.’ That is why it is such a queer-sounding phrase; it is
not false, but it is also not a conclusion arrived at by evidence, despite seeming
like one. ‘In the beginning was the deed…’ suggests that certain activities came
first, and only afterwards were certain expressions used.18

I do not explicitly learn the propositions that stand fast for me. I can
discover them subsequently like the axis around which a body rotates.
Wittgenstein, religious belief, and On Certainty 45

This axis is not fixed in the sense that anything holds it fast, but the
movement around it determines its immobility.

We engage in certain activities, having learned them without thinking about

their ‘basis.’ Afterwards we discover, perhaps while doing philosophy, the
propositions that stand fast for us, that is, the propositions that are
presupposed by the activities in question—e.g. conducting experiments,
playing chess, doing history. These propositions, the ‘axis’ in the passage
above, are not themselves held in place by anything. It is the activities
themselves that determine which propositions hold fast for us. Wittgenstein
offers us an example of this in his description of Lavoisier’s experimentation
as an example of an activity that rests on the idea that the same event will
occur under the same conditions (§167). This is part of a ‘definite world-
picture,’ which Lavoisier did not invent, but learned as a child and which
functions as the ‘matter-of-course foundation’ of his research. Similarly,
playing chess rests on the idea that the pieces are not moving themselves
unbeknown to the players (§342). This is the axis around which the activity
of chess rotates. It is the activity of playing chess that renders the idea fixed.
The idea that the Earth is more than 50 years old is part of the ‘matter-of-
course foundation,’ the axis on which all of our study of history turns.
Although these propositions themselves have no grounds, no evidence that
supports them, they stand fast: they are part of the riverbed because of the
activities we engage in in our lives.

Disputing a Weltbild
A Weltbild is acquired or changed through upbringing, conversion, or
persuasion—a theme which runs throughout Wittgenstein’s remarks.

Men have believed that they could make rain; why should not a king
have been brought up with the belief that the world began with him?
And if Moore and this king were to meet and discuss, could Moore really
prove his belief to be the right one? I do not say that Moore could not
convert (bekehren) the king to his view, but it would be a conversion
(Bekehrung) of a special kind; the king would be brought to look at the
world a different way.

Moore and the king have different world-views. Moore could not prove to
the king that his view is wrong; he could only ‘convert’ (bekehren) him by
getting him ‘to look at the world a different way.’ A proof is not possible
because Moore and the king do not share a common background against
which truth and falsehood are distinguished, and against which criteria for
proof are established.
46 Iakovos Vasiliou

To imagine someone asking seriously about the recent coming into

existence of the earth, it is perhaps easiest to think of a child who has not yet
been initiated into ‘our’ picture of the world. Wittgenstein imagines a child
asking him whether the Earth existed before he was born (§233). He describes
the answer as a step in imparting a world-picture to the child, not simply
relating an ordinary fact. In answering the child in this case you are imparting
to him or her the background against which to determine the truth and
falsehood of other claims.
Wittgenstein describes an interaction with a figure like the king in terms
of a kind of persuasion (Überredung) aimed at getting him to accept our
world-picture (Weltbild) (§262). As Wittgenstein considers disputes about
propositions that are part of our Weltbild, part of the riverbed, he shows that
such a conception is not brought about by reasoning. It is rather a
‘conversion’ or ‘persuasion’ into proper reasoning, reasoning which is not
itself effected by reasoning19 but by one’s upbringing, by the language games
into which one has been initiated. To effect a change in it, one must effect a
change in one’s world-view, be rebrought up, converted, or persuaded.
Wittgenstein returns to this frequently in On Certainty. He imagines people
of a contemporary scientific culture confronting people who believe in
oracles, saying that one group would ‘combat’ the other (§609).

Where two principles really do meet which cannot be reconciled with

one another, then each man declares the other a fool and a heretic.
I said I would ‘combat’ the other man,—but wouldn’t I give him
reasons? Certainly; but how far do they go? At the end of reasons comes
persuasion (Überredung). (Think what happens when missionaries
convert [bekehren] natives.)

Wittgenstein argues, then, that beliefs which are part of our Weltbild are
not individual beliefs that are imparted piecemeal and established by the
‘ordinary’ means of providing evidence. In forming part of the backdrop,
part of the ‘inherited background’ against which we determine what is
true and false, they are beliefs we acquire through upbringing, conversion,
and persuasion. Reasons give out in arguing for such propositions because
these propositions determine what we count as reasons in the first place.
Between systems, as it were, Wittgenstein tells us that we are left with
persuasion (Überredung) and conversion (Bekehrung): see things the way
I (we) do. When natives are converted, one world-view is replaced by
another. Wittgenstein uses the term ‘combat,’ implying that one picture
or system wins out over another. But thinking of the conversion of natives
we see that it is not simply that beliefs are changed or revised, in the way
that we might revise our beliefs about, say, some historical figure or event,
but a whole way of life. The conversion of natives involves not simply
getting them to believe in something that they do not believe in, but to
Wittgenstein, religious belief, and On Certainty 47

live in a wholly new way, to give up certain practices and certain activities,
and take up others.


On Certainty claims, then, that all empirical propositions are not in fact the
same type of proposition. Some propositions—M-propositions—count as
propositions that form part of our ‘frame of reference’ and are like ‘hinges’
(§§341, 343) on which other propositions and activities turn. M-propositions
are not held on the basis of evidence. Rather, we engage in certain activities
which hold these propositions fixed for us. We only formulate M-
propositions after we have already been initiated into a particular way of
life. Together they reflect and describe a ‘world-picture’ or a ‘system of
reference.’ To challenge or dispute an M-proposition, then, is not like
disputing an ordinary empirical belief. Reasons and evidence cannot be
provided for M-propositions because the M-propositions themselves
determine what count as evidence and reasons in the first place—indeed what
count as empirical facts. Disputes about world-pictures are ‘resolved’ in
‘combat’ by the conversion or persuasion of one of the parties.
How are ‘The Earth is more than 50 years old’ and ‘Christ was resurrected
from the dead’ similar? One respect of the similarity is the origin of the beliefs.
In neither case does the belief arise as the result of evidence. Rather, the
expression of the beliefs is part of the expression of one’s picture of the world.
They are expressed after that view of the world is already there, and represent
the culmination of certain ways of living and thinking. Further, they are
imparted as the result of a particular kind of upbringing. One is not argued
into believing them, in the way that one might argue that the earth is 5 not
4.5 billion years old, or that Washington crossed the Delaware on such and
such a day. In these latter cases, cases I have called ordinary empirical beliefs,
what counts as evidence and as lack of evidence is clear. M-propositions or
‘propositions of the frame’ have been arrived at non-rationally: that is not to
say irrationally, for they determine what counts as rational and irrational.
The language-games we have been initiated into are ‘not reasonable (or
unreasonable)’; rather, they are simply there, ‘like our life’ (§559). Other than
bringing someone up, educating someone to believe in God and so supplying
them with ‘arguments’ from that world-picture,20 a person, like the king, can
also be ‘converted’ to a different system of reference.
I have tried to show the similarities between the way Wittgenstein talks
about M-propositions and the way he discusses religious belief. The account
that emerges appears to be one that implies some sort of radical relativism. A
believer has a certain world-picture that is ‘based on’ a particular way of
living his life. In this respect the believer is no different than the secular
scientist. We all hold certain beliefs fixed. These beliefs are held in place by
the sorts of activities we engage in, the sorts of lives we lead. At bottom,
according to Wittgenstein, the core beliefs of the secular person are no more
48 Iakovos Vasiliou

well-grounded, absolutely, than the beliefs of the religious person.

Wittgenstein speaks of conversion and persuasion, and people with different
world-pictures ‘combating’ one another. So where does this leave matters?
The non-believer can declare the believer’s beliefs false, but it seems that he
does this only within his own system of reference, within his own world-
view. Between world-views Wittgenstein appears to have left us with only
non-rational methods of conversion. Is Wittgenstein, then, at least in On
Certainty, a conceptual relativist of some sort? I cannot answer this larger
question here, but it is an issue that he himself sees as a concern.

‘But is there no objective truth? Isn’t it true, or false, that someone has
been on the moon?’ If we are thinking within our system, then it is certain
that no one has ever been on the moon.
It would strike me as ridiculous to want to doubt the existence of
Napoleon; but if someone doubted the existence of the earth 150 years
ago, perhaps I should be more willing to listen, for now he is doubting
our whole system of evidence (System der Evidenz). It does not strike me
as if this system were more certain than a certainty within it.
It strikes me as if someone who doubts the existence of the earth at that
time is impugning the nature of all historical evidence. And I cannot say
of this latter that it is definitely correct.
(§§108, 185, 188)

These passages express a clear reluctance to make claims that cut across
systems of evidence or world-pictures. It remains an open question whether
Wittgenstein has the resources to deal with this sort of relativism and, indeed,
whether he is in the end dissatisfied with it.21

1 I shall be considering primarily ‘Lectures on Religious Belief’ in Wittgenstein:
Lectures and Conversations on Aesthetics, Psychology and Religious Belief, ed.
C. Barrett (Berkeley: University of California Press, 1972) and Culture and Value
(CV) (Chicago: University of Chicago, 1980).
2 I am assuming that we have a rough idea of examples of religious belief without
needing to attempt any sort of definition. Of course, there are many different
systems of religious belief, and beliefs between and within those systems can and
do conflict.
3 Nothing special is intended by using an example of a future event. We could just
as well discuss a past event, Christ’s resurrection, or a present one, say the
weeping of an icon, as an example of a miracle.
4 I am not claiming that these are exhaustive, and clearly there are many versions
of the views under each of these types.
5 This is perhaps a claim that few actual believers would make themselves. It is
clear that Wittgenstein thinks most believers are confused about the nature of
their own belief; cf. his exasperation with attempts to put religion on a ‘scientific
foundation’ (LC, pp. 57–9, remarks on Father O’Hara).
Wittgenstein, religious belief, and On Certainty 49

6 Although in most cases it is better translated ‘belief’.

7 All italics are Wittgenstein’s own, unless otherwise indicated.
8 This sentence and the next are about what Wittgenstein thinks about religious
belief in a positive way, although they are by themselves quite cryptic. I include
them here simply so as not to break up Wittgenstein’s remark, but I will not
discuss them until later. At the moment I am concerned with Wittgenstein’s
criticism of other accounts of religious belief.
9 I should note that I understand ‘the belief’ in the second paragraph to refer to
‘religious belief,’ thereby developing further the remark in the first paragraph.
Robert Arrington has pointed out to me that one might also take ‘belief’ to refer
to any ordinary belief, which would significantly alter the meaning of the
passage. This is certainly a possible reading, but I do not think it fits best with
the rest of the passages in LC or with Wittgenstein’s remarks in CV. I should say
here that LC itself is a highly questionable text in so far as it is assembled from
notes taken by students, and not written by Wittgenstein himself; there are many
ambiguities, grammatical infelicities, and sometimes apparently contradictory
statements. This is especially distressing as a source for a philosopher who was
notoriously fastidious and agonized about the precision of his expression, editing
his remarks many times (and of course holding back his work from publication).
I have therefore generally treated LC as a second-rate source and as much as
possible supplemented any claims that rely on it with remarks in CV, which are
in Wittgenstein’s own hand.
10 Below we will see that the expression ‘form of an empirical proposition’ is one
that Wittgenstein both used and thought inadequate. I will discuss why, but for
now it is useful shorthand.
11 Unless otherwise indicated, all section numbers are those of On Certainty.
12 See especially ‘Proof of an External World’ and ‘A Defence of Common Sense’ in
his Philosophical Papers (London: Allen & Unwin, 1959).
13 In fact, Wittgenstein himself clearly distinguishes the logical possibility of being
mistaken about the truth of a proposition from the logical possibility of the
proposition’s being false (§155). For a discussion of On Certainty as a reply to
the skeptic, see M.McGinn, Sense and Certainty (Oxford: Blackwell, 1989),
especially Chapters 6–8, and M.Williams, Unnatural Doubts (Princeton, NJ:
Princeton University Press, 1996).
14 Recall Wittgenstein’s remark from Culture and Value ([1937] p. 32), quoted in
the previous section, about there being a kind of truth that is neither historical
truth (the truth of an ordinary empirical belief about the past) nor a ‘truth of
reason’ (an a priori analytic truth).
15 See §185 for this expression; also §105.
16 The ‘certain empirical propositions’ of §83 refers to M-propositions. §84 begins:
‘Moore says he knows that the earth existed long before his birth…’
17 I want to separate this point from any argument against radical scepticism. We
might agree that Wittgenstein has shown us something illuminating about the
nature of Moore-propositions and the roles that they play in our lives and
thoughts, without his having thereby shown us that they are immune to skeptical
doubt. See Williams, Unnatural Doubts:

[The] difficulty is to make it clear why these Wittgensteinian observations

show that the traditional epistemologist’s questions are illegitimate, rather
than just unusual. It is all very well to claim that our relation to framework
judgments is non-epistemic: the problem is to show that this amounts to
more than the claim that our relation to them is not ordinarily epistemic, or
(an even weaker conclusion) not ordinarily treated as epistemic, neither of
50 Iakovos Vasiliou

which the sceptic is obviously committed to denying. What needs to be ruled

out, if this response to scepticism is to work, is the suggestion that our
relation to such judgments becomes epistemic in the context of the
traditional project of assessing our knowledge of the world as a whole. The
sceptic agrees that his project is unusual but denies that this alone reveals it
as incoherent.
(pp. 25–6).

18 Wittgenstein quotes the same passage at CV(1937), p. 31.

19 For this idea employed in a closely connected way, see J.McDowell, ‘Might There
Be External Reasons?’ in J.E.J.Altham and R.Harrison (eds), World, Mind, and
Ethics (Cambridge, 1995), pp. 68–85. McDowell’s account, however, does not
discuss Wittgenstein.
20 See OC §106, and also §107: ‘Isn’t this altogether like the way one can instruct
a child to believe in a God, or that none exists, and it will accordingly be able to
produce apparently telling grounds for the one or the other?’
21 I would like to thank Robert Arrington and Bill Vasiliou for their comments on
earlier versions of this paper.
4 Creation, causality, and
freedom of the will
William H.Brenner

According to Norman Malcolm, Wittgenstein once said he could not understand

the conception of a Creator. ‘I believe,’ Malcolm continued, that he

was prepared by his own character and experience to comprehend the

idea of a judging and redeeming God. But any cosmological conception
of Deity, derived from the notions of cause or of infinity, would be
repugnant to him…. He was impatient with…attempts to give religion a
rational foundation.1

In this paper I marshal material from various Wittgenstein texts for the
purpose of elucidating a concept of a Creator neither repugnantly rationalistic
nor out of harmony with the idea of a judging and redeeming deity. My
purpose is less to exposit Wittgenstein’s scattered remarks on these matters
than to develop and organize the insights I believe some of them contain.
Perhaps even less than elsewhere, Wittgenstein has not in these remarks
spared his readers the trouble of thinking for themselves. Whether this
reader’s identification and development of Wittgensteinian insights are
legitimate and plausible is for others to judge. The following brief outline
may be of help.
Believing in a Creator requires, not accepting a causal hypothesis (section I),
but prescinding at a certain point from the causal point of view (section II); it is
connected, essentially, with a particular way of acting and reacting (section
III); it requires relativizing the everyday practice of judging ourselves and others
(section IV). This account of ‘believing in a Creator’ is offered as contribution
to philosophical investigation, not religious apologetics (section V).

When someone who believes in God looks around him and asks ‘Where did
everything that I see come from?’ ‘Where did everything come from?’ he is
not asking for a (causal) explanation.

52 William H.Brenner

What I actually want to say is that here too it is not a matter of the words one
uses or of what one is thinking when using them, but rather of the difference
they make at various points in life…. Practices give words their meaning.2
(ROC, III, §317)

Focusing on the form of the words ‘Where did everything come from?’
makes us think of causal problems, just as (for example) the form of the
words ‘Orange is a blend of red and yellow’ makes us think of reports of
experimental results (compare LFM, p. 234). Thus the forms of language
lead us to misunderstand forms of life, human practices. And if we
‘surrender the reins to language and not to life, then the problems of
philosophy arise.’3
Consider the following Blue Book passage on the mind-body problem: ‘“Is
there then no mind, but only a body?” Answer: The word “mind” has a meaning,
i.e., it has a use in our language’ (BB, p. 69). Wittgenstein could have made a
similar remark about the philosophical problem of God and the world: ‘Is there,
then, no God, but only the world?’ Answer: the word ‘God’ has a meaning, a use
in our language. In each case, the answer is really a rejection of the question. The
question suggests that ‘mind’ and ‘God’ are names of hypothetical entities in
(apparently otiose) systems of causal explanation; it leads us into speculations
about the connection between ‘the world of experience’ and occult, super-
physical entities of which we know nothing. What Wittgenstein suggests is that
we need to look at the forms of life in which mental and theological words are
employed, and at what kind of employment they have there. We may find that
they have uses other than pre-scientific, pseudo-scientific, or superstitious uses.
One can use ‘God’—as a causal hypothesis—as (to use a figure from Karl Kraus)
one can use an urn for a chamber pot. But its distinctively religious, spiritual
employment lies elsewhere: ‘Religious faith and superstition are quite different.
One of them results from fear and is a sort of false science. The other is a trusting’
(CV, p. 72).4 Theological vocabulary can be used in a superstitious spirit as ‘false
science,’ but it can also be used in connection with learning, living, and teaching
distinctively religious (rather than secular) ways of life.5 In the latter case it is a
kind of grammar.
‘Theology is the grammar of the word “God”’ (AWL, p. 32).6 But we are
tempted to treat the sentences of theology as if they were not grammatical,
demanding them to be justified on the model of justifying sentences by
pointing out evidence. I think Wittgenstein would reject this demand as
contrary to the spirit and point of God-talk. For example, the sentence
‘We’re safe in the hands of God,’ as he understands it, is the expression of a
particular attitude towards the normal human preoccupation with keeping
safe and secure—the attitude some will express in the words of St Paul, that
‘in everything God works for good with those who love him.’7 But the point
of those words (as I think Wittgenstein and many Christians understand
them) is to express something absolute, some categorical rule and measure
of living and assessing life.8 Based on evidence, the confidence or trust it
Creation, causality, and freedom of the will 53

expresses would be (at best) contingent and hypothetical. Then ‘God’s love’
would signify an object measured, rather than a measure. ‘[Therefore,] if
there were evidence, this would in fact destroy the whole business’ (LC, p.
But if it is not evidence that leads us to trust in God, what is it that does?
Why take God-talk seriously at all? Wittgenstein suggests that certain
experiences might bring this about, for example: ‘sufferings of various sorts.
These neither show us God in the way a sense impression shows us an
object, nor do they give rise to conjectures about him. Experiences,
thoughts,—life can force this concept on us’ (CV, p. 86).9 Analogously:
although certain experiences force the concept color on us, they do not
show us color in the way a sense impression shows us an object. One does
not ‘convince someone of the reality of color’ by showing him colored
patches; for if he has no sureness in practice about that, he will not
understand what aspect of the patches you mean him to attend to. He is
convinced of its reality in and through his practice of the use of the ‘color
language’ he learned as a child.
Saying ‘God exists’ might have the same sort of use as saying that color
exists, namely as a grammatical remark made in the context of linguistic
instruction.10 What is at issue in such a context is whether a way of speaking,
doing, judging, and evaluating is being taught. The use of such propositions
is in teaching a way of judging existence, not in conveying an opinion about
the existence of something.
Wittgenstein suggests that what the theological proposition ‘God’s
essence guarantees his existence’ might really mean is that ‘what is here at
issue is not the existence of something’ (CV, p. 82). One could also say (he
continues) that the essence of color guarantees its existence: because all this
really means is that in this case (as opposed, say, to that of pink elephants)
there is no such thing as explaining what it would be like if colors were to
exist: ‘And now we might say: There can be a description of what it would
be like if there were gods on Olympus—but not: ‘what it would be like if
there were such a thing as God” (ibid.). The transcendent God of Judaism,
Christianity, and Islam is to be contrasted with the immanent gods of
ancient Greece and Rome: unlike the gods, God is not properly represented
as the personification of any of the powers of nature, not even the most
awesome and mysterious. ‘And to say this is to determine the concept
“God” more precisely’ (ibid.).11
Just before that line, Wittgenstein compared the reality of God with the
reality of color. Developing that striking comparison, I would say that while
the color system opens up a ‘logical space’ for describing the appearance of
objects, monotheistic systems open up a ‘logical space’ for challenging false
absolutes—whether they be the individual ego, a human collectivity, or
elemental powers of nature. And I would develop that idea by reference to
the following lines from St. Augustine’s Confessions, a spiritual classic
Wittgenstein is known to have pondered and respected:
54 William H.Brenner

I asked the winds that blow, and the whole air with all that is in it answered:
‘Anaximines was wrong; I am not God.’ I asked the heavens, the sun, the
moon, the stars, and they answered: ‘Neither are we the God whom you
seek.’ And I said to all the things that throng about the gateways of the
senses: ‘Tell me of my God, since you are not He. Tell me something of
Him.’ And they cried out in a great voice: ‘He made us.’ My question was
my gazing upon them, and their answer was their beauty.12

In thus having the powers of nature cry out in a great voice ‘He made us,’ St.
Augustine was simultaneously affirming the beauty of the world and
distancing himself from it; he was expressing a way of living, sometimes
expressed in the words ‘in the world and yet not of it.’ Wittgenstein may
have been trying to express a similar ideal when, in a 1929 journal entry, he
exclaimed: ‘Just let nature speak and acknowledge only one thing higher’
(CV, p. 1). Do we not hear in this an echo of Augustine’s ‘Tell me of my God,
since you are not He’?

The Wittgenstein notes published as ‘Cause and Effect: Intuitive Awareness’
contain a valuable supplement to David Hume’s classic account of our
concept of causality. As Hume pointed out one of its instinctive roots, our
tendency to trust experienced regularities, Wittgenstein points out another, a
certain tendency to look from one thing to another:

Calling something ‘the cause’ is like pointing and saying ‘He’s to blame!’
We instinctively get rid of the cause if we don’t want the effect. We
instinctively look from what has been hit to what has hit it. (I am
assuming that we do this.)
Someone touches me with a pole, I look along the pole. Someone throws
a stone, I feel it and see him in a particular position, I throw it back. This
is a reaction against a cause.
(PO, pp. 373, 410)

I want to connect that passage with the following from Culture and Value:
‘To what extent do we hold ourselves responsible for the future?…If
something unwelcome happens:—do we ask “Whose fault is it?”, do we say
“It must be somebody’s fault,”—or do we say “It was God’s will”, “It was
fate”?’(CV, p. 61).
Asking for the cause of an event and insisting on an answer is expressive
of a different attitude, a different mode of life, from not asking it. Here I
think of an episode from the popular 1994 movie Pulp Fiction. Not even
grazed by a volley of bullets directed at him and his friend, the main
Creation, causality, and freedom of the will 55

protagonist expresses that ‘different attitude’ by calling it a miracle, thanking

God for it, and changing his life accordingly. He might have reacted quite
differently, shrugging it off as no more than the lucky outcome of events
explainable by careful investigation of the ballistic facts (compare LC, pp. 56
and 61). Reacting to an event as the will of God is entirely different from
reacting to it as a curious phenomenon with no obvious explanation;
experiencing a surprising event as religiously significant—‘miraculous’—is
entirely different from observing it from a scientific, that is, a causal, point of
view. ‘Fate is the antithesis of natural law’ (CV, p. 61). We are not meant to
fathom fate and make use of it, as we are causal laws. The same point would
apply to the will of God. As noted above, Wittgenstein defines superstition as
a false science resulting from anxiety, contrasting it with genuine religious
faith, which he calls ‘a trusting’ (CV, p. 72). So, I take it, he would refer to
those who do nevertheless try to fathom and utilize ‘the divine will,’ or ‘the
book of fate,’ as superstitious. In their non-superstitious employment, the
sentences ‘It was God’s will’ and ‘It was fate’ would be ways of expressing
the rejection of a life of anxious striving, and a drawing of limits to one’s
efforts to acquire and utilize knowledge of the causes of things.13
How then are believers who place their trust in the divine will to conceive
the relation between Creator and creation? Comparing God to a potter and
creatures to clay, St Paul asked, rhetorically: ‘Hath not the potter power over
the clay?’14 But what consequences are we to draw from that? Are we to
blame our Creator for our vices? Here—and in the spirit of theology’s via
negativa—Wittgenstein would remark that all religious similes, even those
authorized by the Bible, ‘move on the edge of an abyss’ (CV, p. 29). The
example he uses is from John Bunyan’s Pilgrim’s Progress, where God is
compared to a Lord of the Road. On hearing the story of Christian’s
experiences along the road, our first thought might be:

‘[A]nd all these traps, quicksands, wrong turnings, were planned by the
Lord of the Road, and the monsters, thieves and robbers were created by
him?’ Certainly, that is not the sense of the simile! But such a continuation
is all too obvious! For many people, including me, this robs the simile of
its power.
But more especially if this is—as it were—suppressed.15
(CV, p. 29)

When the limitation of the simile is suppressed, it is liable to reemerge in the

form of ‘the problem of evil’ (How can the world around us be the creation
of a compassionate Lord?) and ‘the problem of freedom and predestination’
(How can the good Lord possibly damn some who by their own strength can
do nothing but sin?). If these problems are not to drain the power from the
simile for the would-be believer, he or she has to be told, frankly, that at
certain points it is not to be given the application it will naturally, in human
terms, seem to demand:
56 William H.Brenner

Someone can be told for instance: ‘Thank God for the good you receive
but don’t complain about the evil; as you would of course if a human
being were to do you good and evil by turns.’ Rules of life are dressed up
in pictures. And these pictures can only serve to describe what we are to
do, not justify it. Because they could provide a justification only if they
held good in other respects as well. I can say: ‘Thank these bees for their
honey as though they were kind people who have prepared it for you’;
that is intelligible and describes how I should like you to conduct
yourself. But I cannot say: ‘Thank them because, look, how kind they
are!’—since the next moment they may sting you.16
(CV, p. 29)

The work done by the sentence ‘All good gifts come from on high,’17 or
something like it, could also be done by a command, including one you give
to yourself. And, conversely, the utterance of a command such as Thank God
for the good you receive but don’t be resentful about the evil,’ may be like the
expression of a belief. Such a belief is not, however, a proposition about which
one can comfortably say: ‘If so and so is discovered, then it’s true or probably
true; if they discover such and such, then it’s false or probably false.’ It is the
expression, rather, of ‘a truth to live by.’ Compare what Wittgenstein said
about belief in a Judgment Day:

If there were evidence, this would destroy the whole business. Anything
I normally call evidence wouldn’t in the slightest influence me. Suppose
we had the best scientific evidence that certain people can foresee the
future, and that these people described some sort of Judgment Day:
queerly enough, belief in this happening wouldn’t be at all a religious
belief. A religious belief might in fact fly in the face of such a forecast,
and say ‘No. There it will break down.’18

A religious belief is the expression not of a belief founded on evidence but of

a way of living and assessing life—a way that sets limits to beliefs founded on
evidence and on the practices in which they are embedded.

It is a most important fact of human nature that a certain image can force
itself on you, an image generally connected with a particular way of acting.
(PO, p. 435, paraphrased)19

The word ‘God’ and the picture of a great and powerful Lord sometimes
pointed to in explaining it are nothing but instruments, and everything
depends on their use (RPP, I §586). For example, consider their use in
modifying our reaction to successfully overcoming ‘overwhelming obstacles.’
Creation, causality, and freedom of the will 57

Where before we would have said ‘Nobody helped us; we must be stronger
than we thought!’, now we say ‘God helped us!’ Although the sentences ‘God
helped us’ and ‘Nobody helped us’ may be used in precisely the same
circumstances, they are not to be construed as just two ways of saying the
same thing. For they have completely different uses—the first to express a
reaction to a situation, the second to describe it.20
‘Our help is in the name of the Lord, who made heaven and earth.’21 Some
of us came to use prayers like that to express what Wittgenstein referred to as
the experience of feeling absolutely safe: ‘the state of mind in which one is
inclined to say “I am safe, nothing can injure me whatever happens”’ (PO,
pp. 41–2).22 Saying this need not be understood as the expression of a childish
or insane sense of invulnerability; it may (and sometimes should) be
understood as the paradoxical expression of a particular attitude toward the
natural human desire for safety—the attitude that ‘in a deeper sense’
something else is more important: ‘Religion is, as it were, the calm bottom of
the sea at its deepest point, which remains calm however high the waves on
the surface may be’ (CV, p. 53). It is on the ‘calm bottom’ of his life that a
believer finds the ‘something else more important’—‘the peace of God that
passes all understanding.’
‘I am carnal…. [T]hey that are in the flesh cannot please God’ (Romans
8:14–15). I take it that Paul was expressing the sort of experience
Wittgenstein would have called ‘feeling absolutely guilty’ (PO, p. 42).23
Such words do not necessarily signify a pathological lack of self-esteem. In
certain circumstances they function as the paradoxical expression of an
attitude toward the normal desire in any individual to feel pride in herself
or himself—the attitude that in a ‘profounder sense’ what one needs is to
humble oneself. And here I think of Wittgenstein’s conception of
Christianity as religion for the individual who ‘feels infinite distress and
needs infinite help’:

Someone to whom it is given in such distress to open his heart instead of

contracting it, absorbs the remedy [prescribed by the Christian religion]
into his heart. Someone who in this way opens his heart to God in
remorseful confession opens it up to others too. He thereby loses his
dignity as someone special…and acknowledges as it were that we are all
wicked children…. Of course you must continue to feel ashamed of
what’s within you, but not ashamed of yourself before your fellow
human beings.24

In this way you can say it is very important what others think of you, but
that in another, profounder way it doesn’t matter at all.

The practice of judging people, of sizing them up morally, is quite an
58 William H.Brenner

important part of everyday life. And when we engage in it, we are supposed
to distinguish what people choose to do from what merely happens to them,
and to judge them only in terms of the former.
‘The process we call “choosing” does take place.’25 But couldn’t all our
‘choices’ be illusory? Imagine, for example, knowing that someone is acting
under post-hypnotic suggestion: wouldn’t that be definite grounds for saying
that he only believes he chooses? (A court of law might excuse a defendant
from responsibility for his act on such grounds.26) Such a use is necessarily
exceptional; for its point is to distinguish this sort of case from normal cases.
An individual’s sincere avowal normally settles questions about his choices.
‘I chose to sit in the front rather than the back’—it is only in special
circumstances that you can sensibly respond ‘You didn’t really choose.’
Although the case of ‘choosing under post-hypnotic suggestion’ would be
quite an abnormal one, certain experiences might incline us to compare more
ordinary cases with it, or with something like it. For example, learning the
results of the ‘cognitive biases’ research in social psychology might
significantly enlarge the range of cases for which it seems appropriate to
exclaim: ‘It’s as if these people have been programmed to react and to choose
as they do!’
Suppose we knew all natural laws and the working of people’s cells acting
on one another, and were able to calculate their choices from moment to
moment: it would not follow that we ought to stop holding them responsible
for their choices and punishing them for some of the things they do (LFM, p.
242, and PO, pp. 429–4427). Such an eventuality might nonetheless affect
the point of the concept of choice in our thinking, however, and the spirit of
the relevant language game. Compare James 4:14–16 (KJV):

[Y]e know not what shall be on the morrow. For what is your life? It is
even a vapor that appeareth for a little time, and then vanisheth away. For
that [reason] ye ought to say, If the Lord will, we shall live, and do this, or
that. But now you rejoice in your boastings: all such rejoicing is evil.

Should we acquire the power to forecast ‘the morrow’ much better than we
can now, that might destroy the point of comparing our life to a vapor. For,
in those circumstances, the temptation to trust—not in the Lord but in science
and technology—would likely be even greater than it is today.
‘People are determined’: this might be the expression of a confused
assimilation of causal determination to compulsion. In the everyday practice
of language, however, saying of an agent ‘He was determined’ might function
to register our unwillingness to pronounce judgment on his action. ‘The thief
moves as inevitably as a stone falling.’ This might mean that (for some reason
or other) you don’t want to be harsh in your judgment of him. ‘My God, I’m
like a falling stone!’: overhearing somebody say this, you might conjecture
that he has been brooding about the course of his sorry life and now expresses
his hopelessness and despair in these words; or you might react with the
Creation, causality, and freedom of the will 59

thought, ‘I know just what he means!’ (Cf. PO, p. 439).

Shall we join with ‘soft determinists’ in proclaiming the compatibility of
free will with determinism? That would be too easy. For, while there is a
scientific-philosophical use in which ‘determined’ is logically compatible with
‘free’ (where ‘determined’ just means ‘explainable by citing a set of natural
laws’) (PO, pp. 429ff.), there is also an everyday use of ‘determined’ whose
whole point is to deny free will. For example, the point of calling the people
described below determined would be to deny that they are free and
responsible agents:

Life is like a path along a mountain ridge; to the left and right are slippery
slopes down which you slide without being able to stop yourself…. I
keep seeing people slip like this and I say ‘How could a man help himself
in such a situation!’ And that is what ‘denying free will’ comes to. That is
the attitude expressed in this ‘belief’. But it is not a scientific belief and
has nothing to do with scientific convictions.
Denying responsibility is not holding people responsible.
(CV, p. 63)

‘He hasn’t given himself weakness and strength. “Hath not the potter power
over the clay?”’ (PO, p. 437).28 We might say this when—stepping back from
our habitual immersion in the cares of daily life—we find ourselves unable to
judge other people. But we might also dismiss such scruples, especially in
cases we feel obliged to judge: ‘Perhaps nothing better could have been
expected of him, given the kind of man he is; but he should nevertheless take
responsibility for himself and his actions, and we must nevertheless hold him
responsible for his choices.’29
We can cite many reasons—moral, political, and psychological—for doing
this. It could be (and has been) denied, however, that any of these
‘commonsense’ reasons—good as they are within limits—is valid without
qualification. To transpose a theme from On Certainty: as soon as we take
the sentence ‘He’s responsible’ out of its everyday practical contexts, it
appears in a false light and sounds presumptuous—as if God himself can’t
say anything about it. It is as if ‘responsible’ did not tolerate a metaphysical
emphasis.30 And—drawing now from Martin Luther—it is also as if ‘freedom
of the will’ did not tolerate a theological emphasis, for:

[T]hough I should grant that free will by its endeavors can advance…
unto the righteousness of the civil or moral law, it does not advance
towards God’s righteousness, nor does God in any respect allow its
devoted efforts to be worthy unto gaining his righteousness; for he says
that his righteousness stands without the law….31
Do they not make God a respecter of merit and persons when they say
that one man is without grace by his own fault…?32
60 William H.Brenner

‘I think yesterday was a crisis in my life,’ wrote William James in his diary:

I finished the first part of Renouvier’s second ‘Essais’ and see no reason
why his definition of free will—‘the sustaining of a thought because I
choose to when I might have other thoughts’—need be the definition of
an illusion. At any rate, I will assume for the present—or until next
year—that it is no illusion. My first act of free will shall be to believe in
free will.33

I suppose that while James would advise us to take hold of the idea of free
will as an instrument of deliverance from despair, Luther (following St Paul)
would admonish us to pray for divine grace—and to thank God when
deliverance comes. Being religious is not the same as being self-assertive or
self-reliant. It is more (as I want to put it) a ‘theological suspension of self-
‘In religion every level of devoutness must have its appropriate form of
expression which has no sense at a lower level,’ remarked Wittgenstein. ‘At
my level the Pauline doctrine of predestination is ugly nonsense,
irreligiousness,’ he continued (CV, p. 32). We can imagine that neither
Wittgenstein nor James would have been ready to hear Luther’s Pauline
admonition and to use it in the good and godly way for which it was doubtless
Believers in divine providence and predestination do not, of course, join
with fatalists in denying the importance of individual initiative, or in ruling
out the everyday human practice of holding ‘normal adults’ responsible for
their actions. What they want is to set limits to those practices, and to the
(often judgmental) attitudes that animate them—so that secular life does not
become a cage for us, separating us from everything sacred. What is wanted
is access to a place where we can stand at a contemplative distance from
‘common life and practice,’ and—like Arjuna before the great battle35—view
it all sub specie aeternitatis:

(The thought forces itself upon one): The thing seen sub specie
aeternitatis is the thing seen together with the whole logical space.
As a thing among things, each thing is equally insignificant; as a world,
each one equally significant.
(NB, p. 83)36

Our freedom and moral responsibility distinguish us from animals incapable

of reason and deliberation. Some of us are honored for developing rational
self-control, others despised as slaves of passion or social pressure; some of
us take responsibility for our actions, others habitually make excuses.
Religion teaches a way of ‘flying above’ these and other everyday distinctions
and of viewing every life, higher and lower, as ‘a work of art created by God
and, as such, certainly worth contemplating’ (CV, p. 4). It is from this
Creation, causality, and freedom of the will 61

contemplative viewpoint, when we have retreated for a time from ‘the battle
of life,’ that the world is revealed as a ‘limited whole’ independent of our
will. Then we feel that, while the facts of the world are sometimes subject to
causal analysis and human control, the existence of the world is always (as it
were) ‘dependent on an alien will’ and ‘mystical’ (see NB, p. 74, and TLP
Friedrich Waismann once asked Wittgenstein whether there is a connection
between the existence of the world and the ethical, to which Wittgenstein

Men have felt that here there is a connection and they have expressed it
thus: God the Father created the world, the Son of God (or the Word
that comes from God) is that which is ethical. That the Godhead is
thought of as divided and, again, as one being indicates that there is a
connection here.
(WWK, p. 118)

The Son reconciles us to God, teaching us to address him as ‘Father,’ rather

than ‘Alien Will.’ The Word that comes from God is (let us suppose): ‘Love
your enemies and pray for those who persecute you, so that you may be
children of your Father who is in heaven; for he makes his sun rise on the evil
and on the good, and sends his rain on the just and on the unjust.’37 We are to
return from our contemplative retreat to active participation in the battles of
life with a less ruthless, more compassionate view of all our fellow combatants.
The Word that proceeds from the Father, the ethical, is that we should shine
the light of our attention and concern on everyone we meet—not keeping it
under a legalistic basket and letting it shine only on those ‘with a right to it.’
‘How things are in the world is a matter of complete indifference for what
is higher…’ (TLP 6.432). Here God is revealed as transcendent Creator. But
through the incarnation of his Word, God is also revealed as the ethical—and
to that ‘the how’ cannot be of complete indifference. By following the Word,
we imitate our Creator at the human and the personal level. The sons of God
imitate their Father in heaven through the self-limiting impartial love of
neighbor taught by Jesus, and by other spiritual teachers as well. This love
combines the universality of justice with the passion of friendship into a
compassion that goes out to all our fellow creatures, always mindful that ‘All
flesh is grass, and all its beauty like the flowers of the field’ (Isaiah 40:6).
That is perhaps how Wittgenstein might have thought of the Godhead as
at first divided and then united. And perhaps the following beautiful lines
from Simone Weil will suggest how the thought might be developed and

God is absent from the world, except in the existence in this world of
those in whom His love is alive…. Their compassion is the visible
presence of God here below.
62 William H.Brenner

Compassion is what spans the abyss which creation has opened

between God and the creature…. [It] should have the same dimension as
the act of creation. It cannot exclude a single creature.
Every created thing is an object for compassion because it is

In an important passage from Culture and Value I made use of earlier,39
Wittgenstein remarked that the work done by the declarative sentence ‘It’s
God’s will’ can also be done by the commandment ‘Don’t be resentful’ or
‘Don’t grumble,’ and that conversely the commandment can be uttered like
the affirmation of a truth. That remark continues in the new, revised and
augmented, edition of Culture and Value as follows:

Now why am I so anxious to keep apart these [theological] ways of using

‘declarative sentences’ [from other, more familiar, uses]?…It is simply an
attempt to see that every usage gets its due. Perhaps then a reaction
against the over-estimation of science….40
But of course the words ‘see that they get their due’ & ‘overestimation’
express my point of view. I could have said instead: ‘I want to help this
& this to regain respect’; only I don’t see it like that.41

Although many of Wittgenstein’s remarks reveal a personal respect for

theological usages (and some, perhaps, a struggle to enter a particular
community of faith42), he should not be read as an apologist—as Simone Weil,
in many of her best-known writings, possibly should. Here as elsewhere,
Wittgenstein’s writings are essentially expressions of wonder at the existence
of language and attempts to understand it in all its variety.43
‘[I]f you want to stay within the religious sphere you must struggle’ (CV,
p. 86). It is also a struggle, a distinctively philosophical struggle, to keep the
variety of language in view—including the possibility of ‘paltry and neglected’
non-scientific, non-rationalistic uses. I have been elucidating one such use in
this paper.
To sum up: ‘God created the world’ may be the expression, not of a
(would-be scientific) causal belief, but of a way of living and assessing life
that sets limits to all causal beliefs and to the human practices in which they
are embedded. One such practice, that of sizing ourselves up morally, requires
distinguishing what we ‘freely cause to happen’ from what ‘just happens to
us.’ Now while believers of the kind I have in mind have no wish to deny the
importance of this everyday practice, they do want to set limits to it—and to
the judgmental attitude naturally animating it. They find that through ‘trust
in the Creator,’ they are given a place from which to view this and every form
Creation, causality, and freedom of the will 63

of common life and practice from a contemplative distance, and given, finally,
the grace to return to active participation in life with a less ruthless attitude
towards themselves and others.44

1 Ludwig Wittgenstein: A Memoir, 2nd edn. (New York: Oxford University Press,
1984), p. 59.
2 Compare ROC, I, §6: ‘[H]ere language games decide.’
3 Quoted in Nicholas Gier, Wittgenstein and Phenomenology (Albany, NY:
SUNY Press, 1981), p. 70, from Wittgenstein’s ‘Big Typescript’ (MS 213, p.
4 Compare the Gospel parable of the lilies of the field and the birds of the air.
5 Compare Socrates’ use of religious words with Euthyphro’s.
6 Wittgenstein attributes this conception of theology to Martin Luther. While the
Luther scholars I consulted recalled no textual confirmation of Wittgenstein’s
attribution, they did find it plausible. The closest text I could find was in vol. 26
of Luther’s Works, ed. Jaroslav Pelikan (Philadelphia: Fortress Press, 1967).
Commenting on Galatians, Luther remarks that ‘doing’ and ‘making’ become
completely new words in theology and acquire a new meaning: ‘When you read
in Scripture, therefore, about the patriarchs, prophets and kings that they worked
righteousness, raised the dead, conquered kingdoms, etc., you should remember
that these and similar statements are to be explained according to a new and
theological grammar…’ (p. 267).
7 Romans 8:28 (RSV).
8 It is this ‘absolute’ and ‘categorical’ use that turns the original, presumably
spontaneous utterance of the saint into a grammatical rule. Compare RFM, p.
437: ‘Every empirical proposition may serve as a rule if it is fixed, like a
machine part, made immovable, so that now the whole representation turns
around it and becomes part of the coordinate system, independent of facts.’
Cf. OC§98.
9 On the cryptic concluding line of that remark—‘So perhaps [the concept “God”]
is similar to the concept “object”’—, see my Wittgenstein’s Philosophical
Investigations (Albany, NY: SUNY Press, 1999), p. 143, note 5.
10 See OC§36: ‘“A is a physical object” is a piece of instruction which we give only
to someone who doesn’t yet understand either what “A” means, or what
“physical object” means. Thus it is instruction about the use of words, and
“physical object” is a logical concept. (Like color, quantity,…).’ Compare Z,
§477, and PR, p. 72.
11 The God of Micah, Jesus, Augustine, Anselm, Aquinas, Descartes, etc., is on a
different logical level from that of any cause postulated in any science—or in any
pseudoscience, for that matter. On this, and on the evolution of the concept of
God in the Bible from tribal deity who walks in the garden with Adam to ‘high
and lofty one who inhabits eternity,’ see B.F.Tilghman. An Introduction to the
Philosophy of Religion (Oxford: Blackwell, 1994), pp. 26–34.
12 From Augustine’s Confessions, X:6, translated by F.J.Sheed (Indianapolis, IN:
Hackett, 1993). Whatever else we may want to say about it, the idea of a Creator
in that passage surely involves no ‘repugnantly rationalistic’ use of the concept
of causality. (Cf. my ‘Chesterton, Wittgenstein, and the Foundations of Ethics,’
Philosophical Investigations 14 (1991):311–23.) ‘But is it true that nature is
created, and that she speaks to us of her creator?’ While I think I know how to
show how these paradoxical word formations—‘maker of nature,’ ‘speech of
64 William H.Brenner

nature’—can come to have a meaningful and indeed fundamental place in a

person’s thought, I do not think I know what to make of that objection.
13 Theologians will of course want to stress the difference between their way and
fatalism: to explain—as I will be only suggesting—how it accords greater scope
and dignity to individual choice than does fatalism, and how it is a way of
spiritual pilgrimage rather than of quietism or inertia.
14 Romans 9:21 (KJV), emphasis added.
15 This passage makes me think of what Wittgenstein says about ‘hidden
contradiction’—for example in LFM, pp. 209–11.
16 Compare CV, p. 77:

Just look at an allegory like The Pilgrim’s Progress and notice how nothing
is right—in human terms…. (On railway stations there are dials with two
hands; they show when the next train leaves, they look like clocks though
they aren’t; but they have a use of their own.)…If anyone gets upset by this
allegory, one might say to him: Apply it differently, or else leave it alone!

17 James 1:17 (KJV).

18 LC, p. 56, condensed.
19 Compare CV, p. 20: ‘A preference for certain similes could be called a matter of
temperament and it underlies far more disagreements than you might think.’
20 This is an instance of what Wittgenstein calls ‘the use of a word in a secondary
sense.’ See Philosophical Investigations, §282 and p. 216, and my discussion in
Wittgenstein’s Philosophical Investigations, pp. 74, 144–51. Compare Gareth
Moore, Believing in God: A Philosophical Essay (Edinburgh: T. & T.Clark,
1988), passim—a neglected work to which I am much indebted for my
understanding of Wittgenstein and religious uses of language.
21 Psalms 124:8 (RSV).
22 From Wittgenstein’s ‘Lecture on Ethics.’ See Ray Monk, Wittgenstein: The Duty
of Genius (New York: The Free Press 1990), p. 116.
23 ‘Lecture on Ethics.’
24 Here I quote from p. 52 of the revised edition of CV, ed. Alois Pichler (Oxford:
Blackwell, 1998), because here I prefer Winch’s new translation. (The passage is
on p. 46 of the 1980 edition.)
25 From the Wittgenstein Nachlass, Ms 115, quoted in Stuart G.Shanker,
‘Wittgenstein versus James and Russell on the Nature of Willing,’ in John
Canfield and Stuart G.Shanker (eds),’ Wittgenstein’s Intentions (New York:
Garland, 1993), p. 226.
26 Shanker mentions a case in Canada of a man’s being acquitted of homicide on
the grounds of somnambulism, ibid., p. 239.
27 ‘Lectures on Freedom of the Will.’
28 Cf. Romans 9:21.
29 Compare O.K.Bouwsma, Wittgenstein: Conversations 1949–1951
(Indianapolis, IN: Hackett, 1986), pp. 15–16.
30 Compare with:

If I say, without any special occasion, ‘I know I’m sitting in a chair’, this
seems to me unjustified and presumptuous. But in its language game it is
not presumptuous: there it has no higher position than, simply, the human
language game, for there it has a restricted application. But as soon as I
make the statement out of its context, it is as if I wanted to insist there are
things I know—things even God himself can’t say anything to me about.
(OC §§553–4, condensed)
Creation, causality, and freedom of the will 65

It is as if ‘I know’ did not tolerate a metaphysical emphasis.

(OC §482)

31 Erasmus-Luther: Discourse on Free Will (New York: Ungar, 1961), p. 135. For
an illuminating Wittgensteinian reading of Luther on predestination, see John
H. Whittaker, Matters of Faith and Matters of Principle (San Antonio, CA:
Trinity University Press, 1981), pp. 69–91.
32 Martin Luther, The Bondage of the Will (Grand Rapids, Mi.: Fleming H.Revell,
1957) pp. 292–3. Compare O.K.Bouwsma’s Wittgenstein: ‘In the context of the
Gospel injunction “Be ye perfect”—I ought does not imply I can. I take it that
the proper response to that injunction would be to try and keep trying—and to
learn humility from your inevitable failures’ (pp. 37–9, condensed). And compare
that with Wittgenstein’s own—I think somewhat less incisive—reflections on that
topic in CV, p. 54.
33 The Letters of William James, ed. Henry James, (Boston, MA: Atlantic Monthly
Press, 1920), p. 148.
34 Compare Kierkegaard’s ‘teleological suspension of the ethical’ in Fear and
35 See the Bhagavad Gita, passim.
36 Compare CV, p. 80: ‘The Sabbath is not simply a time for rest, for relaxation.
We ought to contemplate our labours from without and not just from within.’
37 Matthew 5:43–47 (RSV).
38 Simone Weil, First and Last Notebooks, trans. R.Rees (London: Oxford
University Press, 1970), pp. 102–03.
39 Although I did not quote the passage, I made use of a modified version of its
contents on p. 85.
40 Compare CV, pp. 60–1 (older edition): ‘Science: enrichment and
impoverishment. One particular method elbows all others aside. They all seem
paltry by comparison, preliminary stages at best. You must go right down to the
original sources so as to see them all side by side, both the neglected and the
41 CV (1998 edition.), p. 70 (not present on the corresponding page, p. 61, of the
1980 edition). Compare ‘Lectures on Freedom of the Will’: ‘All these arguments
might look as if I wanted to argue for the freedom of the will or against it. But I
don’t want to’ (PO, p. 436).
42 I am thinking especially of CV, p. 33, when he asks: ‘What inclines even me to
believe in Christ’s Resurrection?’
43 See Philosophy in a Cool Place (Ithaca, NY: Cornell University Press, 1999),
where D.Z.Phillips develops this important Wittgensteinian theme at length. And
recall that in ‘A Lecture on Ethics,’ Wittgenstein speaks of being tempted to say
that the right expression for the miracle of the existence of the world is, not any
proposition in language, but the existence of language itself (PO, pp. 43–4).
44 For their many helpful comments, I want to thank Robert Arrington, Curtis
Brooks, and Mary Brenner.
5 Faith
Themes from Wittgenstein,
Kierkegaard and Nietzsche
Michael P.Hodges

My subject is the nature of religious faith and particularly what can be said
for its claims. I begin by laying out the route that the argument will follow.
First, using some of Wittgenstein’s remarks I want to disentangle the claims
of faith from dubious historical and metaphysical claims. I am not at all
interested in defending a faith that would require that we believe what, by
normal standards of evidence, are highly problematic historical assertions,
nor do I think that the legitimacy of faith should be a matter of metaphysical
truths which lie just a bit beyond our normal capacities for verification. To
be very clear, if one cannot disentangle faith from such understanding, I think
its claims do not deserve commitment. So after offering such an analysis, I
propose to move to another stage. Wittgenstein’s discussion focuses us on a
sort of authority, and to develop more completely that idea we need to turn
to Kierkegaard’s justly famous account of Abraham in Fear and Trembling.
Kierkegaard leads us through an examination of faith that places it outside
the confines of rationality conceived as prudence or self-interest, even outside
the ethical understood as a form of tragic heroism. For Kierkegaard, faith is
a matter of an absolute relation to the absolute. As such, there cannot be
reasons for such a relation outside of that relation itself. Now it might seem
that given such an account of faith, it will be beyond criticism, and in a certain
sense it is—but, at the same time, I hope to show that there are at least three
critical strategies that can be deployed. One is the traditional strategy of
genealogy, for which I turn to Nietzsche. A second is the strategy of
reinscribing one discourse within another—the religious, say, within the
ethical. The third strategy is in some ways a more radical approach that
depends on the ‘claim’ that all structures of meaning are contingent. To
understand this last claim, I again take up certain remarks of Wittgenstein
and follow out their implications for faith and the religious.
First, let me explain how I conceive my task here. Certainly I do not
propose to evaluate particular religious claims. For example, I do not propose
to adjudicate the dispute between Anglicans and Baptists over the question
of total immersion or between Christians and Jews concerning the divinity of

Themes from Wittgenstein, Kierkegaard and Nietzsche 67

Jesus. As I understand them, these are differences of belief which happen

within religious life. These are differences which inform the lives of the
variously religious, and they must be discussed and settled, if they can be, at
that level. They are roughly analogous to the disputes between scientists over
the particulars of sub-atomic theory. 1 Philosophers have no special
competence to adjudicate either scientific or religious disputes at this level.
However, the utterances of religious believers contain not only specifically
religious claims, e.g. ‘My God is with me most closely in the most difficult of
times’ and ‘Praise be to God,’ but also reflections on the status of the former
claims. Here I would include claims that may bring religious assertions into
conflict with scientific or historical claims. So, for example, those who
contend that we must either accept the Genesis account of creation or the
theory of evolution, but not both, express not only a religious commitment
but also a philosophical view of the nature and status of that commitment.
My own discussion is located on this ‘second-order’ level, and that means
that what I will have to say will offer a way of understanding religious
assertions and will therefore come into conflict with other consciously or
unconsciously held ways of understanding religious claims. But, I hope, it
will not come into significant conflict with religious claims themselves as they
actually function in the conduct of our religious lives.2
It is often claimed that religious truths must be believed ‘on faith.’ It is
said that this special ‘mode of apprehension’—faith—can fill the gap between
evidence (or counter-evidence) and truth, yielding legitimate cognitive assent.
‘Maybe I don’t have reasons but I have faith,’ the zealot says, and that puts
an end to all the rational challenges to religious belief. What seems both so
appealing and so disconcerting about this line of argument is that it lays claim
to victory while abandoning the race altogether. That is, it claims cognitive
success but refuses to be judged by its usual standards.
There is behind all this a certain picture of what faith must be. The claims of
religion assert historical, physical or metaphysical truths which unfortunately
fall ‘slightly’ beyond the normal cognitive powers of reason. However, each of
us is in possession of a faculty of nonrational cognitive assent by means of
which we can travel the intervening distance. This faculty, although different
from reason and other standard evidence-gathering faculties, is a fully
legitimate mode of knowing. It yields apprehension of the truth just as surely
as, perhaps more surely than, any other human cognitive capacity.
This ‘cognitivist theory of faith’ has as its natural corollary a view of the
nature of religious truth. As Rush Rhees puts it, ‘the language of religion is
seen as in some way comparable with the language in which one describes
matters of fact.’3 Since the language of religion is supposed to express facts or,
better, super-facts about entities constituting a realm beyond that of ordinary
experience and just out of reach of our ordinary capacities, faith must be
conceived as a faculty or super-faculty which puts us in touch with those facts.
Wittgenstein explicitly rejects the picture of religious believing that I have
just outlined. In ‘Remarks on Frazer’s Golden Bough,’ he says: ‘No opinion
68 Michael P.Hodges

serves as the foundation for a religious symbol. And only an opinion can
involve an error’ (PO, p. 123). He develops this fundamental insight in a
passage written in 1937:

Christianity is not based on a historical truth; rather, it offers us a

(historical) narrative and says: now believe. But not, believe this narrative
with the belief appropriate to a historical narrative, rather: believe, through
thick and thin, which you can do only as the result of a life. Here you have
a narrative, don’t take the same attitude to it as you take to other historical
narratives! Make a quite different place in your life for it.
(CV, p. 32)

Wittgenstein places ‘historical’ in parentheses in this passage so as to bracket

its usual implications. The point is that while the narrative may be historical,
that is not relevant for its religious function. More will be said about this point
shortly. First, we should note that the nature of the believing is somehow
different for religious truth than for ordinary historical narrative. In Lectures
and Conversations Wittgenstein points out that people who have faith don’t
apply the doubt which they would ordinarily apply to historical propositions,
especially propositions of a time long past. There is a natural skepticism about
ordinary historical claims for which we have very little independent evidence,
especially when such claims are filled with specific detail and removed to the
far distant past. In fact, those scholars, such as Ernest Renan and Albert
Schweitzer, who have attempted to assess the historical evidence have
uniformly concluded that the evidence is slim indeed.4 Interestingly, however,
such conclusions do not dampen their level of religious commitment. And, at
the same time, Wittgenstein also says ‘indubitability is not enough in this case.
Even if there is as much evidence as for Napoleon. Because the indubitability
wouldn’t be enough to make me change my whole life’ (LC, p. 57). Belief as
faith is not located on the usual continuum between indubitability and normal
doubt, simply because religious believing is not an epistemic state. However,
the believer ‘has what you might call an unshakeable belief. It will show, not by
reasoning or appeal to ordinary grounds for belief but rather by regulating for
in all his life’ (LC, p. 54).
We must be careful here not to confuse Wittgenstein’s view with an all too
familiar thesis. Priests and ministers often tell us that true religious
commitment is not a matter of mere intellectual assent—something else must
be grafted onto that assent, some change in our way of life. However, this
position does not really reject the picture of religious truth underlying the
cognitivist theory of faith. It merely tries to add another element to
intellectual assent. Faith is intellectual assent together with some change of
life. But the two are externally related so that one brings about the other.
That is, it is because we grasp certain truths that we are to change our ways.
For Wittgenstein, none of this will do. His more radical position becomes
clear in the following passage:
Themes from Wittgenstein, Kierkegaard and Nietzsche 69

Queer as it sounds: The historical accounts in the Gospels might,

historically speaking, be demonstrably false and yet belief would lose
nothing by this: not however, because it concerns ‘universal truths of
reason’! Rather because historical proof (the historical proof-game) is
irrelevant to belief. This message (the Gospels) is seized on by men
believingly (i.e. lovingly). That is the certainty characterizing this
particular acceptance-as-true, not something else.
(CV, p. 32)

This is a radical claim—the Gospels might be historically false and belief

would lose nothing. What does Wittgenstein mean? To answer this question,
we must draw a distinction between the Jesus of the Gospels and any
historical person or persons that stand behind those texts. Jesus is present in
the Gospels, not merely described by them. For the believer, it is the Jesus
revealed in and through the Gospels who transforms human life and with
whom one can come into relation. Whatever historical elements are or are
not included recede into irrelevance. Wittgenstein does not assert or deny the
historical truth of the Gospels. He claims that historical truth is irrelevant to
the ‘particular acceptance-as-true’ which is religious believing.
Both poles of the believing situation have now been transformed. Not only
is it wrong to see the relation of the believer to the content of belief as one of
intellectual assent, but it is also wrong to see the content believed as
information, either historical or metaphysical.
John Dewey in his fine little book A Common Faith draws a very
important distinction between two notions of belief. He says: ‘Apart from
any theological context there is a difference between belief that is a conviction
that some end should be supreme over conduct, and belief that some object
or being exists as a truth for the intellect.’5 This distinction is marked in
ordinary language by way of the difference between ‘believing in’ and
‘believing that.’ I believe in democracy and I believe that Washington was the
first US president. While it is no doubt true that there are interesting and
important connections between what I believe in and what I hold to be true,
the two notions are quite distinct. Evidence, in the usual sense, is not always
relevant to believing in. In fact, one can, and perhaps often should, continue
to believe in something, as Wittgenstein puts it, ‘through thick and thin’ or
against all odds and in the face of the evidence. Of course, to believe in
something requires content, but that content is not mere information. We
believe in ideals, i.e. images of human possibility embodied in stories of
human interactions. In their role as transforming possibilities, the whole
question of their historical truth is simply displaced.
It is clearly this notion of belief (believing in) that Wittgenstein is getting
at when he says: ‘It strikes me that a religious belief could only be something
like a passionate commitment to a system of reference. Hence, although it’s
belief, it’s really a way of living, or a way of assessing life’ (CV, p. 64).
Wittgenstein’s reason for calling attention to the use of the term ‘belief’ in
70 Michael P.Hodges

religious contexts is to cancel our inclination to treat it as ‘intellectual assent.’

To believe, in this context, is not to accept certain propositions as true. Rather
it is to be passionately committed to a way of living. This seems to be of
overriding importance for Wittgenstein. Throughout his discussions he
focuses, over and over again, on the role which religious beliefs play in the
conduct of our lives. This means that our relation to religious commitments
is not epistemological but practical. In the face of a total lack of evidence, in
the usual sense, we are still willing to risk everything (see LC, p. 54).6 This
aspect of his view is brought clearly into focus by connecting Wittgenstein’s
talk of ‘passionate commitment to a system of reference’ to the notion of
belief as ‘believing in.’ To believe in something, in the religious sense, is
precisely to take it as a standard by which to conduct my life.7 There is no
logical gap between belief and action in this sense. To believe passionately in
democracy, is, among other things, to be willing to work to bring it about
and to assess the legitimacy of other governments by its standards.8
The ‘cognitivist theory of faith’ treats this aspect of religious believing as
an addendum to its ‘accepting as true’ approach. But for Wittgenstein
‘believing means submitting to an authority’ (CV, p. 45). To submit to an
authority is to accept its dictates, be they those of an individual (e.g. Jesus
Christ) or an ideal (e.g. Capitalism) as the final legitimate ground for action
or opinion. Here we are focused on basic or ultimate commitment and not
something more preliminary like ‘I believe in my stockbroker; after all, he
has never let me down.’ It is in this sense that it is utter nonsense to require
reasons for religious commitment. It is the commitment that gives me reasons;
I cannot have independent reasons for commitment. If I did, then such a
commitment would not be final but contingent upon some other ground.9 If
I have submitted to an authority, then I cannot ‘reserve the right to an
independent judgment.’ As Wittgenstein says: ‘Having once submitted, you
can’t then without rebelling against it, first call it in question and then once
again find it acceptable’ (CV, p. 45).
On what grounds could an independent judgment be made? Any grounds
which might lead to different actions or opinions would not merely be in
error, they would be evil. After all, from the point of view of the committed
Christian, those who would undermine the faith are leading people away
from God.
There are three interrelated points that Wittgenstein makes. First, faith is
not ‘intellectual assent’ in the light of evidence or by means of an extrarational
faculty. It is a ‘passionate commitment.’10 Second, by faith we do not stand in
relation to a ‘proposition’ as a picture of a fact or a conveyor of information.
Rather we are passionately committed to an authority as determinative of
our lives. Third, this authority does not have its authority in virtue of some
further authority (in which case it would be possible to call it into question
without ‘rebelling against it’). Religious language presents us with the
authoritative as immediately and intrinsically so. In this light it is interesting
to note a comment which Wittgenstein made in ‘On Schlick’s Ethics’:
Themes from Wittgenstein, Kierkegaard and Nietzsche 71

Schlick says that theological ethics contains two concepts of the essence
of the Good. According to the more superficial interpretation, the Good
is good because God wills it; according to the deeper interpretation, God
wills the Good because it is good. I think that the first conception is the
deeper one: Good is what God orders. For it cuts off the path to any and
every explanation ‘Why’ it is good…. The first conception says clearly
that the essence of the Good has nothing to do with facts and therefore
cannot be explained by any proposition. If any proposition expresses just
what I mean, it is: Good is what God orders.11

If we remember that the notion of ‘proposition’ operative in this passage is

that taken from the Tractatus, we can place matters in context.
To use familiar terminology, the language of religion is a very different
language-game from those languages in which the point is to state facts. In
the Tractatus Wittgenstein said: The general form of a proposition is: This is
how things stand’ (TLP 4.5). That is, the single role of language is to convey
information—to picture facts. Since religious language does not do that, it is,
on the early view, literal nonsense. What it attempts to say is beyond saying.12
However, in the later work, he rejects the idea that language has an essence
and with it the idea that we can specify a priori what can be said and what
cannot. We must look and see. What this means here is that the proof of the
possibility of sense is the actual use of the language. We look not at what
information is conveyed but at what role the language plays in the lives of
those who use it. Since the language of religion has a place in our lives, it
cannot be dismissed as meaningless nor can it be assimilated to some
preconceived model of sense.
The clue for developing a positive notion of religious believing lies in the
notion that such believing is ‘submitting to an authority’ and with the idea
that in religious language and experience authority occurs as immediate and
intrinsic. These notions are surely at the very center of Kierkegaard’s
discussion of Abraham, and so we need to turn there to continue our
Everyone knows the story of Abraham and God’s command that he should
sacrifice his son Isaac, so I will not take time to retell it. However, it is
necessary to understand the structure of Kierkegaard’s discussion of the story.
He takes, as a beginning point, the popular assessment of Abraham—that he
is the ‘father of faith.’ Now, of course, this is true in two senses. In the less
interesting sense, he is the progenitor of the faithful people through his
offspring. More importantly, Abraham is the father of faith by being the first
exemplar of a faithful one—what Kierkegaard calls ‘a Knight of Faith.’ The
issue for Kierkegaard is what we must be willing to say about Abraham if we
give him this praise. Kierkegaard’s argument is an ad hominem directed
against those who make the popular assessment. He says: ‘Let us then either
consign Abraham to oblivion, or let us learn to be dismayed by the
tremendous paradox which constitutes the significance of Abraham’s life…’13
72 Michael P.Hodges

Kierkegaard considers various possible accounts of the events of the story

but finds that each fails to capture the essence of Abraham as a Knight of
Faith. First we cannot see Abraham as a man of practical wisdom, for as such
he becomes either a faker or a monster who is willing to trade his son for his
own eternal well-being. On the other hand, if we evaluate him by normal
ethical standards he is at best a madman, at worst a willing murderer, saved
at the last moment by a lucky happenstance. Nor can we seek a larger ethical
context—some greater good that might be achieved by this unfortunate
action—in terms of which to understand Abraham’s actions. To attempt this
way out is to transform the ‘father of faith’ into a tragic hero. Finally, we
cannot suggest that Abraham’s intention to kill his son ought not to be judged
by present ethical standards. That interpretation totally destroys the
significance of the action. It may provide us with interesting historical or
sociological information, but we have nothing personal to learn from it. It is
axiomatic for an understanding of the story, as that of a Knight of Faith, that
it is ethically wrong to kill one’s son.
We seem forced to conclude that by ethical standards Abraham’s actions
are wrong and at the same time in so far as we believe, with Abraham, that
he is acting as God commands, we believe him to be acting, in some sense,
rightly. As members of the faithful we might wish never to be given
Abraham’s task, but if it comes to us we must wish that we can carry it out as
he does. The sense in which Abraham is acting rightly cannot be an ethical
one, so it must involve something which goes beyond the ethical. The essence
of faith is thus seen in an all-encompassing and completely overriding
‘submission to an authority’—a suspension of the ethical in favor of an
immediate commitment to the will of God as authoritative, or, as Kierkegaard
puts it, ‘a ideological suspension of the ethical.’
This is the basic structure of Kierkegaard’s discussion. But he leaves us
with what might appear to be an unanswered question. The whole story as it
occurs in the Bible begins, ‘And God tempted Abraham and said unto him
…’ How is it that Abraham recognized the voice that commanded him to
sacrifice Isaac as the voice of God? It is an unquestioned presupposition of
the story that Abraham is talking to God, and yet it is a point of absolutely
vital importance. Why doesn’t Abraham take the command to be a
temptation by the devil?14 Or why doesn’t he take it to be the voice of the
dark side of his own soul. In either case, the clearly correct course would be
to resist, to reject the false command, to refuse to give in to the primordial
desire to destroy his son. However, these possibilities are settled before the
story begins. This seems a strangely naive position, for surely the very content
of the command should make Abraham suspect its alleged source. How could
it be God who is now speaking when it was God himself who gave Isaac as
the fulfillment of his promise.
The natural way to understand the question of how Abraham recognizes
God’s voice is as a question about the objective source of the command. That
is, we take the question on the model of the question whether a voice that
Themes from Wittgenstein, Kierkegaard and Nietzsche 73

sounds like Richard Nixon is really the voice of Richard Nixon. But this leads
back in the direction of the ‘cognitivist theory of faith.’ Now we seem to be
confronted with a gap between what we can know by normal means and
what we need to know in this special situation. Only faith could fill that gap.
And this seems immediately to force two conclusions and one observation on
us. First, it ought to make sense to ask for evidence or reasons for supposing
the voice to be the voice of God. But, second, no evidence could possibly
assure Abraham of the objective source of the voice. No evidence could ever
be adequate to the level of commitment that is required of Abraham. And,
third, Abraham does not seek any such evidence.
Perhaps we will be inclined to avoid the problem by saying: ‘He just knows
it’s God’s, that is the very essence of faith.’ But this response leaves the poles
of the problematic unchanged. Belief is treated as intellectual assent and that
to which we give assent as information—a fact about the objective source of
the voice in question. Suppose, instead, that what distinguishes a divine
command is not the source of the injunction but the place which the
command has in the life of the person commanded. Faith is the context in
which the voice of God is heard—in which an authority has absolute and
overriding significance. On this view Abraham is not credited with a possibly
false belief concerning the source of an order. Abraham’s belief that this is
God’s command is simply his submission to it.15 That is, there are not two
things in this situation—first, an intellectual assent to a particular truth, and
second, a response based on that assent. In this way the components are
merely externally related. But such a bifurcation seems antithetical to
Kierkegaard’s own rejection of the idea that our relation to God is primarily
an epistemological one.16
How must we live in relation to a divine command? What place must we
make for it in our lives? Surely the point is that such a command must have
ultimate authority. It says what finally and ultimately ought to be done. There
can be no higher court of appeal here. So if, for example, one confronts a
divine command by attempting to place finite human judgment above it, one
succumbs to temptation, even if the judgment is an ethical one. Kierkegaard
says: ‘What ordinarily tempts a man is that which would keep him from doing
his duty, but in this case the temptation is itself the ethical….’ (Fear and
Trembling, p. 40). Or if one sees the occasion in terms of individual purposes,
one has succumbed to temptation. Thus if Abraham sees God’s command in
terms of the crass ‘Christian’ view that by this act of sacrificing Isaac he is
buying his salvation, he is lost. What is critical in understanding Kierkegaard
is to understand the nature of Abraham’s response, and by doing that to see
that although Abraham must believe that he is to lose Isaac at the same time
he believes that he will not. Only by complete submission to the divine will
can one gain freedom, only by a perfect willingness to give Isaac up is he
regained. The final truth is: ‘By faith I make renunciation of nothing, on the
contrary by faith I acquire everything…. By faith Abraham did not renounce
his claim upon Isaac, but by faith he got Isaac’ (Fear and Trembling, p. 59).
74 Michael P.Hodges

So, of course, Abraham does attain salvation by submitting his will to

the will of God, but that salvation cannot be the desired object of such a
submission. Under such a condition no submission has occurred. Moreover,
Abraham’s salvation is not to be understood as something that will take
place in an eschatological time and place.17 Rather it consists in a here and
now transformation of Abraham’s relation to himself, to his son, and, in
fact, to the totality of finite existence. In short Abraham ‘lives completely
Abraham cannot be wrong about the source of the command to sacrifice
Isaac; he can only fail to be a ‘Knight of Faith.’ That failure would not be an
epistemological error; it does not involve a false belief. It is an existential
failure. The truth of faith has nothing to do with objective facts. Rather it is
a question of response and the way in which that response is life-
transforming. The venerable Wittgensteinian O.K.Bousma, speaking of the
utterances of the religious, says:

Those sentences are not intended to describe anything as though those

who speak them, with their seeing eyes had discovered the divine hidden
in this man visible to all men. They are not seeing the invisible clothed in
this otherwise man like us. Those who say that in making these
declarations then men are projecting their longings, their wishes, might
be considered nearer the truth than are those who say that they are
talking nonsense. When, accordingly, K[ierkegaard] says: ‘Keep off!
Stop! It is a paradox,’ this is his way of saying ‘A confession, a
declaration, is not a piece of information. This is holy ground. Remove
your clumsy shoes’.18

It may now appear that we have been lead to the conclusion that the religious
life has a character that places it beyond criticism. We have seen that such
language, in so far as it presents an authority in an immediate and absolute
way, stands outside of all prudential and ethical contexts. We have also seen
that the nature of religious believing locates it outside the bounds of the usual
standards of evidence and rationality. Does it follow that all critical
assessment must be internal to the religious project? Does Abraham, in his
relation to God, transcend all standards of evaluation? I don’t think so, but
to see why will require a reassessment of some of what I have said.
So far I have limited myself to a description of the ‘logic of religious
discourse.’ But there is another sort of issue that might be raised at this point.
It was Nietzsche who taught us to distinguish between two very different
questions: (1) What is the analysis of a particular concept? (2) Under what
conditions—natural, historical or individual—does a concept such as ‘the
religious’ come to be formed? While it is true that it is part of the analysis of
religious discourse that authority is presented absolutely and immediately,
given the above distinction it does not follow that we should accept such
authority. Instead we might decide that, in light of the circumstances which
Themes from Wittgenstein, Kierkegaard and Nietzsche 75

gave rise to such a concept, we should abandon it altogether. As Wittgenstein

has pointed out:

[I]f anyone believes that certain concepts are absolutely the correct ones,
and that having different ones would mean not realizing something that
we realize—then let him imagine certain very general facts of nature to
be different from what we are used to, and the formation of concepts
different from the usual ones will become intelligible to him.
(PI II, p. 230)

While Wittgenstein here speaks of the circumstances of concept formation

as ‘certain very general facts of nature,’ he also goes on to ‘compare a
concept with a style of painting,’ where the question of human history
clearly emerges. He says: ‘Compare a concept with a style of painting. For
is even our style of painting arbitrary? Can we choose one at pleasure? (The
Egyptian, for instance.) Is it a mere question of pleasing and ugly?’ (PI II,
p. 230). The possibility that Wittgenstein is rejecting here is that we might
somehow step out of our own historical moment to ‘choose’ another style.
Our appreciation of the ‘Egyptian style’ depends on its being recognized as
just that, a style. But how far are we to take this comparison? We know
that while we may not be able to choose a style at pleasure, nevertheless
there is a history of style, and individuals, such as Giotto or Monet, have
contributed to the transformation of style. We know that while it makes no
sense to suppose that such contributors produced something ‘absolutely
correct’, nonetheless they overcame tensions and problems in previous
The comparison between concepts and styles of painting is provocative,
for now we can see religious discourse as a particular structure of meanings
situated among others. It creates possibilities of experience just as a style of
painting does, but at the same time it excludes others. Think for a moment
about the ways in which impressionists have transformed our ways of seeing.
If we examine religious discourse in this light we may come to reject it on the
grounds of what it excludes. In other words, looking at religious discourse in
this suggestive way seems to open it to critical assessment.
Would Wittgenstein see the insight in this evaluative approach?
Wittgenstein would, I believe, reject the whole project of assessment as a
product of what he calls the ‘main current of European and American
civilization’ for which progress is its very form (CV, pp. 6–7). This is not an
altogether implausible claim, for surely many in the American philosophical
tradition would take up the evaluative project here.19 In so far as they see a
particular practice as one among various possibilities, it becomes available
for critical assessment, transformation or rejection. But for Wittgenstein the
task of philosophy has come to an end once we ‘command a clear view of the
use of our words’ (PI §122). After all, ‘[p]hilosophy may in no way interfere
with the actual use of language; it can in the end only describe it. For it cannot
76 Michael P.Hodges

give it a foundation either. It leaves everything as it is. It also leaves

mathematics as it is…’ (PI §124).
If philosophy leaves mathematics as it is, surely it also leaves religious
discourse as it is. But what does this mean? Certainly Wittgenstein is rejecting
the idea that our practices require the sorts of foundation that traditional
philosophy has sought to supply in the mistaken belief that without them our
practices are problematic on their own terms. Certainly, if we read
‘philosophy’ to mean just the various methods employed in the Philosophical
Investigations to help us overcome the above sort of mistake, then it is quite
clear that philosophy must leave everything as it is, since its task is completed
just when we are returned to those practices free from the demands of
philosophy in the traditional sense. This seems correct, but does it imply that
there is no sort of reflective criticism that can be brought against our
practices? There will, of course, be no neutral perspective from which to
launch such a critique, but the supposition that only such a value-free critique
would do is itself a prejudice of traditional philosophy in just the sense that
Wittgenstein’s methods are meant to displace. Any such assessment would
itself be rooted in values. What we cannot do is to find a transcendental perch
from which to examine our practices in their totality, but we can begin to see
how they impinge on each other and how they enrich or diminish the quality
of our lives. Perhaps Wittgenstein would not engage in this sort of criticism,
but at least he leads us to the door behind which the possibility lies.20 If, as I
have argued, for Wittgenstein religious discourse is ‘a form that human life
has taken,’ that a religious belief is not an hypothesis or theory, as he so
clearly contends in ‘Remarks on Frazer’s Golden Bough,’ then it seems that
the only relevant question to ask about this form of life is what values it
serves and what it suppresses—and this is the genealogical question.
Nietzsche was not interested in an ‘analysis of religious discourse.’ That
he could leave to Kierkegaard or Wittgenstein. He was interested in the
question ‘what historical forces conspired to produce such a set of concepts
and what interests were served and what violated in that formation?’ And he
tells a story of powerful interests at war with each other. That we can ask the
second sort of question means that whatever the analysis of a concept may
be, there is still room for a further evaluative question. Should we continue
to operate with that concept in light of what forces were and are at work in
its formation and specifically of what is excluded by its deployment. If what
is experienced as religious is historically conditioned, then such experience is
open to question. One could, of course, remove the possibility of any critique
here by claiming contra Wittgenstein and Nietzsche that certain concepts are
the absolutely correct ones having a non-historical, non-worldly, basis.
However, this return to Platonism is perhaps unattractive in light of much
recent philosophy.
In addition to the genealogical approach, there is a second way to re-
open the question of critical evaluation which will become apparent if we
return to the story of Abraham. This second way proceeds by asking how
Themes from Wittgenstein, Kierkegaard and Nietzsche 77

one form of life—the religious life—looks from the standpoint of another

form of life or language-game, such as the ethical or prudential one. When
the question arises whether we should operate with religious concepts, this
question places religious experience within the domain of either the ethical
or the prudential, depending on the nature of the ‘should.’ When the story
of Abraham begins for Kierkegaard, we hear that ‘God tempted Abraham
saying….’ The story begins from a transcendent perspective. It is an
unquestioned assumption of Kierkegaard’s treatment that it is God who
speaks to Abraham.21 This assumption is not called into question. The
experience as described already contains, one might say, half of the absolute
relation to the absolute, namely the absolute. The only issue is whether the
other half—Abraham’s half—will be forthcoming. And for him the only
language in which that issue can express itself is whether he will do God’s
bidding or whether he will sin.
Now in this context it is a well-known feature of the story that Abraham
himself cannot know that he will not be required to sacrifice Isaac. If he
does know that, his actions are mere sham and play-acting. He would not
have given up Isaac at all. It is in this feature that we find the ‘suspension of
the ethical.’ If he did know, then he would not have been called out of the
ethical at all. But can the same thing be said for the rest of us as readers of
the story? Isn’t it essential to the point of the story that we know how it
turns out. If we are to take this as a story about faith—true faith—can we
suppose that Abraham does sacrifice Isaac? Suppose, for a moment, that
the story ended with a bloody description of the evisceration of Isaac.
Would this be included in the biblical canon? Would it be taken as the ‘word
of God’? With this question we locate ourselves outside the story and
demand that it meet certain ethical standards. In this sense, the story of
Abraham does not represent a complete suspension of the ethical. While
the ethical may be suspended in the story—for Abraham—the story itself
still occurs within the ethical. That is, there are ethical limits at work which
shape the course of the story. Of course, there is no need to pass judgment
on God per se. Rather we simply refuse to accept that God would have
commanded such a horrible thing. That there are ethical limits at work here
implies that the absoluteness of Abraham’s relation to the absolute is itself
limited—if not for Abraham, then for us.
Remember that it is the burden of Kierkegaard’s discussion of Abraham
to show that in Abraham’s relation to God, he is higher than the ethical. As
Kierkegaard puts it: ‘For faith is this paradox. That the particular is higher
than the universal.’22 But if what has been said above is true, that is no longer
clear, for we have seen how there are or may be ethical limits at work. I do
not mean to suggest that the ethical takes absolute priority over the religious
but rather to point out that each can ‘stand in judgment of the other,’ and in
so far as the debate between them is broached, there will be no neutral ground
from which to settle the question. It is the burden of Kierkegaard’s argument
to show that the religious cannot be contained within the ethical since the
78 Michael P.Hodges

religious is constituted by a relation to intrinsic and immediate authority—

an absolute relation to the absolute. But without a prior hierarchy which
somehow ranks claims from competing domains, we can only speak from
within a structure of claims which claims for itself absolute authority. For
one within the grasp of the religious, there can be no room for the question of
the authority of God; but for one outside that grasp there can be a question
of its legitimacy. It does no good to insist that the Bible must be true because
it says that it is!
Perhaps the way to see the point here is in relation to Wittgenstein’s famous
duck/rabbit example (PI II, xi). When one initially views the object, one simply
sees a duck (or a rabbit). All the elements are parts of a picture of a duck. But,
of course, we all know that in a moment it can change to a rabbit and then
everything is a part of a rabbit picture. Now this change is not a replacement of
an incorrect view with a correct one. Neither is more correct than the other.
And once we have seen each, it is never possible to take either as exclusively
correct. Until the second comes into view, the first or only view simply does not
exist in a field of other possibilities—it just is what it is. Seeing the other
possibility thus does not change our view by replacing an incorrect view with a
correct one; it changes our view rather by breaking the hold of the one upon us
so that we can no longer inhabit either with the same naivety as before. We
might wish that such an experience had never happened, but now it is too late.
We might wish that we had not seen the possibility of reinscribing the religious
within the ethical, but now it is too late.
Perhaps there is something to be learned from the way the different
discourses take hold of us and situate other claims. We have seen how from
the point of view of each language the claims of the other are experienced as
contingent and secondary. That is, from within the experience of an absolute
relation to the absolute—Abraham’s experience—every other claim, even the
ethical, loses its validity. But the possibility of the reinscription of that
discourse within another interrupts any claim to validity beyond the discourse
itself. Once we have seen the possibility of the duck over against the rabbit,
we can no longer take the duck as the way to see the thing. Similarly, having
seen the possibility of a wider ethical context in which Abraham’s experience
is situated, we can no longer see the religious description as exclusively valid.
But the same thing can be said of the ethical form of life with its discourse: it
too is contingent. This insight raises a possibility that has not yet been
examined, the experience of the contingency and therefore the possibility of
the absence of all structures of meaning. This possibility gives rise to a third
perspective from which we can engage in a critical assessment of particular
If the validity of a claim is a function of the discourse or language-game in
which it has a place, then, of course, within that sphere there can be no
question of its application. For Abraham the experience of God just is that of
an absolute relation to the absolute. However, the possibility of an
interruption of that language-game by reinscription also shows the limit of
Themes from Wittgenstein, Kierkegaard and Nietzsche 79

such claims, and this limit can be read in two ways.23 On the one hand, it
guarantees validity within the sphere of its application. 24 But while
contextualizing meaning insures the validity and intelligibility of what is
claimed within the specified language-game, it also implies that claims to
absolute validity make no sense. For all such claims presuppose the language-
game in which they occur and, as Wittgenstein has pointed out, no such
structure of meaning can be seen as ‘absolutely correct.’25
There are, at least, three distinct responses to an acknowledgment of the
contingency of all language-games and forms of life and the possible absence
of all of them. The Nietzschean, or perhaps the reform-minded American
philosopher, would welcome such contingency and even contend that the
absence of the religious would free up our lives. At a minimum, if we abandon
religious discourse new possibilities of experience emerge. We might even
grow into those ‘adults’ Freud seems to envision in the final chapters of The
Future of an Illusion.26
The possibility of the absence of the religious form of life will be seen in a
very different light by the believer. In fact, one might even wonder if such a
possibility can really be accepted by the religious person? The answer here is
yes, but not in the terms given. For the religious person what we are discussing
is the terrifying possibility that the people of God might forget God—the
meaning of God’s word might be lost—and this certainly is a human
possibility. Religiously it is the loss of the possibility of a relation with God.
Again we see that even this radical possibility of the contingency, and
therefore the non-existence, of religious discourse can have no absolute
authority over other structures of meaning. The religious has its own ways of
recasting even the possibility of its own non-existence. But just as with the
previous cases of reinscription, matters will appear somewhat differently if
we begin from a position which does not presuppose the availability of
religious discourse.
These considerations lead us to a third response to the contingency, and
possible non-existence, of all structures of meaning. (As we shall see, a
religious believer might try to take this moment as the very occasion for the
introduction of religious language.) It is very difficult to remain clear at this
point, because what seems to come into view is not just the open-ended
possibility of imagining ‘the interruption’ of one discourse by the process of
reinscription. While, as we have seen, that is clearly possible, we are always
left within a structure of meaning. To describe such possibilities is to remain
within the structures of meaning even as it involves the substitution of one
discourse or language-game for another. With the third response, what seems
to come into view is a much more radical possibility—the absence of meaning
altogether. The experience of the religious shows the contingency of the
claims of the ethical and the prudential, but the possibility of reinscription
shows the contingency of the claims of the religious as well. And it is this
contingency of meaning taken as a pervasive feature of all structures of
meaning that concerns us here.
80 Michael P.Hodges

There is an obvious sense in which such a condition—the absence of all

meaning—could not be articulated. It can only ‘seem to come into view’
because, of course, any articulation of such a view will involve the
deployment of some structures of meaning and will thus ‘paper over’ the
backdrop of supposed non-meaning. This ‘experience’ is a possibility which
cannot strictly speaking be given voice. It might be useful to recall the way in
which Wittgenstein begins the ‘Lecture on Ethics’ by talking about two
‘experiences’ which are such that ‘the verbal expression we give to these
experiences is nonsense!’ (PO, p. 41). There he refers to the experience of
wonder at the existence of the world and the experience of absolute safety.
And, of course, his point is that while we can wonder at things being one way
as opposed to another, it makes no sense to wonder at there being anything
at all. In the same way ‘the absence of meaning’ can have no language. There
may be much that we are tempted to say in the face of such an experience,
but nothing can force itself upon us here. Nothing can be accurate to the
‘experience’ itself. While Wittgenstein discovered two such ‘experiences’ in
the ‘Lecture,’ we seem to have found another.27 Or perhaps it is simply the
same ‘experience’ under a different ‘misdescription.’ Wittgenstein speaks of
wondering at the existence of the world, and we might speak of wonder at
the existence of meaning.
We can be badly misled here because, it must be remembered, the ‘Lecture’
is from the Tractarian period in which the notion of a limit to meaning plays
a central and absolute role. For the later Wittgenstein there is no such
absolute limit. We cannot set a limit except de facto, and such a limit is always
moveable—for where no speech has occurred, in the future it may. The
importance of this point will become clear shortly.
How then should we speak in the ‘presence’ (intentionally bad choice of
words) of the absence of meaning? As noted above parenthetically, at this
point some philosophers may be tempted to speak in religious terms.28 After
all, they may say, it is just God who is at the limit of all human structures of
meaning—always outside and beyond the grasp of human thinking. In fact it
might be argued that such an experience of the radically limited character of
human meaning is just the experience that funds an awareness of the divine.
Here we have a negative theology—a view that contends that it is God’s
radical difference from all human structures of meaning that is God’s
Now, on the face of it, an appeal to some (traditional) structure of
meaning, e.g., the language of theology, would also be an attempt to paper
over the rift. What I mean here is that theological language, at least as it
has functioned in the past, has been used precisely to deny the limit to
structures of meaning. Theological discourse is, after all, a continuation at
just the point at which we are supposed to be faced with a limit beyond
which there was no continuation. It follows that the appeal to God has
been used to reject the contingency of meaning and to invoke a perspective
from which meaning is grounded in the eternal. As such it functions in
Themes from Wittgenstein, Kierkegaard and Nietzsche 81

contradiction to the very experience that it is supposed to take as its point

of departure.
And yet it might be argued that we cannot determine this limit of meaning
in an anti-theistic or even anti-ontological sense either. And this is certainly
correct in one sense for, of course, such notions as ‘anti-theistic’ and ‘anti-
ontological’ invoke structures of meaning just as much as do ‘theistic’ and
‘ontological.’ But there are the seeds of real confusion here. One cannot argue
from an absence of meaning to any claim at all. If, as the negative theologian
claims, language fails us, then that is his insight and nothing more. What this
means, on the negative side, is that any deployment of language here cannot
be judged from the perspective of accurate representation. There is no
meaningful system of representation and nothing to represent. What could it
mean to believe in or deny the existence of ‘a something’ (and to say even this
is to say too much) that is radically other to all structures of meaning?
Wittgenstein says in another context: ‘So in the end when one is doing
philosophy one gets to the point where one would like to just emit an
inarticulate sound.—But such a sound is an expression only as it occurs in a
particular language-game [structure of meaning] which should now be
described’ (PI §261). For the negative theologian, it would seem that what
the theist and the anti-theist have in common is an uncontrollable urge to
emit an inarticulate sound.
That is why there is something wrong if the original claim about anti-
theism (that its discourse is contingent) is taken as a defense of theological
talk. It seems to win by theft what it could not get in a fair encounter
because it leaves the debate between theist and anti-theist within the
domain of the representational model. In that context there is simply
nowhere to go. Wouldn’t the only legitimate move be to follow
Wittgenstein’s famous advice and admit that what we cannot speak about
we must pass over in silence? But this injunction must now be understood
in a way consistent with the later thought. Given that the later Wittgenstein
has rejected an essentialist view of language, we cannot require the speaker
to be silent on pain of transgressing the limits of language per se. Rather we
must take the speaker’s speech as proof that there is no limit being
transgressed. The proof of the possibility of speech is speech. But this
presents us with a new problem. What drives the theologian, or anyone
who would speak at this point, to speech?
The desire to employ a language of divinity or any other language cannot
be forced upon us by the ‘very nature of the case,’ for the case we are
considering is the supposed ‘death’ of all meaning. So that desire must arise
from another source, and the particular content of any discourse spoken in
‘the absence of meaning’ must be borrowed from other domains. In short,
how this experience will be taken up is a function of what ‘structures of
meaning’ are available to the speaker. But, as we have seen, there cannot be
a question of truth as accurate representation here simply because, given the
negative theologian’s own account, there is nothing to represent. In other
82 Michael P.Hodges

words, we must turn away from the transcendental per se and ask what place
such speakings have in life. In this recognition the issues have been
transformed from classical metaphysical matters to questions of ethics and/
or politics. We, who are observers and not necessarily participants, must take
up the language of the religious life in a new way.29 We must ask what values
are embodied herein. What aspects of human life are highlighted and which
forced into the shadows? What political and social agendas underlie the urge
to speak at this point and in this way? To refuse to take up religious language
in this way is quite literally to condemn oneself to silence. Otherwise, there is
simply nothing to say.
Where does this leave the religious person as opposed to the theologian?
The answer is that such a person is left with his or her religious practice and
with the language in which this practice expresses itself. We certainly cannot
deny that this practice is one form that human life has taken, and we must be
reminded of Wittgenstein’s comment that what must in the end be accepted
are forms of life (PI II, p. 226).

1 By claiming that they are ‘roughly analogous’ I do not mean to suggest that
religious claims are pseudo-scientific claims or that there is a decision procedure
for religious claims like that for scientific claims. My point is only to distinguish
between first-order claims within a science or religion and second-order claims
that offer accounts of the meaning and status of the former. I would add that
sincere commitment to first-order claims is no guarantee of an adequate second-
order story about their status. In fact, it may be a hindrance. Scientists are not
always the best at providing an account of what it is that science is up to. In the
same way, those engaged in the religious life may not be the best at giving an
account of what that practice is all about.
2 Complete neutrality is not possible. I do not suggest that what I have to say is a
matter of pure analysis of a preexisting structure. My views are rather
recommendations about how to construe religious language which have certain
advantages but clearly do not ‘leave everything as it is’ (PI §124).
3 R.Rhees, Without Answers (New York: Shocken Books, 1969), p. 121.
4 See E.Renan, The Life of Jesus (New York: Modern Library, 1927), and
A.Schweitzer, The Quest for the Historical Jesus (New York: Macmillan, 1959).
5 John Dewey, A Common Faith (New Haven, CT: Yale University Press, 1962),
p. 20.
6 To say that there is a lack of evidence is not to say that there are not documents
to which to refer. For Christianity, certainly the Bible is such a text but it does
not function as a source of hypotheses to be independently tested. Rather it is
authoritative. The believer does not read the Bible to see if it is true but to
discover the truth.
7 Paul Tillich, for example, speaks of faith as ‘ultimate concern’ and contrasts that
with all sorts of preliminary concerns which may have to be sacrificed in the
name of what has ultimate authority. Being ultimately concerned, in this sense, is
a defining commitment. See The Dynamics of Faith (New York: Harper & Row,
1957), Chapter 1.
8 The interesting question of whether it is possible to provide an independent
philosophical argument for democracy is discussed by Rorty in a paper entitled
Themes from Wittgenstein, Kierkegaard and Nietzsche 83

‘The Priority of Democracy to Philosophy’ in Alan R.Malachowski (ed.),

Reading Rorty (Cambridge, MA: Blackwell, 1990).
9 The ultimacy of religious authority here discussed is not inconsistent with the
idea that while such an authority may be absolute (experienced as such) it is not
unconditional since it is historically situated and from that perspective
contingent—under the right circumstances Christianity might fall into the same
oblivion as Zoroastrianism, but that is irrelevant to the engaged believer. Or
rather it would be experienced as a tragedy of cosmic import. From within faith
such a state of affairs would be nothing less than the loss of the possibility of
10 ‘Commitment’ can be misleading in this context because it suggests something
that I do, while faith befalls or overtakes one. Phenomenologically it comes from
“without” and transforms.
11 In F.Waismann, ‘Notes on Talks with Wittgenstein’, Philosophical Review, vol.
LXXIV, no. 1 (October, 1965):15.
12 While it may be beyond saying, much is ‘suggested’ in the final propositions of
the Tractatus and in the Notebooks. See my Transcendence and Wittgenstein’s
Tractatus (Philadelphia: Temple University Press, 1990), particularly the later
13 S.Kierkegaard, Fear and Trembling (Princeton, NJ: Princeton University Press.
1974), p. 63 (all other citations in the text).
14 Some of these possibilities are examined by Kierkegaard in the four stories with
which he begins Fear and Trembling.
15 This reading is consistent with Wittgenstein’s own view where we are told to
look to how a term functions in the lives of the users if we are to understand it.
16 Wittgenstein offers an interesting discussion of the relation of evidence to
religious belief in LC (pp. 53–64). A complete account would carefully
distinguish at least between ‘evidence for the faithful’—evidence as it functions
within the faith context—and ‘evidence for faith,’ i.e. evidence which might be
offered for accepting or rejecting the faith. Any appeal to a particular passage in
the Bible will, perhaps, provide sufficient evidence for the faithful but it goes
nowhere in providing evidence for faith since it is the very status of the Bible that
is in question there. A great deal of confusion surrounds this important
17 Of course there are various reasons why such possibilities suggest themselves
and of which Wittgenstein would try, by way of his ‘philosophical methods,’ to
disabuse us.
18 O.K.Bousma, ‘Notes 1’, in R.Bell and R.Hustwit (eds.). Essays on Kierkegaard
and Wittgenstein (Wooster, OH: College of Wooster Press, 1978), p. 39.
19 This, of course, is the sort of project that Dewey undertakes in A Common Faith.
20 An earlier version of these comments appeared in my article ‘The Status of Ethical
Judgements in the Philosophical Investigations,’ in Philosophical Investigations,
vol. 18, no. 2 (April 1995):99–112.
21 It would be more correct to say that the unquestioned assumption is that of
Johannes de Silento, but I will not broach the issues of the pseudonymous
authorship here.
22 Kierkegaard, op. cit., p. 65. Here we need to have in mind the Kantian notion of
the universal as defining the ethical.
23 I am indebted to David Wood for this point.
24 Wittgenstein very often makes such a claim when he points out that ‘this is how
the game is played’ or ‘that is just what we do’—see, for example, PI §480.
25 PI II, xii, p. 230.
26 S.Freud, The Future of an Illusion (New York: W.W.Norton, 1961), pp. 46–56.
27 He also mentions the ‘experience’ of guilt before God as well as the two others.
84 Michael P.Hodges

28 One might be tempted to speak in a variety of ways—of mystery, of wonder, of

the beauty of the world. Each of these language-games might be invoked as, no
doubt, might others. The religious response is here taken as an example. It is an
instructive example because of the pervasive nature of this response in our
cultural history. Why that is so is, of course, an interesting and perhaps
answerable question.
29 Here I refer to those of us who would think about, not simply deploy, a particular
language. If you will, this accounts for Kierkegaard’s distance from Abraham,
but the point also applies to those who deploy but also want to step outside and
come to the defense of a way of speaking.
6 D.Z.Phillips’ fideism in
Wittgenstein’s mirror
Mark Addis

Wittgenstein’s treatment of religion has been the subject of controversy with

regard both to its nature and applications to problems in the philosophy of
religion. There has been profound disagreement about how to handle his
writings in this area. At present there is no consensus over the direction which
should be taken. Many philosophers and theologians regard Wittgensteinian
fideism1 as essentially defining what his view of religion is. Significant
disputes about the reading of Wittgenstein’s work on religious topics arise
from examining fideism.2 For the purposes of this discussion, fideism3 will be
deemed to be exemplified by the writings of D.Z Phillips.4 Consideration of
Phillips’ treatment of the concept of internalism and the role it plays in his
understanding of Wittgenstein provides a perspective from which to
investigate interpretative issues.5
The fideist position is that religious language is intelligible only to those
who participate in the religious form of life. Fully understanding the language
of a religious believer is inseparable from comprehending a religious form of
life. This kind of understanding is about making sense of human beings.
Religious language constitutes a distinct linguistic practice which non-
participants in the form of life could not grasp and show to be incoherent or
erroneous. For example, an atheist could not find religious language
intelligible, much less criticise it.6 Religious concepts are available only to
those who partake in the form of life that they are used in. Cognisance of a
religious form of life is necessary in order to perceive what it means to apply
the ideas of truth and falsehood to religion.
Insight can be gained into fideism by reviewing how it relates to the
distinction between cognitive and non-cognitive accounts of religious
language. A representative statement of the cognitivist type of position is
that due to what religious language refers to its employment commits the
believer to particular claims about existence. These claims are expressed in
theologies. If the theologies are not verifiable, there is in principle no
method of determining whether these existential presuppositions are
rational. Two consequences of this are that religious language has no clear

86 Mark Addis

reference and that the believer’s claims fail to meet acceptable standards of
The fideist is opposed to this kind of cognitivist view. A representative
formulation of the non-cognitivist sort of view is that religious language does
not refer to anything while ordinary language refers to something. It follows
that the employment of religious language does not commit the believer to
certain existential claims. Emphasis is not directed towards the idea that there
are reasonable norms of rationality about religious assertions but rather to
the metaphysical significance of the notion of expressing an attitude about
religion. Fideism holds that the non-cognitivist account does not assist in the
understanding of what religious language is actually like. It offers an
alternative conception of how non-cognitivism about religious language can
be characterised.
The difference between cognitive and non-cognitive views about religious
language can be illustrated by reflecting on the Book of Job and the picture
of the nature of God presented in it. When Job suffers his misfortunes his
first response is to seek an explanation for the events which have occurred,
but he gradually realises that he must abandon the desire for answers. In
becoming aware that events do not have a rationale. Job ceases to be
dependent upon them. The arbitrary character of his misfortunes leads Job
to see that he is not the centre of his world but that God is. For the cognitivist
sort of position, the assertion that the actions of God are not necessarily
explicable to man is making a claim about the existence of properties of God.
However, the non-cognitivist type of standpoint regards the assertion as
rather expressing an attitude towards God.

The notions of externalism and internalism play a prominent part in
Phillips’ account of his interpretation of Wittgenstein’s philosophy of
religion. According to him, the ideas of externalism and internalism are
place holders or schemas for particular theories and their content is defined
by that of the theories. Phillips defines externalism from the standpoint of
external criteria of meaning being imposed on religion.7 Externalism is
characterised as the idea of ‘constructing theories of meaning which
determine what is to count as meaning…’. His treatment of internalism is
from the perspective of internal criteria of meaning obtained from religion.8
Internalism is the view that ‘all external criteria of meaning are irrelevant
to religious belief’.
Phillips presents the idea of religious beliefs as being beyond criticism and
beyond change as an internalist position.9 He offers a possible argument for
this conclusion using the following sequence of premisses:10

1 Various aspects of human life, including religion, have distinctive

meanings which must not be reduced to a spurious unity.11
D.Z.Phillips’ fideism in Wittgenstein’s mirror 87

2 Human life is divided into strict compartments, each autonomous as far

as its meaning is concerned.
3 Religious belief is logically distinct from other kinds of belief.
4 It is claimed that what is and what is not meaningful in the context of
religious belief is to be determined solely by whatever is called religious
5 Religious belief is an absolute measure on people’s lives and cannot by
itself be influenced by them or any kind of social or cultural events.

The perspective which Phillips offers on internalism plays a significant role

in his characterisation of his interpretation of Wittgenstein’s philosophy of
religion.12 Phillips rejects the view that his approach to this topic is an
internalist one on the grounds that he does not subscribe to the conclusion of
the internalist argument.13 This repudiation is part of his attempt to expound
his fideism. When objections to internalism are considered in aggregate,
however, it is not clear what Phillips’ denial that his treatment of
Wittgenstein’s philosophy of religion is internalist in nature really signifies.
Phillips’ employment of the notion of internalism is problematic.14 There
are difficulties with the internalist argument. The conclusion needs further
elucidation. For example, religious beliefs have altered over time. Medieval
Christian theology discussed the question of how many angels could fit on
the head of a pin, but this issue is not of current interest. If the assertion that
religious beliefs are beyond change is deemed to be consistent with the view
that religious beliefs have modified over time then it ceases to be clear what
the phrase ‘beyond change’ means. Assuming that the claim contradicts the
position that religious beliefs change over a period of time then an implication
of this is that religious beliefs do not alter over time. The implication is clearly
false if it is supposed to be an empirical claim. However, if the implication is
regarded as being a philosophical assertion then it is stating that the idea that
religious beliefs change over time should stop being connected with the
concept of religious beliefs (which is a deviation from the way in which the
concept is normally regarded). The claim that religious beliefs are beyond
criticism and change is insufficiently precise as it stands to decide the exact
strength of the support which the internalist argument provides for it.
Another difficulty generated by the lack of precision of the claim is that it is
difficult to determine whether commentators on Wittgenstein have supported
it. It is possible that when Phillips criticises the internalist argument’s
conclusion he is attacking a view which is not a widely held interpretation of
Wittgenstein.15 If so then he needs to provide more explanation as to exactly
how rejection of this conclusion serves to elucidate his account. Arguably,
what this discussion shows is that the usefulness of the conclusion of the
internalist argument as a foil for Phillips’ claim that his handling of
Wittgenstein’s philosophy of religion is not an internalist one is limited.
The internalist argument leads to the strongest form of internalism
possible.16 However, widely varying accounts of internalism can be produced.
88 Mark Addis

Weakening the premisses used in the internalist argument leads in turn to

weaker versions of internalism. For example, in premiss two the emphasis
upon distinct religious language raises the question of whether internalism
requires an adherence to the idea of distinct religious language as opposed to
the concept of identifiable religious language. Distinct religious language is
the notion that there is an autonomous and isolated religious language-game.
What identifiable religious language is is a language-game which can be
identified as religious but is not distinct. A commitment to the notion of
distinct religious language is not necessary for a weaker form of internalism.
This is because it is possible to have a version of internalism where it is not
postulated that human life is divided into strict compartments. That is to say,
there could be overlap or interaction between the compartments of human
life. As noted above, Phillips seeks to deny that his reading of Wittgenstein’s
philosophy of religion is internalist in character. Accepting Phillips’ claim that
he is not proffering the strongest possible version of internalism does not
exclude the possibility that he could be propounding internalism of a weaker
kind.17 A consequence of this is that Phillips still appears vulnerable to the
charge that his treatment of Wittgenstein’s philosophy of religion is
internalist. If his approach to this topic is internalist in character then it could
be argued that his interpretation of Wittgenstein’s writings in this area is not
entirely satisfactory.18

The internalist argument does not go through as it is presented because of
problems with premisses one and two and the resultant weakening of the
justification of premiss two by premise one.19 Aspects of Phillips’ rendition of
some of the themes which are used in the internalist argument militate against
the interpretation of Wittgenstein which is preferred here, and can be
challenged. A part of this is that there could be criticism of the reliability of a
number of Phillips’ readings. Many of Wittgenstein’s remarks on religion are
found in conversations which have low exegetical status. A desirable
exegetical methodology for attempting to deal with this problem is to
correlate the interpretation of the conversations with texts of higher
exegetical status. However, Phillips does not adopt this favoured exegetical
The first premiss in the internalist argument is that various aspects of the
forms of human life have distinctive meanings which must not be reduced to
a spurious unity. Phillips observes that religious beliefs would not have the
significance that they do if they were severed from the life which they are
part of.21 However, this does not make what is in religious beliefs any less
distinctive. Religious beliefs cannot be comprehended at all unless their
relation to other modes of life is taken into account.22 The meaning of
religious beliefs is partly dependent on non-religious facts, but this is not
claiming that those beliefs are justified by the facts in question.23 It is essential
D.Z.Phillips’ fideism in Wittgenstein’s mirror 89

to distinguish inquiry which occurs within a form of life from inquiry which
is about it.24 Religious concepts have their meaning within a particular form
of life but this does not imply that these notions make any claim about a
form of life.25 The life of religious institutions shows the meaning of religious
beliefs but those beliefs need not say anything about those institutions.26
There are problems with Phillips’ approach to the concept of forms of life.
There is a difference27 between the preferred reading of Wittgenstein on forms
of life28 and that which Phillips employs. Given this, it is evident that he
misappropriates Wittgenstein’s idea of forms of life. Wittgenstein employed
the notion of forms of life in at least two distinct ways. One use is to
summarise the common human way of acting—that which is particularly and
universally human. There are very general facts about human biology which
are common to the human race as a whole, for example, that a person has
one body. This use refers to the biological aspects of human nature. Another
usage is to stress differences between societies. Wittgenstein was concerned
with cultural practices in a broad way that encompassed both anthropology
and sociology. An example of cultural practices is the use of the word
‘responsibility’ in an ancient Greek legal case about a boy who threw a javelin
and accidentally killed another boy. Consideration was given to whether the
judges of the contest, the person who threw the javelin or the javelin itself
was responsible for the accident. This usage refers to the cultural aspects of
human nature.29
It is crucial whether Phillips is using a cultural or biological concept of
form of life, or whether he is not distinguishing between the two notions. If
Phillips is employing a biological idea of a form of life then he requires more
extensive arguments than he provides to support his ideas about the
relationship between understanding religious language and the religious form
of life. This is because the biological concept of form of life is about what is
universally human and thus is not specific enough to be used to identify
different particular forms of life. Phillips’ interest in the connection between
religious language and form of life would be better served by the cultural
concept of form of life as this is concerned with the identification of diverse
forms of life. It matters that it is unclear which concept of form of life he is
A more specialised issue is how the cultural concept of a form of life is to
be viewed. That is to say, is this form of life held to be a collection of religious
beliefs rather than a set of cultural practices in which questions about
religious beliefs can be asked? It is possible for a form of life to encompass
both a particular set of religious beliefs and a context in which inquiries about
religious belief can be mooted. A question raised by this is whether Phillips’
arguments would most effectively be prosecuted by a concept of a form of
life which is a collection of religious beliefs, although he is actually using the
idea of form of life in a non-specific fashion. Phillips argumentative purposes
would best be facilitated by a notion of a form of life which is a collection of
religious beliefs. A reason for this is that if a form of life is regarded as a
90 Mark Addis

context in which inquiries about religious belief can be posed it could allow
communities of different persuasions to engage in meaningful discussion
together. For example, it permits the possibility that Jews, Christians and
atheists could have intelligible debate on the existence of God. The difficulty
which possibilities of this kind creates for Phillips is that he wishes to claim
that an atheist could not find religious language comprehensible.
Phillips’ employment of the concept of forms of life to defend religious
beliefs from criticism is a central part of his fideism. A major kind of difficulty
with Phillips’ treatment of the idea of forms of life is the extent to which a
religious form of life is beyond criticism from other aspects of human life.
The Winch-MacIntyre debate about primitive societies can be applied to
Phillips’ employment of the concept of forms of life.30 The source of the
debate is Winch’s book The Idea of a Social Science.31 In this work he stated
that one cannot apply the criteria of logic to ways of social life as such.32 For
example, science is one such way and religion is another, and each has
guidelines of comprehensibility peculiar to itself. However, within science or
religion actions can be logical or illogical. Winch developed these ideas in his
subsequent writings. He criticised some versions of the view that there is (or
could be) a concept of making sense of the world which is wholly separate
from modes of social life.33 Winch also objected to the position that this
notion of making sense of the world can supply standards by reference to
which certain ways of life can be criticised as lacking rationality. Any concept
of making sense of the world is comprehensible only against the background
of ways of life and understandings of them. That is what is involved in cross-
cultural attempts to comprehend a society. Every judgment of whether a way
of life is rational has to be relativised to the culture being considered. This
claim about culturally relativised judgments may be construed as asserting
that one should take care not to ascribe one’s own standards of judgment but
ascribe instead those standards to which the other culture would give rise.34
Alasdair MacIntyre agrees with Winch to the extent that an understanding
of a culture with respect to their own notions and beliefs does actually tend
to rule out comprehending it in any other terms.35 Where MacIntyre disagrees
with Winch is in his view that in order to investigate a culture it is necessary
to inquire about what needs and aims cultural practices serve. For instance,
different cultures reflecting different schemes of concepts can be regarded as
serving the same social needs. MacIntyre’s position allows criticism of
cultures other than one’s own because the functions of cultural practices can
be considered.
In the Winch-MacIntyre debate, the idea of a primitive society can be
directly replaced by the notion of the forms of life of religious believers.36
There would be grounds which strongly militate against this application of
the debate to the notion of forms of life if Wittgenstein is read as supporting
the claim that all forms of life are equally valid and no particular form of life
could be criticised. It is useful to consider a few brief examples of
Wittgenstein’s attitude to criticising certain forms of life. In the ‘Lectures on
D.Z.Phillips’ fideism in Wittgenstein’s mirror 91

Religious Belief’, he wanted individuals to pause and reflect about when they
should combat a religious language that is not theirs (or their culture’s).
Another instance is Wittgenstein’s approach to groundless beliefs. The ideas
that some beliefs are groundless ones which a community cannot live without
and that a number of these groundless beliefs are primitive are found in On
Certainty. The application of criticisms from the Winch-MacIntyre debate to
Phillips’ use of the concept of forms of life shows that it is possible to think
that a religious form of life can be critiqued from outside of that particular
form of life. The discussion of the Winch-MacIntyre debate about criticism
of forms of life leads to the question of what role the cultural aspects of
religious beliefs have. Arguably, there is much insight to be gained from
looking at the cultural dimensions of religious beliefs without going as far as
Phillips does with his employment of the notion of forms of life. For example,
MacIntyre maintains that there is much to learn from studying the history
and anthropology of religious beliefs.37 Phillips acknowledges the cultural
dimension of religious beliefs. An instance of this is that he comments that in
current culture the settings of religious beliefs are increasingly seen as cultural
divergences38 and that for many people these beliefs have become alien
cultural practices.39 Difficulties begin to occur with Phillips’ handling of
Wittgenstein’s philosophy of religion when he tries to go significantly beyond
merely recognising the cultural aspects of religion.
Phillips claims that a language is associated with a form of life.40 For
example, in western Europe the language-games of religious believers and non-
believers are associated with different forms of life. Belief and non-belief do
not appear to be opposites within a common framework of belief. Becoming a
believer seems to be an alteration of direction rather than an alteration of
opinion.41 A major strand of critical thought about Phillips’ handling of
religious language-games is concerned with the consequences of his views for
how the notion of the form of life is regarded. Nielsen is opposed to what he
judges to be the excessive compartmentalisation of the forms of social life
involved in maintaining that there are distinct religious language-games.42 He
thinks that this excessive compartmentalisation leads to the undesirable
consequence that the language of distinct religious groups, such as different
denominations of Christians, is incommensurable.43 Nielsen suggests that the
implications of distinct religious language-games do extend across the
boundaries of the language-game into assertions concerning matters which lie
beyond them. He holds that religious language is not something which is
isolated and sufficient to itself. What constitutes evidence for the truth or
reliability of particular assertions is not completely peculiar to the context or
activity being considered. Activities are not that separate from one another.
An issue which arises out of Phillips’ treatment of language-games is that
of the correspondence between distinct language games and forms of life.
There is a complex relation between a language-game and the forms of life it
is associated with.44 The form of life is part of the framework within which
language occurs. Merely recognising important linguistic differences between
92 Mark Addis

various areas of discourse, such as those between religious beliefs and

supporting a football club, does not automatically lead to a commitment to
the view that human life is divided into a number of distinct forms of life. If
there is no automatic association between distinct language-games and forms
of life then an assertion about the former having their own form of life needs
justification. Additionally, it becomes possible that recognition of the
distinctness of language games might have no bearing at all on whether there
are (potentially identifiable) distinct forms of life. The viewpoint which is
adopted on how the relationship between a language-game and a form of life
is characterised has implications for a variety of issues. It could be asked
whether Phillips is sufficiently careful when considering this matter, and if
alternative formulations of the relationship affect the efficacy of his
arguments. If there are problems with Phillips’ treatment of the idea of forms
of life then this creates difficulties for his use of the concept of language-
games because of the association of language-games with forms of life. A
consequence of this in turn is that it may no longer be clear what it means to
claim that there are religious language-games.45

The second premiss in the internalist argument is that there is division of human
life into strict compartments, each autonomous as far as its meaning is
concerned. The autonomous meanings of the divisions of human life are
distinct language-games, and each distinct meaning is a distinct language-game.
Phillips discusses the diversity of religious language in general terms.46 He
remarks that it should not be thought that different linguistic employments are
all variations of a single paradigm use. Frequently, the uses of language are
distinct and will be not be properly comprehended if their particular nature is
not noticed. Phillips maintains that religious beliefs are distinct language-games
because they are not linked to what lies outside religion as what is justified is to
its justification. He claims that justification for religious beliefs stops, and the
attempt to find justification for religious belief is a case of not being aware of
where to stop.47 Phillips contends that religious beliefs are not accountable to
justification by some philosophical standard, and that he arrives at this position
as a consequence of reflection rather than having it as an assumption from
which he starts. It does not make sense to request proof of the validity of
religious beliefs.48 The desire for a foundation for religion is an instance of
seeking justification beyond the point where it is sensible to do so.49
Phillips considers that doubts about the philosophical representation of
religious beliefs as distinct language-games lead to attempts to demonstrate
that their religious conclusions are reached by rational guidelines.50 From the
perspective of there being rational guidelines, disagreements about religion
between believers and non-believers are disputes between opposing
hypotheses. The truth is elicited by the testing of hypotheses and the rejection
of the false.51 Unless it can be demonstrated that believers and non-believers
D.Z.Phillips’ fideism in Wittgenstein’s mirror 93

are employing common guidelines of rationality the misgivings concerning

religious beliefs as esoteric games cannot be averted.52 An objection to
Phillips’ view that religious language-games do not require further
justification is that this position has the consequence of making these
language-games invulnerable to external critique.53 Another criticism which
is made of Phillips’ attempt to preclude justification is that of how it is to be
known that religious language-games are not some kind of disguised nonsense
(that believers are unable to recognise).54 This is because it is inappropriate
to ask for grounds of religious belief.
A source of difficulty for Phillips’ treatment of distinct religious language-
games is how his views relate to Wittgenstein’s conceptions of the
distinctiveness of religious language-games. It is important to observe that
Wittgenstein did not make any remarks about the uses to which Phillips put
the idea of distinct religious language-games and that therefore this usage lacks
exegetical support. Phillips comments that recognising the limitations of the
analogy between games and language is significant. The different language-
games which are played are not part of one big game. Wittgenstein wished to
claim that a language is a family of language-games and this is the sort of unity
which a language has.55 What Phillips emphasises in relation to distinct
religious language games is that there is participation in different-language
games in the same language. A question which the stress upon distinct
language-games raises is the issue of what principles or methods there are for
identifying a distinct religious language-game. Arguably, there is no (good)
exegetical evidence in Wittgenstein about what principles or methods should
be used to assist in the identification of distinct religious language-games.
Perhaps the most promising interpretative justification that can be found
for Phillips’ treatment of the distinctiveness of these kinds of language games
is that in the Brown Book there are analogues to distinct language games. A
significant instance of Phillips’ attempt to invoke exegetical support for his
conception of distinct language-games is when he discusses the idea that
although Wittgenstein stressed the distinctiveness of language games he also
talked of the links between them.56 According to Phillips, Wittgenstein was
aware of the bearings which the various utterances made have on each other
and recognised that these bearings do not amount to the same thing for
everyone. However, in the Brown Book the relationships between language
games are not treated in the way that Phillips thinks Wittgenstein treats
them. 57 That is to say, concern is not with how utterances made in
autonomous and isolated language-games affect each other. For example,
when Wittgenstein was inquiring into the connections between language-
games, he was interested in how actually existing and potential but non-
existing language games might interact, such as in the treatment of whether
someone can swim the river (BB §49). What the above discussion suggests is
that there are no readily available analogues in Wittgenstein to the conception
of distinct language-games which Phillips favours.
Phillips appears to wish to handle distinct language-games in a single
94 Mark Addis

uniform way. However, Wittgenstein did not treat language-games in the same
fashion in all his writings. Due to this there could not be a single conception of
a language-game of the kind which Phillips seeks in Wittgenstein. There are
differences between how language-games, when regarded in the sense of a
model for understanding language, are viewed in Wittgenstein’s middle-period
and later work. In the Brown Book, Wittgenstein experimented with the ideas
of seeing language-games as miniature models of what language is, and the
extent to which the language-game method could provide an account of what
language is. In the Philosophical Investigations, he thought that (invented)
language-games can illuminate our language by exhibiting similarities and
differences with it. An important consequence of these differences is that if
Wittgenstein’s changing perspective on language-games is to be taken seriously
then language-games cannot be properly treated by just one account, as Phillips
attempts to do. For instance, Phillips discusses the idea of the completeness of
language-games. He observes that a reason why Wittgenstein remarked that
every language game is complete is that he wanted to remove the assumption
that all propositions have a general form.58 However, Wittgenstein was not
concerned with the notion of the completeness of language-games in the Brown
Book. In his later work, he tended to think that focusing on the completeness
of language-games obscures the way in which they should be used in the
understanding of language because the goal of the philosopher is not
completeness or exactness but the resolution of philosophical problems. What
this example points towards is that Phillips’ uniform treatment of distinct
religious language-games would not fit comfortably with Wittgenstein’s ideas
about language-games.
A significant issue is whether there could be substantive justification for
the idea of distinct religious language-games from Wittgenstein’s concept of
grammar.59 The grammatical ideas found in Wittgenstein could have varying
degrees of applicability to distinct religious language-games, but stating that
there is this applicability is not the same as providing the required
justification. In order to supply this justification it is necessary to extend his
idea of grammar to produce a grammar of a distinct religious language-game.
This raises the question of whether there are any exegetical grounds for the
extension of Wittgenstein’s notion of grammar in this fashion. Examples of
the way in which there would be extensions are the ideas that there is a
distinct identifiable grammar for a distinct language-game, and that there
are grammars for distinct language-games which are clearly separable from
one another. What these cases suggest is that exegetical backing for the kind
of extension required of Wittgenstein’s concept of grammar is problematic.
On a more speculative note, it is worth considering why Phillips might be
tempted to adhere to the interpretation that he does about Wittgenstein’s
attitude to the idea of distinct religious language-games. A reason for this
might relate to one of Phillips’ overall goals in the philosophy of religion,
namely to defend religious beliefs against criticism. If there are distinct
religious language-games (taken in conjunction with other views that Phillips
D.Z.Phillips’ fideism in Wittgenstein’s mirror 95

holds) which are only comprehensible to those who share in the form of life,
it follows that anyone who does not partake in the relevant form of life cannot
criticise these distinct religious language-games. The desirability of this
position for the defence of religious beliefs against criticism is evident, and
therefore a possible motivation for Phillips’ reading of Wittgenstein on
distinct religious language-games emerges.

Wittgenstein’s objective when doing philosophy was to dissolve philosophical
problems.60 He exclusively used grammatical investigation when considering
philosophical problems as he regarded it as the only method which could
clear up philosophical confusion. For Wittgenstein philosophical inquiry is
not the most general of the sciences (PI §109): rather philosophy makes itself
redundant by exposing the abuses of language which have generated it (PI
§§109 and 255). He thought that philosophy was purely descriptive. Phillips’
view is that Wittgenstein’s concept of philosophy leaves everything where it
is.61 He was not altering anything that lay before him, but was aiming to be
clear about it. There is a difference between the idea that Wittgenstein’s
concept of philosophy leaves everything as it is and the view that it does not
permit criticism of any kind at all. Wittgenstein did not leave certain kinds of
rationalism and scientism, and the criticism, justification and explanation of
religion stemming from them, where they are. He achieved this by appealing
to what already lies before us. Wittgenstein sought to reflect and elucidate
the grammar of religious notions. It is inappropriate to request the meaning
of a religious language-game if asking this implies that there could be some
other answer than explication in religious terms of the content of this
language. Wittgenstein’s concern with religious practice was closely linked
with his view that philosophy leaves everything as it is. His focus on religious
practice has been deeply misunderstood in contemporary work on the
philosophy of religion. Phillips maintains that it is not possible to give a
simple account of what Wittgenstein’s philosophy amounts to.
A significant question is whether Wittgenstein’s concept of philosophy
permits the possibility of criticism of religion in a way which goes against
Phillips’ position. For example, in the first of the ‘Lectures on Religious
Belief’, Wittgenstein was at pains to stress that the relationship between the
believer and the atheist is not one of contradiction. However, when he
proceeded to consider whether the believer and the atheist understand each
other (LC, p. 55), he was careful to avoid an explicit commitment to the view
that neither comprehends the other. The mere possibility that the atheist could
understand the believer in this fashion cuts directly across Phillips’ attempts
to defend religion against criticism. It remains to be shown that this manner
of viewing the atheist is incompatible with Wittgenstein’s concept of
philosophy. That is to say, there is no immediate move from his notion of
philosophy to the exclusion of this position about the atheist.
96 Mark Addis

It is important to investigate whether Phillips’ account of Wittgenstein’s

concept of philosophy can be challenged, and if so what this implies about
how the application of this concept to religion could be viewed. Phillips
contends that Wittgenstein’s philosophical methodology and his practice are
consistent.62 However, consideration of the latter’s work gives rise to the issue
of whether Wittgenstein’s claim that grammatical investigation is the only
method which can clear up philosophical problems and his actual
philosophical practice are (generally) consistent. For example, it is arguable
that Wittgenstein’s handling of the philosophy of mathematics, was
inconsistent with his professed philosophical methodology. If the matter of
inconsistency occurs with regard to the philosophy of mathematics then
inconsistency with respect to the philosophy of religion ought to be seen as a
serious possibility. The possibility is particularly pressing when taken in the
light of Wittgenstein’s fragmentary remarks about religion. For example, in
the ‘Lectures on Religious Belief’ Wittgenstein discussed the issue of evidence
as part of his overall inquiry into religious belief. He contrasted the grounds
upon which empirical and religious beliefs are formed. Wittgenstein
distinguished religious beliefs in part by what he regarded as their
unshakeability. He considered the evidence for religious belief in the first and
second lectures, such as in the discussion of the Catholic priest (LC, pp. 57–
9) and Lourdes (LC, pp. 60f.) respectively. When reviewing Wittgenstein’s
attitude to the evidence for religious belief, there seem to be at least four
possibilities with regard to what is deemed to be acceptable evidence. These
are what counts as evidence for the believer, and what for the non-believer,
plus what counts as evidence for the believer but not for Wittgenstein, and
what counts as evidence for Wittgenstein. The alternative possibilities suggest
that there is dispute over the standards of evidence. In the case of what counts
as evidence for the believer but which Wittgenstein found unacceptable, it
looks as though he was not following his professed philosophical
methodology but was rather making a substantive judgment about what
acceptable evidence is. What Wittgenstein potentially appears to permit as
suitable philosophical practice in the philosophy of religion puts the topic of
consistency into sharp focus. The use that Phillips makes of the claim that
Wittgenstein’s philosophical methodology and practice are consistent has to
take account of the significant dispute over whether Wittgenstein is actually
consistent. Phillips needs to be careful about how much weight he places on
this claim about Wittgenstein’s consistency when he is constructing his
account of the latter’s philosophy of religion.
Phillips professes to be following Wittgenstein’s suggestions in the
philosophy of religion.63 He defines his objectives in the philosophy of religion
with respect to what he takes Wittgenstein’s conception of philosophy to be.64
Phillips remarks that Wittgenstein is not altering anything that lies before us
but is trying to be clear about it, and that this should be the philosophical
ideal in the philosophy of religion. Phillips claims that he has a purely
descriptive approach to the philosophy of religion.65 However, it may be
D.Z.Phillips’ fideism in Wittgenstein’s mirror 97

questioned whether Phillips actually does. That is to say, he could be doing

more than just describing religion. For instance, Phillips states that the
meaning of God’s reality is to be located in His divinity which is expressed in
the role that worship plays in people’s lives.66 It is possible that a religious
believer would not think that Phillips’ view would leave everything as it is.
This is because a believer could regard religious practice, such as prayer, and
theological justification—such as that about the nature of the Trinity—as
both being essential parts of religion. It could be argued that Phillips
emphasises the aspect of religious practice at the expense of what are
normally deemed to be fundamental components of religion.67
There is dispute over whether it is reasonable to attribute a philosophy of
religion to Wittgenstein. Kerr, for example, holds that he did not have one.68
It is plausible to maintain that Wittgenstein did not possess a philosophy of
religion as such but that his philosophical ideas, like those about the
relationship between mind and body, are applicable to questions about
religion. Wittgenstein does not appear to have been sure how his religious
and philosophical ideas meshed. This is evinced by his equivocal attitudes
towards religion, and his lack of explanation about how his religious and
philosophical views were connected. Conceivably what we have may be
nothing more than a collection of remarks about religion by Wittgenstein
which have no meaningful interplay with his philosophical views.
Wittgenstein’s comments about religion could possibly be misconstrued if
they are investigated from the perspective of presuming that he had a
philosophy of religion. If this is so then the question of the appropriateness of
Phillips’ effort to construct a philosophy of religion is put into sharp focus.
Phillips’ treatment of the concept of internalism and the role that it plays
in his understanding of Wittgenstein are problematic. When objections to his
handling of internalism are considered it is not obvious what Phillips’ denial
that his approach to Wittgenstein’s philosophy of religion is internalist in
nature really signifies. This in turn affects his attempts to characterise his
own position by contrasting it with internalism. Developing a view of the
place which fideism could have in the understanding of Wittgenstein’s
remarks about religion opens up the way to different readings of them.69

1 The phrase ‘Wittgensteinian fideism’ (referred to here as ‘fideism’) appears to
have been introduced by Kai Nielsen, in ‘Wittgensteinian Fideism’, Philosophy,
vol. 42, (1967) 191–209.
2 There is a great deal of dispute over fideism which it is not possible to consider
here. A representative contribution to the debate is Patrick Sherry, Religion,
Truth and Language-Games (London: Macmillan, 1977).
3 Other prominent fideists are Malcolm and Winch.
4 Phillips’ writings are voluminous, so it is necessary to consider only selected
aspects of them. His ‘Belief, Change, and Forms of Life: The Confusions of
Externalism and Internalism’, in F.Crosson (ed.), The Autonomy of Religious
98 Mark Addis

Belief (Notre Dame: University of Notre Dame Press, 1981), pp. 60–92 (which is
a contracted version of his Belief, Change and Forms of Life [Basingstoke:
Macmillan, 1986]) is selected as a key text because of his claims that it is a
representative summary of his views.
5 The question of whether Phillips’ critics have been fair to him will not be covered
here. For his responses to his critics, see his ‘Belief, Change, and Forms of Life’.
6 The position presented here is a kind of fideism (as understood in its general
sense). This is because having religious faith is a prerequisite for taking part in
rational discussions of that faith.
7 ‘Belief, Change, and Forms of Life’, p. 61.
8 Ibid., p. 62.
9 Ibid.
10 Ibid, p. 61. This argument will be referred to as the internalist argument. It is
worth observing that the premisses of the internalist argument constitute an
argument for the conclusion that religious beliefs are beyond criticism from other
aspects of human life. However, none of these premisses give any reason for
holding that religious beliefs do not change. Being beyond criticism is different
from being beyond change. Arguably, Phillips should not be putting these two
ideas together to constitute the internalist position.
11 Provided the premiss ‘Philosophical confusions may be generated by the
obscuring of these distinctive meanings’ is taken in conjunction with this.
12 See ‘Belief, Change, and Forms of Life’.
13 Ibid., p. 62.
14 Ibid.
15 Ibid.
16 Even if this should turn out not to be the strongest form of internalism possible,
it is clearly a very strong version of internalism.
17 ‘Belief, Change, and Forms of Life’, p. 62. Exploring fully the matter of whether
Phillips is an internalist of a weaker kind would lead too far afield and will not
be pursued.
18 The interpretations of Wittgenstein’s writings will not be defended in any
systematic way and the view of these issues will draw upon Gorden Baker’s and
Peter Hacker’s work—see their Wittgenstein: Meaning and Understanding
(Oxford: Blackwell, 1980) and Rules, Grammar and Necessity (Oxford: Blackwell,
19 The meaning of premisses 3–5 of the internalist argument requires further
elucidation. However, this matter will not be considered here.
20 For example, he likes using conversations which are not matched up with
writings that have higher exegetical authority.
21 See ‘Wittgenstein’s Full Stop’ in his Wittgenstein and Religion (Basingstoke:
Macmillan, 1993), p. xiii.
22 See his ‘Religious Beliefs and Language Games’, Ratio, vol. 12 (1970), p. 39.
23 Ibid., p. 41.
24 See Phillips, ‘Mysticism and Epistemology: One Devil of a Problem’, Faith and
Philosophy, 12, vol. 2 (1995):170.
25 See Phillips, ‘Religion and Epistemology: Some Contemporary Confusions’,
Australasian Journal of Philosophy, vol. 44 (1966):318.
26 See Phillips, ‘From World to God’, Aristotelian Society, vol. 41 (1967):151.
27 For instance, Fergus Kerr argues that the concept of forms of life cannot be used
as some fideists like it to be, given Wittgenstein’s texts—see his Theology After
Wittgenstein (London: SPCK, 1997), p. 28.
28 For general views of how Wittgenstein employed the concept of forms of life, see
Rudolf Haller, Questions on Wittgenstein (London: Routledge, 1988), pp. 129–
36; also see Baker and Hacker, Rules, Grammar and Necessity, Chapter V.
D.Z.Phillips’ fideism in Wittgenstein’s mirror 99

29 Baker and Hacker (in Rules, Grammar and Necessity, pp. 238–43) argue that
the concept of forms of life is fundamentally about the cultural aspects of human
nature. They would disagree with the distinction between the two strands of the
notion of forms of life.
30 A different sort of reason for considering the Winch-MacIntyre debate is that it
shows that this major controversy in the philosophy of social science has far-
reaching implications for the philosophy of religion.
31 Peter Winch, The Idea of a Social Science and its Relation to Philosophy
(London: Routledge & Kegan Paul, 1988).
32 Ibid, p. 100.
33 ‘Understanding a Primitive Society’, in his Ethics and Action (London: Routledge
& Kegan Paul, 1972), pp. 8–49.
34 Winch further discusses these themes in his ‘Language, Belief and Relativism’,
Trying to Make Sense (Oxford: Blackwell, 1987).
35 See his Against the Self-Images of the Age (London: Duckworth, 1983), p. 229.
36 This replacement is possible because primitive and religious forms of life are
both identifiable forms of life.
37 See his After Virtue, 2nd edn. (London: Duckworth, 1985).
38 ‘The Devil’s Disguises: Philosophy of Religion, “Objectivity” and “Cultural
Divergence”’, Philosophy, vol. 83 supplement (1984):71.
39 Ibid, p. 62.
40 ‘Belief, Change and Forms of Life’, p. 60.
41 ‘Primitive Reactions and Reactions of Primitives’, Religious Studies, vol. 22
42 ‘Wittgensteinian Fideism’, pp. 207f.
43 There are many theories of incommensurability. The notion of
incommensurability being used here is that speakers are not able to
communicate because their words do not have the same meanings. This is one
version of denying the commensurability of religious and non-religious
44 For a discussion of language-games, see Baker and Hacker, Wittgenstein:
Meaning and Understanding, Chapter III.
45 Phillips offers a variety of different grounds for the idea that there are distinctive
religious language-games. For example, in ‘Wittgenstein’s Full Stop’
(Wittgenstein and Religion) he remarks that the claim for the distinctiveness of
religious language-games grows out of what actual usage demonstrates. It can
never be any sort of dogma or methodological postulate forced on language.
However, these claims are unclear as they stand and thus do not counter the
criticisms of Phillips’ usage of language-games raised above. What this example
suggests is that there is no plausible and easy way to demonstrate that religious
language-games are distinct.
46 See his ‘Religious Beliefs and Language Games’, Ratio, 12 (1970), pp. 26–46.
47 ‘Belief, Change, and Forms of Life’.
48 ‘Religious Beliefs and Language Games’, pp. 45f.
49 ‘Religion and Epistemology: Some Contemporary Confusions’, p. 317.
50 ‘Religious Beliefs and Language Games’, p. 30.
51 ‘Primitive Reactions and Reactions of Primitives’, pp. 167f.
52 ‘Religious Beliefs and Language Games’, p. 30.
53 Ibid., p. 27.
54 Ibid.
55 Ibid., p. 37.
56 See ‘Religious Beliefs and Language Games’.
57 Ibid.
58 Ibid., p. 37.
100 Mark Addis

59 The most complete study of this notion is by Baker and Hacker in Rules,
Grammar and Necessity.
60 For an account of his conception of philosophy, see Baker and Hacker,
Wittgenstein: Meaning and Understanding, pp. 259–93.
61 See his ‘Religion in Wittgenstein’s Mirror’, Philosophy, vol. 90, supplement
(1991): 135–50.
62 Ibid.
63 See ‘Belief, Change, and Forms of Life’.
64 See ‘Religion in Wittgenstein’s Mirror’.
65 ‘Belief, Change, and Forms of Life’.
66 ‘Religion and Epistemology: Some Contemporary Confusions’, p. 320.
67 For example, he comments that there are clashes between different concepts of
what religious truth comes to in different religions, and embracing one of these
concepts as true is itself a religious act (‘Mysticism and Epistemology: One Devil
of a Problem’, p. 183). It is worth noting that there is a connection between the
notions of act and practice. It seems reasonable that instead of his view,
acceptance of one of these concepts could be characterised as adopting a religious
68 See Fergus Kerr, Theology After Wittgenstein.
69 Thanks are due to Bob Arrington and Alan Bailey for comments.
7 Wittgensteinian religion and
‘reformed’ epistemology
Paul Helm

In this paper I wish to explore some of the similarities and differences between
the philosophical approach to religion that is characteristic of Wittgenstein
(WR), as this approach has been identified and developed by some of his
interpreters, and what has come to be called ‘reformed’ epistemology (RE).
The exploration is confined to epistemological and metaphysical issues1

Call ‘externalism’ the view that to be rationally tenable religious belief
requires vindication by considerations which are non-religious. Both RE and
WR2 deny such externalism. In the case of RE the denial of externalism is an
epistemological claim arising out of a critique of classical foundationalism,
while in the case of WR the denial of externalism is more principled and
more radical, being maintained on conceptual or ‘grammatical’ grounds.
The sense in which RE denies externalism can be brought out by briefly
reviewing its critique of foundationalism in religion.
Foundationalism is the view that lies at the heart of classical natural
theology, and at the heart of foundationalism lies a principled contrast
between propositions which are foundational (because, for example, they are
not inferred from any other propositions because they are evident to the
senses) and further propositions inferred from such foundational
propositions. The relation between foundations and superstructure is thus
asymmetrical. If p supports q and r, then q and r cannot support p. And if S
is the set of foundational propositions, then any justifiable proposition is
supported by S. Foundationalism asserts that a proposition can be justifiably
believed only if it is either (in Plantinga’s terminology) ‘properly basic’ or is
suitably related to a proposition which is properly basic. We find such a view
in Aquinas: ‘From effects evident to us, therefore, we can demonstrate what
in itself is not evident to us, namely, that God exists.’3 In this case the evident
is what is foundational, the non-evident is what is inferred from what is
evident. Aquinas attempts such demonstrations in his Five Ways, moving in

102 Paul Helm

typical foundationalist fashion from what is evident to us to what is not.

Thus, it is evident to us that some things move, and from this (he believes) it
can be soundly argued that there must be an unmoved mover, God. Whether
Aquinas also holds that it is necessary, in order for faith to be epistemically
justified, that the existence of God can be demonstrated in this way, or merely
sufficient, is an interesting additional question.
So let us suppose that Aquinas, together with a host of other philosophers,
argues that one is justified in believing that God exists only if there is evidence
from which the existence of God can be derived. Or perhaps it is more
accurate to suppose that he argues that one is justified in believing that God
exists only if someone can derive the existence of God from evidence available
to anyone. One’s belief that God exists is not justifiable without such
evidence. According to Plantinga, this means that Thomas must have a
certain view about what constitutes a rational noetic structure, the structure
possessed by the set of propositions that a person believes in so far as he is
rational, together with all their logical connections. A rational person’s noetic
structure is constituted by a basis, his or her basic beliefs, together with what
is built on this basis, the beliefs which are justified by the basic beliefs.
Nothing justifies the propositions in the basis. It is only by coming to believe
in accordance with such foundations that rationally justifiable beliefs can be

According to the foundationalist a rational noetic structure will have a

foundation—a set of beliefs not accepted on the basis of others; in a
rational noetic structure some beliefs will be basic. Non basic beliefs, of
course, will be accepted on the basis of other beliefs, which may be
accepted on the basis of still other beliefs, and so on until the foundations
are reached. In a rational noetic structure, therefore, every non basic
belief is ultimately accepted on the basis of basic beliefs.4

So foundationalism is a theory about a type of structure of beliefs, some

beliefs being basic, others being built upon and supported by these basic ones.
Plantinga gives the name strong foundationalism to the view that a
proposition is properly basic, properly among the foundations of a person’s
noetic structure, if and only if that proposition is either self-evident to that
person or (less demandingly) is evident to that person’s senses. So on this
interpretation of Aquinas, he was a strong foundationalist in respect of
theology. To be strongly foundational a proposition must have such a
character that it would be irrational for any rational person to deny its truth:
so that, according to strong foundationalism, the superstructure of belief
must be built upon propositions that are, say, evident to the senses or of which
the mind has clear and distinct ideas.5 This is the dominant tradition in
modern Western philosophy, the tradition inaugurated by Descartes and
Locke. Let us briefly consider Plantinga’s critique of strong foundationalism,
which begins his critique of externalism in the justification of religious belief.
Wittgensteinian religion and ‘reformed’ epistemology 103

Plantinga makes two criticisms of this view. The first is that it is too
restrictive. For if strong foundationalism were true, then it would rule out as
irrational myriad beliefs which we accept unquestioningly—for example,
beliefs about other people than myself, and about other times than the
present, to look no further.

I believe, for example, that I had lunch this noon. I do not believe this
proposition on the basis of other propositions; I take it as basic: it is in
the foundations of my noetic structure. Furthermore I am entirely
rational in so taking it, even though this proposition is neither self-evident
nor evident to the senses nor incorrigible for me.6

There are two points being made here. One is that as a matter of fact there
are many propositions we believe, and are entitled to believe, which do not
rest upon other, more basic, propositions that make them rationally
believable. Plantinga is making a move here that is typical of that type of
epistemologist which Roderick Chisholm has called ‘particularist’. He is
appealing to particular examples of belief, noting their character, and in effect
saying: ‘If anything is a case of rational belief, this is. And so no theory or
criterion of what rational belief must be like which calls the rationality of
such beliefs into question can be warranted.’ The other point is that such
propositions are to be believed, i.e. we are entitled to believe them even
though they are not evident to the senses. In other words, strong
foundationalism is unnecessary for epistemic justification. As far as one can
see, Plantinga does not give an argument for this view, but makes an appeal
to our normal procedures. He does not say that we cannot provide a
justification for our everyday beliefs, but that we need not do so. (Perhaps he
could appeal to a principle of credulity here.) We do not first have a theory
and then accept those beliefs which accord with the theory; we have certain
beliefs which we take to be paradigmatically rational.
His second criticism is more serious. Plantinga claims that such strong
foundationalism is self-referentially incoherent. That is, it does not satisfy, in
itself, the conditions it lays down for the rationality of any belief. It states
that a proposition is properly basic for a person only if that proposition is
evident to that person’s senses. But this proposition, the one that I have just
identified, is itself none of these things. Strong foundationalism is an
interesting philosophical proposal. It is one that many philosophers have
accepted. But it is not self-evidently true and therefore we are not required,
even as strong foundationalists, to accept it. The paradox is that no strong
foundationalist need accept strong foundationalism. It is neither self-evident
nor incorrigible nor evident to the senses.
So strong foundationalism is indefensible as offering a necessary condition
for any rational belief. Though we may opt for strong foundationalism, there
is nothing compelling about it. But, as I have noted, having rejected strong
foundationalism Plantinga does not for that reason reject foundationalism in
104 Paul Helm

all its shapes and sizes. He reckons that our noetic structures do have a
foundationalist character. But if strong foundationalism is not required, if it
cannot be required for the reasons given, then other versions of
foundationalism, if not required, are most certainly permitted. And among
these other versions, according to Plantinga, is what might be termed theistic
foundationalism. A person is entirely within his epistemic rights in believing
that (say) God has created the world, even if he or she has no argument for
this. The belief that God exists and has created the world can be basic and
thus form part of the foundations of a person’s noetic structure.
The presence of such beliefs in the foundations of a person’s noetic
structure is not arbitrary, according to Plantinga, if they are properly
grounded, as they might be as a result of certain experiences that person has
had, or having certain beliefs ineluctably formed in one on certain occasions.
In Faith After Foundationalism7 D.Z.Phillips offers a root-and-branch
critique of RE, but it has to be said that this critique contains inaccurate and
misleading characterisations of RE and consequently makes the differences
between RE and WR sharper than is in fact the case. He claims that by
comparison with the sort of embeddedness that basic propositions in the
language-game of religion have, according to WR, the basic propositions of
RE seem isolated and even arbitrary.8 But this is to misunderstand the
position, for the following reason.
Take the situation where a person believes in God, such belief being at the
center of a whole host of other beliefs and activities, indeed at the center of
his life. Plantinga himself suggests such connectedness in relation to personal
beliefs about guilt and deliverance, appreciation of the natural world, and so
on. In view of this it seems plainly inaccurate for Phillips to say, about RE,
that ‘we are not shown the way in which belief in God underlies other things
in the noetic structure’.9 That is the very thing we are shown! For what RE
characteristically asserts is that numerous claims abut God, about the fact
that he exists but more especially that he is the creator and that he forgives
sin are properly foundational for the believer, and that such claims are as
central to his or her life as are claims about memory, say, or about other
Phillips at times comes near to seeing this connectedness. He recognises
that for the RE ‘belief in God seems to be placed in the context of a living
faith in which it has its sense’.10 And it is plainly false to say, as Phillips does,
that Plantinga treats the conditions (in which it is appropriate to say that
God exists) as no more than prima facie justifications of belief in God. They
are such justifications, but they are a lot more than that. And they are prima
facie justifications in the sense in which my belief that I am sitting on a chair
is prima facie justification for believing that there is a chair in the room. My
belief may be defeated, for I could be mistaken; but such a belief can also be
reinstated, since the defeater can itself be defeated. In the meantime, in the
absence of defeaters, I continue to justifiably believe it is a chair that I am
sitting on. In the same way the religious believer justifiably continues to shape
Wittgensteinian religion and ‘reformed’ epistemology 105

his life by his religious beliefs. From a theoretical point of view, according to
RE belief in God is probable, like the belief that I’m sitting on a chair, and my
belief that mountains have existed for a very long time. But in the absence of
defeaters that belief amounts to certainty for all practical purposes, like the
belief that I presently have that I’m sitting on a chair.
As RE characterises religious belief, such belief does not start from the
basic propositions in a temporal sense but only in a logical sense. When
Phillips says11 that one cannot start from these propositions because they have
their sense and are held fast by all that surrounds them, he appears to be
denying the possibility of being able to reflect rationally upon one’s belief
and their grounds while still holding them fast, the second-order activity that
is characteristic of much philosophy, including RE. In rationally reflecting on
the foundations of religious belief, RE is not proposing that the believer starts
from such propositions in the sense that he first establishes, in isolation, the
foundational propositions and then proceeds step by step to build the
superstructure. Rather the reverse. For Plantinga religious belief is prima facie
justified just in case it is part of the way of life of the believer. In no sense is
Plantinga holding that for the believer the belief that God exists is a
provisional belief, provisional until he can discover the appropriate criterion
of justification. Rather, the believer is fully and permanently entitled to have
the proposition that God exists in the foundations of his noetic structure.12
There is also evidence of confusion in Phillips’ discussion over what exactly
RE’s version of foundationalism implies. For Plantinga, the claim that under
certain circumstances I see a tree is properly basic, but not strongly
foundational. That is, under certain circumstances a person is entitled to have
such a proposition in the foundations of his noetic structure. Nor is Plantinga
searching for ‘an incorrigible proposition of sense experience’. 13 He
specifically denies that incorrigibility is necessary for proper basicality. Nor
is Plantinga searching for ‘conceptions of minimal experiences, experiences
which are immediate and cannot be mistaken’. Nor does he say that certain
experiences are the justifications for saying ‘I see a tree’.14
So WR and RE have this in common: the belief that God exists is not an
isolated belief, contingently related to other beliefs about him and other
matters and the practice of religion. Where they differ is over the sense of
asking second order questions about such beliefs. RE admits a high degree of
reflectiveness about religious belief and its epistemic foundations, but for
Phillips the very idea of critically reflecting upon one’s beliefs in this vein
betrays a failure to understand the sense of religion. Another difference is
over the propriety of admitting defeaters. For Plantinga there is distinction
between what is necessary to make religious belief reasonable and what might
defeat the reasonableness of the belief. Nothing that has a non-religious
character is required for making religious beliefs reasonable, but such a
reasonable belief may be defeated by evidence; for example, the
reasonableness of the belief that God exists may be defeated by the existence
of evil. The reasonableness of the belief will be reinstated if and when the
106 Paul Helm

defeater provided by the existence of evil is in turn defeated by

counterargument. Finally, RE claims that although natural theology is not
necessary to establish the reasonableness of belief in God, sound instances of
natural theology may be possible, and they may serve to strengthen the
believer’s faith.
Let us now look at Wittgensteinian philosophy of religion. What does the
denial of externalism mean here?
Fundamental to WR is a general claim about meaning as use. The
meaning of any expression is to be sought by understanding the part it plays
in the language of the language-game of which it forms a part. Every
meaningful utterance is so only because its utterer participates in some
language-game. It is no different with religion: an expression such as ‘God’,
or ‘eternal life’ or ‘creation’ is to be understood solely by reference to the
way that it is used, to the ‘form of life’ in which it is embedded. So
discovering what an expression means is not only the first step in its
epistemic justification (though, as we shall see, WR has reasons for disliking
the idea of epistemic justification as distinct from establishing the meaning
of an expression): it is also the last step. Manifesting the use of the
expression in its users’ language-game, their form of life, is both necessary
and sufficient for its legitimacy as a religious utterance. In this sense
justification is an ‘internal’ matter.

There is no question of justifying the criteria for our use of ‘physical

object’: that is how we do use the concept. The comparison with the
reality of God was meant to be at this grammatical level. In each case
there would be no question of a general justification of the criteria for
distinguishing between the real and the unreal.15

To this rebuttal of the need to provide a justification of what establishes the

reality of God there corresponds a point about the sovereignty or
transcendence of God. Is it not impious to require an explanation of the
existence of evil, say? That is, to the denial of externalism in justification
there corresponds a distinctive religious sensibility, one in which the very
thought of calling God into account by seeking some justification or
explanation constitutes a deep-seated (albeit pervasive) religious error. And
there is the corresponding claim by WR that when religious people submit to
the will of God, or attribute some evil to the will of God, this is not an
explanation of why there is evil but is rather an attempt to bring the anguish
of the evil to an end.16 This point is in turn merged with the idea that
Christianity is not a doctrine or a set of doctrines, not a theory, but a way of
living, and that the job of philosophy is to offer not justifications—of religious
belief or of anything else—but descriptions.
‘Words have meaning only in the flow of thought and life.’17 Language-
games delimit explanations. There cannot, in turn, be an explanation of
language-games. They are just there, because that’s how human life is.
Wittgensteinian religion and ‘reformed’ epistemology 107

(Though it has to be said that there is not much suggestion in Wittgenstein

himself, as far as one can see, that religion is a language-game).
So the denial of externalism in WR is not simply a denial of externalism in
justification—as it is in RE (with a minor qualification, in that RE defends
the foundationalist pattern of justification)—it is a denial of externalism in
meaning and understanding. It is because the meaning and understanding of
religion is displayed only within the language-game of religion that the
question of its epistemic justification cannot arise. The meaning of religion
within the language-game is a sufficient justification for its use, and so the
further demand for epistemic justification is idle. Religion has its own rules
of meaning; it is conceptually autonomous, quite discontinuous with other
forms of life.18 So that justification for a religious concept or practice is to be
sought ‘internally’, by reference to the form of life of religion.
Take the question of the proofs of the existence of God. WR resists the
idea that the existence of God, in order to be reasonably believed, must be
justified by sets of general considerations, as in classical natural theology.
And this is because such general considerations are alien to religion; appealing
to them is in fact importing into religion ways of thinking and reasoning
which compromise its autonomy and so corrupt it. The ‘justification’ of ‘the
existence of God’ has solely to do with how that expression is used. Within
religion it makes no sense to raise the question of the existence of God; it is
the cornerstone of the practice. Persistent religious doubt never signals
epistemic doubt but rather a religious reluctance to participate positively in
the game, though such doubts gain their sense from the practice of religion.
Such doubts are to be resolved not by attempting to provide evidence to show
the reasonableness of religion, but by making a renewed effort to show how
religion might continue to address a person’s deepest concerns and needs.
And similarly with historical research, regarding the historical claims made
in Christianity or Judaism. To suppose that the results of such research might
support, or undermine, religion would be to confuse the language-game of
history, or of historical research, with the language-game of religion and so
make religion a prey to external factors.
So the language-game of religion is invulnerable to ‘outside’ considerations
and is only vulnerable to the question of whether religion has a role in a life.
Someone who says that religion means nothing to him is simply claiming
that he finds no nourishment or satisfaction in the characteristic concerns
and claims of religion. And if a situation arose in which no one any longer
was religious, in this sense nothing would have been shown about the
reasonableness of belief in God. Nor would God have ceased to exist.
William Alston has suggested that the thesis of the autonomy of language-
games is due to the epistemising of truth,19 the defining of truth in epistemic
terms, but this is not clear. For one thing, as I have already noted, WR does not
have much room for the concerns that look epistemological in the conventional
sense, i.e. concern for ‘epistemic standards’.20 According to WR there are no
epistemic standards which the language-game of religion has to meet (any such
108 Paul Helm

proposal would be externalism, compromising autonomy) and (as far as I can

see) no epistemic standards which speech within the language-game must meet;
only, as in the grammar of any language, standards of appropriateness, rules
for what to say and what not to say. Contrary to Alston, while WR has a
concept of truth, it is not an epistemised concept, one that can be only defined
in terms of human cognitive states.21 It would be more accurate to call it a
‘semanticised’ concept of truth; for, as has already been stressed, the truth of
religious utterances and the appropriate use of religious concepts are internally
related to their meaning, their place in the language-game of religion. As
Phillips puts it: ‘the religious believer is not prepared to say that God might not
exist. It is not that as a matter of fact God will always exist, but that it makes
no sense to say that God might not exist.’22
What WR seems to deny (among other things) is the possibility of
detachment, the possibility of both having an attitude to God which is
affective and engaged and of being prepared to consider the existence of God
as a hypothesis for which grounds might be given. To give the existence of
God such consideration is to remove God from his role in religion and to
treat his existence like a scientific hypothesis. It is to become abstracted from
God, and so to display an attitude to the existence of God that is
fundamentally irreligious.
Such a view is sometimes called ‘attitudinalism’ or ‘expressivism’.23 But
this is, once again, inaccurate as far as WR is concerned. For an attitude may
be contingently related to what it is an attitude to. I might fear the raging sea,
you might exult in it; we each recognise the raging. But, for WR, religion is
not a set of attitudes like that, attitudes to some experience to which another
person might have different attitudes. Rather (as Phillips repeatedly says),
the very sense or idea of a life is given in those attitudes. It is the fact that a
person considers the existence of God as a hypothesis that rules out this
attitude as religious and makes it foreign to religion. Such a person has a
different concept of God from the concept which (according to Phillips)
functions in religion. Certainly this is not metaphysical realism, but neither is
it ‘attitudinalism’ as, say, emotivism in ethics or R.B.Braithwaite’s account of
religious belief may be understood.
Since the religious form of life gives sense to religious utterance, such sense
may be given to an expression which has the surface grammar of a truthclaim
which is most naturally to be understood as making a claim about reality as
such, a sentence such as: ‘God is the creator of all that is and would exist
whether or not anything else existed.’ Such an expression may have a use in
WR but its sense is not a metaphysical one but an expression of God’s
transcendence as opposed to our contingency and transience. (Phillips might
say at this point: Think how expressions such as “Before the mountains were
brought forth, or ever thou hadst formed the earth and the world, from
everlasting to everlasting, thou art God” [Psalm 90:2] function within
religion.’) Such expressions have a sense within religion: they are to be
understood in terms of awe and wonder at God’s permanence.
Wittgensteinian religion and ‘reformed’ epistemology 109

Furthermore it is not clear that Phillips is motivated by the need to

purchase ‘absolute epistemic autonomy’,24 or the need to find a stopping place
for the epistemic assessment of particular religious beliefs.25 He is not
concerned to take up any attitude to traditional religious epistemology (as it
might be called) other than saying that such epistemology and its
characteristic concerns are misguided in principle. There is certainly no
evidence that WR is designed to ‘protect’ religion from epistemological
examination, even though it may have this effect.
Further, WR has a ready reply to Alston’s suggestion that there are
‘language-games specially designed for examining, describing and evaluating
other language-games’.26 This cannot be so; for, according to WR, one of the
ways in which the autonomy of language-games is made manifest is that it is
only in the playing of such games that the meaning of the language (e.g. of
religion) is disclosed. A ‘language-game’ designed to examine religion by, for
example, assessing its epistemic foundations would, as we have seen, rest on a
misunderstanding of the distinctive character of religion. And similarly for all
other language-games. To suppose that there is a language-game which might
examine and evaluate the language-game of religion is to suppose a common
language and so a common meaning between the two, and in doing so to violate
the conceptual autonomy of each, something which WR explicitly resists.


Foundational versus fundamental beliefs

Both RE and WR have an account of the role of basic propositions in religion,
but these accounts differ significantly, even allowing for the fact that there is
some difference of approach among those who make an appeal to
Wittgenstein. One approach to such basicness may be seen in Anthony
Kenny’s response to RE.
Kenny takes as an instance of a fundamental belief Wittgenstein’s example
‘The earth has existed for many years past’; and Kenny himself offers the
example ‘Human beings sleep’. Kenny contends that while reasons can be given
for such propositions, we do not believe them on the basis of such reasons but
take them to be fundamental. What makes a proposition fundamental is not
merely that it is basic (as the belief that there is writing on the computer screen
in front of me now is basic for me) but that so much turns on it were it to be
denied. If we seriously supposed it to be false, this would wreak havoc in our
noetic structures. For truly to give up, or seriously to question, the proposition
‘Human beings sleep’, for example, would be to throw doubt on the reliability
of the very processes of enquiry. For if one cannot rely upon the truth of a
proposition such as ‘Some people sleep,’ what can one trust?
When Kenny applies this criterion of what counts as a belief being
fundamental to the case of belief in the existence of God, he claims that belief
in God cannot be fundamental, ‘in the sense of being something that is
110 Paul Helm

accepted as basic by all those who have an opinion on the matter’.27 ‘While it
is possible for an individual to accept the existence of God as basic…it is
equally clear that the existence of a God with attributes resembling those of
the God of Western theism is not something which has been universally
believed by the human race’.28 It is not therefore a fundamental belief. It is
not universally believed, and therefore cannot be a fundamental belief in the
required sense. Here Kenny comes close to equating ‘fundamental’ to basic
in the sense in which this is used in strong foundationalism.
But ‘universally basic’29 is ambiguous. It can be taken in a distributive
sense, as a remark about the belief structures of everyone. When taken in this
sense, it is doubtful if there is a common set of such fundamental beliefs (in
Kenny’s sense) held by all intelligent and perceptive human beings, i.e.
universally believed by the human race. Take, for instance, the claim that
Australia exists (another of Kenny’s examples), or that the earth has existed
for many years. I suppose that not every intelligent and perceptive human
being who has an opinion on the matter of the existence of Australia or the
age of the earth holds that Australia exists or that the earth has existed for
many years. The earth is approximately spherical’ seems another good
candidate for a fundamental proposition, but I don’t suppose that all
intelligent and perceptive human beings at present believe that the earth is
approximately spherical, and in any case this seems to be a belief that has
only fairly recently been acquired by human beings. So there are very few, if
any, universally basic propositions.
And certainly Wittgenstein did not think that in order for a proposition to
be fundamental it must be held by everyone in an unvarying fashion. For
Wittgenstein: The same propositions get treated at one time as something to
test by experience and at another as a rule of testing.’30 So a fundamental
belief for Wittgenstein is always a fundamental belief held at a time.
According to Wittgenstein, in any given situation of life there are propositions
that we presuppose, propositions such as: ‘I was born’; The earth has existed
for many years’; ‘There are human beings’. The truth or falsity of such
propositions is not investigated simply because their acceptance underlies the
very business of investigating questions of truth and falsity in regard to other
matters. The fact that we presuppose them in this way does not mean that
such propositions are necessary truths; they may have been discovered and
then become part of the stock of presuppositions. Nor does it mean that they
are ‘self-evidently true’, for there may be those for whom such claims are not
self-evidently true. And (presumably) as a result of cultural and other changes
certain such propositions may cease to be presuppositions. Such propositions
underlie our normal activities in a tacit and unquestioned way; they are taken
for granted.

I am told, for example, that someone climbed this mountain many years
ago. Do I always enquire into the reliability of the teller of this story, and
whether the mountains did exist years ago? A child learns there are reliable
Wittgensteinian religion and ‘reformed’ epistemology 111

and unreliable informants much later than it learns facts which are told it.
It does not learn at all that that mountain has existed for a long time: that
is, the question whether it is so doesn’t arise at all. It swallows this
consequence down, so to speak, together with what it learns.31

Anyone who has had experience of the curiosity of children will think that
what Wittgenstein says here is plainly false!
Phillips understands the role of such fundamental propositions rather
differently from Kenny. Kenny envisages them as one of a disjunctive set of
foundational beliefs which might justify belief in God (though Kenny thinks
that they fail to do so).
According to Phillips, by contrast, it is precisely their role as basic
propositions (in whatever language-game they feature, the language-game of
religion or some other) that exempts them from doubt for anyone who
participates in the language-game. So the basic propositions are
nonhypothetical,32 not liable to be overturned, not because they are logically
necessary33 but because they are groundless34 but given in the form of life.
The basic propositions show how things are,35 for there is no conception of
how things are that is independent of all human practices.
But the basicness of such propositions is not such that they cannot change.
Some propositions can be demoted from performing the role of being basic
in this sense, and others can be promoted. Phillips allows that there is two-
way traffic between grammar, and facts and theories. It is particularly
obvious in the case of science—but presumably it is true also of religious
language-games, though perhaps less obviously so.

Think of chemical investigations. Lavoisier makes experiments with

substances in his laboratory and now he concludes that this and that
takes place when there is burning. He does not say that it might happen
otherwise another time. He has got hold of a definite world-picture—
not of course one that he invented: he learned it as a child. I say world-
picture and not hypothesis, because it is the matter-of-course foundation
for his research and as such also goes unmentioned.36

So one big difference between the epistemology of RE and WR is this. While

RE readily participates in second-order debates about how the reasonableness
of religious belief is to be established, the typical approach of WR is to be
nonreflective in the sense that the philosopher’s activity is focused on the
removal of what is regarded as harmful second-order discourse, i.e. to
eliminate a certain kind of harmful reflection on our beliefs and actions—
and to remove a certain kind of doubt, the sort that rests upon the proposition
in question being hypothetical and so open to refutation. Thus the relation of
basic propositions to the rest in WR is not that they are self-consciously
chosen as a result of rational deliberation and reflection, i.e. as a result of the
use of a criterion in terms of which certain propositions are accepted into the
112 Paul Helm

foundations of a person’s noetic structure, and others not accepted. Rather,

they are just there as part of the very character of human life.
There is some difference of interpretation here. For instance, Kenny seems
to be happy with an interpretation of Wittgenstein’s remarks in On Certainty
that treats them as foundational in a more conventional sense than does
Phillips. But both seem to agree that such basic statements are for
Wittgenstein relative to time and place; but, when in place, hosts of other
propositions which people believe logically depend upon them such that it
would make no sense to doubt them.37

The reality of God

Metaphysical realism38 can be variously characterised. It is sometimes

understood in terms of the existence of things which have their reality
independently of the human mind. They are in no sense constructed by the
mind, or depend otherwise for their existence on the mind. In more recent
discussion the notion has been further refined, so that the issue of realism
versus anti-realism has been taken to be a controversy over how a certain
discourse—in our case religious discourse, but the same issues may arise, say,
over scientific discourse—may be understood. According to the realist what
the discourse is about exists independently of the existence of the discourse.
Were there to be discourse or human thought about some matter, if that
matter exists it really exists. The assertions that the discourse makes are true
or false depending on whether or not they correspond to what exists. For the
anti-realist the sense of what exists and does not exist is given only in the
discourse about it.
On the realist understanding of the existence of God, God exists (or not)
independently of our discourse about him. And the claim that God exists is a
claim about a matter of fact, as is the claim that God does not exist. Whether
that claim is true or not thus depends upon whether some divine reality
outside our discourse exists, and whether what we say about that reality
corresponds to the state of affairs outside our discourse, or not. So that a
statement that God exists is true, on the realist understanding, if and only if
God exists. And, most crucially, the fact that God exists, the fact that makes
our statement that God exists true, if it so be that God does exist, is a fact
that obtains quite independently of any and all of our attempts to gain access
to that truth. We gain access to truths by theorising about them, by inventing
concepts, by conducting certain sorts of investigation, and so on. The realist
claims that what makes any of the assertions within these investigative
procedures true is independent of any of these procedures.
So realism is a metaphysical thesis, a thesis that what is true is so
independently of human knowledge and belief. Not independently of what
human beings can in fact know at any one time, but independently of what
can in principle be known.
The proponents of ‘reformed’ epistemology such as Alvin Plantinga clearly
Wittgensteinian religion and ‘reformed’ epistemology 113

commit themselves to the metaphysical realism of God in this sense, making

a sharp contrast between the epistemic status of claims about God’s existence,
and the truth status of such claims. As Plantinga puts it, ‘Particularism does
not imply subjectivism.’39 The fact that some belief is not shared by everyone,
or is not available to everyone, does not mean that it is simply a report of the
believer’s own subjective states.
In ‘How to Be an Anti-Realist’40 Plantinga develops a more nuanced view. On
the one hand he argues against what he calls the ‘creative anti-realism’ of Hilary
Putnam and Richard Rorty, claiming an unwarrantable conflation between what
is verifiable and what is true. On the other hand he recognises the force of the
impulse that leads to one version or another of anti-realism, the force of
wondering how whatever is true but unverifiable could be true. His resolution of
this situation is to suppose that while there can be truths independent of human
minds there cannot be truths independent of God’s mind. Here he cites Aquinas:
‘Even if there were not human intellects, there could be truths in relation to the
divine intellect. But if, per impossible, there were no intellects at all, but things
continued to exist, then there would be no such reality as truth.’41 So, for
Plantinga, we might say that anti-realism with respect to mundane matters is
true (or at least appealing), but that is so only because supra-mundane anti-
realism is false. God, the ens realissimum, whose existence is necessary, exists
whether we are able to verify his existence or not. If anything, this makes the
metaphysical contrast between RE and WR even starker.
Such metaphysical realism and RE are connected in the following way.
Religious belief, belief that God exists, is held to be reasonableness insofar as
the proposition that God exists forms part of the basis of a person’s noetic
structure. Such a person has grounds for thinking that God exists, even
though he may have no evidence. He does not simply trust that reality is as it
is by an act of blind faith. Nor does he have infallible grounds, knowing
beyond any doubt that God exists. He has fallible grounds which make the
belief reasonable. And things are no different with my reasonable belief that
I am sitting on a chair or that I have existed in the past.42
It might, at first glance, be less clear whether WR is committed to
metaphysical realism in this sense. There is much talk in WR of the ‘reality’
of God, and so we must move with caution. It is clear, for instance, that WR
wishes to distance itself from out and out expressivism or fictionalism in
religion: from a position of the sort avowed, for example, by Robin Le
Poidevin in his Arguing For Atheism. Le Poidevin says:

To engage in a religious practice, on this account, is to engage in a game

of make-believe. We make-believe that there is a God, by reciting, in the
context of the game, a statement of belief. We listen to what make-
believedly are accounts of the activities of God and his people, and we
pretend to worship and address prayers to that God…we allow ourselves
to become emotionally involved, to the extent that a religious service is
capable of being an intense experience.43
114 Paul Helm

The difference may be seen in the fact that according to WR God really exists.
Here is a typical passage from Phillips:

The question of the reality of God is a question of the possibility of sense

and nonsense, truth and falsity, in religion. When God’s existence is
construed as a matter of fact, it is taken for granted that the concept of
God is at home within the conceptual framework of the reality of the
physical world. It is as if we said, ‘We know where the assertion of God’s
existence belongs, we understand what kind of assertion it is; all we need
to do is to determine its truth or falsity’. But to ask a question about the
reality of God is to ask a question about a kind of reality, not about the
reality of this or that, in much the same way as asking a question about
the reality of physical objects is not to ask about the reality of this or that
physical object.44

Two things about this. It is clear from such passages that WR asserts the
‘reality’ of God, and if to assert the reality of A is to be realist about A then
WR is realist about God. But, on the other hand, the reality of God is a kind
of reality, distinctive from other kinds; not that the reality of God is necessary
while the reality of other objects is contingent, but that the reality of God is
internal to the form of life of religion: it makes sense to affirm God’s
existence, or to doubt it, only within the language-game of religion. But then
the reality of physical objects is, for WR, on this view, a parallel kind of
reality, bound up with the language-games of common sense, and of sciences
such as physics and chemistry.
It is illuminating to compare the language-game of science and the
language-game of religion as these are understood in WR. For Phillips, the
reality of God is internal to the language-game of religion; the reality of
electrons is internal to the language-game of science; that of tables and chairs
to the language-game of physical objects, and so on. This is a denial of one
form of realism. But the denial of this form of realism in science, say, does
not mean that Phillips advocates or is committed to instrumentalism or
pragmatism in science. For there are truths about science, but these truths are
to be understood—get their meaning—from the language-game of science.
So there are realities, plural. Is this realism or anti-realism?
One test would be this. On the view of RE, every particular thing that
exists either is God or is created by God. The most basic action of all is the
creation of the universe. According to RE, God has created the heavens and
the earth. There is thus one order of mundane reality that owes its existence
to God. Mundane truth is a unity, unified, ultimately, in the one creative
action of God. How does WR handle divine creation? Not in this way, but by
treating creation as an exclusively religious concept which allows no
possibility of intersection or conflict with, say, natural science or metaphysics.
So the sort of realism affirmed by RE and WR, in turn, is different.
Another test would be this. A fundamental feature of Judeo-Christian
Wittgensteinian religion and ‘reformed’ epistemology 115

religious realism is that it is believed that there is action and interaction

between God and the world that he has created. Take petitionary prayer as
an example of this. In petitionary prayer God is entreated for certain things
which (the entreater believes) he may or may not answer. But the
Wittgensteinian view of prayer is very different, for it allows no place for
petitionary prayer, regarding it as ‘superstitious’.45 So for all the talk of reality
of God, God’s reality is constrained at least to this extent, that he cannot
take action independently of ourselves, he cannot respond—or not, as he
chooses—to our prayers, he cannot intervene miraculously, he cannot give us
life after death.

As briefly noted earlier, there is a further sense in which some interpreters of
WR have argued for realism in religion and also at the same time upheld an
externalist view of religion. This involves postulating a close connection
between religion and ethics, and in fact insisting on a particular ethical
interpretation of religion. According to this interpretation, morality, and
particularly religious morality, does not depend upon the way things go, as,
say, consequentialism in ethics does. Morality is severed from consequence
and from teleology. Thus prayer, as I have noted, is a way of reconciling the
one who prays to the contingencies of his life, a way of overcoming life by
submitting appropriately to its vagaries. So the integrity of religion does not
depend upon what happens; to link devotion to God to how things turn out,
to make the one hinge on the other, is a mark of superstition, not of true
religion. This same approach is to be found in Phillips’ treatment of the
problem of evil, according to which the very idea of seeking justification for
evil is regarded as deeply irreligious and corrupt.
On such a strongly ethical view of religion the possession of eternal life,
and the nature of the truly religious life, are ways in which a person may see
his or her life from this non-teleological perspective. God is not a cause of
anything in life, and to invoke him in a truly and consistently religious fashion
is not to seek for an explanation of what is occurring in terms of the divine
purpose. Much less is it to seek a divine intervention in one’s life. Life has no
purpose, and good and evil have no purpose, and the truly religious response
to this state of affairs is to reconcile oneself to this fate by living selflessly,
dying to oneself within this purposelessness. Nothing outside such morality
can recommend it. And a person who lives her life in a way that does not
depend upon how things go for her is at the same time invulnerable. She
cannot be harmed.
The realism that this approach to religion through ethics delivers is pretty
thin by comparison to the robust metaphysical realism of Alvin Plantinga. It
consists largely in an endorsement of an approach to ethics and to religion
which views it from a self-less vantage point, ethics sub specie aeternitatis.
But this is not because there is an eternal vantage point; rather, the one who
116 Paul Helm

adopts such an idea in a regulative fashion will be able to detach ethics from
The externalism of morality here, its independence from any other
consideration than its own character, is at the same time the affirmation of
an intrinsic connection between morality and religion. These claims are
clearly connected with the contextual character of the meaning of religion
discussed earlier.
Quite apart from the problems with this account of meaning, such a view
of morality raises further interesting questions in the context of our overall
discussion. For one thing, there are different senses in which WR’s denial of
the connection between ethics and how things go might be taken. The most
obvious way is to take it as rejecting any possibility of an ethical justification
of the existence of evil—the rejection of any justification of evil by reference
to the good that God will bring out of it, for example. Or it may be taken as
placing a ban on any response to the fact of evil other than the response of a
certain kind of submission to it.
The earlier discussion of defeaters will have made it evident that RE could
not consistently accept this second position. But RE may be consistent with
the first position. Individuals who are prominent in RE, notably Alvin
Plantinga, have been very active in developing responses to the problem, or
problems, of evil. But these are not moral responses. Rather they rest on
logical and metaphysical claims, claims about the nature of human freedom
and about the logical consistency of sets of propositions concerning God and
human evil. Perhaps there is something intrinsic to RE which rules out a
theodicy, as opposed to a defence. Or perhaps theodicies are possible, but not
necessary. But now I am moving out of my chosen territory, a straight
comparison between RE and WR in respect of epistemological and
metaphysical issues, into areas which are not of the esse of RE, and may not
be of its bene esse46

1 Which is not to say that there are not other significant points of comparison,
given that ‘reformed’ epistemology finds its inspiration in an aspect of the
Reformer John Calvin’s references to the sensus divinitatis. Wittgenstein’s own
belief that there is something basically wrong with human beings has striking
points of contact with Calvin’s belief in radical human evil. (See Norman
Malcolm, Wittgenstein: A Religious Point of View?, edited, with a response, by
Peter Winch, Ithaca: Cornell University Press, 1994, pp. 87–8). A still further
area of comparison is between the idea of religious language-games and William
Alston’s idea of doxastic practices as developed in Perceiving God (Ithaca:
Cornell University Press, 1991).
2 I shall take Alvin Plantinga as the spokesman for RE, and confine my remarks to
the earlier, internalist phase of his epistemology, that to be found in ‘Reason and
Belief in God’, for example (in A.Plantinga and N.Wolterstorff (eds), Faith and
Rationality Notre Dame: University of Notre Dame Press, 1983). And besides
Wittgenstein (of course) I shall take D.Z.Phillips and Sir Anthony Kenny (and,
Wittgensteinian religion and ‘reformed’ epistemology 117

to a lesser extent, Norman Malcolm) as the spokesmen for Wittgensteinian

religion. Thus the discussion has the prospect of being focused, if not
comprehensive. The sense in which Wittgenstein himself was a ‘religious’ person
and thinker (which he believed himself to be) is explored by Norman Malcolm
in Wittgenstein: A Religious Point of View? Strangely, the Cambridge
Companion to Wittgenstein (ed. Hans D.Sluga and David G.Stern, Cambridge:
Cambridge University Press, 1996) is almost totally silent on the religious aspects
of Wittgenstein’s thought.
3 Summa Theologiae, 1A 2.2. (p. 66).
4 Plantinga, op. cit., p. 52.
5 Ibid., p. 59.
6 Ibid., p. 60.
7 D.Z.Phillips, Faith After Foundationalism (London: Routledge, 1988). The
critique of RE is to be found chiefly in the first four chapters.
8 Ibid., p. 41.
9 Ibid.
10 Ibid., p. 49.
11 Ibid., p. 40.
12 For this mistake, see Phillips, Faith After Foundationalism, p. 33.
13 Ibid., p. 44.
14 Ibid., p. 47.
15 D.Z.Phillips, Faith and Philosophical Enquiry (London: Routledge & Kegan
Paul, 1970), p. 70.
16 Malcolm, op. cit., pp. 85–6. On the face of it, this looks arbitrary and stipulative.
Why should not reference to the will of God be an explanation if it is believed
that God has purposes which his will expresses? To say, as Malcolm does, that if
everything is explained by reference to the will of God then nothing is explained
looks inadequate. For there may be different senses in which events are due to
the will of God, different purposes to be fulfilled in the willing. But WR reveals
an absolute aversion to becoming involved in such distinctions, reinforcing the
impression that rather than describe religion it is prescribing what genuine
religion must be like—all else being superstition—and giving some plausibility
to the otherwise not altogether fair charge that its view of religion is ‘expressive’.
As Peter Winch observes, to say that reference to God’s will is not an explanation,
and that it makes no sense to ask why something should be the will of God, is to
make claims about a particular feature of a certain kind of religious language-
game, not about religious language-games as such (Winch, in Malcolm, op. cit.,
p. 112).
17 Ludwig Wittgenstein, Zettel, p. 193.
18 ‘Religion’ here is very abstract. There is no such thing as the practice of ‘religion’,
only of religions such as Christianity, or Islam or Buddhism, and then only of
one version of whichever of these is practiced. Perhaps this abstraction is another
signal of the tendency in WR to prescribe what religion must be like, despite the
disclaimer that philosophy ‘leaves everything as it is’, including religion.
Wittgenstein expresses a disinterest in and indeed a disdain for religious
doctrines; for him they are too close to being explanations.
19 William P.Alston, Taking the Curse Off Language-Games: A Realist Account of
Doxastic Practices’, in T.Tessin and M.Von der Ruhr (eds), Philosophy and the
Grammar of Religious Belief (London: Macmillan, 1995), p. 21.
20 Ibid., p. 22.
21 Ibid., pp. 23–4.
22 D.Z.Phillips, Faith and Philosophical Enquiry, p. 2. The italics are in the original.
23 Alston, op. cit., p. 28.
24 Ibid., p. 31.
118 Paul Helm

25 Ibid., p. 32.
26 Ibid., p. 33.
27 Anthony Kenny, What Is Faith? (Oxford: Oxford University Press, 1992), p. 34.
28 Ibid., p. 35.
29 Ibid., p. 22.
30 Quoted by Kenny, op. cit., p. 23.
31 Wittgenstein, On Certainty §143, quoted in D.Z.Phillips, Faith After
Foundationalism, p. 40.
32 Phillips thinks that an interpretation of Wittgenstein’s On Certainty which takes
the basic beliefs to be hypothetical should appeal to RE (Faith After
Foundationalism, p. 58), but there is no evidence of interest on the part of RE in
Wittgenstein on this topic.
33 Faith After Foundationalism, p. 65.
34 Ibid., p. 63.
35 Ibid., p. 59.
36 Wittgenstein, On Certainty §167, quoted by Phillips, Faith After
Foundationalism, p. 64.
37 The relativism of basic propositions to times and places is clearly brought out by
Phillips, Faith After Foundationalism, pp. 59–61.
38 ‘Reformed epistemology’ is a particular approach to religious epistemology.
As such the approach says nothing directly about the nature of God’s reality. But
reformed epistemologists as a matter of fact have views on the reality of God,
and it is these views that I shall seek to identify and comment upon. These views
are shared by those who take other epistemological positions, e.g. classical
natural theology. And for all I know there are those who take a different view of
the nature of the reality of God that are nevertheless committed to a version of
39 Plantinga, op. cit., p. 78.
40 Proceedings of the American Philosophical Association (1992–93), pp. 47–70.
41 De Veritate Q. 1, A. 6 Respondeo (quoted in Plantinga, ‘How to Be an Anti-
Realist’, p. 68).
42 Though problems are surely raised by the fact of religious pluralism. For
Plantinga on pluralism, see ‘A Defence of Religious Exclusivism’, in Thomas D.
Senor (ed.), The Rationality of Belief and the Plurality of Faith (Ithaca, NY:
Cornell University Press, 1995).
43 Robin Le Poidevin, Arguing For Atheism (London: Routledge, 1996), pp. 118–
44 Phillips, Faith and Philosophical Enquiry, p. 3.
45 D.Z.Phillips, The Concept of Prayer (London: Routledge & Kegan Paul, 1965).
46 I am grateful to Peter Byrne for his comments on an earlier draft of this paper.
8 Wittgenstein and the
interpretation of religious
Alan Bailey

In his post-Tractatus writings Wittgenstein vigorously rejects the supposition
that philosophical progress can be achieved through the construction of
intricate chains of deductively valid reasoning that establish that it is necessary
for their conclusions to be true if their premisses are true. Wittgenstein explicitly
asserts in the Philosophical Investigations that ‘philosophy simply puts
everything before us, and neither explains nor deduces anything’ (§126), and
he subsequently goes on to say: ‘In philosophy we do not draw conclusions.
“But it must be like this!” is not a philosophical proposition. Philosophy only
states what everyone admits’ (PI §599). Moreover Wittgenstein is equally
hostile to the suggestion that philosophy can imitate the methods of science
and construct theories about the hidden nature of things on the basis of
arguments to best explanation. In the Blue Book Wittgenstein insists that this
approach to philosophy is inevitably disastrous: ‘Philosophers constantly see
the method of science before their eyes, and are irresistibly tempted to ask and
answer questions in the way science does. This tendency is the real source of
metaphysics, and leads the philosopher into complete darkness’ (BB, p. 18).
And the same attitude continues to manifest itself in the Investigations:
Wittgenstein maintains that ‘we may not advance any kind of theory. There
must be nothing hypothetical in our considerations’ (PI§109).
Wittgenstein is not content, however, with condemning some specific ways
of arriving at philosophical conclusions: he apparently wishes to claim that it
is not the philosopher’s task to advance or defend any novel or surprising
theses! Most people tend to think of philosophers as attempting to provide
us with good reasons for accepting beliefs hitherto regarded as disputable
and not entirely certain. Yet Wittgenstein asserts, as already noted, that
‘philosophy only states what everyone admits’ (PI §599), and he also claims
that ‘if one tried to advance theses in philosophy, it would never be possible
to debate them, because everyone would agree to them’ (PI§128).
If we accept, though, that Wittgenstein is right to deny that the function of
philosophy is to put forward controversial propositions and then show that
they are true, we are immediately confronted by a pressing question about

120 Alan Bailey

what benefit we can hope to derive from philosophical inquiry. If philosophy

does not have the goal of adding to our stock of true beliefs, then what is it
supposed to do for us? Moreover Wittgenstein’s oracular remarks do not
concern themselves exclusively with the aim of philosophy: they appear to
impinge also on matters of philosophical technique. How can a philosophical
work achieve anything if it consists exclusively of propositions that the reader
already accepts as true?
If we are to have any chance of arriving at adequate answers to such
questions, it is essential to recognize that Wittgenstein holds the aim of
philosophy to be the eradication of conceptual confusion. The following
remarks, written in 1929, serve to make it plain that Wittgenstein holds that
if philosophy is undertaken in a correct manner, it concerns itself exclusively
with issues of meaning: ‘I myself still find my way of philosophizing new, &
it keeps striking me so afresh, and that is why I have to repeat myself so
often. …This method consists essentially in leaving the question of truth and
asking about sense instead (CV, p. 3). However Wittgenstein shows no sign
of being interested in the construction of systematic accounts of the meaning
of words as a worthwhile end in itself.1 His inquiries into the nature of our
concepts are directed exclusively towards the goal of eliminating intellectual
muddle and confusion. Thus he says: ‘A philosophical problem has the form:
“I don’t know my way about”’ (PI §123), and at PI §309 we find the
following question and response: ‘What is your aim in philosophy?—To show
the fly the way out of the fly-bottle.’
It is clear, moreover, that if the goal of philosophy is to provide a cure for
‘puzzlement and mental discomfort’ (BB, p. 59), then it becomes much easier
to see how the achievements of philosophy might not manifest themselves in
alterations in people’s opinions. If one succeeds in untying a knot in
someone’s thinking, then the end result is simply the disappearance of the
knot (see Z §452). Nevertheless this is still real progress.
We do not seem, though, to be significantly closer to an understanding of
how a work like the Investigations is supposed to unravel intellectual muddles
without putting forward any arguments on behalf of contentious theses. In
order to make progress on this issue, we need to look at the way in which
Wittgenstein attempts to eliminate our philosophical perplexities.
Wittgenstein holds that the confusions the philosopher needs to eradicate
arise ‘when language is like an engine idling, not when it is doing work’ (PI
§132). When we first attempt to survey our use of language, we are prone to
form badly distorted models of the ways in which certain components of our
language are normally used. However, once a model of this kind has seized
hold of us, we are unable to relinquish it easily. Consequently we continue to
try to apply the model even when it does not fit, and we descend into
bewilderment and muddle of the type that characterizes someone who
drunkenly persists in trying to open a lock with the wrong key.
The remedy Wittgenstein proposes for this pervasive malaise is an accurate
appreciation of the complicated pattern of use associated with the terms or
Wittgenstein and the interpretation of religious discourse 121

expressions generating our difficulties (see PI§§90, 109 and 122; BB, p. 56).
In many cases this can be achieved by persuading us to look afresh at concrete
examples of our linguistic usage in real-life situations. On other occasions
Wittgenstein constructs imaginary languages that serve to throw into sharp
relief features of our language that are usually obscured by the complexities
around them:

I shall in the future again and again draw your attention to what I shall
call language games. These are ways of using signs simpler than those in
which we use the signs of our highly complicated everyday language….
When we look at such simple forms of language the mental mist which
seems to enshroud our ordinary use of language disappears. We see
activities, reactions, which are clear-cut and transparent.
(BB, p. 17)

And sometimes he exposes a mistaken account of the way a particular

fragment of our language works by sketching a possible but non-existent
pattern of use that is accurately described by that account (see PI §6 and BB,
pp. 34–5).
It is clear, therefore, that Wittgenstein holds that the cure for philosophical
perplexity lies in our acquiring an ability to arrive at a fully conscious
understanding of how we do employ the words and sentences that are potential
sources of confusion and intellectual muddle: ‘Philosophy is a battle against
the bewitchment of our intelligence by means of language’ (PI §109). However
in arriving at this understanding, we simply acquire an overview of a pattern
that always lay before our eyes. Consequently Wittgenstein compares the
philosopher’s situation with that of someone trying to complete a jig-saw
puzzle: ‘It seems to us as though we had either the wrong pieces, or not enough
of them, to put together our jig-saw puzzle. But they are all there, only all
mixed; and there is a further analogy between the jig-saw puzzle and our case:
It’s no use trying to apply force in fitting pieces together. All we should do is to
look at them carefully and arrange them’ (BB, p. 46). But once we have traced
a philosophical problem back to its roots in a linguistic misunderstanding and
we have come to appreciate how that particular fragment of our language
actually functions, the problem in question disappears completely (PI §133).

What, then, can we expect of a philosophical treatment of religious belief
that respects the methodological principles espoused by Wittgenstein in his
later writings? It would, of course, be concerned with trying to understand
the meaning of the statements characteristically made by religious believers.
In A Treatise Concerning the Principles of Human Knowledge, Berkeley
unhesitatingly interprets statements about God as statements about a causally
efficacious being whose existence is wholly independent of human desires,
122 Alan Bailey

beliefs and aspirations.2 Similarly, we find that although the discussion in

Hume’s Dialogues Concerning Natural Religion is presented as being a
discussion of the nature of God,3 the three principal protagonists always
speak of God as an independently existing being with extensive causal
powers. Thus Wittgenstein would maintain that both Berkeley and Hume
are failing to address themselves to the questions that are properly of concern
to the philosopher. Religious believers do indeed talk about God and profess
to believe in God; but it is not the function of the philosopher to assume that
such discourse must be interpreted as discourse about a causally efficacious
being and then put forward reasons for and against the supposition that this
being exists. Instead the philosopher’s task is to investigate without prejudice
the way statements about God are used by religious believers.
Unfortunately Wittgenstein’s own writings fail to provide us with the
detailed account of the usage of religious statements that his philosophical
methodology prescribes. Some of the remarks collected in Culture and Value
do draw our attention to interesting aspects of the employment of religious
claims and affirmations. On the whole, however, the remarks on religion do
little more than assist us in identifying the interpretation of religious discourse
favoured by Wittgenstein: they do not give us any substantial grounds for
accepting that this interpretation adequately represents the content of
religious statements as these are used by most people who think of themselves
as religious believers. A similar conclusion seems appropriate also in the case
of the material that can be found in Remarks on Frazer’s Golden Bough.
Wittgenstein’s discussion of Frazer’s views does help us to realize that Frazer
fails to appreciate the extent to which language and ritual have an expressive
function; and there are clear indications that Wittgenstein is inclined to
believe that the recognition of the expressive role of language holds the key
to an adequate understanding of religious discourse (see, for example, RF,
pp. 7 and 8). Nevertheless it cannot plausibly be maintained that the
examples Wittgenstein assembles in this brief work are sufficient to determine
how we should interpret the statements made by self-professed religious
believers. Only in the lectures on religious belief contained in Lectures and
Conversations on Aesthetics, Psychology and Religious Belief do we
encounter any sustained and reasonably orderly exploration of the way
religious statements are used. And even here it is conspicuous that the aspects
of this usage that seem to support the kind of interpretation accepted by
Berkeley and Hume receive almost no serious discussion.
In the Lectures and Conversations Wittgenstein places great emphasis on
the apparent disparity between the strength of religious beliefs and the
cogency of the evidence that can be offered in support of these beliefs.
Wittgenstein says that religious believers ‘base things on evidence which taken
in one way would seem exceedingly flimsy’ (LC, pp. 57–8). Nevertheless
Wittgenstein tries to persuade us that it would be inappropriate to respond
to this phenomenon by classifying religious believers as unreasonable in the
way that we might characterize a scientist as unreasonable if he or she were
Wittgenstein and the interpretation of religious discourse 123

to embrace a scientific theory on the basis of patently inadequate evidence.

Wittgenstein maintains that whereas there is no difficulty about saying that
religious believers are certainly not reasonable, the accusation that they are
behaving unreasonably would be understood by everyone as a rebuke, yet
many religious believers are entirely sanguine and unconcerned about the
seemingly flimsy nature of the grounds for their beliefs:

I want to say: they don’t treat this as a matter of reasonability.

Anyone who reads the Epistles will find it said: not only that it is not
reasonable, but that it is folly.
Not only is it not reasonable, but it doesn’t pretend to be.
(LC, p. 58)

When we are attempting to interpret someone else’s statements, it is often a

sound procedure to eschew a particular interpretation if that interpretation
involves ascribing irrational beliefs to the person in question. Furthermore,
the desirability of avoiding such ascriptions is increased if that person usually
manifests a good grasp on issues of justification but remains willing to make
these problematic statements even when the grounds for our concerns about
lack of justification have been explicitly drawn to his or her attention. And it
is very striking that in the Remarks on Frazer’s Golden Bough, Wittgenstein
repeatedly appeals to these interpretative principles in order to reject Frazer’s
explanations of the significance and origins of particular rituals.
Wittgenstein notes, for example, that Frazer writes of a rain-king in Africa
who supposedly receives gifts from people in order to secure his help in
making it rain when the vernal rains are due. However Wittgenstein argues
that Frazer misinterprets these gifts:

But surely this means that they do not actually think he can make rain,
otherwise they would do it in the dry periods in which the land is ‘a parched
and arid desert’. For if we do assume that it was stupidity that once led the
people to institute this office of Rain King, still they obviously knew from
experience that the rains begin in March, and it would have been the Rain
King’s duty to perform in other periods of the year.
(RF, p. 12)

Similarly Wittgenstein firmly rejects Frazer’s explanation of rituals in which

a priest-king is ultimately put to death by his own subjects:

All that Frazer does is to make this practice plausible to people who think
as he does. It is very queer that all these practices are finally presented, so
to speak, as stupid actions.
But it never does become plausible that people do all this out of sheer
(RF, p. 1)
124 Alan Bailey

Moreover, Wittgenstein neatly draws our attention to the fact that the very
people who supposedly embrace these outlandish causal hypotheses
invariably show themselves to be competent reasoners in other areas of their
lives: The same savage who, apparently in order to kill his enemy, sticks his
knife through a picture of him, really does build his hut of wood and cuts his
arrow with skill and not in effigy’ (RF, p. 4).
The conclusion Wittgenstein draws from his examination of the rituals
described by Frazer is that they are not the product of beliefs about the causal
mechanisms at work in the world. Instead they are direct expressions of the
participants’ desires and emotions: ‘Burning in effigy. Kissing the picture of a
loved one. This is obviously not based on a belief that it will have a definite
effect on the object which the picture represents. It aims at some satisfaction
and it achieves it. Or rather it does not aim at anything; we act in this way
and then feel satisfied’ (RF, p. 4). It is surely plausible to suppose, therefore,
that the foregoing pattern of thought will recur in Wittgenstein’s lectures on
religion. As we have seen, Wittgenstein holds that if we take the statements
made by religious believers about the existence and nature of God as
indicating that they believe that God is a causally efficacious being who exists
independently of our beliefs and desires, then the beliefs we are ascribing to
these people are unreasonable beliefs because no substantial evidence can be
assembled in favour of the hypothesis that such a being exists. Thus the
situation here parallels the one generated by Frazer’s attempts to explain
magical and religious rituals as the product of mistaken causal hypotheses.
Frazer’s stance commits him, in Wittgenstein’s judgment at least, to the
conclusion that the participants in these rituals are acting on the basis of
unreasonable beliefs. Wittgenstein, however, is very keen to keep to a
minimum such attributions of irrationality, and he consequently offers an
alternative account of the rituals in question that rejects the assumption that
they are based on tendentious causal beliefs in favour of the suggestion that
they have an expressive role. If we assume, then, that Wittgenstein is equally
keen to avoid the need to condemn religious believers as holding a mass of
unreasonable beliefs, it seems likely that he will offer an interpretation of
statements about God that gives them an expressive function and rejects the
assumption that ‘God’ is supposed to be the name of a causally efficacious
being. Moreover Wittgenstein’s comments in both the Lectures and
Conversations and the Remarks on Frazer’s Golden Bough seem to show that
this is indeed the kind of interpretation he favours.
In one passage Wittgenstein specifically discusses the application of
Frazer’s preferred way of explaining magical and religious rituals to the
appeals to God scattered throughout St Augustine’s spiritual autobiography:

Frazer’s account of the magical and religious notions of men is

unsatisfactory: it makes these notions appear as mistakes.
Was Augustine mistaken, then, when he called on God on every page
of the Confessions?
Wittgenstein and the interpretation of religious discourse 125

Well—one might say—if he was not mistaken, then the Buddhist

holyman, or some other, whose religion expresses quite different notions,
surely was. But none of them was making a mistake except where he was
putting forward a theory.
(RF, p.1)

And at another point in the Remarks on Frazer’s Golden Bough, Wittgenstein

compares the religious activities of a sacrificial priest-king to more recent
devotional practices: The religious actions or the religious life of the priest-
king are not different in kind from any genuinely religious action today, say
a confession of sins. This also can be “explained” (made clear) and cannot be
explained’ (RF, p. 4).
It appears, therefore, that when Wittgenstein was writing the Remarks on
Frazer’s Golden Bough in 1931, he was already beginning to develop an
expressive account of religious discourse. And Wittgenstein’s commitment to
such an interpretation of the statements made by religious believers continues
to manifest itself in the Lectures and Conversations. Wittgenstein explicitly
rejects the supposition that religious statements and scientific statements
function in the same way:

In a religious discourse we use such expressions as: ‘I believe that so and

so will happen,’ and use them differently to the way in which we use
them in science.
Although, there is a great temptation to think we do. Because we do
talk of evidence, and do talk of evidence by experience.
(LC, p. 57)

Moreover the positive characterization he provides of statements about a

Judgment Day places the emphasis firmly on the attitude of the person who
finds it natural to talk in such terms:

Here believing obviously plays much more this role: suppose we said that
a certain picture might play the role of constantly admonishing me, or I
always think of it. Here, an enormous difference would be between those
people for whom the picture is constantly in the foreground, and the
others who just didn’t use it at all.
(LC, p. 56)

Although the way in which religious believers persist in affirming the

existence of God, even though evidence for the existence of a causally
efficacious being of the requisite power and goodness appears to be
remarkably sparse, is undoubtedly one of the main features of the use of
religious statements that persuades Wittgenstein to interpret these statements
expressively, Wittgenstein does draw attention in the Lectures and
Conversations to some other aspects of their usage that might be regarded as
126 Alan Bailey

adding to the plausibility of an expressive interpretation. Wittgenstein claims,

for instance, that denials of God’s existence tend to attract moral
condemnation whereas the existence of anything or anyone else can be denied
without attracting such criticism:

If the question arises as to the existence of a god or God, it plays an

entirely different role to that of the existence of any person or object I
ever heard of. One said, had to say, that one believed in the existence,
and if one did not believe, this was regarded as something bad. Normally
if I did not believe in the existence of something no one would think
there was anything wrong in this.
(LC, p. 59)

Wittgenstein also maintains that certain common responses to ordinary

statements of fact are plainly inappropriate or carry an entirely different
significance when directed towards religious statements. He notes that there
is something very odd about saying that these statements indicate the
opinions or views of the people who make them: only the word ‘beliefs’ seems
to fit here. And whereas we would normally judge that someone who
responds to a statement by saying that it is possibly true is thinking along
similar lines to the person making that statement, we would appraise the
situation differently if someone made such a response to a religious statement:
‘Suppose someone were a believer and said: “I believe in a Last Judgement,”
and I said: “Well, I’m not so sure. Possibly.” You would say that there is an
enormous gulf between us. If he said “There is a German aeroplane
overhead,” and I said “Possibly. I’m not so sure,” you’d say we were fairly
near’ (LC, p. 53). It does seem, however, that it would be relatively easy to
account for the foregoing features of the usage of religious statements even if
one continued to interpret references to God as references to a causally
efficacious being.
Wittgenstein himself notes that the word ‘God’ is ‘used like a word
representing a person’ (LC, p. 59). Let us assume for the moment, then, that
‘God’ is the name of a causally efficacious person. All the theistic religions
agree that ‘God’ is not used as the name of an embodied person. So if we are
to persevere with the assumption that ‘God’ is the name of a causally
efficacious person, we will have to say that ‘God’ is the name of a causally
efficacious person who does not have a body. But any adequate
characterization of God as a causally efficacious being must also
acknowledge that God is supposed to be a fit object of human obedience and
worship. Worship, though, is at least in part a matter of respectfully
acknowledging excellencies and being properly grateful for benefits freely
conferred.4 Consequently if ‘God’ is the name of a causally efficacious person
who is a fit object of worship, then ‘God’ is the name of a causally efficacious
person of surpassing power and goodness who has freely chosen to create
and further the welfare of all human beings.
Wittgenstein and the interpretation of religious discourse 127

It is clear, accordingly, that if ‘God’ is the name of such a being and the
statement ‘God exists’ is true, then this is a fact about the world that is of
overwhelming importance to everyone. Let us consider, therefore, a
nonreligious statement about something that would be regarded by most
people as a matter of considerable importance. If A says to B ‘There is a sack
buried a few inches below the surface of your lawn that contains £20 million
in untraceable banknotes’ and B’s only reaction is to reply ‘Possibly. I’m not
so sure myself, it seems obvious that B actually believes that A’s statement
has no chance at all of being true. Thus someone who holds that statements
about the existence of God or the coming of a Last Judgment are statements
about a causally efficacious person or an event with extensive causal
implications can readily argue that it is the potential importance of the claims
made by these statements that accounts for the way in which a response like
‘Well, possibly’ constitutes a dismissive rejection of such statements rather
than an acknowledgement that there might be something of substance to be
said on their behalf.
The observation that any causally efficacious being capable of being
appropriately called ‘God’ must be worthy of obedience and worship also
allows someone who does not embrace Wittgenstein’s expressive
interpretation of religious discourse to provide a plausible explanation of the
fact that denials of God’s existence tend to attract moral condemnation.
Religions like Christianity and Islam portray God as the rightful ruler over
all human beings. Thus someone who denies that God exists is, in part,
rejecting a claim about external authority. Such a person acknowledges no
obligation to obey any laws supposedly originating from God’s decrees. Yet
many human institutions attempt to legitimize their coercive powers by
asserting that God has directly delegated some of his authority to them or
claiming that they have some peculiarly reliable insight into God’s will.
Consequently these institutions are strongly inclined to regard denials of
God’s existence as a threat to their own position and status within society,
and they naturally respond to this perceived threat with moral excoriation
and warnings about the imminent collapse of the social order.
It seems, therefore, that we are forced to conclude that in the Lectures and
Conversations the full burden of supporting Wittgenstein’s expressive
interpretation of religious statements ultimately falls on the considerations he
brings forward about the desirability of adopting an interpretation that allows
us to avoid ascribing unreasonable beliefs to religious believers. Before we turn
to an examination of the intrinsic strength of that line of thought, however, we
do need to survey briefly some of the features of the usage of statements about
God that actually appear to offer strong support for the supposition that ‘God’
is intended to function as the name of a causally efficacious being.
The religion at the centre of Wittgenstein’s attention is Christianity, and the
following supplementary examples of the way religious language is used are
taken from the same source. However Christianity is sufficiently similar to
Islam and Judaism to make it grossly implausible to suppose that the sense of
128 Alan Bailey

the word ‘God’ as used by Christians is radically different from the sense it has
in these other monotheistic religions. Thus any evidence that Christians use
‘God’ primarily as the name of a causally efficacious being is automatically
evidence that this is how the word ‘God’ is employed in Islam and Judaism.
One important aspect of the way many Christians talk about God is their
willingness to construct or deploy in argument with non-believers alleged
proofs of God’s existence. In the case of St Anselm’s ontological argument,
the explicit aim is to persuade us to move from the premiss that we can form
a conception of a being greater than which none can be conceived to the
conclusion that such a being exists outside our minds.5 And other proffered
proofs argue for the existence of a specific kind of causal agent, and then
identify God with that agent. Bishop Berkeley, for instance, argues that a
mind of superhuman power must be causally responsible for the orderly
succession of perceptions that constitutes, in his opinion, the external world.6
William Paley, on the other hand, dispenses with Berkeley’s idealism but
argues that the adaptation of means to ends that we find in the natural world
provides us with an inference to the conclusion ‘A superhuman designer
exists’ that is as secure as the inference from the discovery of a watch to the
conclusion that it is the product of human contrivance.7 And St Thomas
Aquinas concludes the first of his ‘Five Ways’ as follows:

Of necessity therefore anything in process of change is being changed by

something else. Moreover, this something else, if in process of change, is
itself being changed by yet another thing; and this last by another. Now
we must stop somewhere, otherwise there will be no first cause of the
change, and, as a result, no subsequent causes…. Hence one is bound to
arrive at some first cause of change not itself being changed by anything,
and this is what everybody understands by God.8

It appears, then, that an expressive account of the meaning of statements

about the existence of the God of Christianity commits us to the view that
the bishop, archdeacon and saints mentioned in the preceding paragraph have
failed to understand correctly the status of their own religion. Now it is, of
course, true that people rarely acquire a religious faith as a result of
dispassionate reflection on arguments of the kind under discussion here.
Nevertheless the fact that Berkeley, Paley, Anselm and Aquinas are all content
to hold that the word ‘God’ is used in the conclusion of their arguments in
the same sense as it is used when they are engaged in prayer and worship is
strong evidence that even in these latter circumstances they are using ‘God’
as the name of a causally efficacious being. It is also important to bear in
mind that although other Christians encountering these arguments have often
accused them of being fallacious or useless as a means of bringing about a
genuine religious conversion, the allegation that they misuse the word ‘God’
is a relatively rare response and one that has attained a significant degree of
prominence only in the twentieth century.
Wittgenstein and the interpretation of religious discourse 129

Another interesting use of the word ‘God’ by some Christians occurs when
they are trying to justify their opinions about matters of morality. Many
Christians who are pacifists attempt to vindicate their position by appealing
to the Ten Commandments. In the King James’ Bible (AV) one of these
commandments is translated as ‘Thou shalt not kill’, and the relevant verse
(Exodus 20. 13) is often cited as proscribing any participation in armed
conflict. In fact this verse should be translated as ‘Do not commit murder’,
but the significant point for our purposes is that people appeal to it as an
authoritative pronouncement that supports their own attitude to war. If an
expressive interpretation of statements about God is correct, however, then
an appeal to God’s commandments to justify one’s moral stance fails to do
any constructive work. Instead of offering an independent reason for one’s
stance, one is simply reaffirming that stance in different words.
It is also instructive here to consider the reports in the Gospels of Jesus’
miracles. These events are presented by the Gospel writers as one of the
principal signs of Jesus’ special relationship with God, and large numbers of
Christians, even today, would maintain that these events took place exactly
as described in the New Testament. However they would also insist that no
physician or any other person relying on his or her own powers could have
healed people in the way Jesus did. Consequently when they say that Jesus’
power to heal was given to him by God, they cannot be speaking
metaphorically or expressively. In the situation postulated by these Christians
there is a real shortfall of power that can only be overcome through the
intervention of a causally efficacious divine being.
According to Wittgenstein, we discover what is meant by people who
speak and write of God by looking at the way they use the word ‘God’: ‘The
way you use the word “God” does not show whom you mean, but what you
mean’ (CV, p. 58); ‘Grammar tells us what kind of object anything is.
(Theology as grammar.)’ (PI §373).
So far, however, our examination of the way ‘God’ is employed suggests
that the overwhelming majority of people who think of themselves as
religious believers intend it to function as the name of a causally efficacious
person who has no physical body. If this account of the meaning of ‘God’ is
to be overturned, the only implement that seems to lie to hand is the
contention that it would be inappropriate to adopt an interpretation of
statements about God that compels us to ascribe large numbers of
unreasonable beliefs to religious believers. Thus it is now time to undertake
an evaluation of the force of that line of argument.
There clearly are circumstances in which we would unhesitatingly
conclude that we needed to change our interpretation of words and sentences
so as to avoid ascribing unreasonable beliefs to people. Suppose, for example,
we have provisionally decided to translate the word ‘gip’ using the English
word ‘pig’. If we later find that we have good grounds to hold that the people
who use the word ‘gip’ believe that gips are common visitors to their gardens,
have feathers, and can fly, we would undoubtedly decide to revise our
130 Alan Bailey

translation. However there also appear to be occasions when we are

significantly less willing to revise our interpretations merely in order to
minimize the extent to which we see other people as having unreasonable
beliefs. Evidence from surveys indicates that the majority of people in both
Britain and the United States believe that there is some truth in astrology, and
three-quarters of the adults living in the Western world profess to accept at
least some psychic phenomena as genuine.9 However the supposition that
real psychic phenomena exist contravenes well-established physical laws; and
although it is possible that these laws might ultimately be overturned, it is
surely more reasonable, given the evidence currently available to us, to place
our confidence in those laws rather than poorly substantiated reports of
psychic phenomena. In the case of astrology, moreover, the predictions
yielded by this supposed art are sufficiently precise to allow for a significant
degree of empirical testing. Where this has been done, the evidence
overwhelmingly indicates that the position of the constellations at the time
of a person’s birth makes no discernible difference to that person’s personality
or the pattern of events in his or her life.10
If we are prepared to accept, however, that large numbers of people have
an unreasonable belief in astrology or the existence of psychic phenomena,
then why should we baulk at accepting that large numbers of people, namely
religious believers, have an unreasonable belief in the existence of a causally
efficacious being with no body? And this becomes a particularly pertinent
question when we consider the explanations psychologists offer of the origins
of belief in the existence of psychic phenomena.
Psychologists point out that people have a marked inclination to seek
explanations for the events that occur around them. This inclination
obviously plays an important part in ensuring the survival of the human
species. However it is so powerful a drive that where no genuine explanation
can be found psychic powers or phenomena are often invoked to fill the gap.
Furthermore there is strong empirical evidence that the tendency of human
infants to interpret all striking events as the products of some agent’s will
often reasserts itself even in adults when they do not have a firm grasp on the
relevant causal mechanisms. Consequently people find it difficult to refrain
from ascribing paranormal powers to otherwise ordinary agents and
postulating extraordinary agents who exist but cannot be observed.11
Let us consider, then, the likely outcome if someone begins to reflect on
the existence of the Earth and all the animals and plants that live upon it.
Any explanation of the existence of these things that is even approximately
the right explanation requires some understanding of physics, the nature of
the stars and planets and evolutionary processes. For much of human history,
then, the reasonable response to questions about how the Earth and living
beings came into existence would have been a refusal to commit oneself to
any definite opinion. However, in the light of the first of the psychological
mechanisms outlined in the preceding paragraph, we can expect that few
people would have been able to content themselves with suspending judgment
Wittgenstein and the interpretation of religious discourse 131

on this matter: in most instances some explanation or other would have been
intemperately embraced, despite its lack of rational credentials. Moreover,
the second psychological mechanism, utilized above to account for belief in
paranormal phenomena, is likely to ensure that the most popular explanations
would have presented the existence of the Earth and living beings as the result
of the will of some exceptionally powerful agent. Consequently the obvious
limitations of human beings would have led to a belief in the existence of an
intentional agent who is far more powerful than any mere human but who
cannot be observed by means of our physical senses.
It seems, therefore, that the same psychological mechanisms used to
explain present-day belief in the existence of psychic phenomena would also
readily account for the generation in earlier times of widespread belief in the
existence of a causally efficacious but incorporeal being capable of creating
the Earth and all its inhabitants. Moreover, once a belief of this kind was
well-entrenched, the power of education and indoctrination would amply
explain its ability to survive in a less favourable intellectual climate.
At this point, however, it might be suggested that the true force of
Wittgenstein’s objection to ascribing unreasonable beliefs to religious
believers emerges only when we recognize that ascriptions of belief do not
make sense unless we are dealing with what Dennett calls an intentional
system. Dennett holds that it is only in the context of the assumption that a
particular organism has the beliefs ‘it ought to have, given its perceptual
capacities, its epistemic needs, and its biography’,12 that the supposition that
this organism possesses beliefs and other intentional states has any meaning.
And Dennett concludes that if the attribution of the belief that p to some
person would force us to construe that person as unreasonably believing that
p, then we are not entitled to regard that person as holding the belief that p.13
It is clear that if Dennett is correct in his account of the constraints on
meaningful ascriptions of beliefs, we would be mistaken if we interpreted
religious discourse as a manifestation of unreasonable beliefs about the
existence of a causally efficacious divine being. However, Dennett’s position
also offers us some new alternatives to an expressive interpretation of the
statements of religious believers. Dennett relativizes rationality to a person’s
biography: consequently it seems that it would be possible to say that belief
in the existence of a causally efficacious divine being is unreasonable for any
well-informed person who has studied the supposed evidence carefully while
simultaneously conceding that some people might be sufficiently ignorant to
ensure that such a belief is not unreasonable for them. Even more importantly,
Dennett imposes such strong constraints on the concept of belief that it
becomes plausible to hold that if he has accurately specified what is required
of a mental state in order for it to constitute a belief, then religious discourse
might simply be the product of a state of mental confusion that falls well
short of belief.
Another problem with using Dennett’s line of argument to buttress
Wittgenstein’s contention that we cannot appropriately explain religious
132 Alan Bailey

discourse in terms of unreasonable beliefs is that this line of argument delivers

the conclusion that religious believers do not have unreasonable beliefs only
at the cost of also committing us to the conclusion that those people who
seem to believe in the existence of psychic phenomena or the value of
astrology do not, despite appearances, possess any unreasonable beliefs. In
this latter instance, however, the suggestion that we can explain the
statements and actions of these ‘believers’ as expressive in nature is singularly
implausible. We would need to avail ourselves of some other option; and if
an explanation in terms of mental confusion or woeful ignorance about
matters of fact is applicable here, it seems that it would also be a plausible
way of explaining the statements and actions of religious believers.
There are, in any case, rival accounts of what is involved in meaningful
ascriptions of belief. At first sight Dennett’s position appears to be very
similar to the one defended by Davidson in ‘On the Very Idea of a Conceptual
Scheme’.14 Davidson maintains that if we want to understand other people,
‘we must count them right in most matters’.15 And if we are committed to
viewing them as right in most of their beliefs, we must also view most of
these beliefs as reasonable beliefs. However, the crucial difference between
Dennett and Davidson is that Davidson accepts that we can be justified in
ascribing irrational beliefs to another person as long as the overwhelming
majority of the beliefs we ascribe are rational beliefs.16 In some circumstances
there can be a trade-off between maximising rationality and maximising the
elegance and simplicity of our theory concerning the sense of a person’s verbal
performances: we cannot legitimately interpret those performances so that
the person in question emerges as radically irrational, but we are sometimes
entitled to purchase a major improvement in the simplicity of our semantic
theory at the cost of committing ourselves to ascribing a small number of
unreasonable beliefs to that person.
It follows that Davidson’s view of the constraints on meaningful
ascriptions of belief allows us to interpret religious believers as having
unreasonable beliefs about the existence of a causally efficacious being as
long as we are also in a position to interpret them as having predominantly
reasonable beliefs on most other topics. However this requirement is surely
easily met. As Davidson points out, we tend to concentrate our attention on
areas of dispute and controversy: those opinions that everyone shares are too
familiar and dull to attract our interest.17 And even if it is true that religious
believers have some striking and unreasonable beliefs about the existence of
a causally efficacious being called ‘God’, they also share with everyone else a
massive body of beliefs about other people and the objects and causal
mechanisms to be found in the physical world.
A still more radical departure from the interpretative constraints proposed
by Dennett is provided by Stich’s suggestion that the meaningful application
of intentional descriptions to an organism simply requires us to treat that
organism as forming its beliefs in similar ways to the ones we ourselves
employ.18 On this view of the matter it remains essential that we should try to
Wittgenstein and the interpretation of religious discourse 133

minimize the ascription of inexplicable beliefs. However, if an unreasonable

belief can be explained as the product of an aberrant inferential heuristic or
a cognitive deficiency analogous to one that sometimes manifests itself in our
own reasoning, then it can be ascribed to another person without generating
any threat to the legitimacy of regarding that person as an intentional system.
According to Stich: ‘Inferential errors that we can imagine ourselves making
…can be described comfortably in intentional terms. It is only the sort of
error or incoherence that we cannot imagine falling into ourselves that
undermines intentional description.’19
Obviously Stich’s account of the constraints on the meaningful
employment of intentional descriptions leaves even more room than
Davidson’s for legitimate ascriptions of unreasonable belief. As long as we
can see a person’s unreasonable beliefs as arising from cognitive failings
similar to those that sometimes afflict us, there are no grounds at all for
holding that anyone who is appropriately treated as constituting an
intentional system must have more reasonable beliefs than unreasonable
beliefs. Thus if we have been successful in showing how belief in the existence
of a causally efficacious divine being could arise from psychological
mechanisms that tend to exert some influence on everyone’s inferences—
namely the impulse to commit oneself to some kind of explanation even when
firm evidence is lacking and the bias towards explanations in terms of an
agent’s desires and choices—then we could acknowledge that belief in a being
of this kind would be unreasonable without thereby finding ourselves under
any pressure to avoid ascribing such a belief to religious believers.

This examination of Wittgenstein’s extant writings on religion has amply
confirmed the expectation that they would manifest a preoccupation with
questions about the meaning of religious discourse. However it is important
to remember that we do not have access to any work on religion that
Wittgenstein intended for publication.
In the Investigations, for example, we are presented with a carefully
crafted study of the way we talk about minds and mental phenomena, and it
is clear that Wittgenstein has gone to great trouble to ensure that his remarks
are consistent with the method he recommends for eliminating philosophical
problems. Thus the contrast with Culture and Value and the Lectures and
Conversations is striking. In the case of Culture and Value, we simply have a
set of notes written by Wittgenstein for his own benefit. There is, accordingly,
no guarantee that any particular remark would have been endorsed by
Wittgenstein after considered reflection. And when we turn to the Lectures
and Conversations, we are confronted by a redaction of notes compiled by
students listening to Wittgenstein lecture. Thus there is a danger that these
students have set down what they thought Wittgenstein meant rather than
what he said; and it is, in any case, essential to bear in mind that these lectures
134 Alan Bailey

were simply an initial attempt to put Wittgenstein’s general philosophical

methodology to work on topics conventionally subsumed under the rubric of
philosophy of religion. If Wittgenstein had personally reworked his lectures
for publication, it is quite possible that the end result would have looked very
It seems, therefore, that there is scope to conduct an investigation into the
use of religious language that conforms more closely to Wittgenstein’s official
methodology than anything he himself managed to write. Wittgenstein’s own
works after the Tractatus appear to establish that he holds religious discourse,
even when it adopts the surface form of reports on the existence and
properties of causally efficacious supernatural entities and events, to be
essentially expressive in character. In particular, it seems clear that he wishes
to maintain that the religious believer who announces ‘God exists and is my
creator’ or ‘I know that my redeemer liveth’ is not thereby committed to a
belief in the existence of a causally efficacious divine being. Unfortunately
the examples of the use of religious discourse that Wittgenstein actually
brings to our attention are unlikely to persuade many people that his
interpretation of religious statements possesses much plausibility. Moreover
when we direct our attention to aspects of the use of religious discourse that
Wittgenstein virtually ignores, we find a pattern of use that appears to cry
out for explanation in terms of the supposition that religious believers are
committed to a belief in the existence of causally efficacious beings and events
that lie outside the natural order of things. It is, for example, easy to be
impressed by the way in which an injunction like ‘Ye cannot serve God and
Mammon’ can be seen as expressing a moral commitment; but, as I have
emphasized above, the person who quotes this injunction with approval
during a religious service is also in most cases someone who finds nothing
amiss about way the word ‘God’ is used in such contexts as the ontological
argument and the argument to design.
If it is true, however, that a balanced survey of the way in which religious
language is used yields the conclusion that Wittgenstein’s own investigations
have led him to a misleading account of the meaning of religious statements,
then we obviously need to ask why Wittgenstein’s methodology has failed to
repeat in this field the success that it seems to have achieved when applied to
discourse about language and mental phenomena. Part of the answer
undoubtedly lies in the fact that Wittgenstein’s examination of religious
language was not as intensive as was his examination of these other areas of
discourse. It is also true that Wittgenstein was not someone who had what
one might term ‘an insider’s grasp’ on religious discourse. Not only did he
have little inclination to make religious affirmations himself, but he was equally
disinclined to engage in attempts to controvert such statements. However,
the chief reason for Wittgenstein’s relative lack of success in the case of religious
language may well lie in the fact that his philosophical methodology is directed
primarily towards the dissolution of philosophical problems rather than the
disinterested clarification of meaning for its own sake.
Wittgenstein and the interpretation of religious discourse 135

The problems Wittgenstein hopes to eliminate are those that manifest

themselves when we move away from putting language to use in real-life
circumstances. We normally have no difficulty, for example, in telling other
people what time it is or how much time a particular activity is likely to take.
However when we stand back from such mundane activities and ask
ourselves ‘What is time?’, we are suddenly plunged into confusion. In the
case of religious discourse, though, this phenomenon is almost unknown. If
someone who is at ease using the word ‘God’ in prayer and catechisms asks
‘What is God?’, that person rarely has any difficulty in arriving at an answer
with which he or she is fairly comfortable. It is also noteworthy that
philosophical arguments for the conclusion that time is unreal or that we do
not know that time exists do not disrupt our ordinary practices of consulting
watches, making appointments, and allocating ourselves a certain amount of
time to complete particular tasks. However the person who maintains that
God does not exist or that God has no interest in the activities of human
beings normally refrains from prayer and worship unless coerced into taking
part in these activities.
It seems, therefore, that the specific form of intellectual confusion that
Wittgenstein’s therapeutic techniques are intended to eliminate is not
widespread in our thought about matters of religion. However, Wittgenstein’s
descriptions of the way we use particular words and sentences are guided by
these therapeutic considerations. If we are in the kind of confusion generated
by questions like ‘What is time?’ and ‘How can we think about things that do
not exist?’, then Wittgenstein is committed to eradicating this confusion by
giving us an overview of the way we actually put various fragments of our
language to use. Consequently the persistence of the intellectual confusion
provides all the confirmation necessary in order for Wittgenstein to recognize
that he has either misdescribed some aspect of our usage or has failed to
provide a sufficiently extensive survey of that usage. But when there are no
characteristically philosophical perplexities to be dissolved and one is merely
trying to decide between competing interpretations of a particular set of
statements, one can make a choice based on a inadequate survey of the
linguistic evidence without automatically receiving any warning that crucial
aspects of usage remain unconsidered or have not been accorded their due
Ironically, then, the way of understanding religious statements which
Wittgenstein condemns in the Lectures and Conversations as superstition
(LC, p. 59) appears to be the way in which the overwhelming majority of
religious believers intend their religious affirmations to be understood.
Wittgenstein offers an interpretation of religious discourse that is compatible
with many of the features of the way religious discourse is used. Moreover, it
is quite possible that in the case of a few people who would choose to describe
themselves as religious believers Wittgenstein’s interpretation of the content
of their statements is entirely accurate. However, if we are concerned to
understand the phenomenon of religious faith as this manifests itself amongst
136 Alan Bailey

most adherents of such religions as Christianity and Islam, then it would be

grossly misleading to maintain that this faith does not contain as an
important component a belief in the existence of a causally efficacious divine
being. And this conclusion also enables us to insist that such faith is
susceptible to external criticism. Any causal claim can be assessed as
reasonable or unreasonable in the light of the principles of causal reasoning
we all employ in our everyday lives. Wittgenstein may wish to delimit the
boundaries of philosophy in such a way that such assessments do not fall
within the special jurisdiction of philosophers, but any person who is capable
of making competent judgements about the humdrum causal mechanisms
that operate in his or her immediate environment is potentially someone who
can arrive at an appropriate determination of the rationality or irrationality
of the kind of religious faith embraced by most religious believers.

1 See M.McGinn, Wittgenstein and the Philosophical Investigations (London:
Routledge, 1997), pp. 27–8.
2 G.Berkeley, A Treatise Concerning the Principles of Human Knowledge, in
Philosophical Works, ed. M.R.Ayers (London: Dent, 1975).
3 See D.Hume, Dialogues Concerning Natural Religion, ed. N.Kemp Smith, 2nd
edn. (London: Thomas Nelson, 1947), pp. 141–2.
4 See R.Swinburne, The Coherence of Theism (Oxford: Oxford University Press,
1977), pp. 283–8.
5 St Anselm, Proslogion, in Anselm of Canterbury, ed. J.Hopkins and
H.Richardson, vol. 1 (London: SCM Press, 1974), pp. 93–5.
6 Berkeley, op. cit., p. 85.
7 W.Paley, Natural Theology, in The Works of William Paley, Archdeacon of
Carlisle (Edinburgh: Peter Brown and T. & W.Nelson, 1825), pp. 435–9.
8 T.Aquinas, Summae Theologiae, vol. 2 (London: Blackfriars, 1964), pp. 13–14.
9 See S.Sutherland, Irrationality (Harmondsworth: Penguin, 1994), p. 309.
10 See J.A.Paulos, Innumeracy (Harmondsworth: Penguin, 1990), pp. 57–8.
11 See Sutherland, op. cit., p. 310.
12 D.Dennett, ‘Making Sense of Ourselves’, Philosophical Topics, 12:63–81.
13 D.Dennett, Brainstorms (Cambridge, MA: MIT Press, 1978), p. 20.
14 D.Davidson, ‘On the Very Idea of a Conceptual Scheme’, in his Inquiries into
Truth and Interpretation (Oxford: Oxford University Press, 1984).
15 Ibid., p. 197.
16 See D.Davidson, ‘Belief and the Basis of Meaning’, in his Inquiries into Truth
and Interpretation (Oxford: Oxford University Press, 1984), p. 153.
17 Ibid
18 See S.P.Stitch, ‘Could Man Be an Irrational Animal?’, in H.Kornblith (ed.),
Naturalizing Epistemology (Cambridge, MA: MIT Press, 1994).
19 Ibid., p. 344.
9 Wittgenstein and
Wittgensteinians on religion
Kai Nielsen

Wittgenstein once remarked, ‘I am not a religious man: but I cannot help
seeing every problem from a religious point of view.’1 Though he wrote very
little about either religion or ethics, it is true that a sensibility to and concern
for broadly speaking ethical and religious matters is pervasive in almost all
of his work. He wrote extensively about language, meaning, intentionality,
mind, consciousness, the self, logic, mathematics and necessity, but woven
into all these considerations, which have been central to the main historical
tradition of philosophy, is a religious and ethical concern. Perhaps it is better
characterized as an intense ethico-religious concern, for when he speaks of
ethics it is always in a distinctively religious way. But this would be badly
understood if it were taken, after the fashion of Richard Braithwaite and
R.M.Hare, to be a reductive view of religion in which religion is viewed as
morality touched with emotion associated with certain traditional narratives
which may or may not be believed.2 Wittgenstein linked ethics and religion
tightly. But, as we shall see, his thinking here was very different from that of
the reductive, basically straightforwardly ethical accounts of religion of
Braithwaite and Hare.
It should also be noted that Wittgenstein did not write treatises or even
articles on either ethics or religion and that he did not even discuss the topics
that moral philosophers normally consider. Moreover, it is clear that he would
have regarded both philosophy of religion and ethical theory with great
suspicion and even with disdain. John Hyman rightly observes:
‘Wittgenstein’s influence in the philosophy of religion is due to scattered
remarks, marginalia, and students’ notes. He never intended to publish any
material on the subject, and never wrote about it systematically.’3 But all of
that, as I will try to make plain, does not gainsay the import of my opening
quotation from him.
In understanding what Ludwig Wittgenstein has to say about religion, or
indeed about anything else, it is crucial to understand how Wittgenstein
proceeded in philosophy and why he proceeded in that way. Here we must
see that and how Wittgenstein was remarkable in generating and carrying

138 Kai Nielsen

out two revolutions in philosophy, the latter one dismantling the

philosophical practices, techniques and conceptions of the former while
keeping a very similar metaphilosophical conception of the aim of
philosophical activity. It is not an exaggeration to say, as P.M.S.Hacker does,
that ‘Ludwig Wittgenstein…was the leading analytical philosopher of the
twentieth century. His two philosophical masterpieces, the Tractatus Logico-
Philosophicus (1921) and his posthumous Philosophical Investigations
(1953), changed the course of the subject.’4 Hacker goes on to observe that

the first was the primary origin of the ‘linguistic turn’ in philosophy, and
inspired both logical positivism and Cambridge analysis in the interwar
years. The second shifted analytic philosophy away from the paradigm
of depth-analysis defended in the Tractatus and cultivated by logical
positivists…and Cambridge analysts toward the different conception of
‘connective analysis’, which was a primary inspiration of Oxford analytic

However, this remark of Hacker’s, while saying something importantly on

the mark, is also in a way misleading, for not only in tone and attitude, but in
method and aim Wittgenstein was very different from Rudolf Carnap or
Hans Reichenbach (positivists) on the one hand and Gilbert Ryle or Peter
Strawson (Oxford analysts) on the other. Wittgenstein would have rejected
the ‘scientific philosophy’ of Carnap and Reichenbach and the ‘descriptive
metaphysics’ (more descriptive than metaphysical) of Strawson as well, and
at least the avuncular, complacently confident, tone of Ryle’s ordinary-
language philosophy. Both positivism and Oxford analysis would have struck
him as scientistic—though Carnap’s and Reichenbach’s plainly more overtly
so. Moreover, the system-building of Carnap and Strawson would have been
regarded by him as impossible (more ‘houses of cards’) and, even if possible,
unnecessary and indeed harmful.
Through both revolutionary turns, Wittgenstein held a therapeutic and
anti-scientistic conception of philosophy with a deep underlying ethico-
religious intent. (Hence the word ‘harmful’ in the previous sentence.) But it is
important that we do not misunderstand Wittgenstein here. It is not at all
that he wanted to replace logic, metaphysics, epistemology or semantical
analysis with moral philosophy, reformist moralizing or some
lebensphilosophie. Nothing could be further from his intent. Rather, he
thought philosophy itself, as a particularly bad species of intellectualizing,
was bad for human beings since it stands in the way of our coming to grips
with our lives. This coming to grips with our lives—something which he took
to be supremely important—had, in his view, as well as in Kierkegaard’s,
nothing to do with philosophy. Philosophy just gets in our way here.
Philosophical perplexities, both traditional and those arising in contemporary
‘scientific philosophy’, arise from the often obsessively gripping but still
misleading pictures of the workings of our language that we come to have
Wittgenstein and Wittgensteinians on religion 139

when we reflect on it, though often we do not recognize that it is certain

pictures of our language that are generating our perplexities. And it is where
that happens that we get in philosophical trouble: we catch the philosophical
disease. We do not command a sufficiently clear view of the workings of our
language when we try to think about (for example) consciousness, thought,
sensations, truth, warrantability, intentionality and the like. The idea is not
to provide some general descriptive account of our language (Strawson) or
some formal scientific account of the semantics of our language (Carnap),
but to provide, at our conceptual trouble spots, where we are experiencing
mental cramps, a sufficiently clear representation of how our language works
in order to break that perplexity. It will not, of course, cure all perplexities
forever, but it might cure the particular one that is befuddling us—and so we
proceed on from case to case. In this way philosophy is to be therapeutic. It
does not (pace Carnap or Strawson) yield a theory of any kind—the search
for one is perhaps the philosophical illusion—but it is an activity which,
where successfully pursued, yields a sufficient understanding of the workings
of our language and, with that, of our practices and forms of life to break the
spell that a misleading picture of the workings of our language at some
particular spot exerts on us. Philosophy is taken by Wittgenstein to be an
activity and not something which constructs some theory to explain our
language or the forms of life in which our language is embedded.

I shall very briefly say something about what Wittgenstein’s second
revolution consists in and then will turn to a detailed consideration of what it
comes to for religion. Again there is a paucity of material directly on religion;
during this later period, as well as in the earlier, Wittgenstein wrote nothing
for publication specifically and in detail about religion. But there are many
things that are very suggestive, though often only indirectly, for thinking
about religion in quite different ways than have traditionally been done—
ways which I think cut through or rightly bypass much of the cackle that
goes for ‘the philosophy of religion’. Fortunately, as far as texts go, we have
in a recent work written by a former student, close friend and well-known
interpreter of Wittgenstein, Norman Malcolm, a work (Wittgenstein: A
Religious Point of View?) which provides a detailed collection of remarks on
religion made by Wittgenstein along with an analysis by Malcolm of those
remarks followed by a substantial critique of Malcolm’s account by Peter
Winch.6 In this account of Wittgenstein on religion by two prominent
Wittgensteinians, who are also philosophers of importance themselves, we
have a perceptive and faithful rendition of Wittgenstein’s views on religion,
plus, particularly on Winch’s part, the beginnings of a probing critique of
them. (Winch is less of an uncritical disciple than Malcolm is.) I shall build
on this material seeking to etch out (a) a portrayal of Wittgenstein on religion
in his later philosophy and (b) an account of some emendations provided by
140 Kai Nielsen

Wittgensteinians (principally Winch) that will not only bring out the force of
Wittgenstein’s later account, but will, pointing to its vulnerabilities, enable
us better to assess its soundness and import, both in its pristine form and in
its critical Wittgensteinian reformulations. Here we can hopefully examine
Wittgenstein’s account of religion at its full strength. I shall attempt to do
something of this.
But first for a thumbnail general account of what the later Wittgenstein was
up to. In Philosophical Investigations (1953), the central work of his later
philosophy, as well as in work beginning as early as 1930 and in work following
Philosophical Investigations, and most particularly in his last work, On
Certainty (1969), Wittgenstein articulates his changed conception of how to
proceed in philosophy and applies it to a range of philosophical problems.
Propositions are no longer construed as having a fixed logical form and, more
generally, language is no longer construed as having a fixed and timeless
structure, but is viewed as changeable, and not infrequently changing, and these
forms of language are now seen as our historically and culturally contingent
forms of life. The picture theory of meaning of the Tractatus is completely
abandoned in his later work. The conception that words stand for simple
objects that are their meanings is now regarded by Wittgenstein as a bit of
incoherent philosophy. Instead the notions of language-games and practices
are introduced. In being socialized—in learning, as we all must if we are at all
to navigate in the world, to be human—we come to have practices in which
words and actions are interwoven. In this activity, in learning to play these
language-games, we come to understand words by coming to know their uses
in the stream of life, and with this we come to know how to use words in the
course of our various practice-embedded activities.
With this, Wittgenstein abandoned his earlier formalist Tractarian demand
that language, if coherence is our goal, requires determinacy and exactness and
that the sole function of language is to describe. Rather language is seen as an
activity that has many different functions, is embedded in different practices
which answer to and structure our different needs, interests or purposes. For
someone to understand a word, it is not sufficient to bring the learner face to
face with its putative referent while repeating the word. In many cases nothing
like this is possible and in all cases, or at least almost all cases, the learner must
come to understand what kind of word he is being taught; to grasp this an
extensive training needs to have taken place in which the learner comes to be at
home with the everyday activities—the social practices—of which remarks
using the word are a part. As Wittgenstein put it in an oft-quoted remark from
his Philosophical Investigations: ‘For a large class of cases—though not for
all—in which we employ the word “meaning” it can be defined thus: the
meaning of a word is its use in the language’ (PI §43).
There is, on Wittgenstein’s account, no standing free of our practices and
forms of life or escaping the context, including the historical contexts, in
which they are embedded. Both the Tractarian (on the traditional reading)
and the metaphysical realist conception of an independently articulated world
Wittgenstein and Wittgensteinians on religion 141

are incoherent on Wittgenstein’s later account. We have no coherent

conception of a world that we can describe by accurately copying it or
mirroring it or even representing it in our thought. There are no referents
‘out there’ which simply force our concepts on us. We rather understand our
concepts by coming to understand their use in our life activities. Concepts
are aspects of our forms of life. They are not items forced on us by the world.
To understand a concept is to ‘assemble reminders’ concerning how it works
in our lives. And these will be various things, depending on the particular
concept, as part of the varied contexts and the various purposes we have.
These varied activities and ways in which we talk form our practices, and
they build together into our forms of life. We have no concepts or conceptions
which stand independently of them.
Wittgenstein’s earlier views—more accurately his metaviews—on religion,
at least on the standard interpretation of the Tractatus, could not withstand
his changed conceptions about language. As I have noted, the idea of a
general prepositional form is illusory. There is no common property or set of
common properties that all and only propositions have. There are many
different kinds of structures that we call propositions. As P.M.S.Hacker has
put it, many things count as ‘propositions’:

avowals of experience (such as ‘I have a pain’ or ‘It hurts’), avowals of

intent, ordinary empirical propositions, hypotheses, expressions of laws
of nature, logical and mathematical propositions, ‘grammatical
propositions’ (in Wittgenstein’s idiosyncratic use of this term) which are
expressions of rules (such as ‘red is a color’ or ‘the chess king moves one
square at a time’), ethical and aesthetic propositions and so on.7

In the regimented, austere conception of the Tractatus, religious utterances

are pseudo-propositions lacking bipolarity. They are, that is, not capable of
being true or capable of being false. They, on that conception, describe
nothing, are without any cognitive content at all and thus are nonsensical. By
contrast, given Wittgenstein’s later philosophy, they are not, at least on these
grounds, nonsensical. On Wittgenstein’s later, more relaxed and more
realistic, conception of propositions, many religious utterances are
propositions. ‘Bipolarity is a feature of an important member of the family,
but not a defining property of propositions as such.’ 8 Moreover,
Wittgenstein’s earlier conception that it was the sole function of a proposition
to describe is mistaken, and importantly so. The ‘roles of many kinds of
propositions, such as logical propositions and mathematical propositions, are
not to describe’.9 Yet for all of that, they are in order. We cannot take such a
short way with dissenters and simply rule out religious utterances carte
blanche as expressions of pseudo-propositions and nonsensical because they
fail to have the general form of a proposition. The shoe is on the other foot.
The error—the illusion—is to believe that there is such a general prepositional
form: that there is something that propositions essentially are.
142 Kai Nielsen

Wittgenstein continues, the above notwithstanding, to believe that

religious beliefs are very different from factual beliefs. Surface appearances
to the contrary, quite ordinary religious propositions are unlike factual
propositions. They function very differently. But they are not, Wittgenstein
now has it, any the worse for all of that. They are not therefore nonsensical.
A pervasive and, Wittgenstein believes, a pernicious error of our scientistic
culture is to try to assimilate the use of religious concepts to those of
hypotheses, predictions or theoretical explanations. To do that, he has it, is
to completely misunderstand their use. It is to be fettered by one kind of use
of language and to try to read it into other uses. When, for example, a
religious person says ‘I believe that there will be a Last Judgment’, it is a
complete mistake, according to Wittgenstein, to take that utterance as a
prediction. That is not the use, or even anything like the use, it has in religious
language-games. In believing in a Last Judgment a person is not, Wittgenstein
has it, thinking that there will be, or even that is probable that there will be,
a certain kind of event which will occur sometime in the future. The religious
person—or at least Wittgenstein’s religious person—is not thinking any such
thing.10 He is not trying to make any kind of prediction at all (LC, p. 56).
Rather, Wittgenstein equates having religious beliefs with (a) using religious
concepts and (b) having the emotions and attitudes that go with these
concepts. He remarks that ‘a religious belief could only be something like a
passionate commitment to a system of reference’ (CV, p. 64). But these
beliefs—beliefs such as a belief in the meaning of life or the meaning of the
world—can be neither true nor false. The question of their truth or falsity
cannot even meaningfully arise. Moreover they are beliefs which are neither
reasonable nor unreasonable. But what Wittgenstein does regard as
unreasonable are apologists either for or against religion who assume that
religious beliefs can in any way be tested: can be shown to be either true or
probably true or false or probably false by evidence or by argument. That
view he regards as ludicrous (LC, p. 58).
Now that something of Wittgenstein’s later conception of how to proceed
in philosophy is before our minds, I turn to an examination of Wittgenstein:
A Religious Point of View?, starting with some central considerations by
Norman Malcolm. They consist in a rather orthodox but still well thought
out articulation of a Wittgensteinian point of view.
A leitmotif of Malcolm’s discussion of Wittgenstein on religion is
Wittgenstein’s remark in Philosophical Investigations that ‘philosophy simply
puts everything before us, and neither explains nor deduces anything’ (PI
§126). Concerning this Malcolm remarks:

Wittgenstein is here proposing a radical change in our conception of what

philosophy should be doing. To say that philosophy does not seek to
explain anything is certainly not a true description of philosophy as it
has been, and still is, practiced. Many philosophers would be dumb-
founded or outraged by the suggestion that they should not be seeking
Wittgenstein and Wittgensteinians on religion 143

explanations. The traditional aim of philosophy has been to explain the

essential nature of justice, right and wrong, duty, the good, beauty, art,
language, rules, thought. A philosopher may well ask: ‘What am I
supposed to do if not explain?’
In Wittgenstein’s later thinking there is an answer. The task of
philosophy is to describe. Describe what? Describe concepts. How does
one describe a concept? By describing the use of the word, or of those
words, that express the concept, that is what philosophy should ‘put
before us’.11

There is no independent access to concepts, Wittgenstein is at pains to argue

in Philosophical Investigations, and Malcolm follows him here. Malcolm
continues, ‘The description of the use of a word is called by Wittgenstein
describing the ‘language-game’ with that word.’12 Then, without highlighting
the therapeutic side of Wittgenstein’s conception of philosophy, but in effect
remaining faithful to it all the same, Malcolm remarks that it is not the task
of philosophy to describe the use of a word in its totality, as if we had an
understanding of what it would be like to do that, but only those features of
the word that give rise to philosophical perplexity, and indeed, for any
description we might give, to a certain philosophical perplexity in a certain
situation. We assemble reminders to break a certain perplexity where we have
mental cramps concerning the workings of our language. (Here again we see
how very different Wittgenstein is from Strawson.) Describing the use of an
expression is called, rather eccentrically but harmlessly, by Wittgenstein
describing the grammar of the expression. But this, as by now should be
evident, is not just giving an account of sentence construction or syntax. The
point of speaking of language-games is to bring into focus, and clear
prominence, ‘the fact that the speaking of language is part of an activity or a
form of life’ (PI §23). Malcolm, uncontroversially and rightly, takes this to
mean ‘that in describing the language-game, or some part of the language-
game with a word, one is describing how the word is embedded in actions
and reactions—in human behaviour’.13 ‘Words’, Wittgenstein remarks in his
Zettel, ‘have meaning only in the flow of thought and life’ (Z §173). ‘Our
talk’, he adds in On Certainty, ‘gets its sense from the rest of our actions’
(OC §229). Our language-games, embedded as they are in forms of life,
provide us a place for explanations, for giving reasons and for justifications
inside the framework of these language-games or forms of life. But there is,
Malcolm has it that Wittgenstein has it, neither explanation nor justification
for the existence of these forms of language or language-games themselves.
Illustrating the way language-games work and their link with forms of
life. Malcolm comments on our use of ‘motive’. We not infrequently wonder
about the motives of people. Normally the quickest and surest way to find
out is to ask them. ‘Now of course he may not reveal it: perhaps he himself
does not understand it, or perhaps he misrepresents it both to himself and to
others.’14 But then Malcolm goes on to observe: ‘what is highly interesting is
144 Kai Nielsen

that if he does disclose his motive, typically his acknowledgement of it will

not be based on any inference from the situation, or from his own behaviour
or previous actions—as would be the conjecture of others. He tells us his
motive without inference.’15
We can, when we reflect about how this language-game works, just be
struck by its sheer existence and contingency. And this is true with the
language-games we play both with ‘motive’ and with ‘intention’ or with any
other language-game.16 We have contingency here, not necessity. Gone are
the supposed necessities of the Tractatus and indeed of the whole
philosophical tradition. Reflecting on how Wittgenstein is reasoning and how
Wittgenstein thinks we should reason if we would be realistic, Malcolm
remarks that we ‘cannot explain why this use of language exists. All we can
do is describe it—and behold it.’17 He quotes from On Certainty where
Wittgenstein makes a general comment about language-games: ‘You must
bear in mind that the language-game is, so to speak, something unforeseeable.
I mean it is not based on grounds. Not reasonable (or unreasonable). It stands
there like our life’ (OC §559).
Religions, that is Judaism, Christianity, Islam, Hinduism, Buddhism, etc.,
are ancient complex forms of life with their distinctive but purely contingent
language-games. Within these language-games there can be the giving of
reasons, explanations and justifications, but for the language-games and forms
of life themselves, as I have noted, there can be no explanations or justification,
and no foundations for them either. They are human activities that are just
there, and religious forms of life like other forms of life are neither reasonable
nor unreasonable. They do not rest on some deeper metaphysical or theological
foundations or any kind of grounding theory. They have neither some
foundationalist epistemological grounding nor any other kind of grounding,
nor do they stand in need of such grounding, rationalization or theorizing.
They are, Wittgenstein argues, in order just as they are. They are just there, as
we have already noted Wittgenstein saying, like our lives. There can be, and
indeed sometimes should be, internal criticisms within religious language-
games. Some expressions of faith are less adequate than others, less adequately
capture the aspirations of a particular religion, but there can be, Wittgenstein
has it, no intelligible standing outside these forms of life and assessing them.
Justification comes to an end when we come up against them. This is true for
all forms of life, religion included. As Malcolm puts it, giving what he takes to
be Wittgenstein’s account: ‘Wittgenstein regarded the language-games, and
their associated forms of life, as beyond explanation. The inescapable logic of
this conception is that the terms “explanation”, “reason”, “justification” have
a use exclusively within the various language-games.’18 Or again: ‘An
explanation is internal to a particular language-game. There is no explanation
that arises above our language-games and explains them. This would be a
super-concept of explanation—which means that it is an ill-conceived
fantasy.’19 What we can and should do as philosophers is observe and describe
language-games; and, with hard work and luck, we will come to see more
Wittgenstein and Wittgensteinians on religion 145

clearly, by a perspicuous representation, the use of the terms of our language-

games and the role they play in our lives. Philosophy, the kind of therapeutic
philosophy that Wittgenstein, Malcolm and Winch practise, enters when we
become entangled in our concepts—the use of our terms. There, in such
particular situations, philosophy can, by assembling reminders for a particular
purpose, enable us to command a clearer view of our use of these terms, and it
can dispel our confusions about them. Philosophy, Wittgenstein has it, as do
neopragmatists as well, ‘cannot explain why anything happens or exists’ and
‘it cannot reveal the essential nature of anything’, for there are no such essential
natures. Its way of proceeding is descriptive and elucidatory, elucidatory in the
service of dispelling the confusion we almost invariably fall into when we reflect
on our concepts. We normally can operate with them without difficulties, but
we often fall into confusions—suffer from mental cramps—when we try to
operate upon them.
All of this, of course, applies to our religious concepts as much as to any
other concepts. When the engine isn’t idling, when we work with our
concepts—operate with them rather than upon them—we understand them
well enough if we have been enculturated into such forms of life, but when
we think about them, as when we think about other concepts as well, we
almost irresistibly fall into confusion about them. The task of philosophers,
for themselves as well as for others, is to dispel such confusions by providing
in situ a perspicuous representation of these concepts. We move about,
usually effortlessly, in our grammar, in our everyday practices. But in thinking
about what we do with words we not infrequently fall into perplexity. In
order to remove our misconceptions, Malcolm has it, no theorizing is called
for, neither scientific nor philosophical. What is required is only that we look
carefully at the grammar which is at our command. We can think with it
even if we stumble, while still thinking with it, when we try to think about it.
But in doing this Wittgenstein’s counsel is ‘Don’t think, but look’ (PI §66).20
Philosophy, that is good philosophy, should replace our age-old
metaphysical theorizing and its ersatz scientific replacements in a so-called
‘scientific philosophy’ bent on formulating theories about the nature of
meaning, thinking, reference, belief, knowledge, mind, good and God. By
contrast, good philosophy—therapeutic philosophy practised after the
fashion of Wittgenstein—cannot interfere with the actual use of language.
For good philosophy elucidation comes to accurate description in the service
of dispelling confusions about our use of language. We have a mastery of our
language—of the use of our terms—but we fall into confusion when we try
to think about those terms, when we reflect about our use and try to grasp
‘the essence’ of our concepts expressed by these terms. Wittgenstein remarked
in his Philosophical Investigations that our ‘mistake is to look for an
explanation where we should see the facts as “primary phenomena
[Urphänomene]”. That is, where we should say: this language game is played’
(PI §654). Or again: ‘The question is not one of explaining a language-game
by our experiences, but of observing a language-game’ (ibid.).
146 Kai Nielsen

A language-game, as Malcolm well puts it, ‘is an employment of language

that is embedded in one of the innumerable patterns of human life’.21 Some
forms of life are forms of life that not all people in all cultures share.22 We
cannot, Malcolm has it, explain why this is so: that is, why some people have
them and others do not. He remarks: ‘Neither philosophy nor science can
explain this. What philosophy can do is to correct our inclination to assume,
because of superficial similarities, that different language-games and forms
of life are really the same.’23 Some words refer to or stand for something.
They have a reference. But ‘Hans’, ‘blue’, ‘2’, ‘the Empire State Building’,
‘grace’, and ‘God’ do not all refer in the same way. We must, in particular,
not assume that ‘God’ refers like ‘Hans’ or ‘the Empire State Building’.
Wittgenstein, and for that matter Norman Malcolm and Peter Winch, both
following Wittgenstein, are as much set against the idea that there could be a
single true description of the world or some ultimate explanation which
would show us what reality really is as are neo-pragmatists such as Richard
Rorty and Hilary Putnam. Such notions, they all believe, are without sense.
Natural theology and natural atheology, as much as metaphysical realism,
are incoherent. We can have no such knowledge and we do not need it.
Religious beliefs can neither have any backing from metaphysics or natural
theology or science, nor do they need it. (Here there is no difference between
the earlier and the later philosophy of Wittgenstein.) But, by parity,
atheological metaphysical theories or so-called scientific theories of a so-
called scientific philosophy or a ‘scientific world-view’, which are really
metaphysical theories in disguise, are also incoherent and can provide no
intelligible ground or basis for rejecting or criticizing religion. Such
activities—theology and atheology—take in each other’s dirty linen. Both
should be set aside as houses of cards.
However, Malcolm is quick to remind us that Wittgenstein’s account is
not a form of irrationalism or nihilism which says goodbye to reason or
reasonableness, though Wittgenstein does, as much as does Paul Feyerabend,
say farewell to Philosophical Reason or Scientific Reason (so-called
‘Philosophical Reason’ or ‘Scientific Reason’ would be more apt). But to be
against Reason is not to be against reasoning and justification within
language-games and against the reflective effort to make sense of our lives
and to be reasonable. And that reasons, falsifications, explanations come to
an end ‘does not mean that there are no reasons, justifications, explanations
for anything’.24 Within many of our language-games, when we are operating
with them, and reasoning and reflecting inside their parameters, reasons,
justifications, explanations often can be given and often are perfectly in place.
What, however, Wittgenstein does stoutly argue for, and Malcolm and Winch
follow him here, is that the giving of reasons, justifications and explanations
comes to an end somewhere: ‘Where is that? It is at the existence of the
language-games and the associated forms of life. There is where explanation
has reached its limit. There reasons stop. In philosophy we can only notice
the language-games, describe them and sometimes wonder at them.’25 There
Wittgenstein and Wittgensteinians on religion 147

we see what has been called Wittgenstein’s ‘quietism’. Quietism or not, for us
here it is a key question whether, and if so how, it applies to religion—to
Christianity, Hinduism, Islam, Judaism and the like. What is at least initially
unsettling in this context in thinking about Wittgenstein and Wittgensteinians
such as Malcolm and Winch is that it seems that, if their way of characterizing
how to proceed in philosophy is correct, this means that no philosophical or
any other kind of reasonable criticism, or for that matter defence, is possible
of forms of life or, indeed, any of form of life, including Hinduism,
Christianity and the like. Is this where we are at? Is this the end of the line?

It can be responded to such Wittgensteinianism that religions, and most
strikingly Christianity with which Wittgenstein and Malcolm are most
concerned, are inescapably in part metaphysical religiosities.26 Moreover, the
part that is metaphysical cannot be excised from the rest, leaving the rest
intact. Without a metaphysical part as a settled element (component) in that
form of life, the form of life will not even be recognizable as Christianity,
Hinduism, Judaism, or Islam. There are no doctrineless or creedless religions.
Religion is a doing, a committing yourself to act or try to act in a certain way,
but it is not only that. In Christianity, for example, God is said to be the
ultimate spiritual being—the very ground of the world—transcendent to the
world and, in being so, eternal and beyond space and time. And it is an
essential part of that very religion to believe that human beings have immortal
souls such that they—that is we—will not perish or at least will not perish
forever when we die: when, that is, we lose our earthly life. And in addition
there is what Kierkegaard called the scandal of the Trinity, but still, he
believed, a scandal to be accepted trustingly on faith. These are central beliefs
for Christianity, and without them Christianity would not be Christianity. It,
of course, is not only a doctrinal system. It is also, as Wittgenstein and
Kierkegaard stress, a demanding way of life that requires of believers—
genuine believers—a reorientation of their lives. But it is also, and
inescapably, a belief system with a set of doctrines.
This belief system is a metaphysical belief system and Christianity
integrally is a metaphysical religiosity. It simply comes with the religion. But,
if what Wittgenstein, Malcolm, Winch and the pragmatists say is so,
metaphysical belief systems are all incoherent: ‘houses of cards’, as
Wittgenstein said. But then that very form of life, metaphysically infused as it
is, should be said by Wittgenstein and Wittgensteinians to be incoherent. But
that is not at all what they say.
Still, that anti-metaphysical strain is central to their accounts. But, on
another equally central part of Wittgenstein’s account, Christianity can’t be
incoherent, for Christianity, as other religions as well, is a language-game—
an employment of language embedded in a pattern of human life—and thus
a form of life. But forms of life and language-games cannot on Wittgenstein’s
148 Kai Nielsen

account be incoherent or illusory or even, in any central or crucial way, in

error. Such notions have no application with respect to forms of life. So we
can see here that something has to give. Two central points of Wittgenstein’s
account—or so at least it seems—are incompatible with each other. Religions
are metaphysical schemes and metaphysical schemes are incoherent, but
religions are forms of life and it makes no sense to say of a form of life that it
is incoherent.
Wittgenstein, I think, would respond—and here I think Kierkegaard would
respond in the same or a similar manner—that these doctrinal elements, these
metaphysical or metaphysical-theological beliefs, as important as they have
historically been to Christianity and other religions as well, are nonetheless
incoherent and should be set aside while still keeping, for example, the kernel
of Christianity. These religious metaphysical beliefs are not what is really
important in religion, and religiously sensitive people have—though
sometimes inchoately—always recognized that.
What Wittgenstein saw as important in religion is that if one could have
faith—if one could trust in God—then that will turn around one’s life,
enabling one to be a decent person and to do good in the world without
vanity or arrogance. He took faith without works to be utterly vain. Indeed
it would not, as he saw it, properly speaking even be called ‘faith’. Moreover,
as he says in his Notebooks of 1916, ‘to pray is to think about the meaning of
life’ and ‘to believe in God means to see that life has meaning’ (p. 73).27 These
remarks are, against most of the philosophical temper of Wittgenstein, utterly
reductionistic. If what they say were so, it would make, by implicit stipulative
redefinition, many reflective and sensitive atheists into believers in God. By
verbal magic, all sensitive, reflective, caring people become religious believers.
It is to take what may very well be a necessary condition for genuine religious
belief and turn it into a sufficient one. Is this the end of matter? Perhaps not
quite. Let me proceed indirectly by first recording some of Wittgenstein’s
specific comments about religion. Moreover, it is important to keep in mind
the fact, that historically, religions have changed over time; moreover, there
is no reason to believe ‘history has come to end’ and to think that they will
not continue to change.
Wittgenstein had, as I have remarked, scant patience with philosophical
theology or the philosophy of religion, but throughout his life he read and re-
read the Gospels, thought at one time seriously about becoming a priest, and
was deeply taken by the ancient liturgical prayers of the Latin rite and their
translation in the Anglican prayer book, remarking that they ‘read as if they
had been soaked in centuries of worship’.28 Speaking to his close friend
Maurice Drury, who had formed the intention to be ordained as a priest,
Wittgenstein remarked:

Just think, Drury, what it would mean to have to preach a sermon every
week. You couldn’t do it. I would be afraid that you would try and give
some sort of philosophical justification for Christian beliefs, as if some
Wittgenstein and Wittgensteinians on religion 149

proof was needed. The symbolisms of Catholicism are wonderful beyond

words. But any attempt to make it into a philosophical system is

It was the activist life-orientation, involving not the speculative-cosmological

side of Christianity, that appealed to Wittgenstein. What gripped him was
Christianity’s call to radically alter the manner of one’s life—to be just and
caring with one another, to clearly see what a wretched person one was, to
atone for one’s sins, and to struggle to be a decent human being.
The influence of Kierkegaard on Wittgenstein was very deep. It shows itself
in the above remarks about guilt and sin and, again quite differently, in his
attitude towards the historical claims of Christianity and in what he thought
philosophy could achieve vis-à-vis religion. Wittgenstein (echoing Kierkegaard)
wrote: ‘Christianity is not based on a historical truth; rather it gives us a
(historical) narrative and says: now believe! But not, believe this narrative with
the belief appropriate to a historic narrative—rather believe, through thick and
thin, and you can do that only as the result of a life’ (CV, p. 32).30 Wittgenstein,
again like Kierkegaard, saw religion not only as something that makes extreme
demands on one, but as something which answers to the needs of genuinely
religious people—people who see themselves not only to be extremely imperfect
but as wretched. ‘Any half-way decent man’, Wittgenstein wrote, ‘will think
himself extremely imperfect, but a religious man believes himself wretched’
(CV, p. 45).31 Somewhat earlier in Culture and Value, Wittgenstein wrote ‘faith
is faith in what my heart, my soul needs, not my speculative intelligence. For it
is my soul, with its passions, as it were with its flesh and blood, that must be
saved, not my abstract mind’ (CV, p. 33).32
Wittgenstein, given his sense of what religion really is, is fully, intellectually
speaking, on the fideist side coming down to us from Tertullian, Pascal,
Hamann, and, most fully, from Kierkegaard. But in his very conceptualization
of fideism there was also for a religious person an intense activist side very
distinct from his quietism in philosophy. This comes out strikingly in a remark
he made in 1946:

One of the things Christianity says, I think, is that all sound doctrines
are of no avail. One must change one’s life. (Or the direction of one’s
That all wisdom is cold; and that one can no more use it to bring one’s
life into order than one can forge cold iron.
A sound doctrine does not have to catch hold of one; one can follow it
like a doctor’s prescription.—But here something must grasp one and
turn one around.—(This is how I understand it.) Once turned around,
one must stay turned around.
Wisdom is passionless. In contrast faith is what Kierkegaard calls a
(CV, p. 53)
150 Kai Nielsen

For Wittgenstein, as for Tertullian, Pascal, Hamann, and Kierkegaard,

religion was not a question of proving anything or even the articulating of
doctrine, even a doctrine that orders one’s life:

[Wittgenstein] objected to the idea that Christianity is a ‘doctrine’, i.e., a

theory about what happened and will happen to the human soul. Instead,
it is a description of actual occurrences in the lives of some people—of
‘consciousness of sin’, of despair, of salvation through faith. For
Wittgenstein the emphasis in religious belief had to be on doing—on
‘amending one’s ways’, ‘turning one’s life around’.33

He came to have, mixed together with this striving to turn his life around and
his realizing that this was what religion was about, an intense desire for purity
together with an equally intense sense of his own impurity, his sinfulness and
guilt, his standing under divine judgment, his need for redemption and
forgiveness. He had a keen sense of a judging and redeeming God, but the
conception of a creator was foreign to him and, as Malcolm put it, ‘any
cosmological conception of a Deity derived from the notion of cause or of
infinity would be repugnant to him’.34
In spite of Wittgenstein’s statement ‘I am not a religious man’, I think that
it is, as Malcolm puts it, ‘surely right to say that Wittgenstein’s mature life
was strongly marked by religious thought and feeling’.35 Kierkegaard had
percipiently shown how difficult it is to be religious, how many people are
deceived in thinking they are religious when they are not, and that some
people who would honestly say they are not, and even some—say, militant
atheists—who would vehemently assert that they are not, are nonetheless
religious, and indeed deeply so. It is also the case that with his clarity of
intellect, together with his deep religious sensitivity, Wittgenstein is likely to
have had a keen sense of what a religious form of life is. I have claimed, as
have many others, that there is no doctrineless religion and that religion
inescapably involves making cosmological (metaphysical) claims. 36
Wittgenstein firmly rejects this. Is he right to do so?

Concerning what was discussed in III and what I continue to discuss here, it
will be necessary, as Winch reminds us, ‘to observe the distinction between
Wittgenstein’s own religious reflections and his philosophical comments on
religious discourse’.37 I shall centrally be concerned with the latter and show
concern with the former principally to help us, if it can, to gain a purchase on
how we should think and feel about religion. I want to try to see what kind of
form of life it is, what kind of language-game it is, and what role it can and
should play in our lives. And what philosophically we are justified in saying
about these matters.
Malcolm’s account of how Wittgenstein understands religion and how he
Wittgenstein and Wittgensteinians on religion 151

understands philosophy in relation to religion is an important one. That

notwithstanding, Peter Winch, I believe, has brought out some important
ways in which Malcolm’s account is flawed. I want to highlight them and
then comment on them.

1. As we have seen, Malcolm claims, and claims for Wittgenstein, that

there is no explanation for the existence of language-games or forms of life.
Winch says that this is misleading. I think he actually shows something
stronger, namely, that, taken straightforwardly, the claim is just false. Still,
though false, taken straightforwardly, we can give it a very specialized reading
in which it is not false and, so understood, it makes an important point that
is frequently lost sight of in thinking about what religion is.
Winch does not disagree with the general understanding that Wittgenstein
firmly maintained that explanation has an end and that explanatory theories
are inappropriate in philosophy.38 Good philosophy, he agrees, should be
descriptive in the way Malcolm, following Wittgenstein, characterizes. That
is not at all in dispute—and rightly so—as being something that Wittgenstein
firmly maintained. Moreover, Winch, like Malcolm, thought that
Wittgenstein was right about this. But Malcolm overlooks, Winch has it, ‘the
very different issues that are at stake in various of the contexts in which
Wittgenstein insists that “explanation has an end”’.39 Winch writes that it ‘is
misleading to say that “Wittgenstein regarded the language-games and thus
associated forms of life, as beyond explanation.” Language-games are not a
phenomenon that Wittgenstein had discovered with the peculiar property that
their existence cannot be explained!’40 Malcolm maintains that neither the
‘hard’ sciences nor the ‘soft’ sciences can explain why various practices exist.
But, as Winch points out, that is simply false. There are many cases, he
observes, ‘in which historians, anthropologists or linguists give well founded
explanations of the existence of this or that practice. Why ever not! The
important question for us [that is, for we philosophers] to ask is: what
relevance would such explanations have to the resolution of philosophical
difficulties?’41 What Winch takes it that (pace Malcolm) we should not
maintain is that language-games are intrinsically beyond the power of these
sciences to explain; but rather what we should say is that any explanation
they might offer would turn out to be quite uninteresting—and useless as far
as the philosopher’s characteristic puzzlement is concerned.42
Wittgenstein, Winch has it—and it seems to me correctly—was not
concerned to deny that there was any reasonable sense in which explanations
of practices could be given. He was concerned, rather, ‘with the peculiar
pseudosense in which philosophers seek “explanations”’. Spinoza, for
example, thought ‘that because explanations have to come to an end there
must be something which has no further explanation, a causa sui’. 43
Wittgenstein was concerned to combat that, to show that that kind of
rationalism is senseless: that it makes no sense to say that there is something
beyond explanation—intrinsically unexplainable—on which all ordinary
152 Kai Nielsen

explanations depend or that there is, if we push matters resolutely, some

ultimate explanation—some super-explanation as it were—which finally
explains everything and brings inquiry to an end. Wittgenstein does not think,
Winch observes, ‘that explanations come to an end with something that is
intrinsically beyond further explanation. They come to an end for a variety
of quite contingent and pragmatic reasons, perhaps because of a practical
need for action, perhaps because the puzzlement which originally prompted
the search for explanation has evaporated (for one reason or another)’.44
There are many situations, perhaps most situations, in which we have no
need ‘at all’ to explain a practice. The practice seems to us—and sometimes
rightly so—unproblematic. But then, as C.S.Peirce and John Dewey stressed,
circumstances might arise in which we need, or at least want, an explanation
for one or another specific pragmatic purpose—political, moral, sociological
or some combination of them—or perhaps because the practice does not seem
for some reasonably specific reasons to be working so well and indeed might
not be working well. Such situations do arise, and there is no reason to think
such explanations, answering to such problematic situations, are impossible,
always or even generally undesirable, or that they will invariably, or even
standardly, degenerate into philosophical pseudo-explanations. Moreover, we
do not have good textual grounds for thinking that Wittgenstein thought that.
Suppose, however, we stop talking about explanation and talk of justification
instead. Wittgenstein also famously said that justification must come to an end
or it would not be justification. Malcolm has stressed, as a view which is both
Wittgenstein’s and right, ‘that reasons, justifications, explanations, reach a
terminus in the language-games and their internally related forms of human
life’. Let us set aside explanation and just concentrate on the giving of reasons
and the justifying (if such is in order) of a form of life. Winch takes it, correctly
as a bit of Wittgenstein exegesis, ‘that the expression of religious belief is itself a
language-game for which it makes no sense to ask for…rational justification’.45
Within a form of life, a justification of particular beliefs or particular conceptions
in accordance with the constitutive norms and conceptions of that form of life
can sometimes be given. But a request for a justification of the constitutive norms
and conceptions—the very framework beliefs of a religious form of life—is
another matter. Wittgenstein has it that to ask for justification here is senseless.
Job’s seeking to require God to answer to him is seen to rest on a mistake for
one who has faith. The questioning of why God’s will is sovereign and should
never be doubted—the challenging of the whole framework—is, given
Wittgenstein’s conceptions, out of place. Indeed, not simply out of place, but
incoherent. Malcolm had remarked, Winch reminds us, that even in this
technological and materialistic age, there are people who are inside the practices,
the language-games of, say Christianity or Judaism, who pray ‘to God for help,
asking Him for forgiveness, thanking Him for the blessings of this life—and
thereby gain comfort and strength, hope and cheerfulness. Many of these people
would have no understanding of what it would mean to provide rational
justification for their religious belief—nor do they feel a need for it.’46
Wittgenstein and Wittgensteinians on religion 153

And indeed Wittgenstein has it—and here both Malcolm and Winch
follow him—justification there is not possible and, moreover, there is no need
for it. Asking for it is not only obtuse but is wrong: morally wrong.
There are at least three issues here. First, it seems fair enough to say that a
plain untutored person—say a minimally educated person living in an isolated
community of believers—is being reasonable—or at least not unreasonable—
in so believing. Moreover, it would, in most circumstances, be sadistic to
challenge such a person’s faith—a faith that that person regards as an
undeserved gift from God. It would be unnecessary and pointless cruelty,
causing, if it was at all psychologically effective, unnecessary and pointless
suffering. Second, there is the question whether, if that person began to feel—
say quite without wishing it—the irritation of doubt, whether (a) there are
considerations available to an honest, reflective person sufficient to still,
without subterfuge or self-deception, those doubts or (b) whether this is even
an intelligible or legitimate possibility: whether it makes sense to have such
doubts? They may themselves rest on philosophical confusions. Moreover,
perhaps concerning something so basic—something so much a part of the life
of some people—we have something which does not admit of such
rationalization, such a reasoning out of things? Third, whether, that isolated
person aside, for anyone in our modern cultures there are considerations
which that person, or several persons reasoning together and sensitively
feeling through the matter, could articulate that would show such beliefs to
be not only coherent but not unreasonable? Or to come to the opposite
conclusion? Are these, as it seems Wittgenstein and Wittgensteinians
believe—must believe?—bad questions? But if that is claimed, it seems to be
in order for us to ask: Just why are these bad questions? Or are they really
bad questions? Do we just have, in maintaining they are, Wittgensteinian
dogma here?
I think any Wittgensteinian would respond to this last query, and the second
one as well, by rejecting them out of hand. It is practices which give the
intelligibility and coherence to talk—words as they are used in their living
contexts, in this case the context of a living engaged faith. If theorizing, he
would say, makes the talk seem incoherent or unreasonable, then so much the
worse for the theorizing. Moreover, and in addition, religion is something
special, for it is not a matter, except peripherally, of the intellect but of the
heart. The intellect in this context can only dispel bad philosophical reasoning
that gets in the way of faith. There is in such fideistic reasoning a great distance
between the confident doing of natural theology by Aquinas and Scotus and
the fideistic reasoning of Kierkegaard and Wittgenstein: between the confident
claim that if we reason carefully and attend to the facts we can see that it is
irrational not to believe in God and the acceptance of God simply on faith—on
a faith, or a trust, that eschews all search for or recognition of the
appropriateness or even the very possibility of justification, except in the purely
negative sense of showing the mistake of those who would say that without
justification your faith is in vain. For to say that—to demand justification
154 Kai Nielsen

here—is not only unjustified but is unjustifiable. Philosophical clarity,

Wittgensteinians will argue, shows such argumentation is at best mistaken. If
Wittgenstein and Wittgensteinians—such as Norman Malcolm, D.Z.Phillips,
Rush Rhees, Peter Winch, Stanley Cavell and James Conant—and neo-
pragmatists such as Hilary Putnam and Richard Rorty are right about the
incoherence of metaphysics and foundationalist epistemology, then the
rationalistic arguments of the philosophy of religion or natural theology or
atheology cannot get off the ground. Then isn’t the conclusion we should come
to about religion such a Wittgensteinian one? Though this, of course, does not
mean that we ourselves should become religious, but that we should desist from
making claims about religious belief resting on a mistake. That is itself, they
would argue, a mistake—a very big philosophical mistake. We might continue
perfectly appropriately, if we are, to be atheists. But we should not engage in
atheology—philosophical arguments for atheism. Philosophy has nothing to
say here either for or against religious belief. Isn’t this the conclusion we should
be drawn to?
2. Perhaps what has been said above should be sufficient to put such
matters to rest, to lead us, if we would be reasonable, to react and view
things in such a manner. Still such an equilibrium is seldom the case in
philosophy over something so fundamental. So let us look at things from
another angle. Malcolm, correctly catching something that Wittgenstein
stresses, remarks that what for Wittgenstein is ‘most fundamental in a
religious life is not the affirming of creeds, nor even prayer and worship—
but rather, doing good deeds—helping others in concrete ways, treating
their needs as equal to one’s own, opening one’s heart to them, not being
cold or contemptuous, but loving’.47 Surely someone with any religious
sensitivity—or indeed with just plain human sensitivity—will feel the force
of that. That said, Winch’s cautionary remarks are very important here.
Winch says that the link between faith and works ‘is by no means as
straightforward as Malcolm’s discussion may suggest’.48 There are people
with just the doings and feelings described above—people having exactly
those attitudes—who lack religious sensibility, who, as Malcolm himself in
Chapter 7 of his book reminds us, ‘take a serious view of religion, but
regard it as a harmful influence, an obstacle to the fullest and best
development of humanity’.49 Are we, to return to something mentioned
earlier, to turn them into religious believers—people with religious faith—
by stipulative redefinition? Wittgenstein remarked to Drury that it was his
belief ‘that only if you try to be helpful to other people will you in the end
find your way to God’.50 Winch tersely and correctly remarks:
It is important because Wittgenstein did not say that being helpful to
other people is finding one’s way to God, nor that it is a sufficient
condition of doing so. He said it is a necessary condition of doing so.
One cannot live a godly life without ‘good works’; but all the same there
is more to the godly life than that.51
Wittgenstein and Wittgensteinians on religion 155

Moreover, as Winch also emphasizes, we cannot, as Malcolm sometimes

seems to think, understand the ‘works’ Wittgenstein stresses—understand the
role they play in the believer’s life—independently of their connections ‘with
a particular faith on the part of the doer’.52 The doing of good, the being
loving and humble, are for the religious believer internally connected to the
‘use of the language of faith in the life of the believer’. This seems to me, but
perhaps not to Winch, to imply that they cannot in the thought and actions
of a believer be disconnected from certain doctrinal strands and the creedal
expressions of a particular religion. But this at least seems to run against
Wittgenstein’s own setting aside of doctrines as not being what religion is.
It is not difficult to surmise how Wittgenstein, and Kierkegaard as well,
would respond. ‘There you go again’, they would no doubt snort, ’with your
stubborn and even arrogant intellectualism, turning religion into a theory—
failing to see what is there before your eyes that gives religion its importance.
It is not doctrines or creeds that count but commitment and concern turned
into action on yourself, though at the same time with a certain inwardness,
and for others. Religion is ultimate commitment and concern. Brush aside all
this sterile intellectualism. Theorizing about religion is not the way to God:
thinking of great intellectual mansions while you live in a little shack’
(Kierkegaard’s comment on Schelling and Hegel).
Theorizing about religion is, indeed, not the way to God, if there is a way
to God. The way is your action on yourself and for others, but, if it is done
religiously, it is embedded in words integral to a form of life that would not
be the form of life it is without the doctrines and the creeds. Religions are for
the sake of life—for the very things Wittgenstein stresses—but the religious
believer, immersed in those forms of life, sees and feels his commitment and
concern and his deeds in terms of these very forms containing these doctrines
and creeds. He does not have religious feelings which swing loose from
religious concepts. Both his very understanding and his deepest reactions are
tied up (internally linked) with doctrines and creeds and the distinctive
concepts that go with them. And his reactions and understanding here cannot
be split apart (as if there were a ‘cognitive’ and a ‘noncognitive’ side to them).
There is no religious understanding without the reactions and no reactions
which are intelligibly religious without that understanding.
To try to reduce religion and religious belief to some basic deep commitment
and a concern to be a decent human being, to really care about others and do
good, even if we add—probably with very little understanding—‘ultimate’ to
‘commitment’ and ‘concern’, just takes what, as I have already observed, is a
necessary condition for being genuinely religious (note the implicit persuasive
definition here) and turns it into a sufficient condition. On such a view of things
Marx, Engels, Luxembourg, Durkheim, Freud, Dewey, Weber, Gramsci all
become religious. But that is a reductio.
3. Wittgenstein, under the influence of William James’ Varieties of
Religious Experience, came to recognize that the ways in which people
156 Kai Nielsen

express their religious beliefs differ enormously.53 Even within a given

confessional community there are ‘vastly diverse forms of religious sensibility.
And these different forms of diversity crisscross in bewilderingly complex
ways’.54 Even if we avoid any attempt to so define ‘religion’, such that the
term captures all the great historic religions as well as those group activities
and beliefs anthropologists firmly regard as religious activities of recognizable
religions (e.g. the religion of the Dinka or the Neur), and concentrate only on
those religions—Christianity and Judaism—in which Wittgenstein took the
most interest, we still get very diverse forms of religious sensibility and
conceptualization and interpretation of doctrine, and even doctrine itself.
Wittgenstein saw life as a ‘gift’ from God for which one should be grateful,
but life, he firmly believed, was something that also imposes inescapable
obligations. He also thought that in his work and in his life he required help—
some ‘light’ from above, as he put it. These attitudes, Winch observes,
unlinked as they are with specific confessional commitments, are from the
viewpoint of a ‘developed theological doctrine’ inchoate.55 But this, as Winch
is perfectly aware, would not have bothered Wittgenstein one bit. He set
himself, as we have seen, against theological and religious doctrine. More
worrisomely, from Wittgenstein’s point of view, there are, considering the
above attitudes—the above expressions of religious sensibility—some serious
and reflective people whose very seriousness manifests itself in opposition to
such attitudes.56 Some people will have an attitude that accepts ‘one’s fate as
“the will of God”, an attitude which neither pretends to provide any
explanation of that fate nor seeks to find one’. 57 This attitude
characteristically goes along ‘with an attitude of gratitude for life’.58 But
Wittgenstein remarks, commenting on the expression of a very different

We might speak of the world as malicious; we could easily imagine the

Devil had created the world, or part of it. And it is not necessary to
imagine the evil spirit intervening in particular situations; everything can
happen ‘according to the laws of nature’; it is just the whole scheme of
things will be aimed at evil from the very start.
(CV, p. 71)

But the reference to the Devil here is, of course, no more an explanation—
nor does Wittgenstein think that it is—than is a reference to the will of God.
Either viewed as an attempt at an explanation would be what Wittgenstein
called an unnecessary and stupid anthropomorphism (ibid.). But faced with
all the horrible contingencies of life, the suffering, cruelty, indifference, pain,
jealousy, failures of integrity, the breaking of trust—the whole bloody lot—
some would speak of neither God nor the Devil, or of the goodness, in spite
of it all, of the world, or of the malignancy or maliciousness of the world.
Indeed they would think (pace William James) that such talk makes no sense.
Some would say, as I would, ‘That’s how things are’ without reference to
Wittgenstein and Wittgensteinians on religion 157

God or the Devil.59 I think (to abandon for a moment a Wittgensteinian

commitment to description and to speak normatively) this austere approach
is a more proper frame of mind. We see that the plague is always with us,
sometimes rather dormant but at other times raging, and always as something
that will return, and we resolve to fight the plague. (Recall this was Albert
Camus’ figure of speech, and his resolve as well.) But again this expression of
attitude makes no more an attempt at an explanation than does the
expression of the attitude that goes with speaking of God’s will, or of the
Devil having created the world. Winch remarks perceptively that one ‘might
want to single out the reference to the will of God as the only one that
expresses a religious attitude; or one might want to single out “that’s just
how things are” as the only attitude genuinely “free of all superstition”’.60
Our language-games and forms of life, he observes, let us do either. And
people, of course, do either. People, including reflective people of integrity,
often differ here. And, as Hilary Putnam stresses, this is something to take to
heart.61 Moreover, it is not at all evident, to put it minimally, that there would
be anything even approaching a consensus about which attitudes are the more
appropriate or which run the deepest. Indeed, not a few will think there is no
answer to such ‘questions’. And others would think that, even if in some sense
there were, it would be inappropriate to ask them.
‘It’s God’s will’, ‘It’s the work of the Devil’ and ‘That’s how things are’ are
all non-explanatory and in some language-games are where not only
explanation stops, but where justification and the giving of reasons stops. I
think myself ‘That’s how things are’ is by far the more adequate way of viewing
things. It is cleaner, with less mystification, and comes closer to—or so I think—
telling it like it is. However, it should be immediately sceptically queried: How
can I consistently say anything even remotely within that ball-park, given my
pragmatist and Wittgensteinian perspectivism and contextualism, with their
rejection of the idea that there can be a one true description of the world and
my arguments to the effect that it makes no sense to say that one vocabulary
is closer to reality than another, or that we can coherently speak of standing
outside all our practices and assessing them, or that there is some unifying
comprehensive practice that, like the Absolute, encompasses everything?62 I
could say that for certain purposes ‘That’s how things are’ is the more adequate
response, and that for other purposes the other ways are better, but I could
not consistently, it is natural to respond, flat-out say ‘That’s how things are’
is the more adequate conception. I could not say this because some non-
contextualist conception of ‘That is how things are’ is unintelligible. And,
even more plainly and less controversially, I cannot even consistently say that
that is so because it comes closer to telling things like they are and is less
mystificatory and obfuscating. There is no way things are überhaupt, or, even
if there is, even if such talk somehow makes sense, we have no way of knowing
or even plausibly conjecturing when this is so. So we are back with my
pragmatic contextualism and how Wittgenstein sees things.
I think I can consistently have my pragmatist perspectivism and my claim
158 Kai Nielsen

about the greater adequacy of what I call my more austere That’s how things
are’ way of viewing things. I will now argue that this is so: that it is not a case
of having my cake and eating it too. We have genuine descriptions and
explanatory practices, which are alternatives to each other: for example, the
giving of a physiological description of bodily movements or a description in
terms of actions and intentions; or, to take another, the giving of a
commonsense description of tables, bits of mud, water flowing, the moon
being pink on a given night, in contrast to giving a scientific physical
description where we will say different things about solidity, colour and the
like. These are alternative descriptive and explanatory practices utilized for
different purposes. But none of these descriptions are ‘closer to reality’ or
more adequate sans phrase than any other. We can say only that for different
purposes one is more adequate than another; not that one is a more adequate
or a better telling-it-like-it-is than another—period. There the story about
my perspectivism and contextualism is perfectly in place. It is also the account
that Richard Rorty and Hilary Putnam would give of things.
In saying, ‘That’s God’s will’, ‘That’s the Devil’s work’, ‘That’s how things
are’, do we not also have, in a way similar to describing things in terms of
bodily movements and describing things in terms of human actions, different
perspectives answering to different interests with none of them being in some
general, ‘perspective-neutral’, sense more adequate? We can and should retort
by remarking that with ‘That’s God’s will’ or ‘That’s the Devil’s work’ we
have metaphysical utterances penetrating into our common life. They are
metaphysical conceptions. And they, as metaphysical conceptions, are,
Wittgenstein and both Malcolm and Winch following him argue, and, as we
neo-pragmatists argue as well, utterances which, in being metaphysical
utterances, are incoherent, yielding pseudo-descriptions and pseudo-
perspectives from which no intelligible descriptions, interpretations or
explanations could flow. They yield nonsense, but not ‘intelligible nonsense’
somehow conveying cognitive depth as traditionalist interpretations of the
Tractatus claim Wittgenstein obliquely hints at. If Wittgenstein, the
Wittgensteinians and the neo-pragmatists are right in seeing metaphysical
claims as houses of cards that require philosophical therapy to break their
spell, we do not have three alternative perspectives here but only one—one
that (a) in effect summarizes a bunch of empirical observations and more or
less concrete moral observations and (b) makes a morally freighted
generalization about them. On the other hand, we have two metaphysical
fantasies that have crept into the language-games of some people. These
metaphysical fantasies are, as Wittgenstein puts it in other contexts, wheels
that turn no machinery, conceptions that do no work in these practices, and
the people who use such phrases are only under the illusion that they have
some understanding of what they are saying and that these metaphysical
conceptions are functioning parts of our social practices with their embedded
language-games. There are no metaphysical forms of life. (If it is replied that
they do rhetorical work, this is in effect to concede the case.) It is not like
Wittgenstein and Wittgensteinians on religion 159

saying that we use physiological descriptions for certain purposes and action-
intention descriptions for other purposes and that both can be perfectly in
place for their own purposes but no one of them is just telling it the way
things are. The three allegedly alternative characterizations under discussion
consist in one actual characterization and two pseudo-characterizations; and,
of course, if this is so, we can, and should, say the one genuine one is more
adequate. But that is not at all to say that it gestures at ‘the one true
description of the world’. There is no such thing.
Some (including Wittgenstein) might deny that ‘It’s God’s will’ or ‘It’s the
work of the Devil’ are metaphysical utterances. If ‘God’ and ‘the Devil’ are
taken to denote Zeus-like entities, then these utterances are not metaphysical.
They are implicit, very vague, empirical hypotheses. They are, that is, just
crude, plainly false, empirical propositions plainly disconfirmed. Such
religious beliefs are superstitions, and Wittgenstein was keenly aware of that
and rejected such religious beliefs and such a way of looking at religion with
disdain. But it is very unlikely that many Christians, Jews or Moslems so
superstitiously conceive of God and the Devil. Indeed by now most of them
do not. And where they do, Wittgenstein would have no sympathy at all with
that. Where, alternatively, ‘God’ is construed as an infinite individual
transcendent to the universe, we plainly do have a metaphysical claim—and
a very esoteric one at that—and as such it is held to be nonsensical not just by
positivists but by Wittgensteinians and neo-pragmatists such as Putnam,
Rorty and myself.63 If to that it is said that is not how to construe ‘God’
either, then it is difficult to know, unless we want to go back to the crude
anthropomorphic construal or to a purely symbolic construal, how we are in
some non-metaphysical way to construe ‘God’. Just what is this non-Zeus-
like, non-purely symbolic, non-metaphysical construal of ‘God’? Do we really
have any understanding of what we are talking about here?
If instead it is said ‘That’s how things are’ is itself a metaphysical
statement, this should be denied, for it functions as a summarizing, somewhat
moralizingly emotive, proposition standing in for (a) a lot more particular
propositions such as people suffer, the wicked often flourish, starvation and
malnutrition are pervasive, droughts and devastating earthquakes occur,
people are struck down in their prime, alienation is pervasive, tyranny often
goes unchecked, and the like, and (b) the comment that this goes on at all
times and in all places without much in the way of abatement. This—(b) in
particular—may be an exaggeration, but that surely does not make it a
metaphysical statement.
Suppose someone retorts that Jews and Christians do not have to treat
‘That’s God’s will’ or ‘That’s the Devil’s work’ in either the superstitious or
the metaphysical way I attributed to them. Keep in mind, the response goes,
that practice gives words their sense. Some mathematicians, when they speak
of numbers, say they are abstract entities: real things but abstract things. And
with this they become entangled in metaphysics. Indeed we have the shadow
of Plato here. But they could, and most do, legitimately refuse to so theorize
160 Kai Nielsen

and just go on proving theorems, setting up axiomatic systems or, as applied

mathematicians, grinding out calculations for particular purposes and the
like. Why cannot Jews, Christians and Moslems do a similar thing? Why
could they not, and indeed why should they not, just stick with their practices
in saying and thinking the things about God that their language-games allow
them to say? They need no more theorize about God than mathematicians
need theorize about numbers. Indeed it is not only that they need not theorize,
Wittgenstein and Kierkegaard would insist, but that they should not theorize.
That is destructive of faith. It is my intellectualism again—and here I am a
token of a familiar type—that is leading me down the garden path: that is
making me mistakenly think that practices which actually are not
unreasonable—indeed, are compelling for the people who engage in them—
are unreasonable and irrational.
It should be responded in turn that there are at least two disanalogies
between the language-games of mathematics and the language-games of
religion. First, we know, without any meta-mathematics, without any theory
of numbers, at least if we are mathematicians, how to establish truth-claims,
or at least assertability claims, in mathematics. Mathematics is a theory that
structures practice, and mathematicians in doing mathematics cannot but use
theory and in that way theorize. We need not theorize about mathematics,
but we, not infrequently, theorize, often to good effect, with it. Second, we
have in mathematics some ability to say what we are talking about. We often
talk nonsense in talking about mathematics but not always. But actual
mathematical talk is another matter. We have no such ability with our talk of
God, the Devil or the soul. It is not just the meta-talk that is troubling.
Suppose it is in turn responded that this only shows some of the differences
between the language-games of mathematics and those of religion. We
understand, if we are religious, that God is a mystery and—or so
Wittgensteinians have it—that the very demand to be able to warrant our
religious claims shows we do not understand them or understand what faith
requires, including what it is to believe in God. Anything that we could
warrant—establish the truth of—wouldn’t be a genuine religious claim. To
make such a rationalistic demand shows, Wittgenstein et al. would have it,
that we do not understand religious language-games and that we are not
operating from inside them. It would be like in logic to demand that an
inductive argument be deductively valid. It would show that we understand
neither what induction is nor what logic is. We are just senselessly asking for
induction to be deduction.
Still, if this is what religious language-games are like, would it not be better
not to engage in them? We do not know what counts for truth or falsity or what
counts as reasonable or unreasonable here; indeed we do not even understand
what we are saying. We are just in a fog. Nonsense engulfs us. Isn’t talk of mystery
just a high-fallutin’ way of saying that? Once we see this clearly should we not
desist—close up shop, so to say? Moreover, it is not just that we do not
understand: we are forced, if we would play that language-game, to say things
Wittgenstein and Wittgensteinians on religion 161

that we, if we reflect a bit, would not wish to say. Consider again Wittgenstein’s
remark in Culture and Value that we ‘might speak of the world as malicious’ or
‘easily imagine the Devil created the world, or part of it’ or that ‘the whole scheme
of things will be aimed at evil from the very start’. We not only cannot (pace
Wittgenstein) easily imagine these things: we do not understand these utterances.
We only, if we do not think, have the illusion of understanding them by extension
from some familiar utterances we do understand. We understand what it is for a
person to be malicious or an action or attitude to be malicious. We have truth-
conditions or assertability conditions for such claims. But for the world to be
malicious? We can’t intelligibly impute intentions to the world. That makes no
sense at all. Speaking of the world being malicious is but a misleading way of
making the perfectly secular utterance: ‘Many people are malicious and this
maliciousness is pervasive in our lives.’ Similarly, while we understand ‘Sven
created a new recipe’ or ‘Jane created a more efficient electric car’, we do not
understand ‘The Devil created the world’ or, for that matter, ‘God created the
world’. The former two sentences have truth-conditions or assertability
conditions. The latter two do not. Similarly language has gone on a holiday with
the claim: ‘The whole scheme of things will be aimed at evil from the very start.’
Aside from not understanding what ‘the very start’ comes to here, more
importantly we are, with such a remark, again imputing intentions and aims to
what it makes no sense to say has or can have intentions or aims. To say
Shakespeare’s Richard III aimed at evil or the Nazi regime or the Reagan regime
aimed at evil makes sense, but neither the whole scheme of things nor the world
can be intelligibly said to aim at things either for good or for evil. A scheme of
things or a world cannot have aims, form intentions, have desires, goals, and the
like. There is and can be no such teleology of nature. There is no such functional
language-game. Language is idle here. In support of this, I have supplied what
Wittgenstein has called grammatical remarks. But would not Wittgenstein, of all
people, perfectly well realize that? That is the way he repeatedly reasons. And
the grammatical remarks I have assembled above seem to be plainly so. It looks
like Wittgenstein is in a double bind.
Of course Wittgenstein is right, as he says in a sentence following the one
quoted above, that ‘things break, slide about, cause every imaginable
mischief. But that, minimally hyperbolic though it is, is a purely secular
utterance. We have not even the hint of a religious language-game here. If
that is what we ‘really are saying in saying that the whole scheme of things
will be aimed at evil’, we have turned it, by stipulative redefinition, into an
utterly secular platitude without a whiff of religion or religious sensibility.
Where we understand what we are saying we do not have a religious
language-game at all; where we have one we do not—the superstitious
anthropomorphic ones aside—understand what is said and thus cannot
understand what it is for something to be, for example, God’s will, and thus
we cannot do God’s will or fail to do God’s will.
Suppose someone says that that is a philosopher’s hat trick. People do
God’s will. People, following God’s will, make pilgrimages to Lourdes, go to
162 Kai Nielsen

confession, give up philosophy, lead a life of celibacy, go to the Congo or

Haiti to alleviate suffering, etc. But to this, it in turn can be responded, that
this—this doing of God’s will—is but to do things that some people take to
be obligatory, the right thing to do, desirable to do, and the like, and that
some of them associate these moral commitments with their avowals that
that is doing God’s will without understanding what God is or what His will
is or how one could ascertain what is or what it is to do God’s will. It is just
a formula they recite with, if they are genuinely theistically religious, great
conviction and sometimes with intensity of feeling. But that does not, and
cannot, turn it into sense: into an intelligible utterance.
Your intellectualism continues to get in the way, some will respond or at
least think. The aim in speaking of religion as Kierkegaard and Wittgenstein
do is to expose the roots of the intellectual’s compulsion in approaching
religion ‘always to reflect upon the task of living (a certain sort of life) rather
than to attend to the task itself.64 The thing to do is to go to church, to pray,
to confess, to sing songs in praise of God, to alter your life by becoming more
open and loving, to fight against your arrogance and pride, and above all
help your fellow humans by engaging with them in their life struggles. Don’t
think, act! Thinking will never lead you to faith. To think that it can is a
grand illusion of much of the philosophy of religion business. Philosophy will
not lead us to God or help us in our religious endeavours.
There is, both Wittgenstein and Kierkegaard have it, no summing up ‘the
sense of a religion in philosophical or theological doctrines’.65 Kierkegaard
stressed that religious belief stands at a very great distance from philosophical
clarity. Such clarity is of no avail in coming to a religious life or, for that
matter, in turning away from it and combating it. Wittgenstein, as we have
seen, had scant use for religious doctrine, theology or the philosophy of
religion. He took it to be one of the things that Christianity teaches us that
even sound doctrines are useless. The thing is to change your life or the
direction of your life. Even achieving wisdom, if indeed we could do this, is
of little value. Wisdom is cold and does not connect to your passions, does
not grip your life, as religion does by taking hold of you and turning your life
around (CV, p. 53). Wittgenstein, in a deeply anti-intellectualistic way, wrote
that ‘Wisdom is cold and to that extent stupid. (Faith, on the other hand, is a
passion.) It might also be said: Wisdom merely conceals life from you.
Wisdom is like grey ash, covering up the glowing embers’ (CV, p. 56).
Religious faith is a passion yielding a trust that grips your life and turns it
around. Trying to be intellectual about religion—trying to rationalize
religion—will never get you anywhere. People who are gripped by religious
forms of life will not try to show how the religious life is reasonable, though
they need not say it is unreasonable either. They will see all argument and
attempts at reasoning here, on the part either of the believer or of the sceptic,
as utterly pointless.

Is this the end of the line? Should we, vis-à-vis religion, take some such anti-
Wittgenstein and Wittgensteinians on religion 163

intellectualist stance and claim that philosophical thinking, or any kind of

thinking, only stands in the way of coming to grips with religion whether by
way of faith or by rejection of religious faith?
I think not. For one thing, we should be more holist than Wittgenstein or
most Wittgensteinians are prepared to be. We should not take distinct language-
games to be autonomous, yielding their own wholly distinctive criteria of what
it makes sense to say, what is justified, what is acceptable and the like. We need
repeatedly to attend to how our various language-games and practices relate,
criss-cross and affect each other, or would affect each other if we saw with any
clarity how they are related. Though no doubt, without any clarifying
articulation, these different practices just do affect each other. But with a clearer
understanding of how they relate and affect—or could affect—each other, we
may gain a more adequate understanding of how things hang together and of
the import of it. This may not happen, but it is not impossible that it could and
it is worth struggling to attain. Here there should be no quietism. Such a
struggle is both reasonable and worth the candle.
If we look at our religious practices, including those containing rather well-
firmed-up secular knowledge-claims, we can come to see without any theory
at all that certain religious notions make such a bad fit with other things that
are very pervasive in our culture and important to us that, in coming to
understand this, we will come to see that there is very little, if any, sense in
these religious notions. It is certainly right to tell us not to be so sure of
ourselves and to look again to see if we are being blind to a fit that is there
before our eyes which we simply do not see.66 (Perhaps we are ideologically
blinkered here.) Wittgenstein has shown, regarding the language-games he
agonizes over, how often this is the case. But it is also possible that there is no
fit—just clashing irreconcilable beliefs (sometimes just attempts at belief) and
conceptions—or that a better fit can be made of the various things we know,
reasonably believe and care about, by jettisoning religious beliefs and
practices: setting them aside so that they—though no doubt this takes time—
will no longer play a part in our thought and behaviour and in our conception
of how we should guide our lives. It may be the case that there is a severe
strain, and indeed even a clash, between different elements in our forms of
life and that the religious element will, if we really press things with integrity,
be the odd man out. It may be that, in the attempt to overcome the tension by
making our religious beliefs and conceptions fit with the rest of our beliefs
and conceptions, we will have to resort to increasingly ad hoc assumptions
or esoteric readings of our religious beliefs and conceptions. It seems to me
that something like this is actually happening, and indeed has been happening
for some considerable time.67
Holistic description also serves here as criticism. Philosophy, little ‘p’
philosophy, utilizing the method of wide reflective equilibrium, need not, and
sometimes should not, leave everything as it is. A critical philosophy will
utilize Wittgensteinian elucidation principally to break the picture that certain
philosophical conceptions seek to impose on us, but it will also engage in
164 Kai Nielsen

critical assessments, engaging with our lives as well as just with our
cogitations—critical assessments that pass without metaphysical
extravagance or any other kind of extravagance beyond Wittgensteinian
philosophical quietism and neutralism. This is done without trying to have
some ‘ultimate vocabulary’ or some ultimate point of reference or claiming
that there is one and only one true description capturing how the world just
is anyway. Indeed such talk makes no more sense than William James’ talk of
an ‘ultimate datum’.
Wittgenstein shows us well the incoherence of such conceptions. But we
have seen how we can, and sometimes should, criticize practices, and not just
stop with the reminder that this language-game is played. But our criticism
will itself rest on other practices. There is no Archimedean point, independent
of all practices, from which to criticize any of them. But from this—to make
a good Peircean point—it does not follow that any practice is immune from
or beyond criticism. We can’t criticize them all at once or stand free of all of
them and criticize them all at once. But where there is a clash among the
practices or where the irritation of doubt is at work—real live Peircean doubt,
not what Peirce well called Cartesian paper doubt—concerning any one, or
several, of our practices, criticism is possible and in order. So we can see how
a pragmatist need not, and should not, acquiesce in quietism. And we can see
also how we can be pragmatists and consistently say that the Christian faith
or any other faith or any set of beliefs and responses embedded in practices
can rest on a mistake or (pace Putnam) be in deep and massive error.68 And
this holds true not only for religious forms of life, but for any practice or
form of life. We start with practices, and it is important to see that and how
many of them are crucial for our understanding and our lives and are
irreplaceable. There is no place else for us to be than to start with practices
and to remain with practices. Moreover, taking them together, we are stuck
with them. There is no perspective outside of or beyond our practices as a
whole. There is, that is, no leaping out of our skins. But for any one or several
or particular clusters of practices, where for specific reasons we come to have
trouble with some specific practice or specific cluster of practices, it or they
can either be reformed (sometimes deeply reformed) or sometimes even set
aside. There is, to repeat, no practice which is immune from criticism. And
the same is true, at least in principle, of clusters of particular practices. So we
can repeatedly, relevantly and intelligently criticize our very practices and
the beliefs and attitudes that are a part of them. This includes our faiths—
that is, our trustings. It is just that (1) we cannot criticize them all at once or
stand free of all of our practices, and (2) that in criticizing a practice or a
cluster of practices we must also be using practices. Thus we have Peircean
fallibilism and Peircean critical common-sensism—something that was fully
incorporated into the texture of Dewey’s philosophical practice.69 With this,
and without falling into philosophy and the conceptual confusions
Wittgenstein was concerned to dispel, we can do something critical
concerning our forms of life. We can reasonably engage in an activity here
Wittgenstein and Wittgensteinians on religion 165

for which Wittgenstein did not make space and indeed did not envisage. With
his feeling for a religious sense of life he would probably have thought it all
hubris. But need it be?

1 Rush Rhees (ed.), Ludwig Wittgenstein: Personal Recollections (Oxford: Oxford
University Press, 1984), p. 79.
2 See Richard Braithwaite. ‘An Empiricist’s View of the Nature of Religious Belief’,
in Malcolm L.Diamond and Thomas V.Litzenberg (eds). The Logic of God
(Indianapolis, IN: Bobbs-Merrill, 1975), pp. 127–47; and R.M.Hare, ‘The
Simple Believer’, in Gene Outka and John R.Reeder. Jr. (eds). Religion and
Morality (New York: Anchor Press, 1973), pp. 393–427.
3 John Hyman, ‘Wittgensteinianism’, in Phillip L.Quinn and Charles Taliaferro
(eds). A Companion to Philosophy of Religion (Oxford: Blackwell. 1997). pp.
4 P.M.S.Hacker, ‘Wittgenstein’, in Robert L.Arrington (ed.), A Companion to the
Philosophers (Oxford: Blackwell, 1999), p. 538.
5 Ibid.
6 Norman Malcolm, Wittgenstein: A Religious Point of View?, ed. with a Response
by Peter Winch (Ithaca: Cornell University Press, 1994).
7 Hacker, op. cit., pp. 545–6.
8 Ibid. 546.
9 Ibid.
10 Kai Nielsen, Introduction to the Philosophy of Religion (London: Macmillan,
1982), pp. 43–64.
11 N.Malcolm, Wittgenstein: A Religious Point of View?, p. 74.
12 Ibid.
13 Ibid., p. 75.
14 Ibid., p. 76.
15 Ibid., first italics added.
16 Ibid., pp. 75–7.
17 Ibid., p. 76.
18 Ibid., p. 77, italics added.
19 Ibid., p. 78.
20 See Ibid., pp. 79–80.
21 Ibid., p. 81.
22 Ibid., p. 82.
23 Ibid.
24 Ibid.
25 Ibid.
26 See Axel Hägerström, Philosophy and Religion, trans. Robert T.Sandin (London:
Allen &Unwin, 1964).
27 See N.Malcolm, op. cit., p. 10.
28 Ibid., p. 17.
29 Ibid., p. 11.
30 Ibid., p. 32.
31 Ibid., p. 17.
32 See N.Malcolm, ibid.
33 Ibid., p. 19.
34 Ibid., p. 10.
35 Ibid., p. 21.
166 Kai Nielsen

36 See K.Nielsen, God, Scepticism and Morality (Ottawa: Ottawa University Press,
1989); and K.Nielsen, ‘Atheism Without Anger or Tears’, in Hendrik Hart et al.
(eds), Walking the Tightrope of Faith (Amsterdam: Rodolpi, 1999), 80–127.
37 Peter Winch, ‘Discussion of Malcolm’s Essay’, in Wittgenstein: A Religious Point
of View?, p. 133.
38 See Winch, Ibid., p. 100.
39 Ibid.
40 Ibid., pp. 104–5.
41 Ibid., p. 106.
42 Ibid.
43 Ibid., p. 104.
44 Ibid.
45 Ibid., p. 111.
46 N.Malcolm, op. cit., p. 84.
47 Ibid., p. 92.
48 P.Winch, op. cit., p. 121.
49 Ibid.
50 N.Malcolm, op. cit.. Chapter 1.
51 P.Winch, op. cit., p. 121.
52 Ibid., p. 124.
53 Ibid., p. 108.
54 Ibid., p. 109.
55 Ibid.
56 Ibid., p. 110.
57 Ibid.
58 Ibid., p. 113.
59 Ibid., p. 114.
60 Ibid .
61 H.Putnam, Pragmatism (Oxford: Blackwell, 1995), pp. 27–56.
62 See H.Putnam, Renewing Philosophy (Cambridge, MA: Harvard University
Press, 1992), pp. 80–107.
63 See K.Nielsen, Introduction to the Philosophy of Religion; Philosophy and
Atheism (Buffalo, NY: Prometheus Books, 1985); and God, Scepticism and
64 James Conant, ‘Kierkegaard, Wittgenstein and Nonsense’, in T.Cohen et al. (eds),
Pursuits of Reason (Lubbock, TX: Texas Tech University Press, 1993), p. 207.
65 P.Winch, op cit., p. 128.
66 See R.Rhees, Rush Rhees on Religion and Philosophy, ed. D.Z.Phillips
(Cambridge; Cambridge University Press, 1997).
67 See K.Nielsen, Naturalism Without Foundations (Amherst, NY: Prometheus
Press, 1996), pp. 75–155.
68 See H.Putnam, Pragmatism.
69 See K.Nielsen, op. cit., pp. 295–328.
10 ‘Theology as grammar’
Wittgenstein and some critics
Robert L.Arrington

Wittgenstein’s philosophy of religion, as found in his brief remarks on

religious belief and on magic, is as controversial as his philosophy of
mathematics and his philosophy of mind. In fact, many scholars who tend to
follow Wittgenstein in these latter areas are reluctant to accept what he has
to say about religious belief and related topics. Wittgenstein seems to insulate
religion from standard forms of rational criticism, and this is unacceptable to
many philosophers who think they have good reasons for rejecting any form
of theism. Several of the preceding papers in this collection contain
repudiations of Wittgenstein’s philosophy of religion, and most if not all of
them are written by philosophers who are openly sympathetic to many
fundamental aspects of Wittgenstein’s philosophy of language. In this
concluding essay, I wish to examine some of their arguments for thinking
Wittgenstein is wrong in his philosophy of religion. I shall argue that these
arguments are unconvincing, largely because they ignore the implications of
Wittgenstein’s suggestion in Philosophical Investigations (§373) that
theology be construed as an expression of religious grammar.
Kai Nielsen, who is well known for his earlier criticisms of what he called
Wittgensteinian fideism,1 takes yet another look at Wittgenstein on religion.
Expressing an acceptance of what Wittgenstein has to say about the need to
place assertions in the context of language-games and forms of life, and about
the senselessness of speaking of ‘the one true description of the world,’
Nielsen nevertheless finds fault with Wittgenstein’s defense of religious belief
and with his quietism, the latter being Wittgenstein’s view that philosophy
can only describe language and cannot in any way mount a criticism of
language-games. Nielsen thinks he detects an inconsistency in Wittgenstein’s
remarks on religion, and he believes that a certain form of philosophical
criticism is possible with respect to the religious form of life.
First, the alleged inconsistency. According to Nielsen, Wittgenstein thinks
that all metaphysical beliefs are ‘incoherent,’ that they are merely ‘houses of
cards.’ This is one of the areas in which Nielsen professes to agree with
Wittgenstein. However, Nielsen claims, religion in part is inescapably

168 Robert L.Arrington

metaphysical, i.e. inextricably intertwined with metaphysics. ‘There are no

doctrineless or creedless religions,’ he writes (p. 147). These doctrines and
creeds Nielsen takes to be metaphysical claims or to incorporate such claims.
Therefore, he asks, shouldn’t Wittgenstein conclude that religious belief,
containing as it does these unacceptable metaphysical doctrines and creeds,
is incoherent? As Nielsen sees it, Wittgenstein escapes the necessity of
drawing this conclusion only by claiming that the doctrinal elements of
religion are not really important and should be set aside—that what is really
important is changing one’s life and doing good deeds. Nielsen emphasizes
that for Wittgenstein life-orientation, not the speculative side of religion, is
what is important. The language-game of religion is played, but this means
only that people commit themselves to a life of doing good deeds. Hence,
Nielsen takes Wittgenstein to say, Christianity is not a doctrine, and to be a
Christian is hence not to be hobbled by incoherent metaphysical beliefs.
According to Nielsen, Wittgenstein’s escape from the charge that the
religious life is incoherent amounts to nothing more than victory by
stipulative redefinition, stipulating that being committed to doing good deeds
is sufficient for being religious. This would mean that many secular and
atheistic people are religious, and this implication amounts to a reductio of
Wittgenstein’s position. Moreover, Nielsen claims, it is just wrong to think
that being committed to doing good deeds is sufficient for being religious,
since the religious person must interpret these deeds in terms of religious
doctrine. So the theoretical metaphysical element comes back into play—
doctrine is integral to the religious form of life. But this fact entails, on
Wittgenstein’s own assumption that metaphysical utterances are senseless,
that the religious way of life is incoherent. According to Nielsen,
Wittgenstein, given his repudiation of metaphysics, should agree with this,
but of course does not. So much the worse for Wittgenstein.
Let us now look at the way Nielsen proposes to criticize the religious form
of life. This form of life incorporates the idea that whatever happens in the
world is the result of God’s creative act and will—his providence, one might
say. Many people see the world this way, and, Nielsen notes, Wittgenstein
even thinks it is possible to imagine that the events of the world—especially
the tragedies, the disasters, and the suffering—are the doings of a malicious
being. Nielsen himself, however, looks at the devastation and the pain that
constantly occur in the world and proclaims ‘That’s just the way things are.’
He takes this latter remark to be an empirical generalization about the
occurrence of hardship and suffering in human life—and an emotive response
to such empirical facts. Moreover, he thinks ‘That’s just the way things are’ is
a better way of viewing things than to say ‘That’s God’s will’ or ‘The world
is malicious.’ Nielsen’s reason for this claim is that ‘That’s just how things
are’ (as interpreted by him) is a meaningful empirical generalization (even if
an exaggeration), while the other two ‘claims’ are metaphysical ones which
are senseless. It is not that ‘That’s how things are’ is the one true description
of the world (he agrees with Wittgenstein that there is no such thing) but
‘Theology as grammar’ 169

rather that it is not just one sensible option among three sensible ones—the
other two are not meaningful. ‘That’s God’s will’ is, according to Nielsen, a
metaphysical fantasy. ‘God created the world,’ he claims, has no truth-
conditions; therefore we can’t understand it.
Nielsen does not give us in his essay a precise picture of the reasons behind
Wittgenstein’s discontent with metaphysics. At times it appears he thinks
Wittgenstein rejects metaphysical claims (such as ‘God is outside space and
time’) on the grounds that they are unverifiable, that they make appeal to
transcendent entities beyond the grasp of experience. At other times he takes
Wittgenstein to say that metaphysical claims (like ‘The world has an aim’)
violate the grammatical rules of language, rules that specify proper usage. With
regard to some metaphysical propositions, Wittgenstein might indeed argue in
these ways (although not, I think, with regard to the two examples Nielsen
mentions). But these accounts miss the distinctive faults that Wittgenstein finds
with metaphysics. In the Tractatus, metaphysical statements are meaningless
because they are abortive attempts to describe elements of logical form, which
can only be shown, not said. In his later philosophy, Wittgenstein claims that
metaphysicians misinterpret their assertions as factual propositions when in
reality they are conceptual ones. ‘Philosophical investigations: conceptual
investigations. The essential thing about metaphysics: it obliterates the
distinction between factual and conceptual investigations’ (Z §458). Many of
the houses of cards Wittgenstein wishes to destroy (PI §118) are metaphysical
statements which are confused because the metaphysician attributes to them
the properties of both factual and conceptual propositions. They are viewed by
the metaphysician as being about the world (factual) and also as necessarily
true (conceptual).2
Nielsen is right to this extent: in his later philosophy Wittgenstein does
continue to regard metaphysical statements as senseless. Take the realist
metaphysical claim that there are physical objects. “There are physical
objects” is nonsense,’ Wittgenstein tells us in On Certainty §35. And he goes
on to ask: ‘Is it supposed to be an empirical proposition?’ The implication
here is that There are physical objects’ is not an empirical assertion—what
empirical evidence could prove or disprove it? So far, what Wittgenstein says
is consistent with Nielsen’s interpretation of him. But there is more to be
taken into account. What if someone (perhaps G.E.Moore) tried to prove
that there are physical objects by pointing to an example of one and saying
‘This is a physical object’? Wittgenstein’s reply is noteworthy:

‘A is a physical object’ is a piece of instruction which we give only to

someone who doesn’t yet understand either what ‘A’ means or what
‘physical object’ means. Thus it is instruction about the use of words,
and ‘physical object’ is a logical concept. (Like colour, quantity,…) And
that is why no such proposition as ‘There are physical objects’ can be
(OC §36)
170 Robert L.Arrington

In denying that the metaphysical statement in question can be formulated

because ‘physical object’ is a logical concept—what elsewhere he calls a
‘grammatical category’ (BB, p. 19)—Wittgenstein sounds like he is still in the
mood of the Tractatus. But the earlier part of the quotation throws a different
light on things: ‘A is a physical object’ is given a use, but this use is not one of
describing some thing in the world but of instructing someone in the use of a
word. The proposition is not a meaningful description of a fact, but it is
meaningful as a way of expressing a rule about what a word means or how it
is to be used. ‘This is a physical object’ is, therefore, not a factual or empirical
proposition. The philosopher who took the statement as empirical proof that
there are physical objects has confused a conceptual utterance with a factual
Similarly, Wittgenstein argues, ‘Red exists’ is also nonsense if it is taken to
be a factual claim. But ‘Red exists’, might have a use and hence a meaning: it
might be taken to say that ‘red’ has a meaning (PI §58). In other words, it
might be used to license or authorize the use of the term ‘red’ in our ordinary
empirical discourse. The metaphysical statement about red—that it exists—
is misleading because it is not a factual statement about red but a conceptual
one about the word ‘red’ or its use: ‘If “X exists” is intended simply to say:
“X” has a meaning,—then it is not a proposition which treats of X, but a
proposition about our use of language, that is, about the use of the word
“X”’ (PI §58).3
Late in his Remarks on the Foundations of Mathematics, Wittgenstein
makes an explicit comparison between his earlier views concerning
statements about logical form and his later view of such statements as
grammatical propositions. He writes:

What sort of proposition is: ‘The class of lions is not a lion,…’? How is it
verified? How could it be used?—So far as I can see, only as a
grammatical proposition. To draw someone’s attention to the fact that
the word ‘lion’ is used in a fundamentally different way from the name
of a lion….

At this point, he makes the contrast with the Tractarian conception of the
sentence in question:

Even though ‘the class of lions is not a lion’ seems like nonsense, to which
one can only ascribe a sense out of politeness; still I do not want to take it
like that, but as a proper sentence, if only it is taken right. (And so not as
in the Tractatus.) Thus my conception is a different one here. Now this
means that I am saying: there is a language-game with this sentence too.
(RFM, p. 182)

The language-game involving the controversial sentence is one in which it is

used as what Wittgenstein now calls a ‘grammatical’ proposition or
‘Theology as grammar’ 171

statement. Such statements are used to explain the meaning of certain terms.
The sentence about ‘class’ employed in this fashion is perfectly all right. It is
only when one attempts to employ it with a metaphysical emphasis, namely
to describe a fact about the essence of ‘classes’, that one goes wrong.
We might make a similar suggestion regarding ‘There are physical objects.’
This proposition might be used, not to make a factual, empirical claim, but
to say (very misleadingly) that ‘physical object’ has a use and meaning in our
language. ‘There are physical objects’ is not a sensible claim if it is taken to
be a factual one. If it were a factual claim, we could specify the conditions in
experience that, if they occurred, would show that the claim is true, and we
could specify the conditions in experience that, if they occurred, would show
us that it is false. But the debate between the idealist and the materialist (or
Cartesian dualist) is not of this sort. An idealist, like Berkeley, claims that it is
senseless to speak of matter, and hence of material or physical objects. We
have, Berkeley asserts, no idea of matter. Berkeley does not argue that it is
simply factually false to claim that there are physical objects; he claims that
all talk of physical objects and matter is meaningless. Berkeley’s idealism,
then, consists in a recommendation about which concepts make sense: he
rules out the concept of a physical object and uses instead the concept of an
idea in the mind. Of course, this isn’t exactly the way Berkeley saw it. He
thought he was telling us something about the world—i.e., making a factual
claim about reality (esse est percipi)—even while his arguments clearly show
that it is a conceptual one he is making. Hence Wittgenstein’s diagnosis that
the metaphysician (in this case Berkeley) is confusing factual and conceptual
issues. Likewise, the realist or materialist who claims that there are physical
objects thinks of himself as giving us a truth about the world. But what his
claim amounts to—if we want to say that it amounts to anything—is a
comment authorizing the use of physical object terminology. There are
physical objects’ really means something like ‘It makes sense to speak of
objects having place, motion, weight, etc.’ The materialist is simply confused
in taking ‘There are physical objects’ as a factual assertion rather than a
conceptual rule licensing a certain kind of talk.
As we have seen, in his later writings Wittgenstein calls such conceptual
rules ‘grammatical statements’ or ‘grammatical remarks.’ There is nothing
wrong with metaphysical statements, I understand him to say, if they are
taken for what they really are, namely, grammatical remarks. These
grammatical remarks don’t tell us anything about the world—rather, they
provide us with the ‘means of representation’ that we are to employ in talking
about the world. These means of representation are rules for the proper uses
of words. Using words in accordance with such rules, we are able to make
judgments that can be assessed as true or false, or as acceptable/unacceptable
in some other way. The normativity involved in any language is the result of
the fact that the grammar of the language authorizes the use of some
utterances and not others, and at the same time authorizes the mode of
assessment (true or false, acceptable or unacceptable) appropriate to this
172 Robert L.Arrington

form of discourse. ‘There are colors,’ for example, is a grammatical remark

authorizing the use of color terms (‘red’, ‘green’, etc.) and the assessment of
color judgments as true or false. Essence—that quintessential metaphysical
notion—is expressed by grammar, Wittgenstein writes in PI (§371).
Metaphysical statements about essence—that which exists in its own right—
are disguised grammatical propositions.
In the Philosophical Investigations, Wittgenstein, after telling us in PI §373
that ‘Grammar tells us what kind of object anything is,’ makes the famous
parenthetical remark: ‘(Theology as grammar)’. And with apparent approval
he describes Luther as saying that ‘theology is the grammar of the word
“God”’ (AWL, p. 32). The last remark in Zettel is a bit of theological
grammar: ‘“You can’t hear God speak to someone else, you can hear him
only if you are being addressed”.—That is a grammatical remark’ (Z §717).
What these comments suggest is that the standard assertions of the
theologian—‘God exists,’ ‘God is omnipotent, omniscient, and eternal,’ ‘God
is love’—are to be understood not as factual claims about a being in the
world, a being called God, but as grammatical remarks expressing rules for
the use of theological terms in everyday religious discourse. These rules
constitute or create the language of religious belief. One is simply not talking
of God if there is any question about his existence, if, that is to say, it makes
sense to wonder whether he exists. For the religious believer, God, by
definition, exists. Likewise, the religious believer’s concept of God (in
Christianity, Judaism, and Islam) is such that God is eternal, omniscient, and
omnipotent. ‘God is omnipotent, omniscient, and eternal’ is therefore a
conceptual claim, or better, a rule defining the concept of God. There is
nothing wrong with theology, precisely because it consists of a set of rules for
the proper employment of religious terms. Things would go wrong if the
theologian thought he was making true factual claims about God. Doing so
would turn him into a metaphysician who had confused grammatical and
factual investigations. And there would be nothing wrong with atheology if
the atheologian who says ‘God does not exist’ were simply ruling out as
meaningless all talk of God. But the atheologian would be making a confused
metaphysical claim if he thought he was giving us a truth about the world: in
fact there is no God. He would be confusing grammatical and factual
Nielsen, I think, does not take into account this aspect of Wittgenstein’s
thought. He tells us that religious discourse, in its doctrinal aspect, is
meaningless, but he does not take the next step and acknowledge that the
senseless metaphysical proposition ‘God exists’ hides a grammatical
proposition which is far from meaningless. Thus he does not acknowledge
that his own metaphysical assertion, ‘There is no God,’ is in effect an
atheological grammatical proposition: ‘The term “God” has no use’. To
assert the latter grammatical statement is simply to refuse to speak a certain
language—to participate in the religious language-game. That atheists (or
some of them) subscribe to this grammatical proposition tells us something
‘Theology as grammar’ 173

about atheists—that for them religious discourse makes no sense—but it tells

us nothing about the world at large, nothing about the facts or about reality.
And it doesn’t in any sense constitute a refutation of the religious believer’s
way of talking. The religious believer follows the rules that have been
articulated by the theologian. ‘God exists’ for the religious believer is just a
way (highly misleading, to be sure) of authorizing talk of the divine, or, more
specifically, of indicating that to use religious language one must talk of God
as existing (and not as possibly existing and possibly not existing).
Wittgenstein does not accept or reject this way of talking; he simply wants to
ensure that it is not misunderstood—misunderstood as a factual claim when
in reality it is a conceptual, grammatical one.
Now the grammatical statements coming from theology authorize and
constrain a certain way of talking—the religious way. It does not follow from
this that the kind of talk religious believers engage in will always or even
generally be ‘factual’ or ‘descriptive’ in nature. We can see this by reflecting
on grammar and its application in the realm of physical object discourse. The
grammatical rules governing the use of physical object terminology often
constrain factual utterances about particular physical objects—as well as
empirical generalizations about them of the sort found in science. But they
also operate in other areas of discourse, for example, prescriptive discourse
as in ‘Bring me the red book on my desk.’ Such requests are governed by the
grammar of physical object talk no less than are physical descriptions (e.g.
‘There is a red book on my desk’). The order to bring me the book on my
desk must be understood in terms of the criterion of identity for a physical
thing, in this case a book. ‘Bring me the book on my desk’ is not a request or
order that someone bring me a similar object, but that this person bring me
the identical one (the numerically identical one). This criterion of identity for
an object is part of the grammar of physical object talk.
Likewise, the fact that theological statements are grammatical ones does
not entail that all religious statements are factual, descriptive ones. Some may
be—for example, ‘I am a sinner’ and ‘God loves me,’ and ‘God had a purpose
in taking this child’4—but others, many others, will be prescriptive in nature,
giving one edicts for how to live one’s life. In locating such utterances as
central to the religious life, Wittgenstein is simply saying that ‘living a certain
kind of life’ is at the heart of religious discourse and action—not an
investigation into or speculation about the nature of the world.
According to Wittgenstein, theological statements are grammatical rules
guiding religious action and feeling—as well as guiding occasional descriptive
claims about particular persons or events. He thinks the most egregious error
of the philosopher would be to take theological propositions as factual claims
about the world. But this is exactly how Nielsen interprets them. He takes
them to be factual claims that, unfortunately, appeal to or make reference to
transcendent realities—and therefore go beyond the bounds of empirical
evidence—or that violate the grammatical rules for non-religious naturalistic
discourse. Hence, for Nielsen, these claims are meaningless. If we apply
174 Robert L.Arrington

Wittgenstein’s factual-grammatical distinction to Nielsen’s own claims, we

see that he is simply ruling out one way of speaking, the religious way, and
authorizing another, the naturalistic way. But that Nielsen speaks this way,
and wants everyone else to do so as well, in no way shows that everyone
must speak this way. Nielsen simply won’t guide his discourse in terms of the
grammatical remarks emanating from theology. But other people do, and
quite successfully so.
It might be objected here that a philosopher like Nielsen can demonstrate
that one way of speaking—one grammar—is superior to another.
Wittgenstein responds to such a move by declaring the autonomy of language
and the arbitrariness of grammar.5 Grammar, he thinks, is not accountable to
reality, and any attempt to describe experience or reality in a way that could
justify or refute a particular grammar will always beg the question in favor
of the grammar used in the evidentiary description.6
Next, let us turn to the criticisms that Alan Bailey and John Hyman make
of Wittgenstein’s philosophy of religion. Several of these criticisms will find a
sympathetic response among many readers, even those who are inclined to
follow Wittgenstein in other areas of philosophical inquiry. Bailey explicitly
and forcefully rejects Wittgenstein’s treatment of religious belief. He thinks
that Wittgenstein presents us with a ‘misleading account of the meaning of
religious statements’ (p. 134) or at least many of them. Wittgenstein’s
interpretations, according to Bailey, possess much implausibility. In fact.
Bailey claims, the ‘overwhelming majority’ of religious believers mean by
their religious utterances precisely what Wittgenstein denies they mean (p.
135). Religious faith, be it of the Christian, Jewish, or Islamic variety, contains
‘a belief in the existence of a causally efficacious divine being’ (p. 136), and it
is this belief, Bailey thinks, that Wittgenstein falsely denies the religious
believer has. After all, Bailey argues, Anselm, Aquinas, and Paley can’t
reasonably be accused of misunderstanding the meaning of religious
assertions, and they clearly try to prove that the belief in a causally efficacious
divine being is true. But, because their arguments are weak, and because the
evidence for such a being is scant, Bailey concludes that religious belief is
unreasonable. He writes: ‘Any person who is capable of making competent
judgements about the humdrum causal mechanisms that operate in his or her
immediate environment is potentially someone who can arrive at an
appropriate determination of the rationality or irrationality of the kind of
religious faith embraced by most religious believers’ (p. 136). The
‘appropriate determination’ Bailey has in mind seems to be of the negative
variety. In other words, anyone who can assess whether a can opener works
or whether his car battery is dead can see that the claim that there is a causally
efficacious divine being is unreasonable!
Similarly, Hyman thinks that religious belief needs to be brought to
account for the existential claims it makes, both of a metaphysical and a
historical variety. Hyman finds it difficult to accept what he sees as
Wittgenstein’s claim that religious belief is nothing more than ‘a passionate
‘Theology as grammar’ 175

commitment to a system of reference.’ Hyman takes this kind of commitment

to be a commitment to a way of life, to leading a life in which questions will
be asked, obligations acknowledged, decisions taken, and actions performed,
a way of life which can be explained or understood only by the use of religious
concepts. But this can’t be all there is to religious belief, he counters, because
the belief that God exists is one of the reasons a person has for making and
retaining this commitment.
Hyman grants to Wittgenstein that the belief in God’s existence is different
from the existential propositions professed in science and history. It does not
follow, he argues, that it is not an existential proposition at all. It may simply
be a distinctive kind of existential proposition. The arguments and evidence
that can be offered in its favor are not appropriate to the propositions of, say,
organic chemistry or ancient history. But, ‘Evidence and argument are not
the exclusive property of science,’ he argues (p. 8). Hyman alludes to the
sorts of proof one finds in Anselm and Aquinas—these, presumably, are
appropriate with regard to the peculiar existential belief that God exists. ‘God
exists,’ he concludes, may be a different kind of existential proposition from
those found in science and mathematics (which themselves operate with
different kinds of proof), but for all that it is still an existential proposition.
And thus (p. 8) he rejects Wittgenstein’s claim that, considering God’s
existence, ‘what is here at issue is not the existence of something’ (CV, p. 82).
Hyman thinks it is incorrect to claim that religious faith is not a matter of
assenting to doctrines. Both historical and metaphysical propositions form a
part of religious belief, he claims, and some of these beliefs may turn out to
be false (p. 9). Indeed, some of them may turn out to be incoherent, since
logical scrutiny is as possible with regard to religious beliefs as to any other
kind. A variety of forms of rational scrutiny, he concludes, are appropriate in
the context of religious belief.
Bailey claims that Wittgenstein rejects the idea that the religious person’s
‘God exists and is my creator’ is the expression of a belief in the existence of
a causally efficacious being. But Bailey should be warned by Wittgenstein’s
assertion, which he quotes (p. 125), ‘In a religious discourse we use such
expressions as: “I believe that so and so will happen”, and use them
differently to the way in which we use them in science’ (LC, p. 57). Yes,
Wittgenstein could say, the religious utterance ‘God exists and is my creator’
expresses a belief, but it is not the belief in a contingent causal connection
between God the creator and myself as his creature. It is not the kind of belief
that requires evidential support of the kind we demand for beliefs in the
reality of contingent causal connections. Likewise, ‘God exists’ does not
express a belief in the contingent reality of God, a belief that as a matter of
fact this being exists. But for Wittgenstein it nevertheless expresses a belief—
in another sense of ‘belief.’ (And ‘God exists,’ as Hyman claims, may be an
existential proposition without being one used to express a belief in the
contingent existence of God or a belief of the kind needing proof.) The other
possibility is this: the belief in the existence of God as a causally efficacious
176 Robert L.Arrington

divine being is the acceptance of the grammatical statement ‘God exists and
is my creator.’ Such a statement tells us how we are to talk of God, namely as
an existent being—not as one who may or may not exist—and as a being
who is my creator. To talk of God at all is to talk of such a being, for that is
what the religious believer means by God. So the belief that God exists and is
creator of the world is a distinctive kind of belief—and, of course, if one
follows the later Wittgenstein at all, one should be ready to acknowledge
that a word like ‘belief need not always be used in the same way and hence
may not have the same sense or meaning. Wittgenstein need not deny that
‘God exists and is my creator’ is expressive of a belief in the reality of a divine
and causally efficacious being. It may not be expressive of a certain kind of
belief (expressed in a contingent proposition) but indeed be expressive of a
different kind (expressed in a grammatical proposition).
In his various remarks on religious belief, Wittgenstein stresses the fact
that religious believers do not hold their central beliefs with probability or
well-grounded confidence; they hold them with certainty, ‘unshakably’ as
Wittgenstein puts it. Surely this is the case with ‘God exists and is my creator.’
The religious believer does not think that he has good or convincing evidence
that God is the creator of the world. Belief in a causally efficacious divine
being is not acceptance of an hypothesis, not even an exceptionally
wellgrounded one. It is a belief held unshakably, one totally removed from
the traffic of debate and argument, one that has no uncertainty attached to
it. But it is precisely this type of belief or proposition that Wittgenstein labels
‘grammatical’: ‘To accept a proposition as unshakably certain—I want to
say—means to use it as a grammatical one: this removes uncertainty from it’
(RFM, p. 81). ‘God exists and is my creator’ is, for the religious believer, a
grammatical proposition.
If ‘God exists and is creator of the world’ is a grammatical statement, it is
not a presupposition of religious discourse in the sense of being a proposition
that is merely assumed to be true of the world. It is presupposed as a way of
giving meaning to everything we say about the world: all aspects of nature
and human nature are to be understood in terms of their source in God and
in terms of God’s providential relation to his creatures. The import of this
grammatical rule is to require that all descriptions, decisions, etc., be
formulated or completed in terms of the notion of God’s creative power,
God’s judgments, God’s grace, or God’s love and anger. What is at issue here
is a system of representation, a way of talking and thinking about all things.
And to say that the religious believer is passionately committed to this system
of reference is to say that it stands fast for him, that he takes it as a matter of
course. The passionate aspect of the commitment perhaps suggests that in
taking the religious grammatical system for granted, the religious believer
also knows that many other people (heathens, savages) do not. But these are
the very people who, according to his grammatical scheme, are alienated or
separated from God and who therefore live in sin. Hence the religious
believer’s commitment to his grammatical system is also a commitment to
‘Theology as grammar’ 177

seeing human history as a conflict between goodness and sin, salvation and
damnation—and these are matters provoking the most intense passions.
Nielsen, Hyman and, by implication. Bailey are right to say that religious
belief involves doctrine. But the components of this doctrine are not
contingent propositions requiring support and proof. They are grammatical
ones instructing us in how to engage in religious discourse and the form of
life containing it. Thus the believer who moans ‘I am a thoroughly worthless
wretch’ or ‘I have sinned’ is simply applying to himself the religious
grammatical rules defining human nature. For the wretched one, God’s law
is simply given—this law and the correlated notions of fallen and sinful
human nature are simply ‘in the archives,’ accepted as a matter of course.
Likewise, the Christian view that Jesus rose from the dead is not a disputable
historical claim that might possibly be in error and hence be in need of more
historical defense. The resurrection of Jesus is accepted, as believers often
say, as a matter of faith, not a matter of theory or knowledge. The central
tenet of Christianity—that God became man in Christ—is part of the
conceptual scheme Christians use in talking about all of history: it defines
what history means for the Christian.
Bailey and Hyman think that support for the idea that ‘God’ is treated by
believers as the name of an existent and causally efficacious being comes from
the fact that believers refer to the arguments of philosophers like Anselm,
Aquinas, and Paley both in support of the existence of such a being and in
response to the denials of his existence made by non-believers. If believers try
to prove the existence of God by appeal to the ontological, cosmological and
teleological arguments, this would seem to imply that these believers think
the belief in God’s existence needs proof—and indeed that such proof can be
In response, I think it is fair to say that most of the believers who do appeal
to these arguments are of a philosophical bent, and this rules out most
believers. Wittgenstein would not be impressed with the reactions of the
philosophers and would-be philosophers, since he would see their very
philosophy as a source of distortion and confusion. With regard to religious
discourse, the philosophers themselves, or at least Anselm and Aquinas,
would be held responsible by Wittgenstein for much of the confusion that
surrounds the question of the existence of God. These ‘metaphysical’
theologians would be guilty in his eyes of the same confusion he attributes to
metaphysicians in general, namely that of confusing conceptual and factual
inquiries. Indeed, these two medieval thinkers might stand as paradigm cases
of metaphysicians who confuse grammatical statements with factual ones.
‘God exists,’ which should be construed as a grammatical remark, might be
confusedly taken (and is so taken by these medievals) as asserting a matter of
fact—and then one wants some support for it, some proof that it is true.
Anselm and Aquinas oblige and offer their proofs. But the very project is
wrong-headed if ‘God exists and is creator of the world’ is a grammatical
rather than factual claim.
178 Robert L.Arrington

There is another important wrinkle here. Grammatical propositions are

‘necessary’ in the sense that they provide inflexible rules (inflexible in their
application) for the proper use of words. If one talks of God at all, one must
talk of him as existent, else one has misused the term ‘God.’ Not recognizing
that ‘God exists’ is a grammatical remark, the medieval theologians
nevertheless perceive the necessity that surrounds it. They interpret this
necessity as a form of existence—necessary existence—and this they attribute
to God. ‘God exists,’ they claim, expresses the belief that God necessarily
exists. And when they prove (to their satisfaction) that God exists, they think
they have proved the reality of a necessary being. Such philosophizing is a
perfect example of a confusion of the grammatical and factual realms.
Now it is certainly true that many religious believers appeal to something
like the teleological argument or argument from design. They will say such
things as ‘Anyone who beholds the order and harmony in the world will have
to believe in God.’ A neighbor once said something like this to me as we
observed the birth of several puppies—as we observed, as she put it, the
‘miracle of birth.’ But in observing the miracle of birth or the splendor of the
heavens at night or the majesty of the mountains, are believers identifying
evidence of design that provides some sort of causal support for the reality of
a divine creator? I think not. They are not led, as a contingent matter of fact,
from the observations to the religious belief. They already have the belief,
and they see God’s handwriting in the birth, in the sky at night, in the high
mountains. It is not so much a matter of proving the existence of God as of
beholding God, of coming face to face with God through his handiwork. As
Wittgenstein would have it, the religious believer’s conviction and awe are
too great to be the result of some proof that at best provides only probabilistic
support for the proposition that informs the belief and the awe. So the type
of appeal to order and design on the part of religious believers that
philosophers like Bailey and Hyman interpret as providing these believers
with evidence for the reality of a causally efficacious divine being is not that
type of appeal at all. For religious believers, it is not the case that belief in
God’s existence needs proof: they witness God’s existence every day. In
beholding the ‘miracle of birth,’ religious believers celebrate God’s existence.
Upon examination, then, the fact that religious discourse contains some
philosophical elements does nothing to show that religious believers in
general believe in the contingent reality of a divine cause. The presence of
this philosophy just shows how some religious believers—those with a
philosophical bent—become confused. Good theology does not offer proofs;
it offers grammar.7
Bailey thinks that the religious person’s belief in miracles also provides
support for saying that this believer asserts that God is a causally efficacious
being. But, let us ask, what is the belief in a miracle? It is the belief (1) that a
highly unusual event took place which is outside the sphere of normal
causality and cannot be explained in terms of natural laws, and (2) that this
event was caused by God. But note: believers generally believe (unless they
‘Theology as grammar’ 179

subscribe to the existence of a very active and independent devil at work in

the world) that God causes everything, including the natural laws, so it is not
the belief in divine causality that is the distinctive factor in the case of
miracles. What is distinctive is the notion that we mortals cannot understand
why or how God produced these anomalous events. This is a logical cannot,
not an empirical or contingent one: it is not that, try as we might, we cannot
quite pull it off but might do so if we were considerably more intelligent, but
rather that it is incoherent to say that we mortals can understand these
miraculous events. Given the meaning of ‘miracle’ as an anomalous event
caused by God, it makes no sense to say that a mortal person can understand
such an event. At best, our understanding is limited to a grasp through reason
of the natural laws that God created to govern the normal course of events.
Such a notion of limited or partial understanding is in the background of
what is meant, in most theistic religions, by ‘human understanding’ and it is
in the background of all talk of ‘miracles.’ Such a limit on human
understanding is imposed by the very grammar of God-talk. So the belief in
miracles is not the belief in the contingent reality of a causal connection
between God and particular events. It is part of a commitment to a way of
talking and thinking about things that dictates that one see the world as
divinely created (in a way that, in general, we can’t at all understand) and as
the occasional locus of divine intervention (in ways that equally defy human
comprehension). If we can talk of a belief in such divine causality, it is
radically different from a belief in any contingent causal connection. It is a
commitment to a way of conceptualizing the world, a commitment to a set of
grammatical propositions. It is such a distinctive conceptual scheme that
Wittgenstein has in mind when he speaks of a passionate commitment to a
system of reference: ‘It strikes me that a religious belief could only be
something like a passionate commitment to a system of reference. Hence,
although it is belief, it’s really a way of living, or a way of assessing life’ (CV,
p. 64).
The discussion here relates to a claim made by Hyman that religious belief
may, and in many cases does, contain propositions that are false or
incoherent. He remarks that this fact does not render religious belief stupid
or worthless. Nevertheless, accusations of such falsity or incoherence seem
to Hyman to be part of a perfectly acceptable rational scrutiny of religious
belief. But in saying such things, Hyman does not take into account the fact
that Christian belief contains within itself a way of conceptualizing reason,
according to which reason has a role to play in human life but a very limited
one.8 Some things—the most important ones—cannot be understood by
reason, for instance God’s plan or God’s ways with the world. As Martin
Luther bluntly put the matter:

God’s divine essence and his will, administration and works—are

absolutely beyond all human thought, human understanding or wisdom;
in short, that they are and ever will be incomprehensible, inscrutable and
180 Robert L.Arrington

altogether hidden to human reason. When reason presumptuously

undertakes to solve, to teach and explain these matters, the result is
worthless, yea, utter darkness and deception. If anything is to be
ascertained, it must be through revelation alone; that is, the Word of
God, which was sent from heaven.9

It is this attitude toward reason which leads Luther, in The Bondage of the
Will, to speak of reason as ‘the stupid thing’ and as ‘foolish’ and ‘blind.’ For
the Christian, the limitations of reason come out in some very clear ways.
Paradox, as Kierkegaard stressed, is at the heart of Christianity. Perhaps the
greatest paradox is that of Christ, who is both man and God, mortal and
non-mortal. One cannot rationally accept this paradox, Kierkegaard argued,
and yet, he maintained, without accepting it one is not a Christian. But such
a proposition (‘Christ is both man and God’) is not open, for the believer, to
rational refutation or any other kind of refutation. It is one of those
propositions that are placed in the archives, out of the traffic of debate and
refutation. Anyone who rejects this proposition is simply one who rejects the
Christian scheme for conceptualizing the world. Anyone who appeals to
reason to reject this proposition as incoherent is someone who operates with
a concept of reason wholly distinct from the concept that operates within
Christianity. Christianity does not tolerate the kind of rational scrutiny found
in logical criticism of the Trinity, such rational scrutiny itself being incoherent
within the Christian grammatical scheme. Moreover, to engage in such a
rationalistic attack on Christianity is to be guilty, within the Christian scheme,
of the sin of pride.
It remains to look at Bailey’s claim that we have every right, pace
Wittgenstein, to attribute unreasonable beliefs to religious people. According
to Bailey, Wittgenstein’s strongest argument for his conviction that religious
believers are not engaged in bad science—making claims about the world
that are woefully lacking in evidential warrant—is that in general it is a bad
interpretive practice to attribute widespread unreasonable beliefs to people.
Something like the principle of charity should guide our efforts to understand
religious belief, and suggesting that beliefs in a causally efficacious divine
being are unreasonable violates this principle. Thus Bailey is led to think that
Wittgenstein offers an expressive account of religious belief as a way of
avoiding the conclusion that religious believers are unreasonable.
Bailey grants that it makes no sense to attribute nothing but unreasonable
beliefs to people, but since religious believers are perfectly normal in their
beliefs once they are outside the sphere of their religious concerns, the
attribution of unreasonable religious beliefs against the broad background of
reasonable secular beliefs is perfectly legitimate. Just as we are willing to
attribute unreasonable beliefs in astrology and psychic phenomena to people
who are otherwise normal, we can do this with religious beliefs as well.
Moreover, the very psychic mechanisms we appeal to in explaining the beliefs
in astrological and psychic events—the need for some explanation of events
‘Theology as grammar’ 181

and the tendency to impute voluntary intentions to the causes of whatever

happens, etc.—are just those that we might reasonably think to be operative
in the cases of religious belief. Thus we have good reason to dismiss religious
beliefs as unreasonable.
What, we might ask, would a Christian say to this kind of argument?
Perhaps she would point out to Bailey and the psychologists exactly what
Luther said in the passage cited above: that Christianity is a revealed religion.
Its ‘truths’ are not ones we come to know through ordinary empirical
mechanisms—and hence not as a result of any attempt to explain the world.
These ‘truths’ are imparted through the revelation of God and God’s law in
the words of the prophets, the stories of the Bible, and the teachings of the
Church. Moreover, God reveals himself through miracles and through those
forms of contact called ‘religious experience.’ One accepts these ‘truths’ on
faith, perhaps out of a feeling of the presence of the being who imparts them
and out of a trust in this being. If Christianity can be said to have an
epistemology, it is an epistemology of revealed truth, which is played out
against the background of a total trust in the existence of God. It is also an
epistemology, as we have seen, that places definite limits on human
If we put the kind of epistemology that incorporates the notions of
revealed truth and limited human understanding up against the kind of
epistemology presupposed in the attempt to explain unreasonable beliefs in
astrology and psychic phenomena, we are confronted with a dispute that can
hardly be called an ‘empirical’ one. We have no idea what kind of experience
or evidence would lead both parties to reject the epistemology of revealed
truth and accept the epistemology that assumes a naturalistic story of
standard and deviant ways of explaining empirical phenomena (or vice
versa). We have a difference in concepts of knowledge here, a difference in
world-pictures,11 and there is no intelligible way of adjudicating the dispute.
Some of us will sign on to the naturalistic picture of the world and its
naturalistic theory of belief formation; others will sign on to the idea of
revealed truth and its background notion of a divine being who makes the
revelations and can be trusted. Religious believers, who in all secular respects
are indeed normal believers, will see no plausibility whatsoever in the secular
explanation of their religious beliefs in terms of defective ways of explaining
empirical reality. Such an explanation fails to acknowledge—fails even to
understand—revealed truth.
It should be noted that the kinds of explanation to which Bailey appeals—
the human need to understand events and the human tendency to attribute
volitions to the causes of events—are not just part of a neutral explanatory
framework. They are normative and debunking in their very nature. This
need to find an explanation can get us into trouble by leading us to formulate
an explanation before we have adequate empirical grounds to do so; and the
appeal to causal intentions and volitions is just an instance of hasty and
sloppy generalization, proceeding from what is true in the human case to
182 Robert L.Arrington

what is clearly not true outside this sphere, in the nonhuman and inanimate
order of nature. But instead of seeing the sources of religious beliefs as failed
instances of general human attributes, one might want to consider the
possibility that these sources are found in ways in which human beings
manifest their human nature—in distinctive human drives and actions for
which we have no notion of failure or success.12 These may be something like
the human need to engage in ceremonial activities; they may not be far
removed from the human gesture of kissing a picture of the face of a loved
one; or they may be close to the fact that we express our pains through cries
and (later) verbal utterances. Are ceremonies false? They may be improper in
some ways—not properly conducted—but falsity does not seem to apply
here. Can one make a mistake in kissing the picture of a loved one? Can one
make an error in crying out in pain? These are just ways human beings
behave. Likewise, having a trusting, or fearful, response to the world we find
ourselves in—or a combination of the two!—may best be characterized as a
basic human response to life as everyone knows it. Perhaps it is a response
that is not perfectly universal, but it is one that surely is immensely
widespread—not a response expressed everywhere in terms of the same
stories and ceremonies, but some drive or action that might be seen faintly in
all of them, as it were in its myriad manifestations. To reject such a response
as unreasonable is to reject this aspect of our humanity—but for what
I have tried to show in the above reflections on some of the criticisms of
Wittgenstein found in this volume that in various ways the critics do not take
sufficiently to heart and mind the idea of theology as grammar. They do not
perceive that religious assertions are made against the background of a
distinctive system of reference or representation, whose rules are expressed
in numerous grammatical propositions. This failure is part of a double one,
for alongside it is the failure to recognize that their own remarks and
criticisms are framed within the context of a different grammar. Whatever
the critics say may be perfectly reasonable to those others who operate within
the same grammar. But for those, like the religious believer, who operate
outside it, and who think and talk differently, the criticisms carry no weight
whatsoever. Wittgenstein, in his remarks on religious belief, is trying simply
to provide a perspicuous representation of some of the elements of religious
grammar. The upshot of this grammatical investigation will be the dissolution
of the philosophical problem of the existence of God—there is no such
problem; there is only confusion over what is being said and done in the area
of religion. Removing this confusion allows religious believers to ‘go on’
doing what they have been doing, without being confronted by philosophical
obstacles. Equally, to be sure, this exercise in Wittgensteinian philosophy also
allows the secularists to ‘go on’ in what they are doing, thinking and talking
about the world in ways that make no reference to a divine sphere or to a
fear of or trust in an awesome presence, the presence of the divine. The two
practices are simply two of the ways in which human beings are human, part
‘Theology as grammar’ 183

of the many-splendored—or what someone has called the many-splintered—

thing called humanity.13

1 See Kai Nielsen, ‘Wittgensteinian Fideism’, Philosophy, vol. 42 (1967):191–209,
2 The later Wittgenstein gives us as well other diagnoses of the errors of
metaphysicians. Some of them, he argues, produce false or misleading
grammatical models, e.g. the Cartesian model of how we talk about the mind.
The proper response to these metaphysical errors (these grammatical illusions) is
to give a perspicuous representation of the actual grammar of language, and
thus ‘to bring words back from their metaphysical to their everyday use’ (PI
3 To be sure, Wittgenstein locates another use for ‘Red exists’: it is used, he says,
to assert that something that exists has that color (PI §58). This latter statement
is straightforwardly an empirical one. Hence it is not a metaphysical proposition
that tries to say something like: red exists in its own right. Moral: many alleged
metaphysical propositions which are, in their metaphysical use, senseless, are
meaningful if seen either as empirical propositions or grammatical guidelines.
4 But it could be argued that these assertions also express grammatical
5 See L.Wittgenstein, Philosophical Remarks, p. 55; Philosophical Grammar, pp.
97, 184; Zettel §331.
6 See my ‘The Autonomy of Language’ in J.V.Canfield and S.G.Shanker (eds),
Wittgenstein’s Intentions (New York and London: Garland, 1993).
7 We might say that theology offers proofs in the same way mathematics does. But
mathematics, Wittgenstein thinks, generates grammatical propositions. What we
need to understand is how proofs function is such grammatical domains as
theology and mathematics.
8 But see Hyman’s final endnote (p. 11).
9 Martin Luther, ‘Epistle Sermon, Trinity Sunday’ (Lenker Edition, vol. IX:#2–
18), included in A Compendium of Luther’s Theology, ed. Hugh T.Kerr
(Philadelphia: Westminster Press, 1966), p. 39.
10 Perhaps we should not say that Christianity has an epistemology, a theory of
knowledge. Christianity incorporates a concept of knowledge, not a theory of it
which needs to be defended against alternative theories.
11 See Wittgenstein’s On Certainty §§93ff, 162, 167.
12 See the essay in this volume by Brian Clack.
13 I am grateful to Mark Addis, John Hyman, and Grant Luckhardt for comments
on an earlier draft of this paper.

Addis, Mark viii Diamond, Malcolm L. 165

Alston, William 107–9, 116–18 Drury, Maurice 148, 154
Altham, J.E.J. 50 Durkheim 155
Anaximines 54
Anselm 8, 63, 128, 136, 175, 177 Eliade, Mircea 23, 28
Aquinas 8–9, 63, 101–2, 113, 128, emotivism 19–20
136, 153, 175, 177 Engels 155
Aristotle 10 Euthyphro 63
Arrington, Robert x expressivism 23, 108, 125–7; against
Augustine 18, 53–4, 63, 124 ascription of 18–21; appraisal of
Ayer, A.J. 19, 28 13–18

Bailey, Alan ix, 174–5, 177–8, 180–1 faith 4, 66–7; cognitivist theory of 66–7
Baker, G.P. 98–100 Feyerabend, Paul 146
Banner, Michael 20, 28 fideism 12, 85–100
Beattie, J.H.M. 27 figurative language 9–10
Bell, R. 83 foundationalism 101–2, 104, 109–12;
Berkeley 121–2, 128, 136, 171 strong 102–3
Bouveresse, Jacques 6, 11 Frazer, James 12–14, 16–18, 20–2, 24,
Bouwsma, O.K. 64–5, 74, 83 27–8, 122–4
Braithwaite, R.B. 19, 28, 108, 137, 165 Frege 1, 4
Brenner, William viii Freud, S. 79, 83, 155
Bunyan, John 55
genealogical account 76
Calvin, John 116 Gier, Nicholas 63
Camus, Albert 157 God 3, 5, 7–8, 122, 124–9, 174–9; as
Canfield, John 64, 183 causally efficacious 127–9; creator
Carnap, Rudolph 138–9 52–7, 61–3, 107–8; and free will
Cavell, Stanley 154 58–60; reality of 112–15
Chisholm, Roderick 103 Goethe 44
Clack, Brian vii–viii, 28, 183 Gramsci 155
Conant, James 154, 166
Crosson, F. 97 Hacker, P.M.S. 98–100, 138, 141, 165
Hägerström, Axel 165
Davidson, Donald x, 132–3, 136 Haller, Rudolf 98
Democritus 8 Hamann 149
Dennett, Daniel x, 131–2, 136 Hare, R.M. 137, 165
Descartes 63, 102 Harrison, R. 50
Dewey, John 69, 82–3, 152, 155, 164 Hart, Hendrik 166

186 Index

Helm, Paul viii Pascal 149

Hegel 13, 155 Paulos, J.A. 136
Hick, John 27 Peirce, C.S. 152, 164
Hodges, Michael viii Pelikan, Jaroslav 63
Hume, David 54, 122, 136 Phillips, D.Z. viii, 9, 11, 27–8, 65,
Hustwit, R. 83 85–99, 104–5, 108, 111, 114–15,
Hyman, John vii, 137, 165, 174–5, 117–18, 154, 166; on externalism
177–9, 183 86; on internalism 86–8, 92
Pichler, Alois 64
instrumentalism 21 Plantinga, Alvin viii, 101–5, 112–13,
James, Henry 65 Plato 159
James, William 60, 155–6, 164 primitive religion 13–18, 21–2, 24–6
Jesus 7–8, 61, 63, 69–70, 129, 177 Putnam, Hilary 113, 146, 154, 157–8,
164, 166
Kelly, J.N.D. 11 Quinn, Philip 165
Kenny, Anthony viii, 109–12, 117–18
Kerr, Fergus 97–8, 100 Reeder, John 165
Kierkegaard viii, 8, 30, 34, 65–6, 74, reformed epistemology 101–5, 109–16;
76–7, 83–4, 138, 147–50, 153, 155, criticism of 104–5; and externalism
160, 162, 180; on authority 71–3 101; and metaphysical realism
Kraus, Karl 52 112–13
Reichenbach, Hans 138
Last Judgment 5, 29, 34 religion: ethical interpretation of
Lavoisier 45 115–16; language as cognitive and
Litzenberg, Thomas 165 non-cognitive 85–6
Locke 102 religious belief 29–30, 167–9, 172–4;
Luckhardt C.G. 11 constraints on ascription of 131–3;
Luther, Martin 59, 60, 63, 65, 172, factuality of 30–1; as fundamental
179–81, 183 109–11; as unreasonable 129–33,
Luxembourg 155 180–2
Renan, Ernest 68, 82
McDowell, J. 50 Rhees, Rush 67, 82, 154, 165–6
McGinn, M. 49, 136 ritual 12–18, 21–6, 123–5
MacIntyre, Alasdair 90–1, 99 Rorty, Richard 82, 113, 146, 154, 158
magic 12–18, 21–2, 24–6 Russell 1, 4
Malachowski, Alan 82 Ryle, Gilbert 138
Malcolm, Norman x, 51, 97, 116–17,
139, 142–7, 151–5, 158, 165–6 St. Paul 52, 55, 57, 60
Marx 155 Schelling 155
metaphysical realism 112–15 Schlick 71
Mitchell, Basil 28 Schweitzer, Albert 68, 82
Monk, Ray 63 science 13–16
Moore, Gareth 63 Scotus 153
Moore, G.E. 39, 45, 49, 169 Senor, Thomas 118
Shanker, Stuart G. 64, 183
Nielsen, Kai x, 91, 97, 165–9, 172–4, Sherry, Patrick 97
177, 183 Skorupski, John 27
Nietzsche viii, 66, 74, 76, 79 Sluga, Hans 117
Socrates 36, 63
Outka, Gene 165 soul 9
Spinoza 151
Paley, William 128, 136, 177 Stern, David 117
Index 187

Stevenson, Charles 19, 28 75–6, 95–7, 119–21, 134–5,

Stitch, S x, 132–3, 136 137–42; externalism 106–8; faith
Strawson, Peter 138–9, 143 68–70; form of life 8, 77, 82, 88
Sutherland, S. 136 (absence of 79; biological 89;
Swinburne, Richard 20, 28, 136 cultural 89–90; and justification
151–4; and language-game 91–2;
Taliaferro, Charles 165 and Winch-Macintyredebate 90–1,
Tertullian 11, 149 99); grammar 94 (as arbitrary 6;
Tessin, T. 117 metaphysics as 169–72); language-
Thomas, Keith 17, 28 game 4–5, 71, 77–9, 92–5, 106,
Tilghman, B.F. 63 109, 143–6; later philosophy of
Tillich, Paul 82 language 20–1; limits of meaning
80–2; logical truths 1; metaphysical
realism 113–14; Moore propositions
Vasiliou, Iakovos viii
39–42, 47–8; mystical 2–3, 19;
Von der Ruhr, M. 117
nonsense 2; On Certainty 39–48;
pain behaviour 24–5; primitive
Waismann, Friedrich 61, 83 reactions 22–3; reference system
Weber 155 42–3; religious belief 5–9, 31–9,
Well, Simone 61–2, 65 47–8, 121–7, 142, 147–50, 154–5;
Whittaker, J.H. 65 (and authority 70–4; expression of
Williams, M. 49 155–64; immunity from rational
Winch, Peter 11, 27, 64, 90–1, 97, 99, criticism 10); very general facts of
116, 139–40, 145–7, 151–5, 157–8, nature 75; world picture 42–5
166 (disputing a 45–7)
Wittgenstein: conception of philosophy Wolterstorff, N. 116