This article was downloaded by: [UQ Library]

On: 08 June 2013, At: 16:28

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Inquiry: An Interdisciplinary Journal of Philosophy

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/sinq20

On the attitude of trust

a
Lars Hertzberg
a
Åbo Academy, Smedsgatan 4, SF‐10600, Ekenäs, Finland
Published online: 29 Aug 2008.

To cite this article: Lars Hertzberg (1988): On the attitude of trust, Inquiry: An Interdisciplinary Journal of Philosophy, 31:3,
307-322

To link to this article: http://dx.doi.org/10.1080/00201748808602157

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 Inquiry, 31, 307-22

On the Attitude of Trust

Lars Hertzberg
Åbo Academy

In On Certainty, the emphasis is on the solitary individual as subject of knowledge.

The importance of our dependence on others, however, is brought out in
Wittgenstein's remarks about trust. In this paper, the role and nature of trust are
discussed, the grammar of trust being contrasted with that of reliance. It is shown
that to speak of trust is to speak of a fundamental attitude of one person towards
others, an attitude which, unlike reliance, is not to be explained, or assessed, by an
Downloaded by [UQ Library] at 16:28 08 June 2013

appeal to reasons. It is, rather, because we have such a fundamental readiness to

accept what we are taught by others that we can come to develop an understanding
of reasons. The idea that believing something without evidence is always a
weakness is shown to be a philosophical prejudice. Trust is always for something
we can rightfully demand from others: misplaced trust, accordingly, is not a
shortcoming on the part of the trustful person, but of the person in whom the trust
was placed. The destruction of trust is a tragedy of life; in Culture and Value,
Wittgenstein suggests a connection between distrust and madness.

I
A reader who comes directly to On Certainty* from the classical writings
on scepticism, without an acquaintance with Wittgenstein's earlier work,
Philosophical Investigations, may have an illusion of being at home in it
at least in one respect: for long stretches it may seem to him that he is
here once more confronted with the solitary subject of traditional epis-
temology struggling to determine what, if anything, he can know with
certainty about the world in which he finds himself. For undoubtedly the
single observer is prominent in On Certainty, much more so than in
Philosophical Investigations. A central category in the discussion is that
of judgments that stand fast for me. These will not in some cases be
the same for others (e.g. that this is my name, that I have never been to
China). As Norman Malcolm notes in discussing On Certainty: '[I]t is an
individual who asserts that something is certain. If I am certain about
the truth of something it is / who am certain. . . . If I have occasion to
declare, "I can't be mistaken about my name", this assertion "relates
essentially to me".' 1

* L. Wittgenstein, On Certainty, ed. G. E. M. Anscombe and G. H. von Wright, trans.

D. Paul and G. E. M. Anscombe (Oxford: Blackwell, 1969). Section references prefixed
OC are to this work.
 308 Lars Hertzberg

If, on the other hand, we approach On Certainty with a familiarity

with Philosophical Investigations, it will, or should, look differently to
us. To such a reader it will go without saying that the issues discussed in
On Certainty are to be taken as arising among speakers who share a
language and who use this language in telling one another things, asking
questions, conducting investigations, etc. And of course there are
pointers to remind him, such as: 'In order to make a mistake, a man
must already judge in conformity with mankind' (OC, §156).
On the whole, I should like to suggest, On Certainty is to be read as
an afterthought to Investigations, as a bringing into focus of issues and
kinds of language-game that were left out of consideration or were not
given sufficient emphasis in that work, rather than as a new departure.
And if On Certainty is, indeed, a corrective to the earlier work, our
Downloaded by [UQ Library] at 16:28 08 June 2013

reading of the later work must, in turn, be kept in check by what is said
in Investigations.
One of the most important of the themes linking the later work to the
earlier is the role of our believing others in learning to make judgments
and learning about the world:
As children we learn facts; e.g., that every human being has a brain, and we
take them on trust. I believe that there is an island, Australia, of such-and-such
a shape, and so on and so on; I believe that I had great-grand-parents, that the
people who gave themselves out as my parents really were my parents, etc. . . .
The child learns by believing the adult. Doubt comes after belief. (OC, §§159-
60)
I really want to say that a language-game is only possible if one trusts something
(I did not say 'can trust something'). (OC, §509)
What I wish to do in this paper is to inquire into the nature of the atti-
tude of trust and its role in learning. More broadly speaking: what is the
importance of our depending on others in various contexts of learning,
and what does this dependence consist in? The paper is not exegetical.
Its aim is to illuminate the context of some of the remarks in On
Certainty.

II
There does not seem to have been a great deal of discussion about the
concept of trust in recent philosophy. What discussion has been carried
on seems to have focused mainly on the relation between the concepts
of believing someone and believing what he says.2 It may have been
thought that this is the basic question to consider, that believing another
is ultimately reducible to a matter of believing what he says (though
writers on the topic have been careful to point out that there is not a
 On the Attitude of Trust 309

strict identity between the two). For one thing, however, such an
approach disregards the fact that trust also enters into human relations
in ways that have little to do with accepting the truth of statements. And
besides, even when we consider the role of trust simply in the context of
our learning to make judgments where the concept of truth is relevant,
it should be clear that an account along those lines will not do. For
learning to judge cannot begin with our accepting what others say as the
truth. Regarding a statement as true is itself an exercise of judgment,
which becomes possible only on a comparatively advanced level of
understanding: the decision to accept a judgment cannot be meaning-
fully attributed to someone who is not yet in a position to reject it.
The attempt to account for the role of trust in human relations as a
matter of the accepting of statements, I would argue, suffers from a cog-
Downloaded by [UQ Library] at 16:28 08 June 2013

nitive and intellectual bias. Believing what others say is a refinement of

other, more basic forms of trust. Only in a context constituted by trust,
we might say, do truth and the making of statements have a place. We
must begin by trying to understand the nature of trust as a primitive
reaction.

Ill
Let us, for instance, consider trust in relation to fear. If I trust someone,
this may mean that I am not frightened of him; also, in his presence I
am perhaps less frightened of the things that otherwise frighten me. The
child does not fear the dark when his mother is in the room. The knife
in my friend's hand is not frightening to me; I know he will not attack
me. Perhaps I know too that he will protect me from the attacks of
others, and that the knife will help him.
In these cases, it seems, the role of trusting can be understood in
relation to a person's needs, and desires. Trust, here, though it may
involve no beliefs, is like the confident belief that I shall be carried
safely past the danger or the pain; it enters into my relation to the
object of fear.
The point I wish to make, however, is that in some cases more than
this has to be involved in a trustful attitude. I should like to illustrate
this point with the help of the story of Abraham and Isaac. Consider, on
the one hand, the trust that God, according to the Bible, demanded
from Abraham, commanding him (without any explanation) to sacrifice
what meant more to him than his own life. In preparing to obey, Abra-
ham showed his faith in the inscrutable goodness of God; he was pre-
pared to embrace a way of acting that promised nothing but pain and
grief for himself and his wife, and nothing desirable for anyone else. His
 310 Lars Hertzberg

obedience expressed the thought that God's command was able by itself
to constitute the action as a good, although to his mind nothing else
spoke in its favour. Now consider, on the other hand, the trust that
Isaac had for his father. He went along on the journey to Moriah with-
out fear, trusting his father would not use his knife on him. But suppose
he had known what was on his father's mind, and had still gone on with-
out resisting. Acting in that way would have required a deeper sort of
trust: not just a belief that nothing he feared for could befall him from
his father, or at his father's side, but the conviction that whatever came
to him through his father, even if it was precisely that which he feared,
was still something to be accepted. His trust would then have been simi-
lar to his father's trust in God.
Downloaded by [UQ Library] at 16:28 08 June 2013

IV
This situation presents in dramatic form what I would contend must nor-
mally be the nature of a child's relations to his elders. Isaac's attitude
towards his father, as I have imagined it, does not simply enter into his
relation to the objects of his fear, but rather establishes a new relation
to them: while he may still fear them, he has now at the same time
come to see them as things to be accepted, maybe even sought for.
Something now speaks in favour of enduring the fear, or the pain. What
makes this important, I would argue, is that the possibility of something
speaking in favour of or speaking against a way of acting is first intro-
duced in this sort of way. Differently put: in this sort of way an under-
standing of what it means to have reasons for acting or not acting in a
certain way becomes possible.
There is a way of thinking about reasons that may incline us to over-
look this point. On this manner of thinking (of which we find an
instance in Hume), an understanding of what it means to have a reason
for acting is a natural development out of our desires and inclinations.
The thought that we have a reason for doing something is simply the
thought that this action is required for, or conducive to, some desired
objective.
I would argue that such an account fails to advance our understanding
of the matters at hand.3 Its apparent plausibility is connected with an
obscurity in the role that desires are taken to have in the context. I may
explain the point of something I am doing by indicating an object I
desire and expect to get through my action. So a desire may, at least in
a sense, be given as a reason. But what I do is not simply to report that
something is an object of my desire; rather I express an attitude towards
this desire, implying that, in these circumstances, I regard it as accept-
 On the Attitude of Trust 311

able (or as making sense, or the like) to act on it. So I must already
have learnt to take part in discussions about actions being or not being
acceptable, making or failing to make sense, if I am to make an utter-
ance that can be understood in this way. And this is, of course, different
from simply having learnt to give expression to my desires.
What may tempt us to overlook this point is the fact that we may
attribute desires to someone even in cases in which there is no question
of his having an attitude towards them. Consider the ways we sometimes
speak about infants and animals. We may say, for instance, that a cat
behaves in a certain way because he is hungry. And there may be a kind
of learning involved here, as the cat gradually comes to look for food in
the right places, etc. Hence it is tempting to think that what is going on
here is similar to the case of adult people acting on reasons, 'only with-
Downloaded by [UQ Library] at 16:28 08 June 2013

out language'. (This was probably the way Hume thought about it.) But
I hope I have said enough to make it clear that the cases are different in
important ways.
If this is correct, it would seem that a child's relation to those of his
elders on whom he is dependent for being brought up into membership
in a human community cannot be understood simply in terms of his hav-
ing learnt to look to them to fulfil his needs and desires. Of course their
willingness and ability to fulfil them may influence the quality of that
relationship. But the child's readiness to go along with what is intimated
to him cannot be thought to derive from any more basic consideration:
what others suggest will matter to the child simply because they suggest
it, or because it is they who suggest it. It seems, then, that the idea of
something speaking in favour of an action can only acquire meaning for
a child through someone's speaking in its favour. In this way, coming to
have an understanding of good and bad, of things mattering, pre-
supposes a fundamental dependence on other people.

Contemporary philosophers have, I believe, been inclined to overlook

these points, because of a tendency to assimilate the attitude discussed
here to a different sort of attitude that one person may have to another.
This tendency is closely connected with the cognitive and intellectual
bias mentioned before. For convenience, I shall describe the distinction
that needs to be made here as a contrast between the attitudes of trust
and reliance. It is not my ambition, however, to draw a map of the
actual use of the words 'trust' and 'reliance', let alone recommend a way
of using them. My concern is rather with drawing attention to what
might be called a grammatical contrast between various forms of human
 312 Lars Hertzberg

relation. While, the words 'trust' and 'reliance' may often mark this dif-
ference, I am not maintaining that they do so systematically. Besides, as
will be made clear, there are cases fitting the grammar of trust in essen-
tial respects to which the word 'trust', nevertheless, does not have an
application.
First, consider the grammar of reliance. It can, I believe, be briefly
characterized by saying that to rely on someone is to exercise one's
judgment concerning him. The expression of one's reliance may be the
fact that one tends to accept claims made by that person within a given
area; it may also be a matter of depending on him for getting certain
things done, or depending on him not to do certain things. Calling this
an exercise of judgment serves to bring out two points. On the one
hand, it says something about the role of the concept of reliance in
Downloaded by [UQ Library] at 16:28 08 June 2013

explaining a person's thought and action: the fact that someone depends
on another in this way can be seen to be intelligible, in so far as it is
intelligible, in the light of what he knows or believes about the other, or
of his past experience with him. On the other hand, calling reliance an
exercise of judgment is a comment about the terms in which such an
attitude will be discussed and criticized: in trying to decide whether we
should rely on someone in a given situation, we consider the evidence
concerning his competence, character, etc., and weigh it against the
stakes involved.
The grammar of reliance is like the grammar of believing or knowing
certain facts about a person, though it cannot be assimilated to it, since
it may not be possible to specify, in the form of a proposition, what one
person's reliance on another consists in. Still, the comparison of reliance
to factual belief brings out some important features of this attitude.
Reliance has a more or less specific content: one relies on a person for
particular purposes. This means that there is a more or less definite
range of things concerning which I am prepared to take his word, or a
more or less definite range of things I expect him to do or not to do.
Correspondingly, relying on someone involves the thought of inde-
pendent standards by which it is to be judged whether or not my reliance
on him was misplaced; reliance is conditional on those standards being
met.
Reliance is also like factual belief in the sense that my relying on
someone is conceptually independent of whatever attitude I take to him
in other respects. I may be suspicious of his character, or despise him
for his childish innocence, etc., and yet rely on him in this or that par-
ticular respect. With this is connected the point that reliance seems not
to be, essentially, an attitude towards a person. The ways in which one
may rely on a tool, a measuring instrument, a horse, etc., seem to be
analogous to the ways in which one may rely on people.
 On the Attitude of Trust 313

VI '
Such, I want to say, is the grammar of reliance. The point I wish to
make is that a number of situations involving a trustful attitude to others
are ones in which that grammar does not apply. This is so, for example,
in the following types of situation (the cases of trust mentioned in On
Certainty belong to one or another of these):
(i) The child's first initiation into the form of life of his community. The
child's spontaneous activities and responses are guided in particular
directions or given particular shape: the child is given certain things to
eat and is taught not to eat certain other things; his movement in certain
directions and his exploration of certain kinds of objects are encour-
aged, while other movements and explorations are stopped; his expressive
Downloaded by [UQ Library] at 16:28 08 June 2013

behaviour evokes responses, is encouraged and gradually shaped into

what will constitute ways of using language. These developments are
guided primarily by the individuals whom the child recognizes and with
whom he feels secure.
(ii) Initiation into particular social institutions. E.g. learning to walk in
the street, to use public transportation. Learning about the use of
money, about buying from shops, etc. Learning to read newspapers, to
use the telephone and postal systems, to deal with public officials, etc.
(iii) Initiation into particular forms of discourse and inquiry. Learning
mathematics, history, geography, etc. Only in these situations does the
learning of 'truths' occupy a central position. Concerning these cases,
Wittgenstein writes:

I believe what people transmit to me in a certain manner. In this way I believe

geographical, chemical, historical facts etc. That is how I learn the sciences. Of
course learning is based on believing.
If you have learnt that Mont Blanc is 4000 metres high, if you have looked it
up, you say you know it.
And can it now be said: we accord credence in this way because it has proved
to pay? (OC, §170)

The latter two types of situation differ from the first through the fact
that the trustful attitude being developed here is not directed at any
given individuals: rather, our learning to trust people in contexts like
these is bound up with learning to identify the position that they occupy
or the institution or subject for which they speak. Nevertheless, the
distinction between these three types of case is not a sharp one; on the
contrary, the nature of these forms of initiation and the role of our
dependence on others in connection with them will be more clearly
understood if we think of them as forming a continuum. Thus, there will
 314 Lars Hertzberg

be a continuous transition from the learning of general to the learning of

specialized skills and attitudes, and from situations in which the empha-
sis is on doing things to situations in which it is on knowing truths (lan-
guage gradually acquiring a greater role in the process, a role which is
not coextensive with, but rather preliminary to, the learning of truths).
And on the other hand, the dependence involved in intimate human
relations will gradually be extended into the kind of trust that constitutes
the basic framework of life in a human society.
It should be obvious that in these cases the features of reliance that
were pointed out before do not apply. Coming to trust one's teachers is
not, in these cases, a matter of exercising one's judgment. The learner
has no advance understanding of the aim of the instruction, hence
nothing will constitute, for him, evidence of his teachers' reliability or
Downloaded by [UQ Library] at 16:28 08 June 2013

unreliability in the matter at hand. The learner is not in a position to impose

conditions on his trust, since the standards for assessing the teachers'
performance are only made accessible to him through their performance.
Later developments may, of course, come to undermine one person's trust
for another. This may come about, for instance, through a conflict between
his trust for that person and his trust for someone else. But the way in
which this happens cannot be determined by the nature of the trust itself.
Nothing can force a person to give up his trust in another; in order to do
so he must change.
This also means that, in so far as I trust someone, there will be no
limits, given in advance, of how far or in what respects I shall trust him.
If I trust someone, I cannot at the same time reserve for myself the
judgment concerning the purposes for which he is to be trusted. It is
from him that I learn what he has to teach me. I go along.
Suppose I am initiated into a new subject such as history. In the pro-
cess I learn a new way of speaking: a use of assertions about the past
which are not expressions of memory. When I take my teacher's words
on trust it would not make sense to say that I rely on the information he
gives me. I do not simply accept what he says as the truth, but rather I
learn from him a new application for the concept of truth. If I am to
learn history at all, I must permit what he tells me to become for me the
standard of truth in this area. The truth of what he says can be chal-
lenged only by others who also speak for the subject, and whom I
believe in the way I believe him. Wittgenstein writes:
If someone asked us 'but is that true?' we might say 'yes' to him; and if he
demanded grounds we might say 'I can't give you any grounds, but if you learn
more you too will think the same'.
If this didn't come about, that would mean that he couldn't for example learn
history. (OC, §206)4
The grammar of reliance fits a situation in which I take up an attitude
 On the Attitude of Trust 315

towards a person because I believe certain things about him; a situation

involving trust, on the other hand, is one in which I may come to
believe certain things because I have an attitude towards a person.
When I trust someone, it is him I trust; I do not trust certain things
about him. If I trust someone there cannot be certain respects in which I
distrust him. Distrusting him would mean that I myself retained the ulti-
mate judgment concerning the respects in which he was to be trusted; but
then I would not really have placed my trust in him. (I may, however,
consider someone I trust partly unreliable: this is so because I may not
rely on him to be himself in all circumstances. On the other hand, of
course, I may rely on someone I do not trust.)
Perhaps this could also be expressed as follows: when I trust someone,
he as it were embodies goodness, or reason, for me. And, as was sug-
Downloaded by [UQ Library] at 16:28 08 June 2013

gested before, it is only because there are people to whom I have this
attitude that I can acquire an understanding of goodness or reason. (On
the other hand, my idea of what is good or reasonable cannot be formed
or influenced through my relying on someone, since my reliance is
simply an expression of what I judge to be good or reasonable to begin
with. Of course reliance may grow into trust.)
The contrast between reliance and trust could be summed up as
follows. In relying on someone I as it were look down at him from
above. I exercise my command of the world. I remain the judge of his
actions. In trusting someone I look up from below. I learn from the
other what the world is about. I let him be the judge of my actions.

VII
What has just been said about the attitude of trust would not seem alto-
gether to fit, say, a normal child's attitude towards his teachers at
school. If one can speak of trust in the case of a school child, it is
usually of a heavily qualified sort. So the case seems to constitute a
middle ground between trust and reliance which I apparently have failed
to allow for in my contrasting account of the grammar of the two
attitudes. The point, however, is that a child starting school is not
wholly innocent of the world. For the teacher to demand trust in certain
areas might place him in conflict with the child's trust in his parents, for
instance. Because of this the child's trust in his teacher will, in a sense,
be circumscribed from the start; on the other hand, school-teachers, in
our culture, are not normally expected to challenge the limits of that
trust.
Of course a teacher may establish a rapport with the child which
enables him to win over some or all of the child's allegiance from his
 316 Lars Hertzberg

parents. In some such cases it may be debated whether the teacher is

misusing the child's trust. On the other hand, it could be suggested that
such conflicts of trust are an essential part of the child's growth towards
independence.

VIII
It is important to add another qualification here. In some of the cases I
have cited as instances of the grammar of trust, it would be quite mis-
leading actually to apply that word. Thus, if one were to explain the
conduct of a small infant in mimicking his elders, in being receptive to
their warnings and encouragements, etc., by saying that he trusts them,
Downloaded by [UQ Library] at 16:28 08 June 2013

one would seem to be presupposing that there was something like a

question of choice in the matter. This in turn would presuppose on the
part of the child an understanding of the identity of those to whom he
took up this attitude, that is, some recognition of those confronting him
as human beings, or as these particular human beings. He would also
have to be taken to understand that they had certain intentions, etc. Of
course, the idea that such an understanding might lie at the basis of the
child's relations to his elders is absurd. What is in question are simply
the child's natural ways of reacting to other people. In fact, were the
child to react to his environment in any way other than this, we should
be at a loss to recognize in his activities the behaviour of a human
being. The human infant is not, as it were, an independently intelligible
living unit, and not simply because of the physical cares which he must
receive from others, but because the sense of his activity depends on the
way in which it is interwoven with the activity of others. (The child, we
might say, only gradually becomes independent of his mother's body.)
The upshot of this is that it would be misleading to say, 'The child
behaves in this way because he trusts the adults', rather he simply
behaves in this way, and out of this relation (perhaps it could be less
misleadingly described as an absence of distrust) there gradually evolve
attitudes which may be called trustful.5

IX
An alternative way of expressing the point just made would be to say
that, when a child develops attitudes towards his elders that are justly to
be described as trustful, the trust does not arise, as it were, as some-
thing added to the child's earlier relation. It arises, on the contrary, to
 On the Attitude of Trust 317

the extent to which .the child's relations remain unaffected in spite of the
changing surroundings. (If there is a difference in quality between the
infant's dependence and the trust of the maturer individual, we might
say, it simply consists in the fact that, while the relation itself remains
the same, it has now become conceivable for it to be different.)
Again, a different way of making this point is to say that what stands
in need of explanation, normally, is not the fact that someone trusts
another person, but the fact that he distrusts him. Here, again, is a con-
trast between the grammar of trust and reliance. We saw that, for
reliance to be intelligible, it must be seen as an exercise of a person's
judgment concerning others. Trust on the other hand is implicit in many
of the primary reactions of one human being to another. It is the loss of
this way of reacting that has to be made intelligible, by invoking, say,
Downloaded by [UQ Library] at 16:28 08 June 2013

the effects of experience or instruction.

Being prepared to go along with what another intimates - to comfort
him if he expresses pain or grief, to return a smile, to approach if he
beckons, to follow the direction of his eyes, to stop short if he frowns -
is what basically constitutes understanding these human forms of
expression, and is accordingly part of what it means to see another as a
human being.
In this connection, consider the following passage from Wittgenstein's
notes, 'Cause and Effect: Intuitive Awareness':

The game doesn't begin with doubting whether someone has a toothache,
because that doesn't - as it were - fit the game's biological function in our life.
In its most primitive form it is a reaction to somebody's cries and gestures, a
reaction of sympathy or something of the sort. We comfort him, try to help
him. -
. . . [I]magine a mother whose child is crying and holding his cheek. One kind
of reaction to this is for the mother to try and comfort her child and to nurse
him in some way or other. In this case there is nothing corresponding to a doubt
whether the child is really in pain. Another case would be this: The usual reac-
tion to the child's complaints is as just described, but under some circumstances
the mother behaves sceptically. Perhaps she shakes her head suspiciously, stops
comforting and nursing her child - even expresses annoyance and lack of
sympathy. But now imagine a mother who is sceptical right from the very begin-
ning: If her child cries, she shrugs her shoulders and shakes her head; sometimes
she looks at him inquiringly, examines him; on exceptional occasions she also
makes vague attempts to comfort and nurse him. - Were we to encounter such
behaviour, we definitely wouldn't call it scepticism; it would strike us as queer
and crazy.6

Let us bring this point into connection with the following remarks,
occurring in a later context:
 318 Lars Hertzberg

Do I pay any mind to his inner processes if I trust him? If I don't trust him, I
say, 'I don't know what is going on inside him'. But if I trust him, I don't say
that I know what is going on inside him.
If I don't distrust him, I don't pay any attention to what is going on inside
him.- 7

Wittgenstein seems to be suggesting that one of the sources of the idea

of the inner as something hidden, that is, of a possible discrepancy
between what a person really is and what he appears to be, is the
distrust of one human being for another. It is not that, in trusting some-
one, we take for granted that the inner is in alignment with the outer
(such a description would rather fit a case of reliance: say, the way two
businessmen might rely on one another in closing a deal); rather, we
Downloaded by [UQ Library] at 16:28 08 June 2013

take his conduct at face value, giving no thought, so far, to the inner.
Or better still: to this extent there simply is no distinction for us between
the inner and the outer. (There are, of course, additional sources of the
idea of the inner, such as the ambiguity of a person's conduct. That is,
there may be independent difficulties arising out of the question: what ¿s
the face value of his conduct?)
Distrust is not conceivable, however, until a fairly advanced stage of
human relations.8 The little child is no more capable of suspicion than
he is capable of deceit. Distrust is an outcome of experience, that is of
disappointment. The child's shyness of strangers is not a case of distrust.
If the child is merely shy, he is not awaiting proof of the stranger's good
intentions. When the child finally does come forth, the fact that his
responses to the other were delayed does not mean that they are not
spontaneous.9 Being on one's guard is a different thing altogether. The
child learns to be on his guard through being shaken out of his trust;
say, through finding himself an object of ridicule, malice, envy or
suspicion. Whether he will have that experience depends partly on the
circumstances, partly on his character; for as was said before, circum-
stances by themselves are never sufficient to force one to give up one's
trust. When it happens the child learns to conceal himself and learns to
think of others as not being open with him. (The interdependence here
is all-important.)

X
It belongs to the grammar of reliance, we have argued, that the ration-
ality of one person's relying on another is judged on the basis of the
grounds available to him. How are we to judge the merits of trust, since
it is not based on grounds? (I assume the case we are considering is that
 On the Attitude of Trust 319

of someone at a stage sufficiently advanced for the possibility of his

withholding his trust from someone to be intelligible.) It would be in
line with what seems to be a prevalent attitude among philosophers to
conclude that, since trust is not based on grounds, it is never a rational
option. Malcolm discusses such a view in his paper, 'The Groundlessness
of Belief, in which he criticizes the idea that '[t]he objective, mature,
strong attitude is to hold beliefs solely on the basis of evidence'.10 I
would argue that ideas like this are due precisely to a failure to distin-
guish the grammar of trust from that of reliance. The prevalence of this
failure, it appears, is not due solely to a lack of philosophical acumen,
but seems to be in keeping with a widespread attitude to life, with what
might be called a belief in the general superiority of suspicion over trust.
This attitude is castigated by S0ren Kierkegaard in the following passage
Downloaded by [UQ Library] at 16:28 08 June 2013

of Works of Love:

All men have a natural fear of making a mistake - through believing too well of
a man. The mistake, on the other hand, of believing too badly of another man,
is perhaps less feared, at least not in comparison with the first. But then we do
not fear most of all being in error, then we are, on the contrary, in error
through our one-sided fear of a certain kind of error. It mortifies our vanity and
our pride to have thought too well of a swindler, to have been foolish enough to
believe him - for it is a contest between brain and brain. One is vexed at oneself,
or at least one finds that it is . . . 'so stupid' to have been fooled. But ought it
not to seem equally stupid to us, to say the least, to have believed evil, or sus-
piciously to have believed nothing, when there was good! . . . But here in the
world it is not 'stupid' to believe evil of a good man; it is a superciliousness by
which one adroitly gets rid of the good; but it is 'stupid' to think well of an evil
man, that is why one protects oneself - since one fears so much being in error.
He who has love, on the other hand, truly fears being in error, therefore he
believes everything.11

The inclination to view trust as a weakness, I would contend, is based on

a failure to take note of another facet of the contrast between the gram-
mars of trust and reliance. This concerns the way responsibility is to be
allocated. If I was wrong in relying on someone, this was a failure in my
judgment: it would have been better had I been more astute. When
someone's trust has been misplaced, however, it is always, I want to say,
a misunderstanding to regard that as a shortcoming on his part. The
responsibility rests with the person who failed the trust. The reason for
this is that, unlike reliance, the grammar of trust involves a. perspective
of justice: trust can only concern that which one person can rightfully
demand of another. So if I claim that A trusts B to do X, I commit
myself to thinking that (in the absence of exonerating circumstances) B
is open to blame if he fails to do it. If on the other hand I think it would
be wrong for B to do X, or if I do not agree that A can demand it, then
 320 Lars Hertzberg

I will not think about A's expectation as a case of trust, even if that
were the way in which A represents it to himself.12
The responsibility involved in being an object of trust is apparently
what gets expressed in our speaking of a person as having (possessing)
another's trust, as well as of trust being misused - figures of speech
which do not seem to have any natural application in the context of
reliance.
The distrustful person is someone who has been damaged by other
people. No matter that this, to an extent, is what happens normally in
the process of growing up: the destruction of our trust in others is a tra-
gedy of life, a fall from grace. We can find expressions of such a view in
Wittgenstein's notebooks. In some of the remarks published in Culture
and Value,13 he comments on the attitudes of openness and distrust.
Downloaded by [UQ Library] at 16:28 08 June 2013

Thus, in a moving passage, written in 1944, he speaks about the effect

that religious belief may have on a man's relations to others:
Someone who . . . opens his heart to God in confession lays it open for other
men too. In doing this he loses the dignity that goes with his personal prestige
and becomes like a child. That means without official position, dignity or dis-
parity from others. A man can bare himself before others only out of a particular
kind of love. A love which acknowledges, as it were, that we are all wicked
children.
We could also say: Hate between men comes from our cutting ourselves off
from each other. Because we don't want anyone else to look inside us, since it's
not a pretty sight in there.
Of course, you must continue to feel ashamed of what's inside you, but not
ashamed of yourself before your fellow-men. . . .

In a passage written two years later, on the other hand, he speaks about
the growth of distrust in a context which suggests that he sees it as
linked to the descent into madness:

Madness need not be regarded as an illness. Why shouldn't it be seen as a

sudden - more or less sudden — change of character?
Everybody is mistrustful (or most people are), perhaps more so towards their
relations than towards others. Do they have any reason for mistrust? Yes and
no. Reasons can be given, but they are not compelling. Why shouldn't a man
become much more mistrustful towards others? Why not much more withdrawn?
Or devoid of love? Don't people get like this even in the ordinary course of
events?

XI
These connections between trust, character and the quality of human
relations are not discussed in the remarks brought together in On
 On the Attitude of Trust 321

Certainty. The emphasis in that work is an intellectual one, and these

issues would not seem to have a natural place in that context. However,
if in reading it we keep at the back of our minds aspects of the attitude
of trust like the ones that have been discussed in this paper, it will per-
haps be easier to keep in sight the continuity of On Certainty with
Wittgenstein's earlier work.14

NOTES
1 Norman Malcolm, Nothing is Hidden (Oxford: Basil Blackwell, 1986), p. 234. The con-
cluding phrase is a quotation of OC, §637.
2 I might mention G. E. M. Anscombe, 'Faith', in her Collected Philosophical Papers III
(Oxford: Basil Blackwell, 1981), pp. 113-20; Stanley Cavell, The Claim of Reason -
Downloaded by [UQ Library] at 16:28 08 June 2013

Wittgenstein, Skepticism, Morality and Tragedy (Oxford: Oxford University Press, 1979),
pp. 390-3; D. W. Hamlyn, The Theory of Knowledge (London: The Macmillan Press,
1970), pp. 89-90; Anthony Kenny, Faith and Reason (New York: Columbia University
Press, 1983), passim. An exception is Annette Baier's paper 'Trust and Antitrust', which
unfortunately came to my attention only after I had completed mine (Ethics 96 [1986], pp.
231-60). She, too, notices the philosophical neglect of the topic and offers some interesting
suggestions concerning the reasons for it. As for her own account of trust, while there are
a number of points that I agree with and many that I find illuminating, her general view
seems to be rather different from mine. Putting my disagreement concisely, she seems to
me not to do full justice to the basic nature of trust.
3 I thank Peter Winch who made me realize the necessity for restating my argument on
this point.
4 Cf. also OC, §283.
5 The difficulty of formulating a just description of the infant's relation to other people has
analogies with the difficulty incurred by G. E. Moore in using the word 'know' to
characterize his relation to the proposition that he was a human being, had two hands,
etc. Wittgenstein's grappling with this difficulty, of course, is what gives On Certainty much
of its peculiar tension. In doing philosophy, we are tempted, in these and similar cases, to
apply descriptions to phenomena which are too basic, or primitive, for the contraries of
those descriptions to have any intelligible application. (Cf. 'The baby trusts no one'; 'He
trusts only himself, etc.) Yet in the absence of intelligible contrasts, the appearance that
anything is being said in using those descriptions must evidently be an illusion. This
difficulty is connected with issues concerning the language of philosophy and its relations
to language in use which it would take me too far afield to discuss here. These issues are
addressed in the interchange between John W. Cook and Norman Malcolm, in Philo-
sophical Investigations 3-4 (1980-81), by Elizabeth Hankins Wolgast, in Paradoxes of
Knowledge (Ithaca/London: Cornell University Press, 1977), ch. VI, as well as by Peter
Winch in his contribution to the present collection.
6 Philosophia 6 (1976), p. 414. (Trans. Peter Winch.)
7 Ludwig Wittgenstein, Remarks on the Philosophy of Psychology II (Oxford: Basil
Blackwell, 1980), trans. C. G. Luckhardt and M. A. E. Aue, §§602-3.
8 Once the idea of a discrepancy between the inner and the outer is grasped, however, it
can also be used to preserve trust. This point is made by S0ren Kierkegaard as follows: 'If
anyone thinks that one ought not to believe even the best of men because of the possibility
that he might prove a deceiver, then this is also true of the converse, that you can expect
good in even the worst of men, for it would be possible that his baseness was only an
appearance of evil' (Works of Love [Princeton, N.J.: Princeton University Press, 1946],
trans. David F. and Lillian M. Swenson, p. 184).
9 An infant at a certain age is frightened by strangers. Perhaps it will be argued, in view of
this, that the child's primary reaction might be distrustful rather than trustful. There is,
 322 Lars Hertzberg
however, an important asymmetry here. Fear, in a sense, is barren: the child will not be
brought to see, through fear, that some people are to be trusted. From those he trusts, on
the other hand, he may learn that some people should be feared. (The child who is only
fearful will not even learn distrust.)
10 In Norman Malcolm, Thought and Knowledge (Ithaca/London: Cornell University
Press, 1977), p. 204.
11 Op. cit. I have in some points deviated slightly from the Swenson translation.
12 What makes B's failure blameworthy is sometimes, though not always, the fact that A
has the expectation. This is why some cases of trust presuppose that A and B are in a
direct relationship. Promises are a case of this.
13 Culture and Value (Oxford: Basil Blackwell, 1980), ed. G. H. von Wright, trans. Peter
Winch. The passages quoted appear on pp. 46 and 54 respectively.
14 I thank David Cockburn, as well as the participants at the seminar in Skibotn, for use-
ful comments. I have not been able to do all of them justice.
Downloaded by [UQ Library] at 16:28 08 June 2013

Received 22 January 1988

Lars Hertzberg, Smedsgatan 4, SF-10600 Ekenäs, Finland