THE COLLECTED

PHI LOS OPHI CAL PAPERS OF

G. E. M. ANSCOMBE
VOLUME THREE
Ethics, Religion and
Politics
Basil Blackwell Oxford
First published in 1981 by
Basil Blackwell Publisher
108 Cowley Road
Oxford OX4 iJF
England
All rights reserved. No part o f this publication
may be reproduced, stored in a retrieval system,
or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording
or otherwise, without the prior permission of
Basil Blackwell Publisher Limited.
in this col l ect i on G. E. M. Ans c ombe 1981
British Library Cataloguing in Publication Data
Anscombe, Gertrude Elizabeth Margaret
The collected philosophical papers o f G. E. M. Anscombe
Vol. 3: Ethics, religion and politics
1. Philosophy, English Addresses, essays
lectures
I. Title
192'.08 B1618
ISBN 0-631-12942-1
Typeset in Photon Baskerville
Printed and bound in Great Britain at
The Camelot Press Ltd, Southampton
12 Faith
In the late 1960s some sentence in a sermon would often begin: We used to
believe t h a t . . . . 1 always heard this phrase with an alarmed sinking o f the
heart. I had alternative expectations. The more hopeful one was f o r some
absurd lie. For example: We used to believe that anyone was safe for
Heaven i f he kept the Church rules. We used to believe that there was no
worse sin than to miss mass on Sunday. The worse one was o f hearing
something like We used to believe that there was something special about
the p r i e st h oo d ; We used to believe that the Church was here for the salva
tion o f souls.
Now there was a We used to believe . . . which I think could have been
said with some truth and where the implied rejection wasnt a disaster. There
was in the preceding time a professed enthusiasm fo r rationality, perhaps
inspired by the teaching o f Vatican I against fideism, certainly carried along
by the p romotion o f neo-thomist studies. T o the educated laity and the
clergy trained in those days, the word was that the Catholic Christian faith
was rational, and a problem, to those able to feel it as a problem, was how it
was gratuitous. - a special gift o f grace. Why would it essentially need the
promptings o f grace to follow a process o f reasoning? It was as i f we were
assured there was a chain o f proof. First, God. Then, the divinity o f Jesus
Christ. Then, his establishment o f a Church with a Pope at the head o f it and
with a teaching commission from him. This bod y was readily identifiable.
Hence you could demonstrate the truth o f what the Church taught. Faith,
indeed, is not the same thing as knowledge but that could be accounted for
by the extrinsic character o f the proofs o f the defide doctrines. The knowledge
which was contrasted with faith, would be knowledge by proofs intrinsic to
the subject matter, not by proofs from someone s having said these things
were true. For matters which were strictly o f faith intrinsic proofs were not
possible, and that was why faith contrasted with knowledge .
This is a picture o f the more extravagant form o f this teaching. A more
sober variation would relate to the Church that o ur L or d established. In this
variant one w ou l d n t identify the church by its having the Pope, but
otherwise; and one would discover that it had a Pope and that that was all
right. This more sober form had the merit o f allowing that the believer was
committed to the Christian faith, rather than suggesting that he had as it
were signed a blank cheque to be filled out by the Pope in no matter what
sum.
A yet more sober variant would have avoided trading on the cultural in-
This paper has not previously been published. It was delivered as a Wiseman Lecture at
Oscott College.
The Philosophy of Religion
heritance for which the name o f Jesus was so holy that it was easy to go
straight from b e li e f in G o d to b e li e f in je s us as G o d s Son. In this more sober
variant one would be aware o f the dependence o f the New Testament on the
Old: one would be clearly conscious o f the meaning o f calling o ur Lord
Christ .
The sober variants would have a disadvantage fo r the propagandists o f
the rationality (near demonstrability) o f faith - though a great advantage in
respect o f honesty and truthfulness. The disadvantage was that no one could
suppose it quite easy fo r anyone to see that what Jesus established was
matched by the Catholic Church that we know. I f it was j ust a matter o f his
having f ounded a Church with a Pope, then it was easy indeed! Butotherwise
it was obvious that learning and skill would be required to make the iden
tification. And the considerations and arguments would be multifarious and
difficult to be sure about. Hence the problem most commonly felt, amongst
the more intellectual enquirers, as to the character o f the certainty ascribed to
faith. The so-called preambles o f faith could not possibly have the sort o f
certainty that it had. And i f less, then where was the vaunted rationality?
But there was a graver problem. What about the faith o f the simple ?
They could not know all these things. Did they then have some inferior
brand o f faith? Surely not! And anyway, did those who studied really think
they knew all these things? No: but the implication was that the knowledge
was there somehow, perhaps scattered through different learned heads,
perhaps merely theoretically and abstractly available. In the be lie f that this
was so, one was being rational in having faith. But then it had to be
acknowledged that all this was problematic - and so adherence to faith was
really a matter o f hanging on, and both its being a gift and its voluntariness
would at this point be stressed.
I sometimes hear accounts o f the times o f darkness before Vatican II
which strike me as lies. I hope that I have not been guilty o f lying in what I
have said here. This at least is my recollection o f how it was in some presen
tations, some discussions, some apologetic.
Was, and is no longer, not necessarily because better thoughts about faith
are now c o m m o n ; there is a vacuum where these ideas once were prominent.
But all these considerations, proofs, arguments and problems are now out o f
fashion, f o r various reasons which I w on t discuss.
The passing away o f these opinions is not to be regretted. They attached
the character o f rationality entirely to what were called the preambles and
to the passage from the preambles to faith itself. But both these preambles
and that passage were in fact an ideal construction - and by ideal I d o n t
mean one which would have been a good development o f thinking, i f it had
occurred in an individual; I mean rather fanciful , indeed dreamed up ac
cording to prejudices: prejudices, that is, about what it is to b e reasonable in
holding a belief.
The right designation f o r what are called the preambles o f faith is not
that but at least f o r part o f them, presuppositions . Let me explain this in a
Faith
115
simple example. You receive a letter from soneone y ou know, let s call him
Jones. In it, he tells you that his wife has died. Y o u believe him. That is, you
now believe that his wife has died because you believe him. Let us call this just
what it used to be called, human faith . That sense o f faith still occurs in
our language. Why , someone m*ly be asked, do y ou believe such-and-
suchP , and he may reply I j ust took it on faith - so-and-so told me .
Now this believing Jones, that his wife has died, has a number o f presup
positions. In believing it you presuppose (1) that y our f r i e n d j o n e s exists, (2)
that this letter really is f rom him, (3) that that really is what the letter tells you.
In ordinary circumstances, o f course, none o f these things is likely to be m
doubt, but that makes no difference. Those three convictions or assumptions
are, logically, presuppositions that>'o have i f y our b e li e f that Jones wife has
died is a case o f your believing Jones.
Note that I say they areyour presuppositions. I do not say that your believ
ing Jones entails those three things; only that y our believing Jones entails
that y ou believe those three things.
In modern usage faith tends to mean religion, or religious belief. But
the concept o f faith has its original home in a particular religious tradition.
I f a Buddhist speaks o f his faith , saying f o r example that his faith ought
not to be insulted, he means his religion, and he is borr owing the word
faith which is really alien to his tradition. In the tradition where that
concept has its origin, faith is short for divine faith and means believ
ing G o d . And it was so used, among the Christian thinkers at least, that
faith, in this sense, could not be anything but true. Faith was believing God,
as Abraham believed God, and no false b e l i e f could be part o f it.
I want to say what might be understood a bout faith b y someone who did
not have any; someone, even, who does n ot necessarily believe that God
exists, but who is able to think carefully and truthfully about it. Bertrand
Russell called faith certainty without p r o o f . T hat seems correct. Ambrose
Bierce has a definition in his Devil s Dictionary; The attitude o f mind o f one
who believes without evidence one who tells without knowledge things
without parallel . What should we think o f this?
According to faith itself faith is believing God. I f the presuppositions are
true, it is, then, believing on the best possible grounds someone who speaks
with perfect knowledge. I f only the presuppositions were given, Bierce
would be a silly fellow and Russell would be confuted. But is there even the
possibility o f believing G o d ? This is hard to grasp: it is itself one o f the
things without parallel .
Anyway, in general, faith comes by hearing , that is, those who have taith
learn what they believe by faith, learn it from other people. So someone who
so believes believes what is told him by another human, who may be very
ignorant o f everything except that this is what he has to tell as the content o f
faith. So Bierce s Devil may be right. Qne who has not evidence believes one
who has not knowledge (except o f that one thing): at least, he believes what
the latter says and he gets what he believes from the latter; yet according to
The Philosophy of Religion
faith he believes God. I f so, then according to faith a simple man - a man
with no knowledge o f evidence - may have faith when he is taught by a man
ignorant o f everything except that these are the things that faith believes.
More than that, according to faith this simple man and his teacher have a
belief in no way inferior to that o f a very learned and clever person who h a s
faith.
//faith is like that, even i f it is believing God, then it follows that the Bierce
definition is right after all. For everyone is to have faith and few can be
learned, and their learning doesn t give them a superior kind o f faith.
Everyone is to run: and few are road-sweepers.
It is clear that the topic I introduced o f believing somebody is in the middle o f
our target. Let us go back to Jones, and investigate believing Jones, when you
read in his lettei that his wife has died. You can t call it believing Jones j ust i f
Jones says something or other and you do believe that very thing that he says.
For you might believe it anyway. And even i f it s someone s saying something
that causes you to believe it, that doesn t have to be believing him. He might
just be making you realize it, calling it to your attention - but y ou j u d g e the
matter for yourself. N o r is it even sufficient that his saying it is y our evidence
that it is true.1 For suppose that you are convinced that he will both lie to
you, i.e. say the opposite o f what he really believes, and be mistaken ? That is,
the opposite o f what he thinks will be true; and he will say the opposite o f
what he thinks. So what he says will be true and you will believe it because he
says it. But you w o n t be believing him!
Ordinarily, o f course, when you believe what a man says, this is because
you assume that he says what he believes. But even this doesn t give us a
sufficient condition f o r your believing to be believing him. For here again one
can construct a funny sort o f case - where you believe that what he believes
will be true, but by accident, as it were. His b e li e f is quite idiotic, he believes
what h e s got out o f a Christmas cracker for example. In fact, unknown to
him but known to you, what has been put in the crackers for their party arc
actual messages with some practical import. You know that the messages in
the blue crackers are all true, and the ones in the red crackers all false. He
believes any o f them. And now he tells you something, and you believe it
because he says it and you believe he is saying what he believes, and because
you know that this thing that he believes comes out o f a blue cracker. That
wouldn t be believing him. But when you believe your history teacher, for
example, it is enough that you believe what he says because he says it and you
d on t think h e s lying and you think what he believes about that will be true. I
mean, that is enough f o r you to be believing him.
So the topic o f believing someone is pretty difficult. O f course i f you could
put in that you believe the person knows what he is telling you, these
difficulties d o n t arise. You believe what he says because he says it and you
believe that he knows whether it is so and wori t be lying. T h a t s why this p a r
1 I asked Mary Geach to construct a case to show this. She responded with what follows.
Faith 117
ticular problem w o n t normally arise about that letter from your friend
Jones telling y ou his wife has j ust died.
Now there is another question about what it is to believe someone, which
concerns the presuppositions. I said that jyou presuppose that Jones exists,
did write the letter, and did say that in it, i f you believe him to the effect that
that was so. I d id n t say that the mere fact o f y our believing Jones presupposed
those things. Now what in fact are we to say here? Suppose someone has a
hoax pen-friend - I mean, the pen-friend is really a contrivance o f his
school-fellows who arrange for their letters to be posted from Chicago to
England and make the non-existent correspondent tell their friend all sorts
o f things. And suppose he believes the things in the usual sort o f way in which
people believe things they are told. Is he believing the pen-friend? What are we
to say? Would n t we say that some ancient believed the oracles o f the gods?
And w ou l dn t it then be right to say he believed the god whose oracle it was ?
I f y ou insist on saying that the deluded victim does not believe the pen-
friend because the pen-friend doesn t exist, you will deprive yourself o f the
clearest way o f describing his situation: he believed this non-existent p e r
son . And, somewhat absurdly, you will have to say that his own expression
o f b e li e f I believe her , is not an expression o f belief, or not a proper one.
What then would be the proper one? We had better settle for saying that the
victim believes the pen-friend, and that the ancient was believing A po l l o -
who does not exist. And doesn t the same point hold for the case where the
letter-writer does exist, but you have misunderstood what he wrote, or mis
takenly supposed that this letter is from him? Especially i f the mistakes were
quite reasonable ones.
Now let us think some more about the presuppositions. Ordinarily the
presuppositions o f believing N simply do not come in question. I get a letter
from someone I know; it does not occur to me to doubt that it is from him.
Suppose that the doubt does occur for some reason. The letter perhaps says it
is from him - the very thing to raise a doubt! Now I take it as obvious that, i f I
decide to believe that the letter is from him, I w on t do so on the grounds on
which I believe him when the letter says his wife-is dead. For I believe his wife
is dead because he says so. But the reason why I believe the letter is from him
is n ot that he says so. His credibility is not my warrant for believing that the
letter is from him. Even i f the letter begins This is a letter from your old
friend Jones and I j ust believe that straight o f f and uncritically, I believe the
sentence, and believe that the letter is from Jones because the sentence says
so, but I could never say I believed it because I believed him. This is the sense
in which the presuppositions o f faith are not themselves part o f the content
o f what in a narrow sense is believed by faith.
Now let us change the case. Suppose a prisoner in a dungeon, to whom
there arrives a letter saying: This letter comes from an unknown friend, N . 2
2 This development o f a case which I considered (see What is it to believe someone? in
Volume II), o f a letter from an otherwise unknown person, is taken from Peter Geach. See The
Virtues (Cambridge, 1977).
x 1 8
The Philosophy of Religion
It proposes to help him in various needs which he is invited to communicate
by specified means. Perhaps it also holds out hope o f escape from the prison
The prisoner doesn t know i f it is a hoax or a trap or is genuine, but he tries
the channels o f communication and he gets some o f the things that he asks
for; he also gets further letters ostensibly f rom the same source. These letters
sometimes contain information. We will suppose that he now believes that N
exists and is the author o f all the letters; and that he believes the information
as coming from N. That is, his b e li e f in that i nformation is a case o f believing
N. His be lie f that N exists and that the letters come from N is, j us t as in the
more ordinary case, not an example o f believing something on N s say-so
On the other hand, as we are supposing the case, he does not have prior
knowledge o f N s existence. And it could happen that he, like the man who
uncritically accepts the letter beginning This is from J ones , believes the
opening communication This is from an unknown friend call me N
straight off: j us t as he d likely believe straight o f f that a whispered o r tapped
communication purporting to come from the next cell is a communication
from another prisoner. Even so the beliefs which are cases o f believing N and
the be lie f that N exists are logically different. This brings out the difference
between presuppositions o f believing N and believing such-and-such as
coming from N. Pre-suppositions d o n t have to be temporarily prior
beliefs.
Suarez said that in every revelation God reveals that he reveals. That
sounds like saying: every letter from N to the prisoner informs him that the
information in the letter is from N, and every bit o f information from N is
accompanied by another bit o f information that the first bit was from N. Put
like that there is an absurdity, an infinite regress. But it should not be put like
that. Rather: in every bit o f information N is also claiming (implicitly or
explicitly, it doesn t matter which) that he is giving the prisoner information.
And now we come to the difficulty. In all the other cases we have been con
sidering, it can be j n a d e clear what it is f o r someone to believe someone. But
what can it mean to believe G o d ? Could a learned clever man inform me
on the authority o f his learning, that the evidence is that G o d has spoken?
No. The only possible use o f a learned clever man is as a causa removens pro-
hibens. There are gross obstacles in the received opinion o f my time and in its
characteristic ways o f thinking, and someone learned and clever may be able
to dissolve these.
Forgetting that a bout hearing - i.e. from teachers - should we picture it
like this: a man hears a voice saying something to him and he believes it is
God speaking, and so he believes what it says - so he believes G o d ? But what
does he believe when he believes it is God speaking ? T hat God has a voice-
b o x ? Hardly. In relation to the be lie f that it is G o d speaking, it doesn t
matter how the voice is produced. There is a Rabbinical idea, the Bath Q ol ,
the daughter o f the v o i c e . You hear a sentence as you stand in a crowd - a
few words out o f what someone is saying perhaps: it leaps out at you, it
speaks to your c o n ditio n . Thus there was a man standing in a crowd and he
Faith 1 1 9
heard a woman saying Why are you wasting your time? He had been
dithering about, putting o f f the question o f becoming a Catholic. The voice
struck him to the heart and he acted in obedience to it. Now, he did not have
to suppose, nor did he suppose, that that remark was not made in the course
o f some exchange between the woman and her companion, which had
nothing to do with him. But he believed that G o d had spoken to him in that
voice. The same thing happened to St Augustine, hearing the child s cry,
Tolle, lege .
Now the criticial differentiating point is this. In all those other cases it is
clear what the one who believes X means by X speaking , even when we
judge that X doesn t exist. For example, what the believer in the oracle
means by A po l l o speaking . But it is not clear what it can mean for God to
speak.
For A po l l o , or Juggernaut, is simply the god o f such and such a cult. Note,
I am not here following those who explain deity as the object o f worship .
That definition is useless, because they have to mean by worship the
honour paid to a deity . Divine worship is the special sort o f honour
intended to be paid, the special sort o f address made, to a deity. This may be
offered to what is not divine, to a stone or another spirit or a man; or to what
doesn t exist at all.
So when I say A po l l o was the god o f such-and-such a cult , I am calling
attention to the question: what would it mean to say These were n o t - n o n e
o f them were - the temples and oracles o f A p o l l o - precisely of the temples
and oracles o f A p o l l o ? What would it mean to say that Shiva was not the god
o f destruction? Shiva is the god o f this worship, which is the cultus o f a god
o f destruction.
In this sense, G o d is not the god o f such-and-such a worship. This is
something that can be seen by an atheist too, even though he holds that there
is in any case no such thing as deity. For he can see, i f he thinks about it, that
G o d is not a proper name but is equivalent to a definite description (in
the technical sense). That is, it is equivalent to the one and only true g o d ,
the one and only real deity . The point o f putting in true and real is
that those who believe there is only one deity have so much occasion to speak
o f deities that they do not believe to exist. We then speak o f A po l l o , Shiva
and Juggernaut as gods who are not gods. An atheist believes that God is
among the gods who are not gods, because he believes that nothing is a deity.
But he should be able to recognize the identity o f G o d with the one and
only g o d . ,
It is because o f this equivalence that G o d cannot be formally identified as
the go d o f such-and-such a cult or such-and-such a people. T o say that God
is the god o f Israel is to say that what Israel worshipped as god was the one
and only g o d . So it could significantly be denied. And it could be seen to be
true - even by one who believed that the description the one and o nly g o d is
vacuous. . .
And so we can say this: the supposition that someone has faith is the sup-
120
The Philosophy of Religion
position that he believes that something - it may be a voice, it may be
something he has been taught - comes as a word from God. Faith is then the
belief he accords that word.
So much can be discerned by an unbeliever, whether his attitude is poten
tially one o f reverence in face o f this phenomenon o r is only hostile. But the
Christian adds that such a be lie f is sometimes the truth, and that the con
sequent belief is only then what he means by faith.