The Nature Of Knowing

THE BOOK WAS
DRENCHED

<s^ -

w ^

[<OU 164482 ^
>Qi ~ ;o

OSMANIA UNIVERSITY LIBRARY

l No. &/ &* /^^Accession No.

Author

This book should be returned on or before the date last marked below

THE NATURE OF
KNOWING

THE NATURE OF
KNOWING

By

R. I. AARON

M.A. (Wales), D.Phil. (Oxon.)

Of University College, Swansea

Sometime Fellow of the University of Wales

LONDON

WILLIAMS & NORGATE LTD
1930

PR1NTFD IN GRFAT BRIT UN' BY
UNVVIN BROTHERS LIMIJLD, LONDON AND \\OKING

Contents
PREFACE page 7

INTRODUCTION 9

THE SENSORY EXPERIENCE

Section i. The Nai've View 19

Section 2. Critical Theories 24

Section 3. A Further Critical Position 42

DISCURSIVE REASONING

Section [. The Prior Knowledge of Prin-
ciples 69
Section 2. The 'Experience' Presupposed 82
Section 3. The Nature of Mediate Know-
ledge 95

INTUITIVE APPREHENSION

Section i. 'Transcendent' Knowledge 121

Section 2. The Intuitive Character of the

Knowing Act 138

CONCLUSION 151

Preface

IN the following pages I have set forth in essay form the fruit
of six years' research upon the problem of knowing. These
researches were pursued for the most part at Oxford, and the
present essay is a restatement in, I hope, more explicit and more
adequate terms of the argument contained in a dissertation
which I submitted for the degree of D.Phil, in 1926. I decided
not to publish the dissertation itself. It was lengthy, cumber-
some, and most serious defect of all contained so much
historical detail as to bewilder the reader and to distract his
attention from my main argument. I thought it wiser, there-
fore, to discard the historical matter altogether, especially as
most of it was already familiar enough to any serious student
of philosophy. In the argument of the present essay there are,
for this reason* few historical references ; though a reader who
is acquainted with the philosophical speculations both of the
past and of the present will quickly realize my indebtedness
to others. As to the past, I find my debt greatest to Plato,
Aristotle, Descartes, Spinoza, and Kant.

I have been singularly fortunate in my teachers throughout.
In connection with the present work I have to express my
thanks for suggestions to Professor J. A. Smith and the Master
of Balliol (Dr. A. D. Lindsay), the examiners of my disserta-
tion; to Professor H. A. Prichard; to my colleague, Dr. A. C.
Ewing; to Professor II. H.Joachim; and, finally, to the Provost
of Oriel (Dr. W. D. Ross). The three latter gentlemen have
aided me very considerably, and I am much obliged to them.
My thanks are due to my publishers and their reader for
valuable guidance; also to my sister, Miss E. G. Aaron, for
helping with the manuscript and proofs. Finally, I am grateful
to the Court of the University of Wales for electing me a
Fellow of the University, and to the Provost and Fellows of
Oriel College, Oxford, for a special research grant.

R. I. AARON

YNYSTAWE, SWANSEA
December 1929

Introduction

IT will be the aim of this essay to describe as accurately as
possible the nature of knowing. We say 'knowing' rather than
'knowledge' because the two terms are not always synonymous.
In everyday language the term 'knowledge* may mean one of
three things: firstly, the actual knowing of the object; secondly,
the whole object known; and thirdly, a stock of information
possessed by the mind which it can recall whenever the neces-
sary conditions are realized. In this essay, our immediate con-
cern is with knowledge in the first sense. It is our purpose to
describe knowing; though, obviously, frequent reference to
knowledge in the other two senses will be necessary.

Now the strictest description would be definition, but no
satisfactory definition of knowing is available. This fact may
be attributed tt> one of two causes. It may be held that knowing
is something elemental and therefore cannot be defined in
terms of anything other than itself. For in definition so it may
be argued we express the essential feature or features of the
thing defined in terms whose meaning is already familiar to us.
Thus it belongs to the essential nature of a triangle that it
should consist of three straight lines in certain definite spatial
relations, and the definition of a triangle is only possible when
we know, amongst other things, the meaning of the phrase 'a
straight line*. Now if I were to try to define knowing by saying
that it is, for example, the co-presence of mind and object, this
would presuppose a knowledge of what I wanted to define,
because, if I were asked w r hat the word 'mind' signified here,
the only possible answ r er would be that mind in this context is
the knowing pow r er; that is to say, my definition would be
circular. And in this manner, since knowledge is something
elemental, every attempt at its definition, it may be argued,
would necessarily involve the use of the very term we wish to
define, and, in so far, would prove invalid.

In the second place, the mind seeks definitions because it
hopes thereby to gain both in thought and statement the
precision that it needs. But this clearly implies that definition

10 THE NATURE OF KNOWING

is possible only when we already possess a certain acquaintance
and even familiarity with the thing to be defined. Definition is
the crystallization of our thoughts about any matter by expres-
sing them in terms which are already precisely fixed in our
minds. It can occur only at an advanced stage of knowing.
Now with regard to knowledge of knowing itself we can hardly
claim to be at a sufficiently advanced stage to attempt definition.
For though we are all acquainted with knowing at least, we
believe that we know frequently yet when we begin to reflect
we find it difficult to make clear to ourselves the nature of this
knowing, and the more we reflect, the more difficult does the
task become. None the less, in order to define knowledge it
would seem necessary for us to possess some prior notion of
what it essentially is, for it is only then that we can attempt to
clarify this notion still further by definition, if, that is to say,
we grant that definition is at all possible in this case. But at
our present uncritical stage, we have so vague a notion of what
we mean by knowing that immediate definition is out of the
question. For these reasons, therefore, it is impossible at this
point simply to define knowing and, having thus completed our
task, put down the pen.

Realizing that definition is here out of the question, we must
proceed to our goal, namely, the accurate description of know-
ledge, by a different route. The procedure we propose to adopt
is the careful scrutiny of what are taken to be examples of
knowing in order that by such scrutiny we may, if possible,
lay bare the nature of the knowing involved. We shall, there-
fore, find it necessary as we proceed to subject such instances
of knowing to a critical examination, and to inquire into the
validity of the claims made for them. We shall search for those
experiences which are in the fullest and completest sense
instances of knowing, since it is the scrutiny of such that will
reveal most to us about its inmost nature. Having discovered
such examples, we shall then seek by fair and accurate analysis
to describe the knowing contained therein.

Thfcs it will be understood that our primary aim is not to
explain knowing, but rather to describe it. The question, 'How

INTRODUCTION u

does knowing occur?' can be answered in one of two ways.
Firstly, we may describe the process or processes present in
the mind when it knows ; secondly, we may set forth a thorough-
going metaphysic so as to show the ultimate source of such
knowing and the nature of the universe within which alone
knowing can become possible. The second answer would
provide an explanation of knowledge, whereas the first would
be descriptive only. Now it is not the primary aim of this essay
to supply the reader with 'explanations' in this sense, nor to
answer these profounder questions. The attitude we intend
to adopt is that of 'first things first' ; since it seems foolish and
rash to seek for the metaphysical explanation of that which
as yet we cannot even describe. At the same time, we readily
grant the possibility that no completely satisfactory account
of knowledge can ultimately be given without a metaphysic of
some kind or qther. The distinction, that is to say, between
description and explanation may not hold in the last resort;
every description to be complete may also have to be an explana-
tion. Nevertheless, as we begin on our search for an adequate
description of knowledge, it is as well to point out some of the
more obvious truths first, and to grasp firmly the things that
lie, comparatively speaking, ready to hand, before venturing
into the cavernous depths of metaphysics. For then we shall at
least safeguard ourselves against the unhappy fate of those who
plunge heedlessly into the gloomy darkness of deep speculation
without first securing for their guidance such illumination as a
careful study of what lies in the open can provide. We there-
fore make no apology for the fact that this essay is mostly
'surface' work. The reader should not look to it for ultimate
explanations.

But these words must not be taken to mean that the limited
task before us is an easy one ; on the contrary, real difficulties
present themselves from the first. No sooner do we begin upon
our search than we meet with a serious problem, namely, what
may and what may not be taken for granted at the outset.

It is sometimes supposed that philosophy is unlike th$ par-
ticular sciences in that it takes nothing for granted. This state-

12 THE NATURE OF KNOWING

ment, however, is not absolutely true. Philosophy, perhaps,
takes less for granted than does any particular science ; never-
theless, it always begins by assuming certain positions as yet
undemonstrated. In this matter of epistemology, for example,
unless one is a confirmed and complete sceptic, one must take
the fact of knowledge itself for granted ; while actually, as has
been pointed out long since, even the sceptic who flatly and
explicitly denies this fact is at the very moment of his denial
implicitly presupposing it. For he is claiming to know that there
is no such thing as real knowledge. Either he does not know this,
and then his flat denial becomes impossible; or he does, and
then he himself is possessed of knowledge a fact that contra-
dicts his own denial of it. The case of the agnostic who doubts
without definitely denying the actuality of knowledge is different.
He may refuse to assert anything whatever, preferring to sus-
pend his judgement throughout. He would not then be pre-
supposing the fact of knowledge ; but neither would any inquiry
ever be possible in his case. To carry out an inquiry we must
make assertions which we hold or imply to be true. 1 In order
to philosophize significantly, therefore, it is necessary to take
knowledge for granted from the outset ; we already believe that
the human mind is capable of knowing.

Thus we find it essential in an effort to describe the nature
of knowing to start from a basis which is taken as true, without
being demonstrated to be such. Obviously, we should assume
in this way only the barest minimum necessary, and we must,
furthermore, make clear and definite to ourselves what exactly
it is which we do thus assume. In the first place, as we have seen,
we take for granted the fact of knowing. Our quest is one into
which a person who refuses to make this assumption cannot
enter. We may be sceptical about many epistemological tenets

1 Some of these assertions may be mere opinions about which we
do not feel completely certain. Yet judgements of probability pre-
suppose some certainty. And, in any case, when we express an
opinion, we imply that it may be true that is to say, that we may
have gained knowledge, which in turn implies a belief in the possi-
bility of our gaining knowledge.

INTRODUCTION 13

which are now generally accepted as true, but we cannot be
sceptical about the fact of knowing itself if we wish to proceed
with our inquiry ; for by such scepticism we should be depriving
ourselves of the one faculty whereby the pursuit of the inquiry
becomes possible. If we assert at the outset that we cannot
know, it is then foolish to try to discover w r hat knowledge could
ever be, since the discovering would itself be an example of
knowing.

Belief in the actuality of knowing as a fact of human expe-
rience is thus essential for the further progress of our inquiry;
but if we hold this belief it follows that we already know some-
thing as to the real nature of knowing. The truth seems to be
that, however far back we go, the inquiry into its nature is never
begun from a point at which we know absolutely nothing about
it, as if our minds were in this respect vacant and empty,
waiting to be filled. On the contrary, since in our inquiry we
propose to proceed by seeking for valid examples of knowing
amongst our experiences, we could never begin on this task did
we not already possess some method whereby we might test
the various experiences, so as to discover which were true
instances of knowing and which not. This method may be
modified frequently as we proceed, but the capacity for testing
must be in our possession from the outset of our inquiry; for
otherwise \ve should be incapable of recognizing any instance
of kno\\ing to be such. And, as is evident on reflection, what
we really mean here is that we could never pick out a single
instance of knowing did we not know beforehand some one or
many of its characteristics.

Now what characteristic or characteristics do we look for in
an experience when we seek to discover whether it be an
instance of knowing or not? First of all, clearly, we demand that
the experience should give us assurance amounting to con-
viction. Here, it would seem, is as universal a feature of know-
ing as any. Knowledge is always marked by unwavering con-
viction, and if we doubt, however slightly, we realize that our
state then qua doubting is not one of knowing. That experience
alone can be termed 'knowing' in which we are convinced

i 4 THE NATURE OF KNOWING

beyond the shadow of a doubt. 1 Thus our first test as to whether
we are knowing or not at a particular moment is this one : the
presence or absence of a feeling of conviction. Furthermore,
the judge as to whether I now possess or do not possess this
conviction is, of course, myself. I affirm that such and such an
experience is a true instance of knowing, simply because I am
convinced that it is so. In this matter the mind itself has the
last word. We are not dealing for the moment with the more
difficult question as to the demands that must be satisfied
before the mind can attain to such conviction, for example,
the demand for consistency and such like. We are merely
pointing out the fact that knowing invariably involves a
feeling of conviction, and that the presence or absence of
the latter is indeed our first test of the presence or absence
of knowledge.

From the outset, therefore, we take it to be Mie, firstly, that
knowing is a fact, and, secondly, that a characteristic mark of
it is a sense of certainty. There is no knowing without a feeling
of conviction. At this point, however, an important reservation
must be made, namely, that a like feeling of conviction seems
frequently to occur when actually we are not knowing. Con-
sequently, it is in itself no infallible sign of knowledge. This
greatly complicates our problem, for though admittedly there
can be no knowing without conviction, there may yet (it would
seem) be conviction without knowledge ; that is to say, we may
feel sure that we are knowing at a time when actually, as we
ourselves may be brought to confess later, we are not knowing. 2

1 Incidentally, it is because this is so that one finds so much
difficulty with the phrase 'knowing vaguely'. Knowing is such, we
feel, that it leaves no room for vagueness. If an experience is marked
by a sense of vagueness, then it is not knowledge, whatever else it be.
It does not give conviction.

a At present, however, we have no right to be dogmatic on this
matter. For in spite of first appearances, the question may still be
asked : Is our acquiescence in error absolutely identical in character
with the conviction we feel when knowing, for instance, that two
parallel straight lines will never meet? We must postpone the dis-
cussion of this question. We shall return to it when considering

INTRODUCTION 15

If, on every such occasion, we really were knowing, there would
be no difficulty in discovering any number of instances, but
unfortunately we often believe we know when we do not. It is
this fact of error which makes the problems of epistemology
so desperately difficult.

For, even at this early stage, we realize that knowing cannot
involve error. That is to say, if in any experience we were
convinced of something and thought ourselves to be actually
knowing but later realized that we had been in error, we should
not continue to think of the first experience as an instance of
knowing. Consequently, in seeking for an example of knowing,
it is not correct to accept as such any chance experience in which
we find ourselves convinced of something. On the contrary,
such an experience may easily fail to provide us with real
knowledge, in spite of the fact that at the time we seem to be
convinced that it does. And therefore we need to examine such
instances carefully and not conclude too hurriedly that they
are sound examples of knowing; we need to scrutinize our
convictions and to hold our judgements in suspense that is to
say, to become sceptical, though not with regard to knowing
in general, since, as we have seen, such scepticism would make
all advance impossible, but only with regard to these supposed
instances of knowing. We must learn to stand aloof from them
so as to examine them in as detached a manner as possible.
And as we proceed with this examination, we may hope to gain
greater insight into the nature of knowing itself.

These considerations seem to necessitate a distinction which
we propose to consider more fully at a later stage of the argu-
ment. We must distinguish between a knowing, on the one hand,
which, if it occur at all, is infallible, and a cognitive experience,
on the other, which is definitely fallible. And this distinction
is all the more necessary if we say that the cognitive experience
may contain within it infallible knowing. Our cognitive experi-
ences are fallible, yet they give us occasionally, we must believe,

the character of error. On a prima facie view, however, it seems
almost obvious that we feel just as certain about things when we err
as when we actually do know.

16 THE NATURE OF KNOWING

certain knowledge. But how can this be? How can man gain
certainty when it is patent that his cognitive experiences as a
whole are fallible ? In the third chapter of this work we shall
face this difficulty and suggest a solution. But for the sake of
clarity we recognize here, at the outset, the fallibility of our
cognitive experiences even though we continue to demand an
infallibility for knowing as such.

But, again, if the state of knowing is one in which the knower
is completely and absolutely convinced then we must also
recognize from the very beginning of our inquiry the existence
of a further mental state, which is still cognitive, but in which
we are not fully convinced. It is convenient to term this state,
one of opining, and in terming it so we adopt the practice of
many earlier thinkers. We certainly possess the capacity to
make a judgement whose truth is only probable, and we fre-
quently find ourselves in the state of believing something
without being quite certain about it. Were it to be shown false
(though we hardly think it will) we should not be altogether
unprepared. There are, of course, degrees of probability.
Sometimes we put forward a statement in a very tentative
manner expecting every moment to find evidence brought
against it which will completely refute it. On other occasions,
while w r e admit the absence of theoretical certainty, and while a
measure of doubt still lingers in our minds, we should be very
surprised, indeed, if our belief proved false. None the less, we
are still opining. In both cases our state would be different
from that in which we are completely convinced. This would
have to be recognized even by those thinkers who confine
human knowledge to probability and who deny that man ever
can be completely convinced about anything. Indeed, their
denial assumes the difference, and is not significant unless the
assumption be made. But we, on the other hand, believe that
the human mind does know with certainty, and it is this know-
ing with certainty which provides the subject-matter for the
present inquiry. Therefore, we shall not be directly concerned
with the nature of opining in these pages. Nevertheless, it is as
well to distinguish it from knowing at the very outset.

INTRODUCTION 17

Lastly, we may here add, no knowledge would ever be recog-
nized by us to be such unless its object were the real. This
assumption is so very obvious and trite that it seems hardly
worthy of mention. At the same time, however, there might be,
and indeed is, much disagreement as to the exact meaning and
reference of the term 'real' in this context.

THE NATURE OF KNOWING

THE SENSORY EXPERIENCE
i

The Naive View

WE are engaged upon the search for a fair and adequate example
of knowing, and it might at first be thought that such an enter-
prise need prove neither arduous nor protracted. For the naive
person who has hardly begun to reflect about the nature of
knowing can provide us with a ready example that appears
perfectly satisfactory to him. I have only to open my eyes, he
would say, to know. Seeing is knowing; so also are listening,
smelling, tasting, and touching. I see this paper on which I now
write, 1 hear the bird in the garden, I smell the rose which is
before me, I touch the table and taste the fruit all such
experiences are instances of knowing, and in searching for
examples we need go no farther. Moreover, sense-perception,
such a person might continue, is not merely one example but,
clearly, the only possible example of knowledge, if we confine
the latter term to the knowing of physical objects and of the
world around us. For I can only come into contact with the
external world through the senses. If I were blind, deaf, dumb,
unable to taste and unable to touch, my state would be pitiable
indeed ; not only because I lacked these capacities whilst other
men possessed them, but even more because all knowledge
would be denied me, excepting at most the vague 'inward*
knowledge of my own feelings of pleasure and pain, of joy and
sorrow if it were permissible to suppose that a creature in
this unhappy condition could ever be capable of such emotions.
This immense world in which I live would remain unknown to
me in its entirety. I only begin to know it \<*hen I sense. The
only outlet to the external world is the one afforded by sensation. 1

1 The nai've person is, of course, no philosopher, and I have no
philosophical school in mind. Furthermore, I doubt whether it

20 THE NATURE OF KNOWING

Consequently, it would seem to him, if I deny that perceiving
or sensing 1 is knowing, by that very act I also deny the possi-
bility of any knowledge of the real world outside. And he never
doubts that the real is what I see, touch, taste, and so on. Of
course, I myself do not see all real things; there are many
existences in this vast universe of being which I have not
experienced, and which I am not likely to experience, but if I
do ever come to a knowledge of them it will only be by way of
the senses. Again, the sensing in question need not be mine,
for I can learn by listening to, or reading about, the experiences
of others in conditions completely different from mine. Know-
ledge by hearsay is a very valuable means of widening one's
spiritual horizon. But first-hand knowledge of the physical
world whenever it occurs is, according to the naive person,
invariably sense-perception. And as the latter thus tells us all
we know about the external world it must bb as sound an
example of knowing as is to be found anywhere. Hence an
accurate description of sensing or perceiving would be an
accurate description of knowing, and we need only carry it out
to complete the task we have here set ourselves.

The naive position, it must frankly be admitted, is not with-
out its strength. It satisfies our first demand for conviction on
the part of the knower, for there cannot be a greater degree of
conviction than that possessed by the naive person. In sensing,
he supposes, we experience the physical world exactly as it is,
and each object as it is. In "all the choir of heaven and furniture
of earth" there is nothing that remains hidden from me as a
sentient being once I am in a position (spatially and tempor-
ally) to sense it. At such a time reality lies open before me,

would be at all fair to foister such crude views as these even upon
the much maligned 'man in the street'. I have simply set forth
explicitly for the purposes of the argument a position which is as
such rarely held. Many apparently less crude reflections, however,
when analysed are seen to originate with the implicit adoption of
the above view, and it is the obvious position from which to start
our present inquiry.

1 For the naive person there is no distinction between the two
terms.

THE SENSORY EXPERIENCE 21

and I know it directly and completely without altering it in
any way.

But, what is it that provides the ground for such conviction ?
Clearly the undeniable fact that all sense-experiences are, from
one point of view, infallible. I look out through the window
and see the blueness of the sky and the greyness of the house
opposite. At the moment when this occurs, it is simply impos-
sible to deny that I am seeing blue and grey. On this point
scepticism can never arise. However long I reflect over the
matter I can never bring myself to doubt that I am now seeing
blue. Here surely is something about which I am completely
convinced, for no greater degree of conviction can ever be
possible; and here, the naive person would urge, is an excellent
example of knowing. Furthermore, its excellence is the greater
in that it gives complete certainty without involving me in any
sustained intellectual effort, and is thus as valuable for its ease
and spontaneity as for the conviction it inspires.

In passing, its possession of these features is well illustrated
by the readiness with which certain thinkers make use of
seeing as a metaphor for the supreme kind of knowledge, some-
times termed 'intuition'. The latter is the knowledge w T hich it
is customary to attribute to God and to those higher spiritual
beings who, like Him, do not proceed by way of laborious
processes of reasoning but see all things directly and infallibly.
"In heaven," it was well said by one such thinker, "each being
is, as it were, an eye/' 1 Just as I cannot doubt that I am now
seeing blue, so, too, I could not doubt this supreme knowledge
if I possessed it; and, again, the ease with which I see blue is
analogous to the ease with which I should know if I could
'intuit' in this way. 2 Of course, this does not mean that such
thinkers have ever supposed that seeing itself is in any way
identical with such divine knowing, or even that seeing is
actually a case of knowing. All we wish to point out is that the

1 Plotinus, Ennead IV, iii. 18. ricei Ac . . . oiov 6(f>6a\iiu<;

3 Most of the great philosophers have held that man also may
possess this supreme knowledge, that such 'intuition* indeed is
human knowledge par excellence.

22 THE NATURE OF KNOWING

use of the metaphor would not be possible did not the thinkers
who use it recognize in the sensory experience a certain
infallibility and directness.

Here, therefore, \ve contend, is the strength of the naive
position. I am thoroughly convinced that 1 now see blue, and
there is no room for doubt. If this were all that the position
affirmed there could be no possible objection to it. Actually,
however, the naive person is not content with so limited a
statement. Not only do 1 see blue, but, he would add, I see the
blue sky. Not only do I see red, but I see a red rose. My senses,
it is true, give me colours, sounds, tastes, smells, and certain
touch-feelings. Yet, the naive thinker would say, it is absurd
to suppose that all they give is of such a nature. What I see now
is this table with the books, papers, and vase which are on it.
I see the man in the street, and hear the birds sing. My senses
give me knowledge of physical objects as they are*; they open out
before me the panorama of this actual world in which I exist.
And such knowledge is as completely certain as is the seeing
of blue, being identical \\ ith it since both are examples of sensing.

Now, it is with regard to these additional claims that doubts
arise on reflection. Just as the strength of the position lies in
the presence of a certain infallibility in the sensory experience,
so its weakness lies in a failure to point out carefully enough
where the sensory experience is infallible and where it is fallible.
For, actually, the view that the experience is an indubitable
direct knowledge of the real physical world cannot hold its
ground against the first breath of critical reflection that comes
its w r ay. For, while I cannot doubt that I now see red, it is not
impossible to doubt that I am seeing a red rose. On closer
observation, for example, I may discover that what is before
me is no real flower but only an artificial one, and that I was
deceived at first in thinking it a real one. Or, again, I look across
the Bristol Channel and see in the distance what appears to
me to be land and what I take to be the Devon coast. But as I
look, what I took to be land gradually moves away and I realize
that I have been looking at a cloud. These examples are sufficient
to cause us to doubt the infallibility of the sensory experience,

THE SENSORY EXPERIENCE 23

as providing us with indubitable knowledge of the actual
external world. Whether through the sensory experience we
ever come to know anything about the physical world or not,
it is clear that it does not invariably provide us with certain
knowledge about it, and once we realize this the first naive
position is no longer tenable. 1

And this is true even though we now admit that there is
room for a distinction within the whole sensory experience
between a sensing and an act of judgement. 3 For as against the
above examples it may be objected: "Your whole sensory
experience of, for example, seeing a grey patch and mistaking
it for the Devon coast is much broader than just the bare
sensing. What you actually saw was a grey colour. There is no
error in your seeing the grey ; the error entered when you tried
to give greater significance to this grey patch by judging it to
be the Devon coast. But if you had contented yourself with the
affirmation 'I see grey' there could not possibly have been any
error".

Now if, for the moment, we granted such criticism to be
sound, it still could not be used to bolster up the naive position
in any way, for such criticism cannot have come from the
mouth of the naive thinker; on the contrary, its occurrence is
a definite sign that we are being forced away from that position
and arc leaving it behind us. The naive person has no room
for such distinctions; the possibility of our senses misleading
us has never entered into his head; and the assertion that we
see the colour grey and it alone would appear absurd to him.
We shall suggest later that just seeing grey cannot in itself be
held to be knowledge of the physical world, so that if such a
person did accept the criticism made above he would be giving

1 It, of course, remains none the less true that I do see the colour.
It is also true that I know that I see it. The cognitive act of 'enjoying'
my own experience is already present. But the naive person is inter-
ested solely in my knowledge of the external world, and so disregards,
as we shall do for the time being, this other element of knowing
present in the sense-experience.

3 The name then given to the whole experience as including both
the sensing and the judging is usually 'perception*.

24 THE NATURE OF KNOWING

up his fundamental tenet in its entirety, namely, that in sensing
we know the actual physical world. This would signify, of
course, a complete reversal of his former beliefs, and he would
no longer represent the unreflective type first portrayed by us.
But even if we suppose that seeing grey is actually an example
of knowing the physical world that is, knowing a quality of
(one part of) it the position of the critic is still very different
from that of the naive person who affirms that sensing is a know-
ing of the physical world in its fulness, and who certainly does
not wish to limit that external physical world to a few colours,
sounds, tastes, and the like. Whatever, therefore, we mean by
'seeing grey', the criticism just put forward, if accepted by the
cruder type of thinker, would seriously modify his former
position, so much so that he could no longer be held to be a
representative of the naive view.

We see, then, that the confidence of the naive person in his
belief is indeed justified to some extent by the presence of a
certain infallibility in sensing: we shall reconsider this feature
with more care in the third section of this chapter. On the other
hand, we realize that the sensory experience is no infallible
knowledge of the external world. For the testimony of the
senses is often misleading, and never wholly trustworthy. And
therefore, while we have not decided as yet in what way, if any,
the sensory experience can be accepted as an example and type
of knowing, we feel that the claims made for it by the naive
person are invalid. It is no easy, infallible, and exact knowledge
of the world around us, as the slightest reflection will show.
On the contrary, the deceptiveness of the senses compels us
to reject the naive position as totally inadequate and to demand
a more critical account of sensation. It is to this consideration
that we now turn.

2

Critical Theories

In the previous section the fallacious character of the naive
position has been made clear. It is crude, over-simple, and

THE SENSORY EXPERIENCE 25

completely untenable. Dissatisfaction with it has found outlet
in many and varied critical theories of perception, but it is not
our purpose to set out in detail the history of these theories.
Such an historical account lies beyond the scope of the present
work, though it will be foolish not to make whatever use we
can of the lessons learnt by past philosophers in our own efforts
at describing knowledge as adequately as possible. Sufficient
use, however, will be made of earlier thought if, guided by it,
we lay bare the basis common to all the many criticisms of the
naive position, and if, following the same guidance, we proceed
to discover for ourselves the soundest critical standpoint with
regard to the nature of perception. Our problem, therefore,
may be set forth thus: Can we, rejecting the naive position,
continue to term perception an instance of knowing, and, if
we do so, will the term 'knowing' retain exactly the same
meaning in the* new speculations as it did in those of a cruder
kind?

All critical theories of perception start from the realization
of the occasional deceptiveness of the senses. The view that
the sensory experience provides us with an infallible and exact
knowledge of physical objects in the external world is thoroughly
fallacious. In the example given earlier I affirmed that I saw
the Devon coast. Later I was myself forced to the conclusion
that the affirmation was erroneous, and that my senses had
misled me. This possibility of error, as w r e pointed out previ-
ously, is sufficient of itself to demonstrate, once and for all,
that perception is no infallible knowledge of the world around
us, and that the position of the naive thinker, in spite of his
conviction and dogmatic assurance, is wholly untenable.

But how did it come about that my senses, on this occasion,
misled me, and what exactly do I mean when making that
assertion? In this respect an objection arose in the first section
which, while illustrating the beginnings of the new critical
reflection, throws much light also on the source of error in
sensing. Actually, it was objected, I did not see the Devon
coast. I only saw the colour grey, or speaking still more
accurately a grey patch. Now the error in the whole sensory

26 THE NATURE OF KNOWING

experience described as seeing the Devon coast arose not in
the seeing of the grey patch, but in the judgement which
immediately followed upon the seeing. Given the grey patch,
I endeavoured to discover its significance, and judged it to
be the Devon coast, on the ground of, for example, its similarity
to a grey patch which I had often seen before in that direction
and which I had invariably taken for the Devon coast. And it
was in this effort to discover the significance of the 'given' that
I fell into error.

There is here a clear distinction between a seeing and a
judging within the whole sensory experience of seeing the
Devon coast. We are not usually cognizant of any such dis-
tinction when we actually experience the sensation, for what
we have called the judging requires so little effort on our part,
on account of the very numerous occasions upon which we
have made like judgements, that we are hardly *ware of it until
we are compelled, by some means or other, to focus our atten-
tion upon it. But this distinction, once realized, makes possible
a better understanding of what occurs when, as we say, our
senses mislead us. The error in sensing now reveals itself to
be a case of false judgement ; we affirm something to be that
which it is not. Hence, as the result of this first rough analysis
of the whole sensory experience, we may now more clearly
understand why it cannot possibly provide us with infallible
knowledge. Any and every sensory experience of the kind
illustrated in the above example, we now argue, involves an
act of judgement; it is not mere sensing, it is perceiving; and,
as human judgement is fallible, it follows that the sensory
experience taken as a concrete whole cannot be infallible.
That is to say, being misled by sensing means, in this case,
making an erroneous judgement, which is part of the whole
sensory experience.

Thus, when once the distinction between mere sensing and
judging is set up within the whole sensory experience the
occasional error we discover must be attributed not to the
sensing but to the judging. It might seem to follow from this
that whereas in the judging we sometimes know and sometimes

THE SENSORY EXPERIENCE 27

err, in the mere sensing we never err but always know if we
are agreed that 'knowing' is the correct term for seeing grey,
for example. But here a difficulty ensues. If the mere sensing
is itself knowing, what necessity can there be for this additional
judgement ? Why should we go out of our way to involve ourselves
in judgements which are frequently erroneous if we already know ?

The answer can only be that, whatever we do know by mere
sensing (if we know anything), we do not know all we wish to
know; we are left with a gap in our knowledge which sensing
cannot fill. Consequently, we have recourse to judging. For
though we were for the moment to grant that the mere sensing
does give knowledge, it clearly does not give that full and
complete knowledge of the physical world which alone would
satisfy us. If it did, of course, judging within the sensory
experience would be wholly unnecessary. We should know by
mere sensing ail that we chose to know. The opposite, how-
ever, is clearly the case. Mere sensing, it would seem, simply
provides us at most with information which is not itself full
and exact knowledge of things as they arc, but is rather a help
of some sort towards the attainment of that fuller and exacter
knowledge. If we take it to be full knowledge we deceive
ourselves. At most, it is only a beginning needing the aid of
further mental operations before an exact knowledge of the
object can occur.

On this point all critics of the naive view are agreed. Here
their theories find a common starting-point, however much
they may diverge later. Thus the critical view which most nearly
resembles the naive position avoids the latter's naivete by
seriously modifying the claims made for sensing. In sensing,
it asserts, we do not know the full physical world, nor indeed
do we know any one particular existent as it really is. Sensation
(that is, sensing) merely provides us with knowledge of certain
features or qualities of existents. This knowledge, however, is
exact, direct, and infallible in character; but it is not complete
knowledge of the real. There must also be other capacities and
faculties at work. In sensing I only know that this apple before
me is red in colour, sweet, smooth, hard, and has a peculiar

28 THE NATURE OF KNOWING

smell of its own. That it is one, a unity, that it is a body, or,
again, that it is like another apple, and comparable with it these
things I know not through sensation but through some other
means. Therefore in gaining a knowledge of the one physical
object I do not merely sense it, for sensing gives me knowledge
only of its 'sensory' qualities, as instanced above, but I must
also set those other faculties to work in order to know all its
qualities and relations. Thus my final complete knowledge
of the object can only be the outcome of a combination or syn-
thesis of many features known by many faculties into one whole. 1
At this juncture we do not wish to discuss the general theory
of knowledge implied in this first type of criticism. Our interest
at present is confined to the problem of sense-experience. And
since this is so we need only attend to the following considera-
tions. The adherents of this first critical position have to pre-
suppose that, if knowledge of the whole physioal object is ever
to be possible, then the knowledge we get of the 'sensory'
features in sensing must be certain and infallible, since we
depend upon it solely for knowledge of these features, in such a
way that if it gave us error we should never be able to recognize
it as error. We must take what the senses 'give' with respect
to such features, and if they do not give the truth that is, if
they do not reveal the real features of real things then, however
true our knowledge of the other features may be, the final whole
that we make by combining the different features will necessarily
be false. It is, consequently, essential for this view that sensing
should give indubitable knowledge of the particular features
in question, for with regard to them there is no appeal beyond
the senses. I do not see (by bare seeing) the rose itself in so
far the view rejects the na'ive theory but I do see its redness,
and thus know finally and absolutely this real quality pertaining
to the real existence.

1 This act of combination may, however, occur spontaneously,
being, indeed, not so much a combining on the part of the mind,
it might be said, but rather the apprehension of the qualities as
being in combination. (Even so, the position is clearly a difficult one
to defend.)

THE SENSORY EXPERIENCE 29

But just as it could not be granted that sensing provides us
with exact knowledge of physical objects, so now the claim that
it invariably provides exact knowledge of a real feature of the
physical object must also be held invalid. Error seems to be
possible even when we merely affirm that a certain physical
object has this or that colour. For instance, I may look at a
wall and say, "This wall is red". On the view under considera-
tion, though the seeing gives me no other knowledge of this
object, it yet does convince me beyond the possibility of doubt
that the colour of the wall is red. But if, now, following some
pre-arranged signal, a friend looks at the wall at the same
moment as myself, but a great distance off, he will probably see
it to be grey, and will have to say, "That wall is grey", A third
person looking through a powerful microscope upon some
portion of the wall will simultaneously see it to be, perhaps,
brown, and will say, "This wall is really brown", and there will
be as much evidence to justify his assertion as there is to justify
mine, or that of my friend. Now, in point of fact, the wall can-
not itself be red, grey, and brown at one and the same time. It
is quite true that, as the thinkers who hold this view would
speedily contend, the difference in what we see is due to a
difference in the circumstances in which we see. But, even so,
it is clear that seeing is not invariably an exact knowing of the
colour of a thing. If the wall is red, then when seen as grey or
brown its real colour is not being known. Consequently, seeing
a colour is not always the direct knowledge of a real feature of
an external thing as this particular theory would claim.

"But", it may be objected, "you are causing unnecessary
confusion. In every case of sensing the 'given* has to be
'adjusted'. In this instance we must 'adjust' by allowing for
the differences in the circumstances of the three persons when
seeing. Once this allowance is made, the real colour of the object
can then be determined with ease." Now, if we adopt this
position, then clearly what I see is, on occasion at least, no ex-
ternal existence, nor even a quality of an external existence, but
rather something which may provide me with a clue to know-
ledge of the external existence a very different theory. Seeing

3 o THE NATURE OF KNOWING

itself, we should then be admitting, does not always give direct
knowledge even of a feature of the physical thing. Furthermore,
once I begin to 'adjust', my conviction that I know the real
colour of the thing depends not on the seeing of the colour,
but on the 'adjusting'. I should say, then, that I know that this
wall is red, not so much because I see it to be red (for I know
that in other circumstances I may see it to be grey or brown),
but because in normal conditions I see it to be red. My con-
viction is based not on the fact that I saw red, but on the con-
clusion to which I have come, as a result of a process of reason-
ing, that conditions were in that case normal (and abnormal if
I saw grey or brown). If then we do know in such a case, the
knowledge, even of this one feature of the real, does not occur
by way of mere sensing that is, in this case, by merely seeing a
colour. More is present. Seeing a colour in itself does not
provide me with certain knowledge of physical objects. Con-
sequently, this first critical position, we are led to conclude,
is no more tenable on reflection than is the naive position, for
it cannot rightly be granted that we know, in sensation, directly
and exactly, even a feature or quality of an externally existing
thing.

But if now what we see is neither the real physical object
nor a quality of it, what else can it be, and how can it aid us
to know the real ? An answer is provided by the second main
critical position, which holds that sensing presents us with a
content illustrative of the real independent of us while not
itself being that independently real. We are provided with
'sensa' which are related in some way to the external existences,
while yet not revealing them as they are. By working on such
content, however, adjusting, relating, classifying, and so on,
we may hope in time to gain adequate knowledge of the inde-
pendently real itself. This position is, we believe, in essentials
one with the famous Theory of Representative Perception,
which produced so much bewilderment in the minds of six-
teenth-, seventeenth-, and eighteenth-century thinkers. That
theory, however, was sometimes expressed in a cruder and
more naive fashion. The ideas in the mind, the mind's imme-

THE SENSORY EXPERIENCE 31

ttiate objects, were conceived as exact copies of the real ; though,
indeed, most thinkers soon realized that the 'ideas' of secondary
qualities, colours, tastes, sounds, and the like were not exact
copies, but only represented in a vague way certain 'powers 1
in things, one of which, for example, caused me to see a particu-
lar colour. The still more critical position now being put forward,
however, holds that none of these 'ideas' or 'sensa' are exact
copies, but simply that they are sufficiently representative to
guide me to the real. We cannot say, without being naive, that
sensing is itself a knowing of the externally real ; but we can
say that it provides us with a content which inexactly copies the
real, and that if by a process of thinking the inexactitude of the
copy could be determined it would be an easy task to correct
the copy where necessary, so that knowledge of the real might
ensue. The basis of this view, however, is identical with that
of the Theory of Representative Perception; namely, that in
sensation we have to deal in the first place with representations
of the independently real and not with that real itself.

Once we have arrived at this stage a distinction of very great
importance becomes possible and, indeed, necessary. This is
the distinction between what is and what appears, between
reality and appearance, between the actual existence and the
idea or image. In sensing we do not experience the real as it is,
but an appearance of it. The content given by sensation, accord-
ing to the new critical theory, is phenomenal or what appears
only. We have to seek with its aid for the independently real.
But if this be taken to mean that we are to search amongst the
representations for that one which images the external with
sufficient exactitude to give us knowledge of it, a serious
difficulty immediately arises. For how can we ever know whether
a representation or copy of X is a good one or not unless we
have known X itself? Now on the theory under consideration
we sense only the representations or copies and never the
originals; it is, therefore, impossible to test these copies, since
we cannot finally test a copy's worth except by the original.
This is an old criticism, but none the less sound; for, so long
as we continue to think of the problem in terms of copy and

32 THE NATURE OF KNOWING

original, then without a knowledge of the original we cannot
know whether a representation does adequately represent what
it is supposed to represent or whether, on the contrary, the
representation is false. If I say that what I see is appearance
only, then when I see the colour of a wall to he red, grey, and
brown in different circumstances I must be able to see the real
colour itself before I can say which of those appearances red
wall, grey wall, or brown wall, is the truest copy. Yet colour as
an actual property of a real external existent if there be real
colour in this sense is just what, according to this theory, I
cannot see.

Against any theory, therefore, which holds that the imme-
diate object of sensation is a representation, and that we only
know the real object when we have discovered that one repre-
sentation among the many 'given' which exactly represents it,
the following criticism may be urged. Never,by any process,
can we learn the degree of exactitude with which such repre-
sentations mirror the external world beyond the representations
if they do mirror it at all unless we succeed in directly
apprehending that external world itself. Yet such direct appre-
hension is, on this view, impossible by way of sensing. While
if another approach to the external real were posited providing
us directly with perfect knowledge of the real object, then the
search within the content of sensation for the best copy would
be wholly unnecessary. The dilemma in this case is real enough.
For if such direct access to the external is indeed possible, then
the representations of sense are superfluous; they are copies
(most often inexact copies) of what we already know directly.
But if, on the other hand, we refuse to admit the actuality of
such direct knowledge, we then can never test the copies in the
light of the original and so can never discover the true copy. Such
a dilemma shows the unsatisfactory nature of this theory and
necessitates its rejection. Knowing cannot be a search for the
best copy of the real amongst the many representations provided
in sensation. Here, again, is a view which a little reflection
shows to be completely untenable.

It is consequently characteristic of those philosophers who

THE SENSORY EXPERIENCE 33

attempt a new approach to this question, by which they hope
to overcome the difficulties of Representative Perception and
like theories, that they seek no longer to discover in the 'given'
of sense representations which adequately mirror the external
real. On the contrary, they try to discover truths which are
true simply because they carry with them a validity that is
necessary. They try to find within their sensory experiences
what must be, what is so necessary that it cannot be denied.
And it is this act of discovery which they term knowing. They
no longer expressly seek for the external 'represented' in sense-
data, but for the inevitable, whether it be external or not. This
profound difference in attitude between what we have termed
the second main critical standpoint and the third, which we are
now beginning to consider, must be fully realized and con-
stantly borne in mind if we would understand some further
developments ijjt epistemology. A sensory experience taken
as a whole is to be conceived as revealing truths only in so far
as there is within it a knowing of the necessarily valid. The
knowing involved in the sensory experience is simply the
discovery of those truths which hold necessarily within the
content of sense-experience.

Now if we are to know the inevitable in the 'given' of sense
we shall succeed in doing so only as we lay bare inevitable
relations. For, clearly, what we know in this case must be
relations which hold universally within the appearances pre-
sented by sense. And I judge them to be inevitable because I
know them to be universal. Consequently, to say that I know
the necessarily valid, or what must be, simply means that I
know a law or laws which a particular set of experiences must
always obey. Knowing, indeed, on this view, is no longer to
be conceived as the apprehension of one particular physical
object in its particularity, whether directly or indirectly by
way of representations ; it is rather the realization of a definite
law as a fact in a particular experience, which law is the 'truth'
of that experience. And this new meaning of truth makes it
wholly unnecessary for us who desire true knowledge to seek
in our sensory experience for representations of external reals.

34 THE NATURE OF KNOWING

But at this stage an additional consideration of very great
importance must be brought forward. We still conceive the
'given* of sense as 'appearance', as only partially real in a vague
way which at present we do not understand. We also continue
to admit the existence of an independently real which, we
suppose, somehow or other produces in us sensations, though
sensing gives us no knowledge of that real. Consequently, the
law known in the sensory experience does not hold of this
independently real world; it only holds of the phenomenal
sphere. Now a law holding among external real things is on
the face of it wholly independent of the mind knowing it,
and if I ever know it I shall only do so by apprehending it
directly as it really is. But, clearly, a law known in the pheno-
menal world need not be thus independent. It is not a real law
of real things. To say so would be to revert at this crucial
moment in the argument to the naive position, without being
able in any way to justify the reversal. On the contrary, the law
I know in the phenomenal world is itself phenomenal. It is no
law of really existing things, but is at least as dependent upon
the mind that knows it as is the phenomenal of which it holds.

A further consideration, however, may now show that
actually such laws are even less independent of the mind than
is the manifold 'given 1 in sensation. For while the law, to know
which is to apprehend the necessarily valid in our experience,
holds of and in a phenomenal world, yet it itself is not 'given'
with the phenomena in sense-experience. What are 'given' are
colours, sounds, tastes, smells, feelings of resistance, of smooth-
ness, of roughness. But, for instance, that one 'appearance' is
produced by another as effect by cause is never 'given'. I have
never seen, heard, smelt, tasted, nor touched such a relation.
We never sense any definite connection within the manifold
presented by sense, though we know many such connections.
To take another example ; by just seeing a deep yellow, or, if
it be preferred, a yellow patch, I cannot see its connection with
the sweetness of an orange, I am only 'given' the colour yellow,
though I immediately think of sweetness. These connections
then are never 'given' with the content of the phenomenal

THE SENSORY EXPERIENCE 35

world; they are not 'presented' in our sense-experience; so
that they do not appear to possess even that meagre measure
of independence of the mind which phenomena still possess
in that they are 'given' to the mind rather than created by it.

Therefore, according to the present view, we do not know
the 'truth' about phenomena that is to say, their laws and
their necessary order (the existence of which in the phenomenal
world enables us to predict certain events as inevitable) by
simply apprehending such laws in the way in which a naive
person would say we 'apprehend' the real. Nor, again, are they
'given' in the content of our sense-experience, as colours and
sounds are 'given'. Only one alternative, consequently, remains:
namely, that the source of such laws is inward, that we, ourselves
create out of our experience an ordered whole in which we may
justifiably expect certain things to happen, since we who thus
expect phenomena to follow laws have ourselves set down the
laws they have to follow. If we are to speak consistently and
significantly, bearing in mind all that has been said up to the
present about the nature of the sensory content, then it is thus
alone that knowledge of the necessarily valid can occur, and
only thus can we explain our possession of ordered systems
of such knowledge in the various sciences. But it also follows,
if this be the true account, that human knowledge is not of the
real world itself, 1 since all we have to deal with is the pheno-
menal, and this alone is the content of our experience. Sensation
never 'gives' us the real, though what it 'gives' is somehow
remotely connected with the real. Therefore, if we, who have
only the 'given' of sense to work upon, desire knowledge, it
can only be knowledge of an ordered phenomenal world,
wherein the order itself is neither 'grasped' nor 'given', but
introduced into the 'given' by the mind itself. That the 'given'
should in this way allow itself to be ordered by mind is cer-
tainly strange enough, for, though phenomenal, it is still in a
measure independent of the mind. Nevertheless, this must be
the case if human knowledge is ever to be possible. While,
indeed, the fact that the phenomenal is still, in however slight
1 That is, 'real* as opposed to 'phenomenal'.

36 THE NATURE OF KNOWING

a measure, independent of the knowing mind must be reckoned
a very fortunate one for us, since it gives our science a faint
external reference which saves it from complete subjectivity.

This position 1 certainly avoids the central fallacy of the
Theory of Representative Perception ; we no longer seek the
best representations or copies of we-know-not-what originals
a task of Sisyphus. But it avoids it at a price that few would be
prepared to pay who fully understood the implications of the
new theory as it stands. For by it the human mind is adjudged
incapable of ever coming to kno\v the real external world.
Instead, we are shut up within a world of appearances, vaguely
suspecting the existence of a real world which has produced
and is producing these experiences within our consciousness,
but which is nevertheless unknown to us and unknowable.
Prisoners within the confines of the phenomenal, the urge
for knowledge which possesses us can receive 'only such satis-
faction as comes from knowing this shadow-world of appear-
ances ; nor is there anywhere a path that can lead us to the real
world beyond.

In spite, therefore, of the necessity which permeates this
limited sphere, making human prediction possible, and in spite
of the 'objectivity', as being equal to necessity and universality,
which has thus been assured us, we cannot rest satisfied with
this position. However certain our knowledge be, if we are
quite explicitly conscious of the fact that it only applies to the
semi-real world of phenomena, then it cannot be an example of
that perfect knowledge for which we seek. Indeed, using words
significantly, we do not feel justified in terming this creation of
a systematic world of phenomena (which we know to be pheno-
mena and nothing more) 'knowing' at all. On the contrary,
our awareness in this instance of the fact that we are confined
and limited, that what is before us is the phenomenal and half-

1 The position we have in mind, of course, is the Kantian. We do
not put the argument forward in an historical form, however, because
we do not wish to argue as to whether Kant did hold this view or
did not. Our attitude throughout this essay is different. Here is a
view, whoever first put it forward. Is it itself sound?

THE SENSORY EXPERIENCE 37

real only, would approach nearer to our ideal of knowledge,
for it would actually be a knowledge of what really is. Although,
incidentally, such knowing would remain \vholly unexplained
and inexplicable on the theory under consideration. For human
knowing, according to the latter, is more akin in nature to the
dream and to the fantasy than to knowledge of the real, though
it be a dream that greatly helps us in our dealings with the
'appearances' of everyday life. To accept the position as it
stands would be to accept an intolerable bondage; we should
lose all confidence in the mind's power to know; we should be
plunged into a state of hopeless and despairing scepticism.
How could it be otherwise if we knew beforehand that the
object of human knowledge must invariably be phenomenal in
nature ?

As a consequence a further development of the argument
becomes inevitable if only to help us regain our faith in ourselves
as beings capable of real knowledge. For the new theory that
now emerges re-establishes in philosophical speculation the
common-sense conviction that what we know /.v, that the real
is the knowable, and not, as with the earlier theory, an unknow-
able. The new development, however, is no fresh start. It
re-endorses earlier criticism and makes them its basis. For it
the 'given' of sense is appearance only; there can be no return
to the naive position. Secondly, a 'true' object cannot be found
by searching for exact representations of the real in the content
of our sense-experience; consequently, the Theory of Repre-
sentative Perception and any other theory that approximates
to it must be rejected. Thirdly, the 'principles' whereby the
'given' is ordered are not themselves 'given' but are ways in
which the mind thinks appearances. In a word, the new theory
is faced with this difficult task: it must confirm such prior
criticism, while, at the same time, it must free the human
mind from the limitations which those very criticisms have
seemingly shown to be necessary.

It carries out this act of emancipation, to its own satisfaction
at least, by adding one more criticism in its turn to those
already made. For it asserts that the theory we have just been

38 THE NATURE OF KNOWING

discussing fails lamentably in one respect. Though certainly
revolutionary in many of the changes it introduces, nevertheless,
as the result of too close an allegiance to the past, the theory
misses the real and vital consequence of its own position. A
complete break away from the earlier standpoint, which,
emphasized the reality of an external thing-in-itself , is necessary
on the lines already set down for us by the theory just discussed.
That the latter, however, failed to carry out completely this
revolution in thought is proved up to the hilt by the phenomen-
alistic nature of its epistemology.

Yet it securely established the grounds for this further
advance. For once the three criticisms set forth above have
been admitted, then surely one cannot long continue in the
belief that the 'appearances' of sense in any way point to an
external real, and that the only real knowledge would be the
apprehension of this thing-in-itself outside, although actually
such an apprehension is wholly impossible for us. For if the
real is what lies outside, influencing us in our sensing, but yet
unknown to us as sensing, and if again no other point of con-
tact with that real is possible for the human mind, then no
amount of 'adjusting' and no measure of intellectual labour
can lead us to knowledge of the real. Indeed, all such working
upon the 'given' of sense would lead us directly away from
rather than towards the real, since ex hypothesi our nearest
approach to reality occurs in sensing itself. And worst con-
sequence of all if we persist in this view a most confusing
distinction between the real and the true must be made. The
'real', as such, it will be necessary to assert, belongs to a sphere
which is transcendent and cannot be experienced by us; the
true, however, is the necessarily valid within the world of our
experience. Hence the paradoxical position, that we may gain
truth while still remaining wholly ignorant of the real.

But all these disquieting consequences, it is now pointed out,
are due to the simple fact that we persist unnecessarily in the
belief that the real is this thing-in-itself, this existence that
transcends human experience. If we take our experience as it
is and forget for the time being all the various theories whereby

THE SENSORY EXPERIENCE 39

we have tried to interpret it, there is nothing, the new view holds,
that makes this concept of the transcendent necessary ; nothing
that makes it impossible for us to hold that reality is more
apparent in thinking than in sensing, that our 'true' object
(which indeed is not so much found in appearance itself,
but is rather the thought-out 'truth* of such appearance in the
sense we have already explained) is itself what really exists,
that, lastly, the principles of the understanding are capable
of ordering not only the phenomena of sense, but the real world
of actual existence. There is, as we say, nothing that makes
impossible such a belief if we once free ourselves from the
tyranny of earlier thinking.

Certainly, such a theory of knowledge involves a radical
change of outlook that makes the whole universe of being appear
in a new and perhaps at first strange light. For if the mind's
principles are fche laws of what really exists, and not only of
a world of phenomena, such complete accord between mind-
created truth and the real must mean that the real itself is
spiritual in nature in this sense, at least, that mind provides
the ultimate explanation of its being. That is to say, actual
existence in all its forms must be identical in its ultimate nature
with this mind of ours, otherwise we could never by working
upon the manifold provided in sensing, under the guidance of
the principles of the understanding, hope to gain for ourselves
ii complete knowledge of the real world, nor could we account
for the fact that the necessarily valid is the real. All this may
appear strange, but if the facts be otherwise, then, according
to the present theory, it can only mean that the real is not
rational and so cannot be known by us, and we are plunged
again into the scepticism which we are trying to avoid. We are
forced, nolentes volentes, either to deny that human beings ever
can possess real knowledge or to assert as our one way of escape
that the real is spiritual and that, therefore, our thinking can
legislate for it. The real conforms to those principles which
my own thinking sets forward as true; so that the mind can
construct truth, not of itself creating the world which it knows,
for that world is already created, but rethinking the very

40 THE NATURE OF KNOWING

thoughts which created it. It follows that in the seeing of colours,
the hearing of sounds, and in all those experiences which we
term sensations we are at the stage of Appearance, wherein the
mind's knowledge of the real is vague and indistinct, but that
this knowledge becomes more and more adequate and com-
prehensive as we proceed to discover by thought, guided by its
laws or principles, the inmost 'truth' of what appears.

Therefore we needs must reject the 'sensationalist' pre-
suppositions of the former position. If knowledge is to be in
any way possible, what is real is not something outside affecting
us in sensation whilst yet remaining unknown. For the con-
sequence of such a view is that we find ourselves compelled
to turn away from the real, since we know beforehand that it
is utterly impossible for our finite minds to 'get out' to it. As a
result we are confined, against our will, within the limits of a
world of appearances, which we do our best to order as well as
we can. But if real knowledge is ever to be possible, we must
deny this conception of a transcendent external world and
free ourselves from the bogey of the merely phenomenal. We
must assert that the real itself lies 'inward' and not 'outward';
that, indeed, the distinction between 'outward' and 'inward'
loses its meaning: that the mind can order and legislate for
the real because reality itself is spiritual in nature: that if we
seek the essence of anything, what it exactly and most truly
is, we shall discover that it is something spiritual: and that,
finally, this something is better known by conceiving and
thinking than by sensing.

This, then, is no idle and fantastic theory. On the contrary,
it is the position to which we are driven by the logic of the
argument. It is the inevitable conclusion of a trend of thought
that begins with the rejection of the naive view set forth in the
first section. At present we do not mean to discuss the soundness
of this final position (that is to say, of an Hegelian type of Ideal-
ism) as a whole system of philosophy. Our more immediate
concern is the problem of sense-experience. And from this
point of view, what follows asserts an obvious truth: If it be
said, criticizing the first naive view, that the content of the

THE SENSORY EXPERIENCE 41

jensory experience cannot be the real as it is, but must rather
DC thought of as some partial presentation of the real, then, if
consistently worked out, the argument proceeding from such
a starting-point must lead us step by step either to a thorough-
going scepticism or to this spiritually realistic theory of being
and of knowledge which we have just been expounding. Reality,
we must say, if we wish to avoid scepticism, is not some-
thing, to be looked for outside mind, but something through
and through 'mental', 'ideal', or, again, 'spiritual* in its
character. 1

We believe it possible, therefore, to draw the following
important conclusion. If we assert that sensation is the only
outlet to the real, and that what is 'given' in sensation is appear-
ance, even though we add that such an appearance is the real-
as-appearing, or the physical-as-appearing, or, again, the actual-
as-appearing, rrfeaning thereby something which only just fails
to be the real or the physical or the actual itself, and if we thus
base our criticism upon the distinction between appearance
and reality, then inevitably we are closed up within a phenomenal
world, however much we long for the noumenal. And, unless
we succeed in finding the noumenal by holding that the real
is spiritual in its ultimate nature, we shall have cut ourselves
off for ever from the real. So long as Appearance and Reality
is thus the predominating antithesis in men's minds when the
attempt is made to give an adequate account of knowledge, so
long also must the Idealist interpretation of knowledge and of
its object of necessity ensue to save mankind from the deepest
and the most despairing forms of scepticism. Such is the con-
clusion which we believe to be fully warranted by the inquiry
carried out by us in this section.

1 It may be objected that a third alternative is possible, namely,
that we can 'infer' the nature of the external real from the pheno-
menal. But if what is 'presented* is phenomenal containing no clue
or hint as to the true nature of the external real, it would seem
logically impossible to 'infer' the real. Where such inference occurs
it must be because we already know something of the nature of that
external real for example, that it coheres. We shall endeavour to
show later how this knowledge comes about*

42 THE NATURE OF KNOWING

3
A Further Critical Position

The Idealist position at which we have now arrived
confirms our faith in our cognitive powers and dispels our
scepticism. It also frees us from the insufferable bondage of
phenomenalism. It assures us that the real itself can be known
by mind. But the fully real, it would say, is not known by all
minds, since some minds are less developed than others. For
instance, the mind which puts its trust in perception is confused
in its knowledge. Perception is a low stage of experience at
which the content experienced is only half true. It is the stage
of Appearance, illusory and deceptive. But a rational being
cannot and, Idealism adds, need not remain at this stage.
Within the content experienced in perception there are germs
of a fuller truth and the mind of its own power, working
according to its own principles, can develop them, constructing
(or reconstructing) by this development the real itself. For
as thus developed what appears loses its illusory character.
The half-real becomes more completely real. Appearance is
becoming Reality. If the shadow- realm of Appearance be likened
to a prison, it is a prison within which the prisoner is not
confined against his will. Its doors are open wide to any being
whom Reason has taught to walk.

Are we then to adopt the Idealist solution as here set forth ?
One point has already been made clear. The moment we feel
compelled to assert in a critical mood that in the sensory
experience what we gain is wholly phenomenal in character,
then we have, ultimately, only two alternatives from which to
choose, either complete scepticism as to the knowledge of the
real, on the one hand, or Idealism, on the other. And while, no
doubt, many thinkers consider that the assumptions made by
the Idealist are so unjustified as to vitiate his solution of the
problem, so that it cannot really be considered as an alternative,
yet no serious inquirer after the truth ever delights in scepticism

THE SENSORY EXPERIENCE 43

as such, nor will he adopt it while there is any other satisfactory
solution at hand. Hence the strength of the case for an Idealism.

Nevertheless, in spite of its strength, we do not intend to
adopt the Idealist solution in this essay; and we must now try
to show why. We do not adopt the solution because we feel
that the problem which Idealism thus attempts to solve is an
unreal one. It arises from misapprehension and faulty analysis.
For we suspect that the whole difficulty consequent upon the
adoption of phenomenalism ought to have been avoided at the
outset. As yet, we suggest, full justice has not been done by us
to the facts of the sensory experience. And in this section we
propose to begin criticizing the naive position afresh. We hope
to show that, actually, the facts when rightly examined do not
drive us into phenomenalism, and that the validity of the whole
antithesis between Appearance and Reality as set forth in the
last section is fo be questioned. And if it can be shown that
phenomenalism is unnecessary, and even unjustifiable, then
surely the Idealist solution, which is to save us from the evil
consequences of phenomenalism, is in so far equally unneces-
sary. There may, of course, be other powerful arguments for
Idealism. We are here attempting no refutation of Idealism in
general. But if the \vhole phenomenalist position is shown to be
the outcome of insufficient attention to the evidence available,
and if on paying greater attention to this evidence the difficul-
ties and problems of phenomenalism vanish with it itself, then
clearly the need for an adoption of Idealism on our part as a
solution of such difficulties no longer exists.

Now we assumed throughout the last section that bare sen-
sation that is to say, the seeing of colours, hearing of sounds,
and so on has always some cognitive value. 1 The fact that the
assumption was made so readily (and, in a sense, so unwittingly)

1 It ought to bo emphasized, perhaps, that we mean by 'sensation'
the mere hearing a sound, tasting a taste, seeing a colour, and not
the whole sensory experience, as it occurs, for instance, in the adult
human mind. The bare hearing of a sound no doubt is only a part
of the whole auditory experience. But by a process of analysis and
abstraction we shall consider this element in itself in order to show
that it itself is not knowledge.

44 THE NATURE OF KNOWING

by us suggests that we were still under the spell of the first
naive view, which looks upon sensation as exact and perfect
knowledge of the real. And because we made this assumption
then doubt about the cognitive character of sensation was not
carried to the extremes that certain considerations would seem
to demand. We have throughout presupposed that sensing is a
form of knowing. In the last section, certainly, it was, granted
that the knowledge given was very vague; indeed, the word
'appearance' was used in order to suggest the distressing
obscurity that characterized the content of sensory knowledge.
But yet we never doubted that sensation did provide us with
knowledge, however indistinct and inadequate. We must no
longer, however, withhold our minds from this supreme doubt ;
for no satisfactory theory of sense-experience can be gained
while we ourselves are conscious that certain difficulties
remain unfaccd, and one such difficulty lies ifi the possibility
that sensing is never on any occasion an instance of knowing,
even of the vaguest kind a possibility which is accentuated
by certain facts about our sense-experience which we must
now consider. 1

The naive position itself clearly cannot be accepted. Neither
can that modified form of it which holds that we sense directly
and infallibly real features of real physical existents. The
argument urged against the latter position was that if, for
example, looking at a wall, we all see a real feature of it, namely
its colour, then one and the same wall must be red, grey, and
brown at one and the same time, since it is seen to possess
these three distinct colours by three observers observing
simultaneously. It is true, as was pointed out, that the difference
in the seeing is due purely to a change in the circumstances in
which the seeing occurs. But the fact that such circumstances

1 In the discussion which follows I have been much influenced
by the teaching of Professor H. A. Pri chard on the nature of the
object in sensation (as yet unpublished). I readily and gratefully
acknowledge the debt. I should add, however, that the consequences
I draw from the theory (as also the actual formulation) are my own.
And I cannot say how far Professor Pri chard would agree with me.

THE SENSORY EXPERIENCE 45

can change what is seen was sufficient to prove our point at
the time, namely, that seeing does not always provide us with
reliable knowledge even of the colour of a thing, and that,
therefore, our sensing of features, for example, our seeing of
colours, is no infallible knowledge. By this means the theory
that in seeing a colour I am invariably knowing a feature or
quality of a real physical object was overthrown. 1

But we must now consider the more important consequences
that follow upon the difficulty illustrated here. Quite clearly,
a red wall, if it be red, is neither grey nor brown, and yet we
see it to be grey from a distance and brown again under a
microscope. Therefore, we concluded, seeing a colour is no
complete and infallible knowledge of a real feature of a real
existent. Now, however, we need to ask a further question: Is
it even partial knowledge? Common to all the critical theories
discussed in the last two sections was the implication that sens-
ing is knowing, but nevertheless an incomplete partial knowing
that called out for completion by an intellectual effort of
'adjusting' and judging. But can we rightly say that this instance
of seeing a grey wall is 'partial' in this sense of providing us
with what might of itself lead us to see the real colour of the
wall? If we had never been near enough to see the wall as red,
would the greyness we saw ever of itself suggest the redness ?
Clearly not, for the error if 'error' it is of seeing the wall
to be grey is no 'mistake' on my part that can be rectified by
reflecting. However hard I reflect I shall not succeed in seeing
the wall to be red from that distance. There is, as a matter of
fact, no mistake about the experience; for I did see grey; and

1 It might be argued by some thinkers that all three colours do
exist someivhere in the real physical world, though not in one and
the same space. The apparent contradiction would then be resolved,
while the colours would still be physical. But even if this signified
anything and we must confess that we find it difficult to attach
any meaning to it yet the same problem remains. How are we to
determine which of the three colours belongs to that space which
the wall itself fills ? And the fact that the problem arises at all shows
that we do not know in sensation directly and infallibly the real
colour of the real wall. This is all we seek to show.

46 THE NATURE OF KNOWING

yet it is equally clear that I was not then knowing, nor even
'partially' knowing, if the colour of the object is indeed red,
or any colour other than grey.

As this is an important point, we may illustrate it further,
taking on this occasion an instance of hearing for we must
make certain that what we say applies not merely to seeing, but
also to all other instances of sensory experience. Suppose X
and Y both hear (speaking unphilosophically) one and the same
noise. X says, "That is a pistol being fired. " Y might reply,
"No, that was not a sharp enough report to have come from a
pistol. You have inferred wrongly." X might then reflect and
perhaps admit that Y was right. He would realize that he had
mistaken a dull sound for the sharper-sounding report of the
pistol. But now, we may suppose, X suddenly becomes par-
tially deaf, unknown to himself. Again they both hear a noise ;
for Y it is loud, for X it is not. If Y now tells Xthat the noise is
loud, X will not believe him, and however much X reflects in
this case, by no reflection whatever will he come to think that
the sound was loud. That is to say, both X and Y are absolutely
certain about what they heard; ex hypothesi they hear the
sound produced by the same external something; and yet they
do not hear the same sound. "But", it will be said, "X is deaf;
the conditions under which he hears are abnormal and, if he
was aware of this, he would give way to Y." That is certainly
so. But the very fact that a change in the physiological con-
ditions can thus produce a change in the sensation means that
X, at least in this case, was not knowing one and the same
external object with Y when he was hearing, for he was abso-
lutely certain (and not only partially) that he was not hearing a
loud noise, whereas if normal he would have heard a loud
noise. X, therefore, we must conclude, is in this case simply
not knowing; and by this we do not mean that he has fallen
into error, or has inferred wrongly. Certainly, such false infer-
ences may well occur later as a part of his whole experience of
sensing, which, taken thus as a whole, is invariably wider than
simply seeing a colour or hearing a sound. We only mean that
in the bare hearing of the sound as such X is neither knowing

THE SENSORY EXPERIENCE 47

nor, strictly speaking, erring. 1 In that particular his experience
is not a cognitive one. What X hears is not what Y hears, yet
neither is making a mistake. X as truly heard the slight sound
as Y the loud. Therefore, if there is a real noise which they
might be said to be 'knowing,' one of them (in this case, we
suppose, X) is not knowing it when he hears a noise, not even
'partially', or vaguely, or, again, half-erroneously.

When, that is to say, I see in abnormal conditions, I am not,
as barely seeing a colour, knowing any external real thing or
any real feature of such an external. I do not even 'half-know'
this feature, or know it vaguely. But clearly, when once this
position is established, doubts must immediately arise about
the character of sensing in normal circumstances. For can we
admit the implication involved in the preceding paragraph,
namely, that a difference in the external circumstances, physio-
logical and physical, of itself produces this radical difference
in the inward nature of sensing, making it in the one case
knowledge and in the other an experience which is purely non-
cognitive? Does it not seem absurd to suppose that sensing
is knowing when I see the colour red under normal conditions
to revert to the first example and that it is not knowing but
something else when the other observers see the colours brown
and grey? Again, if I look at a rose in light that is gradually
fading, when I first look it has a fresh red tint, then I see it take
on a darker shade, and ultimately it will become indiscernible
from the blackness around it. Now in such a case it would
surely be false to hold that when I saw the red rose I was
knowing, while as soon as I began to see it 'changing its colour'
I was no longer knowing, but experiencing a completely differ-
ent experience. If it be correct to hold that in abnormal cases
of sensing I am, purely in so far as I see a colour or hear a
sound, not knowing a real physical object nor a real feature of
it, it would seem necessary to add that in normal cases the
experience and content being of the same general character as

1 When I talk of 'knowing' here I mean, of course, knowing in
the sense in which the experience claims to be knowing that is,
knowing the externally real.

48 THE NATURE OF KNOWING

in the abnormal case I also have no such knowledge. In a
word, seeing a colour, hearing a sound, tasting a taste, smelling
a smell, and feeling a resistance, are none of them cognitive
experiences as such', they give us no knowledge of external
physical objects. When, for example, I see a colour, just in so far
I know no physical object, either as a whole or in part.

"But", it may be objected, "what I know in sensing is nothing
external; it is a sensum or sense-datum, something merely
in the mind. Yet sensing is none the less definitely a knowing.
I know the sensum." Such a view, of course, would be very
different from what we mean ordinarily by calling sensation
knowing. Ordinarily, we should mean that in sensation we
know the real external world and no mere mental 'world'. But
can we accept this much modified view ? Are we justified in
calling sensation knowing even in this sense? We doubt it.
Once again we must recall that the discussion io about the mere
seeing of a colour, hearing of a sound, and so on. Now, no doubt
there is knowing in the sensory experience taken as a whole.
We shall shortly point out what elements of the experience
are distinctly cognitive. But is just seeing a colour or hearing a
sound knowing? Certainly, my consciousness that I now see
is a knowing. Certainly, again, if 1 say "This is blue" I make a
judgement, involving recognition, which at least claims to be
true and is cognitive. But is seeing blue itself a knowing?
We cannot admit that it itself, as such is. We do not know*
'sensum' here, even if we see one. Seeing a colour, hearing a
sound, belong rather to the realm of imagination than to that
of knowing. And though the imaginative experience again
taken as a whole involves cognitive elements, yet few would
assert that imagining an image and knowing are synonymous.

We might contrast knowing and sensing in greater detail.
From the outset we have taken it for granted that a truth 'once
true is always true*. Consequently, if we ever did have an experi-
ence of knowing as such, then, we suppose, momentary changes
of circumstance in the whole process by which knowing comes
about would not affect the content known in any way. Truth,
we suppose, is independent of the knower's particular location,

THE SENSORY EXPERIENCE 49

and of whether it is known by this person or by that. Likewise,
it does not depend upon the state of the knower's bodily organs
nor upon the present condition of his immediate physical
environment. But the content of sensation is completely lacking
in the independence and absoluteness which thus characterise
truth. What I see, for example, is dependent upon the particular
set of conditions in which I happen to enjoy the sensory experi-
ence, conditions having to do with my spatial position, the
presence of light, the state of my eyes and optic nerves, and
so on. Clearly enough, changes in the physico-physiological
process accompanying each sensory experience have an effect
upon the content 'given' me in sensation. But if seeing the
colour, hearing the sound, were truly knowing, this would not
be the case. For then the only changes of which we should
be cognizant would be changes in the object known. The fact
that changes iij the physico-physiological process produce
changes in what we see, hear, and so on, is, in itself, sufficient
proof that seeing colours, hearing sounds, are not instances of
knowing.

On these grounds, therefore, we feel obliged to conclude that
the seeing of colours as such must not be conceived as a know-
ing. 1 The same is true also of hearing a sound, smelling an
odour, tasting a taste, and of any touch-sensation. As such
these elements of sensory experiences are not cognitive. If I
were merely seeing a colour I should not, in the first place, be
directly knowing any actual feature of any real existent outside
me (that is, of an independently real physical object), nor, in
the second place, should I be providing myself thereby with
content, which could be so worked upon and so developed by
mind that in time I came to gain a knowledge of truth. This
latter point is as evident as the former once we reflect upon the
content of abnormal sensation. A colour-blind person, who
sees (and always has seen) all things as grey, will never by
reflecting upon the greyness, or by 'working' upon it in any

1 Again, for the sake of safety, I had better repeat I do not mean
that the whole sensory experience in an adult human mind of, for
instance, seeing the red rose is completely non-cognitive.

SO THE NATURE OF KNOWING

way, make out the real colour of a real thing, if a real thing
actually has a colour. His seeing of grey has, in itself, no cogni-
tive value in determining the actual colour of the real thing
if, again, it has a colour. But, clearly, what applies in the
abnormal case applies equally well in the normal. For, to repeat
a former question, how can we allow that identically the same
experience is in the one case knowing (when circumstances
external to it are normal) and in the other not knowing (when
such external circumstances are abnormal)? And if it is not
knowing not even knowing vaguely then colours, sounds,
tastes, and so on, which make up the content of such sensing
do not represent or illustrate the real in any way. Hence we
see no need, and have no room, for a phenomenalist theory
of knowledge. What sense 'gives', if we are thinking of colours,
sounds, tastes, and so on, is nothing that either exactly or
inexactly pictures the real, nor again can reflection upon a
colour, or a sound, as such, lead to knowledge. We do not deny,
however, as we shall proceed to make clear, that the sensory
experience, taken as a whole, is definitely cognitive, and that
we find in it the basis of much future knowledge. We hasten to
mention this point in order to avoid misunderstanding. But,
while we merely see colours, hear sounds, and so on, in so far
we know nothing, either directly or indirectly. Consequently,
we simply cannot say that we are dependent upon the seeing
of colours, the hearing of sounds, and so on, for our knowledge
of the external world, nor again that we are at first confined
and limited in our knowledge to a world of sensory-appear-
ances, or to the phenomenal, reflection upon which may lead
us to full knowledge.

It may, however, be objected that if we definitely adopt the
position now being suggested, our difficulties will be infinitely
more serious and acute; very soon all advance will become
impossible. For we now affirm that experiencing the content
of sensation (colours, sounds, smells, and so on) is not knowing
the real, since such content is, so far as we know, neither the
real itself nor an appearance pointing to the real and in some
way illustrating it, whether distinctly or indistinctly. But, it will

THE SENSORY EXPERIENCE 51

be asked, if we thus wholly deny that sensing is knowing do we
not shut ourselves off from the possibility of ever knowing the
external real ? What could we come to know of the world around
us if we were incapable of sensing? It is a very old dictum that
knowledge begins in sensation, and now we seem to be denying
the assertion outright. Are we not really in a worse position
than were those critics of the previous section who found them-
selves shut up within a phenomenal world? And shall we not,
in order to provide ourselves with some outlet, either have to
accept their phenomenalism or wholly forget our doubts about
sensation and return to the naive position ?

By way of answer to this objection, and in order to make our
position clear and definite, we may now set forth explicitly two
further considerations relevant to the present issue. In the first
place, we do not agree that sensation is the only outlet to reality.
On the contrary^ if by 'outlet to reality* we mean knowledge
of reality, and if by * sensation' we mean just the seeing of
.colours, hearing of sounds, and so on, then we contend sensation
is not even an 'outlet to reality', one amongst many. Seeing
a colour, we have agreed, is not knowing; and, therefore, in
this sense, it cannot be regarded as an outlet at all. But here an
important reservation must be made. For, in the second place,
it seems clear to us that the whole of any concrete sensory
experience, as it occurs, is never the mere seeing of colours,
the mere hearing of sounds, and so on. More is always involved
in the experience. Consequently, to deny that seeing a colour
is knowing, neither necessitates nor in any way justifies the
further assertion, namely, that no particular sensory experience,
as a concrete whole, can ever be a knowing.

That the sensory experience as a whole is not merely a seeing
of colours, or a hearing of sounds, and so on, can be proved
conclusively by reconsidering a point already made. In merely
seeing a colour there can be no mistake. I see the colour which
I actually do see, and that is the end of the matter. Neverthe-
less, it is a patent fact that any concrete sensory experience,
taken as a whole, may, and very often does, contain error. But,
if so, such a sensory experience must be more than the mere

52 THE NATURE OF KNOWING

seeing of colours, hearing of sounds, and so on. For, once we
admit that some of the experiences are erroneous, we imply
that they might have been true that is, we implicitly affirm
their cognitive character. But if seeing the colour were the
whole rather than a part only of the experience, this would be
impossible, since we have already granted that seeing a colour
is an instance neither of knowing nor of erring. The sensory
experience as a whole, therefore, is a complex of which seeing
the colour is only a part.

The source of the error in this complex is in some instances
fairly obvious. When, for example, I see a coloured patch in
the distance and say "That is Jones,'* the whole experience,
though its duration be exceedingly short, is a very complicated
one, involving at least recollection, comparison, and recognition.
Into such a complex act error may easily enter. For instance,
I may have recalled wrongly or inadequately. took the person
in the distance to be Jones because he is just that height, but
I ought also to have remembered that Jones is not so broad-
shouldered. In other instances, however, the source of the
error is not so obvious. Especially difficult to understand and
explain, is that more fundamental type of error, whose exis-
tence has already been noted by us. We believe, however, that
much light can be thrown upon the character of the sensory
experience in general by a thorough-going analysis of this
latter error. We refer to the error involved when X and Y
to revert to the former example hear what we should describe
(when off our guard) as one and the same sound differently.
There is no mistaking, yet both cannot be correct, if it really is
the same sound. The truth is, of course, that all talk about one
and the same sound in this context is false. X hears one sound
and Y another. Their experiences differ in content and, in so
far, neither errs. Nevertheless, we should naturally say, X is
making a mistake. But on what grounds do we make this asser-
tion? Why should we tend to think that X is in error? Clearly,
because, in the first place, we assume throughout that X in
hearing is knowing (or erring about) some externally-existent
noise, and that Y in hearing is also knowing (or erring about)

THE SENSORY EXPERIENCE 53

this same external existence; and because, in the second place,
we take it for granted that in knowing the object is independent
of the mind knowing. Consequently, it seems evident to us
and to X and Y themselves that they ought both to hear exactly
the same sound. If they do not, one of them is erring. In this
case we judge that X is erring, because the noise is loud not
only for Y but for many others as well.

The error, that is to say, is connected with, being indeed the
immediate consequence of, the conviction, natural to us, that
when we see a colour or hear a sound we are knowing something
existing really in total independence of our sensing. If X and Y
believed that the sounds they heard were dependent on the
sensory experience, if they supposed them identical in nature
with the images of our imaginations and dreams, there could
have been no cause for disagreement between them. For it is
in no way necessary for two men to imagine the same noise
at the same time. The disagreement came about only because
both of them claimed, in hearing the noise, to be knowing
something real, independent of them, something 'objective'
as opposed to the 'subjective'. That is to say, the real cause
of the disagreement lay in their common conviction a con-
viction shared by all unreflecting persons that in hearing a
sound we arc ipso facto knowing the physically real world
which is independent of us as knowing it. In the sensory
experience we feel convinced of the immediate presence of the
external, and the alleged error of X is only to be understood
in the light of this conviction. X is not really more mistaken
than Y, but we should naturally consider him to be mistaken
because both X and Y, we suppose, are hearing the same external
something and so ought to hear the same noise and because
most of us hear the same noise as Y. 1 That is to say, when actu-
ally we are justified only in the belief that we hear a noise or see
a colour, we invariably talk and act as if we were hearing some-
thing real outside us and seeing a coloured something, a some-

1 Even here a difficulty arises on reflection. For how can we possibly
know that, for example, hearing a loud noise, or, again, seeing a red
colour, is the same experience for all of us ?

54 THE NATURE OF KNOWING

thing which exists independently of our seeing it. And we talk
in this manner because we are convinced at the time that the
external real is being known by us. It is the independence which
we attribute to this 'something', which leads us to expect a
conformity between our sensory experiences and to suspect
ourselves of error when this conformity is missing, when, for
example, on different occasions we see the same external
something (as we think) to be grey and red, or hear the same
something to be both loud and not loud.

It is this conviction then which accounts for the fact that we
so readily assume the world to consist of objects, which possess,
as real qualities, colours, sounds, smells, tastes, resistances.
We actually see colours, but we feel convinced at the same
time that we are in contact with the real other than, and inde-
pendent of, ourselves as sensing, and we consequently assume
that we are seeing a world of independently real entities. The
same is true of hearing, tasting, touching, and smelling. Here
lies, surely, the deepest and most fundamental error of human
experience, that when we see, for instance, a red patch we
straightway believe we see a red physical object. In these pages
the error is attributed to the fact that in this complex sensory
experience we do not merely see the colour, but are also at one
and the same time convinced of the existence of an externally
real physical object and we immediately apply to it as a quality
the content seen.

Other explanations may be attempted. It may be said that
colours look as if they were physical objects, and that sounds
seem to be real things. But such an explanation of the source
of the error seems absurd; a colour looks like nothing but a
colour; a sound cannot seem to be anything other than a sound.
This cannot be the true explanation.

Another explanation that may be urged is that we become
convinced of external things in so far as we realize the character
of the space within which we see colours, since this space is so
far independent of us as to make it impossible for us to arrange
things spatially according to our private wishes. If we once
realize this, it may be argued, it then becomes evident to us

THE SENSORY EXPERIENCE 55

that we are dealing with an independent world, and it is natural
for us to suppose that this independent world is filled with
coloured objects. But this explanation is really no more adequate
than the former. For even if we grant the first point that spatial
arrangements are not created by us, but are known simultane-
ously with the seeing, even so, our conviction that there exists
before our eyes in sensation a world of independently real
coloured objects cannot be the outcome of our realising that
spatial arrangements are not our private creation. For when
we imagine we are equally well compelled to set our imagery
forth in space, yet there is in imagination no conviction that
we have before us here and now an external independently
real world of coloured objects. And we have no right whatever
to assume that the space of the imagined world is a private
creation of the mind, while that of the sensed world is indepen-
dent of the mirld .

One further explanation might be offered. We become con-
vinced that an independently real world of coloured objects
exists the moment we realize that the mind is not completely
master of its sensory experiences. For instance, I cannot see
what I want to see. I can imagine the deep purple of a kingly
robe at the present moment, but I can only see the white which
I (unphilosophically) take to be the colour of the paper before
me and the brown of the table. However strong my desire to
see purple, and however clear the image of purple before the
'inward eye', I yet do not see purple. Once 1 realize this, it is
urged, it then becomes natural for me to infer the existence
of an independently real world outside consisting of so many
coloured objects, and to conclude that I shall not see purple
until something which is purple comes within my range of vision.

Now it seems fairly obvious that what we see is in some
measure independent of us. 1 Yet our belief in the existence of an

1 Though, indeed, in certain abnormal cases the opposite seems
to be the truth. For in hysteria the patient thinks he sees outside
him what exists only in his imagination, and lunatics no doubt have
often been convinced that they see the purple robes of sovereignty
upon them although they are really dressed in ordinary clothes. But
for the moment we can postpone consideration of these cases.

56 THE NATURE OF KNOWING

independently real world in sensation can hardly have arisen
from our realization of this fact. If the mind becomes aware of
external existence only when it has explicitly realized that in
sensation it is not free as to what it senses most of us would
still be without the conviction that there exists an indepen-
dently real world of coloured objects. Yet this conviction as a
matter of fact seems always to be present, however far back
we go in our experience, and certainly it exists long before
we become aware of the mind's determination in sensation.
Indeed, it seems rather ridiculous to suppose that at some early
date in my mental history I chanced to be seeing yellow but
wished to see, say, blue, and so concluded, since my wishes
were frustrated, that this experience of seeing was, in part,
out of my control, and that therefore there necessarily existed
an independently real world of coloured objects.

On the contrary, we seem to be convinced of'thc existence of
an independently real world consisting of so many separate
coloured objects simultaneously with our seeing the colour,
and not by any inference from it. Therefore, w r hile this argu-
ment from our determination ab extra in sensation may be of
great value as confirming our conviction that there exists an
independently real world, we cannot admit that the conviction
originated with it. It is obviously prior to it. From the first
we seem to feel certain that there is a world around and outside
of us as sensing, and we implicitly believe that the colours we
see belong as qualities to the world about whose existence
we are so convinced. In a word, we straightway see, so we
suppose, a coloured world out there.

These attempted explanations, therefore, have all proved
unsatisfactory. None the less the facts which they seek to
explain remain indisputable. The sensory experience is not
merely a seeing of colours, a hearing of sounds, and so on.
There is always present in each and every sensory experience
that which claims, at least, to be knowledge. Throughout we
are convinced that we are here and now apprehending a world
of objects. The latter are not created by us, and are not of the
character as the images of our imagination, but are

THE SENSORY EXPERIENCE 57

independent of us and, in this sense, outside us waiting to
be apprehended. Furthermore, we are also convinced that these
objects possess various sensory qualities, colours, sounds, tastes,
and so on which we apprehend in sensation.

Now how far are these convictions valid ? Are they justified
and warranted by the rest of our experiences, and by our
later reflections upon the sensory experience? It was pointed
out earlier that we are frequently convinced without actually
knowing at the time. Error would not be error were we not
convinced whilst erring that we were knowing the truth. 1 But
do we know the truth or are we in error when we are convinced
in the sensory experience that we now apprehend an inde-
pendently real world of coloured objects?

To this one question there are two parts. Firstly, do I see a
world of independently real coloured objects ? Or to consider
another of the Senses do I taste an independently real object
which has, as one of its qualities, a certain taste? At the un-
reflective stage we are convinced that such is the case. The
colour, the sound, the taste, the odour, the resistance belong
to the independently real object as so many qualities. The
apple before me is itself red, sweet, hard, and so on, quite apart
from my seeing, my tasting, and my touching. But is our con-
viction sufficient evidence that the independent real does possess
these properties? We have already answered this question. In
the argument of the present section we have tried to make it
clear that such a conviction is wholly unwarranted by the facts
of our sensory experience. The fact that I see colours does not
justify me, surely, in holding that the independently real world
is a world of coloured objects. I only believe that it does when
I take it for granted that the seeing is itself knowing, whereas
reflection shows that it is nothing of the kind. It may scil! be
admitted, however, that the object possesses some property or
quality which causes me to see a certain colour at the present
moment and in the present circumstances. (Were it not so I

1 Consequently, doubt is not error, nor is suspension of judgement,
for in such experiences there is no conviction that we now know. For
the moment I leave out of consideration error in probable judgements.

58 THE NATURE OF KNOWING

could not begin to explain why it is that I see one object to be
red and another to be blue in like circumstances. It may even
prove true that actually independently real objects are coloured,
though at present I certainly do not know this.) Our only point
now is that in merely seeing a colour we do not know any real
quality of an independently real existence directly, and that it is
pure assumption on our part to suppose that we do. We are
not, therefore, justified in our first conviction that the indepen-
dently real consists of so many coloured objects. We have not,
of course, proved this conviction of a coloured world to be
definitely erroneous, but we have shown that it is pure assump-
tion and so quite possibly erroneous. That we immediately
see an independently real world of coloured objects must be
written dow r n as a completely unverified assumption.

But, secondly, a more important question remains to be
considered in this connection. Do we really apprehend an
external existence whilst sensing? We certainly feel convinced
that we do in an unreflective mood. If I close my eyes and
begin to imagine I know that all the occurrences imagined
by me belong to a world of imagination within my mind. But
once I open them I am convinced that I look out upon a world
other than, and independent of, my mind. Now even though
it be admitted that we do not see and hear actual qualities of
the real objects outside, can we not assert without fear of con-
tradiction, that in the sensory experience we apprehend the
existence of such objects and of an external independently
real world ?

Our conviction that such is the case seems so natural, so
universal, and so certain, as to appear unassailable. Nevertheless,
certain considerations which we must now make cannot but
throw doubt even upon this conviction. In certain cases of
hysteria the patient is convinced that he sees what can only
be the creation of his imagination. He sees occurring externally
to him and, as he believes, independently of him, happenings
which we know could not occur anywhere except in his dis-
ordered imagination. Here then is a case of imagining in which
the person imagining is convinced that he is now knowing

THE SENSORY EXPERIENCE 59

occurrences in an independently real world. Or we may take
the more normal experience of dreaming. I may dream of many
strange events and 'see' them all occur before my eyes in the
world outside and around me, but when I wake up I realize that
I have been dreaming and have actually seen nothing outside me
in independence of my own imagination. But if this is so, then
clearly in dreaming and in hysteria my conviction actually
misleads me. And if it be proved at fault in hysteria and dream-
ing, my faith in my conviction whilst sensing must also be shaken.
May not the whole of my experience be a dream which I myself
spin out of my imagination ? May not my conviction of an out-
side world be as illusory in the sensory experience as it is in
dreaming or in hysteria?

Now as long as these questions remain unanswered we
have no right to claim for our conviction absolute validity.
We can no longer be theoretically and absolutely certain that
the sensory experience is in part an apprehension of an inde-
pendently real world around us. But again and this now
becomes a matter of vital importance for our argument we
have still less right to deny this assertion outright. Because our
conviction is illusory in dreaming and in hysteria we have no
right to assume that it is equally illusory when sensing ; because
our experience is defective in one respect we cannot argue that
it must be defective in all respects. And certainly we have no
right to infer that the world about whose independent existence
we are so convinced in the sensory experience is actually
dependent upon me and is identical in character with the world
of imagination. With full confidence, then, we can neither
assert nor deny the existence of an independently real world
if all the information we possess is that given us in the sensory
experience.

Nevertheless, though we are conscious of the fact that com-
plete knowledge in this connection is out of the question at
present, we can yet claim that one position is more likely to be
true than another. Indeed, it is now becoming obvious that
most of what we ordinarily call knowing is not knowing in
the strict sense that is, knowing apodeictically but is rather

60 THE NATURE OF KNOWING

the determining of what is most probably true in any set of
circumstances. It would be absurd, therefore, to deny the
possibility and the actual existence of what w^e term 'opinion'
retaining the term 'knowledge* for absolute and certain
knowledge. Now our first conviction that there exists an
independently real world the conviction that makes solipsism
appear from the first an absurd theory becomes, upon reflec-
tion, an opinion whose truth is extremely probable. Doubt
creeps in when we recall such experiences as dreaming and
hysteria. Yet, \vhile we cannot be certain, it seems exceedingly
probable that an independently real w r orld exists, and exceed-
ingly improbable that the whole of our life should be nothing
but a dream. All the rest of my experience seems to confirm
my belief that there are other existences, independent of me.
Certainly, metaphysical speculation may lead me to say that
the Real in, or the Essence (and so the explanation) of, all that
lies about me as well as of my own existence is Mind. Even so,
it would remain extremely probable (though not certain) that
/ am not the mind that brought into being these things around
me, as I bring into being an occurrence in a dream. And though
I have not complete and absolute certainty on this point I live
out my life on this hypothesis, and thus far apart from the
doubt already mentioned no other has arisen. Here, therefore,
is something which seems exceedingly probable. But it is not
absolutely certain knowledge. It also must be written down
in the last resort as something taken to be true, without our
being completely certain of its truth. 1

Is there, then, anything of which we are completely certain
in sensation? Obviously, the moment we reflect, we are com-
pletely certain of one fact, namely, that we do have this experi-
ence. For whether awake or dreaming, in a normal or an abnor-

1 There are, we repeat, degrees of probability. And we may safely
say that this latter assumption is more likely to be true than our
earlier assumptions, namely, that independently real objects actually
possess as qualities the colours we think we see them to possess. On
reflection, and as our knowledge increases, this latter assumption
seems to become less and less probable.

THE SENSORY EXPERIENCE 61

mal state, I do now see this particular colour, and I know that
I see the colour. But this knowledge hardly occurs in the most
primitive forms of the sensory experience. It comes later. It
presupposes a reflective mood, and a capability of distinguish-
ing between what is external to me as experiencing and what
is internal. But this explicit distinction is not made at the lowest
levels. It is certainly assumed in thinking of the physical object.
Yet it is only later, surely, that it becomes explicit. The truth
seems to be that the question of the dependence or indepen-
dence of these physical objects and real entities which I think
I see upon my mind does not really arise in these primitive
experiences. It only arises explicitly for the reflective philo-
sophical person. And in just the same way we have at this
level no explicit idea of self, and cannot be said then to feel
certain that we now enjoy a particular experience for example,
seeing a colons. Thus it would be hardly correct to say that
throughout the sensory experience we find a knowledge of the
self as enjoying a sensation. But is there, then, any certain
knowledge in these lowest sensory experiences ?

We believe an affirmative answer must be given to this ques-
tion. We doubt the existence of coloured objects. Furthermore,
the moment the matter becomes explicit we find it possible to
doubt that in the sensory experience we have a knowledge of
independently real physical objects. But present in all sensory
experience we find one conviction that we have never yet
succeeded in doubting. It is the conviction of real being. In
the sensory experience we are aware of existence (whether
internal or external to mind). We believe we know much more,
we 'see', so we suppose, coloured things. This, however, can
be doubted. But that sense-experience involves an awareness
of existence cannot be doubted. It is so far cognitive. That is
to say, if we take the sensory experience of the most unreflec-
tive person and abstract from it all that he thinks he knows,
but that we see reason to doubt, there will still remain a core
of certain knowledge which cannot be doubted. Viewed cogni-
tively, sensory experience is largely erroneous, but it is not
completely so.

62 THE NATURE OF KNOWING

"But", it will be objected, "what does this person know?
One cannot know a that unless one knows its what. One cannot
know bare existence, one can only know something having
definite qualities. But the only certain knowledge you allow
the mind in sensation is that of bare existence. Yet no mind
can ever know bare existence as such. For as such bare existence
is just nothing." We admit the strength of the criticism and
do not wish to avoid it. Our answer is that the sensory experi-
ence is never merely the knowledge of bare existence. Bare
existence is not the content of the experience. It is the result
of our abstraction and analysis. What we believe we know
when we sense is a world of sensible objects. But we have
shown that we are not justified in our belief. All we can definitely
and justifiably be certain of is that actually something exists.
The mind, however, unjustifiably but very naturally, 'sees'
in the content 'given' by the senses the 'stuff 'of the real, and,
as a consequence, believes itself to possess a more extended
knowledge of the real world in sense-experience than it does
actually possess.

If it now be added that every other human experience
involves so much knowledge, since there is no experience in
which we are not aware of existence and are not apprehending
some reality of some sort, we shall not deny this. What we
claim for the barest sensory experience that can be imagined,
from the point of view of true and certain knowledge, is very
little indeed. But we believe that it was necessary in a full
and complete analysis of the sensory experience to mention
this essential component of each sensory experience, however
bare.

We are now in a position to carry out a three-fold analysis
of the sensory experience taken as a whole. In the first place,
we note a certain determination or affection of the mind in
sensation. We find ourselves compelled to see blue on one
occasion and yellow on another. However strongly we will we
cannot will to see a particular colour. The mind in sensing is
not completely master of the situation ; in some measure it is
determined. Here is one essential feature of the sensory expe-

THE SENSORY EXPERIENCE 63

rience which it is always dangerous to disregard. A second
feature is the seeing of the colour, the hearing of the sound,
and so on. We are not merely determined, but we also see. This
seeing of a colour could never occur were we not beings who
possessed the qualities and capabilities (both mental and
physiological) necessary for its occurrence. So also with hearing
a sound, tasting a taste, and so on. 1 To be determined is not
enough. But when we, who possess these qualities, are deter-
mined in this manner, then we see, hear, taste, and so on. In
the third place, there is a cognitive side to the sensory experi-
ence as a whole. This is not the seeing of a colour, but is to
be distinguished from it; for seeing a colour is not knowing,
neither is hearing the sound as such. The sensory experience
as a whole, however, is never the mere seeing of a colour.
Throughout we are at least making claims to know. We have
been led to conclude that most of these claims are invalid.
Much of the so-called 'knowledge' in the sensory experience
is really probable only and ought better to be termed 'opinion'.
Its defect is that it does not carry with it complete certainty.
Moreover, some portion of what we naively claim to be know-
ledge is definitely error. Indeed, only one claim seems to be
unquestionably valid for all cases of sensory experience, namely,
that in it w r e know real being. (For the sake of completeness,
however, we should also add that already we, at least, opine,
if not know, 2 temporal and spatial relations, identities, causal
relations, and the like in the sensory experience.) Thus the
sensory experience taken as a whole possesses definitely a
cognitive feature ; and this is especially true if within 'cognition'
we include, as is usual, opinion, and even error.

Also, it is necessary to mention the activities of mind qua

1 I have not sought to give any account of the nature of these
necessary qualities. Such an account would definitely go beyond the
scope of this work. All I wish to show is that this feature of the whole
sensory experience is distinguishable (though not separable) from the
knowing present.

J At present we need not discuss whether this is knowledge or
opinion.

64 THE NATURE OF KNOWING

comparing, relating, and recollecting within the whole sensory
experience. These clearly pertain to its cognitive side. We have
been considering the sensory experience in its barest forms,
yet in no case, it would seem, are the above activities wholly
absent. In the more developed forms of the sensory experience,
as, for instance, in adult experience, they are obviously present.
We gain knowledge and come to an opinion mediately after
much comparison, recollection, and reflection. We shall con-
sider the mediation involved in greater detail in the next
chapter. But it also seems to be present even in our earliest
sensory experiences, especially so if these call forth any
explicit and significant statement on our part, however
simple; for instance, ''How hot!" "This is blue." A
discussion of this matter, however, would be irrelevant at the
present moment.

We thus acknowledge the existence of these* activities in the
sensory experience, but postpone consideration of them to
the next chapter. It is as well, how r ever, to point out at once
that we do not intend to adopt the further position, sometimes
held, that the purpose of these activities is to help us create out
of the content 'given' by the senses a significant sensible object.
We have throughout combated this view. There is no more
serious error in the realm of epistemology, we should like to
suggest, than to suppose that knowledge comes about by
'working* upon the sensory qualities provided by our senses.
The ultimate consequence of such a view, as we have pointed
out, is always the same; we find ourselves shut off from the real
world and confined to the phenomenal. And this consequence,
we think, results from the failure to analyse the sensory experi-
ence properly. It is due to mistaking what is not cognitive in
the experience for a cognitive feature of it. The colours seen,
the sounds heard, and so on, may be termed, if we wish it,
'appearances'. But they are not to be thought of, on any account,
as in themselves leading us to, or in any way suggesting,
knowledge of the real, they provide such knowledge neither in
part nor wholly, neither, again, directly nor indirectly by way
of copy and illustration. Nor, lastly, can we gain any knowledge

THE SENSORY EXPERIENCE 65

whatever of the real by reflecting on such colours and sounds
in themselves, however hard we reflect. Whatever knowledge
itself is, it is not the seeing of a colour, nor is it any kind of
intellectual 'working' upon the bare colours, sounds, and so
on, provided in sensation. We should rid ourselves of the
misleading idea that we know a half-real shadow-world in the
sensory experience, and that by 'working* upon this shadow-
world we may in time come to know the real which lies behind
it. As much as we know in the sensory experience is the real
itself and no half-real. And if anything helps us in future, then
it can only be the knowledge we actually do gain.

At present, then, as the result of our reflections in this chapter,
we have little to say about the nature of knowing. For that
which is knowing in the sensory experience is so minute and
so inextricably woven into the texture of the whole experience
that it is exceedingly difficult to detach it in order to observe
it as it is in its own nature. Hence, though we may feel
convinced that every sensory experience involves a knowing,
namely (at least), the apprehension of the existence of a real
world, yet we cannot hope to gain much positive information
about the nature of knowing from scrutinizing it as it is em-
bedded in the sensory experience. It would indeed be foolish
to expect the nature of knowledge to be best revealed in that
experience wherein it seems least present. In the sensory
experience knowledge is at a minimum. Consequently, if we
wish to discover the nature of knowing it would be wise to turn
away from the sensory experience. There are other human
experiences where the knowing involved can be more easily
studied. In the sensory experience the cognitive element does
not even predominate; for, undoubtedly, the characteristic
feature of any sensory experience is the sensing of colours,
sounds, tastes, and so on. Thus the knowing in it is hidden
and encumbered by what is not knowing. We need to find an
experience, in which knowing will be freed from some at least
of the encumbrances that surround it in the sensory, so that it
can be more easily examined and observed.

For these reasons we must content ourselves at the end of

66 THE NATURE OF KNOWING

this chapter with the negative information which is all that we
have thus far obtained. The sensory experience, we conclude,
certainly does not present us with that example of complete
and perfect knowledge for which we seek. At most, it can only
provide us with one brief glimpse of our quarry in the distance.

DISCURSIVE REASONING

THE argument of the preceding chapter has thrown little
light upon the character of knowing. On the whole, it would
seem untrue to say that the sensory experience is ever com-
pletely non-cognitive. But what is not knowing in the experience
is so closely bound up with the actual knowing present, and
is so frequently confused with it, that the use of the sensory
experience as an instance of knowing is unsatisfactory and even
dangerous. For it is only with difficulty that we free ourselves
from the naive and completely misleading tendency, so deeply
rooted in our minds, to regard the bare seeing of a colour as
itself an infallible knowledge of the external world. The first
truth that must be learnt is this negative one that knowing is
something wholly different in character from merely seeing a
colour, hearing a sound, and so on. And since such bare
sensing is the most prominent and the most characteristic
feature of the sensory experience (though never the whole
of it) the latter cannot but prove a very poor instance of
knowing.

Now a far more valuable instance, it may be said, is to be
found in discursive reasoning. Here, certainly, is a conscious
attempt at knowledge. Prima facie, it seems to consist in the
effort to attain further knowledge indirectly (that is, through
another or other known truths) when the direct approach is
impossible. Discursive reasoning is in this respect synonymous
with mediation, and most of the knowledge (and opinion)
gained by mankind seems to be gained in this way. For
instance, in our developed sensory experiences the presence
of discursive reasoning is obvious. We succeed in making new
assertions through comparing and relating the present * given*
with our other experiences; we discover unity in differences
and differences in unity; we 'see* things in a wider context
and relation; and thus argue from truths already known to
others not yet known. The sensory experience in its concrete-
ness, as it occurs in the ordinarily developed adult mind, is as
much discursive reasoning as it is sensing. And as our know-

67

68 THE NATURE OF KNOWING

ledge increases so we find greater need for discursive reasoning.
Now, if such be the case, a careful examination of discursive
reasoning or mediation should give us real insight into the
nature of knowledge. It must be the instance of knowing for
which we seek.

In the present chapter we propose to examine this claim.
What is discursive reasoning? Is it through and through
knowing and nothing but knowing ? Or is it a complex process
like sense-experience, of which knowing is part only? If the
former, how best can we describe the knowing? If the latter,
then, again, what characteristics belong to the knowing
involved, and, in addition, what other feature or features
different from knowing are present and what function do they
fulfil ? These are the problems that must engage our attention
in the present chapter.

Before we can proceed, however, to thek- detailed con-
sideration, and before we endeavour to give a more definite
meaning to the terms used thus far in describing discursive,
reasoning, certain very important preliminaries must be dis-
cussed. These are themselves of the greatest interest from our
point of view, because of the further light they throw on the
nature of knowledge.

For, firstly, how do we come by that prior knowledge pre-
supposed once we conceive discursive reasoning to be the
gaining of new knowledge through truth already known ? The
prior knowledge, it may be answered, was itself known in
some earlier mediate process based upon still earlier know-
ledge. But clearly this cannot go on for ever. There must be
a basis, which is known not by mediation but by some other
means, upon which all our mediate knowledge rests as on a
foundation. But how did this first knowledge come about?
And is it wholly different from the knowledge found within
the process of discursive reasoning? If it be so, is it a better
instance of knowing than discursive reasoning? And are there
then many kinds of knowing, each distinct from the others ?

In the second place, what exactly is the content of the
knowledge presupposed in mediation ? It is evident from what

DISCURSIVE REASONING 69

has been said that some knowledge must always be presupposed
in mediation. But knowledge of what? The difficulty in seeking
to answer this question arises from the fact that the prior
knowledge is required for at least two purposes, the one
radically different from the other. In the first place, we require
a logically secure foundation for the structure we are about
to build by discursive reasoning. Unless the basis and principle
of procedure are logically valid, the whole structure will be
unsound. We must know beforehand what is and what is
not logically valid. It is this kind of prior knowledge which
we shall discuss in the first section. But, in addition, the
discursive reasoning cannot proceed unless some * stuff* is
known beforehand. The thought-process begins with some-
thing already known, not only with a knowledge of the
logically valid, but also with a certain definite content. This
content, it is usual to suppose, has already been gained by the
mind in 'experience'. Before the process of mediation could
ever begin the mind has already 'gained certain data'. Now
such language is distressingly vague. What is this 'experience'?
And how does it provide the 'stuff' beforehand? The more
nominalist type of answer to these questions will engage our
attention throughout the second section, but no final answer
can be given until the nature of mediation or discursive
reasoning itself has been made clear in the third section of
this chapter.

I

The Prior Knowledge of Principles

In this section we shall consider the knowledge of the
basic principles presupposed in all discursive reasoning. We
shall show, firstly, why such prior knowledge is necessary.
In the second place, we shall defend the thesis that such
principles are necessarily universal and all-pervasive features
of real being. Thirdly, we shall discuss the nature of the act
of knowing by which the principles are apprehended.

70 THE NATURE OF KNOWING

We may best begin by reiterating a point already made,
that any mediate process which is to provide us with demon-
strable knowledge must rest ultimately on a sure foundation.
And this means, in part at least, that as we advance step by
step in the argument we are throughout guided by sound
principles. Long before we bother to formulate these principles
precisely, however, we apprehend the truths they embody,
for example, that a thing cannot contradict itself, and we
make use of them when reasoning discursively. Indeed, it is
only when we are compelled to defend our position or when
the sceptic in us awakes that we realize how all along we have
proceeded on the understanding that certain primary principles
which we now seek to formulate in precise terms were true,
and how, throughout, the validity of our reasoning depends
upon their truth. But it is clear that the ultimate source of
our assurance as to the validity of any reasoning process lies
always in some principle (or principles) which we see to be
indubitably true. 1 Consequently, if we affirm that mediation
yields demonstratively valid knowledge, then we presuppose
the existence of a prior knowledge, namely, that of the first
principles of reasoning, upon which our discursive reasoning
depends and by which it is guided throughout. The necessity
of this prior knowledge is obvious.

Moreover, this prior knowledge of the principles must be

1 We do not mean that these principles are primary premisses in
the sense of starting-points for any particular science. They are
rarely, if ever, premisses at all. Every particular science has its own
primary premisses. For instance, if we take the particular ground for
any statement in arithmetic, this will be grounded on another state-
ment, and ultimately we shall come to one about perhaps the nature
of the unit used, or again about arithmetical progression and scale.
These final statements may be regarded as primary premisses of the
science. But they, and their like in other sciences, are not what we
have now in mind. The principles according to which we reason in
any mediate process are not confined to any particular science. Nor
are they first premisses or definite starting-points. They are rather
principles in accordance with which and not from which we invariably
argue.

DISCURSIVE REASONING 71

completely certain and beyond all doubt. For if any demon-
stratively valid knowledge is ever to be possible it must be
based on our conviction that the foundations of our reasoning
are secure; and if we are not certain of these foundations, if
we have the slightest cause for doubt, then we can put no
confidence in the structure erected on them, however securely
built it appear; we cannot, that is to say, but be sceptical.
Furthermore, in the present case, we cannot save ourselves by
making a show of suspending our judgement. It may be true
that these principles are not valid; but if we once suspect
this, then, while we continue to suspect, we must also doubt
all knowledge gained discursively. We must, therefore, if we
wish to affirm that discursive reasoning does give knowledge,
also affirm that the first principles necessarily presupposed
within it are valid. Thus, to all intents and purposes, the
implicit denial present in the hesitation of one who withholds
his judgement is in this case of like nature with and produces
.the same result as the explicit denial of the sceptic. 1

But, granting the necessity of a prior indubitable know-
ledge of such principles, what do we know when we know
them? In the first place, since we are dealing with an instance
of knowledge, the object in this case must be identical with
the object of all other knowledge, namely, reality. In the
second place, this view is confirmed by a further reflection.
Unless the principles are valid, as holding of the real world,
then no knowledge of that real world can be gained discur-
sively. We have taken it for granted that knowledge must be
of what is. Consequently, the mediation could not provide us
with knowledge if the principles in accordance with which it
proceeds did not hold of the real. If we started with principles
not holding of the real and allowed ourselves to be guided by

1 We may here add, as something that follows obviously upon the
above reflection, that the first principles cannot be 'ordering con-
ceptions' in the sense of hypotheses made by us that may or may not
be true. The first principles must be true, or else all the rest of our
thinking is invalid. We do not seek to establish them by mediate
reasoning, as is the case with hypotheses, but we establish all that is
established mediately by reasoning according to them.

72 THE NATURE OF KNOWING

them, then discursive reasoning would become, at best, a form
of intellectual entertainment, pleasant enough in itself perhaps,
but possessing, from the point of view of knowledge gained,
no value whatsoever. It could tell us nothing about reality.

It follows, therefore, that in knowing the principles accord-
ing to which we reason we must be knowing certain charac-
teristic features of the real world as it is. We cannot avoid
this conclusion without denying the very possibility of know-
ledge through discursive reasoning. We certainly could not
avoid it by asserting that the principles hold of a half-real
phenomenal or 'sensory' world immediately before us in
thinking. For we have earlier 1 agreed to reject all pheno-
menalistic accounts of knowing. Knowledge, we contend, if it
exists at all, is of the real, and not of what we know to be
merely phenomenal. Hence, unless these principles, knowledge of
which is presupposed as providing the foundation for mediation,
are of the real, we cannot hope to gain knowledge discursively.

In order to illustrate the matter further, we may consider
the principles individually. The principle of Non- Contradiction
is formulated thus: <( The same attribute cannot at the same
time belong and not belong to the same subject and in the
same respect." As thus formulated this is one of its earliest
formulations (and surely its best) it is evident that the
principle has to do with the real, and not with any merely
ideal world. It does not apply merely to thought processes;
though it does apply to them, since it applies to all things real.
But the point is that in apprehending it we are apprehending
a characteristic mark of what is. Reality, that is to say, is of
such a nature that we are compelled never to admit patent

1 In the second section of the first chapter. If the 'objective* world
is conceived as a world of phenomena largely dependent upon the
knower, no self-contradiction need then be involved in a theory
which asserts that the principles holding in the 'objective' world are
themselves the consequence of mind's synthetic activity. But we
reject phenomenalism, and certainly should be contradicting ourselves
if we put forward any such theory. The only possible position for us
is the one set forth above. The principles known must be universal
features of the real which we discover,

DISCURSIVE REASONING 73

contradictions into our thinking about it. The same is true of
the principle of Identity "whatever is, is", or "a thing is
what it is". The reference here is to real being, to everything
that is. Anything possessing reality is what it is, and we cannot
think it otherwise. Thirdly, we have the principle of Excluded
Middle: "a subject either has or has not a certain attribute";
there is no third possibility. Here, again, knowledge of the
principle is an apprehension of a fact about the real.

Such are the three principles recognized by logicians as
being the bases of all thought the so-called Laws of Thought, 1
The use of this latter term must not be taken to imply that
the laws are merely of thought, or ideal only as opposed to
real. For, once again, what is meant when we emphasize the
need for such Laws of Thought as a basis is that human
thinking and reasoning could never provide knowledge of the
real unless th$ mind already possessed in reasoning a clear
apprehension of certain facts about the real. We never carry
forward our processes of mediation in complete ignorance of
the matter under consideration. Throughout we have know-
ledge of the real, but in discursive reasoning we pass on to a
fuller knowledge of it. And we may legitimately press this
point further. Such prior knowledge gives us information as
to the character which reality everywhere and always, uni-
versally, possesses. Before we can begin to reason discursively
w'e must (speaking logically) first know the real, and also know
something which holds universally of it. The real must be
knowable and known by the human mind, even with regard
to certain universal features of it. This is the ultimate pre-
supposition of all discursive reasoning.

Such a conclusion seems inevitable. But what exactly does
it signify? All knowledge gained demonstratively necessarily

1 We emphasize the fact that they are the bases of all thought
because their universality and all-pervasiveness is their distinguishing
mark. And this must mean, we contend, that they hold of all reality.
Whatever is real, whether it be 'subjective' or 'objective', must
possess these features, and the mind cannot think of any real exist-
ence without conceiving of it as identical and self-consistent.

74 THE NATURE OF KNOWING

presupposes a knowledge of certain universal characteristics of
reality, and if the former occurs, then clearly the latter already
exists. Yet these obvious truths, we must now point out,
throw no light on our actual knowing of the first principles.
They simply show the necessity of that knowledge if further
knowledge is to occur. Now, in the first place, it is still possible
to doubt the occurrence of any further knowledge ; and in such
a case the necessity for a prior indubitable knowledge of
principles would not arise. This, however, is a position we
have agreed to reject. But, in the second place, even if we
admit the existence of an ever-growing body of certain know-
ledge based ultimately on the principles, w r e can only argue
from this to the necessary existence of the prior knowledge.
We are given no information about the nature of that prior
knowledge. The above argument can at most only confirm its
existence by showing its necessity if knowledge has been
gained demonstratively. Hence, to say that knowledge of the
first principles is necessarily presupposed by all further know-
ledge is not to describe its nature. And this point should be
emphasized, since it is sometimes carelessly assumed that we
have said all there is to say about knowledge of the principles
when we say that it is necessarily presupposed in all demon-
stration and therefore must exist.

How, then, are we to describe this knowing? We have
already learnt that it cannot be mediate knowledge. Obviously,
we do not know the principles with the help of principles
that is, through any process of demonstrative reasoning. Nor,
again, can we say that the 'objective' world is itself mind-
ordered, and that, therefore, the principles according to which
the mind thinks are the very principles that order the 'object-
ive' world. Finally, such knowledge cannot be 'given' by the
senses; for they, if they 'give' any knowledge at all and we
deny this 1 could only 'give' knowledge of the present, the
'here-now', whereas in knowing the principles we know

1 At the same time we, of course, admit that the determination of
the mind in sensation is an occasion for knowing, and that the sensory
experience, taken as a whole, is cognitive,

DISCURSIVE REASONING 75

universal and all-pervasive characteristics holding throughout
reality. What, then, are we to say? How do we know, for
instance, that one and the same attribute cannot both belong
and not belong at one and the same time to the same subject
in the same respect? We can only answer that we know it as
something self-evident. We directly 'perceive' the truth of the
statement, and know beyond the possibility of doubt that
reality never contradicts itself. It is immediate apprehension
of the truth. The mind apprehends directly by that knowing
power which belongs to it this universal feature of the real.

From the time we first begin to reason discursively, then,
we know that the real world does not contradict itself, and we
reason on this basis. Moreover, the fact that we do not at the
time formulate this truth in precise logical terminology in no
way deprives the knowledge we possess of its certainty. Our
knowledge herE is no opinion to which we hold even though
we realize its possible falsity; nor, again, is it Vague' or
'indefinite'. It is definite enough, as our use of it in reasoning
proves. It is certain knowledge that w r e may or may not possess.
We either know or do not know it, but we cannot be said to
know it dimly, for if we once see that what is real cannot
contradict itself, then we cannot later gain any greater clearness
or any greater certainty on this point. 1

' Though we believe that this knowledge comes early in the mental
history of a rational being, we do not wish to imply that it is innate.
And, certainly, we should reject the view that because it is a self-
evident truth that the real cannot contradict itself it must there-
fore be innate. A truth's self-evidence in no way makes it innate.
For, obviously, what is perfectly sclf-c\ident to a trained mind may
be incomprehensible to the untrained. In the last section of this
chapter we shall point out how much mental preparation is frequently
necessary before the mind is enabled to apprehend the self-evident.
And, no doubt, we should not have learnt that the real is non-
contradictory had we not first experienced certain experiences (for
instance, the sensory experience), so that we can hardly be said to
know even this truth innately. It is a later discovery. Our poinr,
however, is that from the first discovery of it our knowledge here is
certain and indubitable a direct apprehension of one characteristic
feature or mark of the real world.

76 THE NATURE OF KNOWING

If, then, we ever have certain knowledge, we have it when
in discursive reasoning we proceed on the understanding that
the real does not contradict itself. But what proofs have we
to offer whereby this claim may be substantiated? Obviously
none, except simply the proof of our own compelling con-
viction when we consider the matter. And if it be objected
that the convictions of a fallible creature are an insufficient
criterion, we can only reply that we cannot bring ourselves to
admit the possibility of these principles being untrue. To deny
them is to talk patent nonsense. To deny them significantly
and not merely in words would be to conceive and to
assert their opposite. But we simply cannot conceive a world
in which a thing is what it is not. And what is true here is
true also of the other principles. We cannot deny them. We
may adopt, of course, the agnostic's attitude and believe
nothing. But though we should thus cease to tiffirm the prin-
ciples, we should not be denying them. Moreover, this attitude
is itself hardly possible. We know that the real does not con-
tradict itself, and on this matter we cannot not believe. Yet we
cannot give any reasoned ground for our belief if it were at
all necessary. We believe because we know the real and arc
convinced that we know.

This is clearly an important matter, and we shall return to
it later. In the meantime, it is obvious that if we have any
knowledge at all of the real, we have it when we know the
principles, and we shall proceed on this understanding. It is
now, however, necessary to consider a further problem that
arises in this connection. Arc we to regard this knowledge of
principles as something unique? In particular, does it com-
pletely differ in kind from the mediate knowledge which
presupposes it? There will be no need to stress the prime
importance of this problem for the present essay.

Knowledge of the first principles, we have agreed, cannot
themselves be known demonstratively. It was the recognition
of this truth that first led to the explicit formulation of the
distinction, which has since played an important part in
cpistemological discussions, between mediate knowledge on

DISCURSIVE REASONING 77

the one hand, which involves a process, and immediate on
the other. The knowledge of the principles must be of the
latter type. It cannot be the product of any inferential process,
for every process of inference is based upon it. So much is
clear. But this whole distinction is frequently set forth as if
there were two completely distinct kinds of knowledge, capable
of functioning apart from and wholly independent of each
other, and as if mediate and immediate knowledge were abso-
lutely different in their nature. Now, is this view sound? If
accepted, the complexities of the epistemologist's problems
are so increased that it becomes wellnigh impossible for him
to give a satisfactory and consistent account of knowing.
This, however, should not restrain us from accepting the view,
if we can prove it to be sound. But we can hardly claim that
we are now in a position to offer any such proof. For, assuredly,
before we can* assert with any show of authority that the
knowledge involved in mediation is absolutely distinct from
that of the first principles, our understanding of both kinds
of knowing ought first to be intimate and thorough. But
actually, while we have very little information as to the know-
ledge of first principles, we have at the present stage even less
with regard to mediation. Therefore, until fuller information
is obtained, the absoluteness of the distinction remains simply
a conjecture, premature and as yet unjustified, which the facts
when revealed may wholly fail to warrant.

Indeed, one interpretation of the distinction can easily be
refuted on reflection. It may be said that the knowledge of
the principles is unique (completely distinct from mediate
knowledge) in the sheer purity of its intuition, an intuition
which is so pure and so distinct that it needs an object which
is itself distinct and purely simple. Now the raison d'etre of
such an interpretation is not far to seek. So closely is object
related to subject in knowledge that if we once conceive of
the knowing as distinct and unique we tend naturally to think
of its object as being so also. A unique act of knowing suggests
an equally unique object. Furthermore, it may be argued that
the uniqueness of the object is not only suggested but made

78 THE NATURE OF KNOWING

necessary by the uniqueness of the knowing. For 'intuitive*
knowing, it may be (and often has been) said, is unique in the
sense that, unlike the laborious knowledge gained after long
processes of reasoning, it takes in the whole truth at once; it
'grasps' the complete real directly at one grasp (to use a
physical analogy). But if this is so, then it seems to follow
that the real or the true, the object in this case, must be that
which can be so grasped. It must, that is to say, be something
complete in itself, totally distinct from all else, an isolated
and discrete unit. What is known must be pure in the sense
that it is throughout one and the same; it must be simple,
having no parts and no connections ; it must be a perfect unit
which, being wholly independent of all else, is either known
by this one act of apprehension or remains unknown for ever.
Thus the exigencies of this view would compel us to see the
ideal of all knowledge, the perfect example for* which we have
sought, in a sheer intuition of sheer simples, an atomistic
intuitionism. 1

We shall devote the whole of the next chapter to a discussion
of intuitive knowledge. In this section, all we wish to show
with regard to it is that, whatever we assert about other
cognitive experiences, knowledge of the first principles cannot
be such sheer intuition of sheer simples. Against this position
three objections might be urged.

In the first place, it has been seen that the principles are
simply so many pervasive features or characteristics of the
real known by us. Now, clearly, a feature of anything cannot
itself be described as 'a discrete and isolated unit' without
talking nonsense. At most, such a description could only be
applicable to that of which it is a feature. If the above argu-
ment, therefore, demands a discrete unit as the object of the
sheer intuition, then, obviously, knowledge of the principles
could never be 'intuitive' in this sense. Consequently, the
description a 'sheer intuition of sheer simples' cannot apply
to our knowledge of the principles. We might, justifiably,

1 The writings of Descartes provide us with an excellent example
of this type of thinking; but it is in no way confined to his pages.

DISCURSIVE REASONING 79

advance this argument further. For if it be correct to hold
that such principles are in truth pervasive and common
features of all things real, then it seems to follow that there
cannot be any real object which is a discrete unit or an isolated
atom if, again, we mean by the latter an object different
throughout from all other objects. For the real world, if the
principles do hold of it, could never contain within it any
object having no possible ground of relation with other
objects, since the first principles would apply to each and
every object, and so there would always be something in
common. Such a discrete unit could only be an intellectual
entity created by abstraction from the real (if, indeed, intellect
could ever create a pure unrelated) ; and, therefore, in becoming
aware of it we should not be knowing what really is. It also
follows that the ideal of knowledge can never be a sheer
intuition of she^r simples in the above sense, since knowledge
is of the real, and all real objects must, at least, be related in
this respect, and have this much in common, that the first
universal principles hold of them.

In the second place, our knowledge of the principles makes
further knowledge possible. But an immediate apprehension
of a purely discrete unit could never help us in our knowing,
for though many such discretes were apprehended, no further
knowledge could ever be drawn from them by any process of
thought, and this because knowledge of a discrete unit could
never tell us anything of the world beyond it. A pure discrete,
that is to say, can never serve as a logical starting-point.
Consequently, if knowledge of the principles were a sheer
intuition of completely isolated discretes, then such knowledge
could never play the supremely important part epistemolo-
gically which our knowledge of the principles actually does
play. It could supply neither the logical basis nor the guidance
necessary for discursive thought. On the other hand, if our
knowledge of the principles does supply the required basis,
then obviously it cannot be an intuition of sheer discretes.

Thirdly, however and this is the most important objection
we have to make the position may be criticized on the

8o THE NATURE OF KNOWING

ground that it already isolates the knowledge of principles
from mediate knowledge. For the implication of the theory
under discussion is that the first apprehension of principles
used in reasoning functions apart altogether from the subse-
quent reasoning and is wholly distinct from it. In the first
place, it is implied, we gain knowledge of the principles
intuitively, and then, in the second place, we draw our infer-
ences and work to our conclusions. These two functions of
the mind, it is supposed, are wholly distinct, though the
second presupposes the first. This would mean that nowhere
in the mediate process itself does knowledge of the principles
occur, but that it only occurs before the process begins, and,
being in essence different from it, cannot occur as an integral
feature of it. Yet actually any mediate process that we take
into consideration reveals the presence in it of a knowing of
principles. Each step in a valid mediate process can only be
taken by a mind apprehending simultaneously the first prin-
ciples. Thus in mediate processes the mind adopts one course'
rather than another because it knows that, for example, a thing
cannot possess two contradictory attributes and so contradict
itself. Every movement in the process presupposes a present
knowledge of these principles that is to say, of the funda-
mental structure of the real. It is therefore absurd to suggest
that the knowledge of the principles is not an integral part s of
the knowing involved in mediation, since if it were not
throughout present there would be no mediation whatsoever.
Knowledge of the principles, therefore, is not something com-
pletely distinct from discursive reasoning. On the contrary, it
is itself a feature and a part of that reasoning, and without it
the reasoning could not occur.

Thus the knowledge of principles presupposed in mediation
cannot be conceived as an intuition unique in character and
differing absolutely from every other instance of knowing ; nor
is its object a mere isolated and discrete unit. If this were its
real character, mediation could not possibly presuppose it.
We should not, of course, deny that the knowing of the first
principles is 'intuitive' in the sense of being direct or imme-

DISCURSIVE REASONING 81

diate. We consider the description 'a direct and immediate
knowledge or apprehension of the real' an excellent one in
this context. What we should seriously question, however, is
the implication that the same description cannot be applied to
other instances of knowing, particularly to the knowing present
in discursive reasoning. Indeed, so far as we can see at the
present moment, the real difference between the knowing of
principles and that which occurs in and through discursive
reasoning may be external to the actual knowing itself, which
as such may be one and the same in both these cognitive
experiences. It is certainly not permissible to argue that since
the knowledge of the principles is logically prior to any know-
ledge that comes discursively it must be essentially different
from it qua knowing.

Mediation then, we can now say, presupposes as a wholly
necessary precondition a knowledge, in the first place, of exist-
ence a knowledge, as we contend, present in the lowest
forms of the sensory experience and in the second place, of
the general structure of that existence. The necessary priority
of the latter, however, we must hasten to add, is logical, not
chronological. For it is false to suppose that knowledge of the
principles comes first in time and is isolated from the know-
ledge present in mediation. On the contrary, knowledge of
the principles, we have seen, is an essential part of the knowing
in mediation, though we are not at present exactly aware of
how great a part. It is evident, however, that it cannot be the
whole of the knowledge therein contained, since in mediation
we claim to know more than the mere general structure of
what is, whereas knowledge of the principles provides us simply
with this bare general structure.

To conclude : though we do not understand as yet the exact
place of this prior knowledge in mediation, we have learnt
enough about it to enable us to assert without fear of refutation
that the knowledge of the principles cannot be regarded as an
'intuition' or direct apprehension, distinct from all other kinds
of knowing, and having a distinct object. On the contrary,
knowledge of the principles is so intimately related to the

82 THE NATURE OF KNOWING

mediation which presupposes it that it cannot be considered
as an instance of an isolated knowing complete in itself. Though
it itself is clearly not the product of any mediate process, it is
nevertheless embedded in processes of mediation. And before
we can understand its nature we must learn more as to the
meaning of the whole mediate process and as to the way in
which greater knowledge results from it.

The 'Experience' Presupposed

All discursive reasoning, we have urged, presupposes a
knowledge of the principles. Does it presuppose any further
knowledge? Obviously, it presupposes something further
(whether this be knowledge or not), since we could not pro-
ceed very far discursively if we were completely confined at-
the outset to a knowledge of bare first principles. The latter
is certainly a condition sine qua non of all sound argument;
but it does not provide premisses for any particular piece of
reasoning. We must begin our process of discursive reasoning
with definite information of some sort that may serve as
premisses for our argument. But how do we gain this informa-
tion? It is customary to answer that the content upon which
and about which we reason is provided us in * experience'.
Now this ans\ver needs to be examined further, for as it stands
many different interpretations of it are possible.

It is first necessary to set forth and criticize one interpreta-
tion which we believe to be fundamentally unsound, in spite
of the fact that, in some form or other, it appears frequently
in philosophical speculation. For it may be said that if expe-
rience does 'give' the content, then it cannot be the task of
discursive reasoning to 'give' it over again. If we already know
the content by way of * experience', then clearly we do not
need to re-know the same content by way of reasoning. The
view, of course, tacitly assumes that 'experience' and 'dis-

DISCURSIVE REASONING 83

cursive reasoning* are fundamentally distinct faculties of the
mind having different functions, and since it is the function
of the former to stock the mind with information about the
real, it cannot also be the function of the latter. As a conse-
quence, discursive reasoning as apart from 'experiencing*
cannot, strictly speaking, be a knowing. It becomes for such
thinkers something else, namely, a doing. In discursive reason-
ing we do something with the content already known.

The doing is a conceiving of concepts, whose necessity
becomes obvious on a moment's reflection. For 'experience',
on the present view, is supposed to provide us directly with
knowledge of different particulars as they exist in the sup-
posedly real world. But a world of particulars, owing to its
endless complexity, is difficult to handle. Therefore, the task
of reasoning is to conceive and use concepts or universals,
each of which s l cands for a bundle of particulars. By this means
the knowledge we already possess becomes easier to retain, to
apply, and to communicate. Concepts are shorthand notes that
save an immense amount of intellectual labour on our part.
But their conceiving is clearly not a knowing on this view,
even though we grant that greater knowledge may result after
conceiving them because of the real economy secured and
because the mind is then left freer than it would otherwise be.
Also, conceptualizing enables the mind to bring many par-
ticulars together, so that it can provide itself with a more
comprehensive view of the real known by experience. Yet the
conceptualizing itself is a doing here, and not a knowing. The
concept is created, not known. Moreover, it is created in order
to be used. We group together a bundle of particulars, agree
upon a common name for the bundle, and so facilitate future
reference to any particular within this group or class. But we
do not increase our knowledge of the real in any way. At the
most, what is additional is the name. All we know about the
real, however, is already given in 'experience' before the con-
ceptualizing begins.

Such nominalism as is here set forth is no new doctrine.
But, lately, it has taken upon itself a new form the economic

84 THE NATURE OF KNOWING

theory of the concept. This theory concerns itself more par-
ticularly with scientific thinking, one instance of discursive
reasoning. 1 It is held, explicitly or implicitly, in some form
or other for it appears in many guises by a large number
of philosophers and scientists. It is the ground for much of the
anti-intellectualism prevalent in Europe and America during
the present century. And since it is so important, we shall
devote the rest of this section to a critical examination of it.
Here, again, we do not intend to adopt the historical method,
but shall content ourselves with a statement of the theory in
its most explicit form and with an attempted criticism showing
both its strength and its weakness.

Scientific thinking, the theory holds, is in essence practical,
both in itself and in its purpose. It is the creating of a concept
for use. The discursive thinking which is scientific reasoning
gives us and can give us no new knowledge. Any further
knowledge over and above the beginnings 'given' in experience
must be expected not from scientific reasoning but from the
use of another faculty. Thus, some who adopt the present
view declare that an 'intuition' of living reality can take place
in which no reasoning occurs, and that our hope of pure
knowledge lies in this 'intuition'. Others assert that within
'reasoning' in its widest sense we have also to include the
pure knowledge of real being, as depicted, for instance, in
Reason's knowledge of the pure concepts or categories. 3 But
while a door is thus left open for the entrance of pure
knowledge, both schools reject the view that science is pure
knowledge. Science can only give us the pseudo-concept or
artificial creation made for use and for use alone. In scientific
reasoning the mind 'conceptualizes' or 'intellectualizes' the
content given by experience, solidifying the flow of the real,

1 It is not the only instance, obviously, if we mean by 'science', as
is usual to-day, the mathematical description of nature. Philosophical
thinking is another instance of discursive reasoning; for example, the
reasoning about moral, epistemological, and ontological problems.

2 The two schools we have more particularly in mind, of course,
are the Bergsonian and the Italian Idealist respectively.

DISCURSIVE REASONING 85

breaking it up into bits and pieces and disregarding the
differentia of each particular as compared with any other
within a species. It does this because a 'conceptualized' or
'intellectualized' content can more readily and more adequately
satisfy the demands made upon it by the mind (which needs
must use its knowledge) than can an 'unintellectualized'
content. The aim of scientific reasoning is not the gaining of
new knowledge. Its real purpose is to secure an easier applica-
tion and a more efficient use of the knowledge we already
possess; we can do more with 'conceptualized' knowledge than
with that which is not 'conceptualized'.

Now this theory, we readily admit, is a very important
contribution to epistemology. It stresses the modern tendency,
prevalent even amongst scientists themselves, to doubt the
absoluteness of the results of scientific inquiry. It takes this
tendency to its extreme and denies science all cognitive value.
At the same time, however, it recognizes its supreme economic
value. No instrument has ever served mankind better. But the
value of such reasoning, the theory asserts, is practical rather
than theoretical. The real knowledge in so far as there is
real knowledge has already occurred before we begin to be
scientific. 1 The scientific concepts are economic 'goods', made
to be used. They must not be conceived as providing truth.
They exist in order to serve a practical purpose. Science does
not give truth: it controls nature. If we desire knowledge of
reality we must seek for it elsewhere.

In fairness to such thinkers, however, we must here note a
further point. Their position cannot be identified with the
Pragmatic school of thinkers who teach that knowledge itself
is wholly utilitarian (or pragmatic) in character. Knowledge,
the latter would assert, is merely 'what works'. It is completely
subservient to action and is determined by our practical needs.
Whatever best suits our convenience at any moment, that, for
the time being, is knowledge. Now the theorists whom we are
at present considering do not hold this view r . Some of them

1 As one exponent of the view neatly asserts : Science is fructifcra,
but not lucifera. Its soul is utility.

86 THE NATURE OF KNOWING

would most emphatically reject it. Knowledge for them is not
merely 'what works'. It may prove to help us practically in a
particular case; in another it may not. Its utility is merely an
accidental quality of it. Their real point, however, is that
scientific reasoning is definitely and essentially utilitarian,
definitely a doing and not an instance of knowing in any sense.
The conceptualistic process is simply an economy. Its aim is
not to know but to save mental labour. The most we can say
for it is that it seeks to make greater knowledge possible, but
we are never to say that it of itself knows.

This doctrine is in certain important respects sound. It sets
forward explicitly one feature of scientific reasoning which
needed recognition and emphasis. Beyond a doubt such
reasoning is a labour-saving device. The concepts that it uses
have a practical value. They are often enough simply 'ordering
conceptions' which help the scientist as he struggles to control
nature. 1 To conceive them in themselves would not be to know
the real. Undoubtedly, again, the scientist does not concern
himself with, does not seek to know, many real details which
could, he thinks, prove of no value to him. He ignores them.
He is compelled to select. Science is certainly an economy,
and, as such, considering only this side of its nature, it cannot
be described as a knowing. So much seems true. In the next
section we shall suggest that some such doctrine as this must
be true not only of scientific reasoning, but of all mediation
and of all discursive thinking. It invariably involves an element
which is merely a doing.

Yet, though there is obviously an element of truth in this
doctrine, we nevertheless believe that, as it stands, it is
unsound. As an account of scientific reasoning, taken as a
whole, it leaves too much unsaid. It emphasizes one feature
only, and consequently fails to present a true view of the
whole. Scientific reasoning is admittedly an economizing, but,

1 The history of science gives us many examples of 'ordering
conceptions', some of which have long since been discarded. Such
are, for instance, the concepts of caloric and ether which dominated
eighteenth-century scientific reasoning.

DISCURSIVE REASONING 87

as we hope to show, it is also a knowing. And we cannot
describe it fully unless we recognize the presence of this
additional feature. For, like the sensory experience, scientific
reasoning also, we shall contend, is a cognitive experience,
even though there be that within it which is non-cognitive
in character.

By w r ay of preparation for the fuller description of discursive
reasoning which we hope to proffer in the next section, we
may here consider in greater detail the practical or economic
feature of scientific reasoning. We shall also show how impos-
sible the position becomes if we think of scientific reasoning
merely in terms of this one feature of it.

We may begin by pointing out that the economy under
consideration can be secured in, at least, two ways: firstly, by
arranging and ordering the knowledge we already possess in
such a manner as to facilitate access to it; secondly, by learning
how to deal with and how to control that of which we are
partially ignorant. The second process enables us to act upon
an object without first gaining full and adequate knowledge
of it. It is clearly the more difficult task and demands the
greater ingenuity.

The first is a feature of thought everywhere. Indeed, the
classifying and grouping, the ordering and systematizing,
which provide formal logic with much of its subject-matter,
may, from this point of view, be regarded as the effort present
in all thought to lessen the mind's labours by presenting it
with a connected system of knowledge. If there were no such
systematizing in thought, the mind would always be faced
with the completest disorder. Its knowledge would be heaped
up in a chaotic mass, so chaotic that it could prove of little
use in everyday life. Thought, as being an organizing and
ordering activity, is that which safeguards the mind from a
danger which is for ever threatening it, namely, that the
knowledge it already possesses may become so inaccessible as
to be worthless for all practical purposes. To avoid this the
mind thinks in an orderly fashion (tabulating according to
its concepts of species, genus, class, particular, universal,

88 THE NATURE OF KNOWING

individual, and so on), in order that it may be enabled to
make as full a use of such knowledge as it possibly can.

This, therefore, is the first type of economizing connected
with scientific thought. Now clearly, in such a case, thinking
is not merely a doing. It is also a knowing. With each step
forward in the work of systematizing and ordering we gain new
knowledge of what can and of what cannot be linked together
and related. We recognize, to say the least, in the present
experience some feature or features common to many past
experiences. We apprehend an identity, an apprehension which
is definitely something cognitive. Moreover, our general divi-
sions and systematizings (according to which we classify this
present particular under its universal) would be valueless from
a practical point of view if they were merely arbitrary. We
must 'divide at the joints', and this implies that we know
where the joints are. The doing in this case rrtast at one and
the same time be a knowing if it is to achieve its purpose.
And to repeat a tautology it must be a knowing of the real ;
that is, of the world within which and upon which we act.
Hence, to emphasize the very real economic value of such a
systematizing process is a necessity, for it does present us with
an orderly rather than a disorderly world, and so retains as
accessible Avhat might otherwise become inaccessible. But this
does not justify the further assertion that the process is purely
a doing. For knowing, as we have shown, is inevitably a part
of it, and, indeed, the main problems connected with it are
essentially theoretical.

But, perhaps, we are overthrowing a mere man of straw.
For those who teach the economic doctrine of the concept
might themselves readily admit that the particular kind of
doing now under consideration must be accompanied Jby a
knowing if it is to occur at all. Yet this, they might argue, is
not the doing which is par excellence scientific reasoning. The
latter is a doing of quite a different order, and is in no way a
knowing. The concept (or pseudo-concept) of science, they
might say, remains as yet wholly unconsidered by us. For it is
not a feature of, or, again, a division in, reality. It is a symbol

DISCURSIVE REASONING 89

(for something in reality) used by the mind when it reasons
scientifically. The scientist, they say, usually supposes, rightly
or wrongly, that reality consists of individual things which
affect us in the sensory experience. He also believes that the
real individual is not revealed immediately and completely in
sense-experience. For instance, he would be more ready to
believe that the table upon which I write is an entity, con-
sisting of a very great number of atoms, somehow knit together,
than that it is the solid, unbroken piece of matter about whose
existence the crude sensationalist would feel so confident.
None the less, he finds in the sensory experience an invaluable
aid from a practical point of view. For in it there recur often
with great frequency certain regularities which he notes, and
in time he is led to assume that such regularities are universal
throughout sense-experience and will invariably recur in his
own history afcd in that of other people. On this assumption
he acts, and frequently his future experiences do justify his
first assumptions, giving him a greater confidence in them.
He seeks for more recurrences of the same type, his method
being that of abstraction. He explicitly disregards all that is
unique and does not recur.

How, then, according to these theories, does he form his
concepts? They are already being formed. He now combines
together certain of these recurring features which have always
gdne together in his experience, and to this combination he
gives a name. This is his concept, an entity which, of course,
does not exist apart from his conceiving of it. Is it not obvious,
they would now ask, that when he does so conceive he knows
nothing? For if he knew, he would know the real; but the real
is the concrete individual, and not this collection of more or
less regular recurrences. For all practical purposes and the
view presupposes that the scientist's purposes are wholly
practical it is sufficient for him to conceive his concept, a
symbol of the real which helps him to handle it. He need not
know the real, and, strictly speaking, never endeavours to do so.
His concept is a general symbol for many real individuals,
and the unique features, the idiosyncrasies of the latter are

90 THE NATURE OF KNOWING

wholly left out of account. They are differentia which do not
matter for the scientist's present purpose.

Moreover, by reasoning conceptually, and not about each
individual as such, he is able to make a fuller use of the
ordering and classifying which is also part of scientific method.
For the unique, which defies classification, is disregarded by
the scientist. A botanist, for instance, when enjoying a per-
ceptual experience, which we call 'seeing a particular butter-
cup', recognizes what he sees as an instance of the species
Buttercup, a concept with which he is already w r ell acquainted.
But this again, he recollects, is a member of an equally familiar
genus, Ranunculus. Consequently, by recognizing it as an
instance of this species and genus, he indirectly learns a great
deal which would help him to deal with this individual butter-
cup, though he knows next to nothing of it itself. When the
scientist conceives Ranunculus, he is, speaking strictly, knowing
no real existence, according to these theorists. The real exist-
ence is the individual buttercup; Ranunculus is a concept I
have conceived. Yet such concept-making enables me to deal
with the real things, namely, these individual buttercups, in a
way that would be otherwise impossible in the circumstances.

But the method permits of a much more extensive use.
Thought's capacity for systematizing and ordering, its practice
of setting out in compartments and of 'pigeon-holing', may
further enable it to connect a concept with a whole order or
system of concepts having its own formulae or conceptual
abridgements of long processes of reasoning. Thus, when
seeking to determine and to define certain variations in physical
phenomena, we may observe that the changes under con-
sideration always occur according to an order which we have
met with elsewhere for example, in mathematics. Having
observed this, we may endeavour to work out our problem
mathematically, making as much use as possible of the formula;
which are the stock-in-trade of every mathematician. In this
manner we may quickly and easily gain precise results, whereas,
if we had continued to search for them by observing the
changes in the physical world alone, much greater mental

DISCURSIVE REASONING 91

labour and energy would have been necessary, while perhaps
the actual result gained would not have been so complete.
Such an application of mathematics to physics occurs more
and more frequently.

But the important point to note in this connection, our
theorists would say, is that no new knowledge has been gained.
Our doing gives no knowledge of the real, but it determines
it conceptually in such a way that we can act upon it efficiently.
The subject-matter of the scientist's thinking is the concept,
and never the real individual, if there be such. For it is only
when dealing with conceptual symbols of the real that he can
make use of the formulae belonging to a particular system of
concepts. The formulae cannot apply to the real individual as
it is, different from all else. Hence it follows that, however
successful the scientist be when using this method, he never
comes to the knowledge of the individual as such by its means.
That is not the purpose of the symbolization. Yet, without
coming to know the individual thing in itself, the symbolization
through concepts enables him to 'handle' such an individual
much more successfully than he would otherwise be able to do.
If, therefore, we expected from science knowledge of the real
individual thing, our expectations could never be fulfilled;
for from the moment we begin to be scientific we are no
longer dealing with the individual thing as it is in its full
reality.

Taking all these facts into consideration, the theorist feels
himself justified in concluding that scientific thought should
no longer be conceived as a knowing. We do not know the
real through such thinking; we merely learn how to handle it.
'Conceptualizing' is never as such concerned with knowing
reality. It is explicitly an economy. When conceptualizing we
do not know anything fresh, but create (with the help of past
knowledge) and make use of the concept. In science, such
theorists contend, we sacrifice exact knowledge in order to
satisfy our more immediate need for action and to secure
control at the earliest possible moment over the forces of
nature. It is for this reason that we bring into being and con-

92 THE NATURE OF KNOWING

ceivc an unreal, split-up, 'pigeon-holed*, but more tractable
world.

Now, how far is the position here outlined correct ? We have
already seen that the success of scientific reasoning in ordering
and classifying its content is itself sufficient proof of its more
comprehensive knowledge of the real. What of the scientific
concept? Are we, in conceiving, also knowing (even though
our primary purpose is practical)? Or is it merely a doing?
That the first alternative in this case is the sounder in spite of
the above theorists is suggested by the fact that purposive
acting in adult human behaviour, such as the scientist's,
invariably implies knowing, and that, therefore, if through
scientific reasoning increased action becomes possible, this in
itself is a sure sign of an increase in knowledge also. It is
quite true that we may and continually do act on things
without knowing everything about them, yet the surest way
to secure a more efficient control over anything is to increase
the sum-total of our knowledge about that thing. For how
could a mere * conceptualizing', that never sought to know the
unknown, help us to 'work' more satisfactorily on the unknown ?
It clearly could not. The hard-and-fast distinction between
'doing' and 'knowing' is illusory. Every doing is in part a
knowing; all human action is that of a knowing mind. The
absolute distinction thus glibly pre-supposed by adherents of
the economic theory of the concept is artificial in the extreme.
To uphold it is to do an injustice to the facts of human expe-
rience. Acting implies knowledge throughout.

At present, however, \ve need not follow out this line of
criticism. Such generalities as these arc apt to produce vague-
ness and indefiniteness, whereas, in this case at least, a clearer
and more apt criticism is at hand. For, according to the theory,
the concept is made for use. Yet once the position is fairly
understood, it is difficult to see how such a 'creation' could
ever be of any use. Instead of aiding the mind in its dealings
with the real, it itself cannot but mislead it on every possible
occasion. For, on this view, the real is unlike the 'conceptual
stuff' created by the scientists. The real is not split-up and

DISCURSIVE REASONING 93

'pigeon-holed' into distinct compartments. It is only our
failing intellects that need the 'split-up* conception of it. It is
we who solidify the ceaseless flow of real life and create the
concept by cutting out a piece of the mass thus solidified and
giving it a name. This unreal symbol of the real we preserve
and communicate for our use. But surely, on the face of it,
there is something absurd in the position. If I ever did make
use of these unreals, how could I hope thereby to deal more
effectively with the real? How can I, carrying on processes
of reflecting about a split-up world, hope to handle the real
life-flow more successfully? Would not such 'conceptualizing'
decrease rather than increase my power of acting on the real ?
Would not such a creation be valueless or even definitely
harmful? The only conceivable manner in which the 'creation*
could be of real use would be as turned back again into the
fluid mass. The concept made for use, that is to say, is useless
until we know or perceive that moment of the flowing reality,
as it actually is, for which the concept stands and of which
it is the solidification. Otherwise, it misleads the mind,
causing it to believe, for instance, that it has to deal with
solid lumps of stuff, rather than with a ceaselessly flowing
real.

But, again, supposing it were possible for the mind to
recollect, with the aid of the concept, that aspect of reality
which it solidifies, this recollection, on the view under con-
sideration, would prove equally valueless. For the ever-
changing real would have taken upon itself in the meantime
a new aspect, and we should therefore recollect what had
already become unreal. That is to say, to act on the real world
at any particular moment we must have knowledge of the
real as it is at that moment. And this is true whatever view we
adopt as to the nature of the real, whether it be a ceaseless
flow, or a block universe, or anything else we choose. If the
concept fails to provide such knowledge, it is worse than useless.
Conceptualizing must be itself a knowing ; it cannot be merely
a doing of something with knowledge gained beforehand. For
if we find it easier to act on the real as the result of a con-

94 THE NATURE OF KNOWING

ceptual process, our increased facility of action must be due
to the fact that the conceptual process has itself increased our
knowledge of that real.

As it stands, therefore, we cannot accept the economic
theory of the concept. So long as it contents itself with
describing scientific reasoning in terms of 'doing' only, its
description must remain inadequate. For the practical success
resultant upon such reasoning is in itself a sign that new
knowledge of the real is being gained in the process. We may
certainly use symbols to help us. Also we have a right to guess
at the truth when we do not know it, and even to assume
certain things to be true for practical purposes and for the
time being. We have never denied this. Certainly, again,
scientific reasoning is an 'economizing', and to think of it as
such is to gain a better understanding of its nature. But if no
more be said the description is totally inadequate, and, as such,
falls away from the truth.

All such description is misleading not because scientific
reasoning is affirmed to be a doing, but because it is affirmed
to be merely a doing. The suggestion that a process of scientific
reasoning nowhere involves as an essential feature of it a
knowing cannot stand examination. Knowing, we contend, is
part of the inmost structure of that intellectual activity termed
scientific reasoning. Furthermore, and this is the most impor-
tant point of criticism we wish to make, the denial of this
assertion would have been wholly impossible were it not that
a serious error had crept into speculations about the intellect's
work. We refer to the erroneous supposition with which we
began this section, namely, that a stock of knowledge lies
ready to hand in the mind, and that then thought or reasoning,
or, again, intellect, begins on its task of 'handling' the already
known.

This is an old fallacy. None the less, the new economic
theory of the concept, as we have before suggested, is grounded
upon it. For the position set forth above is but the modern
counterpart of the nominalistic tendencies of past generations
of philosophers. As such, we reject it. The intellectual process,

DISCURSIVE REASONING 95

if it involves a doing, is also at one and the same time a
knowing; and while emphasis on the economic value of such
a process is permissible and necessary, no true view of its
nature can be attained unless equal emphasis is laid on the
cognitive character of the process.

If we now return to the problem with which \ve started
this section, namely, the problem as to the legitimate use that
may be made of the term 'experience', we may conclude that
we must not speak of it as providing grist for the mill of
intellect, stocking the mind with a knowledge-content in order
that the intellect may work upon it, if by this it be also meant
that the intellect itself need kno\v nothing further. This view
of the relation of 'experience' and 'thought* is unsound, for,
as we have shown, it is difficult to see how in these circum-
stances what the intellect did with the knowledge given it
could prove 01 the slightest value. But if this view is false,
what is the true one? Is there a legitimate use of the term
'experience 5 ? Can it ever mean something necessarily pre-
supposed by a process of mediate thinking? These questions
open up the whole problem of the true nature of mediation,
and we can no longer evade a definite discussion of it. For no
final solution of these preliminary problems concerning the
presuppositions of mediation can be given until we solve the
central problem, namely, "What does the phrase 'mediate
knowledge' itself connote"?

3
The Nature of Mediate Knozvledge

In this section we hope to gather up the strands of the
preceding argument. Up to the present, our search for a
perfect example of knowing has led us, firstly, to deny that
seeing a colour, hearing a sound, and so on, is knowing;
secondly, to assert that the sensory experience is nevertheless
cognitive; thirdly, to reject the theory that the apprehension

96 THE NATURE OF KNOWING

of the first principles (while it obviously is knowledge) is a
knowledge of sheerly discrete simples; fourthly, to dispute
the economic theory of the concept according to which con-
ceiving is merely a 'doing'. But while we have succeeded in
establishing these positions, it is obvious that the perfect
example of knowledge for which we seek remains undis-
covered. The sensory experience is a complex whole whose
most characteristic part is non-cognitive. The knowledge of
first principles is embedded in discursive reasoning, and is
not in itself a whole concrete experience. Scientific reasoning,
the one instance of discursive reasoning as yet considered by
us, has revealed itself to be in part a 'doing', even though, as
we contend, the 'doing' could not occur without a concomitant
knowing. Nowhere have we discovered an experience which is
through and through pure knowing.

Nevertheless, each of these experiences is cognitive, even
though none of them is wholly knowing, and in examining
them we have already gathered certain facts about the cognitive
experience in general. Thus we have agreed that the knowledge
of principles is the immediate apprehension, present in dis-
cursive reasoning, of certain universal characteristics possessed
by the real. An active mind apprehends or grasps the real
straightway. The analogy implied is, of course, that of physical
grasping, as when I grasp a physical object. And though
it does not follow that this analogy most appropriately
expresses the real character of knowing in every other
connection, it seems apt when applied to knowledge of the
principles.

But now the specific problem of the present section arises.
For in outward appearance, at least, most of what we ordinarily
suppose to be knowledge fails to conform with the above
description. It is not direct, but indirect. Are there, then, two
kinds of knowing, the one direct, an apprehending, and the
other indirect, something different in kind? An affirmative
answer seems to be inevitable, but we hesitate. For have we
not seen that our best instance of immediate knowledge is
itself somehow embedded in a process of discursive reasoning

DISCURSIVE REASONING 97

whereby we hope to gain knowledge indirectly or mediately? 1
Can mediate knowledge, then, be immediate, even in part?
Now a full and complete account of mediate knowledge would,
no doubt, include a thorough-going logic of all inferential
processes, both deductive and inductive. But such a logic
would be beyond the scope of the present work. Instead, we
shall confine ourselves to the consideration of two problems,
whose solution ought to throw further light upon the nature
of knowing. Firstly, what exactly do we mean when we talk
of knowing mediately} And, secondly, how does this knowing
compare with that which is immediate ? In answering the first
question, of course, we shall have already, implicitly at least,
answered the second.

Our present problem, then, is to determine the nature of
mediate knowledge and to distinguish it from immediate
knowledge. And in order to make our position completely
clear, we propose to put forward without further delay the
main thesis of this section. We wish to argue that a hard-and-
fast distinction between mediate and immediate knowledge, in
which these are taken to be two distinct types of knowledge,
cannot stand examination. Actually, all knowledge, wherever
and whenever it occurs, is immediate in character; the facts,
when considered, justify no other interpretation of them.
Nevertheless, this statement is in no way incompatible with a
continued use on our part of the term 'mediate knowledge',
and we do not intend to dispense with the latter. For it
expresses something which definitely needs to be expressed.
The unfortunate fact is that when we analyse instances of
* mediate knowledge' we find that the phrase carries with it
two significations, and that neither of these significations can
be disregarded. In the first place, we mean by it an immediate
knowledge of an implication considered not so much in itself
but with special reference to the light it throws upon the
subject of the conclusion implied. As an illustration, we have

1 For the present we use 'indirect' and 'mediate' as synonymous
terms. What we mean to express by them will be further illustrated
by examples in the pages that immediately follow.

G

g8 THE NATURE OF KNOWING

our mediate (or indirect) knowledge about S in a syllogism
that it is P. We shall proceed to consider this instance in the
next paragraph. In the second place, 'mediate knowledge'
frequently connotes a complete thought-process culminating
in a knowing (which is itself immediate) or in an opining. 1
The knowing or opining becomes possible as the result of the
process. To add to our difficulties, we also find that the phrase
is habitually used to connote both significations at one and
the same time, that frequently the culminating knowing or
opining is itself of an implication.

We must consider the matter in greater detail. A very good
illustration of mediation is the perfect syllogism: M is P.
This is the major term, in which we assert a universal truth
or general principle. S is M our minor term, asserting a
particular truth. There follows the conclusion , S is P. We do
not know directly that S is P; we know it indirectly. We see
that the premisses imply it. Now, how can w r e best describe
the mediation present in this syllogism? It would obviously
not be an adequate description of it to say that it is a sequence
of three true judgements ; nor even a sequence of three judge-
ments such that the final could not occur had not the first
judgements preceded it. For though this latter might pass as
a very superficial definition of syllogism, it does not set forth
satisfactorily the inmost nature of the experience. It does^ not
state positively that the final judgement is implied in the
premisses, and it does not show the nature of its necessary
derivation. The perfect syllogism, we should rather say,
expresses verbally one whole movement of thought in which
the premisses are seen necessarily to imply the conclusion. This
statement, of course, is no definition of the syllogism as such,
since it could be applied to any piece of sound inference, and
not all inference is syllogistic. It is none the less valuable for
our purpose, since it draws attention to the mediation present
and seeks to describe it. The description is still superficial, as
we hope to show. But it does bring out the first important

1 In the latter case we have, strictly speaking, not 'mediate know-
ledge', but an opinion arrived at mediately.

DISCURSIVE REASONING 99

consideration. Knowledge can occur mediately because we can
see necessary implications.

By this we mean that, granted a general rule of the type
M is P, which must hold of all instances of M, and granted
also that S is one instance of M, then this system of relations
already implies that S is P. In other words, granted that some-
thing can be predicated as holding of a general term, and
also that S can be subsumed under this general term, then
the predicate must equally apply to S. Now, if we analyse the
above argument, it will be evident that it is based upon our
knowledge of two relations, namely, the subject-attribute and
the member-class relations. Our knowledge of the former
alone makes possible the predication of M by P and S by M ;
while that of the latter enables us to relate S and M. Our
argurpent could not occur without a definite knowledge of
these relations. Furthermore, we know that if M is P and
S is M, then S is P. The sum of our knowledge, then, in the
present case is that of the two relations, 1 together with the
further knowledge that these relations under the conditions
found in the demonstrative syllogism entail or imply a third
relation. This third relation holds between S and P. No
further knowledge is necessary to complete the syllogism.

But we have said that the perfect syllogism is an excellent
example of mediate knowledge. In what manner, then, does
the* knowledge present in it differ from the immediate kind ?
This is a difficult question to answer. For, firstly, knowledge
of the above relations, subject-attribute, member-class, seems
to be definitely immediate. It is a direct apprehension of
objective relations. (Indeed, that these relations cannot them-
selves be known syllogistically is obvious once we remember
that knowledge of them is presupposed by all syllogism, and,
in this respect, they are on the same footing as the co-called
Laws of Thought. Like the latter, they are characteristics of
reality that we must first (speaking logically) know if we are

1 In order to avoid confusion, it is perhaps best to add that by
these two relations we do not mean the two premisses.

ioo THE NATURE OF KNOWING

to syllogize about reality. 1 ) Nor, secondly, is the knowledge of
the implication mediate. For we immediately see that if M is P,
and S is M, then S must be P. But if this is all the knowledge
that occurs in the syllogism, why call it mediate knowledge?
The position seem?, to be that whenever we observe ourselves
knowing, we find that our actual knowledge is immediate in
character, whereas it is also clear that the syllogism as a whole
is an instance of mediate knowledge. How can we overcome
this seeming contradiction ?

It cannot, if we are honest with ourselves, be overcome by
supposing that knowledge of the implication is in some way
not immediate. As far as we can see, it is so patently imme-
diate that to suppose otherwise is really impossible. Conse-
quently, we cannot agree that the actual knowing in the
syllogism is itself a passage or a development from premisses
to conclusion, and that in this sense only is it to be termed
'mediate'. For this latter position seems to rest upon the
erroneous assumption that the implication, the relation between
premisses and conclusion, is itself the knowledge. Whereas it
seems obvious that the actual knowing is not the implication
as such, but rather our 'seeing' or our direct knowledge of this
implication. Of course, it is the fact of implication which
makes syllogism (and all inference) possible. We could not
know the implication unless the premisses actually do imply
the conclusion. But the implication, none the less, belongs to
the objective rather than to the subjective side, to the known
rather than to the knowing. In other words, however much
the premisses imply the conclusion, no positive knowledge is
gained until a mind, possessing the capacity to know, actually
does know the implication. This is a fundamental fact; obvious
perhaps, but yet dangerous to ignore, and fully worthy of our
emphasis. For it makes it impossible for us to admit that
syllogism and for that matter inference in general can be

1 Whether knowledge of them is presupposed in all possible types
of mediation or inference is, however, another question. Can we
say, for instance, that the inference which is arithmetical calculation
presupposes them ?

DISCURSIVE REASONING 101

rightly described as "the ideal self -development of an object". 1
Mediate knowledge, we feel, is falsely conceived if it be
conceived as a self-development. We readily admit that the
conclusion of the syllogism is completely dependent on,
wholly implied in, the premisses, but the knowledge which
is present in the syllogism is o/this dependence or implication.
It is in no way identical with the implication itself. Hence,
we use loose and dangerous language if we say that the know-
ledge present is the passage from premisses to conclusion, or
is the development of the premisses into the conclusion.
Actually, the knowledge which is the core of the syllogism
is the immediate apprehension that the premisses imply the
conclusion."

We must, then, face the major difficulty. Viewed cognitivcly,
the best possible instance of mediate knowledge, namely, the
perfect syllogism, is essentially an act of immediate knowledge,
however much it differ in certain respects from other instances
of knowing. Syllogism most certainly cannot be regarded as
the verbal expression of a type of knowledge distinct in kind
from immediate knowledge. And what is here true of syllogism
is surely true of all inference. What we describe as a process
of coming to know X as the result of knowing Y, where Y is
a premiss or a system of premisses, is really a seeing that Y

' The phrase is Mr. Bradley 's. Cf. Logic, 2nd edit., p. 597.

- In passing, we may add that the occurrence of syllogisms which
are perfectly valid in spite of the fact that they start from false
premisses in itself shows how the apprehension of the implication
'$ the core of the syllogism. What we know categorically in syllogism
r is such is the implication. Given the premisses, then we see imme-
diately that the conclusion must follow. Moreover, because the
knowledge of the implication is the essential element, we can quite
legitimately syllogize even when we are not certain that P holds for
ill M, or, again, that S is actually one instance of M. We can argue
syllogistically to conclusions that are probable only, proceeding from
premisses only probably true. But if our thinking is to be valid, we
nust feel convinced beyond the possibility of doubt that the premisses
io imply the conclusion. Here is the real knowing which occurs at
the heart of the syllogism, and this, we argue, is immediate in
;haractcr.

102 THE NATURE OF KNOWING

implies X. What we call knowing X indirectly is really knowing
directly that X is implied by the premiss or premisses. We
know that S is P indirectly; and this means that we directly
know it to be implied in the premisses M is P and S is M.
That is to say, we know the conclusion hypothetically, //
so-and-so, then so-and-so; but from the standpoint of the
actual knowing this is merely to say that we know the impli-
cation of the premisses categorically. No additional knowledge
is involved. And the categorical knowledge in this case is
direct apprehension. Hence, if what we say is sound, it follows
that the difference between immediate and mediate knowledge
cannot lie in the character of the actual knowing as such in
both cases. The only observable difference at present is a
difference caused by looking at one and the same thing from
different points of view. We apprehend an implication directly,
but by one and the same act learn something (indirectly) about
the subject of the conclusion. Thus, though the actual know-
ledge is immediate, we may mean by * mediate knowledge' th
indirect knowledge we have about the subject of a conclusion,
which conclusion we immediately see to be implied in certain
premisses. 1

But there is another sense in which we can talk of mediate
knowledge, and the syllogism again illustrates this further
sense. It is a sense of the term made possible by the fact that
we frequently denote by the word 'knowledge' not only the
act of knowing itself, but also a whole thought process within
which such a knowing occurs. Consequently, while we continue
to affirm that the actual knowing in the case of the syllogism

1 Have we not here also the key to the understanding of the dis-
tinction between the so-called 'knowledge by acquaintance* and
'knowledge about'? Our suggestion is that all knowledge is actually
'knowledge by acquaintance' if by this be meant what we have
termed immediate knowledge or direct apprehension. But in the case
of our 'acquaintance* with, or immediate knowledge of, an implica-
tion we come to know, by this very act, something about the subject
of the conclusion implied. 'Knowledge by acquaintance* and 'know-
ledge about* arc one and the same thing looked at from different
points of view.

DISCURSIVE REASONING 103

is no process from premiss to conclusion, but is the immediate
apprehension of the implication, we can still admit that the
knowing may result upon a process, and that the whole expe-
rience might be conceived as 'mediate* in so far as the final
knowledge comes about with the help of or through the whole
process.

But, if we use the term in this sense, we must bear two
important facts in mind. Firstly, the actual knowing, the
culmination of the process, is still immediate. Secondly, the
mediation which occurs is not necessarily inferential. The
first point we have already considered, but the second must
now engage our attention. If we think of mediation as the
process leading to knowledge, then inference is only one
instance of such mediation and is not co-extensive with it in
meaning. For we term 'mediate' here any process which helps
us to know. 1 nus, for instance, the actual seeing of figures
in geometry is a help in the gaining of knowledge. Yet we do
not infer our knowledge from what we see. Again, the asking
of certain questions and the clear formulation of problems,
the gaining of new experiences, the recalling of truths (already
learnt) at a certain point in the argument, the removal of
prejudices these may all help to make knowing possible.
For, on the one hand, they may clear hindrances out of the
mind's path in knowing; on the other, they may so enrich
the* mind as to enable it to know where it could not know
before. In so far as they fulfil these functions, they are all
instances of mediation in the broad sense now under con-
sideration. Inference, however, is one particular instance of
such mediation. Its distinguishing mark lies in the fact that it
involves the apprehension of a necessary implication. Thus
the syllogism is mediate knowledge in a double sense. It is
mediate knowledge in the first sense in so far as it is indirect
knowledge about the subject of a conclusion directly seen to
be implied in certain premisses. But it is also mediate know-
ledge in the second sense. For in it we find premisses set out
by thought in such a manner that the mind can immediately
perceive their implication, so that the relating of the premisses

104 THE NATURE OF KNOWING

in this manner enables the mind to come to know an implica-
tion, and, by one and the same knowing act, to know also
something about the subject of the conclusion.

But in whichever sense we use the word, in no case is
'mediate knowledge' a distinct kind or type of knowledge
standing over against the immediate kind. Its core, we repeat,
is always immediate knowledge; in the case of inference, the
immediate knowledge of an implication. It may be objected
that such a view gives no room for certainty in the conclusion
gained mediately. But this criticism could not be justified.
What we do assert is that in syllogistic inference to revert
to the instance of the syllogism we know directly the implica-
tion alone, if M is P and S is M, then S is P. We do not know
that S is P in itself, but only as implied in the premisses.
We know it hypothetically, if ... then . . . We ask too much
from the syllogism, as such, when we ask from it a categorical
statement in its conclusion, for instance, S is P. All it can
tell us with complete certainty is that S is P is necessarily
implied in the premisses. If we do gain a knowledge that
S is P which is certain in itself, categorically, then we do not
gain it merely as the result of the syllogism. We admit that
the knowledge of the conclusion in cases where the premisses
are known to be true would be beyond doubt, but the con-
clusion is not certain in itself; it is still certain hypothetically.
It is certain because we know, firstly, that the premisses are
true; secondly, that the premisses imply precisely this con-
clusion and no other. We also admit that as the result of
syllogizing we may become so familiar and so well acquainted
with the character of S that by a direct act of knowing we
'see' beyond all doubt that it must be P. That S is P would
then become as completely self-evident to us as is the fact
that the premisses imply the conclusion. But we should know
that S is P in such a case not syllogistically but by a further
act of direct knowing, for categorical knowledge of S that it
is P cannot be given syllogistically as such. These, it seems to
us, are the facts of the case.

We ought to point out, however, before turning away from

DISCURSIVE REASONING 105

consideration of the syllogism, that usually we ask no more
from it than probability. Most often, when we actually use
syllogisms in reasoning, our major premiss is a memorized
general rule of whose truth or falsehood we are not directly
aware. Such, for instance, is the rule, all organisms are mortal,
when we have not directly apprehended with apodeictic cer-
tainty that mortality pertains essentially and so necessarily to
organism as such. We have merely accepted the rule as some-
thing generally assumed, and contradicted by nothing in our
own experience, or, again, as 'proved' by us inductively.
Now, in the syllogism, Man is an organism, an organism is
mortal, therefore man is mortal; we make our appeal to this
general rule as a major premiss. We do not directly see that
mortality appertains to organism (for, in such a case, we
should also directly see it to pertain to organism in this one
instance of it, namely, man), but we have earlier established to
our own satisfaction a general rule, and w r e now recall it and
use it as our major premiss. We know with certainty that just
these premisses imply this conclusion. And so, if our experience
leads us to think these premisses on the whole sound, we think
it safe to affirm the conclusion. Thus, though the syllogism
in this, its more usual form, gives no theoretic certainty, none
the less it gives probability. It is one method, of many used
by mind, for bringing past experience to bear on present
problems. When we fail to gain certain knowledge of X
directly, we may yet arrive at probability if we can show that
X belongs to a group each member of which, we have been
led to believe in the past, is conditioned by a certain rule. In
such a case, it is well to note, the probability is grounded upon
a prior certainty, namely, the certainty that the premisses do
imply the conclusion.

Deductive inference, however (of which syllogism is one
instance 1 ), is not the only type of inference, and we must

1 We believe that our argument in the foregoing pages would hold
for all instances of deductive inference, but to attempt a detailed
proof in each instance would be too vast an undertaking for the
present work.

io6 THE NATURE OF KNOWING

now turn to consider the other main type. By inductive infer-
ence we claim to know general laws not directly but mediately,
proceeding from the careful observation of particulars. (There
exists, however, one method of coming to know general laws
which cannot be described as inductive, though it is tradi-
tionally termed 'induction', and in this case we must not be
misled by traditional terminology. The so-called Induction by
Complete Enumeration, or Perfect Induction, is really non-
inferential in character, that is to say, the final knowledge is
gained directly. For if I actually see that so-and-so holds
individually of each and every member of a class, then no
inference obviously is involved in my asserting the universal
application of the so-and-so throughout the class.) Where,
then, shall we find inductive inference and inductive reasoning
proper? Now, in spite of much disagreement amongst logicians
as to the exact character of induction, there is general agree-
ment as to the existence of a definite type of reasoning which
is essentially inductive, and we propose to examine this
reasoning briefly. The whole inductive process presents a two-
fold character. The reasoning in it is carried on about certain
alternatives already suggested to the mind by its observations.
The first task in induction, therefore, is to set forth the alter-
natives. These are hypotheses, possible general laws to account
for the facts of experience. 1 For instance, we note that fre-
quently two facts come in conjunction, and this suggests to*
us their constant conjunction as a general empirical law. Or,
again, we find one fact always preceded by another in such a
way as to suggest that the first is the immediate cause of the
second. In sound induction, of course, this work of making
hypotheses is no mere guess-work. The hypotheses need to be
well-founded. Incidentally, it is important to observe in this
respect that in making the hypotheses we are guided not only
by our present experience, but by much past experience
relevant to the matter under consideration, as also by funda-
mental laws or principles of thinking (both the particular

1 We shall consider later in this section the meaning of the term
'a fact of experience*.

DISCURSIVE REASONING 107

axioms of our particular science and the still more fundamental
and pervasive Laws of Thought).

Now observation of the facts usually suggests more than one
possible general law, and we have to determine which law,
amongst all the possible ones, actually holds in this case.
This further task is carried out by a reasoning which is essen-
tially inductive. It consists in eliminating hypothesis after
hypothesis by the discovery of negative instances. Thus, if
we make the hypothesis that X is the cause of Y, and discover
an instance in which Y occurs in the absence of X, or fails to
occur in the presence of X, then we can be certain that this
one instance is sufficient to overthrow the hypothesis. In this
manner hypothesis after hypothesis can be shown to be
unsatisfactory, until, finally, one alone remain^ which still
satisfies the facts. And since the facts must obey some law or
other, we now conclude that the one remaining hypothesis is
sound and states the general law. In other words, if the facts
disprove every alternative except one, that alternative must
be the right one, and we are justified in placing our faith in it.
Such seems to be the core of the reasoning which is inductive.
It has rightly been pointed out also that, even though we fail
to reduce the number of hypotheses to one, something has
been gained if we reduce them at all. Though still left with
two alternatives, or even three, we are in a sounder position
than when confronted with five or six. And if our partial
reduction is carried out by the method of elimination, the
argument remains definitely inductive. 1

Here, therefore, is a further instance of mediate knowledge,
and we have to ask of it the same question as was asked of the
previous instance deductive inference. Is mediate knowledge
here something completely distinct from immediate? Now in

1 We have not considered here the methods of analogy and incom-
plete enumeration which are usually classed with induction. In the
former case, at least, it is exceedingly doubtful whether the argument
can ever be based on any eliminative process as can the main type
of inductive argument. These methods, however, have no great
epistemological importance, and we need not consider them here.

168 THE NATURE OF KNOWING

this case, again, we believe that a negative answer is the only
possible one once we consider the position fairly. For if we
consider inductive reasoning carefully enough, we must come
to see that its essence or core is again the immediate appre-
hension of an implication as something which is true in itself.
We have satisfied ourselves that of certain laws suggested by
the facts of experience, X alone holds throughout experience.
(All other suggested laws have been contradicted by expe-
rience.) Now we know that some law or other docs hold of
experience. It is then obvious that X is the law which is valid
for experience in this case. We see the implication directly.
Our conclusion that X is the law is, ho\vever, gained infer-
entially. The law, in so far as it is established inductively, is
not self-evident. But the above implication is self-evident, and
it is the implication that we directly perceive Of. apprehend.
In knowing the implication \\ith certainty, however, we gain
the assurance that X is the law which holds if our premisses
are valid. In this sense our knowledge of the law is mediate.
It is also mediate, we should add, in the second and broader
sense. The whole process which is induction is, in part, a
preparation for further knowledge, and the knowledge which
finally occurs presupposes this preparation. The direct know-
ledge of the implication, which, viewed from another angle, is
indirect knowledge as to the validity of the law, can only occur,
for instance, as a consequence of the right use of much we
already know or opine. In reasoning inductively, for instance,
we take it for granted that the world is uniformly ordered and
not chaotic in its character, that it is intelligible; and we make
use of this information. (To deny this, of course, would be to
take away the very foundation of inductive reasoning. For, as
we have shown, it is an argument carried out on the under-
standing that some law docs hold for these facts under obser-
vation, even if we cannot now exactly determine what law,
and that if we succeed in setting forth all the possible alter-
natives, then one of them must be the law which really docs
hold.) Moreover, we have to search for hypotheses and elimi-
nate those which are unsound, and this in turn is a task which

DISCURSIVE REASONING 109

presupposes the right use of much further information if these
hypotheses are to be anything more than wild guesses. We shall
presently return to the question of the part played by expe-
rience in suggesting such alternatives. But the final conclusion,
it is obvious, is reached through the use we make in a process
of discursive reasoning of much prior knowledge, and is
mediate, therefore, in this second sense of mediation.

When we turn to consider the measure of certainty possessed
by the conclusion of such an inductive process, we realize
that induction, though it may give us practical assurance, can
hardly provide theoretical certainty. We can never be wholly
certain of our premisses. If our imagination were keener and
more acute, experience might suggest to us still further alter-
natives than those considered by us, and if we searched far
enqugh we !jnght find a negative instance that would destroy
even this hypothesis which we now accept as the true law.
To derive a principle or law by inductive reasoning is most
emphatically not to know it with certainty. 1 And yet, we may
note in passing, general laws established in this manner con-
stitute the large proportion of major premisses in syllogistic
reasoning. But inductive reasoning, of itself, docs not culminate
in the certain knowledge of a general law. For the most part,
it enables us to make a judgement whose truth is probable.
None the less, it may conceivably suggest a connection whose
necessity w r e may come to see directly. It may fulfil the function
of a propaedeutic to a future knowledge that shall be com-
pletely certain even though it fail itself to give that knowledge.
In such a case, we should know only in so far as we directly
apprehend the necessary connection, but in suggesting this

1 It is, of course, theoretically concehable that we should have,
first, discovered every possible alternative, secondly, known that we
had discovered all the alternatives, and, thirdly, shown all except
one to be false. In this case we should have attained certainty, if the
premisses were sound. Our assurance would still be of the hypothetical
kind which pertains to whatever is known as implied. But actually
inductive reasoning never gives certainty, since we can never be
certain that every alternative has been considered nor that every
negative instance has been brought forward.

no THE NATURE OF KNOWING

particular law and emphasizing its probable truth the induction
would have helped materially. And even when induction fails
to lead to such an act of direct apprehension, as is usually the
case, its value still remains great. For, guided by 'experience'
and past knowledge, it frequently gives probability of a very
high order, and such probability is in itself extremely valuable;
while, on the practical side, we continually act as if induction
gave us not probability but certainty, and find the action, for
the most part, successful.

If we now reconsider the whole argument up to the present
point, we see that the phrase 'mediate knowledge' is used to
convey two distinguishable meanings. In the first place, it
means indirect knowing that is to say, coming to know some-
thing indirectly about the subject of the conclusion in seeing
an implication directly. The latter is all the knowledge actually
present. Simply as the result of our thinking, we do not know
more of the conclusion than that we see it to be directly
implied by the premisses. (Later, of course, by a further act
of knowing, we may come to see directly the relation set forth
in the conclusion.) There exists no indirect knowledge as
something distinct and separate from direct knowledge. We
cannot find any evidence of two distinct types of knowledge,
the one direct and the other indirect, in this sense. Mediate
knowledge qua indirect is simply the direct knowledge of an
implication looked at from the point of view of the information
given about the implied. But, in the second place, the phrase
'mediate knowledge' also denotes a whole thought-process,
together with the knowing that culminates it. Certain thought-
processes possess just this characteristic that they culminate
either in a direct apprehension or in an opining, and without
the thought-process the apprehending or opining could not
occur. (It is worth remembering, also, that a thought-process
of this kind may frequently lead neither to knowledge nor to
opinion, but to a state of suspended judgement.) Now the
whole discursive-process, including its culmination, is an
instance of mediate knowledge. If we think merely of the
direct apprehension, however, in which the process may

DISCURSIVE REASONING in

culminate, we had best repeat that it itself is not a process.
It is direct insight that comes like a flash after thinking about
a matter, and so brings the thought-process to an end. Even
in mediate knowledge it itself is immediate.

We cannot, however, leave the matter of mediation thus,
for though our main interest lies with the actual knowing and
we have already shown what form it takes in inference yet
a further question which arises in this connection must be
faced. How can a process of discursive reasoning ever help us
to know? To this question we should answer: Discursive
reasoning facilitates the task of the knowing mind by pre-
senting it, as the result of its thinking, with a world which is
more systematic, more coherent, and within which a greater
number of relations are already known. It can do this, we add,
bepause it ^ses past knowledge and past opinions, and con-
ceives its world accordingly. Now progress in knowledge is
easier when dealing with the more ordered conceptual world
than when dealing, for instance, with the sensory world. New
relations are more easily apprehended. And often all advance
becomes impossible until some further systematization is made.

But, it will immediately be objected, the conceptual world
is arbitrary, artificial, and unreal, the outcome of a falsifying
abstraction. Therefore, conceiving can help us to know only
in so far as the object we desire to know is unreal. It plainly
cannot help us to know the real, for it hides that real from us.
We, of course, do not agree. Conceiving helps to make greater
knowledge of the real world the only knowledge we recog-
nize possible. To understand how this comes about, we must
first recall the argument of the previous chapter. The assump-
tion underlying the above objection is that the real world is
the sensory world, that the particulars of sense are the real
things which exist, and that when we abstract in conceiving
we are turning our backs upon the real world. But this assump-
tion is totally false, and our theories as to the nature of con-
ceiving, once the above assumption is made, cannot but be
unsatisfactory in the extreme. For, if we once suppose that
the only outlet to the real world is sensation, then clearly

ii2 THE NATURE OF KNOWING

knowledge of that real for us must mean knowledge of this
particular colour, this particular sound and taste. Anything
else is phantasy. How far a perceived ' object' (not sensed as
such) could be real would remain a difficult problem. But
certainly a concept or general idea would only be possible as
an arbitrary creation which itself could never be used to gain
further knowledge. At best, it could only act as an economy,
a 'shorthand-note'. For whatever we chose to do with our
intellectual faculties, the real would always be known in
sensation; and we could never 'abstract' from the sensory
content without definitely moving away from the real; the
image would have to be conceived as a weak or 'decayed'
sensation; while the concept would be one stage further
removed from reality. But when once we understand the true
position, namely, that the content of sense is not itself the
'stuff' of the physically real, we can then, at least, claim the
right to abstract from its recurring manifold and to disregard
certain details without necessarily suffering any loss in know-
ledge of the objective reality. Once, then, we free our minds
from false assumptions, there is nothing impossible in the
suggestion that conceiving may help us to know the real.
Some advance must be made on the sensible world. If we
rest content with a world of sensible objects, which we too
readily assume to be the real world, we shall gain no know-
ledge. The first lesson we have to learn is that things do not
possess just these sensory qualities which we ascribe to them
in the sensory experience. The sensible world, which we
naively claim to be physically real, is the outcome of our
fundamentally false assumption that sensing is knowing with
the consequent ascription of sensory qualities to that real.
And we must first realize the possible falsity of this ascription
before we can hope to understand how reasoning and reflection
can enable us to gain, at least, probable truth about the world
around us, and how they may lead to certainty. In spite of its
conceptual character, the world conceived by the scientist may
be more real (as, indeed, we usually believe) than is the sensory
world of every-day.

DISCURSIVE REASONING 113

The above objection, therefore, cannot stand. That is to
say, it cannot stand if we interpret it to mean that conceiving
is no help in knowing the real because it (conceiving) itself is
a turning away from the real revealed in sensory experience.
None the less, we must admit that it has a certain force from
another point of view. The objection may only mean that the
conceptual world is in itself unreal (whatever the sensible
world be), and so conceiving it cannot possibly help us to
know the real. Here the objector would, at first, appear to
stand on firmer ground. For it is hardly possible to deny that
the conceived world about which we think is in part the
creation of our own minds. In its totality it is not completely
identical with the real world which we are coming to know
through its aid. For instance, the man of science would be
ready to admit that the world he presents to us was not wholly
discovered by him. Quite explicitly, some of it is the fruit of
his own imagination. But this admission cannot rightly be
taken to prove the thesis that conceiving is of no help in
knowing. For though conceiving is in part a creating, and
though there is something arbitrary and artificial in it, we may
yet claim for it that it makes more frequent advance in know-
ledge possible. This advance becomes possible because in
conceiving we abstract. We can (and do) select those elements
in experience reflection upon which is most likely to lead to
farther knowledge. Now the elements we do select are in-
variably recurrent features. Through experience we become
aware of certain common features features which repeat
themselves on more than one occasion, and we frame our
concepts according to these features. The existence of these
common features is obviously important, since in reasoning
we seek to link together and to systematize in the hope of
coming to apprehend new links. But the completely unique
would resist all our efforts at relating it with anything else.
Therefore, we disregard what is unique in our sensory expe-
rience. We disregard the particular time of the sensation which
can never be repeated. We disregard the unique set of circum-
stances in which the sensory experience occurred, but we

ii4 THE NATURE OF KNOWING

fasten upon a feature, which this particular has in common
with many others, and so link this particular up with other
particulars, conceiving the type.

But how does experience (by which we mean sensory expe-
rience) 'give' us the recurrent feature? It would not give it if
the sensory experience were merely a seeing of a colour,
hearing of a sound, and so on. But the seeing of colours, as
we have throughout insisted, is itself an occasion for know-
ledge. When we see a patch of red, we spontaneously know
existence. In seeing the colours, which are now before me, I
know the difference between one and many. Moreover, I can
pick out two that are like, and relate them as being like in
colour. Now I do not see this likeness. I see the colours, but
know the likeness. And it is our power of apprehending like-
nesses in colour which alone enables us to conceive redness,
blueness, and so on. It is not that all reds are identical in
shade. It is questionable whether any two instances of red
are even completely identical. Indeed, we conceive redness by
arbitrarily taking one particular shade of red, not too dark
and not too light, and making it stand for all others. (This is
one instance of the 'doing' present in conceiving.) Yet we are
only able to do this because we have apprehended a likeness
and a similarity between different reds. The concept could
not be made without the prior knowledge. It is based upon it.
Sometimes, again, we come to know in experience not tne
approximate recurrence of a colour, but the more exact recur-
rence of a relation between colours for instance, a sequence.
To take a simple case: In carrying out an experiment I see
two colours, blue and red. I notice in carrying out many such
experiments that the red always follows on the blue (this is a
'fact* of my experience), and I conceive the event as a sequence
of blue-red. 1 Now what happens here, speaking precisely, is
this. I see the colours. I know the sequence, red following on

1 To avoid, for the present, difficulties about memory, we may
think of many experiments being carried out simultaneously before
our eyes, so that we can see and compare them all at one and the
same time.

DISCURSIVE REASONING 115

blue. Finally, I know that this sequence is repeated on each
occasion. The last two items are definitely instances of know-
ledge. That is to say, the feature apprehended as recurring
was itself, in the first instance, known rather than seen, a
known relation between colours. And it is not impossible by
further abstraction to conceive the relation alone here
sequence and to reflect upon it. But whether our concept
be of this extremely abstract kind or not, the point we wish
to make is that in forming it we are guided by knowledge
already gained in sensory experience. Even though conceiving
is, in part, creation, we neither create ex nihilo nor capriciously.
The conceived world embodies in itself knowledge already
gained, and the new ordering in conception proceeds according
to what we know. Thus the more systematic character of the
world conceived, for instance, by the botanist more syste-
matic, that is to say, as compared with our own everyday
non-scientific view of the plant world is ultimately based on
his capacity to know real likenesses and is the consequence of
his abstraction and selection according to this knowledge. All
conceiving is ultimately based upon our knowledge of real
features of that real world which we first know in sense-
experience.

We now begin to understand the role of experience in
knowledge, and how discursive reasoning, through which this
first minimum of knowledge becomes enlarged, most certainly
presupposes it. Discursive reasoning is conceptual in character;
but conceiving can only occur on the basis of what we already
kno,w in experience. Conceiving is the outcome of abstracting
certain real features known in experience in order to consider
them alone and in order to pay greater attention to them.
These features are also known to be common or universal
throughout a particular group. Discursive reasoning is a further
systematizing of these concepts according to what we know
or opine. As such, it facilitates the task of the knowing mind,
because the more closely knit, the more inter-related and
connected, the world we think about, the more likely are we
to see the necessity of new relations, or to opine that such-

n6 THE NATURE OF KNOWING

and-such a relation probably holds of the real. And this con-
ceived world, just as it is more systematic, is also more real
than the sensory world, because in conceiving it we take up
the knowledge already contained in the experiencing of that
sensory world and increase it in our apprehension of new
relations and of greater system. Thus the world conceived by
the scientist, though not wholly real, is yet both more syste-
matic in its character and better grounded than is the sensory
world of the unreflective person. In other words, when
reasoning conceptually, the conditions are more favourable
for the occurrence of knowledge. Here lies the real utility of
discursive reasoning from a cognitive point of view. It increases
the possibility of further discovery. As such it has a side which
is a doing. We readily recognize this. In inference, for instance,
we must seek for, and bring together, the righ* premisses. 1
Also the terms of the premisses are conceptual, and in thinking
any concept we must bring many particulars under one head
through knowing the recurrence of some feature. This bringing
together, as such, is in both cases a doing. Yet it is a doing
which proceeds strictly according to, and is justified by, a
knowing (or an opining) of something in common between
the particulars, and of something which can link the two
premisses. And when this doing, so intimately bound up with
knowing, has occurred, we may find ourselves in a position
to gain greater knowledge, as when we are enabled to see
directly that the premisses of an argument imply a further
conclusion.

And, perhaps, the true meaning of the second type of
'mediation' considered by us in this section may now be
more easily understood. It is mediation in this sense, that
where we at first fail to know a truth, we may, as the result
of discursive reasoning (or thinking), arrive at a position in
which we can know it directly. Reasoning or thinking, from
this point of view, in its systematizing and ordering, its

1 This frequently involves the adoption of 'a trial and error*
method of procedure.

DISCURSIVE REASONING 117

classifying and relating, is simply our effort to work ourselves
into a position in which further knowledge can be gained. It
is our effort so to enrich the mind or the intellect by a wise
use of experience, of prior knowledge and opinion, that the
mind is enabled to apprehend new truths. As such it is the
process which makes the further knowledge possible. And
mediate knowledge is this whole process completed in its
culminating act the act of direct knowing. In other words,
mediate knowing in this sense of it is in essence the appre-
hension of a further truth through the use we succeed in
making of past knowledge and of the whole experiences in
which such knowledge occurred. Mediation is the process of
using the knowledge and opinions we already possess in a
certain definite manner, doing something with them, so as to
gain still greater knowledge of the real, which new knowledge
in its turn may be used again in the same manner. This is the
true dialectic of knowledge. The final knowing, however (if
knowing occurs), is not a doing, nor a process; in each case it
is the immediate apprehension made possible by the process.

With these statements in mind, we are better able to deter-
mine the measure of truth which pertains to the economic
theory of the concept considered in the last section. The
concept, we can now agree, is, from one point of view, 'made
for use'. But the use to which it is put is not the control of
nature, as the theory supposes. It is only useful because con-
ceptual reflection facilitates knowing, leads to further know-
ledge. Its usefulness belongs primarily to the theoretical,
rather than to the practical sphere; although, admittedly, the
greater knowledge which it brings in its train may lead in turn
to a greater control over nature. But conceiving is more
immediately useful in so far as it helps to make possible
further knowledge. And conceiving is useful in this sense, we
argue, just because it is not merely a doing, but embodies
within itself and applies to the present much past knowledge.

It is necessary, however, to repeat that our discursive
reasoning leads us most often not to certainty but to prob-
ability. Occasionally, we are able directly to apprehend the

ii8 THE NATURE OF KNOWING

truth as the outcome of the thought-process. But, usually, we
do not enjoy such good fortune. The assertion of a probability
is as far as we can get. But may it not be argued that the
gaining of such a probability is itself a step forward in the
process which brings us nearer the direct apprehension of the
truth ? For the probable cannot be an end in itself. We cannot
rest satisfied with it. Nor does our search end when we have
attained it. It is true that in our practical life we only desire
an effective control over nature, and if an estimation of the
most probable gave us the desired control we should be
satisfied with it. But it does not do so. Even in the practical
sphere we need certainty; probability is only a makeshift.
The doubt and uncertainty which are present in opining breed
hesitation and awkwardness in action. And on the theoretical
side it is obvious that the end of our inquiry //rust be .the
attainment of certainty. Now, from this larger point of view,
w r e may look upon any mediate process culminating in the
attainment of the probable as itself a part of a larger process
whose natural culmination would be complete certainty. In
terming this larger process 'mediate' we might be stretching
the word beyond its ordinary usage. But it would certainly
be a mediate process in this sense, that through it an end
would be attained, namely, certainty. Viewed thus, the attaining
of probability would be part of a vast mediate process towards
full and complete knowledge of the real.

We may conclude the present section and chapter by sum-
ming up the results of our inquiry as to the nature of mediate
knowledge. We have seen that the evidence available does not
justify the assumption of two completely distinct kinds of
knowing, the one immediate, the other mediate. We have
ample evidence of the existence of immediate knowledge ; we
have none of the existence of a mediate knowledge distinct
from it and opposed to it in character. It is not denied that
X may be known indirectly or mediately. But this knowledge,
on examination, reduces itself to a knowing directly that X is
implied by certain premisses. The actual knowing present
is direct, in spite of the fact that we talk loosely of knowing

DISCURSIVE REASONING 119

X indirectly. If we choose to maintain the term 'indirect
knowledge* as meaning the (direct) knowledge of an implica-
tion considered not so much in itself but according as it gives
information about the implied, then no harm is done.

The issue is complicated, however, by another fact. Knowing,
frequently, cannot occur without a certain preparation in
thought, which involves a right use of much past knowledge.
First, certain conditions have to be satisfied. The act of
knowing (or opining) presupposes a thought-process, fre-
quently prolonged and intricate. This process also is rightly
termed 'mediate', for through it we know, and without it we
should not know; and there can, again, be no objection to
this usage of the term, so long as it is understood that the
knowing is the final act, the culmination of the process, but
not. the pnx?ss itself. Search where we will, we cannot find
any fair instance of a knowing which is a process. And until
we find such an instance we cannot admit that it ever exists. 1

There remains one final question. Can we claim to have
discovered in discursive reasoning that perfect example for
which we seek, that experience which is through and through
knowing? Quite clearly we cannot. Knowing is invariably
present in discursive reasoning, but does not exhaust its nature.
In particular, that flash of illumination which occasionally
brings the reasoning process to an end is certainly knowing.
But it is never the whole of the experience. To consider it in
itself, we have to abstract. None the less, we can claim to have
learnt much in this chapter as to the nature of knowing, and
as to the manner in which and the occasion upon which it
functions within us. Mediate knowledge, we learn, is not
completely and wholly different from immediate knowledge.
On the contrary, the evidence seems to point to the fact that
knowing is one in nature throughout. Throughout it is the
immediate apprehension of the real. Differences between kinds

1 In denying the position that knowing is a process we have, of
course, in mind a logical process. From a purely psychological point
of view knowing must be a temporal process, however short that
process be. It takes up time.

120 THE NATURE OF KNOWING

of knowing are really differences in the types of process which
enable us to know, which liberate our capacity for knowledge.
These differences form the subject-matter of logic. But the
epistemologist and the metaphysician must concern themselves
with the act of knowing itself, and must consider the difficult
problem presented by the fact of knowledge. Whence comes
this wonderful power, none the less wonderful in that it is
meagre and limited in human minds? How can we explain
knowing ? And what must be the nature of reality if within it
there is a knowing mind? These questions, we believe, are as
vital for philosophy to-day as they ever have been, and cannot
be disregarded. In this essay, however, we mean to confine
ourselves to description, and shall not attempt any explanatory
answers. In a last chapter we shall try, by broadening our
outlook, to complete this purely descriptive work/

INTUITIVE APPREHENSION

THE argument of the earlier chapters points to the following
position : the knowing act always remains identical in character
however much the circumstances in which the knowing occurs
may vary. Throughout, it is an immediate apprehension of the
real. But is this position sound ? Is the knowing act one and the
same throughout ? For even though we may now be prepared
to accept the position with regard to the knowing present in
the sensory experience and in discursive reasoning we may
yet wish to make a reservation with regard to certain expe-
riences which are abnormal but yet definitely cognitive. It
is frequently assumed implicitly, and sometimes explicitly
claimed, that there exist types of cognitive experience in which
the actual knowing differs completely from, and definitely
transcends, the knowing present in the sensory experience
and in discursive reasoning, the ordinary knowledge of every-
day. Clearly, if such types do exist, the above position cannot
be sound. The knowing act cannot be one and the same through-
out. The inquirer into the nature of knowing must therefore
face this difficulty. And in the first section of the present chapter
we propose briefly to consider some experiences which have
been put forward as instances of 'transcendent' knowledge.
The term 'intuition' is sometimes used in this connection and
in* the second section we shall proceed to consider this term,
and to give an account of the 'intuition' whose existence we
feel it essential to posit in describing knowledge.

' Transcendent* Knoivledge

Pure knowledge, it may be urged, is enjoyed by few mortals.
The realm of the mundane, where men grope about in darkness,
can provide us with no instance of pure knowing. For, as it is
in its purity, knowing does not belong to the everyday life
of man; on the contrary, it is something 'other-worldly* and

122 THE NATURE OF KNOWING

inspired, something foreign to man's natural estate. It may best
be described as an intuiting. It is the vision of the poet, the
illumination of the artist, the contemplation of the mystic, and
the faith of the saint. It is knowledge of reality, not through
slow and laborious processes of intellectual activity, but through
an immediate and complete apprehension of what most truly
is. Such knowledge does not belong to the common mass of
men: genius alone enjoys it. The gap between it and those
instances of knowledge which have thus far engaged our
attention is as wide as that between the infinite and the finite,
the perfect and the imperfect.

This is the extreme form of a view which is hardly ever held
by philosophers, but which is frequently found elsewhere.
Ordinarily w r e do believe that the genius has his own way of
knowing and that his way is not ours. Now yi seeking to
examine this belief \ve meet, at the outset, with a very serious
difficulty. To discuss the character of the 'transcendent'
knowledge and to compare it with the rest of human knowledge
we need first to have experienced it ourselves. Otherwise, we
shall be discussing that of which we are ignorant. And if, in
our ignorance, we make any assertions whatever as to the nature
of such pure knowledge, the truly inspired person may speedily
turn upon us and hold that our assertions are unsound and
our account thoroughly false. Certainly, it is but natural that
the vision of the greatest minds cannot be transmitted in its
entirety to lesser minds. This is not because the vision is a
mere subjective experience that cannot be shared ; nor because
such men make any effort to keep the vision to themselves;
but simply because other minds are not great enough to partake
fully of that which the inspired person has to give. It is the
mark of true genius that as one increases in appreciation of its
productions and enters further into the mind of the artist, or the
mystic, so one finds still greater depths unplumbed and new
truths left undiscovered. And if such inspired men tell us that
the insight they possess is something very different from
discursive reasoning even at its best we have no right to
disbelieve them. If the seer is convinced that the Spirit of Posey

INTUITIVE APPREHENSION 123

or of Painting has breathed upon him, or that he has been
illumined by the Contemplation of the Light, or has himself
taken part in the mystic dance around the Throne of God,
trying thereby to express to us by metaphor what is otherwise
wholly inexpressible in our everyday language, we must simply
take it for granted that something has happened to him which
does not happen to us. For this reason, what we have to suggest
in this connection and throughout this section is very definitely
tentative and uncertain in character, and we make no pretence
either to a complete thoroughness or to an authoritative finality.
We thus readily recognize that the vision of genius is far
removed from the knowledge of every-day. We are anxious not
to belittle the difference between them, for to do so would be
to miss the real problem altogether. None the less, we cannot
but feel that the view as set out above goes to too great extremes.
A mysterious difference in kind is posited where there may
after all be nothing but a difference in content known. Truly,
the ordinary mortal cannot fully share in the inspiration of
genius. But this fact in no way necessitates the belief that such
inspired knowledge is totally distinct in kind from everyday
knowledge. It may, of course, be so; yet it equally well may
not. For it is surely illogical in the extreme to argue that since
I know nothing of an experience it must be totally different in
structure and kind from those experiences with which I am
already familiar. Yet this seems to be the drift of the argument.
Of this thing I know nothing; therefore, it must be completely
and mysteriously different in its nature from everything I do
know. But to think so is to think fallaciously. The unknown
need be no more mysterious than the known. And in this case,
we suggest the difference may be merely one of content. Does
it not seem that the real ground for the distinction between
'every-day* and 'transcendent' knowledge lies in the difference
in nature between the object known in the two cases? It is,
primarily, the fact that poet, saint, and mystic claim to 'draw
down heaven', whilst the ordinary man in his ordinary mood
does not lift his eyes above the world around him, that leads us
to allot the term 'heavenly* to the knowledge of the former and

i2 4 THE NATURE OF KNOWING

'earthly' to that of the latter. And it may yet prove that changes
in content, the what known, need not involve a change in the
knowing. One and the same power of apprehension may be
capable of knowing both 'earthly' and 'heavenly'. In such a
case, though what he knows is so very different, the actual
knowing of the most inspired poet may be identical in nature
with my own.

But surely, it may be objected, that which knows the finite
cannot also know the infinite? Yet this, perhaps, is what the
objector too readily assumes. At least, he should first prove his
thesis. For it is not inconceivable that the mundane and the
transcendent, the extraneously conditioned and the absolute, the
finite and the infinite may, in spite of all their difference, be
apprehended by acts of apprehension identical in nature.
The 'inspired' person certainly deals with quite a different
objective world, and no doubt quite a different set of circum-
stances may be necessary to enable him to know, yet it is not
absolutely necessary that his knowing, as such, should differ*
in ultimate nature from the knowing act involved in everyday
knowledge.

At the same time, admittedly, it is only too evident that if
certain theories with regard to the nature and status of human
knowledge were sound, the mind that knew the finite could
certainly never succeed in knowing the infinite. As an instance
one might take the theory which confines human knowledge
strictly within the limits of the phenomenal and holds that
valid knowledge is only possible for the human mind in so far
as the mind itself sets out the principles which the world it
knows obeys. It would then be clear that the transcendent
could never be known by such a mind. To know the transcen-
dent it would need a capacity and power wholly different in
nature from that which functioned when it gained knowledge
of the phenomenal world. Thus it would be necessary to assert
the existence of two completely distinct kinds of knowledge ;
the one knowing the phenomenal, the other the transcendent
or ultimately real; the one, in part at least, a creation of its
object, so that the object depends upon it for its very existence,

INTUITIVE APPREHENSION 125

the other a discovery of an independent object. But such a
dualistic interpretation of knowledge is in no way necessary,
we contend, if the description given in these pages is the true
one. For we do not believe that human knowledge is ever of the
phenomenal. No such distinction as that between the pheno-
menal and the real needs to be introduced, on our view. Know-
ledge, if it occurs at all whether it be at a high or a low stage
of mental development is of the real. It is true that the mind
conceives a world in imagination to aid it in its knowing; and
certain of its principles are in reality only well-founded hypo-
theses; but it does not then know the conceptual, it knows the
real with its aid. 1 Wherever knowledge occurs, the object is the
real. And, we add, the knowing act which enables us to know
the real at the finite level may also quite conceivably know the
real t;hat lies beyond it.

> If we now take stock of our position, we can affirm, on the
one hand, that the supreme knowledge of the 'inspired' person
need not 'be totally different in kind from the knowledge of
everyrday, even though its content differs exceedingly; and
on the other, that the account of knowledge given in the earlier
chapters of this essay may, in its general outlines, quite con-
ceivably apply even to the 'transcendent' knowledge discussed
in this section. We may, therefore, ask whether, as a matter of

x * The only occasion upon which the mind can be said to know the
conceptual is in self-consciousness. The mind turns back upon itself
and observes its own conceptualizing. This is knowledge again of
the real, but of a real which is subjective rather than objective.
Incidentally, we may add, the argument of the previous pages also
holds with respect to the subject's knowledge of itself. It is admittedly
difficult to understand and to describe this knowledge. In particular,
the fact that the subject known can never be an object of knowledge,
since as such it would not be subject, seems to make the task of
describing self-consciousness additionally difficult. But surely these
facts do not make necessary the conclusion that the act of knowing
involved in consciousness of the self is completely different from
the act of knowing involved in all other cognitive experiences. And
yet this belief is frequently found. We suggest that it is at least
conceivable, and, as far as we can see, probable, that our actual
knowing of the subject is identical in character with that of the object.

26 THE NATURE OF KNOWING

r act, 'inspired' knowledge is of the same kind as the knowledge
rf every-day, and whether the account given of the latter does
not hold equally well of the former ?

We propose to suggest that an affirmative answer ought to
be returned here, that just as it is with everyday knowledge
so with the knowledge of genius, first, there is an effort of some
sort to enable him to know, and the knowledge itself which
follows is an immediate apprehension of the real. For our
information we have to rely largely upon that which the knower
chooses to tell us and he, most frequently, is not very communi-
cative on this point. Usually his mind is so full of the vision he
has seen, and he is so anxious to share it with others, or at least
to express it adequately, that he has very little time to spare
in which to describe the manner of his coming to knowledge.
For instance, it is but rarely that one finds a p^et discussing
the exact nature of his insight. That is none of his business.
He will be ready enough and anxious to present his reader with
the truth he has discovered. He has learnt of life, and what he
has learnt has so elevated his thoughts that he cannot contain
his emotion but must express it, using in the expression the
language natural to great emotional stress, namely, rhythm.
But he does not reveal so readily his spiritual history from
the point of view of his knowing. He has much to say of life,
but very little of how he learnt all he now knows about life.

The problem is further complicated by the fact that many
poets from time to time have held a theory as to the nature
of poetry which differs essentially from the one implied in the
above paragraph. The poet, they w r ould say, does not, as a
matter of fact, seek for truths at all. If he does so, then he is
no longer a poet, but a philosopher or a scientist. The poet, as
poet, is an artist and, like every other artist, his work is to
amuse and to interest by doing something well though, of
course, the amusement need not be frivolous in character.
It is not his duty to seek new truths about life. His task is to
create Beauty and not to discover Truth. There are thus (at
least) two types of poets : the one strives to know and to reveal
the inward truth of life and holds this to be the proper business

INTUITIVE APPREHENSION 127

of the poet, the other is content to amuse by using words in an
exquisite manner. The latter type does not seek to know.
The ordinary knowledge of every- day is sufficient for his
purposes. His aim is to set it forth in a way that will please.
His real interest is in technique and not in content. 1

Fortunately, it is not necessary for us in this work to settle
the issue as between the two schools. Our interest lies clearly
in that type of poet for whom art involves, in part at least, the
gaining of new knowledge. For the purposes of this argument,
therefore, we may disregard the poet who is interested in
technique and form alone. And of the other we shall ask, Is his
knowledge different in kind from the knowledge of the scientist
to take our best instance of * everyday knowledge'? How
does the knowledge of the one compare with that of the other ?

By way of Answer, we may at least point out certain details
which appear to be analogous in the two cases. Both seek truth
and both find hindrances in their path. We have already shown
how a scientist finds it necessary to work mediately by way of
discursive reasoning. It is by such laborious processes alone
that his immediate apprehension of new truth becomes possible.
So, too, the poet must struggle and strive if he is to gain a more
comprehensive insight into the real. It is no easy, effortless
acquisition. Biographies of this more romantic type of poet
have constantly to mention severe and often bitter mental
conflicts, from which relief is found in the actual composition,
the expression in verse, when it becomes possible. Mental
struggles of this kind with the passionate emotional outbursts
that accompany them seem to be the inevitable counterpart of
the more romantic poetic inspiration. Their source is a failure
to 'gain the vision'. Truth does not flash upon the poet's mind
spontaneously whenever he desires to understand. Poet and
scientist are alike both in seeking for truth and in finding it
only after much effort.

They proceed to overthrow the obstacles in their path,
however, in a different manner. The scientist proceeds logically.

1 The difference here is, perhaps, exaggerated in order to make
it clear.

iz8 THE NATURE OF KNOWING

He reasons discursively, from one relation to another according
to the Laws of Thought. He 'perceives' various implications
and uses his knowledge to systematize his world as completely
and as coherently as he possibly can. Consideration of this more
systematic world enables him to apprehend new truths which
in turn lead on to others. Now though the workings of the poet's
mind remain largely a mystery, it is quite evident that he does
not proceed in this manner. He does not stand aside to argue
from one abstraction to another. Instead, he enters into the
centre of the flood and seeks to live out as complete a life as
is humanly possible. In imagination, at least, he will taste of
all things, and as he tastes and lives he learns. Thus, a more
or less permanent conflict within the one mind between
two strands of temperament, between, for instance, an austere
asceticism and an indulgent laxity, may of itself,Jead to many
experiences which result in greater insight into the real. Out
of the conflict, as a flash, a new truth about the life he lives
comes to him, and it is this truth, now apprehended by him-
for the first time, which he expresses in his poem, finding relief
in the expression. Thus it is the living out of life either actually
or in imagination which provides the medium through which
the poet attains his knowledge. It is not by processes of discur-
sive thought carried forward in a strictly logical fashion that he
usually overcomes the hindrances. It is as if the power that he
possesses of knowing lies captive within him until he experi-
ences in hisown tense fashion thepleasures, the joys, the sorrows,
and the griefs of life, its strain and its conflicts. To know he
must first live intensely and imaginatively. This vivid imagina-
tive experience seems to be the poet's mediation whereby he
attains the end he has in view, namely, a fuller understanding
of life. 1

But does the poet really know? Or does he merely delude

1 Incidentally, imagination is also an essential feature of scientific
thinking, as we have shown. But, in science imagination is the hand-
maiden of logic; in poetry, imagination as a means to an end, namely,
the attainment of knowledge, is supreme. (I am, of course, thinking
of the romantic type of poetry only.)

INTUITIVE APPREHENSION 129

himself in thinking so? To this question we can provide no
definite answer, and that for reasons already touched upon.
When it is asked, Did this thinker gain greater knowledge of
the real through this particular piece of discursive reasoning?
I can often answer, yes, or no. For, starting with the same
premisses and carrying out the same process of reasoning, I
myself come or do not come to the same conclusion. That a
process of reasoning implies such and such a conclusion is
something neither true nor false for me until I have 'seen' its
truth (or falsity) for myself. But it is, to say the least, extremely
difficult for the ordinary man to capture the poet's experiences
in order to verify his conclusions. Yet until he does so he can
neither confirm nor reject that which the poet claims to be true.
Of course, the information may be such that it can also be gained
through a process of discursive reasoning, and in this case I
could verify the poet's assertions without entering into his
experiences. But when this is not possible I am powerless to
pass judgement.

Byt the matter goes deeper. If it be ever true that the poet
does arrive at knowledge independently (either wholly or in
part) of logical processes of discursive reasoning, this would
mean that the real may on occasion be known by a method
that for want of a better name we shall call 'non-logical 1 .
In such a case, knowledge of the real would not be the monopoly
of scientific thought nor even of the discursive process. Logical
reasoning 1 would not always be necessary for knowledge. Is
such a view sound or false ? Is a logically disciplined process of
reasoning essential for the occurrence of knowledge? Or has
the poet his own method of attaining to truth? The evidence
points to the truth of the latter alternative, but one cannot
commit oneself until much further consideration has been
given to the matter. The whole problem is obviously one of
extreme importance. 3

1 We do not say Reason, which may mean something very different,
namely, that which we here refer to as the 'knowing act*.

3 Our knowledge of the principles presupposed in reasoning is an
excellent instance of knowledge gained prior (logically) to all processes

I

tao THE NATURE OF KNOWING

But though we cannot answer definitely, our account of know-
ing would not be proved invalid if the above view were sound.
In particular, we could still assert the identity of the knowing
act itself throughout all cognitive experiences. For in any pro-
cess of discursive reasoning, as we have described it, the know-
ing act is a distinguishable feature within the whole of the
experience. It is not the whole of it. And the fact that the
poetic experience qua cognitive differed from discursive
reasoning would not necessitate the existence of two completely
distinct acts of knowing. The knowing act itself might be
identical in character within both experiences although the
circumstances necessary for knowledge in each case differed
completely. This would suffice to make them radically different,
and we do not wish to minimize that difference. But
in such a case the difference would lie not so -much in the
knowing itself, as it actually occurred, but in the prior efforts,
whereby the mind succeeded in working itself into a favourable
position for knowledge. To conclude, then, whether we speak
of poet or scientist, we can say that each possesses power to
know, but that this power lies in bondage and needs to be
freed before the act of knowing can occur. The freeing takes
different forms, the objects known may also differ exceedingly,
but the knowing act, that flash of illumination, when it occurs,
seems to be identical in both cases, and there is nothing in our
general account of knowledge up to the present which makes it
impossible or even difficult for us to believe this.

We are now in a position to give some sort of answer how-
ever hesitating to the question with which we began this
section. Is there a 'transcendent' knowledge wholly distinct
from the everyday knowledge of ordinary life ? We answer that,
confining our remarks to poetry, the knowledge which the poet
claims to gain differs in objective content from the knowledge
of every-day ; that, furthermore, it also differs on the subjective

of reasoning. So that it would be patently false to assert that all
human knowing occurs as the result of discursive reasoning. The
present question, however, is: Has the poet discovered a method of
procedure ', which is non-logical and yet leads to knowledge?

INTUITIVE APPREHENSION 131

side, since in the poet's experience the power to know which
he possesses is liberated and given freedom to operate in a
manner that is, perhaps, uniquely his; but that the knowing
act itself thus liberated does not, so far as we can see, differ in
the two cases. The 'heavenly' knowledge of the poet is, in its
essential character as knowing, in no way different from the
'mundane' knowledge of the scientist and the ordinary man.
Regarded from the point of view of its content it may be fuller
and more perfect, as the poets themselves would claim, but
from the point of view of the knowing act involved and of its
functioning, it seems to be identical in nature with the types
of knowledge discussed earlier. Such is the suggestion we make.
We have no space left in which to consider other branches
of art, such as painting, music, sculpture, and so on. 1 Nor can
we discuss jiere another type of knowledge, which may be
regarded as 'transcendent' in character, namely, that which
makes possible the moral life. Perhaps, in no sphere are there
problems so difficult of solution as those which face the moral
philosopher, and, at the present stage, we prefer not to venture
any opinion whatever with regard to them. We may, however,
justifiably point out what is, indeed, common knowledge, that
the moral life is not achieved easily. Here again a struggle is
involved a struggle, moreover, w r hich is two-sided. For
while it is no easy matter to know what path we ought to tread,
it* is even more difficult to tread it in actual practice. From the
point of view of an epistemological inquiry, of course, the
gaining of new moral knowledge would be the more interesting
feature, though no doubt the effort at living out what one per-
ceives to be the good life would itself lead to a deepening of
our knowledge and could not be ignored. Further, the impera-
tive which commands us unconditionally to do that which we

1 It may be objected that in taking the case of the poet whose
purpose it is to gain knowledge we have not really been dealing with
an artist at all, since it can never be the purpose of an artist, as
artist, to gain knowledge. If this be true, we agree that the instance
taken is unsatisfactory: But, on that hypothesis, the consideration of
art would not come within the scope of this essay.

i 3 2 THE NATURE OF KNOWING

determine to be the right thing in the circumstances would
itself need examination. With regard to the gaining of new moral
knowledge, however, we should be tempted to say that moral
knowledge is a direct apprehension of truth, though an act of
apprehension w r hich could not occur without, and except
through, the prior occurrence of certain auxiliary processes
which make the apprehension possible. What these processes
are we do not here profess to explain. 1

Finally, to consider the matter of 'transcendent* knowledge
adequately, we should have to devote very serious attention
to religious knowledge and to faith. Now in so far as we use
this latter term to express a kind of knowledge, we may mean
by it one of two things. In the first place, little more may be
meant than hearsay knowledge, as when one knows a matter
not through finding it out for oneself, but by hearing about it.
(Most of our everyday 'knowledge' is of this kind.) Thus by
faith is often meant simply the acceptance of the dogmas
taught by some religious body or other dogmas which pro-
fess to be truths gained earlier in the history of that religious
body. Of course, if the acceptance is sincere and not merely
nominal, faith, even in this sense, does involve some measure
of finding out for oneself. The sincere believer does not blindly
swallow everything offered him. In the last resort, he can accept
nothing which openly conflicts with his own experience and
thinking. He does accept a position without having discovered
the full truth about it for himself; nevertheless, as much truth
as he has discovered seems, taken all-in-all, to point to the
truth of this position. In such a case faith is the theoretical
counterpart of trust. When I say I have 'faith* in a person or
'faith' in some project my use of the word implies that I do not
know for certain at the time how this person will act in the
future, or whether the project will turn out in the hoped-for
manner. But my knowledge of the person, and again of the
project, is sufficient to make me feel fairly confident as to the

1 Where we fail to gain complete certainty on a moral issue, we
may either gain a measure of probability or suspend our judgement
entirely.

INTUITIVE APPREHENSION 133

issue, and though I have no certain knowledge I have 'faith'.
No doubt, much religious 'faith' is of this kind, and for beings
who are not omniscient (but who have yet to live and to act)
such 'faith' is a necessity. We accept on hearsay a dogma and
believe in it because it confirms our knowledge and our own
experiences and even explains them. What we know in no
way establishes the dogma's validity; but it points the way of
the dogma; and so we accept the latter though we ourselves
did not discover it for ourselves but learnt it from another.
Frequently, we mean no more by faith than this acceptance on
our part of another's discovery in the religious sphere.

In the second place, however, we may mean by faith the
first apprehension of such dogmas, the 'inspired' knowledge of
religious genius. Such knowledge, it is only too clear, differs
radically in <*ne important respect from the knowledge of every-
day things. When I know the material world around me I
know it, we usually imagine, by my own efforts. In such a case,
we should not ordinarily say that the object known helps me to
kno\y. When, however, I know some other mind the object
here may help me to know. I know more about my friend than
I do about a perfect stranger, and this because my friend has
in part 'revealed' himself to me. The object known has helped
the subject to know. Now in the case of religious knowledge
the object is God, Omniscient and Most Perfect. Hence, if
man learns of God it can only be because God Himself imparts
the information. Knowledge of God cannot be conceived as
something which we ourselves discover by our own unaided
efforts. On the contrary, we naturally feel that if we know Him
at all, it must be because He himself has chosen to reveal
Himself to us. Man's knowledge of God must be revelation;
in it God Himself discloses to man His own nature. The term
'revelation' should be retained. It aptly describes this most
characteristic feature of man's religious knowledge. Yet, grant-
ing that here the subject-object relation is of a unique kind,
there is still no necessity to suppose that the actual knowing is
different in character from all other instances, for it is quite
conceivable that God might choose to reveal Himself to us by

134 THE NATURE OF KNOWING

way of the ordinary channels of knowledge. Our knowledge
in this instance may proceed in a fashion identical with all other
knowledge, the only difference being that we cannot suppose
this knowledge could ever occur were it not God's will that it
should. But is faith, then, as the religious genius's knowledge
of God, actually identical in nature with other kinds of
knowledge, or is it distinct in kind? 1

Again, we can only suggest the possibility of an affirmative
answer. In the first place, the prior struggle which we have
come to expect is obviously present. However strong and pure
be man's desire to know Him, God does not quickly reveal
Himself. The greatest religious teachers that humanity has
known all unite in this testimony that God is to be found only
by dint of ceaseless search. And, certainly, few ideas show
slower development historically than does the idea of God. If
God is to be known at all, He cannot be known in any effortless
way. First there must come, as a necessary precondition, a
process in all cases arduous and prolonged in which the
mind is prepared for the knowledge of God. What everyone
would wish for, if he once thought it possible, namely, the
immediate attainment here and now of a complete knowledge
of God, is, as a matter of fact, wholly impossible. The religious
genius gains his insight into God's nature gradually. Always,
so it would seem, there must be a preparation of some sort
through which alone that insight can become possible.

But in what does the preparation consist ? Do we seek Him
through the medium of logically constructed processes of
thought, or through our imaginative and emotional experiences ?
Clearly both media have been used. Philosopher and poet have
each sought for God in each his own way, and on occasion both
claim to find Him ; while, often enough, since there is something
both of the poet and philosopher in every man, the two methods
have been combined. The mediation, that is to say, is not

1 Throughout the above paragraph I have been following so closely
upon Professor C. C. J. Webb's argument on this matter which
seems to me very sound that I must here be allowed to acknowledge
the debt.

INTUITIVE APPREHENSION 135

invariably 'logical' in our sense nor invariably imaginative, but
may be either or both. Nor are these the only ways of approach.
On the contrary, it would seem as if every path that leads to
knowledge, of whatever kind, can be utilized in the search for
God. 1

In the second place, the experience would be impossible had
we not power to know, and were not this power actualized,
in the experience. For though we admit the uniqueness of this
instance of knowledge and recognize in it God's revelation of
Himself to man, nevertheless man must be capable of receiving
the information imparted to him, he must himself possess the
power of apprehending the Object. After appropriate prepara-
tion the knowledge comes like a flash to the active mind.
Religious knowledge cannot be a passive experience. Revelation
is only pos^ble in so far as man possesses power to know and
to apprehend. Thus, though the faith of the saint, as a cognitive
experience, may differ greatly from our ordinary everyday
knowledge, there nevertheless pertain to it certain general
characteristics which belong to human knowledge in every
sphere. Through some process or other a capacity is liberated,
what is potential within us is actualized, so that we attain
knowledge of the Object.

All mystic literature is a constant re-emphasis of this truth.
Every man, whatsoever his estate and condition, possesses
within him potentially the knowledge of God. The Light is
within, even though at present it be enshrouded in darkness.
Consequently, the mystic consciously sets himself the task of
actualizing the potential. He seeks a Way, whereby he may
attain the fullest experiences possible. Firstly, he holds, there
must be a purifying, a moral disciplining, a giving up of the
'life of the flesh'. Secondly, Wisdom must be diligently pur-
sued. Not only must the 'flesh' be conquered, but one's place
in life must be learnt. This understanding of life is gained in
many ways. It may be gained by way of science and philosophy,

1 Frequently enough, for instance, a fuller consciousness of the
moral life and of its demands has led to an increased knowledge
and understanding of God.

i 3 6 THE NATURE OF KNOWING

or by way of religious devotion, or by the contemplation of the
beautiful, or even, lastly, by faithful and long-sustained service
to one's fellow-man. By such moral disciplining and by such
acquisition of Wisdom the soul of man is prepared for the
beatific vision of God. If the Way be truly and faithfully
followed, then gradually our faculties will be freed, the obstruc-
tions which encompass the Light within will be removed, and
the highest knowledge together with the noblest emotions will
be ours. 1

At this stage the knowing act within will be freed completely.
Its final emancipation will have occurred. In such a case, if
we know at all, it will be with God's knowledge, which differs
from finite knowledge in that nothing ever hinders its function-
ing. At the finite level it is only with difficulty that we can
conceive of such an experience and we cannot foel sure that
'knowledge' is the right term to apply to it. Yet implicit in the
position of these mystics is the belief that, in the last resort,
our knowledge is not completely different from God's. The
Light within each soul is already something divine. The know-
ing act itself, it is implied, is infinite. It is its opposite, that
which hinders its operation and that which we must first over-
come and remove if we wish to know it is this, which is finite
in the cognitive experience. God's knowledge, on this view,
actually is what our knowledge would be if the power to know
within us were liberated not spasmodically, here and there,
but everywhere and in every circumstance.

But, at present, we do not wish to follow out this extremely
speculative line of thought. Our task is a humbler one. As the
conclusion of our reflections in this section we are not able to

1 It is necessary, however, to note one important modification.
The mystical consummation of man's experience, the last stage of
all, is not the mere knowledge of God. Higher than knowledge of God
is unity with Him. When both the intellect and the emotions of man
are developed to their uttermost, then, at such transcendent moments,
one's self will be merged within the Divine. One's will, one's thought,
one's emotion so the mystics claim become God's; though, indeed,
it may very well happen that we ought no longer to speak of will,
thought, and emotion in this context.

INTUITIVE APPREHENSION 137

offer any demonstrated facts. We can only make a suggestion,
which seems to us to be truer in this connection than any other,
namely, that the highest cognitive experiences of which man is
capable are not altogether different in nature from our more
ordinary cognitive experiences. They share many character-
istics in common with the latter. The mystic's striving through
years of patient labour for a completer insight into the Divine;
Nature, the storm within the artist's mind before the vision
flashes upon him, the conflict of desires and the * inward argu-
ment' which precedes the intuitive apprehension of one's
obligation and duty is there not here something analogous
to the intellectual struggle, the mediation presupposed by each
new act of knowing in the sphere of discursive reasoning?
And does not the analogy hold, we suggest, because in their
ultimate nature all these experiences are one and the same?
They are all finite cognitive experiences, that is to say, experi-
ences in which the mind, already possessing the power to know,
can nevertheless only know in certain definite conditions which
must first be secured. Now, if this suggestion is sound, then
the 'transcendent' knowledge of inspired genius does not differ
fundamentally and in kind from the 'mundane' knowledge of
the ordinary man. Genius, in all these manifestations of it,
seems to be the consequence of a better use of one's faculties,
resulting in a more complete liberation of mind than is usual.
fjut these faculties do not belong to genius alone, they are
latent in all of us. The inspired person follows a path that all
may follow, and that everyone, indeed, actually does follow
whenever he succeeds in gaining new knowledge by whatever
method.

Thus, as we see it, there is no greater mystery in the know-
ledge of the genius than in the knowledge of any one of us.
The real mystery is the act of knowing itself if that which
is so natural to us as to be, perhaps, what we most essentially
are can actually be termed 'mysterious'. If it were possible to
explain this act of knowing, the core of all our cognitive experi-
ences, the further difficulty as to the nature of the higher
knowledge of inspired men could hardly prove insurmountable.

i 3 8 THE NATURE OF KNOWING

The real problem for the epistemologist, who seeks not only
to describe but also to explain, is to discover the source of
the mind's cognitive power, which is as clearly present in the
everyday knowledge of the man in the street as in the superb
vision of genius. From this point of view, the lowest type of
knowledge is no less remarkable as a phenomenon, even though
( it emerge in those experiences which we share with the beasts
of the field, and even though all the information it has to give
is of some drab corner of the world around us.

The Intuitive Character of the Knowing Act

*
In this section we propose to argue that the knowing act is

intuitive in character. On our view, no other term more
adequately expresses the characteristic nature of the knowing
act whenever and wherever it occurs. But if we use the adjec-
tive we must make clear what we mean and what we do not
mean by it. For a loose use of the term is dangerous. There
is a healthy tendency nowadays in philosophical circles to
deprecate the over-frequent usage of the term 'intuition'. Too
often in the past its use has conferred an appearance of wisdom
upon what is actually loose thinking, and on many an occasio'n
it has served as a cloak to hide real failure. It is, assuredly,
one of the easiest terms to misuse; and, whenever it appears,
one should be on one's guard against the intellectual laziness
of which it is a frequent sign. For the term readily lends itself
to false usage. This is, perhaps, due to the fact that by an
'intuition* we frequently mean an experience which we have or
do not have, but which is not further analysable into anything
other than it itself. Consequently, simply to avoid greater
mental effort, a lazy thinker will be tempted to call every
experience, which he finds difficult to analyse, 'intuitive'. The
inevitable result is that the term has become suspect.

Nevertheless, we consider the use of the term justified in thq

INTUITIVE APPREHENSION 139

case of the knowing act. The knowing act is an intuition.
This does not mean, however, that the whole cognitive experi-
ence is through and through intuitive; for the knowing act is
not the whole but a part only of that experience. The fact is
that we have failed to find an instance of the perfect knowledge
for which we seek in this essay, an experience through and
through knowing and nothing else. None the less, we have,
discovered true instances of knowing, and we claim for each
instance that it is intuitive in character, though the knowing
in each case is only a part of a larger experience. This is obvi-
ously true of our everyday experiences. And it seems equally
true of any higher experiences we might enjoy. For even
though we were to admit that the genius enjoys supreme mo-
ments in which the mind is, as we say (speaking loosely), filled
with illumination or inspiration, yet such moments are essenti-
ally parts and parts only of knowing experiences, and each
part is dependent upon the rest of the experience to which it
belongs. The moment of complete insight is the consummation
of a whole experience, and is only isolated from it by a definite
act of abstraction. Though we recognize the presence of real
differences, both objective and subjective, between 'transcen-
dent* and 'every-day' cognitive experiences, these differences
are yet not sufficient to destroy the general identity of character
which, so we argue, persists throughout these experiences.
In the two groups, the whole cognitive experience is a process
involving the liberation of the knowing function on the one
hand, and its actual functioning on the other. Now this func-
tioning, this act of knowing, seems to be identical throughout ;
and throughout it is intuitive in character.

Moreover, we should say that the knowing act is the sole
intuition. By this we mean that it alone satisfies our notion of
what an intuition should be. It is direct and immediate know-
ledge. Its object is the real, not a representation nor a copy of
it. It is no process, but is an act of apprehending and this,
though we admit that whenever we find it it is embedded in a
process. It is mi generis, like nothing other than it itself. It is a
unique form of mental functioning, and, finally, it is infallible.

i 4 o THE NATURE OF KNOWING

These characteristics which pertain to the knowing act do, we
believe, justify us in terming it an intuition, and since the
knowing act is alone in possessing all these properties we shall
use the term exclusively to signify knowing in this sense. For
the sake of precision and consistency, therefore, we shall
reject certain other usages of the term. For instance, we shall
not call the hypotheses of the brilliant scientist 'intuitions'.
He only intuits, in our sense, when he knows with certainty.
Nor shall we talk of a woman's 'intuition', when we merely
mean a form of shrewd guesswork. Nor again shall we use it
in speaking of animal knowledge, if such knowledge be held
to differ in kind from human knowledge. 1 Nor, finally, shall
we continue to talk of 'sensuous intuition', if by this be meant
the 'receiving' into the mind of a 'given' manifold, the affection
of the mind in sensation. We shall reject all these usages f
the term and confine it strictly to the act of knowing the real.
The use of the phrase 'sensuous intuition', indeed, deserves
more than a passing notice; for it leads to much confusion.
By it is meant the seeing of the colour, the hearing of the sound,
and so on. Now if we term these 'intuitions' we already suggest
that they are instances of knowing. For to call something an
'intuition' is to give it a cognitive character. It is impossible
to rid the word of that suggestion. Accordingly, when the naive
person refers to the seeing of the colour as a 'sensuous intuition'
the phrase exactly expresses his meaning, since just seeing a
colour is for him a knowing of the real. The critical person,
however, cannot but be confused. He has realized that seeing a
colour is not in itself an instance of knowing the real, as the
naive person would claim. Yet if it is an 'intuition' it must, he
also realizes, be a knowing of some kind. Hence his difficulty.
As a consequence, he is frequently led to talk of vague knowing,
or of half-knowledge, or of something which is just-not know-
ledge; and the result is confused thinking. In other words, by

1 In these pages we have not thought it necessary to consider at
length the alleged * instinctive knowledge* of the lower animals,
which, so some would urge, differs in kind from human knowledge.
Our concern throughout is with the latter.

INTUITIVE APPREHENSION 141

terming the seeing of a colour and the hearing of a sound a
'sensuous intuition* we are, to a certain degree, prejudicing
the case from the outset. The very terms we use imply (whether
we wish it or not) that seeing a colour is itself a knowing, and
if we wish to avoid the implication we must avoid the use of the
term in this connection, even though we feel that the word
'intuition' does express some of the qualities which can be
attributed to seeing a colour for instance, its directness. Fur-
thermore, the use of the term makes a sound analysis of sensory
experience well-nigh impossible. For its adoption is almost sure
to result in an ignoring of the true knowing act present in the
sensory experience. The full attention is bestowed upon the
mere seeing of the colour, and by terming the latter an intui-
tion' we suggest to ourselves and to everybody else that we are
continuing to recognize the cognitive character of the sensory
experience, although, as a matter of fact, we are completely
ignoring it. The result is that we deceive ourselves, for, becom-
ing critical and realizing that such a sensory experience (think-
ing pf it as merely seeing the colour) provides no direct know-
ledge of the real, we still think that in some vague fashion the
seeing of the colour is a knowing. But if some other term had
been used for the mere seeing of the colour, it would then be
clear that the experience (if it is merely seeing a colour) is not
cognitive at all. It would be clear that some essential element
had been completely ignored. Our use of the phrase 'sensuous
intuition', however, hides this all-important truth from us.
Knowing becomes in part a 'reception' of a 'given', and in
part a doing of something with this 'given', a constructing, a
forming, an ordering of an objective world which exists as a
vague shadowy structure of our own creation. And this con-
fusing consequence, together with the resultant scepticism, is,
we feel, the outcome of a false analysis of the sensory experience,
whose falsity, we suggest, tends to be hidden from us by the
use of the phrase 'sensuous intuition'. For this reason, there-
fore, we consider the term 'sensuous intuition' an exceedingly
dangerous one.
We confine the term intuition, then, to the knowing act. In

i 4 2 THE NATURE OF KNOWING

our opinion the term can be applied with justification to it
alone. The reader, however, may well hesitate on one point
before accepting our view. He may agree that knowing as such
is something direct, immediate, and sui generis. He may also
agree that it is not a process. In so far he would be prepared
to term it intuitive in character. But an intuition, he feels,
ought to be infallible. 1 Now nothing is more obvious than that
human knowledge is fallible. How, then, can he and how can
we call human knowledge 'intuitive'? The reader will have
here touched upon a vexed question. But though the question
he asks is not easily answered, we have no right to shirk it.
For the sake of clearness we shall first put forward in one
sentence the answer we suggest, after which we shall give reasons
for holding it. The knowing act itself, we suggest, whenever it
does occur, operates infallibly; but the concrete human cogni-
tive experience taken as a whole is fallible.

We have assumed throughout this essay that human knowing
is a fact. We admit that we have given no definite proof of
this, and, more, that no completely satisfactory proof of it js
or ever will be possible. Against a thorough-going agnosticism
we cannot bring a single argument. If the fact of knowing be
granted, however, then it means that when a man is convinced
that he knows he does, at least sometimes, know. Now if it were
possible to show that his failures are due not to the functioning
of that which we have called the knowing act but to something
else in the whole experience of which the knowing act is part
only, it would then be clear that the conviction which the
knowing act inspires is completely trustworthy. What we are
saying is that a man may be convinced, that is, satisfied in his
mind and yet err this certainly cannot be denied; but that
there is also a deeper conviction which cannot mislead. To say
that this latter ever does mislead is to adopt agnosticism straight-
way. For, if it fails us once, we cannot trust it on any other

1 Most thinkers would agree that the term 'intuition' should be
used to signify a knowledge which is infallible, and this is the view
we adopt in these pages. To deprive it of this meaning would be, in
our opinion, to emasculate it considerably.

INTUITIVE APPREHENSION 143

occasion. Our present task, therefore, is to show, in so far as we
can, that the act of knowing itself is never erroneous, that error
always enters in some other way. 1

One type of error, frequently experienced, is obviously not
due to any fallibility in the knowing act. We mean the type
found in learning by testimony. The testimony may be of two
kinds, firstly, that of other persons, or, secondly, that of our
own memories. In the former case, error can easily enter. If I
accept as true something which I have not seen to be true for
myself, I may find later that I have fallen into error. Where
we have to rely upon hearsay 'knowledge' we cannot rid ourselves
completely of this possibility. The most reliable source of
information sometimes fails us. Nevertheless, reliance on
others in this sense is a necessity. For practical purposes we
are frequently compelled to take another's word on a particular
point. Especially does this hold true of ages and civilizations in
which learning is advanced and in which specialization cannot
be avoided, for then, because of our inevitable inexpertness in
certain realms, we have to learn many items of knowledge not
by finding out for ourselves, that is to say, speaking strictly,
by knowing them ourselves, but by accepting the information
given by another with regard to them. Now in such a case the
important point is that our error, were we thus to accept what
is not true, cannot possibly be due to the functioning of the
kino wing act and does not make the latter fallacious, for merely
to accept something as true on the \vord and authority of another
is not to see its truth for oneself. Again, one's memory may
fail. I may learn by rote at some time or other an item of
knowledge originally gained either by directly apprehending it
myself or by taking it as true on the authority of another. But
later in recalling what I knew I may falsify it, usually because I
give insufficient attention to the work of recollection. Now the
term memory is, we admit, ambiguous and the problem of

1 It seems hardly necessary to add that we do not mean spoken
conviction in the above paragraph, for we sometimes say that we are
convinced when we are not, and, occasionally, the less convinced we
are the more vehement our speech becomes.

144 THE NATURE OF KNOWING

memory supremely difficult. We do not here propose to discuss
its nature. But when we mean by remembering remembering
by rote, as in the present case, then remembering something
obviously is not the same as knowing it. 1 And since this is so, a
defect in the work of memorizing, as in the above instance,
cannot rightly be attributed to the act of knowing. What we
learn by testimony, therefore, may be erroneous, but in such a
case our taking it as true is no failure of the knowing act.
Slightly different from these instances, but worth mentioning
none the less, is that type of error w r hich arises from a defect in
the media through which we communicate information to
each other. For instance, on a walk, I may see a person in a
field nearby and ask him the distance to the neighbouring town.
He knows that it is ten miles away and shouts it back to me.
But his voice is not clear, or the wind is high,*and I hear,
" Seven' '. Surely the resultant error is not attributable to any
defect in my power to know nor, for that matter, in his. We
need not further analyse these instances of erring, for all we
wish to prove is that the knowing act as such is infallible, and
in these cases it cannot possibly be held responsible for the
occurrence of error.

But in the above experiences the knowing act is absent,
though it may be presupposed, for instance, in the knowing
of my informant, or, again, in my own knowing of what I now
recall. My * knowledge' by hearsay and my bare remembering
by rote, however, are not, as such, acts of knowing. Yet there
are other cognitive experiences which involve intuitive acts
of knowing, and which, none the less, are fallible. We must
now consider these. Is their fallibility due to a fallible knowing
act ? Since we have just been considering the case of memory,
we may begin with the consideration of erroneous cognitive
experiences involving memory. Here is one such case: I see

1 The case of erring in remembering a past event by recalling it
in imagination is more difficult. No doubt cognitive elements are
definitely present here. We cannot consider this case fully, however,
without at the same time essaying an exhaustive analysis of the
memory experience.

INTUITIVE APPREHENSION 145

directly that certain premisses involve a conclusion. I take these
premisses to be true and accordingly assert the truth of the
conclusion. The information contained in the premisses, how-
ever, is simply remembered by me; and it may be false. If I
then use the falsely recalled information my conclusion will be
erroneous. Many errors in calculation can be accounted for in
this way. As an instance, we may take the simplest form of
calculation, namely, arithmetic. No one in his senses would
say that twice one are equal to eight. We immediately 'per-
ceive' that twice one is two. But if we were given a more com-
plex multiplication problem, running into many figures, we
might, owing to the strain upon our attention and the conse-
quent mental fatigue, slip into taking seven times seven as
being equal to fifty-six instead of forty-nine. The error would
be duo to the fact that we were simply recalling, without
'perceiving' the truth for ourselves, as when I 'perceive' that
twice one are two. 1 The defect lies not in the knowing act but
in the memory. Very many errors in calculation (not only in
aritlynetic, but also in other spheres) are of this type. They
result from the fact that memory gives us false premisses. And
frequently, in such a case, our conviction that the premisses
imply the conclusion is so strong that we accept the conclusion
as absolutely (and not only provisionally) true, and even feel
convinced about this. If in such a case, however, the conclusion
is'not true, our error can be attributed to a double defect of
memory. For, firstly, we faultily recollect one (or two or many)
of the premisses, and, secondly, we completely forget that
the premisses are unverified. If we remembered the latter
point, our conviction as to the truth of the conclusion would
speedily vanish. Now in such an experience as the above the
defect which produces the error does not lie in the knowing
act, for, though our premisses are false, it still remains true
that they do imply the conclusion. The defect lies in the
memory.

1 This fact is confirmed by our having frequently to 'run over the
whole table* when in doubt. Clearly, we are simply seeking to recall
something previously learnt by rote.

I 4 6 THE NATURE OF KNOWING

But not all human error originates in a defect of memory.
For, as we see it, the type of error which is most prevalent finds
its source elsewhere. It originates in man's impulse to complete
the incomplete an impulse which in itself is perfectly legiti-
mate. Our curiosity is such that we cannot remain satisfied with
part-knowledge. And in our haste to press forward towards
omniscience we frequently mistake something which is not
knowing for knowing. Hence, error becomes possible. Knowing
a part, we 'take' a whole; but, frequently, the 'taking* is not a
knowing, and if we think that it is we fall into error.

In this way we frequently mistake the probable for the
certain. So long as we are clear in our own minds about the
probability of the probable and neither think of it as, nor claim
it to be, certain no error is involved in its assertion. But the
moment we assert, or even implicitly assume, in our haste for
finality, that the probable is certain, then error has already
entered. So much is obvious. (In much the same way, we
also err if we assert that what is really improbable is probable
or that something has a greater or less degree of probability
than it really has.) But how do we come to make the mistake of
supposing the probable to be certain? Now, knowing and
opining, we hold, are two states of mind distinct from each
other, and if sufficient care is taken it is always possible to
distinguish between them. I may be in a state of knowing
something with certainty, as when, for instance, I see that the
premisses of a syllogism imply the conclusion; or again, I may
be in quite a different state of mind, namely, opining, as when I
believe that a conclusion gained inductively is probably true.
The latter is simply a well-grounded opinion, though an opinion
which may be grounded upon much certain knowledge gained
previously. (The better grounded the opinion the more know-
ledge is presupposed in it.) Now it is possible for a man to
ignore the probability of the probable and to believe, for the
time being, that it is certain. It would, no doubt, be an exag-
geration to say that he forces himself to believe this. But he
ignores that which would establish the mere probability of his
belief, and so imagines that he knows. His state, however, is

INTUITIVE APPREHENSION 147

surely different from that of the man who does know and
knows that he knows. The conviction of the latter is not his.
He has just stumbled into a kind of conviction or into belief,
as the result of ignoring certain evidence. And what we suggest
is that his lapse is the consequence of a desire natural to man
for certainty and finality in knowledge. He has opined that
such and such a position is sound and has slipped into the.
belief that his opining is knowing. But his opinion may be un-
sound, as all opinions may. Though he thinks it to be knowing
it is still fallible. Yet this fact, namely, the fallibility of his
experience, cannot be used as an argument to prove that know-
ing as such is fallible, for knowing as such is ex hypothesi some-
thing different from his state. "But/* it may be objected,
"ought we not to face the possibility that we are always in his
state ? -May we not always be in the state of thinking or believ-
ing that we know without really knowing on any single occa-
sion ?" Here, the objector would be admitting the distinction
between the two states, whether there actually exists an instance
of real knowing or not, and this admission is sufficient for the
above argument. For all we wish to maintain is that we cannot
prove the knowing act to be fallacious by saying that sometimes
we err even when we imagine we know. The latter state is not
really an instance of knowing with certainty, but of imagining
that we know with certainty a very different thing. As to the
actual existence of certain knowledge, we assume throughout
this essay that we do sometimes know with certainty, and are
not always, when we claim to know, mistaking a well-grounded
opinion for certain knowledge. We cannot see that any other
answer to the objection is possible. If the objector persists in
doubting the existence of any certain knowledge, nothing more
can be said on that head.

We occasionally then mistake an opining for a knowing, but
the fallibility of what we thus take to be knowing is no argument
for the existence of a like fallibility in knowing itself. That
we can so mistake something else for a knowing, and that we
do so in order to satisfy our desire for complete knowledge,
is confirmed by the attitude of the naive sensationalist. His

i 4 8 THE NATURE OF KNOWING

error is perhaps the deepest and most fundamental of all. He
mistakes not opining, but sensing for knowing. To the real
actually known by him in the sensory experience he applies
the content gained in the mere seeing of the colour, hearing
of the sound, and so on, which so we argue are not in them-
selves instances of knowing at all. And so he senses a 'real*
world, as he thinks, of things having colours, tastes, smells,
and so on. An activity of the mind, namely, the seeing of the
colour, is assumed to be a knowing, when actually it is nothing
of the kind. The mind desirous of a fully-determined and well-
qualified reality applies the content of sensation to the real,
exactly as if seeing the colour were itself a knowing of the
real. In just the same way and for the same ultimate reasons
the man of science may occasionally dress up the general struc-
ture of the real whose nature he has apprehended* in the garb
of imagery and hypothetical conceptions, and fall into believing
that the skeleton so clothed is the fully real, and so forget that
his 'world' is partially true only. In all these instances the error
lies in our tendency to take as knowing what is actually, not
knowing, and the source of the error is our desire for completion
and totality in the objective world. In no case can the error be
said to result from a defect in the knowing act itself. In no case
do we find that the direct apprehension of the real has itsel
given, instead of truth, error.

We have here considered the main types of error. No doubt,
however, there are other types. To be truly exhaustive we
should have to consider every possible instance of error.
Failing this, it would be a good exercise, if we had the space
to spare, to consider in detail each single instance of the 'logical'
and 'material' fallacies set forth in works on formal logic. We
venture the opinion that here again we should never meet with
an instance of a cognitive experience, in which the error could
be attributed directly to the knowing act. Its source would lie
elsewhere, namely, in the whole mental preparation for the
act of knowing. But to carry out even this reduced task would
be to pass beyond the scope of the present essay, and we shall
not attempt it. We have stated earlier, however, that while we

INTUITIVE APPREHENSION 149

admit the fallibility of the cognitive experience taken as a whole,
we cannot admit the fallibility of the knowing act as such.
And our (admittedly incomplete) consideration of the main
types of error certainly substantiates this position. The reader,
as we have suggested earlier, may feel a certain hesitation in
conceding the intuitive character of the knowing act on account
of the patent fallibility of our cognitive experiences. But if he t
now agrees that what applies to the whole cognitive experience
need not and does not, so far as we can see apply to the
knowing act as such, then we shall have done something to
remove his qualms, and he will be in a better position to accept
our general thesis. What we urge is, firstly, that the whole
cognitive experience is not merely knowing, in the strict sense,
it is also a seeing or an opining, a conceiving, an ordering, a
classifying, and so on; and, secondly, that error has its source
not in the knowing act, as such, but in some other part of that
whole experience.

Hence it is quite possible to hold, so far as the present evi-
dence goes, that the knowing act possesses, together with all the
other qualities mentioned, this further quality of infallibility,
and that it is rightly termed an 'intuition' even in this sense.
The error in the whole cognitive experience can be traced, we
believe, to sources other than the functioning of the knowing
act. We have, of course, never tried to prove that the cognitive
experience in man is infallible; such a project could only be
undertaken by a person who very foolishly closes his eyes to
some of the most obvious facts of our finite experience. We
merely make the claim that the term 'intuitive' can be applied
to the act of knowing itself (which is part only of the whole
cognitive experience) and that it can be so applied even though
we recognize that anything which is intuitive must be infallible
and cannot of itself be the source of error. None the less, the
whole cognitive experience is fallible, for it invariably, so far as
we can see, includes within it a preparation of some kind for
knowing, which preparation may be defective, as when our
premisses are false in reasoning. We can find no instance of a
cognitive experience which is simple, in the sense that the

150 THE NATURE OF KNOWING

whole experience consists of knowing (or intuiting) and nothing
else. Therefore, the infallibility of the knowing act cannot in
any way be taken to imply a like infallibility in the whole
cognitive experience. We may err; but the error does not
originate in the knowing act. The latter is in the full sense of
that term an * intuition*.

CONCLUSION

WE are now in a position to draw our conclusions. These are
hypothetical in character; that is to say, we do not wish to
claim that this essay has finally established their truth. We
are content to put them forward almost in the nature of sugges-
tions, and had best present them in the following form: The
problems connected with epistemology are more likely to b*e
solved, we think, if we accept as working hypotheses two posi-
tions. The first, that knowing, as such, is one and the same
throughout, whatever the form of the whole cognitive experi-
ence; the second, that this knowing, identical in character
throughout experience, is best described as an intuitive appre-
hension of the real. We believe that the inquiry, now concluded,
fully .justifies us in making these suggestions.

"In emphasizing the first point, that knowing as such is
identical in character throughout experience, we definitely
deny the existence of so many types or kinds of knowing, each
distinct from the other. The evidence when carefully con-
sidered supports the denial. We do not believe, for instance,
that sensing is one kind of knowing, and that discursive
reasoning is another, whilst intuiting is still a third completely
distinct type. The real differences that exist between these
cognitive experiences do not lie in the knowing as such. They
lie elsewhere. Thus, the affection of the mind in sensation
however it be explained is, we believe, an occasion for the
occurrence of knowledge. But this knowledge, we affirm, is an
intuitive apprehension of the real and does not differ in kind
and character from knowledge on any other occasion. Or again,
we may consider discursive reasoning. By terming a cognitive
experience 'mediate' we convey the suggestion that the knowing
present could not occur without a prior process of some kind,
which liberates the knowing faculty, and enables the mind to
know. Together with this we may also mean that this knowledge
which we term 'mediate* is of an implication, so that v we come
to know indirectly something about the subject of the conclusion
at which we arrive by knowing the implication directly. But the

152 THE NATURE OF KNOWING

actual knowing in discursive reasoning again is direct and not
at all different, so far as we can see, from knowing in the
sensory experience or from any other instance of knowing
wherever it occurs.

But while we emphasize the identity of knowing as such
throughout these cognitive experiences we think it necessary
to recognize a difference in kind between the two experiences of
knowing and opining. Opining is not knowing become vague.
Knowing does not shade off into opining. The difference be-
tween the two is, in our opinion, absolute. When I opine I am
not certain; when I know I am certain. We have not sought
in these pages to give any account of opining, since it w r ould not
be strictly relevant to the matter in hand. Our concern has been
with knowing and not with opining, however well grounded
it be. Unfortunately, however, the human mind can, as we havje
already pointed out, mistake an opining for a knowing. We
can fall into believing that we know when we are only opining.
And this makes our sole criterion in knowing, namely, our 'own
conviction that we are now knowing, untrustworthy. None the
less, no other criterion exists. 1 Our only method of procedure
is to subject our convictions to every possible test, to free our
minds from all prejudices, to be very careful that we have not
mistaken what is not knowing for knowing. And if after every
possible test is made we are still convinced, then we can rest

1 It would be of no avail to say here that coherence or corre-
spondence is a criterion. For what we mean when we make such an
assertion is that when I learn, for instance, that some theory or
other is inconsistent with itself, I know (and am convinced that I
know) that it involves falsehood. But if I am asked how I know this,
I can only answer that I am convinced of it. My ciiterion is my own
conviction. I know that I know. And I make my appeal to another
on the confident assumption that his mind also possesses power to
know, and that he will be as convinced of the impossibility of the
self-contradictory as I am of it. That is to say, my ultimate appeal is
not to the fact that the self-contradictory is impossible, but to the
fact that I am convinced, and that you too, I confidently assume,
will be convinced, that the self-contradictory is impossible. It is this
conviction which is the ultimate criterion, and it alone.

CONCLUSION 153

assured that we are knowing. It may, of course, be said that
we can never feel sure that every test has been tried and that,
therefore, an element of doubt will always remain. But it is our
assumption throughout that knowing does occur. And we
believe that there are experiences where doubt never enters,
however careful we be. Instances are to be found in the mathe-
matical sciences, but are in no way confined to that sphere. W
know the so-called Laws of Thought with complete certainty ;
but better still we frequently see that one thing implies another
beyond the possibility of any doubt. Our suggestion in these
pages is that the conviction which the knowing act brings in
its train is wholly trustworthy; that the untrustworthy con-
viction arises from a mistaking of an opining for a knowing.
And though it is difficult in actual practice to distinguish be-
tween the two, yet the untrustworthiness of the latter cannot
be attributed to the former. Meanwhile our own experience
leads us to assert though it be an assertion without proof
that the former type of conviction does most certainly exist,
that occasionally we do most certainly know in the strictest
sense of that term. 1

In the second place, it has been our purpose in this essay to
describe knowing as accurately as possible. And the conclusion
to which we have come is this one : that if we do wish to de-
cribe knowing in terms other than it itself, that is to say, if
we wish to say something more than merely that knowing is
knowing, just as seeing blue is seeing blue, then the most
appropriate description of knowing, in our opinion, is 'an
intuitive apprehension of the real.' We have already shown why
we use the word 'intuitive* in this connection and we need not
repeat the argument. Also, we have considered the word
'apprehension', and have decided that the analogy which it
suggests is most suitable for expressing the character of this
knowing act. Finally, we have assumed from the outset of this
essay that the object of knowing, when it actually occurs, is

1 We ought, perhaps, to add that where knowing does not occur,
opining is extremely valuable. We made this plain in the third section
of the second chapter.

154 THE NATURE OF KNOWING

the real, what is; and that it is ridiculous to suppose otherwise.
Knowing, therefore, we suggest, is best described as 'an intuitive
apprehension of the real/

As a final word we shall add that our use of the phrase 'a
knowing acf in this essay is also, in our opinion, justified.
The apprehension is an act. By which we mean to convey,
firstly, that knowing as such is not a process. We believe that
to think of knowing as a process is to misconceive its character.
This misconception, we think, is the outcome of confusing a
process of reasoning which may be a necessary preparation for
knowing with the actual knowing itself. The whole cognitive
experience in discursive reasoning, for instance, always involves
a thought-process ; but we cannot see that the knowing itself is
ever a process. It is a simple act. In the second place, we think
the term 'act' justified because it conveys the further meaning
that the knowing is an actualization of a capacity. We possess
throughout the power to know, but, on occasion, in the right
circumstances, this potentiality is actualized. In this sense,
again, knowing is an act. Finally, in the third place, we have
been trying to confine our attention, so far as was possible, to
the subjective side of the knowing experience, and to the actual
knowing rather than to the object, or to the whole subject-
object relation. And the use of the word 'act' tends, we think,
to keep this fact before the reader. The subject of our inquiry
throughout has been that mental functioning which is knowing
and which we now think best to describe as the intuitive
apprehension of the real.