empiricism. Any view which bases our knowledge, or the materials from which it is
constructed, on experience through the traditional five senses. What might be called the
classical empiricist view is associated especially with Locke, the first of the so-called British
Empiricists, though elements of it go back much earlier. It found itself in a running battle
with *scepticism, which led it to become more extreme, especially in Locke's successors
Berkeley and Hume, with echoes early in the twentieth century. This in turn led to a critical
reappraisal and severe reining-in of empiricism by Kant, and later, after the twentieth-century
revival, by Wittgenstein. A more sober empiricism, however, is much more widespread,
though its very sobriety puts it in some danger of losing its distinctive nature as empiricism.
What follows is intended to fill out this picture, ending with a few miscellaneous points and
distinctions.
Empiricism has its roots in the idea that all we can know about the world is what the world
cares to tell us; we must observe it neutrally and dispassionately, and any attempt on our part
to mould or interfere with the process of receiving this information can only lead to distortion
and arbitrary imagining. This gives us a picture of the mind as a 'blank tablet' (*tabula rasa)
on which information is imprinted by the senses in the form of *'sense-data', to use a
technical term invented in the nineteenth century and not to be confused with the wider and
vaguer term 'the data of the senses'. Previously the term *'idea' had been used in this sense,
though confusingly in others as well. Sense-data were therefore the 'given', prior to all
interpretation, and the mind, which now and only now became active, manipulated these
sense-data in various ways, combining them or abstracting from them, to form the great bulk
of our ideas and concepts, and then went on to discover relations between these ideas, or to
observe further manifestations of them in experience and relations between these
manifestations.
This in varying versions is the classical empiricist view. It leads straight off to problems
involving scepticism, for if the mind is limited in this way and must rely entirely on these
ideas or sense-data, how can it ever know anything beyond them? They are
supposed to 'represent' an outer world, but how can the mind know that they do any such
thing? Indeed how can it know what is meant by talking of an 'outer world' at all? Locke
himself, at least as traditionally interpreted, seems not to have taken these problems very
seriously, but they come fully to the fore in his successors, especially Berkeley and Hume.
Empiricism becomes more extreme when it abandons the claim to know an outer world at all,
and insists that what we call the outer world is simply a construction by our minds,
indistinguishable from a real outer world in practice. But can this view be coherently stated
at all? If we have no knowledge whatever of anything beyond our own experiences, how can
we even envisage the possibility of something beyond them in order to contrast them with it?
How would we understand what it was we were envisaging? This is an example of a move
very common in philosophy, whereby a theory is accused of being unable, on its own terms,
to state itself coherently. It is developed, in various ways, both by Kant and by Wittgenstein.
There is another objection too to this extreme kind of empiricism, because it is not obvious
that sense-data of the kind required by the theory can exist. They are usually supposed to be
things that are exactly as they appear. Since their being just consists in their appearing to
some mind they can have no hidden depths that that mind could fail to know about, and they
cannot fail to have whatever properties they appear to have. Our knowledge of them must be
incorrigible, i.e. it does not make sense to say that we might be wrong about them (about
those that appear to ourselves, that is; we might go wrong in our guesses about those that
appear to other people, but it is not clear how we could know of the existence of other
people). We can, and of course do, have sensory experiences, but what is not clear is that
what these are of is certain objects which we cannot go wrong about. As Wittgenstein
claimed, and surely with some plausibility, what we cannot go wrong about we cannot go
right about either; there is simply no room for anything that could be called judgement or
knowledge. An image can exist on a camera plate, but the camera does not 'know' the image,
and can no more be right about it than wrong about it. Of course when presented with a
brightly coloured object I can hardly in practice go wrong if I claim 'This is red'. But I could
in principle be confused about just what counts as being red—and might be confused in
practice if I ventured as far as 'This is scarlet'. Such confusion need not be merely linguistic,
or about the meaning of the word 'scarlet'; I might well become persuaded that the thing I
called scarlet had not in fact really appeared to me in the same way as things I had previously
called scarlet. We may remember too the difficulty aspiring painters have in 'seeing things as
they really look'; if taken literally this would be an illusory goal to seek (Gombrich).
Extreme empiricism of this sort then seems to be incoherent. By insisting that we know
everything through experience it makes us start from a position of total isolation from the
world, and then it becomes miraculous that we could ever escape from there. We are locked
into a castle surrounded by a moat, and the ideas or sense-data that we hoped to use to bridge
the moat turn into a drawbridge and fly up in our face. Evidently we must start from within
the world itself, which means that in some sense we must already know some things, without
having to find them out. It is not that we must have some magical armchair access to the
world—that would be to put us behind the moat again but supplied now with a magical
bridge, across it. Rather we must come to the world armed i with certain ways of looking at
it, and without insisting that our knowledge must always start-with something we can know
incorrigibly (a view known as *foundationalism). The mind must be active not just, as Locke
thought, in manipulating: and building on an experience already received passively, but also
in receiving that experience itself.
This at any rate is the sort of reaction to extreme empiricism that we find in writers like Kant
and Wittgenstein. But so far we have only been concerned with empiricism taken to its
limits. Many philosophers and many features of a philosophy, or approaches to a question,
can be called empiricist without involving this whole story. Empiricists may, for instance,
confine themselves to opposing the more extreme forms of *rationalism. Or they may allow
that the mind is active in the way mentioned above, but insist that there are no a priori truths,
i.e. truths that can be known without recourse to experience; apparent exceptions such as the
truths of mathematics and logic they will regard as not really truths in any substantive sense
at all, but more like rules of procedure, so that 'Twice two is four' means something like
'When confronted with two things and two things assume you have four things'. Probably
most philosophers would regard themselves as empiricists to some degree, if only because
refusing to do so might suggest adherence to an extreme form of rationalism. But the
distinction between empiricism and rationalism is wearing thin for reasons connected with
the challenges recently mounted against the analytic-synthetic distinction, and one motive for
refusing to call oneself an empiricist (or a rationalist for that matter) is that it suggests that
one accepts that distinction. But even with regard to the older philosophers the traditional
contrast between 'British empiricists' and 'continental rationalists' cannot be regarded as
anything but a rough label of convenience, however true it may be that, as explained above,
empiricism in particular reached a zenith among the former.
Also one should distinguish between empiricism as a psychological doctrine of how the mind
acquires the contents it has and empiricism as a doctrine of justification, about how we can
justify our various claims to knowledge. However, these questions are often run together,
especially in older writers, and indeed they have not always been kept apart in the present
article. Furthermore though the two questions are conceptually distinct, and for much of the
twentieth century in particular the distinction has been rigorously insisted on, more recently
the tendency has been revived, though this time overt and acknowledged, to run the questions
together, or else to assert that the latter (concerning justification) cannot be answered and
must be replaced by the former (concerning origins and development).
One further sphere in which the relevance of empiricism may be mentioned is ethics. If we
cast cheerfully aside the bogy represented by the *naturalistic fallacy we might define 'good'
in terms of something like 'catering for certain interests', and then perhaps define 'right' as
something like 'tending to maximize good'. If we insist that this is what the terms mean, so
that the definitions are simply a matter of semantics, we can then claim that ethics has
become an empirical subject, assuming at any rate that it is an empirical matter what things
count as interests and for whom. Of course whether we should take this line is another
question.
Various types of empiricism have been singled out from time to time and given special
names. *Logical Positivism is a type of empiricism, and indeed is one of the main forms that
extreme empiricism has taken in its revival during the twentieth century. Because it concerns
the meanings of words or sentences it has sometimes been called *logical empiricism, just as
Logical Positivism itself is so called for that reason. One Logical Positivist in particular,
Moritz Schlick, dignified his own version of the theory with the name 'consistent
empiricism'. One philosophy with some kinship to empiricism is *pragmatism, and William
James called his own version of empiricism 'radical empiricism', though, distinguishing it
from pragmatism. Constructive empiricism is the view, associated with Bas van Fraassen,
that science should aim to construct a theory which will be 'empirically adequate', i.e. will
imply the truth of all that we find to be true when we observe entities that can be observed.
The theory may make statements purporting to claim the existence of unobservable entities
(electrons etc.) and such statements must be taken literally, not analysed as 'really' saying
something different and innocuous; but we can treat it as a good theory, and accept it for
scientific purposes, without believing it.
A.R.L.