Can I know what another person is thinking or feeling? If so, how?
The problem of Other Minds is a true
philosophical enigma. It is apt to strike children with no philosophical
education whatsoever, yet remains intractable to many academics. Broadly
speaking, the problem can be divided into three questions. Firstly, how do
I come to believe that there are minds in the world other than my own?
Secondly, how can I justify my belief that there are minds in the world
other than my own? Thirdly, what can I state about the mental states of
minds other than my own?. The question we are dealing with here falls
largely into the third category, although of course issues relating to the
other two will also be involved.
Firstly, it is imperative to assert that, in looking for 'knowledge', we
are not aiming for logical certainties - we are not aiming to show that
any propositions about other minds can be demonstrated with absolute
certainty equivalent to that of mathematical truths. Philosophy ever
since Descartes has tended to be defined by scepticism: either it aims to
produce sceptical theories or it aims to refute them. And sceptics tend
towards extremity in their doubts. It must be stated here and now that
there are not, and never can be, any theories that prove demonstratively
that other minds exist, or that I know others' mental states. This is not
what should be aimed at in attempting to solve the problem. As Austin puts
it "To suppose that the question 'How do I know that Tom is angry?' is
meant to mean 'How do I introspect Tom's feelings?' is simply barking up
the wrong gum-tree."
Most philosophers agree that their theories only bestow a greater or
lesser amount of probability onto statements about other minds (although
there are exceptions, e.g. Peter Strawson's attempt to argue
transcendentally for the existence of other minds through our own
self-consciousness). There have been a number of different attempts to do
this. J.S. Mill, who produced the first known formulation of the Other
Minds problem, used the so-called 'Argument from Analogy' both to explain
how we come to believe in other minds and to justify this belief. Briefly,
the argument holds that I am directly aware of mental states in myself,
and I am aware of the behaviour of mine that results from and is caused by
these mental states. As I can observe similar physical behaviour in
others, I draw the analogy that it is caused by the same (or at least
similar) mental states to my own. As in all arguments from analogy, I
assume that because x is similar to y in some respects, it will be similar
in others. So as I know how I behave if I am feeling, say, angry, I assume
in someone else's case that his behaviour is an indication of the mental
state I call 'anger'. My opinion in this respect is aided by the fact that
most humans' behaviour when they claim to be angry is broadly similar.
The argument from analogy, also employed by Bertrand Russell in a slightly
simplified form, is subject to a devastating criticism. Unlike most
analogies, in the case of other minds, there is no conceivable way of
verifying the conclusion we make. We have no way of discovering whether
someone else is angry or not, and our position means that this is a
necessary disadvantage. The only way to have someone else's experiences
would to become that person, and in doing that, I would no longer be
myself and I would no longer be having someone else's experiences. Thus it
is impossible to conceive of any set of experimental circumstances under
which I would be able to ascertain whether or not the human who is
expressing anger-behaviour really is angry or not. And as Norman Malcolm
has pointed out, as there are no conceivable criteria I could use to
determine whether someone is angry or not, simply claiming that they are
angry is a meaningless statement.
Many philosophers, perceiving this fatal flaw in the argument from
analogy, have attempted to produce theories on other minds that are not
based on analogy. Malcolm himself held that the problem lies in the belief
that in looking for evidence of other minds, we need to start off from our
own case and then look for evidence that other cases resemble my own in
other humans. He claimed, characteristically following Wittgenstein, that
statements about mental states in others have no 'special' status but
rather that they are 'primitive, natural expressions' of the state in
question. In other words, 'my leg hurts' is equivalent to non-verbal
behavioural expressions of having a painful leg such as crying, limping,
or holding my leg. The statements are not propositions as such, and so
have no 'truth-value'. In my view, there are huge problems with this
account. Firstly, its explanatory power is exceedingly limited as it makes
no distinction between those who are pretending to be in a mental state
and those who genuinely have it. How does it help us to believe that our
world is not populated by robots? Secondly, it does not sufficiently
explain how we came to attach the words we do to our mental states. Crying
and limping are 'natural', animistic reactions to pain, but language is
learned from others. How can this be accounted for?
Other philosophers have been less successful in escaping the clutches of
the argument from analogy. H.H. Price, in his article Our Evidence for the
Existence of Other Minds, seems to dismiss it, but then employs it
himself, simply changing the terms of the analogy, claiming that we come
to believe in other minds through other humans' use of informative
language, not through their behaviour.
A.J. Ayer, in his essay One's Knowledge of Other Minds, argues that the
belief in other minds is at least as justifiable as any other inductive
argument. When we refer to the mental states of others, the descriptive
content of that reference need not necessarily include any reference to
the possessor of that mental state. There is no contradiction in asserting
that I could have had that mental state. Implicit in this argument is
Ayer's belief that a person is no more than the aggregate of all his
properties. Thus, as none of those properties are necessarily unavailable
to me, I make no contradiction when I say that I could have had them:
"But even if my friend has no properties which make him an exception to
the rule about feeling pain, may he not still be an exception just as
being the person that he is? And in that case how can the rest of us know
whether or not he really does feel pain? But the answer to this is that
nothing is described by his being the person that he is except the
possession of certain properties. If, per impossible, we could test for
all the properties that he possesses, and found that they did not produce
a counter-example to our general hypothesis about the conditions in which
pain is felt, our knowledge would be in this respect as good as his: there
would be nothing further left for us to discover." (pp 213-4).
And thus, if I could have had the mental states in question, I could be
the person who had them. And if I could be that person, I could verify
whether that mental state actually exists or not. Ayer's reasoning seems
valid enough, but it is hard to know precisely what he means. It seems
certain that in referring to mental states, it is implicit that someone
owns (or is) the mind in which those states are occurring. Although Ayer
is right in his claim that we need not refer to the 'owner' of the state
when we talk about the state itself, and therefore that the owner 'could'
be us, this doesn't seem to address the issue at hand. The problem is one
of other minds, and we are, all of us, in a situation where we find
ourselves confronted with apparent minds other than our own which are
>From the realisation that a belief in other minds can only arise through
observation of the behaviour of others arose the 'cul-de-sac' philosophy
of logical behaviourism. This theory, now largely discredited, holds that
all statements about mental states can be translated, without loss of
meaning, into statements about observable behaviour. Thus to say that
Jones is in pain is to say that (for instance) Jones is wincing, crying
out, grimacing etc. The statements are equivalent, and consequently the
problem of other minds is not so much solved by behaviourists as
dissolved. But the terminal problem for behaviourists lies in the case of
first-person psychological statements. We certainly don't learn about our
own mental states by observing our own behaviour. When I say 'I have a
headache', I don't mean that I am clutching my head, that I am taking
aspirin etc. The feeling of the headache seems in some way to pre-empt all
of this behaviour, and generally to be the primary cause of it. The
behaviourists made a valiant attempt to solve the problem of other minds
by doing away with the asymmetry between my mental states (normally taken
to be learnt through introspection), and the mental states of others
(normally taken to be learnt through introspection), but they ultimately
failed because their account of first-person psychological statements was
Wittgenstein, in his 1953 work Philosophical Investigations, attempted to
show that the construction of a private language (a language that no-one
other than the creator is logically capable of understanding) was
impossible because languages must follow rules, and it would be impossible
for a language with no external reference to follow rules. For instance,
if I have a certain experience x one day and call it 'pain', and then have
another experience y the next day which happens to be different to the one
I had the day before but which seems to me identical, and so I also call
it 'pain', how, as far a I am concerned would this situation differ from
one in which the second experience was actually x? It would not, so I
could conceivably be wrong in every statement I make regarding my own
mental states. The point Wittgenstein is trying to bring out is that,
contrary to the philosophies of Cartesianism and traditional empiricism,
the language we couch our mental statements in is a public language: the
words we use only acquire their meaning through public usage. And thus if
there were no other minds in the world other than our own, we could not
make publicly understandable statements about our mental states. This is a
powerful argument, although it is open to at least two criticisms.
Firstly, it is claimed by some philosophers that it leads inexorably to a
form of behaviourism in which my knowledge of my own mental states through
introspection is not accounted for. Secondly, the argument tells us very
little about the content of other minds. What is the relation between
words and mental states?, and more importantly, how could it conceivably
be discovered? By appeal to our own case? That would just beg the
Nevertheless, I believe that Wittgenstein's approach is the correct one.
It is a truism to say that I cannot have your experiences, and it always
remains logically possible that an malin génie has set me amongst a world
of unthinking, unfeeling robots who have been programmed to exhibit
behaviour (both verbal and non-verbal) leading me to assume that they have
thoughts and feelings. But this seems somewhat unlikely. Rather, it seems
more sensible to believe that there are minds animating all the beings
around me. If I am not to be subject to the sorts of criticisms that
Malcolm makes of the argument from analogy, then I must define some sort
of criteria for asserting that Mr X is feeling y etc. And my only
criteria, of course, can be my observations of his behaviour. To ensure
that my interpretations are as accurate as can be, I must take into
account everything about him: his simple overt behaviour, his environment
and social context, his biological make-up etc. And my reference point can
only be myself, as I am the only example in the world of whose mental
states I can be confident. So as far as X's circumstances and behaviour
mirror my own, I can say that his mental states mirror mine. And I can go
slightly further: I can claim that my knowledge of my own case can give me
an idea of how certain properties and the connections between them can
affect mental states, and so if I notice some property x in X that happens
not to belong to me but of which its significance concerning mental states
I am aware, then I can make statements regarding X's mental states with a
certain confidence. If this conclusion seems rather weak, then I can only
appeal to the enigmatic nature of the problem itself and ask others to
Finally, I would like to mention an observation of mine regarding the
nature of the problem itself. The problem of other minds arises because we
have no certain criteria for ascertaining the possessing of a mind (or
mental states) by a being, and this in turn arises because we do not know
precisely what a mind is. If philosophers of mind ever produce a theory of
mind which provides us with knowledge, to a certain extent, of the nature
of mind, then it would at least theoretically be possible to have criteria
for the possessing of mind, which would provide very good foundations for
the solving of the problem of other minds.
Sartre by Arthur C. Danto. Fontana, 1975.
Strawson's Transcendental Deduction of Other Minds by J.L. Martin in New Essays in
the Philosophy of Mind ed. John-King Farlow and Roger A. Shiner. Canadian
Association for Publishing in Philosophy, 1975.
The Philosophy of Sartre by Mary Warnock. Hutchinson & co., 1965.
Analogy by Bertrand Russell in Essays on Other Minds ed. Thomas O. Buford.
University of Illinois Press, 1970.
Our Evidence for the Existence of Other Minds by H.H. Price in Essays on Other
Wittgenstein's Philosophical Investigations by Norman Malcolm in Essays on Other
Knowledge of Other Minds by Norman Malcolm in Essays on Other Minds.
Other Minds by J.L. Austin in Austin's Philosophical Papers ed. J.O. Urmson and G.J.
Warnock. Clarendon Press, 1979.
One's Knowledge of Other Minds by A.J. Ayer in Philosophical Essays by Ayer.
Macmillan and co., 1953.
The Mind and its Place in Nature by C.D. Broad. London, 1925.
The Concept of Mind by Gilbert Ryle. London, 1949.
The Oxford Companion to Philosophy ed. Ted Honderich. Oxford, 1995.