CO

».
[I

LIL.

PHYSICAL EEALISM

1'IUNTKD BY

SrOTTLSWOODE AND CO., NEW-STUEET SQUAUB
LONDON

PHYSICAL REALISM

J5KING

AN ANALYTICAL PHILOSOPHY

FROM

THE PHYSICAL OBJECTS OF SCIENCE

TO

THE PHYSICAL DATA OF SENSE

BY

THOMAS CASE, M.A.

AXU SENIOR TUTOH CORPUS CHRIST! COI.LKCJE, AND LECTUUBR AT
CHKIST CHL'IICH; FOUMKHLY FEIJ.UW OK BKAHBXOSE AND
TUTOIl OK BALLIOL COLLEGE, OXFOKD

drrodflgui A«y/£rt*caw— AKISTOTLE

LONDON
L 0 X G 11 A X S, GHEE X, A X I) C 0.

AND NE\y YOliK : 15 EAST 10"' STKEET
1888 ,

All right. 3 ff*rrrr,l

' Neque tamen illis nihil adtli posse affirmamus : sed contra,
nos, qui Mentem respicimus non tantuin in facultate propria,
sed quatenus copulatur cum Rebus, Artem Inveniendi cum
Inventis adolescere posse, statuere debemus.'

BACON, Nov. Org. i. 130.

TO

WILLIAM S. SAYOBY, F.E.S.

SURGEON -EXTRAORDINARY TO H.M. THE QUEEN

PRESIDENT Of THE ROYAL COLLEGE OF SURGEONS OF ENGLAND
SURGEON TO ST BARTHOLOMEW'S HOSPITAL

CONTENTS.

PART I.
GENERAL PROOF OF PHYSICAL REALISM.

CHAP.

I. THE PHYSICAL OBJECT OF SCIENCE . . 3

II. IDEALISM AND REALISM . ]3

III. THE PHYSICAL DATA OP SENSE . 40

IV. THE HISTORICAL ORIGIN OF PSYCHOLOGICAL IDEALISM 82

PART 11.
PSYCHOLOGICAL IDEALISM.

V. DESCARTES JQJ

VI. LOCKE J41

VII. BERKELEY .... |fl(j

vin. BERKELEY'S THEORY OF VISION . 225

IX. HUME Ofj(j

x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS . . . .°,19

APPENDIX QQQ

PART I.

GENERAL PROOF OF PHYSICAL REALISM.

' Itaqiie contemplatio fere desinit cum aspcctu ; adeo ut rerum
invisibilium cxigua aut nulla sit observation

BACON, Nov. Org. i. 50.

CHAPTEE I.

THE PHYSICAL OBJECT OF SCIENCE.

' NATURAL PHILOSOPHY, as now regarded, treats generally
of the physical universe, and deals fearlessly alike with
quantities too great to be distinctly conceived, and with
quantities almost infinitely too small to be perceived
even with the most powerful microscopes ; such as, for
instance, distances through which the light of stars or
nebulas, though moving at the rate of about 186,000
miles per second, takes many years to travel ; or the
size of the particles of water, whose number in a single
drop may, as we have reason to believe, amount to
somewhere about 102G, or

100,000,000,000,000,000,000,000,000.

Yet we successfully inquire not only into the composi
tion of the atmospheres of these distant stars, but into
the number and properties of these water-particles ; nay,
even into the laws by which they act upon one another.'
This quotation from Professor Tait's ' Recent Ad
vances in Physical Science ' is a recognition of the
reality of the insensible, and of its knowledge by the
natural philosopher, as facts. No metaphysical theory
of existence can be complete, unless it recognises the
known reality of the insensible physical world ; and
no psychological theory of human knowledge can be
accepted as even a probable hypothesis, unless it

B 2

4 PHYSICAL REALISM PART i.

explains how these scientific objects of human know
ledge are known from the original data of sense.

The distinction between the sensible and the scien
tific-, the apparent and the real, the perceptible and the
imperceptible, is not only a scientific fact but has be
come a commonplace in natural philosophy, without
having produced any marked effect in mental philo
sophy. Astronomy has long opposed the real to the
apparent motions of celestial bodies ; and Sir Isaac
Newton carried this contrast so far as to oppose abso
lute, true and mathematical, to relative, apparent and
common, time and space. In physics, apparent size is
the room which a body seems to occupy, physical size
is the real space taken up by its particles. Not only
physics, but chemistry and biology unite in the anti
thesis of molar and molecular motion, in recognising
therefore motions which are for the most part imper
ceptible, in resolving what seem to our senses to be
heterogeneous qualities into mere varieties of imper
ceptible motion, and in referring these motions to
particles which are as imperceptible as the motions
themselves. In all these sciences the latent structures
and processes of things are opposed to their external
appearances and perceptible changes.

I do not mean that these undeniable conclusions,
very far removed as they are from the original data of
observation and experiment, are at all inconsistent with
the sensations, perceptions, observations, or experiences
which ordinary men have, and from which the natural
philosopher starts. On the contrary, the very untutored
senses themselves are best explained — nay, can be only
explained — by statements at first sight opposed to
them. It is only in appearance that the motion of the
earth round the sun contradicts our senses, for, though

ciiAr. i. THE PHYSICAL OBJECT OF SCIENCE 0

it contradicts one single appearance, the whole sum of
astronomical observations is only to be explained by
means of it. Similarly, when it is said that one thing is
apparently larger and physically smaller than another,
vision is contradicted, but the sense of touch is justified,
and our experience as a whole explained. The latent
motions of particles, into which sensible qualities are
resolved, at first sight contradict but really explain
the whole system of our sensations of touch, vision,
and hearing.

But though the results of science thus explain the
data of sense, it must be remembered that they only
explain them, and are not themselves data of sense.
No man can make himself see the earth going round
the sun, except by standing on the sun itself. No man
can see light at the moment when it starts from a
distant star years before it reaches his senses. Micro
scopes can be multiplied in power, but they are
millions short of the actual (I do not speak of the
potential) divisibility of the particles of things.

Moreover, the natural philosopher gives even greater
reality to the imperceptible than to the perceptible.
The astronomer not only opposes but prefers real to
apparent motion, the physicist physical to apparent
size, and all natural philosophers latent structures and
molecular processes to masses and their molar motions.
It is not too much to say, that the mission of modern
as well as of ancient philosophy is to convince mankind
that sense is unequal to the subtlety of things ; to get
behind the scenes and see the machinery of nature at
work ; to recognise the insensible as real, yes, and more
real, than the sensible. Sense is not science.

Our knowledge is not limited to sensible pheno
mena. We are quite as certain of the existence of

0 PHYSICAL REALISM PART i.

that which cannot be brought within our sensibility
as of that which can, and of objects which we do
not experience as of objects of experience itself.
Further, we are quite as certain that they exist in
space and in time ; for if they are not in space they
have no size, if they are not in time they have no dura
tion, and that which has neither any size nor any dura
tion is nothing ; and, if they are neither in time nor
space, they do not move, for motion is change of
place in space during time. Space and time are not
mere forms of our sensibility, but conditions of things
and their motions beyond the range of our sensibility.

We not only know that the imperceptible exists, and
that it exists in space and time, but also we know im
perceptible attributes both of the perceptible and of the
imperceptible. For example, I know that the hour-hand
of my watch moves, though I cannot perceive it moving,
as well as that the minute-hand moves which I can per
ceive moving with difficulty, or the second-hand which
I can perceive moving with ease. I know that the im
perceptible particles of matter gravitate imperceptibly
towards one another, as well as I know that their masses
gravitate, and that unless gravitation is true of the
former, it is not true of the latter. Still more insensible
are cohesion and chemical affinity, which are imper
ceptible motions exerted between imperceptible particles
and at imperceptible distances. The whole of modern
science is based on the fact that there are numerous
latent structures and latent processes which are known
to be real attributes of particles themselves latent. He,
then, who will venture to assert, as mental philosophers
often do assert, that the attributes which we ascribe to
things are simply the phenomena or the sensations
which they cause in us, must be prepared to deny all

CHAP. i. THE PHYSICAL OBJECT OF SCIENCE 7

the imperceptible structures and motions which are
recognised as attributes of things in natural philosophy.
Natural philosophy does not stop at the reality and
knowledge of imperceptible things and their imper
ceptible attributes. It takes one step further : it regards
the imperceptible as not only real but causal. In the
first place, among imperceptible objects there are latent
processes of cause and effect, no part of which can be
represented by a sensible object. When, for example,
the physicist declares that the medium called aether
remains fixed in space, while each successive part of it
undulates in consequence of the previous undulation of
another part, in the same manner as water communi
cates successive waves, he affirms that the whole of this
propagation of undulations through aether is real, though
the whole of it is imperceptible. Secondly, he affirms
still more ; he affirms that the imperceptible undula
tions not only cause one another, but finally cause our
sensations of light. In this instance of light, as well
as in the parallel case of heat, natural philosophy un
hesitatingly accepts the conclusion that imperceptible
motions of imperceptible things not only exist but cause
our sensations. In other words, secondary qualities as
existing in nature are insensible primary qualities which
are causes of secondary qualities, as sensible in us.

Natural philosophy is not a sham. One or other,
or many, of its propositions, may be untrue. But its
whole fabric of the physical, but insensible, world
which causes the sensible image of it to arise in us,
cannot be an invention. There is a thing beyond sense,
a reality beyond phenomena, not only actual in nature,
but known to science. There is a thing real and known
which is not a sensible phenomenon, because such
things as imperceptible particles are known really to

PHYSICAL REALISM PART i.

exist, though they are incapable of becoming sensible.
There are attributes real and known which belong to
this thing, but are not sensations or sensible phe
nomena, because such attributes as the imperceptible
motions of imperceptible particles are known really to
take place, although they are not capable of becoming
sensible. Finally, these real things by these real
attributes are real and known causes of human sensa
tions because the imperceptible motions of the imper
ceptible are known really to cause sensations of light
and other sensations in men, although the latent pro
cess, by which an imperceptible motion such as the
undulation of ether produces sensible light, is totally
beyond the reach of sense, which perceives not the
undulation but the sensible result. Thus real things
and real attributes transcending yet really causing sensa
tions are, in some way or other, known to the natural
philosopher. The insensible, then, is not a simple
reality, but contains three realities, all insensible : real
substances, real attributes, real causes of sensations.

There are things in themselves. A thing in itself
might mean a thing out of all relations. In this sense
nature contains no things in themselves ; it is a system
of related things the universe of which is alone out of
relation as the sum of all relations. But this is not
what is meant by a thing in itself in philosophy : what
is really meant is not a thing out of all relations, but a
thing distinct from the phenomena it causes in us, a /
thing in itself as opposed to its sensible appearance. In
this meaning, nature contains infinitely more things in
themselves than it contains phenomena ; and man, as a
natural philosopher, knows things in themselves which
are not phenomena, when he knows imperceptible
particles ; knows not merely the phenomena which

CHAP. I.

THE PHYSICAL OBJECT OF SCIENCE (J

they cause in us, but their real attributes, when he
knows imperceptible motions, and knows that the
thing in itself, not as an ' unknown cause,' but by its
real attributes produces phenomena, when he knows
that imperceptible things, by their imperceptible
motions, cause human sensations. There are real things
known, real attributes known, real causes known,
beyond the phenomena of sense. All this knowledge
does man as a natural philosopher possess of things in
themselves.

Two antitheses have been handed down to us from
ancient philosophy, the natural and the supernatural,
the visible and the invisible. These distinctions are
often treated as convertible ; but they are not so. The
natural and the visible are not identical ; and the super
natural and the invisible are not identical : there is a
natural yet invisible world. Between the extremes of
visible nature and the invisible supernatural world
there is an invisible nature, distinct from both ; a world
which is neither in heaven nor in man, but in itself.
If we combine both the antitheses, they cease to be
double, and form this triple division :—

1. The natural and visible, e.g. sensible phenomena.

2. The natural and invisible, e.g. insensible bodies
and imperceptible particles.

3. The supernatural and invisible, e.g. God.
Natural philosophy is the science of nature visible

and invisible. From the former it infers the latter.
But it stops at nature. So far as it is the science of an
invisible nature, it is a philosophy of the suprasensible,
not a theology of the supernatural. It outruns sense,
but walks with reason to knowledge, without flying to
faith. That we know invisible nature beyond sense in
natural philosophy is a simple fact, explicable by logical

10 PHYSICAL REALISM PART i.

reasoning from sense. Can we in theology further know
the invisible beyond nature as well as beyond sense ?
Can we know the supernatural world and God by reason
ing from sense ? These are questions beyond natural
philosophy. But the theologian may be sure that, on
the one hand, unless we can vindicate our knowledge
of insensible nature, we can hardly hope for a know
ledge of an insensible world beyond nature ; and that,
on the other hand, reasoning from sense to nature
encourages reasoning from nature to God. Natural
philosophy is the first step beyond sense into the unseen
world, within which natural theology soars heaven
wards to tell

Of things invisible to mortal sight.

I will conclude this chapter by quoting, from Sir
John Herschel's ' Discourse on Natural Philosophy,' a
passage which is sufficiently near to the existing state
of science for our present purpose. Its value is that it
groups together a number of scientific conclusions,
which, as it seems to me, cannot be explained by any
theory of reality except realism, or the theory that
there is a real and known world beyond phenomena,
or by any process of knowledge except syllogism, or
deductive inference which carries reason beyond sense.

' What mere assertion will make any man believe,
that in one second of time, in one beat of the pendulum
of a clock, a ray of light travels over 192,000 miles,
and would therefore perform the tour of the world in
about the same time it requires to wink with our eve-
lids, and in much less than a swift runner occupies" in
taking a single stride ? What mortal can be made to
believe, without demonstration, that the sun is almost a
million times larger than the earth ; and that, although

CHAP. i. THE PHYSICAL OBJECT OF SCIENCE 1 1

so remote from us that a cannon-ball shot directly
towards it, and maintaining its full speed, would be
twenty years in reaching it, it yet affects the earth by
its attraction in an inappreciable instant of time ? a
closeness of union of which we can form but a feeble
and totally inadequate idea, by comparing it to any
material connection ; since the communication of an
impulse to such a distance, by any solid intermedium
we are acquainted with, would require, not moments,
but whole years. And when with pain and difficulty
we have strained our imagination to conceive a distance
so vast, a force so intense and penetrating, if we are
told that the one dwindles to an insensible point, and
the other is unfelt at the nearest of the fixed stars, from
the mere effect of their remoteness, while among those
very stars are some whose actual splendour exceeds by
many hundred times that of the sun itself, although we
may not deny the truth of the assertion, we cannot but
feel the keenest curiosity to know how such things were
made out.

' The foregoing are amongst those results of scientific
research which, by their magnitude, seem to transcend
oar power of conception. There are others again,
which, from their minuteness, would elude the grasp of
thought, much more of distinct and accurate measure
ment. Who would not ask for demonstration, when
told that a gnat's wing in its ordinary flight beats many
hundred times in a second? or that there exist ani
mated and regularly organised beings, many thousands
of whose bodies laid close together would not extend
an inch ? But what are these to the astonishing truths
which optical inquiries have disclosed, which teach us
that every point of a medium through which a ray of
light passes is affected with a succession of periodical

1'2 PHYSICAL REALISM PART i.

movements, regularly recurring at equal intervals, no
less than five hundred millions of millions of times in a
single second ; that it is by such movements, communi
cated to the nerves of our eyes, that we see — nay, more,
that it is the difference in the frequency of the recur
rence which affects us with the sense of the diversity of
colour ; that, for instance, in acquiring the sensation of
redness our eyes are affected four hundred and ei^htv-

»/ o «/

two millions of millions of times ; of yellowness, five
hundred and forty-two millions of millions of times ;
and of violet, seven hundred and seven millions of
millions of times ? Do not such things sound more like
the ravings of madmen, than the sober conclusions of
people in their waking senses ?

' They are, nevertheless, conclusions to which any
one may most certainly arrive, who will only be at the
trouble of examining the chain of reasoning by which
they have been deduced ; but, in order to do this,
something beyond the mere elements of abstract science
is required. Waiving, however, such instances as these,
which, after all, are rather calculated to surprise and
astound, than for any other purpose, it must be ob
served that it is not possible to satisfy ourselves com
pletely that we have arrived at a true statement of any
law of nature, until, setting out from such statement,
and making it a foundation of reasoning, we can show,
by strict argument, that the facts observed must follow
from it as necessary logical consequences, and this not
vaguely and generally, but with all possible precision
in time, place, weight, and measure.'

CHAT. IJ.

CHAPTEE II.

IDEALISM AND REALISM.

THE problem of this essay is to use the insensible
world of science as a fact from which to find the nature
and origin of knowledge. Science is systematic know
ledge. Yet the mental philosopher usually contents
himself with endeavouring to explain ordinary know
ledge. If he is a mental physiologist, it is true, he
also uses natural science to proceed from the organs to
the functions of sense. But there is another use of
natural science to mental philosophy, which has been
too much neglected : the objects of science are as
important as the bodily organs to the explanation of
knowledge. Natural science should be used to ascer
tain what we know as well as how we know it. More
over, the insensible physical world of the natural
philosopher ought to prove to the mental philosopher
that neither all knowable objects nor all sensible data
are psychical, but some are physical. I purpose to
show that physical objects of science, being objects of
knowledge, require physical data of sense. Hence this
essay is called Physical Eealism.

We must confront natural with mental philosophy.
The former has outstripped the latter. Natural philoso
phers have long ago discovered to a great extent how
physical nature is the causa essendi of sensible data ;
but mental philosophers have failed altogether to show

14 PHYSICAL REALISM PART i.

how sensible data are the causa cognoscendi of physical
nature. The reason is, the data are mainly unknown.
The existing hypotheses of the origin of knowledge do
not explain the facts of science, and too often end "by
denying what they fail to explain. Especially to blame
is the hypothesis that all the data of sense are psychical
facts, such as sensations and ideas, from which there is
no way to insensible but physical objects of scientific
knowledge. This vicious hypothesis is psychological
idealism. Hence this essay is designed to combat
psychological idealism by means of physical realism,
and to appeal from the hypothesis of psychical data to
the physical objects of science. The physical world of
science cannot be explained by the common hypothesis
that all sensible data are psychical, nor without the more
moderate hypothesis that some are physical.

The motto of all idealism is ideate prius reale
posterius. But it has many meanings. Anaxagoras
founded philosophical idealism by the proposition that
the Divine Intelligence is prior to the order of nature ;
and in adding that soul is also prior to body Plato
became its second founder. The Cartesian idealism
means that knowledge begins with psychical ideas, and
the Kantian idealism that it adds a priori mental ele
ments. Of these idealisms two are of supereminent
importance in the history of thought ; that which places
God at the beginning of the world, and that which
places psychical ideas at the beginning of knowledge.
The former is the belief of the majority of mankind, the
latter of most philosophers since Descartes. The former-
is theological, the latter psychological idealism.

Theological and psychological idealism are not
necessarily connected. A philosopher may hold that
God causes physical nature and man apprehends it.

CHAP. Ti. IDEALISM AND REALISM 15

He may be theologically an idealist, psychologically a
realist. On the other hand, he may suppose that all
sensible data are psychical facts, and yet doubt the
existence of God. He may be psychologically an
idealist, theologically an atheist. The founders of
natural theology had no thought of making psychical
facts the beginnings of human knowledge. The
followers of Hume hardly consider themselves supporters
of the doctrine that God created the world. These
distinctions are of importance, because there is a crude
notion in our times that idealism in mental philosophy is
necessary to theology. They are of special bearing on
the scope of this essay, which is aimed, not at theo
logical, but solely at psychological idealism.

Psychological idealism began with the supposition of
Descartes that all the immediate objects of knowledge
are ideas. From Descartes it passed to Locke and
Berkeley. But with Hume it changed its terms from
ideas to impressions. Kant preferred phenomena,
Mill sensations. The most usual terms of the present
day are sensations, feelings, psychical phenomena, and
states of consciousness. But the hypothesis has not
changed its essence, though the idealists have changed
their terms, — Verbum, non animum, mutant. They at
least agree that all sensible data are psychical objects
of some kind or other.

The psychological idealists differ widely about the
origin of knowledge from these psychical data. Some
of them hold that there are a priori elements contributed
by mind to the psychical data of sense, others that these
supposed elements are a posteriori. But this difference
about the origin does not prevent them from agreeing
about the object of sense, which they alike hold to
be some kind of psychical fact, whether idea, im-

16 PHYSICAL, REALISM PART i.

pression, phenomenon, sensation, feeling or state of
consciousness.

There is a further difference among the idealists.
Some of them, beginning with Descartes, believe that,
though the immediate objects of sense are psychical,
reality also includes physical facts. Others, beginning
with Berkeley, reply that psychical data cannot yield
physical objects, and therefore the psychical is all that
is known to be real. The former divide reality into the
psychical and the physical, the latter resolve it wholly
into the psychical. The former have been called
Cosmothetic Idealists, and the latter Absolute or Pure
Idealists. But, while they differ only about the objects
which can be mediately known, they still agree about
the immediate data. Starting from the common hypo
thesis that all sensible data are psychical, the cosmo-
thetic idealist nevertheless believes in physical realities,
but the absolute idealist denies or doubts them.

Cosmothetic idealists further differ among themselves
about the physical world. Descartes held that a physical
world can be known through the medium of ideas ;
Locke, in one of his many moods, that it is a cause
of ideas, but unknown. This difference is important,
because cosmothetic idealism is the usual view of men
tal physiology in our own time, and it is held in both
forms. Mental physiologists have unwarily received
from psychologists the hypothesis of psychical data,
which they usually call sensations, and have at the
same time learnt from nature that the data of sense
are effects of physical structures and motions beyond
sense. Hence they are cosmothetic idealists. But
according as they are rather physiologists or rather
psychologists, they lean to Descartes or to Locke. The
former hold that, starting from psychical sensations as

CHAP. IT IDEALISM AND REALISM 17

data, by inference we know their physical causes ; the
latter, that the psychical sensations are produced by the
physical causes, which are nevertheless unknown and
unknowable. Their differences, however, do not dis
turb the consensus that the immediate objects of sense
are not physical, but purely psychical.

It may be thought that this consensus of idealism
is a proof of truth. But agreement is one of the chief
causes of human error, because it tempts men to dis
pense with further consideration of the question. More
over, we shall find that the inconsiderate assent to this
common proposition is the very reason why opposite
schools of idealists cannot conclusively answer one
another. Lastly, there are two kinds of consensus :
one, assent to a self-evident principle, such as 1-f 1 = 2 ;
the other, agreement in a common hypothesis. Now
the proposition that all sensible data are psychical
phenomena is not a self-evident principle, but a de
batable hypothesis.

Eealism is the philosophy of a reality beyond psy
chical facts. The earliest form in which it was a
conscious doctrine was the belief in the reality of
universals. Plato thought that there were universal
forms existing in themselves, incorporeal and super
natural archetypes, in accordance with which similar
individuals are produced in nature. Aristotle agreed
that there are real universal forms, and even that they
are incorporeal substances. He contended, however,
that they exist not in themselves but only^as belonging
to individual substancesj_which are concretions of matter
and form. TrT the Middle Ages the disciples of Plato
and Aristotle were called Eeales, to distinguish them
from the Nominales, who either contended that uni
versals were merely general names, or else general

c

18 PHYSICAL REALISM PART i.

conceptions. Those who adopted the latter view were
afterwards called Conceptualists.

It is not necessary to be either a Flatonist or an
Aristotelian. There is a third realism of universals
possible ; and that, too, without falling into nominalism
or conceptualism. The theory of the reality of univer
sals, though overlaid with many errors, contains two
important truths. The first is, that science knows of
classes which, have an indefinite number of similarities,
such as triangles, colours, and living beings. The
second is, that of these similarities some are fundamental,
others derivative ; e.g. three-sided rectilineal figure is
the foundation of innumerable other similarities of tri
angle ; undulations of ether produce the facts of colour,
metabolism is the basis of the facts of life. The first
truth shows that a natural class, or real kind, is not a
name, nor a notion, but a real sum of individuals form
ing an indefinite number of similarities. The second
truth shows that the distinction between essence and
property is not a nominal difference depending on the
meaning of a name, nor a notional difference depending
on the analysis of a notion, but a real distinction depend
ing on the fundamental character of the similarities,
on which the rest depend. Without natural classes,
whose similarities can be expressed in laws, there would
be no science ; and without essences, or fundamental
similarities of those natural classes on which other
similarities depend, we could not have the mathematics
of the triangle referring its propositions back to its
being a three-sided figure, nor the physics of light?
referring all the facts of colour back to the undulation
of rcther.

A natural class, then, is the sum of individuals
possessing an indefinite number of similarities. A real

CHAP. IT. IDEALISM AND REALISM 19

essence is the fundamental similarities of the individuals
of a natural class. It is easy to make too much of it
or too little. If we follow the nominalist, and make
sethereal undulation the meaning of the name 'light,' or
the conceptualist, and make it the analysis of the
notion, we make too little of it, because the undulation
of gether began before, goes on without, and will last
after, our names and notions. If, on the other hand,
we follow Aristotle, and make it an incorporeal sub
stance coexisting with matter, we make too much of it,
because it is only a motion of matter after all ; while,
if we try to soar with Plato into the supernatural world
and make it a heavenly archetype of earthly light,
we fail to explain the facts and desert science for
mysticism.

The realism of universals, however, is not the
business of this essay. There is another meaning of
realism, which we may call the Eealism of Individuals.
This is the theory that there is a physical world of
individuals beyond psychical sensations and ideas. It
may be held with any theory of universals ; the realist
of individuals is not necessarily a realist of universals.
It is also a later product. The realism of universals is
rather a doctrine of ancient, the realism of individuals
rather of modern, philosophers. Not that Aristotle
rejected the distinct reality of physical individuals ;
but it never occurred to him that it needed to be
proved. There was, as Brandis remarked, an uncon
scious realism in ancient philosophy. It seldom
doubted a world beyond the psychical; the question
was rather whether there were not three worlds; natural
individuals, supernatural universals, and psychical in
telligences. But in modern times the development of
psychological idealism has brought even the physical

c 2

20 PHYSICAL REALISM PART i.

world of individuals into question. In opposition to
this psychological idealism a conscious realism has
arisen, the object of which is to show that there are
physical things beyond psychical facts. This realism
of physical individuals is part of the business of this
essay, and for shortness will in the sequel be called
simply Realism.

Realism is constantly misunderstood. It is some
times supposed to be a synonym for mere Sensualism,
or the belief that physical things are as they appear to
our senses. But sensualism is only a crude form of
realism. There is a realism which goes beyond sense to
science, and holds that things are not as they imme
diately appear to sense, but rather as they are mediately
inferred by science. A more serious misunderstanding
is the confusion of realism with Materialism. Material
ism is a kind of realism ; it is also more. It is a double
hypothesis : first, that there are physical things ; secondly,
that they are either the only realities, or at least are
prior to psychical realities, whether in nature or in man.

Only the first part of this hypothesis is essential to
realism ; the second part, which contains, too, the real
sting of the materialist, is unnecessary to the realist.
A man ceases to be a materialist, but he remains a
realist, if he holds that God is the Creator and
Governor of the world, while the world is not a
psychical fact of God's Intelligence but a physical effort
of His Intelligent Will ; and that nature is posterior to
God though prior to man. The motto of materialism is,
reale prius ideale posterius : the motto of realism is reale
non est ideale. In short, it is one thing to affirm a
natural world of individual objects beyond sense,
another thing to deny a supernatural world beyond
nature.

CHAP. II.

IDEALISM AND REALISM 21

Hence realism is not the exact contrary of all
idealism. It is not opposed at all to the idealism of
natural theology. It is not even the direct contrary
to all psychological idealism. Idealism centres itself
on the data, realism on the objects of knowledge. The
former says that all sensible data are psychical, the
latter that some objects are physical. Hence a difficulty
in contrasting them, and even in keeping them distinct.
Some idealists, as we have seen, though they regard all
data as psychical, admit the independent reality of
physical objects. As Hamilton has pointed out, the
cosmothetic idealists are also hypothetical, or, as some
would say, transfigured realists. The exact contrary
of realism is not all idealism but pure or absolute
idealism. The pure or absolute idealist denies the
reality of aught beyond the psychical world, the realist
affirms the reality of the physical.

At the same time realism is not a single body of
doctrines. Eealists agree only in one position — the
reality of physical things. In the foundations of that
position, in the sensible data of knowledge, they differ
toto ccelo. It is, therefore, necessary to classify them to
prevent confusion, and that sort ofignoratio elenchi,wbich
idealism and realism alike have to suffer from their
opponents when they are not properly defined.

Of the realism of individuals there are two species
recognised among modern philosophers — the Hypo
thetical Eealism of the cosmothetic idealists, and the
Intuitive or Natural Eealism of the Scotch philosophers,
Eeid, Stewart, and Hamilton. Agreeing about know-
able objects, hypothetical and intuitional realists differ
about the data of sense. According to the former, the
data are psychical ideas or sensations of the ego ; ac
cording to the latter, they include the primary qualities

PHYSICAL HEALISM

TAUT I.

of the physical non-ego. Agreeing in a physical world,
they differ about the way in which it is to be reached,
the former holding that it is inferred from psychical
data, the latter, that it is immediately perceived. Hypo
thetical or transfigured realism is the hypothesis that
our senses present psychical ideas or sensations repre
senting external physical objects ; intuitive or natural
realism, the hypothesis that the senses present the pri
mary qualities of external physical objects themselves.

Modern philosophy exhibits a constant oscillation
between the opposite poles of the ego and the non-ego ;
and the two received kinds of realism are opposite cur
rents in this oscillation. The cosmothetic idealist or
hypothetical realist, learning from natural philosophy
that his senses do not directly perceive external things,
takes refuge in the psychical world of his own soul.
Dissatisfied with this alternative, and conscious that
he somehow apprehends something physical, the in
tuitional realist flies forward to the direct perception of
an external world. Extreme views are usually as untrue
as extreme measures are dangerous. Is there a via
media ? I venture to propose a new Eealism.

When I consider the objects of science, I am struck
by the enormous number of things and attributes
entirely beyond the reach of sense and not even
corresponding to any sensible object, I refer, espe
cially, to corpuscles, their structures and motions.
Secondly, on going further, I find that the whole ex
ternal world has been discovered by sciences, such as
optics, acoustics, and biology, to be insensible, and
that nothing is sensible except what has been impressed
on the body, and in the body on the nervous system,
of a sentient being. Thirdly, I notice that a connection
has been scientifically established between external in-

CHAP. II.

IDEALISM AND REALISM 23

sensible objects and the objects of which I am sensible.
The former are causes of the latter. They are also
found to resemble one another in primary qualities,
such as duration, extension, motion, but not in secondary
qualities, such as light, heat, and -sound; for the se
condary qualities, as they are in external nature, are
found by corpuscular science to be insensible modes
of primary qualities ; light, heat, and sound being all
insensible modes of motion producing a heterogeneous
effect on the senses.

I cannot believe that this whole fabric of insensible
objects can be scientific, yet unknown. But it must be
either physical or psychical. If the objects are psy
chical, they are either sensations or ideas. But they
are insensible and often inconceivable. Now what is
insensible cannot be a sensation, and what is incon
ceivable cannot be an idea. Not all objects of science,
then, are either sensations or ideas ; therefore they are
not psychical objects at all. It remains that they are
physical objects.

Again, I cannot believe that this whole fabric of
physical objects of science can have been inferred
without sufficient data of sense. I therefore proceed
to inquire what data of sense are required to infer a
physical object of science. This is a question of logic.
Now the rules of logic teach me that whatever is inferred
is inferred from similar data. If I infer that all men
will die, it is because similar men have died. Now, as
we have seen, physical objects are scientifically inferred
from sensible data. It follows that the sensible objects,
which are these data, must also be physical. The
similar can be inferred only from the similar, therefore
the physical can be inferred only from the physical.
This conclusion, however, places me in a dilemma.

24 PHYSICAL REALISM

PART

Science shows me that the object of sense is internal,
logic that it is physical. The former evidence might
incline me to cosmothetic idealism, the latter to intui
tive realism. Which shall I prefer ? Am I to say that
the sensible data are psychical objects within me ? No,
because I require physical data of sense to infer
physical objects of science. Am I to say that the
sensible data are physical objects without me? No,
because no external object is sensible. I can be neither
a cosmothetic idealist, because of logic, nor an intuitive
realist, because of natural science.

If, then, natural science requires that the object of
sense must be within my nervous system in order to be
sensible, and logic that it must be physical in order to
infer physical objects of science in the external world,
how can the sensible object be at once physical and
internal ? I answer, it is the nervous system itself
sensibly affected. The hot felt is the tactile nerves
heated, the white seen is the optic nerves so coloured.
The sensible object must be distinguished from its
external cause on the one hand, and on the other hand
from the internal operation of apprehending it : it is
the intermediate effect in the nerves produced by the
external cause, and apprehended by the operation of
sensation. In particular, the operation and the object
of sensation must not be confused, because the former
may be psychical, the latter is physical. There is some
plausibility in saying that the act of consciously touch
ing is psychical, there is none at all in saying that the
hot felt is psychical. Non sequitur. Vision may be a
psychical sensation, but the white seen is a physical
object. Nor is there any reason why a psychical opera
tion should not apprehend a physical object. The sen
sible object then is identical neither with the external

CHAP. ii. IDEALISM AND REALISM 25

cause nor with the internal operation of sensation. It
is the effect in the nervous system produced by the one
and apprehended by the other. For example, the hot
felt and the white seen are produced by external objects
and are apprehended by internal sensations of touch
and vision, but are themselves respectively the tactile
and the optic nerves sensibly affected in the manner
apprehended as hot and white.

From such sensible data, internal, as science re
quires, and physical, as logic requires, man infers
physical objects in the external world by parity of
reasoning. Men in general begin by inferring that
physical objects of sense are produced by physical
causes exactly similar. Thus from the hot within we
infer a fire without. Such objects, directly inferred to
correspond with sensible data, may be called the
originals represented by them. They are inferred,
but are generally said to be perceived ; thus we speak
of perceiving the fire though we only infer it. We
may, perhaps, say then that the originals of the sensible
are insensible objects inferentially perceptible.

Afterwards, scientific men carry on this parity of
reasoning, and infer that these originals beyond sense
consist of further insensible particles similar to the
originals, but not at all represented by sensible data ;
and that many other objects, such, for example, as the
side of the moon always turned from the earth, are
incapable of producing sensible objects in us. These
unrepresented objects may be said to be not only in
sensible but imperceptible, and are objects of an infer
ence which may be called transcendental, in the sense
of transcending both sensitive and inferential perception.

Lastly, science also finds that in another direction
the ordinary man has carried his inferences from

20 PHYSICAL REALISM

PA.RT I.

similar data to similar objects too fur. Physical objects
are found to be like sensible in their primary, not in
their secondary qualities ; for instance, external motion
is like sensible motion, but external heat is an imper
ceptible mode of motion while sensible heat is not
sensibly a motion at all. How is -this inferred?
Because, though at first sight sensible heat would
demand a similar external object, when all the facts
of sensible heat are accumulated they are found to be
the kind of facts that are only produced by motion.
Hence from sensible physical data we scientifically
infer insensible physical objects, like sensible objects in
primary but unlike in secondary qualities.

Such is the realism proposed in this essay. It may
be expressed in two propositions : there are physical
objects of science in the external world ; therefore
there are, as data to infer them, physical objects of
sense in the internal nervous system. It is a via media
between intuitive realism and the hypothetical realism
of the cosmothetic idealist. As it recognises physical
realities, it is realism. As the objects, which it sup
poses to be sensible, are not external but internal, it is
not intuitive realism. As the objects of sense, which
it supposes to be the data of inferring an external
physical world, are not psychical but physical, it is
not hypothetical realism. As they are physical data
within, to infer physical objects without, the realism
which I advocate may be called Physical Eealism.

There are three realistic ways of explaining our
knowledge of an external physical world. The first is
cosmothetic idealism, which supposes that we are sen
sible of a psychical, but infer a physical world. This
is against logic, which shows that all inference is by
similarity. The second is intuitive realism, which

CIIAr. IT.

IDEALISM AND REALISM 27

supposes that we directly perceive an external physical
world. This is against natural philosophy, which shows
that we perceive nothing directly but what is propagated
into our nervous system. The third is physical realism,
which supposes that we sensibly perceive an internal
but physical world, from which we infer an external
and physical world. This agrees with both natural
philosophy and logic.

Physical Eealism must be especially distinguished
from intuitive, or, as it is also called, natural realism.
It is true that the theories have some common points.
This essay owes to Eeid the instructive remark on the
' Sentiments of Bishop Berkeley,' that there is no evi
dence for the doctrine ' that all the objects of knowledge
are ideas in my own mind.'1 The rejection of idealism,
the reality of the physical world, the belief in a phy
sical object of sense, and the possibility that a psychical
subject may apprehend a physical object, are all points
in intuitive realism which find a place in physical
realism. But here the agreement ends. The intuitive
realist holds an immediate perception of a physical
world outside. I distinguish the immediate perception
of the physical world within, and the inferential per
ception of the physical world beyond myself.

The intuitive realist follows the idealist in thinking
too much of the sensible data, and too little of the
insensible objects of science. He gives too much
weight to consciousness, and too little to science, or
rather too much to the ordinary and too little to the
scientific consciousness. He appeals to common sense,
which is the problem rather than the solution of philo
sophy. He elevates the dicta of consciousness and

1 Eeid, Essays on the Intellectual Powers. Essay II., chap. x.
p. 283 (ed. Hamilton).

28 PHYSICAL REALISM PART i.

common sense from unanalysed facts into self-evident
principles. Hence, in asserting an immediate know
ledge of external nature lie contradicts science. But
we must appeal from common sense to universal science,
and from ordinary to scientific realism. The idealist
can never be answered by asserting the reality of the
sensible world, which he admits, and, if it stood alone,
could explain. He must be confronted with the in
sensible world of science.

The intuitive realists have an impossible theory of
the data of sense, comprised of two incompatible ex
tremes. On the one hand, they admit the idealistic
position that secondary qualities, as sensible, are psy
chical sensations ; on the other hand, they assert that
external primary qualities of the non-ego are imme
diately perceived. The admission is fatal, because the
Berkeleian at once points out that primary qualities are
apprehended in the same way as secondary, and there
fore if one set, as sensible, are psychical sensations, why
not the other ? The assertion is equally fatal, because
scientific analysis shows that nothing external is imme
diately perceived. Hence I retract the admission and
reject the assertion. Whether directed to primary or
to secondary qualities, sense apprehends neither a sen
sation nor an external object, but an internal object in
the nervous system. Everything external is inferred.

Perhaps the chief reason of the defect in intuitive
realism is the confusion of object and non-ego. Object
is the res considerata apprehended either by sense or by
reason. It is not always an external object. In sense,
it is always internal, whether it be the hot or the moving,
the white or the extended, secondary or primary. In
reasoning, it is external, whenever we infer something
beyond the sensible object within us. But the intuitive

CHAP. ii. IDEALISM AND REALISM 29

realists, having confused object and non-ego , supposed
that whenever sense has an object it presents the non-ego.
Keally, sense always apprehends an object distinct from
the operation, but never a non-ego distinct from the ego,
that is, the man himself. Hence, also, their erroneous
belief that in apprehending a primary quality, as an
object, sense presents a quality of the non-ego, and in
not apprehending a secondary quality as it is in the
non-ego, it presents no object. Eeally, as sensible, both
primary and secondary qualities are apprehended as
objects, but not as external. For example, the sensibly
hot and moving are both apprehended as objects by
sense, but entirely within the sentient being.

The subordination of secondary to primary quali
ties is not at all in the sensible effects, but in the external
causes. In the external world, secondary qualities are
found by science to be only specific varieties of primary
qualities. In the internal world, all qualities appear
to sense to be equally elementary. As sensible, a
primary quality, such as motion, is not in the non-ego,
and a secondary quality, such as heat, is not a mere
sensation ; nor are they both sensations ; but they are
both sensible objects, both internal to the sentient being,
both physical, both parts of the nervous substance
sensibly affected, both apprehended in the same way as
objects by the operation called ' sensation.' From these
qualities, all apprehended in exactly the same way as
sensible objects in our nervous system, the ordinary
man infers a complete correspondence of qualities out
side, the scientific man partly corrects him by reducing
secondary qualities to primary qualities in the external
world.

The relativity of knowledge has become a common
place. Is it a fact ? A sensible effect is the result of

30 PHYSICAL REALISM PART T.

the combination of two causes. As active or efficient
cause, the external world produces the sensible effect
in the nervous system ; as passive or material cause,
the nervous system receives this effect according to its
susceptibility. Hence the effect is like or unlike to the
efficient causes, according to the varying susceptibility
of the nervous system. There is a variation in different
animals and in different men, and even in the same man
at different times. But in all men there is one differ
ence of main importance. The nervous system is far
more susceptible of similar effects from primary than
from secondary qualities. It is more capable of re
flecting the waves of the sea than the undulations of
gether. Not that the effect is wholly alike in primary
or wholly unlike in secondary qualities. The primary
quality of distance is imperfectly reproduced in sense,
the secondary quality of aerial vibration is to some
small extent represented in the sense of hearing. But,
on the whole, there is a general similarity of the
sensible to the external in primary, and a general
dissimilarity in secondary qualities, because of the
inferior susceptibility of the nervous system to receive
like effects from the latter qualities in external objects.
In the sense, then, that the sensible effect only partly
depends on the external efficient cause, and partly also
on the matter of the nervous system, there is a rela
tivity of knowledge to the structure of the nerves.
There is also an evolution, which consists in the in
creasing adaptation of the nerves to sustain the effect
under the action of the external object.

On the other hand, by the relativity of knowledge
it is generally meant that the sensible effect produced
is a psychical fact, not partly but wholly heterogeneous
to the physical object, if there be one. In this sense

CHAP. ir. IDEALISM AND REALISM 31

physical realism is opposed to the relativity of know
ledge. It is true that red refuses to appear to our
senses as a motion representing the external motion
which produces it. But the cause of this fact is to IK;
found in the construction of the optic nerve, which,
when acted on by a certain imperceptible motion of
aether, receives a sensible colour apparently unlike
motion, just as oxygen and hydrogen in certain pro
portions, when acted on by electricity, become water.
In the same way, when a wheel rotates too quickly, the
sensible effect ceases to be a motion, because the nerves
are insusceptible of taking on so rapid a motion in
sense. The sensible effect is similar or dissimilar to
the external object, so far as the nervous system is
capable or incapable of being affected similarly to the
external object. There is no occasion then to resort to
the hypothesis of a psychical relativity : the nervous
element is sufficient.

Moreover, if there were a psychical relativity, it
would be ineradicable, because the sensible effect would
then be completely heterogeneous, and would there
fore supply no data of inference to an external physical
cause. Eeally, sensible effects are partly like and
partly unlike the external causes, because the nerves
are partly fitted and partly unfitted to represent
them. Being partly like, the nervous unfitness to re
present secondary qualities as they are in nature is
being constantly eliminated by scientific reasoning.
Thus, sense sometimes presents motion as motion, but
cannot help presenting the hot, the red, &c., as
heterogeneous to motion, because of the structure of
the sensory nerves ; science, by comparing sensible
motion with the sensible facts of the hot, the red, &c.,
infers that the external cause of the latter is really a

32 PHYSICAL REALISM PART i.

mode of motion. In secondary qualities the sensible
effect is heterogeneous, but the cause inferred by science
is identical with the external object. Not that scien
tific elimination of the defects of sense ever becomes
so complete as to end in absoluteness of knowledge.
But there is a constant progress towards making science
the mirror of being. Sense starts with physical data
partly like and partly unlike external nature ; science,
by progressive inferences, tends more and more to dis
cover the external qualities which cause not only the
like but the unlike data in the nervous system. The
sensible, therefore, is not a psychical effect completely
heterogeneous to the external physical cause, but a
physical effect partly relative to the nervous system ;
and science is perpetually correcting this partial re
lativity.

It is usual to divide theories of sensation and per
ception into presentative and representative. There
are two presentative theories, respectively characterising
the pure idealist and the intuitive realist. The former
holds that there is no distinction between sensation and
perception : sense, according to him, immediately per
ceives psychical facts, which are the sum of known
existing objects. The latter distinguishes sensation and
perception, because he distinguishes the psychical and the
physical : sensation, in his view, is limited to psychical
sensations, perception immediately apprehends the pri
mary qualities of an external physical world. The
pure idealist says, ' What I see is what exists ; ' the intui
tive realist, ' What exists I see : ' the former reduces
nature to perception, the latter brings perception to
nature ; one holds esse is percipi, the other esse per-
cipitur. But the point is that, according to both, the
real is the sensible world, which is directly presented,

CHAP. II.

IDEALISM AND REALISM 33

not represented, in perception, without an inference to
an external original. The representative theory, on
the other hand, distinguishes the data of sense, as pre
sented, from the external world, as represented, in
perception. It exists in many forms, according to
various theories of the data of sense. But the current
form is that of cosmothetic idealism, which holds that
sense presents psychical data of some kind, representing
physical objects in the external world.

Physical realism must accept the representative
theory, but not in its idealistic form. The data pre
sented to sense are internal, yet not psychical. They
are physical parts of the nervous system, tactile, optic,
auditory, &c., sensibly affected in various manners, repre
senting, but only partly resembling, the external world.
Further, in sense, the object is not the operation, the hot
is not touch, the white is not vision, the loud is not
hearing. From these points I form the following theory
of sensation. In that the sensible object is internal, sen
sation is not the immediate apprehension of an external
object. In that the sensible object is physical, sensation
is not the immediate apprehension of a psychical fact.
In that it is the immediate apprehension of an object,
though internal, it is a kind of perception. I should
define sensation, or sensitive perception, as the im
mediate apprehension of an internal physical object
within the nervous system of a sentient being.

But perception cannot be confined to sensation.
Although it is true that sense feels the hot, and reason
infers the fire, everybody talks of perceiving the fire.
The philosopher will find it vain to fly in the face of
the universal language not only of ordinary life but
even of science. He must recognise this perception
and analyse it. There is, then, besides sensitive or

L>

34 PHYSICAL REALISM

PART I.

immediate perception, inferential or mediate perception.
The former is limited to the internal object of sense, the
latter extends to the external original. Moreover, so
long as we remember that there is an inference in this
latter operation, the term ' perception ' not only does no
harm but serves to mark a most important distinction.
We first infer external originals of sensible objects, e.g.
the fire, the sea, &c. ; we cannot be said to see, but we
may be said to perceive, these external objects, and also
to observe and experience them, though indirectly.
Afterwards, we go on to infer other external objects not
represented by any sensible object, e.g. a corpuscle,
aether : these we cannot be said either to see or per
ceive ; they are not only insensible but imperceptible,
and we infer them by reasoning which transcends per
ception. In short, we must distinguish sensitive
perception, inferential perception, and transcendental
inference.

Hence the following classification of physical objects
knowable, and of the operations concerned with them : —

1. Internal parts of the nervous system sensibly
affected: sensible data : immediately perceptible, objects
of sense, or of sensitive perception, observation, ex
perience.

E.g. the sensibly moving, the sensibly hot.

External parts of the universe: insensible objects :
objects of inference.

(1) Originals represented by sensible objects, and
resembling them in primary not in secondary
qualities: insensible but mediately perceptible
objects of inferential perception, observation,
experience.
E.g. the fire, the waves of the sea.

CHAP. ii. IDEALISM AND REALISM 35

(2) Objects unrepresented, though causing some
sensible objects by imperceptible secondary
qualities : the imperceptible : objects of trans
cendental inference.
E.g. corpuscles, the undulations of aether.

This essay contemplates not only a new realistic hypo
thesis, but a different method from that usually used in
mental philosophy. Every philosophy must have a
beginning. But the beginning must be what is best
known ; and in mental philosophy the present objects
of science are better known than the original data
of sense. The method in use takes too direct a way
of getting at the original data. It is true that the
beginnings of human knowledge are sensible data.
But the philosopher does not stand at the beginning of
human knowledge. Philosophy did not begin with the
infancy of the human race. The philosopher cannot
observe his own infancy. The sensible data have long
since been overlaid with an immense mass of inferences.
Hence, though man may have begun once, it is impos
sible for the philosopher to begin now, with the data.
Yet most books on knowledge begin with the dogmatic
assertion that the immediate objects of the senses are
psychical sensations, from which they proceed to allow
man as much knowledge of nature as can be squeezed
out of the original hypothesis. But the assertion itself
must be proved.

Besides the induction of causation, we may either
reason synthetically from cause to effect, or analytically
from effect to cause. But the latter is the more usual
method, because man knows so much more about
facts than about their causes. Hence the order of
science is usually the reverse of the order of nature.

D 2

36 PHYSICAL REALISM

TART I.

Nature always proceeds from cause to effect, science
usually from effect to cause ; so that science becomes
an analysis of the synthesis of nature.

Similarly, the order of mental philosophy is the
reverse of the order of human knowledge. It is true
that the order of human knowledge is from cause to
effect in the sense that sensible data are the causes cog-
noscendi of physical knowledge. We begin with them
as children ; hence also we are tempted to begin with
them again as psychologists. But the procedure is
fallacious ; we must begin with the more knowable.
Now every mental philosopher is an adult man, and
every adult man is more certain what he now knows,
than how he originally came to know it, of the dis
coveries of science than of ' the secret springs and
principles by which the human mind is actuated in its
operation,' of the known objects than of the sensible
data. Accordingly, as, in the science of nature, we
must generally begin with present facts and go back
wards to the causes essendi, so, in the science of know
ledge, we must generally begin with the facts of
scientific knowledge and go backwards to the causa?
cognoscendi. Modern philosophers have made the mis
take of attempting to repeat the synthesis of know
ledge from the original data of the child and the
race. But we must rather retrace our steps from the
present to the past; instead of trying to follow the
synthesis of knowledge from an unknown beginning,
we must make an analysis from the present objects
of scientific knowledge to the original data of sense.
In a word, our method must be an analysis from science
to sense.

Hence, I began with attempting to give an outline
of the kind of objects recognised in science. This

CHAP. ii. IDEALISM AND REALISM 37

beginning lias several advantages. First, science is
knowledge ; hence to begin with its objects is an
appeal not from knowledge to reality, but from the
data to the objects of knowledge. It is not a dogmatic
assertion of what is, but an historical description of what
is known. Secondly, science is knowledge at its widest
extent, knowledge proceeding from the sensible through
the insensible, but perceptible, to the imperceptible
world. Hence we get a more extended view of know-
able objects than that usually attained by mental
philosophers, who tend to concentrate themselves on
the world of sense and perception. Thirdly, science is
knowledge at its best, whereas the hypotheses of mental
philosophers about sensible data can hardly be called
knowledge at all. In appealing from the hypothetical
origin of knowledge to what is actually known in science,
we are appealing from the less known to the more
known. In short, we are getting the facts of knowledge,
wherewith to test our hypothesis of its causes.

The next step is analytically to find the sensible
data required to cause the knowledge of the objects of
science as facts. All theories of the sensible data and
of the origin of knowledge, idealistic and realistic, must
be treated and compared as hypotheses. We must ask,
indeed, what is their direct evidence, but also and
mainly whether they account for the knowledge of the
objects of science. The general examination of these
hypotheses will follow in the next chapter. Afterwards,
the various hypotheses of Psychological Idealism will
be taken in detail. The elimination of these hypotheses
will finally bring us to Physical Eealism.

Philosophy began with the external object, which
was first of all treated as a pure reality by the Pre-So-
cratic philosophers. Gradually it came to be regarded

38 PHYSICAL REALISM PART i.

as also an object of knowledge, a view which culminated
with Aristotle. Aristotle's method was essentially to be
gin with being as being, then to consider it secondarily as
a knowable object, and thus to proceed from the known
object to the knowing subject. Objective are generally
the foundation of subjective distinctions in his writings.
Descartes revolutionised philosophy by beginning with
the conscious subject and passing through its conscious
operations to the object apprehended. From his time
the general order of mental philosophy has been syn
thetic, from the subjective operations to the objective
world. I propose to revert to the old order, and pro
ceed analytically from object to subject, but in a new ^
spirit.

Ancient philosophy rightly began with the object, but
considered it too much as being, and too little as known.
Consequently, it had a tendency to multiply entities
without considering whether they are knowable. Hence
the Cartesian revolution and the synthetic method
from subject to object. But after the first consciousness,
I think, the object is on the whole better known than
the subject ; else natural philosophy would not be more
advanced than mental philosophy. In order to avoid
at once the dogmatism of ancient, and the doubtfulness
of modern, philosophy, I propose to begin with the
object, not as being, but as known in science, the most
perfect form of knowledge. I proceed to ask what
sensible objects are required as data for science to X
know these objects. Of the knoAving subject I treat
only so far as it bears on the objects known by sense
and reason, because, though I know well that I am, I
know less what I am than what I know. The ancient
method from being to knowing was the right order,
though too dogmatic in application. The modern

CHAP. IT. IDEALISM AND REALISM 3D

method inaugurated by Descartes, from the subject
through the data of sense to the objects of science, was,
after its first step, fallacious, because it then proceeded
synthetically from the less to the more knowable. The
analytic method of physical realism, without neglecting
direct evidences of the data, proceeds, on the whole,
from the more knowable objects of science to the less
knowable data of sense.

TABLE OF IDEALISM AND REALISM.
IDEALISM. KEALISM.

(1)A11 sensible (1) All sensible (1) All sensible (1) Some sensible
data are psy- data are psy- data are inter- data are exter-

chical. chical. nal but some nal and physi-

are physical. cal.

(2) All objects ( 2) Some objects (2) Some objects (2) Some objects
knowable' are are physical. of science are are physical.

psychical. physical.

40 PHYSICAL REALISM PART i.

CHAPTEE III.

THE PHYSICAL DATA OF SENSE.

Nihil est in intellectu, quod non prius in sensu. How
far is this time-honoured' proposition true? As we have
seen, it is not true of the objects of science. The whole
physical world is beyond the reach of sense, insensible ;
the corpuscles, of which it consists, are beyond the reach
of inferential perception, imperceptible. It is true that
objects of science are similar to sensible objects, but
they are not the same. They are objects of intellect
which are inferred from sensible objects but have never
been in sense. But even this more modest statement
must be qualified.

In the first place, it requires Locke's correction that
knowledge has two sources — sensation and reflection,
outer and inner sense, or sense and consciousness. We
immediately apprehend not only the objects of, or
rather in, our senses, but also ourselves apprehending
those objects, and performing many other conscious
operations. Secondly, there is also a simpler source
than sensation — the feelings. We immediately feel
pleased and pained, and that too without apprehending
any object ; as in the pain of hunger, the pleasure of
nutrition. Sensation is more complex than feeling, be
cause it is the apprehension of an object ; touch the
apprehension of the hot, vision of the coloured, hearing
of the sounding, &c. Frequently we have a feeling and

CHAP. in.

THE PHYSICAL DATA OF SENSE 41

a sensation together ; for example, when we feel pleased
or pained at the same time as we taste sweet or bitter.
But it is of the greatest importance to distinguish feeling
as a source of knowledge, especially as it is not at all
improbable that it was the original source even of sensa
tion. Even now that feeling and sensation are distinct,
feelings are still the raw experiences of volitions, passions
the beginnings of actions. We feel pleasure and pain
before we will to pursue the one and avoid the other.
All knowledge, then, does not begin with sensation, but
with feeling, sensation, reflection.

It is true, however, that all knowledge of nature
begins with sensation. Yet even this modified proposition
must be carefully guarded. In the first place, though phy
sical knowledge begins with the operation of sensation, it
does not follow that the object, in apprehending which
the operation of sensation consists, is also a sensation.
Yet this non-sequitur appears in the first few pages of
most books of modern philosophy. The causes of the
confusion of sensation with its object are to be found
partly in the structure of modern languages, which,
being far richer in abstract than in concrete terms,
tempt philosophers to fall into a loose way of speaking
of perceiving a sensation instead of perceiving a sensible
object ; but mainly in another confusion, that of object
and non-ego, which makes philosophers shrink from
speaking of perceiving a sensible object, lest they should
seem to assert an intuition of the external world. But
an object (TO avTLKeiptvov) is merely that which is
apprehended as opposed to the operation of apprehend
ing it, and is not necessarily external to the apprehend
ing subject. In sense, without being external, the
object is still distinguishable from the operation ; the
hot from touch, the sweet from taste, the coloured from

42 PHYSICAL REALISM PART i.

vision, the loud from hearing, the scented from smelling.
Although, therefore, physical knowledge begins with
sensation as an operation, it does not begin with sensa
tion as a sensible object. Given, then, that physical
knowledge begins with sense, we still have to ask, what
is the object apprehended immediately by sensation ;
what is the sensibly hot, sweet, coloured, loud, scented ?
This is the question of the present chapter.

There are two main evidences of hypothesis — the
direct and the indirect. Direct evidence is the best, if
possible, but it is seldom attainable ; for example, there
is no direct evidence for the hypothesis of gether. But
where direct proof fails, indirect should be all the
stronger in compensation. It consists in using the
facts to test the hypothesis, and that in two ways.
First, the facts must be explained by the hypothesis ;
secondly, they must eliminate other explanations. Thus
the hypothesis of an undulating aether, as the vehicle of
light, though wanting in direct evidence, is proved by
its power of explaining all the facts of light, and by the
elimination of the hypothesis of emission, which explains
some, but not all the facts.

I propose to apply these rules to the various
hypotheses of sensible data, stated in the last chap
ter. Are the objects of sense, which form the data
of science, psychical or physical ; and, if physical, ex
ternal or internal ? On the one hand, how far is
there direct evidence for any of these hypotheses ?
On the other hand, how do they stand the indirect test
of the facts of science ? That is, can the objects of
science as facts of knowledge be explained by any
hypothesis of the data of sense ; and can the other
hypotheses be eliminated ? Being hypotheses, idealism
and realism alike must be treated by the logical rules

CHAT. III.

THE PHYSICAL DATA OF SENSE 43

of hypothesis. Sensible data must be made to explain
the scientific facts, as aethereal undulations have been
made to explain luminous facts. We must be on our
guard against synthetic hypothesis. What would be
thought of a natural philosopher, who dared to start
with the hypothesis of emission and denied all the facts
of light, which cannot be deduced from the emission
of corpuscles by a luminous body ? What, then, shall
we think of mental philosophers, who start with the
hypothesis of sensations and deny all the insensible
world which cannot be deduced from the contempla
tion of sensations by sensation? I admit that there
may be direct evidence of an hypothesis. But even
so, unless that evidence be mathematical certainty, the
hypothesis must also be submitted to the indirect or
analytical evidence of explaining the facts. Now it
cannot be pretended that the direct evidence of the
hypothesis of perceiving sensations or any other hypo
thesis of sensible data is mathematically certain. There
fore all the hypotheses of idealism and realism must pass
through the alembic of analysis.

The first direct evidence is that of consciousness.
Consciousness is the immediate apprehension of oneself
performing some operation. Thus I am conscious
that I feel, that I perceive through my senses, that I
imagine, remember, reason, desire, will, act. Unfor
tunately, however, this operation of apprehending other
operations has come to be confused in psychology
with the operations themselves. Hamilton, seeing that
perception requires an object, and consciousness of
perception requires perception, falsely concluded that
the consciousness includes the perception of the object,
whereas it only requires it as a condition. He com
mitted the common fallacy of confusing a thing with

44 PHYSICAL REALISM PART i.

its condition. Eeally, perception is the apprehension
of the object, consciousness of perception the apprehen
sion that I am apprehending the object. Mill, again,
seeing that feeling pleasure and pain are the same as
being conscious of feeling them, falsely concluded that
every operation is the same as its consciousness. He
committed the fallacy of over-generalisation. In feeling
pleasure and pain there is no distinction between opera
tion and object, and hence none between feeling and
consciousness. But whenever there is a distinction
between operation and object, the operation is concerned
with the object and the consciousness with the operation.
Hence to see white is different from being conscious of
seeing white. So with other operations. Reasoning is
a mediate operation from premises to conclusion. The
consciousness of reasoning is an immediate apprehension
that I am performing that mediate operation. Will
is an active operation, the determination to act ; its
consciousness an intellectual operation, apprehending
that I determine to act. To reason and to will, then,
are not the same as being conscious that I reason and
will.

It is not improbable that the lowest potency of sensi
tive life may have been mere feeling, and the beginning
of consciousness mere conscious feeling ; and that as,
in the growth of the senses, the operation and the object
became distinguished, consciousness became distinct
from the operation, the operation being concerned with
the object, and the consciousness with that relation of
oneself to the object, in which an operation about an
object consists. But whatever may have been the
genesis of consciousness, its nature consists not in being
the sense of objects but the sense of operations. When,
as in feeling, there is no distinction between operation

CHAP. in. THE PHYSICAL DATA OF SENSE 45

and object, there is none between consciousness and
operation. When, as in sensation, there is a distinction
between operation and object, the operation is con
cerned with the object, the consciousness with the opera
tion. Not that consciousness has no reference to the
object, but only that it is not the apprehension of it.
The operation, which is the apprehension of the object,
is a certain relation of subject to object : the conscious
ness, which apprehends the operation, is an apprehension
not of the object, but of the relation of the subject to the
object. For example, I see white, I am conscious that
I am seeing white.

It was necessary to have thus defined consciousness
on account of the mass of confusion and inconsequence
imported into psychology by regarding consciousness
as identical with all the conscious operations. Hamilton,
seeing that consciousness is intuitive, but falsely identi
fying it with the perception of an external world, falsely
concludes that perception of an external world is also
intuitive. He ought by the same argument to have
made reasoning immediate, or else consciousness mediate,
either of which alternatives is absurd. Mill, seeing that
consciousness is limited to the apprehension of mental
operations, and falsely identifying it with the mental
operations, falsely concludes that the mental operation
of sensation is also limited to the apprehension of
mental operations. He might as well have said that
will, being identical with its consciousness, is an intel
lectual apprehension of a mental operation. But as
will is an active determination to do something, while its
consciousness is an intellectual apprehension that one
has that active determination, so sensation is an appre
hension of an object, while its consciousness is an
apprehension that one is performing that operation.

46 PHYSICAL REALISM PART i.

Sensation says, ' This is white or sweet ' ; consciousness
says, ' I am seeing something white or tasting something
sweet.'

This being consciousness, one operation of which I
am conscious is that I know objects. What knowledge
of objects am I conscious of possessing ? In answering
this question, it must be remembered that science is a
kind of knowledge of which we are conscious. There is
an ordinary consciousness and a scientific consciousness.
The ordinary man thinks little or nothing about it, but
the man of science is conscious that science passes
beyond sense into the insensible, and beyond the objects
represented by sense into what I have called the im
perceptible world. We are conscious of knowing a
sensible, an insensible, and an imperceptible world by
natural philosophy.

Now, this knowledge does not appear to conscious
ness to apprehend a psychical object. When I reflect
on my inferential knowledge of the number of corpuscles
in a drop of water, or of the distance of the sun from
the earth, or of the size of the earth ; when, again, I
reflect on my indirect perception of a fire, or the waves
of the sea ; when, finally, I reflect on my sensation of
the white object I see or the hot object I feel ; in all
three instances, I appear to my consciousness to be
apprehending not psychical, but physical facts. The
conscious subject maybe psychical, the conscious opera
tions may be psychical ; but I am not conscious that
the vision of white, or the perception of a fire, or the
inference of a corpuscle, apprehends a psychical object.
So far as I am conscious of the sensations of my five
senses, a white object in vision, a hot object in touch,
a scent in my nostrils, a sound in my ears, a flavour
in 'my mouth, cannot but seem to be apprehended as

CHAP. in. THE PHYSICAL DATA OF SENSE 47

physical objects. Consciousness of the apprehension of
objects is in favour of realism.

But when I apprehend the white, the hot, a scent, a
sound, even a flavour, I further appear to be appre
hending an object not only physical, but also external
to myself. This seemingly conscious appearance is the
strong point of intuitive realism, which depends on it to
claim an intuition of an external world. Nevertheless,
the appearance is a delusion, which we can trace to its
source. From my earliest infancy, whenever a sensible
effect has been produced in my nervous system, I have
been accustomed to infer an external object. By asso
ciation, perhaps also facilitated by evolution, the in
ference has become so automatic as to be unnoticed. The
consequence is, I think I am intuitively sensible of the
external object when I am really inferring it. Nothing
can prevent the delusion. I appear to see the paper
and its distance from me. I cannot now consciously
disengage the sensation of the sensible object from the
inference of the perceptible original.

Hence the limits of consciousness as an evidence.
Consciousness does not become reflective, and therefore
a source of psychology, till many operations have
already become automatic in the conscious subject.
The process from the sense of the insensible object to
the inference of the perceptible original has been re
peated an incalculable number of times before any man
is sufficiently adult to consciously reflect on what he
has been doing. Accordingly, consciousness is the
source rather of the nature than of the origin of

o

knowledge ; invaluable for what we know now, delusive
for how we came to know it. I am conscious that
I somehow apprehend a sensible and an insensible
world ; but I am not conscious of the exact point at

48 PHYSICAL REALISM TART i.

which it ceases to be sensible, and becomes insensible
and inferred. Intuitive realists were right in appealing
to consciousness for the nature of knowledge ; only
they should have appealed from the ordinary to the
scientific consciousness. But they were quite wrong in
appealing to consciousness for the ultimate origin of
knowledge. They said truly, ' I apprehend an external
world ' ; they said falsely, ' I apprehend it intuitively/

Nevertheless, the antithesis between the nature and
origin of knowledge must not be exaggerated. Con
sciousness tells us something of the origin of our know
ledge. We are not conscious of the inferences of child-

o

hood : when we are old enough to take notice we
become conscious of new inferences. We are not con
scious of inferring an external world : we are conscious
of inferring corpuscles. The exact limit is that we
are not conscious of the primary data and the first
inferences, but of adult inferences. But, again, con
sciousness has something to tell us concerning even
the primary data of sense. It is not their direct but
their indirect evidence. It tells us what is our know
ledge of objects, and this conscious knowledge must
be explained by the primary data. Thus consciousness,
on the whole, is the apprehension of our knowledge
of objects and the test of the primary data and origin ;
it is the direct evidence of the nature, the indirect
evidence of the origin of knowledge. The facts of con
sciousness must be first described and then explained
by all scientific psychology. The main fact to be ex
plained is our consciousness that we somehow appre
hend a sensible and an insensible physical world.

There is a superficiality of consciousness as there is
of sensation. Yet each is the origin of a philosophy.
Without sensation there would be no natural, without

THAI', in. THE PHYSICAL DATA OF SENSE 40

consciousness no mental philosophy. Sensation is neces
sary to the science of nature, consciousness to the
science of in hid. Sensation apprehends what is sensible,
consciousness what is knowledge. Sensation carries us
some little distance into physical causation, conscious
ness into the origin of knowledge. But sensation leaves
us to infer causes essendi in external nature, conscious
ness causce cognoscendi in internal knowledge. Yet
sensation is the indirect test of all hypotheses to ex
plain the causes of sensible objects, consciousness of all
hypotheses to explain the origin of conscious know
ledge. Nevertheless, both in themselves are superficial ;
for sensation has no immediate intuition of the external
causes of nature, and consciousness none of the internal
data of knowledge. As direct evidence, sensation tells
ns only the bare sensible effect and not its external
causes, consciousness only what we know now, not how
we came to know it. Sensation and consciousness are
twT> senses, the outer and the inner. Neither is false ;
both are limited. Truth is in pro/undo ; yet not in a
bottomless abyss, but in depths to be plumbed only by
reason, and that reason not a priori, but logical inference
from the outer and inner senses. Not sensation, but
reasoning from sensation, discovers external causes ;
not consciousness, but reasoning from consciousness,
discovers the primary data and origin of knowledge.

Consciousness, then, does not aid the idealist in his
assertion that all the immediate objects of sense are
psychical. It tells us that we somehow know physical
objects. It is so far in favour of realism. Having,
however, inferred long ago from sensible data that-
physical objects exist in the external world, we cannot
now help seeming to be conscious of perceiving them
intuitively. This confusion favours intuitive realism.

E

50 PHYSICAL IlEALISM PART i.

But consciousness cannot be used as direct evidence to
tell us what we intuitively perceive, because our intui
tions were overlaid with our inferences long before our
consciousness became attentive. Moreover — and this is
the main point — the confusion of what we intuitively
perceive by our senses with what we mediately infer by
our reason is cleared up by philosophy. What philo
sophy ? This question brings us to the second kind of
direct evidence for the data of sense. The philosophy
which has distinguished the data of sense from their
inferred causes is natural philosophy.

Natural philosophy has shown that the sensible
object is not really identical with, but is an effect distinct
from, its external original. When a person hears a
cannon fired at a considerable distance, his first impres
sion is that he hears the sound at the very moment the
ball issues from the cannon's mouth, and that the cannon
sounds as he hears it. But if he ascends a hill, and the
cannon again fires, he finds that he sees the smoke of the
camion long before he hears the sound, and can count
several seconds between the object seen and the object
heard. There is only one possible explanation of this
distinction. The object seen and the object heard are
neither identical with one another, nor with the external
object which produces them. The smoke ascends from
the cannon and reflects the undulations of light, at the
same moment as the ball leaves the camion and com
municates vibrations to the air. But the undulations
of light travel faster than the vibrations of air, and
produce a visible effect on the person before the audible
effect is produced by the slower mode of motion. The
visible effect produced by the undulations is not the
smoke, and the audible effect produced by the vibra
tions is not the cannon's roar : else thev would be

CHAP. III.

THE PHYSICAL DATA OF SENSE 51

apprehended at the same moment. Both are effects of,
neither identical with, the external object.

Again, natural philosophy in the department of
physics has shown that external things do not in all
their attributes precisely correspond to their effects on
our senses. They have duration, extension and motion
corresponding to their attributes as sensibly perceived ;
but they have not heat or colour in the way in which
we touch what is hot or see what is coloured. On the
contrary, the causes of sensible heat and colour are in
sensible motions. The attributes which are in nature
as they are in our senses, are called primary; while
those which are not in nature as they ara in our senses,
are called secondary qualities.

Again, natural philosophy has proved that external
things affect our senses by the causation of motion. To
begin with motion before it affects the senses ; either a
given external thing may itself move from a distance,
until it comes into contact with a sensitive subject, as
a cannon-ball does when it hits a man ; or it propagates
a motion from particle to particle until the particles im
mediately in contact with the sensitive subject receive
the motion, a process which takes place in the propaga
tion of the undulations of light. In both cases the result
is the same : the object immediately apprehended could
not be the thing at a distance, but the thing immediately
next to the sensitive organ. But we shall find that it is
not even the nearest thing, as a matter of fact, but an
effect within our senses.

Again, biology, from Galen onwards, has shown that
the nervous system is the material cause susceptible of
the effect produced by the efficiency of the external
object. It has discovered much of the structure of the
nervous system. The peripheral terminations of nervous

E 2

52 PHYSICAL HEALISM

PAhT I.

fibres are not actually exposed to external tilings.
Hence the motion lias to be propagated through a non-
sensitive covering before it is actually brought to the
nerve. It is impossible, therefore, to be sensible of an
external object, from which the nervous substance is
divided by a medium in the body of the sentient being.
Moreover, when the peripheral termination of a nervous
fibre has been reached, the effect is still insensible till
the motion has been communicated to the brain. When
a nerve has been cut off from the brain, if the part
between the peripheral termination and the section be
irritated, no sensible effect takes place ; but if the part
between the brain and the section be irritated by pres
sure, or electricity, or disease, the effect is sensible.
The brain, therefore, is an integral part of the nervous
material susceptible of a sensible effect. I say a part,
because there is no evidence that the brain alone would
be sensitive, if a whole nervous structure up to the
point at which it loses itself in the brain were removed.
But the whole evidence together clearly shows that no
sensible effect is produced by an external object until
the propagation of motion from the external object has
passed, not only the external medium, but what may be
called the internal medium of the periphery, has reached
the nervous fibres, and communicated itself to the brain.
The nervous system is the primary matter susceptible
of a sensible effect, and the sensible effect, therefore, is
internal.

It is further evident that, like other material causes,
the nervous system partly determines the effect. It is
susceptible of effects, like the primary qualities of ex
ternal objects, but not like their secondary qualities.
Probably its structure is adequate to the former, but
not sufficiently subtle for the latter. Hence the dura-

CHAP. in. THE PHYSICAL DATA OF SENSE 53

tion, extension and motion of external bodies is able
to produce similar sensible duration, extension and
motion in the nervous system. But when the delicate
vibrations of the air and the still more subtle undula
tions of ajther strike upon the organs of hearing,
touch and vision, the nervous structure of these
organs is too coarse-grained to reproduce them, and
substitutes the heterogeneous effect of sensible sound
and the still more heterogeneous effects of sensible heat
and light.

Lastly, evolution has made it exceedingly probable
that, like other material causes, the nervous system has
itself been modified by the repeated action of the ex
ternal efficient on its structure. It is probable that, by
the frequent operation of appropriate stimuli on parti
cular parts of the general sensitive system, the original
sense of touch has been differentiated into the five senses.
I would make two further suggestions. First, it is
probable that as touch preceded the other senses, so the
feelings preceded touch. In this case, the sentient being
at first simply felt mere pleasure and pain from external
objects; afterwards proceeded to the more complex
operation of touch, in which the sensation of touching
is distinct from the sensible object, hot or cold, in the
tactile nerves, and the consciousness of touching distinct
both from the sensation and the sensible object ; and,
last of all, proceeded to infer external causes. Secondly,
it is probable that, as the nervous system has become
more differentiated, it may also become more subtle, and
therefore more discriminative of secondary qualities.
Some approach to this ideal may be found in sensible
sound, in which there is some trace of vibrations, though
not adequate to the external vibrations. Why, then,
may not the nervous system some day become more

PHYSICAL REALISM PART i.

attuned to represent aethereal undulations to some
extent in the wonderful sense of vision ?

The discoveries of natural philosophy eliminate
intuitive realism, by proving that the external is not
identical with the sensible object, but is the cause
which produces it in the nervous system. Contenting
himself with crude consciousness and common sense,
forgetting how late consciousness becomes reflective,
and that common sense never becomes a science, the
intuitive realist takes the appearance that we have an
intuition of the external world for a fact, and some
times even converts it into a first principle. But he
comes into contradiction with science. Natural philo
sophy shows that the external world affects us indi
rectly, and that we have no empirical intuition except
of ourselves. We might doubt between consciousness
and science, if we could not see that the supposed in
tuition of the external world is a delusion of association,
and that consciousness is put out of court by its in
ability to reflect at the time when the inference of the
external world was being made ; made so often then as
to have become automatic, and now made so quickly
as to seem an intuition. On the strength of science,
then, we must reject the hypothesis that the data of
sense are to be found in the external world, in the
non-ego.

The same scientific discoveries raise a strong pre_
sumption in favour of physical realism, which simply
adopts the scientific account without further hypo
thesis. In the first place, I suppose that the effect
produced on the nervous system is the sensible physical
object, which we are conscious of apprehending, but by
a confusion believe to be an object external to ourselves :
for instance, when we see something white, it seems to

CHAP. in. THE PHYSICAL DATA OF SENSE 50

be the external paper, or what not, but it is really the
effect produced by the paper reflecting undulations on
the optic nerve. Secondly, I suppose that as we know
the external physical cause to produce the sensible
physical effect, and as we must start from sense, we
must use the sensible physical object impressed on o in
ner ves to infer the external physical object, as cause.
The scientific account of the causation of the sensible
effect leads directly to physical realism, which simply
reads the process of causation backwards into a process
of knowledge.

o

Biology has brought the sensible effect within the
nervous system. Has it carried it further ? The
attempt has often been made by biologists. They sup
pose that the physical effect produced in the nervous
system is not yet sensible, even when it has reached the
brain ; that it remains a mere impression, no more
sensible than the external object ; and that when the
motion of the external object has produced the motion
of the medium, the motion of the medium the motion of
the nerves, the motion of the nerves the motion of the
brain, the process is not yet finished. They suppose
that the cerebral motion, which is physical, produces a
sensation which is psychical ; and they do not ordi
narily distinguish the sensible object from the sensation.
From this hypothesis it would follow that the hot felt,
the white seen, the sweet tasted, the durable, extended,
and moving, apprehended by any sense are psychical
affections produced by cerebral motion. The sensible
object will be neither the external object nor the in
ternal effect in the nervous system, but the internal
psychical sensation. If so, realism will have to succumb
to idealism.

The question we now have to ask ourselves is not

PHYSICAL REALISM

TART I.

whether the external object causes our sensation in some
way or other. The scientific evidence of the propa
gation of motion from external objects to our bodies
and the conscious involuntariness of sensation are suffi
cient proofs that the external object does cause our
sensation. It is, however, a different question how one
causes the other. Secondly, the question we now have
to ask ourselves is not whether there is any evidence at
all that the sensation produced is purely psychical.
What is to be said on this point will follow when we
come to the Cartesian philosophy in detail. The
present questions are, first, whether biology proves that
within ourselves nervous and cerebral motion produces
a psychical sensation ; secondly, if so, whether it follows
that the sensible object also becomes a psychical sensa-
The answer is that biologists have gone beyond
biology, and that no affirmative answer can be ^iven to
these questions from the observations, or direct in
ferences from sense, which are the evidences of their
science.

In the first place, the nervous system is imperfectly
known. It is quite clear that external objects propa-
ate motions to the nerves, but it is not at all clear
what happens when the effect has been produced. In
optics, for example, so long as we are reading of the
undulations of light, of the manner in which rays are
communicated to the eye, of the structure of the lens
by which the rays are made to converge on the retina
and of the general structure of the retina, and even of
nervous elements, everything is clear. But the
further we penetrate from the retina along the optic
nerve to the optic centres at the base of the brain, the
darker the subject becomes, and fact seems to pass into
hypothesis. It is the same witli all our senses

CHAP. in. THE PHYSICAL DATA OF SENSE 57

difficulties begin at the very terminations of the nerves.
What, for example, are the precise functions of the
tactile corpuscles, of the rods and cones of the retina,
of the rods of Corti in the ear ?

We know more of nervous structure than of nervous
action. What is nervous action ? This is an unsolved
problem. What is cerebral action ? This is a more
unsolved problem. The structural connections of
afferent nerves with centres, of centres with efferent
nerves, of efferent nerves with muscles, and to some
extent the structural constituents of nerves and muscles
are fairly made out. It is also found that an appre
ciable interval takes place between the stimulation of
an afferent nerve and the muscular motion which it
indirectly but ultimately produces. This interval proves
an important point about nervous action ; it is a motion
because it takes time to go from place to place.

The genus of nervous action, then, is known to be mo
tion. But what is its differentia ? After the first crude
hypothesis of animal spirits moving in the nerves, nervous
motion was supposed to be the simplest form of me
chanical motion by impact, as if the impression were
pushed along to the brain, as a series of bricks knock
one another over. Then it was supposed to be vibra
tion. Later researches tend to show that it has relations
to the motions of electricity and of chemical action. It
is, no doubt, some molecular motion allied to other
motions of the same kind ; but its peculiarity is its
slowness, compared, for instance, with electricity. Its
precise differentia is at present unknown. Cerebral
motion is still more unknown. It has been found,
by experimenting on various parts of the brain, that
different parts are to some extent connected with
different muscular motions, from which it is inferred

58 rriYSiCAL REALISM

TART I.

that they are also connected with different nervous
motions. But how the brain moves between the stimu
lus of an afferent nerve and its effect on an efferent
nerve is unknown. He would be a bold man who
would come forward and say he knows the motion by
which the effect impressed on the nerves is communi
cated to the brain and there made ready for sensation.
How, then, can he say he knows that cerebral motion,
of which in biology he is ignorant, produces a psychi
cal sensation, which is beyond the venue of a physical
science ?

. Secondly, the so-called transmutation of cerebral
motion into psychical sensation is admitted to be per
formed in some mysterious way, unknown and inex
plicable. This point may be made clear by the following
quotation from Professor Huxley's Lay Sermon on Des
cartes' Discourse, in which the Professor is trying to
prove that thought is existence, and, so far as we are
concerned, existence is thought :—

' For example, I take up a marble, and I find it to be
a red, round, hard, single body. We call the redness,
the roundness, the hardness, and the singleness, " quali
ties " of the marble ; and it sounds, at first, the height
of absurdity to say that all these qualities are modes of
our own consciousness, which cannot even be conceived
to exist in the marble. But consider the redness to
begin with ; how does the sensation of redness arise ?
The waves of a certain very attenuated matter, particles
of which are vibrating with vast rapidity, but with very
different velocities, strike upon the marble, and those
which vibrate with a particular velocity are thrown off
from its surface in all directions. The optical appa
ratus of the eye gathers some of these together, and
gives them such a course that they impinge upon the

CHAP. III.

THE PHYSICAL DATA OF SENSE 59

surface of the retina, which is a singularly delicate
apparatus, connected with the termination of the fibres
of the optic nerve. The impulses of the attenuated
matter, or aether, affect this apparatus and the fibres of
the optic nerve in a certain way, and the change in the
fibres of the optic nerve produces yet other changes in
the brain ; and these, in some fashion unknown to us,
give rise to the feeling, or consciousness, of redness.
If the marble could remain unchanged, and either the
rate of vibration of the aether, or the nature of the
retina could be altered, the marble would seem not red,
but some other colour. There are many people who
are what are called colour-blind, being unable to distin
guish one colour from another. Such an one might
declare our marble to be green, and he would be quite as
right in saying that it is green as we are in declaring it
to be red. But, then, as the marble cannot, in itself, be
both green and red at the same time, this shows that
the quality " redness " must be in our consciousness, and
not in the marble.'

Thirdly, the hypothesis of this unknown transmuta
tion is inconsistent with one of the best established
facts of the nervous system— its physical continuity. It
supposes that physical motion of afferent nerves and
brain causes psychical .sensation, which causes psy
chical volition, which causes physical motion of efferent
nerves, which causes physical motion of muscles. But
wherever nervous structure is accessible to observation,
the afferent nerves finally communicate with centres
which communicate with efferent nerves, without any
rupture of physical continuity. It might, indeed, be
urged that the intermediate purely psychical processes
nevertheless intervene insensibly in the centre between
the afferent and efferent nervous processes. But this

00 PHYSICAL REALISM TAUT i.

hypothesis is rendered most difficult by the phenomena
of reflex action. In reflex action, the afferent and
efferent nervous processes are certainly connected with
out any breach of physical continuity. It might again
be objected that only the nerves of reflex processes are
continuous. But we cannot divide the nerves of reflex
action from those of conscious action, and say that
the former nerves are physically continuous, whereas
the latter are interrupted by purely psychical sensations
and volitions, because the very same nerves, which are
used in conscious, are used also in reflex actions. For
example, we may wink either voluntarily or automati
cally. An object strikes the eye, transmits its motion
to the afferent optic nerve, which communicates with
the brain, which transmits the motion to the efferent facial
nerve, governing the orbicular muscle of the eyelids,
which makes them close. The whole of this process often
takes place automatically, without any rupture of phy
sical continuity. When it takes place consciously,
are we to say that the physical motion, having arrived
from the optic nerve to the brain, does not produce the
motion of the efferent nerve, but produces a psychical
sensation instead, which produces a psychical volition,
which at length affects the efferent nerve? There is
not a tittle, of biological or any other evidence that the
physical continuity is sometimes preserved sometimes
broken in this manner in the very same series of nerves.
To escape this gratuitous hypothesis of psychical
interruption, some of the mental physiologists resort to
paradoxes, in order at once to preserve the physical
continuity of the nervous system, together with purely
psychical sensations. Allowing that in all cases the
motion of the afferent nerves propagated through the
centres produces the motion of the efferent nerves in a

CHAP. in. THE PHYSICAL DATA OF SENSE Gl

continuous manner, some suppose that, standing quite
apart from these physical processes, the conscious sub
ject is a sort of impartial spectator, performing purely
psychical operations that have no physical effects, while
others positively go the length of supposing that not
only in sentient beings, but in all nature, there are
always two independent but parallel streams, the well-
known physical motions and supposititious psychical
processes accompanying them. These hypotheses are
exceedingly like the Pre-established Harmony, and like
it in being made to get over a self-made difficulty.
They are hypotheses to cover an hypothesis. The
former alternative does not go beyond conscious beings,
but it fails to explain a fact of consciousness far more
certain than the hypothesis. We are certainly con
scious that external objects somehow affect our feelings
and sensations, that our sensations, desires and infer
ences affect our volitions, that our volitions somehow
affect the motions of our bodies. It is absurd to suppose
that our conscious operations are inert and idle, when
they are consciously both passive and active, and that
the conscious subject is like a child, given his opera
tions like a toy to make believe he is very busy, but
really to keep him quiet. The latter alternative which
carries this inert psychism into everything whatsoever,
without any evidence, except the original hypothesis
of two parallel streams in a sentient being, would have
us believe that the wind blows, the waves swell, the
earth moves, with some obscure sentience. Such a per
sonification of nature was excusable in primitive religion,
but it is not worthy of modern science.

Lastly, to return to the usual hypothesis that nervous
motion produces psychical sensation, which again issues
in nervous motion, one cannot help asking what can be

02 PHYSICAL REALISM TART T.

the source of a biological hypothesis so foreign, nay,
so contradictory to the evidence of biology ? Biologists
have become psychologists, and have fallen under the
dominion of the idealists. Without any criticism, with
out any biological proof, simply because it is the fashion,
and as if it were a first principle, they have accepted
the idealistic hypothesis of purely psychical sensation,
and thereon have reared an hypothesis of their own, that
nervous motion produces this psychical sensation, which
reproduces nervous motion.

Now, the present question, as I said before, is not
whether there is a purely psychical sensation, but
whether there is any evidence that motion propa
gated from the afferent nerves to the brain produces
such a tertium quid, instead of producing motion
from the brain to the efferent nerves. There is no
evidence, either psychological or biological. As a
psychologist, I am conscious that I perform the opera
tion of sensation, which for argument's sake may be
assumed as purely psychical ; but I am not conscious of
my nervous motion. I am not, therefore, conscious of
sensation arising out of nervous motion. A biologist,
not in himself but in another body, can observe a nervous
system, its physical continuity, and the time of its
action proving motion ; but this dissector cannot either
observe or be conscious of the sensation of another
nervous system ; he cannot, therefore, observe nervous
motion issuing in sensation. That there is such a pro
cess from the physical into the psychical and back is
sheer hypothesis, an arbitrary concordance of idealism
and biology,

Nor is this all ; they proceed to suppose that the
effect produced by the external object on the internal
nervous system is not yet sensible, but that, when the

Tin-; PHYSICAL DATA OF SENSE 03

psychical sensation is produced, the effect for the first
time becomes sensible, so that the sensible object is
either identical with the sensation, or at all events is
equally psychical. But, in the first place, even if a
purely psychical sensation is produced in this manner,
it does not follow that the sensible object becomes ,
psychical. There is no reason, except the /old and ex
ploded hypothesis similia similibus cognoscuntur, why
a psychical operation may not apprehend a physical
object. Secondly, whatever may be the nature of the
operation, it is most improbable in itself that the hot felt
through one's body, the white seen through one's eyes,
the loud heard through one's ears, is anything but a
physical condition of the tactile, optic and auditory
nerves in connection with the brain. The idealistic
hypothesis of psychical sensation, then, does not prove
the biological hypothesis of the transmutation of nervous
motion into psychical sensation, nor either hypothesis
the third hypothesis that the sensible object is psychical.

Ter sunt conati imponere Pelio Ossam

Scilicet atque Ossse frondosum involvere Olympum.

The study of the mental physiology of the present
day suggests several reflections. In the first place, the
insensible and imperceptible motions of asther, their
reflections from other bodies, and their impact on the
senses are now well-established discoveries of science.
They are known qualities, which are not sensations, but
the insensible causes of sensations. Not all knowable
qualities, therefore, are sensations. Secondly, as we
recede from the external world behind the periphery
into the nervous system, science becomes more vague.
What are nervous and cerebral motions ? Thirdly, we
are told that cerebral motion, which is physical, pro
duces a heterogeneous sensation, which is psychical.

64 PHYSICAL IIKALISM PART i.

But we are given no evidence of this transmutation. We
cannot observe it in a dissecting-room. If it be said we
are conscious of it, we answer that we are conscious of
sensation, but not of cerebral motion, and therefore not
of cerebral motion producing psychical sensation as a
separate and indeed heterogeneous fact. Fourthly, this
transmutation of one unknown into another unknown
is admitted to take place in an unknown manner.
Fifthly, we are illogically asked to infer from this trans
mutation of cerebral motion into psychical sensation
that the sensible object, e.g. the red seen in vision, is
also a psychical sensation. Sixthly, we are not told
how, if the object of sense thus becomes psychical, we
infer the external causes, which, as we have seen, are
much the clearest part of the whole business. Seventhly,
we often find that, with more logic than consistency,
the external objects which were previously made the
scientific causes of sensation are nevertheless afterwards
declared unknown and unknowable. Meanwhile, the
fallacy of this so-called biology is its assumption of
psychological idealism. All that is really proved by
natural philosophy is that external redness, for example,
is an insensible quality of insensible aether, consisting of
a vibration of a certain velocity ; and that, reflected
by an external object, it produces in the optic nerves of
a sentient being a sensible redness, which is not iden
tical with the external vibrations nor itself a sensible
vibration at all. The simple conclusion from these
scientific facts would be that the nervous effect is the
sensible redness, from which, together with sensible
motion, the external motions of vibration are inferred.
Nothing more is proved by mental physiology.

When we look back at the whole light thrown by
natural philosophy on the sensible object, we shall find

CHAP. in. THE PHYSICAL DATA OF SENSE 65

that it is known that external physical objects produce
internal physical effects in the nervous system, but it is
not known that these internal physical effects in their
turn produce an internal psychical object of sense.
Meanwhile, we are conscious that, when we use our
senses, we somehow apprehend a physical object, which
seems by an illusion to be also external. The simplest
hypothesis, which can be made in these circumstances,
is that the sensible object is neither external on the one
hand nor psychical on the other, but the internal
physical effect on the nervous system. In other words,
there is a via media between intuitive realism and
idealism of all kinds, closer to the scientific facts than
either hypothesis ; namely, physical realism.

At the same time, scientific observation is not a
positive proof of physical realism. It brings the sensible
object within the man : it cannot decide whether it is
or is not within the soul. Its ultimate result is that
the sensible object is not external but internal, not
without but within the sentient being, not identical with
the physical object in the outside world but produced
in the interior microcosm of the animal organism. This
negative conclusion eliminates intuitive, but it does not

o

positively establish physical realism. As a direct
evidence, natural philosophy, being founded on obser
vation, is able to show that the sensible object is not
the physical object outside, but is within the nervous
system ; not being founded on consciousness, it is not
able to decide whether this internal sensible object is
physical or psychical, whether it is the nervous effect,
or something even more internal. It leaves this problem
unsolved. Accordingly, there still remain two pos
sible alternatives — physical realism and psychological
idealism.

F

GO PHYSICAL REALISM TART i.

Nevertheless, scientific observation makes physical
realism the more probable alternative, because this
hypothesis simply accepts the proved nervous effect as
the sensible object, instead of hypothesising a further
psychical object, which is unproved, and breaks the
nervous continuity. When as a mental philosopher one
adds consciousness to scientific observation, the proba
bility of physical realism is increased. Consciousness
tells us that we somehow apprehend physical objects,
which appear also to be sensibly external. Scientific
observation disabuses us of the appearance that the
sensible object is external, but not of the consciousness
that it is physical. Natural philosophy, as a direct
evidence, may be said to remove the physical object
of sense from the external to the internal world, but no
further than the nervous system. The most probable
mental philosophy would simply conclude that it there
becomes sensible — though only the most probable.

We asked for direct evidence that the immediate
object, hot, coloured, &c., perceived by our senses is a
psychical phenomenon, and we find there is none. Con
sciousness is so far from saying so, that it confuses the
immediate and the mediate, and leads us to think that
the immediate object is not only physical but external.
Scientific analysis corrects this confusion, and teaches us
that the immediate object is not external but internal,
but does not go on to show that it is not only internal
but psychical. I suspect that the idealists by a kind of
confusion have changed the truth that the object of
sense is not external but internal into the hypothesis
that it is not physical but psychical.

The idealist may reply that direct evidence is not
required for an hypothesis, and that the psychical object
is like ether — something inaccessible to direct evidence,

CHAT. in. THE PHYSICAL DATA OF SENSE 67

but needed to explain the facts. I accept this issue.'
I admit that, if the idealistic hypothesis of the sensible
object could explain the facts of the known world and
eliminate the hypothesis of physical realism, it would be
proved by this indirect evidence. There would still be
no direct evidence that a hot or coloured object is not a
physical but a psychical fact. But, contrary to all ap
pearance, we should be obliged to conclude that, as
light is paradoxically but really an undulation of aether,
so is the seen, or felt, or heard, or tasted, or smelt, a
psychical, and not a physical, fact.

The gist of the idealistic hypothesis is that not some
but all the immediate objects are psychical, and that
no physical object whatever is apprehended by sense.
The consequence is that all our sensitive experience
will be limited to psychical objects; for, so far as it is
sensitive, experience is merely the sum of our sen
sations. Moreover, the supposition of a priori elements
of knowledge will not help us, for nobody pretends
that we have an a priori apprehension of the physical
to add to an a posteriori apprehension of the psychical :
such an hypothesis would be too great an inversion.
The consequence is that all the data of our knowledge
will be psychical. No doubt different idealists will pro
vide more or less of such psychical data. Some will
have merely psychical sensations, others will add a
psychical subject, and others again psychical apprehen
sions a priori. But at the widest the data will all be
psychical facts of some kind or other.

Now the question arises, what can be known from
psychical data? If all the immediate objects I touch,
see, taste, smell, and hear are psychical, and I am psychi
cal, and all my apprehensions are psychical, if all my
sensitive experience is of nothing but psychical phocno-

F 2

GS PHYSICAL REALISM PART i.

inena, if all the data which form the immediate premises
of my mediate knowledge are psychical, what can I
infer from such facts in the premises ? To answer this
question we must consult the logical rules of inference.
All inference is by similarity. Not to enter into the
question whether there is one fundamental type, there
are three apparent kinds of inference — induction, deduc
tion, and analogical inference. All these are different
modes of reasoning from similar to similar. In induction
we apprehend that similar particulars have a similar
characteristic, and infer that the class, including those
and all other particulars similar to them, have that
similar characteristic. In deduction we start with a
proposition stating the similar characteristics of a class,
either inferred by induction or otherwise known, as
major premise ; we combine it with a minor premise,
asserting that something is one of the class of similar
particulars ; and from this combination we infer that this
new but similar particular has the similar characteristic
already known to belong to the class. In analogical in
ference, which is an imperfect substitute for induction
followed by deduction, we apprehend that a particular
has a characteristic, or several similar particulars have
a similar characteristic; we apprehend by analogy
that another particular is similar to the given parti
cular or similar particulars ; and from the analogy we
infer that this new but similar particular may have the
characteristic similar to that of the given particular
or particulars. Various men are mortal, . • . all men
are mortal : all men are mortal, I am a man, . • . I am
mortal : the earth is inhabited, Mars is like the
earth, . Mars may be inhabited : — these inductions,
deductions, and analogical inferences are nothing but
inferences from similar to similar. They are founded

CHAP. in. THE PHYSICAL DATA OF SENSE GO

also on the reality and knowledge of classes and laws.
But what is a class except similar things, and what is a
law except the fact that similar things possess similar
characteristics ?

From this limitation of inference to similarity it
follows that whatever the character of the data, such
will be that which is inferred. If all the data were
psychical, then, by parity of reasoning, we could only
infer the psychical. If we never had direct experience
of anything physical whatever, then, there being nothing
physical in the premises, nothing physical in the con
clusion could possibly be inferred. From the similar
the similar is inferred ; from the psychical the psychical.
But in order to infer the physical we must have some
physical data.

The universal similarity between the data in the
premises and the inferred in the conclusion requires to
be guarded from misapprehension. I said above that
the old hypothesis— like is known by like — is a fallacy.
I now say that like is known from like. These positions
are not inconsistent. The former refers to the relation
of subject and object, the latter to the relation of object
to object. There is no reason why the object appre
hended should be like the subject apprehending ; but
there are reasons why objects inferred should be like the
objects from which they are inferred— the rules of logic.
If the subject has constantly had physical objects pre
sented to it, it must apprehend them, or be useless.
But when the subject has before it the immediate objects
which can be presented to it, whether a posteriori or
a priori, it has all the data from which reasoning can
start ; and if that reasoning is to maintain the consistency
of truth, it can add nothing in the conclusion which is
not justified by the presence of something similar in the

70 PHYSICAL 11EALISM

PART I.

premises. If reasoning contains, on the Kantian hypo
thesis, a priori apprehensions, these will be part of the
data ; but if it adds anything, not in the data but in
the conclusion, which has no analogue in the premises,
reasoning becomes paralogism. This fallacy is well
known in deduction ; but it is equally true of induction,
which only generalises the subjects and predicates con
tained in the particular instances, and of analogical
idference, which infers that one particular similar to
another may be similar also in a characteristic already
apprehended in that other. Therefore, although like
objects are not necessarily immediately apprehended by
a like subject, only like objects are inferred from like
objects, not by any necessity in the relation of subject
and object, but by the nature of reasoning. Hence a
psychical subject may immediately perceive physical
objects ; but if it were a psychical subject and perceived
psychical objects it could infer nothing but psychical
things, similar either to the psychical subject perceiving
or to the psychical objects perceived.

Again, the logical canon, like is known from like,
must not be confused with the metaphysical hypothesis,
like causes have like effects. Aristotle extended the
principle of the propagation of the species from the
organic to the inorganic world, and thought that every
cause is homogeneous with its effect. Modern science
has discountenanced this view, except in the far-off sense
that all physical causation may be the propagation of
motion in various forms. But when I say that we can
only infer like objects, what I mean is not that we must
infer causes like the effects, but causes like the causes
which we have already known. For example, Newton,
already knowing the effects requiring gravitation to
cause them in terrestrial bodies, when he found similar

CHAP. irr. THE PHYSICAL DATA OF SENSE 71

effects in celestial bodies, inferred that their cause also
is a celestial, similar to terrestrial, gravitation. Now, if
all the data of sense \vere psychical, not only the effects
but also the causes in sense would be psychical : conse
quently, when we came to a sensible effect, similar to
other sensible effects, but not due to any sensible cause,
we should have to infer a similar cause beyond sense ;
and, as all the causes in sense w^ould ex hypothesi be
psychical, we should have to infer, by parity of rea
soning, a psychical cause, not because the effect was
psychical, but because all previously known causes
would be psychical. If, on the other hand, there were
physical causes in the data of sense, we could then, and
only then, infer a similar physical cause beyond sense.

Again, when I say that only like objects are inferred
from like, I do not mean that nothing new can be inferred,
but only nothing new which is not similar to the data.
The conclusion is no mere restatement of the premises.
What is inferred need not have been already experi
enced, nor is reasoning confined to merely reproducing
the immediate data of the senses. But what is inferred
must be similar to what has already been experienced.
What is new, and has never been, nor ever will be, in
experience, such as an ^ethereal undulation, can be in
ferred. But the ethereal undulation is a motion similar
to the experienced motion of waves of water. Nothing
new, which is not similar to the data, can be inferred.
It is true of the Deity Himself, who, though not experi
enced, is inferred to be like man, but infinitely intensi
fied in the attributes which we already know in our
selves. Consequently, if all the data were psychical, we
should be able to draw inferences to similarly psychical
subjects and similarly psychical objects, new but similar
to the data. But we should not be able to infer some-

72 PHYSICAL REALISM PART i.

thing wholly new, dissimilar and heterogeneous, for
which there was no analogue either in the sentient sub
ject or in the sensible objects. Hence, the physical,
for which there would be ex hypotliesi no analogue in
the premises, could not be inferred. If, on the other
hand, as I suppose, the sensible data are physical facts
in my organism, I can then infer new but similar physical
objects outside, although I have never immediately per
ceived them by sense.

Another misapprehension will immediately arise. It
is said that one opposite implies another, and, therefore,
though we experience only one opposite, we infer the
other. Thus, it is supposed, from psychical data we
infer their opposites, physical things. I am almost
ashamed to write down Aristotle's distinction of con
tradictories and contraries ; but it is necessary in an
illogical age. Contradictory opposites are the positive
and its negative, as relative and not relative, finite and
not finite. Contrary opposites are the furthest removed
positives, as white and black. Now contradictory op
posites in a sense imply one another, but contrary
opposites do not. White implies not white : it does not
imply black. We might have apprehended white without
having any conception of black, much less having proof
of its existence. Secondly, great harm is done by such
vague terms as ' imply ' and ' implication,' which some
times mean conceiving and sometimes inferring. The
positive, when apprehended, makes us conceive the con
tradictory negative, but does not make us infer that it
exists. Are we to fall into the old sophism of arguing
that as something is contradicted by nothing, nothin^

o «7 O" O

exists ?

It is a common argument that the relative which
we experience implies the non-relative and absolute,

CHAP. in. THE PHYSICAL DATA OF SENSE To

the finite implies the infinite. This is an utter confu
sion of contradictories and contraries. The relative im
plies the not relative ; but the contradictory, not
relative, is not necessarily the positive contrary, abso
lute, for it also includes nothing ; and the relative, in
implying the not relative, does not decide whether it is
absolute or nothing. As white implies not white, but
not necessarily black or any other particular colour, so
the relative implies not relative, but not necessarily the
particular species of not relative, absolute. The same
remark applies to the opposition of finite and infinite,
except that in this case the term ' infinite ' is ambiguous,
being properly the not finite, but including both that
which is not finite, because it is nothing, and that
which is not finite, because it extends without limit.
The finite implies its contradictory, not its contrary : it
implies the negative not finite, but does not imply
the particular positive species, the infinite which
extends without limit. Secondly, the relative and
finite imply only in the sense of making us conceive
the mere contradictories, not relative and not finite.
The positive sides of the contradictions not only leave
the content of the negatives undetermined, but also
leave the question undecided whether we can infer
that there is anything corresponding to the ideas of
the negatives. Nor do they even give us the ideas
till we have not only apprehended the positives, but
also apprehended that they are relative and finite.
The relative and the finite, then, when apprehended to
be such, make us conceive the ideas of the not relative
and not finite, but give us no idea of a positive some
thing absolute and extending without limits, much less
make us infer that this species of not relative and not
finite is something real as distinguished from nothing

74 PHYSICAL TIEALISM

at all. When we merely experience something which
happens to be finite, we need not think of any opposite ;
if we think of it as finite, we must have an idea of the
not finite ; but we need not form an idea of the positive
infinite, much less can we prove that there is something
infinite, and say, ' I experience the finite and relative,
therefore there is an infinite and absolute.' Men accept
such arguments because they think it helps to prove
the existence of a Deity. But the finite and relative do
not make us conceive a positive infinite absolute, much
less infer its existence ; and theology has better argu
ments for a Deity than the confusion of negative and
positive, of contradictory and contrary opposition, of
conception and inference, of ideas and judgments.

Similarly, the psychical does not imply the physical.
The physical and the psychical are contraries, not con
tradictories. The contradictory of the psychical is the
not psychical, which may be anything else or nothing.
Suppose that I had experienced nothing but psychical
data. If I had never thought of them as psychical, but
only as hot, red, and so on, I should have had no reason
to conceive the not psychical. If I had thought of them
as psychical, I must then have had the bare idea of not
psychical as its contradictory. But I should neither
have been able to have inferred that it existed nor what
it was. The content of the idea would have been the
bare negation or contradictory of the psychical. I
should have had no idea of the physical as a positive
contrary, much less have proved its existence. Just as
the apprehension of white makes me conceive the idea
of not white, but does not infer that there is any other
colour, much less the contrary black, and just as the
apprehension of the relative and finite makes me con
ceive the idea of not relative and not finite, but does not

CHAP. in. THE PHYSICAL DATA OF SENSE 75

infer that there is anything which is not relative and not
finite, much less the contrary absolute and extending
without limits, so the apprehension of the psychical
would make me conceive the idea of not psychical, but
would not tell me that there is anything positive which
is not psychical, much less that it is the contrary,
physical. To infer the existence of the positive con
trary, the physical, I should have required other than
psychical data, which would, however, have been ex
hypotliesi all the data possible.

In all cases the existence of a contrary is a matter not
of implication in the knowledge of the opposite con
trary, but a matter of independent inference. Human
reasoning would indeed be easy, if without further
question the moment one had ascertained a thing, one
knew that its contrary existed ; when one had experi
enced white, one knew black ; when all experience had
been of the relative and finite, one knew the absolute
and infinite ; when all the immediate data of all reason
ing were psychical, one straightway knew that there
are physical things. Why, one contrary does not even
make us conceive the idea of another, much less infer its
existence. The white makes us conceive the idea, not
white : we want other evidence to infer the existence of
the black. The psychical makes us conceive the idea,'
not psychical : we want other evidence to prove the
existence of the physical. A synthesis from psychical
data to physical things must be founded on some better
device than the fallacy of the implication of opposites.
But in reality the whole hypothesis of such a synthesis
is illogical. To infer physical things we require more
than psychical data, and their implications, and their
consequences : we require physical data in the premises
similar to the physical objects in the conclusion.

7G PHYSICAL REALISM TART i.

The canons of inference, then, teach us, first, that
from similars similars are inferred ; secondly, that what
is inferred may be something new so long as it is similar
to some of the data ; and thirdly, that it cannot be the
contrary of all the data. Therefore, on the idealistic
hypothesis that all the data are psychical, in the first
place, what is inferred would also be psychical ; secondly,
it would include other psychical subjects and other psy
chical objects similar to those which ex Jiypotliesi form
the data of inference ; but, thirdly, it would not include
physical things, for which there would be no analogy,
and which are not implied in merely psychical data : for
psychical data would not make us even conceive, much
less infer their contraries, physical things. On the
other hand, if some of the data are physical, what is
inferred can be physical like the data, different yet
similar objects, the data being in our own bodies, the
inferred objects in the external world.

We constantly hear at the present day of two worlds
and their correspondence — the psychical and the physical.
It is not the purpose of this essay to deny this anti
thesis, nor to depend upon it. But it is also commonly
supposed that all the data of our knowledge belong to
the former world, from which the latter is inferred.
Against this hypothesis I direct this essay. If all the
data of sense were psychical, the parity of reasoning
would have no data to infer the physical. But the
physical world is the object of natural science, which is
knowledge. Therefore, not all the data of sense are
psychical. There must be similar physical data to infer
similar physical objects.

Such, then, are the data required by the rules of
reasoning to infer a physical world. We began by
saying that, if the idealistic hypothesis led to the

CHAP. TIT. THE PHYSICAL DATA OF SENSE 77

only possible explanation of the facts, we must accept
it even on this indirect evidence. We now see to
what it logically leads. All that is inferred as well
as all that is perceived, all that is immediate and all that
is mediate, all that is apprehended in us and all that is
known beyond, will be psychical. That is, all known
realities will be psychical facts of some kind or another.
As Berkeley says, the whole known world will be mind
and ideas ; with Hegel, thought will be being and
being will be thought. These are the logical idealisms.
Nothing physical, and not psychical, will be inferrible,
still less knowable.

This logical consequence of all psychological idealism
must be confronted with the discoveries of natural philo
sophy. A survey of these discoveries shows an enormous
mass of insensible and inconceivable realities, which
are scientifically known by inference from sensible data.
But they are physical realities, incapable of being re
solved into any kind of psychical fact ; being insen
sible they are not sensations, being inconceivable they
are not ideas. It follows, therefore, that some things
physical, and not psychical, are knowable, and not all
known objects are psychical.

The physical objects of scientific knowledge directly
eliminate pure idealism. Starting synthetically from
the common idealistic hypothesis that the sensible data
are psychical, the pure idealist draws the strictly logical
conclusion that all known objects, inferred from these
psychical data of sense, must also be psychical. Accord
ing to him, then, there are no physical objects of know
ledge. His logic is consistent, but his conclusion is
false. He has omitted the physical world which, being
beyond our sensations and ideas, cannot be resolved
into sensations or ideas, but which yet is an object of

78 PHYSICAL HKALISM TART r.

8(.ience_the most perfect form of knowledge. Not all
known objects, therefore, are psychical; some are phy
sical. Pure idealism then is false, and some form of
realism true. As intuitive realism has already been
eliminated by natural philosophy, it only remains to
decide between the hypothetical realism of the cos-
mothetic idealist and the physical realism of this essay.
The physical objects of scientific knowledge in
directly eliminate cosmothetic idealism with its hypo
thetical realism. The cosmothetic idealist tries to
reconcile the idealistic theory, that the sensible data
are psychical, with the realistic theory that some objects
knowable by inference from these data are physical.
We have found that the realistic part of his theory is
correct. He has the merit of admitting that there are
physical objects of knowledge : this is his superiority to
the pure idealist. He has the merit of admitting that
they are not intuitively perceived by sense, but inferred :
this is his superiority to the intuitive realist. But he is
illogical. His defect is the inconsequence of supposing
that physical objects, though not intuitively perceived,
could be inferred from purely psychical data. But we
have seen that all inference is by similarity, and there
fore physical objects could not be inferred from purely
psychical data. The physical would be the object
of a new term in the conclusion, absent and un
justified in the premises. If all the data of sense
were psychical, then, by parity of reasoning, all objects
knowable from them would be psychical. But by the
discoveries of science, and by the admission of the
cosmothetic idealist, some objects knowable by inference
from the data of sense are physical. Therefore not all
the data of sense are psychical. Sublata consequent
tullitur antecedens.

CHAP. in. THE PHYSICAL DATA OF SENSE 79

Cosmotlietic and pure idealism are mutually destruc
tive of each other. The former admits that some objects
are physical, which prove that the latter is wrong in
supposing all objects to be psychical. The latter admits
that only psychical objects can be inferred from psychi
cal data, so that the former is wrong in supposing that
physical objects are inferred from psychical data. Pure
idealism fails to recognise, cosmothetic idealism fails to
explain, the knowledge of an insensible and inconceiv
able physical world. If we combine both we destroy
the common data of both. As the pure idealist says,
if all the data were psychical all the objects would be
psychical ; but as the cosmothetic idealist admits, not
all the objects are psychical. It follows that both are
wrong in saying all the data are psychical. Their data
fail to explain the physical objects of scientific know
ledge. Science eliminates all psychological idealism.

Meanwhile the physical objects of scientific know
ledge are not merely destructive of psychological
idealism, but are also constructive of physical realism.
They prove in themselves that some objects of know
ledge are physical, and, in combination with the logical
rules of inference, that some data of sense must be
physical, to infer them. Similars are inferrible only
from similars. Therefore the physical is inferrible only
from the physical. But some objects of science are
physical ; therefore they are inferrible only from physi
cal data. These data of sense, however, though physical,
are proved by scientific analysis to be internal ; there
fore the data of sense are physical objects within our
nervous system, from which we infer physical objects in
the external world. This is the theory of physical
realism, established by the logical rules of hypothesis.

I admit that the direct evidences are not a positive

80 PHYSICAL REALISM PART i.

proof of physical realism. Consciousness, alone, is even
in favour of intuitive realism. But scientific analysis
destroys this hypothesis by separating the sensible
effect from the external cause, and showing that the
sensible object must be internal. On the other hand, it
does not show that the sensible object is not only in
ternal but psychical, and therefore does not favour
idealism. It makes the intermediate theory of physical
realism possible, even probable. I do not believe, how
ever, that the data of sense are recoverable by any
direct method, because from our very birth, and with
inherited power, we overlay them with inferences.
Hence the shipwreck of modern philosophy, which sup
poses its hypotheses of sensible data to be first principles,
and has alternated between the opposite but equally
futile attempts to grasp physical things by sense, or to
leap from psychical data to physical things.

I admit, therefore, that the crucial evidence must
be indirect. That hypothesis of the data of sense must
be accepted, which explains the knowledge of the objects
of science. This insensible, this inconceivable, this
physical world of science is not an object of intuition,
is not a sum of psychical sensations and ideas, is not
inferrible from psychical sensations and ideas. Its
knowledge then must be accounted for otherwise. It
is inferrible from internal and physical data, the nervous
system sensibly affected by external objects. The data
of sense, then, are neither physical objects without,
which are the causes not the objects of sense ; nor
psychical objects within, from which nothing physica
could be inferred ; but physical objects within, from
which physical objects without are inferred by all, and
known by science. Physical realism, therefore, or the
theory of internal physical data to infer external physi-

CHAP. in. THE PHYSICAL DATA OF SENSE 81

cal objects, is, in accordance with the logic of explana.
tion and elimination, the only hypothesis of the data of
sense sufficient to explain the knowledge of the objects
of science. It is a mental philosophy born of natural
philosophy, ' that great mother of sciences.' '

1 Bacon, Nov. Org. i, 80.

82 PHYSICAL REALISM rAiir r.

CHAPTER IV.

THE HISTORICAL OKIGIN OF PSYCHOLOGICAL IDEALISM.

tr

ABISTOTLK remarks that we ought not only to criticise
our opponents, but also to point out the causes of their
errors. The origin of intuitive realism and its presen-
tative theory of perception, is the inevitable tendency
of ordinary man to confound sense with reason, and his
sensations with his inferences. He has so long been
accustomed to infer an external world, that at last he
cannot but fancy his senses perceive it. He seems to
himself even to be conscious that it is so, calls his con
fusion common sense, and at last defies philosophers to
distinguish the sensible and the real. To have dis
abused philosophy of this confusion is one of the
many services owed by mankind to Greek philosophers.
The distinction of sense and reason soon dawned on the
Greeks, and with it the discovery that the object of
sense is not the external thing at a distance from our
selves, but some sort of result on our senses, from
which the external tiling is inferred by reason. In
short, the Greek philosophers founded the representa
tire theory of sensitive perception. But they did not
agree about the nature of the sensible object, or repre
sentative of the external thing impressed on the senses.
Without pretending to give a history of their views, we
may distinguish two great epochs : the first, that in
which the sensible object was regarded as a corporeal

CHAP. iv. OlilGIN OF I'SYCJIOLOGICAL IDEALISM

effect ; the second, that in which it began to be regarded
as an incorporeal essence in our senses. In this second
epoch the Greeks prepared the way for the theory that
the sensible object is an incorporeal idea. But they
never actually reached the idealistic theory.

The first approach to a scientific theory of the
objects of knowledge is to be found in the Atomists,
Leucippus and Democritus of Abdera, the pioneers of
a sound philosophy of nature. To them we owe the
dawn of the truth, afterwards developed into the dis
tinction of primary and secondary qualities, that the
real and original qualities of particles are figure, position
and arrangement, whose different combinations, together
with motion, give rise to qualities, such as heat and
colour, which, though really derivative, appear equally
original to our senses.

The manner, however, in which this important doc
trine was presented to the world was not purely unex
ceptionable. The Atomists, it is true, admitted that
there is for every variety of sensible quality a distinct
mode, or schema in their language, of the original qua
lities ; for example, a sharp taste arises from angular,
a sweet from round schemata. But, to say nothing of
their crude speculations on corpuscular structure and
motion, they fell into the fallacy of confusing the deri
vative quality with its sensible effect in the famous
dictum, ' Conventionally there is sweet, conventionally
bitter, conventionally hot, conventionally cold, conven
tionally colour ; but really atoms and void.' * From
this Atomistic identification of secondary qualities with
their sensible effects, assisted by the Heraclitean identity
of contraries, it was but a short step to the sceptical
theory of Protagoras, that all qualities are merely the

1 Scxt. Einp. Adv. Math. vii. 135.

G 2

84 PHYSICAL REALISM TART i.

appearances in our senses, without any correspondents
in the fluent matter of nature.

The Atomists did not recognise sufficiently, the
Sceptics not at all, the fact that derivative or secondary
qualities are qualities of external things. There is
also a common tendency in modern mental philo
sophy to identify secondary qualities with their sen
sible manifestations. But for every sensible quality,
which is the product of an external object, there is a
distinct quality in the external object. A primary
quality is also like the sensible quality. A secondary
quality, such as heat or colour, is not, indeed, like the
sensible effect, being a mode of a primary quality, such
as motion ; but it is a distinct and specific variety of
that primary quality ; it is the motion of a different kind
of matter, it goes on independently of the sensible effect,
and it is a knowable object of science. Thus, it has
been discovered in natural philosophy that heat and
lio-ht are not molar but molecular motions, that they

o

are motions of sether ; that they are, in rerum natura,
different motions of different lengths, the waves of mere
heat being longer than those of light, and that they are
so disseminated throughout the universe as to produce
no sensible effect incalculably oftener than they excite
touch or vision.

It was perceived by the genius of Bacon that heat is
of two kinds, in or dine ad universum and in ordine ad
sensum, the former being an insensible mode of corpus
cular motion, the latter the same thing but with a rela
tion such as -is competent to sense.1 The Atomists were
too narrow in confining heat to the sensible effect of a
distinct mode of matter in the external world, and
Protagoras quite wrong in denying the distinctness of

1 Bacon, Nov. Org. ii. 20.

CHAP. iv. ORIGIN OF PSYCHOLOGICAL IDEALISM 85

the external quality ; Bacon was right in regarding heat
as a mode of motion in the external world, as well
as a sensible quality in our senses. So with all other
secondary qualities ; they are modes of primary qualities,
but distinct modes ; they have a generic resemblance to
other modes, but they have also specific differences.
Sound is a vibration of air, heat and light undulations
of rcther.

The only plausible objection to this view would be
that the names 'heat,' 'light,' and so on, should be
confined to the sensible effects and not extended to
their external causes. It must be confessed, also,
that so long as distinctions of things are observed,
the use of names is comparatively unimportant. But
names are the vehicles of distinct ideas, and it is the
duty of every science to have some distinct name for
every real distinction of things. The specific modes
of primary qualities must receive some name or other.
It will not suffice to leave the external cause of sensible
sound to the periphrasis, vibratory motion among the
particles of an elastic aerial medium ; or that of light to the
periphrasis, undulations in an ethereal medium per
vading interstellar spaces and bodies formed of ponder
able matter. New names might be invented, but they are
not forthcoming, and it is doubtful whether they would
be superior to, and still more doubtful whether they
would be victorious over, the old names, ' sound ' and

' light.'

Secondary qualities are real, though derivative,
qualities of external objects, as well as qualities of
sensible objects ; and their names should be equally
extensive. In support of this view, let us quote a
passage from Professor Stokes, ' On the Beneficial Effects
of Li^ht,' all the more valuable because it was not

80 PHYSICAL REALISM PAKT i.

written to support any general philosophy of secondary
qualities :—

' Beyond both ends of the visible spectrum there lie
radiations which do not affect the eye, but are never
theless, as we have every reason to believe, of the same
physical nature as those which do, from which they do
not differ by any inherent quality. As the agent which
excites vision has been called from time immemorial
" light," or whatever may be the corresponding term in
other languages, it will be convenient to use the same
word to designate the agent considered in itself, and
irrespectively of its capacity for exciting vision, a
capacity which would be regarded as a mere accident
of light, in the technical logical sense of that word.
Accordingly I shall now use the word " light " to
designate what, for want of a better term, I have just
been calling "radiation," a word which would more
properly denote the process of radiation than the thing
radiated, be it the material or immaterial, be it matter
or undulations ' (p. 6).

Qualities, then, as distinguished by natural philo
sophy, are divided as follows : —

I. External, in or dine ad universum.

1. Primary, original qualities; e.g. duration,

extension, motion.

2. Secondary, specific modes of primary quali

ties ; e.g. sound, heat, light, as modes of
motion.

II. Internal, in or dine ad sensum.

1. Primary, and like external primary qualities,

which cause them.

2. Secondary, unlike external secondary quali

ties, which cause them.

CHAP. iv. ORIGIN OF PSYCHOLOGICAL IDEALISM 87

It is to be noticed, in this division, that the derivative
character of secondary qualities refers not to their
sensible but to their external aspect. As sensible, we
apprehend them in exactly the same way. Again, the
ordinary man infers external qualities alike in both
cases. The difference entirely arises when the scientific
man begins to infer that external secondary are modes of
primary qualities, because their sensible effects are so
similar to those of primary qualities; for instance,
the effects of external sound, heat and light are
effects of motion by the laws of motion.

To the Atomists is due, not only the foundation
the theory of primary and secondary qualities, but also
the discovery that the object of sense is not the external
thin- itself, but an effect produced by the external thin
on the senses. They supposed that effluxes, continual!;
thrown off from bodies, come into contact with on
oro-ans.1 They thus anticipated modern physical inquiry
on the senses, although their necessary ignorance of the
laws of motion prevented them from realising the vibra
tions and undulations, which have taken the place oi
emissions, in the case of hearing, sight, and the perception
of temperature by touch. The consequence of this
position to the theory of knowledge in Greek philosophy
was that its immediate object was henceforward
rally agreed to be not the thing at a distance, 1
result of the thing on the organs of sense.

In the Atomistic theory the immediate object ^ c
sense, though internal and representative, is neither im
material nor psychical : it is a physical object,
point has never been disproved. Modern physiology
we have seen, has brought the motions of matter j

i Arist. DC Divin. per Somn. 2 = 4G4 A G (Berlin cd.) ; cf. Pint. DC Plac.
Phil iv. 8.

88 PHYSICAL REALISM PART i.

as the physical substances of the nerves ; but it has
never shown that this physical object is converted into
a psychical sensation, either at the extremities of the
nerves, or in the nervous fibres, or in the nerve centres,
or in the brain itself, or beyond it. Why, then, should
we not perceive the physical effect in our internal

organs ?

The physical character of the immediate object of
sensible knowledge was not at first forgotten. It sur
vived in the Epicurean philosophy. It even left a relic
in the philosophy of Plato, who always represents
sensation as a motion communicated from matter
through body to soul.1 Hence sense never appears in
any Platonic dialogue as a part of the soul, nor the
sensible object as something purely psychical. It is not
in his theory of sense, but of reason, that Plato becomes
idealistic. The objects of sense are, according to him,
results of material motion communicated from body to
soul ; the objects of rational knowledge are results com
municated from immaterial ' forms ' to the pure soul.

Aristotle was the author of a new theory of the sen
sible object. He had an aversion to atomism, perhaps
because he confused it with materialism. For atoms he
substituted primary matter ; instead of figure, position,
and arrangement, he regarded heat and cold, dry and
liquid, as its primary contrarieties.2 The Atomists
considered the external thing to be wholly corporeal ;
Aristotle divided it into two heterogeneous substances
— corporeal matter and incorporeal form 3 — the former
of which was different for each individual, the latter the
same for all individuals of one kind. While the Atomists
had held that the sensible object which results from the

1 Plato, Phil 34 A ; Tim. 42 A, G4. 2 Arist. DC Gen. et Corr. ii. 1.
3 Id. Met. Z 7 = 1032 B 14.

CHAP. iv. ORIGIN OF PSYCHOLOGICAL IDEALISM 89

external thing is a corporeal efllux, Aristotle persuaded
himself and his followers that it is the identical incor
poreal form transferred without the different corporeal
matter from the external thing into the sensitive faculty,
as an impression is transferred without the metal from
a metallic seal into wax.1 For example, vision, accord
ing to him, receives the essence of white without the
matter of the external wax into the visual faculty.
Hence his distinction of nutrition and sensation : in
nutrition we receive the whole thing, in sense the form
without the matter of the thing. He agreed, indeed,
with his predecessors in the fundamental point that the
external thing is not presented, but that the sensible
object presented is a representative result of the external
thing. But this object in our senses, which, according
to the Atomists, was a corporeal efflux, was, according
to Aristotle, an incorporeal form, called by himself
ala-O^rov elSo?, and by his scholastic followers, species
sensibilis. From his time onwards, the object of sense
began to be usually regarded as not only internal, but
also incorporeal, though not yet as a purely psychical
object.

Aristotle's new theory of the object and nature of
sensitive perception is charged with errors. He substi
tuted for the explanation of the world by particles, the
abstractions of matter and form ; he inverted the real
order of primary and secondary by making heat and
cold original qualities ; he arbitrarily severed a single
corporeal thing into a corporeal and an incorporeal half,
and by this latter figment endeavoured to explain the
object of sense. We see here the beginning of the false
hypothesis that the object of sense is not a corporeal
fact. Aristotle was right in thinking that sense does

1 Arist. De An. ii. 12.

90 PHYSICAL REALISM TART i.

not perceive the external thing, wrong in thinking that
what it perceives within is an incorporeal form.

Hamilton has misunderstood these Aristotelian errors.1
He says truly enough that Aristotle distinguishes proper
from common objects of sense,2 and that the former
agree with the secondary, the latter with primary quali
ties. 13 ut he misses the real point by supposing that
Aristotle meant to derive the former from the latter.
Aristotle distinguished proper and common sensibles
solely in relation to the senses which perceive them.
Heat and cold, for example, are proper sensible objects
of touch ; but so far from being regarded by Aristotle
as secondary qualities, they form one pair of his primary
contrarieties of matter. The classification into common
and proper is not intended by Aristotle for a classifica
tion into primary and secondary ; so far from it, his
primary qualities are falsely taken from what are really
secondary qualities, heat and cold, dry and moist.

Secondly, Hamilton rightly says that Aristotle calls
such qualities as heat and cold affective qualities, be
cause they produce affections in us.3 But we must not
therefore infer that lie meant either that they produce
this effect through insensible primary qualities, or that
they are themselves mere affections in us, or that, being
qualities outside, the affections are not like them. These
are opinions of people who hold an atomistic theory
of primary and secondary qualities, but they are not
Aristotelian. In fact, the most fundamental defect in
Aristotle's natural philosophy is the supposition that
heat and cold are primary contrarieties of matter in
capable of further resolution. His opinion was that

1 See Eeid's Works, ed. by Hamilton, Note D, on Primary and
Secondary Qualities.

2 Arist. De An. ii. G. 3 Id. Cat. 8 = 9 A 28 seq.

ORIGIN7 OF PSYCHOLOGICAL IDEALISM (,H

heat and cold are real and original qualities of matter,
derived from no others, and that they produce in us
affections of heat and cold similar to themselves. This,
morever, was his theory of the perception of all
qualities.

Thirdly, Hamilton is right in saying that, according
to Aristotle, there is an identity between the external
object and the object perceived.1 But he is wrong in
inferring from this identification that, according to
Aristotle, the external object is presentatively perceived
without any intermediate object. The identity is not of
existence but of essence, not numerical but specific,
not numero but specie. Aristotle supposed that in all
members of a kind there is one form, and that, when
one member of a kind produces another member, it pro
pagates the form, or, as we say to this day in organisms,
the species, from its own matter to the matter of the
new recipient of the form or species. Thus he supposed
man to beget man.2 Hence, in sensible knowledge, he
supposed that the external object propagates the form
of the sensible quality, such as heat, without its own
matter into the matter of the sense, which thus receives
the form or species of heat into its own matter without
receiving the matter of the body which propagates the
heat. Therefore the hot body and the hot affection of
sense are the same only as the impression on the seal
is the same as the impression on the wax, or as the
father is the same as the son ; that is, the same in form
or essence, not in matter or existence, the same specie
but different numero, like but not the same objects.

According to Aristotle, then, the sensible object is
not numerically identical with the sensible object, but

1 Arist. De An. in. 2 = 425 B 25-7.

2 Id. Met. Z 7-8, esp. 1033 B 29-1034 A 8.

92 PHYSICAL REALISM PART i.

only identical in essence. It is the form or species,
without the matter of the external object,, propagated
into the senses. Aristotle was no intuitive realist. He
held, indeed, that sense perceives the identical essence
of the external thing, but not the external thing itself;
and he held that it receives this essence into the sensi
tive faculty, and does not apprehend it in the external
world. In short, his theory was a new form of repre
sentation, in which the object of sense was regarded
no longer as a corporeal efllux, but as an incorporeal
essence received without the corporeal matter from a
corporeal object into the senses, and there perceived.

As the objects of sensible knowledge are sensible
species, so the objects of rational knowledge are intel
ligible species, according to Aristotle. The difference
is in the mode of production. The former are propa
gated by external objects into the sensitive faculty, the
latter by active intelligence into passive intelligence.
Aristotle has not explained this mysterious influence of
intelligence on intelligence in the same soul ; nor is it

O o

probable that he proceeded on any other fact than the
consciousness that, while we depend on externals to
perceive, we can command our own thoughts. It would
be, however, useless to go into this question. The
important point for our present purpose is that both
sensible and intelligible species are, in the view of
Aristotle, immaterial, not material, objects. In his philo
sophy, for the first time, we come to the view that all
the immediate objects of knowledge are immaterial facts.
We must not therefore fly to the supposition that
Aristotle thought them to be psychical because they
were immaterial. We have not yet exhausted the mys
teries of the Aristotelian form. A form is supposed
by him to be not only one in connection with many

CHAP. iv. ORIGIN OF PSYCHOLOGICAL IDEALISM 93

matters of different members of the same kind, but also
to be something different from matter, even when so
closely conjoined with matter in fact, and so inseparable
from it in definition, as concavity with nose in snubnose,
and soul with body in an animal. Every form, the form
of a triangle, the form of a stone, the form of a house,
is an immaterial substance, even when conjoined with
matter in a material substance. The form of God Him
self is pure, not in the sense of being less material than
other forms, but only in the sense of never being con
joined with matter. Hence, sensible and intelligible
species or forms are immaterial, not because they are
in the soul, but simply because all forms are immaterial,
according to Aristotle, who thought that if I per
ceive a white paper, I receive from the paper into my
sensitive faculty an identical essence of white, which
was already incorporeal in the paper before it was com
municated to the sensitive faculty of my soul. The
object of sense, then, had, in his philosophy, ceased to
be material, but had not yet become a psychical fact :
it is an essence, which is not matter, whether it is
without or within a soul.

Descartes completed the separation of the sensible
object from the external world. The Atomists had
taken the first step by discovering that the object of
sense is not the external thing, but an internal effect ;
but they admitted that it is, like its external cause,
purely physical, and no more has been proved to this
very day. Aristotle, however, had proceeded to apply
the hypothesis of incorporeal forms to sense, and sup
posed that the object of sense is a sensible species,
similar to the physical cause in identical incorporeal
essence, but not in diverse corporeal matter. It remained
for Descartes to take the final step and destroy the last

94 PHYSICAL REALISM TAUT i.

vestige of resemblance to the physical cause by identi-

O -L " J

fying the object of sense with a psychical idea.

The history of philosophy had insensibly led, or
rather misled, Descartes into his ideal theory. In the
philosophy of Aristotle the incorporeal is wider than
the psychical, because all essences are incorporeal even
in physical things. But in the interval between ancient
and modern philosophy, the hypothesis of the incor-
porealism of essences was discredited, partly by the
attacks of Nominalism, but more successfully by the
revival of natural philosophy, and especially by the
return to Atomism, inaugurated by Bacon, from whom
it passed to Descartes. Bacon discovered that the
essence of anything physical is nothing but a uniform
mode of its matter.1 Descartes thought that it is only
a psychical idea.'2 In these circumstances his hypo
thesis of the sensible object developed itself, as it
were, from the course of history. The sensible object
had been identified by Aristotle with the incorporeal
essence ; the incorporeal had been recently expunged
by Bacon from the physical world ; the essence was
limited by Descartes himself to the psychical idea.
What more natural than to regard the sensible object
also as a psychical idea ?

Descartes, it is true, went back to the Atomists
for the analysis of nature into corpuscles. lie might
also, especially since Galen's discoveries in the nervous
system, have restored the Atomistic theory that the
object of sense is a physical effect on our organs, and
have added that it is an effect on the nervous system.
His writings do, indeed, show that he was not always
certain whether the sensible effect is physical or psy
chical. Sometimes he even seems almost to express

1 Nov. Org. i. 51 ; ii. 17, 20, 52. 2 Princ. i. 58.

CHAP. iv. ORIGIN OF PSYCHOLOGICAL IDEALISM 95

himself as if the idea itself were not distinct from the
nervous imprint. But he finally and deliberately sepa
rated it from the physical effect in the brain in his
Replies to the Objections raised against his Medita
tions. The ' Eesponsio ad Secundas Objectiones ' con
tains a synthetic statement of reasons for the exist
ence of God, arranged in geometrical order, and the
second definition is a formal definition of the idea, as
follows : —

6 By the name, Idea, I understand that form of any
thought, by whose immediate perception I am conscious
of that same thought ; so that I can express nothing
in words in understanding that which I say, but that
from this very fact it is certain there is in me an idea
of that which is signified by those words. And so
it is not only the images depicted in the fancy that I
call ideas : nay, these I here by no means call ideas,
so far as they are depicted in the corporeal fancy,
that is, in some part of the brain, but only so far as
they inform the mind itself turned towards that part
of the brain.'

The influence of Descartes did not at once make
itself felt in all parts of philosophy. English natural
philosophy in this as in other matters took an indepen
dent course, which accounts for one finding Aristotle's
theory of the sensible object surviving in Newton's
Optics. In Quasst. 20 Newton asks : ' Aiinon sensoriuni
animalium est locus cui substantia sentiens adest, et
in quern sensibiles rerum species per nervos et cerebrum
deferuntur, tit ibi prsesentes a prsesente sentiri possint ? '
Similarly English theology did not at first think it
necessary to salvation to consider sensible objects, or
sensation, or even consciousness itself, to be psy
chical, as we may see from the following passage in

96 PHYSICAL REALISM

TART I.

'Tritlieism charged upon Dr. Sherlock's new Notion
of the Trinity,' by a Divine of the Church of Eno*-
land l :-

' I deny that there is any such thing as sensation,
whether internal or external, belonging to spirits not
vitally united to organised bodies. For sensation is
properly the perception of a sensible object by a sensible
species of it imprinted upon and received into the proper
organ by which each sensitive faculty operates and
exerts itself. This, I say, is sensation, and accordingly,
as it is external or internal, so it has external or internal
organs allotted to it; but still both of them corporeal.
And therefore for this man to talk of spiritual sensa
tion is nonsense and a contradiction in the terms, and
consequently not to be allowed' (p. 15).

But mental philosophers, not only on the Continent
but also in England, more quickly received the hypo
thesis of a psychical object of sense. At first, Locke
simply accepted the Cartesian ' idea.' Then Hume dis
tinguished the ' impression ' from the idea. Kant made
6 phenomenon ' the fashionable term. Mill preferred
' sensation.' But all agree in some psychical object or
other. Moreover, mental physiologists have passed over
from Aristotle and Newton to Descartes, when they
ought rather to have retraced their steps from Newton
through Aristotle to the Atomists. The Cartesian hypo
thesis, that the object of sense is a purely psychical idea,
is not so near the truth as Aristotle's hypothesis, that
it is an incorporeal but not psychical species in the
sensitive faculty ; nor is the Aristotelian so near as the
Atomistic hypothesis, that it is a purely physical effect
on the bodily organs. All that is required to make
this last, or rather this first, the truth is to substitute

1 Dr. South.

CHAP. iv. ORIGIN OF PSYCHOLOGICAL IDEALISM 97

for eilluxes mechanical motions, and for the bodily
organs the nervous system sensibly affected.

' Meanwhile,' says Bacon, ' let nobody expect great
progress in the sciences (especially on their productive
side) unless natural philosophy has been extended to
the special sciences, and the special sciences reduced to
natural philosophy. Hence it happens that astronomy,
optics, music, most of the mechanical arts, and (what
may seem more strange) moral and political philosophy,
and the logical sciences, have little or no extent in
depth, but only slide over the surface and variety
of things : because, as soon as those special sciences
have been divided and established, they are no longer
nourished by natural philosophy ; which, from the
sources and true contemplations of the motions, rays,
sounds, texture, and structure of bodies, affections, and
intellectual apprehensions, had been able to impart to
them new force and increase. It is not at all wonderful,
if sciences do not grow, when they have been separated
from their roots.' 1 The revival of Atomism by Bacon,
together with the gradual establishment of the laws of
motion in mechanics, from Galileo to Newton, produced an
instauration of natural philosophy. Let us now, in the
same spirit, return to natural philosophy, in order to
restore mental philosophy. ' Iiiteritus rei arcetur per
reductionem ejus ad principia.'

1 Bacon, Nov. Org. i. 80.

H

PART II.

PSYCHOLOGICAL IDEALISM.

' Pcssimwni cnim omnium cst aiigurium quod ex Conscnsu capitur in rclus
IntcllcctualiltusS

BACON, Kov. Org. i. 77.

H 2

101

CHAPTEE V.

DESCARTES.

PHILOSOPHY ought to begin in doubt. But I cannot
doubt that I think. Cogito, ergo sum. As a thinking I
being, I am a soul, distinct from the body. Soul is ;
thinking substance; body is extended substance; they ;
are heterogeneous to each other. The soul immediately
apprehends ideas, innate, adventitious, and fictitious.
The clearness and distinctness of ideas are a criterion of
truth, and by the veracity of God, enable me to know
objects beyond ideas. Starting from ideas, I infer a
physical world of bodies and insensible corpuscles, whose
qualities are partly like and partly unlike those which I
perceive as sensible ideas, and whose insensible modes
produce sensible ideas. These are the cardinal points
of Cartesian idealism.

Cogito, ergo sum. — I think, therefore I am. This is
the indubitable fact, which Descartes had the undying!
merit of elevating into a principle in mental philosophy.)
The proposition was not new. Aristotle asserted our
consciousness of our operations,1 and even recognised
this fact as a proof of our existence. But he did not
convert the proposition into a psychological principle.
He rightly founded the distinctions of operations on the
distinctions of their objects: hence his discovery of
nearly all that is known in mental philosophy. He

1 Eth. Nic. ix. 9, 9.

102 PSYCHOLOGICAL IDEALISM PART n.

wrongly neglected the consciousness of the operations

about those objects : hence his tendency to dogmatism.

Descartes supplied a defect in psychology when he

discovered the necessity of using as a principle the con-

jsciousness which says to each of us: 'I think; that

is, I feel, perceive, remember, imagine, judge, reason,

^ desire, will : I therefore am.'

It is a principle. Is it the only principle of psycho
logy ? How far will this conscious fact, that I am, carry
me ? I am conscious that I am a thinking subject.
But two further questions immediately present them-,
selves : what am I, and what do I apprehend ? What
is the thinking subject, and what the apprehended ob
ject? Now the mere consciousness that I think will
vnot of itself solve the nature of either subject or object.
The new principle of thinking was no more fitted than
the old principle of contradiction to be a universal
source of all philosophy : it must be accepted, without
pledging us to all the Cartesian deductions.

What" is the thinking subject? What am I? This
terrific question is answered by Descartes, as if it im
mediately followed from the principle, I think therefore
I am, but really by another argument. He cannot say,
I am conscious that I now think, as soul, without a
body. He therefore substitutes the hypothesis, I can
suppose that I had no body and was still thinking. He
then concludes that I, as thinking subject, am not body
but soul. Thus, by an easy transition, he leads his
readers from thinking subject to soul, and makes, not
the original principle, but an hypothesis and a problem
atic conclusion the real premises of his philosophy.

In order that we may feel the weakness of this non
sequitur, let us quote from the ' Discussion on Method,'
Part IV., the passage which immediately follows the

CHAP. v. DESCARTES 103

enunciation of the principle, I think therefore I
am : — -

' In the next place, I attentively examined what I
was, and as I observed that I could suppose that I had
no body, and that there was no world nor any place in
which I might be ; but that I could not therefore sup
pose that I was not ; and that, on the contrary, from
the very circumstance that I thought to doubt of the
truth of other things, it most clearly and certainly fol
lowed that I was ; while, on the other hand, if I had
only ceased to think, although all the other objects
which I had ever imagined had been in reality existent,
I would have had no reason to believe that I existed ; I
thence concluded that I was a substance whose whole
essence or nature consists only in thinking, and which,
that it may exist, has need of no place, nor is dependent
on any material thing ; so that " I," that is to say, the
mind by which I am what I am, is plainly distinct from j
the body, and is even more easily known than the
latter, and is such, that although the latter were not, it
would still continue to be all that it is.'

' I could suppose I had no body.' What is the
nature of this proposition ? It is an hypothesis of what j
might be but is not. I am not conscious that I have no /
body ; I am at best only conscious of the supposition,
which does not become any less a supposition through
my being conscious of making it. Nor is it deduced
from the consciousness that I think, but is a separate
hypothesis. Again, how do we get to the proposition, ;
I am a thinking substance wholly distinct from the
body ? It is a conclusion not from the original principle
alone, but also from the subsequent hypothesis, requiring
also a second hypothesis, that without a body I should
still be thinking.

104 PSYCHOLOGICAL IDEALISM PART IT.

We must, therefore, most carefully distinguish the
original principle, cogito, ergo sum, from the subsequent
conclusion, I am a soul. In the first place, I am con-
£cious of the former, not of the latter. I am conscious
'that I am a thinking subject : I am not conscious that
jthis thinking subject is not body but soul. Secondly,
in order to deduce the conclusion, the principle requires
the intervention of two hypotheses — that I could have
no body and that I should still be thinking ; and in both
/ cases I am conscious of making the suppositions, but
\ not conscious of the facts that I have no body and am
1 still thinking. But sectetur partem conclusio deteriorem.
An hypothetical premise produces an hypothetical con
clusion. The conclusion, then, that the thinking subject
is not body but soul, has not the certainty of the
rinciple, coqito, ergo sum, but is vitiated by the hypo-
heses coy w^fh ^ Thus does Descartes lead his

i

reader to confuse the thinker and the soul, and transfer
tHe conscious certainty of being res coqitans to the hypo
thesiiTof being res a corpore plane distincta.

"That T am a thinking subject is a fact of conscious
ness ; but what I am, as thinking subject, is a matter of
argument. There are three possible alternatives : Jiie
^body, the_soul, the man. Nor can we decide between
these three alternatives by consciousness alone. Con
sciousness, without hypothesis, never made a philoso
pher either a materialist or a spiritualist. We must
not make a fetish of consciousness, but interrogate it
carefully, remember its superficiality, add to it observa
tion, and combine both with reasoning.

In discussions of this kind a false issue is generally
raised at once by speaking of the consciousness of
thoughts. This is an abstraction, useful indeed f°r
some purposes, but still an abstraction, or rather a

CHAP. V.

DESCARTES 105

double abstraction. There is no such a thing as con
sciousness, and no such a thing as a thought ;_ I am
conscious, and I am conscious that 1 think. Conscious
ness and thought are not there, waiting for a subject ;
they already have a subject, or rather subjects — myself,
yourself, every other thinker. Descartes, in a great
measure at all events, avoided this fallacy of hypostasis-
ing abstractions. He was aware that there is no con
sciousness of thoughts, but I am conscious that I think.
He surreptitiously changed the thinker into soul, but
not into abstract thoughts. Those modern philosophers
who suppose consciousness of thoughts are not votaries
of consciousness, but victims of abstraction.

I am, then, not thoughts, but a thinker or thinking
subject. But what is this subject which thinks ? What
part of me is the factor, or what parts are the factors of
thinking ? In this mortal state, in which I cannot ap
prehend myself without my body, I am not conscious
that I think without my body. Nay, I am conscious that
I think with my body. Whatever operation I take, I in
variably find that I am conscious, not of the operation,
which I may afterwards abstract, but of myself per- j
forming it ; I am not conscious that I perform it by my •
soul without my body, because, though I am conscious,'
that I am a thinking subject, I am not conscious that)
this is a soul ; nor am I conscious that I do it by my
body without my soul, for reasons to follow presently.
I am conscious that I perform every operation by my
body, partly, somehow, and somewhere. I consciously
feel pleased and pained in various parts of my body. I *
cannot disengage my consciousness of toothache from /
my mouth, or of headache from my head. LjLHL_coj}-
scious of using my bodily senses in touch, taste, vision,
hearing, and smell. I do not Consciously first feel the

106 PSYCHOLOGICAL IDEALISM PART IT.

sensation and then refer it to the bodily member ; I am
not conscious of these two steps. Seasoning is the
highest kind of thinking ; I am conscious of doing it
, in my head, and by no force of abstraction can I get it
t«»»wJout °f mJ bead. Similarly I am conscious that I will
^ ^ in my head, and I am conscious that my head may ache
with reasoning, and deliberating, and resolving. My
(body is not a mere companion but a conscjou.fi pnrf.npr
°f all jny thoughts.

Bui consciousness is a superficial power. "In speak
ing of the data of sense I remarked that by an illusion,
arising from the confusion of sense and inference, we
^cannot help seeming to be conscious that sense per-
jpeives an external object, though we can make ourselves
independent of the illusion by science, which dis-
jtinguishes the external from the sensible. There is a
similar illusion about our consciousness of the thinking
subject, and fortunately we can explain it and conquer
it by science. The illusion is that we perform some of
our operations on the surfaces by the superficial mem
bers of our bodies. The causes of the illusion are that
we often observe the outer surfaces of our bodies when
we are performing an internal operation, and we are at
the same time unconscious of the inner structures and
] motions of our nerves and brain. The way to make
| ourselves conquer the illusion is by the study of science,
j which shows that what performs the operation is not
I the outer surface but the inner nervous system. For
example, we are conscious that we see something red
somehow by our bodily organs of sight. Now, though
we are sensible of the optic nerve so far as it is sensibly
affected with red, we are neither sensible nor conscious
of it as nervously constituted. But from very early
infancy we observe, i.e. directly infer from sensation,

CHAP. V.

DESCARTES 107

the surfaces of our bodies. By putting our hands on
our eyes we find that they no longer see red, and we
infer that it was our eyes that saw red. It is so with
all our external senses, as they are called from this
illusion of observation. Not consciousness, but obser
vation from very early infancy, made us believe that it
was the periphery that is sensitive. But the inference
became automatic before we were attentively conscious,
and we cannot help seeming to be conscious that our
eyes see. Eeally, however, as science discovers at last,
the eyes are but avenues to vision, and wliaLsees is not
Qur_eyes but the optic nerve in connection with the
brain. A more complicated instance is when a person
who has lost a limb believes that the pain, which he
really feels in the nerves, is still in the limb. His con
sciousness told him but vaguely where he feels the pain,
his observations connected it with the surface of the
limb; hence the illusion. Science alone can conquer
such illusions of observation.

The rough-and-ready way of dealing with this evi
dence is to draw the further inference that we do not
localise any operation except by observation and ana
tomy, and that consciousness has nothing to do with the
body. But this inference goes far beyond the facts.
Observation is limited to the surface of the body, but
the operations, of which we are conscious, are not.
Now, even when they are purely internal, we are still
conscious that they are somehow performed by the
body, without observation and before science. For
example, we are conscious of the pangs of hunger in
the region of the stomach, to descend to the depths of
consciousness : to rise to its summit, we are conscious
of the process of reasoning in the region of the head.
But in neither case does observation of the surface of

108 PSYCHOLOGICAL IDEALISM TART n.

the body reveal the whereabouts of the operation : yet
we are conscious of the body performing it, without
waiting for science.

But there is another defect, for which the conscious
ness of the body as a factor in thinking is responsible.
lit tells me very indefinitely what part is engaged in a
[particular operation. The cause of this indefiniteness
;is the unconsciousness of nervous structure and motion.
The correction of it is the science of nervous structure
and motion. Thus, confining ourselves entirely to in
ternal operations, the locality of which is not accessible
j to external observation, I am conscious of the pain of
I hunger somewhere in the region of the stomach ; sciejq.ce
\reduces this indefinite verdict to definiteness by proving
jthe connection of the nerves of that region with the
/brain. Consciousness again says indefinitely, 'I think
in my head ' ; science tells me, ' Yes, in your brain.'

Here science only corrects consciousness : it does not
contradict it. Consciousness apprehends the indefinite
region at work, science discovers the definite nervous
structure in the direction of that region. Secondly,
unless consciousness apprehended the region, science
could not assign the nervous structure ; if we were not
I already conscious of reasoning in the head, anatomy
1 would not convince us that we reason in the brain.
Thirdly, sometimes consciousness apprehends the region
without science having yet discovered the nervous
structure ; for example, we are conscious, in what is
inadequately called muscular sense, not indeed of mus
cular motion but of the action of our limbs, though
but vaguely and indefinitely ; but on this occasion
science is still more vague and indefinite, having dis
covered the nervous mechanism of muscular motion,
but not of muscular sense. Finally, however wrong

CHAP. V.

DESCATITES 109

consciousness may be in the definite locality of a parti
cular operation, science never disproves that we are
conscious of its being performed somewhere in the
body. I am conscious that I perform all my operations
somehow or another, partly by the body, with more or
less definiteness ; science discovers the definite locality,
still within the body.

There are two points, which sometimes appear in
biological treatises, but are not proved. In the first
place, as we have already seen, there is no biological >
proof that cerebral motion is transmuted into a psy
chical sensation. Secondly, biologists often distinguish
a sensation from its localisation ; at the same time they
sometimes confuse its localisation in the body with the
inference of its external cause. There is a great differ-j
ence between a sensation of an internal sensible object
and the inference of its external cause, as we hav3
already seen in this essay. But there is no difference
between the sensation and its internal localisation in
the sentient subject ; there is no proof of these two
steps. I am conscious of the sensation in a locality of
my body. Neither consciousness nor science proves
that I first have a sensation, then localise it in my
body, and, thirdly, infer its external cause. They prove
together that I first have a sensation located in some
part of my body, and then infer the external causa;
which produces it.

There is another point, which is proved in biology,
but does not disprove the consciousness of the body as
a factor in thinking. I refer to subjective sensations. |/
We have sensations similar to our ordinary sensations,
but not produced by the ordinary external cause. Thus, i
a prize-fighter may be made by a blow to see stars ; a
drunkard under the influence of delirium tremens may

110 PSYCHOLOGICAL IDEALISM PART ir.

:

have a vision of tlie devil. Such sensations are excel-
ent instances to sliow that the sensible object is different
from the external original, and is not always caused by
it ; that there are internal causes of sensations in the
/ nerves ; and that the superficial structure of the eye is
\ a cause, not a subject, of vision. But they do not show
jthat the soul is the sole subject of vision. A prize
fighter seeing stars, a drunkard's vision of the devil,
are odd proofs of psychical sensations. The term ' sub
jective' sensations is misleading, because, in the recent
sense of the word, it suggests 'psychical,' without
proving it.

There is no evidence that sensation, or any other
| operation, is purely psychical. There is evidence that
Jthe body is a factor in all thinking. It is the evidence
of consciousness, interpreted by science. I am not
conscious first of a sensation, and then of its locali-
\sation. I am conscious that I feel, perceive, reason,
/ will, partly by my bod}? . External observation connects
some of these operations with the surface of the body.
Science shows that I do all of them by my nervous
system. Science dispels the illusion of observation, and
corrects the indefiniteness of consciousness. Science
further traces the continuity of the nervous system, and
leaves no gap for purely psychical operations, . Xow,
\ ordinary and scientific observation being limited to the
jbody, if I were only conscious of mere thinking, I should
know my body only as an unthinking cause. But when
I cannot be conscious that I perform any operation
Without being conscious that I perform it somehow in
/my body, that I feel headache, that I use my bodily
senses to see, touch, hear, and so on, and that I reason
\in my head, scientific observation becomes an inter
preter of my consciousness that I use my body to think,

CHAP. v. DESCARTES HI

and shows that the part which I use is the brain in con
nection with the nervous system. The body is a patent
factor of the thinking subject. The neglect of it is the
fallacy of spiritualism.

It does not follow that the body is the sole factor of
thinking. Man does not know the whole of himself,
either by consciousness or by scientific observation ; the
former is superficial, the latter limited. I am conscious/
that I perform my operations partly by my body : science
observes the nervous system, and in combination with!
consciousness, infers that the nervous system is that by
which the body in part performs these operations. But
I am not conscious that my body, nor does science
observe that the nervous system, is the whole thinking
subject. There is no operation which can be traced
throughout its whole course. I am conscious that I
use my bodily senses in sensation and my head in
reasoning. Science observes the nervous system and
brain. But it has not solved the problem of nervous
and cerebral motion. If it solved that problem, it would
still remain to prove that nervous motion is completely
identical writh the operation of which I am conscious.
It is partly so, because I am conscious of partly per
forming the operation by the body, in which science ob
serves the nervous system and the motion it performs
during the operation. . But it is another thing to prove
that the conscious operation and the nervous motion
are completely identical, because I am conscious of the
operation without observing it, and science observes
the motion without being conscious of it. This differ
ence of evidence does not, indeed, prove a complete
difference, because nervous motion and conscious opera
tion may be the same fact approached from different
sides, but the very difference of evidence makes it dim-

112 PSYCHOLOGICAL IDEALISM PART n.

cult to prove a complete identity of fact. Another
evidence might be evoked — the method of explanation.
If all the facts of conscious operations were known, and
nervous motions were known, it might be urged that
the former are explicable by the latter, as the facts of
light are explicable by undulating motion. But there
is a great difference in the two cases. In the case of
li<rht, we can say that its facts are such as the known
effects of undulation by the laws of motion. But the
Operations, of which we are conscious, do not seem to
Consciousness to be the kind of effects produced by any
known motion according to any known laws of motion.
{There is a latent factor in all thinking, the soul. The
/neglect of it is the defect of materialism.

Two opposite errors must be avoided, spiritualism
and materialism. The former neglects the patent, the
latter the latent, factor of the thinking subject. The
former despises the consciousness of the body as a factor,
and the science of the nervous system as the part of
that factor, engaged in every conscious operation : the
latter transgresses the limits of science. Hence the
f former falsely supposes the subject to be all soul, the
latter all body. Both neglect the man ; yet as men we
think. There is room for an intermediate theory of the
thinking subject ; for a theory which is founded on the
combined evidence of consciousness, of observation,
ordinary and scientific, and of reasoning about oneself;
for a theory which avoids the opposite difficulties of
disturbing the physical continuity of the nervous system,
and of inventing a mere parallelism of neurosis and
j psychosis. I suppose that brain and soul are co-opera
tive factors in all conscious operations, in passive opera
tions together affected by external causes, in active
Operations together producing external effects. The

CHAP. v. DESCARTES 113

thinking subject is man, thinking partly by his body,
that is, his nervous cerebral system, and partly by a
latent factor, his soul, co-operating, as by the composi
tion of forces, in every operation.

But what are the objects which I apprehend in think
ing ? This is the second question, suggested by the
consciousness that I think, but not answerable without
further argument. Descartes assumed that all the
immediate objects are psyehiral ideas, while physical
things are only mediate objects known through the
medium of ideas. So far as this theory recognises the
distinction between the internal objects of sense and
external objects of inference, it is correct, and in accord
ance with the scientific evidences already given in the
First Part of this essay. But it contains a further sup
position, namely, that objents^of sense and all other
immediate objects are not only internal but psychical,
arejdeas. Descartes never proved this ideal theory.

In the Third Meditation we find the following
passage : — •

' Nevertheless I before received and admitted many
things as wholly certain and manifest, which yet I after
wards found to be doubtful. What, then, were those ?
They were the earth, the sky, the stars, and all the other
objects which I was in the habit of perceiving by the
senses. But what was it that I clearly (and distinctly)
perceived in them ? Nothing more than that the ideas
and the thoughts of those objects were presented to my
mind. And even now I do not deny that these ideas
are found in my mind. But there was yet another
thing which I affirmed, and which, from having been
accustomed to believe it, I thought I clearly perceivedl
although, in truth, I did not perceive it at all ; I mean
the existence of objects external to me, from w

I

114 PSYCHOLOGICAL IDEALISM TART n.

those ideas proceeded, and to which they had a perfect
/resemblance ; and it was here I was mistaken, or if I
judged correctly, this assuredly was not to be traced to
any knowledge I possessed.' l

Now Descartes does not state precisely how he
arrived at this conclusion that what he perceived were
ideas. No doubt he was unconsciously influenced by
the previous course of philosophy, detailed in my last
chapter, and thought himself entitled to accept the con
clusion much more rapidly than he ought. But he
probably also thought that it followed in some way
from the principle, cogito — from the consciousness, I
think. Now it is true that I think includes the con
sciousness, I sensitively perceive. But I am not con
scious that my senses apprehend ideas. As I walk in
the fields, I am conscious of perceiving something green,
which, so far from being an idea or any psychical fact,
appears to be not only physical but also external.
yScience disabuses me of the externality, but not of the
(materiality of the sensible object. What further evi
dence, then, had Descartes to disprove its physical and
prove its psychical character ?

Descartes derived his ideal theory of the sensible

object apparently from his principle, cogito, ergo sum,

but really from his secondary hypothesis, ' I am a soul.'

'Having convinced himself that the whole subject is soul,

jljlie defined soul as a purely thinking substance, and

jody as a purely extended substance. From these

definitions he deduced the heterogeneity of mind and

natter, of soul and body. Hence he thought it would

follow that the soul by its very essence thinking cannot

•apprehend body by its very essence extended, but is

limited to its ideas. The real Cartesian evidence is this :

1 ' Ex vi meae perceptions,' in the Latin edition.

CHAP. v. DESCARTES 115

the subject is soul, the soul is such as to apprehend only/
ideas,; therefore all immediate objects are ideas. Bun /
neither premise is proved. » **

It is not true that the whole subject is the soul.
Descartes, as we have seen, exaggerated the soul from
a part to the whole thinking subject. The man is the
whole subject : ffine body is part of that by which he
Jhinks; and, being a factor in thinking as well as extended, \
it is not a purely extended substance. The assumption/
at the bottom of the Cartesian definition of body is thatf\
thinking and extension are different. So they are in)
the abstract ; but the same thing may possess both attri- (
butes in the concrete. Number is not extension, but t
the same thing is numerable and extended ; extension is
not thinking, but the same body may be both extended I
and think. When we appeal from abstractions to con- i
sciousness, we find it does think. The body, therefore,
is not purely extended substance, but also thinking.
Again, the smil is a factor in thinking and is in other
respects latent : it does not follow that it is nothing else.
Rather such a supposition is impossible ; for, as Locke
wittily remarked, men think not always, and if the soul
were purely thinking substance, either it must always
think in order to be, or it must have an intermittent
existence, both of which alternatives are impossible and
absurd. Descartes resolved body and soul into the two
opposite abstractions of extension and thinking. But
he did not thereby prove that body is purely extended,
nor that soul is purely thinking, nor their heterogeneity,
nor that the body is no factor in thinking, nor that the
whole thinking subject is the soul.

But if we concede that the soul is the whole thinking
subject, and that thinking is accordingly a purely
psychical operation, whether it be feeling, perceiving,

i 2

116 PSYCHOLOGICAL IDEALISM PART ir.

reasoning, willing, or what not, what do we know about
its nature ? On this point we have a dictum of Sir W.
Hamilton, so admirable that we cannot pass it by. ' We
know,' he says, ' and can know, nothing a priori of what
is possible or impossible to mind, and it is only by obser
vation and generalisation a posteriori that we can ever
hope to attain any insight into the question.' x The-
iiost we know of the soul is that it thinks, whatever
?lse it is ; we cannot enter further into its secret nature
o determine what it thinks. We must, therefore, judge
of it by its fruits. Now, when I appeal to conscious
ness, I am not conscious of perceiving only ideas, but
(physical things apparently external ; and when I correct
the illusion of externality by science, I find that sense
(perceives internal things, but not ideas ; and further,
that it must perceive physical things within in order to
linfer physical things without. I conclude, therefore,
that, as we apprehend the physical as a fact, the soul
must have a power of apprehending it ; for we only
know what the soul must by what it does apprehend,
lit is not true, then, that the soul is such as only to per-
Iceive ideas.

Even, therefore, if the first premise of Descartes be
true, his second is false, so that his conclusion does not
follow. If the soul is the whole thinking subject, it is
not true that its nature is such as only to immediately
i perceive ideas ; for all we know of its thinking is that,
as a matter of fact, it immediately perceives physical,
•though internal, effects on the nervous system. Thinking
we know is not extension, but know nothing about
thinking to prevent it perceiving the extended, nor any
thing about the psychical to prevent it perceiving the
physical. Let vision be purely psychical, white seen can

1 Hamilton's Metaphysics, Lect. xxv. p. 122.

CHAP. V.

DESCAUTES 117

still be physical. Granted, then, that the subject is the /
soul, it is a non sequitur that it perceives only psychical I
ideas.

A fortiori, if the subject is the man, he can perceive
the physical in himself. A certain conditional plausi
bility is given to the idealistic theory of the sensible
object by the spiritualistic theory of the sentient subject.
Although we cannot say that if the subject is purely
psychical the object is psychical, we can say that the
object is not psychical unless the subject is purely psychi
cal. But if the body is a factor of the thinking subject
there no longer remains any plausibility in the idea
theory of the sensible object. The physical can appre
hend the physical, the extended the extended, withi
the nervous system. The thinking subject, body and
soul, does apprehend the physical: it therefore can.j
What we apprehend as a fact is better known than what
we are to apprehend it. Knowable objects must be
explained, not denied, by knowing subjects.

Descartes was a_£lear and distinct_ssxlter ; he was
not so clear and distinct a thinker. His works are
full of confusion. He was the first to confuse the
object with the operation of sense. Hence, when he
speaks of an idea of white, we never feel quite sure
whether he means the white perceived or the sensa
tion of white. Now, if the subject is soul, the operation
is purely psychical ; and, if the object be undistinguish-
able from the operation, it also becomes psychical : if
the white perceived is the same as the vision of white,
and this be psychical, that becomes psychical. But we
found in the First Part that the white seen is not the vision
of white, the sensible object is not the operation of sensa
tion. Hence, it does not follow, even if the operation
of vision be psychical, that the white seen is psychical.

118 PSYCHOLOGICAL IDEALISM PAET n.

A further confusion was necessary before Descartes
could call a sensible object an idea. Confusing it with
a sensation would only have enabled him to call it a sen-
Isation. But why an idea ? Because he merged sensation
(in conception. There are two kinds of simple appre
hension : sensation, the apprehension we have of an
object when the original is present ; and conception,
when the original is absent or non-existent. Aristotle
had clearly distinguished them as cuo-#r?cris and <£ai/-
racrta, and their objects as aio'O'rjpa and <f)dvTao-p.a. But
the poverty and abstractness of modern languages and
the growth of conceptualism obliterated these dis
tinctions, and enabled Descartes first to confuse the
sensible with its sensation, and then the sensation with
the conception or idea. Nothing can be more mislead
ing than the word ' idea,' because it may signify either
the conception or the concept, to use later phraseology.
But Descartes arrived at his theory that the sensible
object is an idea by a fusion of sensible object, sensa
tion, conception, and concept.

A final confusion followed the rest. Wherever there
is no distinction between object and operation, as in
i feeling, there is none between the operation and its con-
( sciousness. Accordingly, Descartes, having first confused
the sensible object with the sensation and then the sen-
'. sation with the idea, having no object left, confused
\ the operation of sensation with its consciousness.1 The
? white, its vision, its idea, its conception, its conscious
ness became all merged : there was no distinction left
between sensible object, sensation, idea, conception, and
consciousness. Thus the sensible object historically be
came a state of consciousness by a series of confusions,

1 Cf. Princ. i. 9.

CHAP. V.

DESCARTES 1 1 9

from which mental philosophy has never quite recovered
itself.

So much for the evidence of the Cartesian theory I
that all immediate objects are ideas. He derived it
not from the principle, Cogito, ergo sum, but from at least )
four hypotheses : —

(1) The subject is the soul.

(2) The soul is such as to perceive ideas.

(3) The sensible object is undistinguishable from
the sensation.

(4) The sensation is undistinguishable from the idea. ;

Not one of these hypotheses is true ; at any rate, all
are uncertain. But if any one of the hypotheses is
false, it vitiates the reasoning ; and if any one is un
certain, it renders the reasoning uncertain. The Car
tesian method is apparently synthetic demonstration,
but really synthetic hypothesis. There is a lesson of
psychological method to be derived from it. We can
not logically start with the subject, and from its sup
posed nature deduce the immediate and mediate
objects of knowledge ; but we must first find what ob-,
jects the subject knows, as a fact, in the sciences, then
the immediate objects of sense, and finally conclude that
the nature of the subject is such that it can know what
it does know. The method must be not synthetical but
analytical, because it must proceed from the more cer
tain to the less certain, not from hypotheses to facts, but
from facts to hypotheses.

We have not yet, however, exhausted the Cartesian
theory of ideas as the immediate objects of knowledge.
Although he thought that all sensible objects are ideas,
Descartes was well aware that there are ideas which
are net sensible. There are, according to him, three 1

120 PSYCHOLOGICAL IDEALISM PART IT.

\Jsorts of ideas — innate, adventitious, and fictitious.
This celebrated theory of the origin of ideas has at all
events two very great merits : first, it called attention to
the important problem of the origin of ideas ; secondly,
hinder the head of innate ideas, it recognises ideas
kvhich are not sensible. He remarks that ' the philoso-
Shers of the schools accept as a maxim that there
s nothing in the understanding which was not pre-
dously in the senses, in which, however, it is certain
hat the ideas of God and of the soul have never
jeen.' l

It is well known that Descartes repudiated the
theory that some ideas are innate in the sense of being
always present. In his replies to the objections raised
against his Meditations, the ' Responsiones Tertias ' con
tain the following passage : — ' Denique quum dicimus
ideam aliquam nobis innatam, non intelligimus illam
nobis semper observari, sic enim nulla prorsus esset
innata ; sed tantum nos habere in nobis ipsis facultatem
'illam eliciendi.' 2 This doctrine of ideas, innate in the
isense of elicited from one's own faculty of thinking, is
developed at length in the c Notes on the Programme
of Regius/ 3 and was the foundation of the celebrated
maxim of Leibnitz : ' Nihil est in intellect!! quod non
\prius in sensu nisi ipse intellectus.' By innate ideas
[Descartes meant ideas, not acquired, like adventitious
jideas, by sense from external objects, nor yet inborn,
but capable of being elicited from the faculty of think
ing, which is supposed to be endued with a capacity of
conceiving them.

This Cartesian mystery of eliciting ideas from the
I faculty of thinking is nothing really but the ordinary

1 Discourse on Method, Part IV. 2 Page 102 (ed. 1685).

3 Id. pp. 184-6.

CHAP v. DESCARTES 121

operations of consciousness and reasoning, hidden under
a fine-sounding phrase. It is quite true that we find
certain ideas in ourselves which have never been in
sense. We arrive at some of them, such as that of
thought and that of truth, not from sensation but from
consciousness. But consciousness is a kind of sense,
and ideas derived from consciousness of oneself are not
elicited from oneself, but apprehended as belonging to
oneself. They are, as Locke afterwards showed, ideas
of reflection, not innate, but acquired by apprehending
ourselves performing conscious operations. There are,
indeed, other ideas of the insensible, which are not
acquired either by sensation or by consciousness. We
neither see God nor are conscious of God in us. Such
ideas also are, according to Descartes, elicited not from
sense, but from our faculty of thinking. In his ' Notes
on the Programme of Eegius,' after disposing of tradi
tion and observation, he thinks himself entitled to con
clude that the idea of God is innate in the sense of
elicited from the faculty of thinking. But there is
another alternative — reasoning from sensation and
consciousness. Logical reasoning is an. indirect origin

When Descartes said that ideas were elicited from
the faculty of thinking, perhaps he had some obscure
unanalysed hint of ideas generated by reasoning. But,
then, such ideas are not innate, but acquired, and are
the most deviously acquired of all ideas. In the first
place, we reason from sensation, and infer that there are
objects, like the sensible, but insensible. We then con
ceive an idea of the insensible. This is plainly the
origin of the idea of a corpuscle, which is an idea
neither sensible, nor conscious, nor innate, but acquired
by reasoning from sense. Secondly, we reason from

122 PSYCHOLOGICAL IDEALISM

PAHT IT.

I sensation and consciousness, and infer other thinkino*

fT>

beings. For example, we are conscious of being able by
reason and will to produce good effects. By reasoning
from sense we infer the goodness of nature. By com
bining these evidences we infer a being who reasons
and wills to produce the goodness of nature. We then
conceive an idea of this being similar to ourselves, but
infinitely more perfect in reason and will. This is at least
(one origin of the idea of God, which is neither an idea
iof sensation, nor of consciousness, nor innate, but ac-
I quired by reasoning from sensation and consciousness.

Before a theory of the origin of ideas can be ad
mitted two conditions of hypothesis must be satisfied.
Other hypotheses must have been eliminated ; and
any hypothesis, which without going beyond known
operations explains the facts, must be preferred to an
hypothesis which supposes an unverified power. But
1 the theory of innate ideas is imperfect in elimination
jand in verification. It shows that not all ideas arise
'immediately from sensation, but it fails to show that
the rest are not due to consciousness and reasoning,
and accordingly the unverified power of eliciting ideas
must yield to the verified powers of sensation, conscious
ness, and reasoning, which are together a sufficient ex
planation of the origin of ideas.

We have now all the immediate data of knowledge,-
and their origin, according to Descartes. They are the
soul known as subject, and its ideas, innate, adventitious
and fictitious, known as objects. What knowledge, then,
even if we have innate ideas, should we get ? What
apprehension of reality, immediate and mediate ? We
should know that we really are souls. We should also
know that we really apprehend ideas. But what else
should we know mediately beyond the soul and its

CHAP. v. DESCARTES 123

ideas ? Descartes replies : from our ideas \ve should [
know that there are real objects beyond them.

Aristotle had distinguished simple and complex ap
prehension, conception and judgment, and had pointed
out that truth and falsity arise with judgment. As long
as we merely apprehend an idea, e.g. of a man or a
centaur, we express no belief about the existence of an
object. But when we judge about objects we appre
hend a relation of combination or separation, which, in
its simplest but not its only form, is a relation of exist
ence or non-existence ; e.g. a man exists, a centaur does
not exist. Hence our judgment may be either true or
false : true, if it agrees with a real relation of combina
tion or separation ; false, if it does not. The question
is, how are we to know that our judgments about objects
are true? How are we to know that there is a real
relation ?

Descartes, though he confused two kinds of simple
apprehension, sensation and conception, was aware of
the distinction between conception and judgment. He
also saw that the possibility of falsity begins with judg
ments.1 But he thought that true judgments can be
derived from ideas themselves, by their own inherent ',
characteristics. He proposed a new criterion of truth,
and a new method of forming true judgments from
ideas themselves.

In the first place, he thought that the mere idea of
God proves God's existence. This theory he applied in
two arguments, one of which proceeded from the idea
of necessary existence. The mind, he says, from per
ceiving necessary and eternal existence to be comprised
in the idea which it has of an all-perfect Being, ought

1 Meditation III.

124 PSYCHOLOGICAL IDEALISM TAKT IT.

I' manifestly to conclude that this all-perfect Being exists.1
This ontological argument, as it is called, which has
become famous from the criticisms upon it, especially
that of Kant,2 and from its revival by Hegel,3 is a most
transparent fallacy. Descartes surreptitiously omitted
the word ' idea.' An idea only comprises ideas, and
ur idea of God comprises, not necessary and eternal
xistence, but only the idea of necessary and eternal
xistence, which only proves that the idea of this all-
perfect Being exists.

The other argument proceeded from the objective

reality of an idea ; that is, according to the proper

meaning of ' objective,' the reality of an idea quatenus

objicitur intellectui. This argument follows the former

argument in the Principles,4 but is stated at greater

length in the Third Meditation. Shortly, it comes to

Jiis : more reality cannot be produced by less ; the idea

>f God has more objective reality than the actual, or

brmal, reality, of a finite substance ; therefore it cannot

je produced by a finite substance, but must be received

rom God Himself. The major is true ; the minor is

alse, because the objective reality of an idea is always less

real than the actual reality of the thinker, and therefore

can be produced by him. God has more reality than

man ; the idea of God has more reality than the idea

bf man ; but man has more reality than his own

jdea of God. We can therefore retort on Descartes

the following syllogism : the less real can be produced

by the more real ; the idea of God is less real than the

thinker ; therefore the idea of God can be produced by

the man who thinks it. As for the way in which a

man produces, not God, but his idea of God, we have

1 Princ. i. 14. 2 Critique, pp. 364-70 (Bohn).

3 Logic of Hegel (Wallace), pp. 91 2. 4 i. 17-19.

CHAP. V.

DESCARTES 125

already described it. He produces it, not by the bare!
idea of his own finite substance, but by reasoning from|
nature to nature's God, and then forming an idea of
a Being, reasoning and willing like himself, but infi
nitely more perfect. Man has no power to produce'
God ; but, by reasoning to God's perfect existence, he '
can produce his very imperfect idea of God. The origin;
of natural theology is reasoning : the origin of the idea[
of God is rational idealisation.

Secondly, conceding the reality of God, though not
proved by these two arguments from ideas, but rather
by a third argument which he adds from our not having
made ourselves, let us proceed to the further use of
ideas as a criterion of truth by Descartes. He accepted
the Christian doctrine that God is not the cause of our \
errors.1 He pointed out that some of our ideas are !
clear and distinct,2 others obscure and confused. He
concluded that those ideas, which are clear and dis
tinct, must be true, otherwise God would be a deceiver.3
In short, he made the clearness and distinctness of ideas,
backed by the veracity of God, a criterion of truth, by
which we may argue from ideas to objects beyond
them.

By this internal criterion of ideas, he supposed
that from our psychical ideas we may infer a physical 1
world 4 :—

' Although we are all sufficiently persuaded of the j
existence of material things, yet since this was before '
called in question by us, and since we reckoned the
persuasion of their existence as among the prejudices
of our childhood, it is now necessary for us to investigate
the grounds on which this truth may be known with
certainty. In the first place, then, it cannot be doubted

1 Princ. i, 29. 2 Id. i. 45. 3 Id. i. 30. 4 Id. ii. 1.

126 PSYCHOLOGICAL IDEALISM TART n.

(that every perception we have conies to us from some
object different from our mind ; for it is not in our
power to cause ourselves to experience one perception
rather than another, the perception being entirely
dependent on the object which affects our senses. It
may, indeed, be matter of enquiry whether that object
be God, or something different from God ; but because
we perceive, or rather, stimulated by sense, clearly and
distinctly apprehend, certain matter extended in length,
breadth, and thickness, the various parts of which have
different figures and. motions, and give rise to the sensa
tions we have of colours, smells, pain, &c., God would,
without question, deserve to be regarded as a deceiver,
if He directly and of Himself presented to our mind the
idea of this extended matter, or merely caused it to be
presented to us by some object which possessed neither
extension, figure, nor motion. For we clearly conceive
(this matter as entirely distinct from God, and from our
selves or our mind ; and appear even clearly to discern
that the idea of it is formed in us on occasion of objects
existing out of our minds, to which it is in every respect
similar. But since God cannot deceive us, for this is
repugnant to His nature, as has been already remarked,
.we must unhesitatingly conclude that there exists a
certain object extended in length, breadth, and tliick-
jness, and possessing all those properties which we clearly
apprehend to belong to what is extended. And this
extended substance is what we call body or matter.'

By the same internal criterion of ideas he thought
that he could infer that distinction in nature which
Locke called ' the distinction of primary and secondary
qualities ' 1 :—

J ' As belonging to the class of things clearly appre-
1 Meditation III.

^

CHAP. V.

DESCARTES 127

bended, I recognise the following, namely, magnitude,
or extension in length, breadth, and deptli ; figure,
which results from the termination of extension ; situa
tion, which bodies of diverse figures preserve with
reference to each other ; and motion or change of situa
tion ; to which may be added substance, duration, and
number. But with regard to lights, colours, sounds,
odours, tastes, heat, cold, and the other tactile qualities,
they are thought with so much obscurity and confusion,
that I cannot determine even whether they are true or
false ; in other words, whether or not the ideas I have
of these qualities are in truth the ideas of real objects.'

The clearness and distinctness of our ideas are facts
of our consciousness ; but we are also conscious that
they do not always correspond to our knowledge.
Many ideas are clear and distinct where the objects
are known to be unreal. Nothing can be clearer or
more distinct than the ideas I have of Achilles and
Agamemnon, of Ulysses and Nestor. Am I, then, to
infer that these Homeric heroes lived in the flesh?
Many ideas are obscure and confused where the objects
are known to be real. The idea of the earth as a vast
globe revolving round the vaster sun is neither clear
nor distinct. Yet the earth revolves round the sun.
Nor do the degrees of truth correspond with the
degrees of clearness and distinctness. My ideas of
Hamlet and Macbeth are clearer and more distinct than
my ideas of the Prime Minister and the Leader of the
Opposition Am I to infer that the persons are also
more real ? I have proofs to prevent such an inference,
you will say. But that is to introduce another criterion,
over and above the clearness and distinctness of ideas.

Ideas are clear and distinct by proximity to sense, \
as well as by accuracy of science : hence, as science

128 PSYCHOLOGICAL IDEALISM

PART II.

Irecedes from sense, it often proves insensible objects
\to exist, of wliicli it forms but obscure and confused
ideas. Science assures me that matter consists of incon
ceivably small particles moving with inconceivable
rapidity on the one hand, and on the other hand that it
Bxtends over an inconceivable immensity of space. These
things are known to exist. Yet the ideas of them are
so far from being clear and distinct that I cannot be
said to have any direct ideas of them at all. If I try to
form an idea of a particle in a drop of water, I divide it
and divide it, but the division baffles my conception.
If I try to conceive the immensity of space, I enlarge
my idea of a limited space, as the frog tried to swell
himself out to the size of the ox, and with much the
same success. Are we to say that these scientific facts
are not real, because we have but obscure and indistinct
'ideas — or rather have no direct ideas of them at
jail?

Again, science is based on the distinction of the

apparent and the real, in time and space, in bulk and

\\ motion. The real in all cases is regarded as more truly

freal than the apparent ; but the apparent is more

^clearly and distinctly conceived. My idea of the ap-

, parent sun moving apparently over the apparent earth

!is clear and distinct : my idea of the real earth movino-

{really round the real sun is obscure and indistinct.

i Am I then to infer that the sun moves over the earth,

|and not the earth round the sun ? Again, the ideas of

secondary qualities in sense are as clear as, if not clearer

than, the ideas of primary qualities. For example, the

idea of sensible heat is as clear and distinct as that of

(sensible extension. The difference is not in the ideas,

but in the inferences of the qualities ; and there must

be some other ground than ideas, which are equally

CHAP. V.

DESCARTES 129

clear and distinct, to justify the differences between our!
inferences of insensible heat and insensible extension.

To say that God does not deceive us is to raise a!
false issue. He does not deceive those who use sensej
and reasoning by the laws of logic. But we should!
deceive ourselves if we were to follow Descartes and
substitute mere ideas for the logical organs of truth.
We are often deceived when we have clear and distinct /
ideas. We are deceived about the bodily locality of
our sensations and the externality of the object of
sense, about the reality of secondary qualities, about
the beautiful and about the good. A man often has
a clear and distinct idea of a duty which is no duty.
I might justify any wickedness, if I allowed myself to
argue, I have a clear and distinct idea of the rightness
of this action, and God will not deceive me, therefore
it is right. The Inquisition, no doubt, had a clear and
distinct idea of the justice of punishing heretics, and a
belief that God never deceived them : on the logic of
Descartes, they would have been justified in punishing
Galileo as a heretic, for saying that the earth goes
round the sun. This lazy logic of ideas would justify I
any arbitrary conclusion, and defy all rational criticism.
For, who is to know whether or not one has clear and
distinct ideas? But Descartes lived in a reaction
against logic.

An idea is an apprehension in the absence of an ex
ternal object. It contains no opinion whether the object
is only absent or also non-existent. Its clearness and
distinctness depend on other causes besides the belief of
existence, and especially on the proximity of the idea to
sense. Hence conception does not in itself guarantee
the existence of an external object. Moreover, we are
conscious of its limits : everybody knows that however

K

130 PSYCHOLOGICAL IDEALISM TART ir.

(well lie conceives, lie is not justified in judging without
further evidence. If we were not conscious of the
frequent disagreement of ideas with facts, God might
have been a deceiver ; but, as it is, He does not deceive
us into thinking that our clear and distinct ideas con
form to facts, but, on the contrary, makes us conscious
that they often do not, and gives us an opportunity of
going beyond them by reasoning. We are conscious
that our ideas do not justify judgments of existence
without rational proof.

But the words, conception and idea, are so vague
that they often get confused with belief, just as the
inconceivable is often confounded with the incredible.
Descartes shows traces of this confusion, in the very
act of drawing a valuable distinction between imagina
tion and pure intellection, in the Sixth Meditation :—

' To render this plain, I first examine the difference
which there is between imagination and pure intel
lection. For example, when I imagine a triangle, not
only do I understand that it is a figure comprehended
by three lines, but at the same time I intuite those three
lines as formed by the glance of my mind ; and this is
what I call imagining. But if I wish to think of a
chiliagon, I equally well understand that it is a figure
consisting of a thousand sides, as I understand that a
triangle is a figure consisting of three, but I do not in
^he same way imagine those thousand sides, or intuite
tlhem as present.'

Everybody would at once suppose that by intellection
he simply meant the belief that there is such a figure as
a chiliagon ; but, if we look a few lines further down,
we find the very reverse. He says that ' the mind,
which understands, turns itself in a way towards itself
/and considers one of the ideas which are in it ; ' and he

CHAP. V.

DESCAKTES lol

repeats the same view in the ' Responsiones Quintae.'
According to Descartes, then, the pure intellection that
a chiliagon is a figure consisting of a thousand sides is
a consideration of ideas. When he thus merged the
conception of ideas and the understanding of objects,
no wonder he thought that clear and distinct ideas
enable us to know objects beyond ideas. But, really,
we judge very clearly that a chiliagon is a figure with
a thousand sides, but we conceive a very obscure idea
of so many sides, if we can be said to have any direct j
idea of them at all. Conception is not co-extensive •
with understanding. The criterion of truth is not in
herent in ideas.

As it requires something more than clearness and
distinctness of ideas to know objects, how do we know
them? How do we know when our judgments agree
with them ? What is the criterion of truth ? Objects
are twofold, internal and external. About the internal
we judge immediately by sensation and consciousness,
the objects of sensation being effects of external objects
on the nervous system, the object of consciousness,
oneself as subject thinking in the widest sense : about
external but similar objects beyond ourselves we judge
from sensation and consciousness mediately by infer
ence. Truth is the agreement of a judgment with the
sensible, the conscious, or the inferred from the sensible
and the conscious, on the logical rule that, if the
premises are true, the conclusion is true. The criterion
of truth is double ; being first the immediate appre- !
hension of the sensible and the conscious within, and
thereupon the mediate apprehension of the similar
but insensible and unconscious without, by parity of
reasoning. Reasoning without the immediate data is
mere consistency, upon them it is the consistency of

K 2

132 PSYCHOLOGICAL IDEALISM TART IT.

,rutli. Knowledge is tlie apprehension of reality, imme
diately by sensation and consciousness, mediately by

ogical reasoning therefrom. To know by reasoning
•equires at least two conditions ; sensible data and

ogical consistency. Whether it requires more, we will
decide, when we come to Hume and Kant.

As, then, we know objects beyond ideas, not by the
clearness and distinctness of these ideas, but by rational
inference, what are the data required for this inference ?
This question is the crucial test of the Cartesian philo
sophy, which aspires to a knowledge of things through
ideas. Descartes did not supply adequate data to infer
the knowledge he admitted. Hence his philosophy
ends in inconsequence. We have already seen, in the
First Part, that it requires like data to prove like con
fusions, and, therefore, physical data to prove physical
conclusions. If all the data were psychical, physical
objects would not be inferrible. If all the data of
a man's knowledge were his soul and ideas, he could
know nothing but other souls and ideas. But Descartes
admitted that physical objects beyond souls and ideas
are knowable. The data of knowledge, then, cannot
fill be, as he supposed, a soul and its ideas.

Descartes was a man of subtle genius, retiring, as it
were, within the chamber of his own soul to survey his
own ideas, and trying to think what they could reflect
of a world without. Let us follow him into this retlre-
pnent, and imagine ourselves each to be a pure soul
contemplating pure ideas. A man must have a diili-

ulty in performing this feat, because he neither is the

me nor does the other. He cannot rid himself of his

jody, nor fail to contemplate the effects on his nerves.

Philosophers take advantage of this superhuman difu-

bulty ; feigning a psychical man, but knowing all the

CHAP. V.

DESCARTES 133

time that each of us will add the physical factor and!
complete his human being. Hence a man fails to realise!
the extraordinary consequences that follow, if he really
has to suppose himself to be a disembodied soul, per-j
ceiving nothing immediately but incorporeal ideas..!
However, let me try.

I should not be able, in the first place, to infer that
the body is a material cause of my ideas, nor that my
ideas are an efficient cause of moving the body. As all
the causes and effects immediately perceived by me
would be my psychical soul and its ideas, all those that
I could mediately infer would be psychical souls and
ideas. Now, Descartes asserted the heterogeneity of soul
and body, but not exactly their incommunicability, still
less the non-existence of the body. His view was that
soul and body are in contact in the pineal gland, that
the motions of the body cause ideas and ideas volitions, |
while this interaction requires the concourse or assistance!
of God. This hypothesis, or series of hypotheses, is \
anatomically false, because it disturbs nervous continuity ;
without proving any connection between the pineal
gland and thinking. Logically, it is false on Cartesian
principles, not merely because soul and body are sup
posed heterogeneous, but because all the causes and
effects immediately perceived being supposed psychical,
a physical body either as cause or effect of ideas could
not be inferred. There is no proof that Descartes him
self ever drew this conclusion, though involved in the
Cartesian theory. He kneiv that the body is scientifi
cally inferred to be cause and effect. Consequently, his
theory that soul and ideas are all the data of inference
must be false, because they cannot be the data of that
scientific inference.

It was left for his successors to draw the logical

134 PSYCHOLOGICAL IDEALISM PART n.

conclusion and contradict science. The Cartesian School
denied that the body is either cause or effect of ideas.
Instead, Geulinx invented occasionalism, or the hypo
thesis that on the occasion of bodily changes God calls
forth an idea of perception in our soul, while on the
occasion of an idea of volition in our soul He moves
our body for us. Malebranche developed this doctrine
into the vision of all things in the Deity. Leibnitz,
rightly characterising occasionalism as a perpetual
miracle, had recourse to a pre-established harmony
between body and soul, established by God before our
creation. But the pure idealists have a more logical way
out of the difficulty than any of the Cartesians. It is
that no body is known to exist at all. If all immediately
known causes and effects were my soul and its ideas, I
should have no data to infer a physical body, much less
that it is wound up like a clock to go with my soul.
Nevertheless, Descartes was right in saying that I have
a body, whose motions science proves to be causes
and effects of thinking. Therefore, immediately known
^causes and effects are not all my soul and its ideas, from
which no body could have been inferred.

Secondly, if all the data were my soul and its ideas,
and I could somehow or other infer the body, at any
rate I could not infer that my body was a part of
myself. How should I know that I have a body ?
Precisely as I should be supposed to know any other
external object, mediately through ideas. I should
have an idea of warmth, and refer it to a fire ; an idea
lof toothache, and refer it to the body. But if I knew
my body in this indirect manner, I should not regard it
•any more than the lire as part of myself. It may be
(objected that I should find it always with me. But so I
ido the earth and the atmosphere. It would seem with

CHAP.

DESCARTES 135

tliem part of my environment ; not a part of me, but
only my nearest and dearest companion. Descartes
vacillated on this point. When he is deducing the con
sequences of his hypothesis, he says, ' I am the mind by
which I am what I am, as distinct from the body.' l
When he is saving facts, he contradicts his hypothetical
deductions. ' Nature,' he says, ' teaches me by those
senses of pain, hunger, thirst, &c., that I am not only
present in my body as a sailor in a ship, but so closely
conjoined with it, and, as it were, intermixed, that I
compose something one with it ; ' !} and, again, ' it is
plainly certain that my body, or rather myself as a
whole, so far as I am composed of body and mind, can
be affected by various advantages and disadvantages
from surrounding bodies.' Quite so ; but he has given ,
us two inconsistent theories of personal identity, of
which the first is false, the second true, but quite incon
sequent, if I am a soul perceiving my own ideas.

If, then, I steadily suppose myself a soul perceiving
its ideas, I find that I cannot infer my own body to be
a part of myself. This is a conclusion so impossible,
so absurd, so ludicrous, yet so common to idealists, that
it is no credit to modern thought to have tolerated for
so long a time hypotheses from which it logically fol
lows. Eeally, Descartes was right in inconsequentlyj
and inconsistently admitting that he is body and soul. 1
But the admission is fatal to the hypothesis that he is
a soul, and to the hypothesis that the objects of all im
mediate perceptions are ideas. If I perceived nothing
but ideas, I could not know my body. Since I do
know my body, I must perceive something else but
ideas. The truth is, I know my body in four ways :
first, I am conscious of it as a factor of myself as

1 Discourse on Method, Part IV. a Meditation VI.

136 PSYCHOLOGICAL IDEALISM PART u.

thinking subject ; secondly, my senses perceive my own
nervous system as sensibly affected, although . I have
long confused this sensible object with the external
'cause I infer; thirdly, from one part of my body
sensibly affected I infer another part ; e.g. I see a re
flection on my optic nerve, and infer that it represents
my hand ; fourthly, by science, founded on all these
evidences, I know that I am a single organism. By
combining all these ways of knowing my body, I know
it better than anything else, and to be a part of myself.

Having feigned myself to be a pure soul contem
plating pure ideas, I could not infer my own body, or
at any rate not as part of myself. But could I infer
any external body ? Descartes, in a passage already
^quoted,1 dwells on the involuntariness of sensible effects,
and many of the idealists have relied on this argument
for an external cause. I freely admit the force of the
argument. But what sort of external cause ? I could
infer only causes similar to those in the data. Either
fcy sensation or by consciousness, or by both, I should
apprehend an interaction of my soul and ideas, and
of my ideas among themselves; and also that some
of my ideas are involuntary ; from which the parity of
reasoning would then allow me only three logical alter-
jnatives : another soul ; this would be Berkeley's Divine
'Spirit : a cause unknown ; this would be Hume's inex-
.plicable something : another idea ; this would be Hegel's
absolute idea.

A logical idealism would further conclude that, so
far from being known to be a physical part of myself,
interacting with my soul and ideas, my body, if known,
is something psychical, and, not being my soul, is a
system of my ideas, while any other soul, if there is

1 Princ. ii. 1.

CHAP. V.

DESCAKTES 137

iSuch a thing, must follow from another similar system
of my ideas. Such a logical deduction escaped Des-

'I cartes, but it has not escaped Mill, who only sub-

« ' stitutes sensations for ideas : 1 — •

' Whatever sensation I have, I at once refer it to
one of the permanent groups of possibilities of sensa
tion which I call natural objects. But among these
groups there is one (my own body) which is not only
composed, like the rest, of a mixed multitude of sensa
tions and possibilities of sensation, but is also connected,
in a peculiar manner, with all my sensations. Not only
is this especial group always present as an antecedent
condition of every sensation I have, but the other
groups are only enabled to convert their respective
possibilities of sensation into actual sensations by
means of some previous change in that particular one.
I look about me, and though there is only one group
(or body) which is connected with all my sensations in
this peculiar manner, I observe that there is a great
multitude of other bodies, closely resembling in their
sensible properties (in the sensations composing them
as groups) this particular one, but whose modifications
do not call up, as those of my own body do, a world of
sensations in my consciousness. Since they do not do
so in my consciousness, I infer that they do it out of
my consciousness, and that to each of them belongs a
world of consciousness of its own, to which it stands in
the same relation in which what I call my own body
stands to me.'

Now, the scientific Descartes knew well that bodies
are neither non-existent nor unknown, neither sensations :
nor ideas. He admitted that involuntary sensible dataj

1 Examination of Sir William Hamilton's Philosophy, pp. 244-5
cf. Lotze, Metaphysics, Book III. chap. iv.

138 PSYCHOLOGICAL IDEALISM PART ir.

[enable us to infer physical bodies as causes beyond
\sense and conception in the external world, that these
bodies consist of insensible particles, that the external
/world is like the sensible in some qualities, unlike in
lothers, and that the modes of insensible particles pro
duce sensible effects on our bodies, which are physical
parts of ourselves ; after the first step, proceeding
logically enough from inference to inference. Let us
add to our previous quotations one passage as a sample
of this profound scientific spirit :—

' But to the insensible particles of bodies, I assign
determinate figures and magnitudes and motions, as if I
had seen them, and yet I confess them to be insensible ;
and therefore some will perhaps ask, whence then I
recognise them such as they are. I answer that I first,
-from the simplest and most known principles, whose
knowledge has been implanted by nature in our minds,
jconsidered generally, what could be the principal differ
ences among the magnitudes and figures and positions
)f bodies, insensible only on account of their smallness,
ind what sensible effects would follow from those various
concourses. And then when I noticed some similar
effects in things sensible, I considered that they arose
from a similar concourse of such bodies ; especially since
no other mode of explaining them seemed capable of
being excogitated.' l

But the psychological Descartes could not logically
[take the first step. He had supposed, as 'the simplest
and most known principles,' hypotheses about the
/subject and its data, which never could have been the
I premises of such a science of bodies and their insensible
particles. If all immediately perceived effects and
Icauses had been soul and ideas, there would have been

1 Princ. iv. 203

CHAP. V.

DESCARTES 139

no primary data to infer bodies — not even one's own'
body, much less other bodies, and their corpuscles!
whose structures and motions cause sensible effects in
one's own body. But, as Descartes admitted, bodies are
known and inferred from sensible data. Therefore the!
data cannot be soul and ideas. From similars dis-
similars cannot be inferred. From soul and ideas, no
thing else follows. But something else is known to
science; therefore, not from soul and ideas. Physical!
bodies and corpuscles, structures and motions, require I
physical data of sense.

After the dogmatism of medieval philosophy, Dggr.,
cartes was right to doubt. He was right also in begin
ning with the certain fact of consciousness; I think,
therefore am. But, at the same time, he forgot that
there are other facts of consciousness. There is a
universal consciousness of the thinking subject, but
there is also a scientific consciousness that the thinking
subject knows physical objects. Instead of this, Des
cartes substituted the hypothesis that the thinking sub
ject is a soul which perceives ideas, and then, in defiance
of logic, attempted a synthetical deduction from this
idealistic hypothesis of psychical data of sense to a real
istic knowledge of physical objects of science. The de
duction may be attacked both by enstasis and elenchus ;
in its premises and in its conclusion. On the one hand,
the subject is not purely psychical, and, if it was, would
not be limited to psychical data : on the other hand, if
the data were psychical, we could not infer physical
objects of science, which are admitted by Descartes, and j
are more certain than any hypothesis of the nature of
the subject and its data. Hence the hypothesis of soul
and ideas must be surrendered, because the thinking)
subject is not the soul but the man, because sensible i

140 PSYCHOLOGICAL IDEALISM PART ir.

objects are not ideas but physical effects on the nervous
system, and because soul and ideas would not enable
man to infer physical objects of science. Descartes, the
original genius of modern idealism, was too introspective.
Of himself he says, ' Totos dies solus in hypocausto
morabar, ibique variis meditationibus placidissime vaca-
bam.' l This seclusion in a hot room is an admirable
way of distilling thoughts, provided only these vapours
of the heated brain can be condensed into a knowledge
of the outside world.

1 Diss. dc Methodo, ii.

141

CHAPTEE VI.

LOCKE.

LOCKE, at the outset -of the c Essay concerning Human
Understanding,' states that it is his purpose to enquire
into the original, certainty and extent of human know
ledge and opinion, without troubling himself about the
essence of mind.1 That js^ he r eject s_tlie Cartesian
method of using the nature of the thinking subject to
cTecluce our knowledge ; and rightly, because it was a
method from the less to tlieTmore certain. But he leaves
the Cartesian deduction, that the data o£ the under
standing are ideas, simply removes the hypothesis from
the premises to the conclusion, and no where throughout
gives any new evidence that ideas are flip ^Q*Q r>f Irnrror.
jedge. The hypothesis of the soul is thus replaced by
The hypothesis ot ideas, as a principle. Now, there had
been some plausibility in the argument — the subject
is the soul, therefore its immediate objects are ideas.
There was iiotliing'butpetitioprincipii in the hypothesis -
the immediate objects of understanding are ideas. Yet
this hypothesis in one form or other has remained ever
since Locke's time as the putative principle of all idealism.
Many a philosopher, who has with Locke recovered
from the Cartesian hypothesis that the subject is soul,
and has followed Hume in correcting Locke's confusion
of sensations and ideas, nevertheless clings to the hypo-

1 Essay, I. 1, 2.

142 PSYCHOLOGICAL IDEALISM run n.

thesis that all immediate objects are some psychical state
or other, without any evidence, whether of Cartesian
deductions, or of psychological consciousness, or of
natural science.

Locke, having begun at a new beginning, pro
ceeds to his method, which is as synthetical as that of
Descartes : —

' I shall pursue this following method.
*- ' First, I shall inquire into the original of those ideas,
notions, or whatever else you please to call them, which
a man observes, and is conscious to himself he has in
his mind; and the ways whereby the understanding
comes to be furnished with them.

'Secondly, I shall endeavour to show what know
ledge the understanding hath by those ideas, and the
certainty, evidence, and extent of it.

' Thirdly, I shall make some inquiry into the nature
'and grounds of faith or opinion, whereby I mean that
assent which we give to any proposition as true, of
whose truth yet we have no certain knowledge. And
here we shall have occasion to examine the reasons and
degrees of assent! l

From this passage w^e can see how vain is psycho
logical synthesis. The smallest mistake at the beginning
vitiates the whole procedure and every consequence. A
man is here said to be conscious of having ideas in his
mind. It is true that he is conscious of having ideas.
But even the followers of Locke himself would deny
that this is all he is conscious of. Hume would say
that he is also conscious of impressions, and Mill would
add judgments. Yet to a philosophical use of the syn
thetic method by Locke it was necessary that ideas should
be all the materials of knowledge ; for the next question

1 Essay, I. 1, 3.

CHAP. vr. LOCKE 143

is -wlial knowledge can be gained by ideas; which is
a false issue, if ideas are not the whole material of
knowledge. But as they are not the whole, it is not to
be wondered at that, in the sequel, Locke oscillates
between two contrary tendencies, a logical but false
^reduction of knowledge to ideas, and an illogical but
true extension of it to things beyond. Moreover, to
inconsequence he adds inconsistency. He tries to begin
with an understanding of ideas and end with a know
ledge of things.

The firstfruits of idealistic hypothesis are at once
manifest. Having assumed that ideas are all the materials,
he consistently assumes that they are all the objects of
understanding : —

' Thus much I thought necessary to say concerning
the occasion of this inquiry into human understanding.
But before I proceed on to what I have thought on this
subject, I must here, in the entrance, beg pardon of my^
reader for the frequent use of the word idea, which he
will find in the following treatise. It being that term
which, I think, serves best to stand for whatsoever is
the object of the understanding when a man thinks, 1^
have used it to express whatever is meant by phantasm,
notion, species, or whatever' it is which the mind can be
employed about in thinking ; and I could not avoid
frequently using it.' l

These words, which, if anywhere, ought to have
come as a proved conclusion at the end, occur as an
undoubted principle at the entrance of the Essay. They
contain a double hypothesis ; first, that ideas are the
immediate, secondly, that they are all, the objects of ^
understanding,(and therefore of knowledge. ?/ The first
part is the ideal hypothesis of Descartes, the second is

1 Essay, I. 1, 8.

144 PSYCHOLOGICAL IDEALISM PART IT.

Locke's corollary. It is a logical corollary, not however
scientific, but hypothetical from an hypothesis. Three
hypotheses started modem idealism; the subject is
psychical, the data are psychical, the objects are
psychical. Never was such a gigantic system of petitio

principii.

The aftermath of idealistic hypothesis appears at the
very end of the Essay. After adopting ^the Stoic divi
sion of the sciences into physics, ethics and logic, he
concludes in the spirit of science, but in utter contra
diction of his original hypothesis, with the following
peroration :—

' This seems to me the first and most general, as well
as natural division of the objects of our understanding.
For a man can employ his thoughts about nothing, but
either the contemplation of things themselves, for the
discovery of truth, or about the things in his own power,
which are his own actions, for the attainment of his own
ends ; or the signs the mind makes use of, both in the
one and the other, and the right ordering of them for its
clearer information. All which three, viz. things as they
are in themselves knowable ; actions as they depend on
us, in order to happiness ; and the right use of signs in
order to knowledge, being toto cado different, they seemed
to be the three great provinces of the intellectual world,
wholly separate and distinct one from another.' l

In the same chapter he has already told us what he
includes under things and signs. On the one hand, by
signs he means ideas and words.2 On the other hand,
under things, he includes, as objects of understanding,
' the nature of things, their relations and their manner
of operation ; ' while physics is, as he says, ' the know
ledge of things, as they are in their own proper beings,

1 Essay, IV. 21, 5. = Ib. IV. 21, 4 ; cf. IV. 5.

CHAP. TI. LOCKE 145

their constitutions, properties, and operations, whereby
I mean not only matter and body, but spirits also.' *
He finally admits, then, that things and ideas are toto
ccelo different, that not ideas but things and their rela
tions are the objects of physics and natural philosophy,
and that not only ideas, but also things, are objects of
understanding, knowledge, and science. Which was
right, the original hypothesis of ideas, or the final admis- ^
sion of things ? The latter, because things inconceiv
able but not incredible are objects of science. Locke, like
Balaam, came to curse, but went away blessing.

To return to the original hypotheses : the conse-
quence is that the whole emphasis of the Essay falls, on
the origin of ideas, as Locke himself admits at the very*-
beginning of the Second Book.2 'Every man being
conscious to himself, that he thinks, and that which his
mind is applied about, whilst thinking, being the ideas
that are there : ' this is his assumption : ' it is in the
first place to be inquired, how he comes by them : ' this
is his hypothetical conclusion. Meanwhile, the origin
of knowledge is postponed till the Fourth Book, and the"
Second and Fourth Books are never welded together.
This is the beginning of a serious evil in modern philo
sophy, the emphasis laid on the origin of ideas in pre
ference to the far more important problem of the origin
of knowledge, and the tendency to let the limits of ideas
determine the extent of knowledge. But ideas do not
dictate knowledge so much as knowledge dictates ideas.

Locke begins the problem of the origin of ideas well by
rejecting innate ideas. No doubt most of his objections
touch the broader form of inborn ideas rather than the
elicited ideas of Descartes. But they have the merit of
pointing out that many ideas, supposed universal, are not

1 Essay, IV. 21, 2. 2 Cf. also I. 1, 8.

L

146 PSYCHOLOGICAL IDEALISM PART n.

possessed by savages, and Locke in this respect has been
confirmed by modern travellers.1 Moreover, he touches
Descartes himself, when he shows that experience has
not really been eliminated,2 and that consciousness, or,
* as he usually says, reflection, is not a mystical revela
tion, but an inner sense. As sensation apprehends sen
sible objects only when present, so consciousness appre
hends one's own operations only when one performs
them. There is no greater source of error in philosophy
Jhan the confusion of the intuitive with the a priori^ and
of the conscious with the innate. Consciousness is an
intuitive, not innate nor a priori, experience of oneself
performing operations.

Locke did a signal service in showing that there are
two kinds of sense, sensation and reflection :—

' Let us then suppose the mind to be as we say,
white paper, void of all characters, without any ideas ;
how comes it to be furnished ? Whence comes it by
that vast store, which the busy and boundless fancy of
man has painted on it, with an almost endless variety ?
Whence has it all the materials of reason and know
ledge ? To this I answer in one word, from experience :
in that, all our knowledge is founded ; and from that it
ultimately derives itself. Our observations employed
either about external sensible objects, or about the internal
operations of our minds, perceived and reflected on by our
selves, is that which supplies our understandings with all
the materials of thinking. These two are the fountains of
knowledge, from whence all the ideas we have, or can
naturally have, do spring.' 3

Modern criticisms of this theory often turn on the
comparison of the mind to white paper, which is said to

1 Cf. Moffat, Missionary Labours in South Africa, chap. ix.
* Cf. Essay, I. 2, 1 ; II. 1, 1. » II. 1, 2.

CHAP. vi. LOCKE 147

be inconsistent with evolution. But Locke rather over
looks evolution than contradicts it. The theory of
evolution is often exaggerated. It shows that our senses
become more and more readily adapted to apprehend
their objects when presented. It does not show that
they ever come to evolve ideas or apprehend anything a
priori. Its evidence is in favour not of a priori ideas, but
only of intuitive perception. For example, the more the
senses have been exposed to light the quicker and clearer
they have apprehended its sensible effects when presented;
by its action a special sense of vision has been gradually
evolved to perceive them ; but there is no evidence that
at last vision will, of itself, without light being presented,
apprehend sensible light a priori. Such a jump from
light presented in sense to light constructed by sense is
not proved by evolution, but arises from confusing the
intuitive with the a priori. Evolution has shown that
we hereditarily tend to use our organs better, and that
by use the organs become more differentiated ; but it
has not shown that they ever evolve an a priori idea.
Now Locke, it is true, overlooked hereditary adapta
tion ; but he was quite right, and would be right to-day,
in resisting a priori ideas, in saying that we begin
' without any ideas,' and in recognising two kinds of
sensitive intuition, sensation and reflection, both pre-
sentative.

The really vital question for the critic of the Essay
is a question seldom asked. All knowledge begins^
with sense ; but what are the objects of sense ?_ Locke's
answer is, ideas ; ideas presented to both senses ; ideas '
of sensation and ideas of reflection. His doctrine of ideas
was modelled on that of Descartes. Perhaps he dis
tinguished sensation from conception better than his pre
decessor, but he left the consequences of the Cartesian

L 2

148 PSYCHOLOGICAL IDEALISM TART n.

confusion of these operations in his doctrine that ideas
are the objects alike of sensation and conception. Again,
he followed Descartes in confusing object and operation,
the idea and its perception.1 The object of sensation,
then, being regarded as an idea, the idea as nothing
but its perception, and the perception as psychical, it
follows that the object of sensation, with Locke as with
Descartes, becomes a psychical result in our minds, dis
tinguished not only, as it should be, from the external
object, but also from the nervous impression, as it should
not. 'If then,' says he, in speaking of the ideas of
primary qualities, ' external objects be not united to
our minds when they produce ideas in it, and yet we
perceive their original qualities in such of them as singly
fall under our senses?/ ''tis evident that some motion *
must be therein continued by our nerves, or animal
spirits, by some parts of our bodies, to the brain, or the
seat of sensation, there to produce in our minds the
particular ideas we have of them."2 Thus the ideal
hypothesis of Descartes was accepted by Locke, and
without further evidence. Moreover, it remains to this
day the current hypothesis, with the sole alteration of
idea into sensation. But, as we have already found, the
sensible object, though internal, is not the sensitive opera
tion. Even, then, if the sensation were a purely psychical
operation, it would not follow that the sensible object
is either a psychical sensation or a psychical idea.

Locke also added to the doctrine of ideas ; and his
fir^t addition was the sensible idea of resistance.3 Des
cartes, with his mathematical genius, had emphasised
the mathematical qualities of body, and especially exten
sion, by which he defined it.4 Locke accepted extension

1 Essay, II. 8, 8 ; II. 10, 2 ; II. 19, 1.
- II. 8, 12. :i II. 4. 4 Descartes, Princ. ii. 4.

CHAP. VI.

LOCKE 149

and the sense of extension, but went on to show that
resistancejs also necessary to body, and therefore added
a sense of resistance. But he spoilt this great contri
bution to the philosophy of matter and sense by his
theory of ideas. If I perceived nothing but sensations
or ideas, I should perceive only a sensation or an idea of
resistance. I should not perceive one part of my body,
JDP. nervous system, resisting jm_other ; I should have to
infer it. But there would be no data for the inference,.

for from psychical sensations or ideas physical resist
ance between parts of a body would not follow. The
sense of resistance, therefore, supplies a new argument
to prove that the real object of sense is the nervous
system, and its various parts resisting one another.

A^ain, Descartes had confined the ideal theory to
sensation ; he had allowed a direct consciousness of
thinking. Locke, with more consistency though with
less truth, interposed an idea not only between outer
sense and the nervous impression, but also between
inner sense audits operations ; so that the direct objects*'
are ideas of sensation and ideas of reflection. But
Descartes rightly regarded consciousness as a direct
apprehension of thinking ; and Locke, instead of trans
ferring the ideal theory to consciousness, should have
retracted its application to sensation, and regarded sen
sation as a direct apprehension of the nervous impression.

Thirdly, Descartes had begun by saying, ' I am
conscious that I think, not of thinking.' The object of
consciousness is not the quality, thinking, but a thinking
subject. Inconsistently with this truth, when he came to
substance, he had fancied that we do not directly per
ceive it, but ' from perceiving that some attribute is
present, we conclude that some existing thing, or sub r
stance to which it can be attributed, is also necessarily

150 PSYCHOLOGICAL IDEALISM PART n.

present ; ' l and he had applied this theory both to soul
and body. Locke developed this hint into a formal
theory that we perceive the simple ideas of qualities,
while 'we accustom ourselves to suppose some sub
stratum, wherein they do subsist and from which they do
result, which, therefore, we call substance.' 2 This theory
he applied to body and spirit ; and from him has de
scended the ordinary hypothesis that the objects of
sense and consciousness are qualities, while substance
is inferred — without data to infer it.

This error will meet us again in this chapter. At
present it will be sufficient to quote a passage from
another part of Locke's Essay : ' Our simple ideas have
all abstract as well as concrete names : the one whereof
is (to speak the language of grammarians) a substan
tive, the other an adjective ; as whiteness, white ;
sweetness, sweet.' 3 This is the well-known logical dis
tinction of abstract and concrete, but its consequences
are often overlooked. Locke, for instance, forgot to
1 ask in which meaning he should call a simple idea. a.r>
object of sense, The abstract whiteness is a quality ;

the concrete white is the qualified. Now, nobody ever

saw whiteness ; the object of vision is the white, the
red, &c. Similarly, the object of taste is not sweetness,
but the sweet ; and so on with all sensible objects.
Universally, then, an object of sense is never a quality^
but always the qualified ; and a quality is an abstrac-
tion; and, though we may sometimes speak of perceiving
it, we do so only for convenience. But the qualified is
a substance ; whiteness and sweetness are qualities, but
the white and the sweet are substances. The object of
jsense, therefore, is always a subatanr.p. I do not mean
that sense perceives a whole substance at once, but only

1 Princ. i. 52. 2 Essay, II. 23, 1. 3 III. 8, 2.

CHAP. VI.

LOCKE 1-jl

so far as it is sensible to a given sense ; sight perceives
a substance so far as it is white ; taste perceives a sub
stance so far as it is sweet, and so on. Nor do I mean
an external substance, for I am a substance, consisting,
too, of an immense plurality of substances, which I per
ceive so far as they are sensibly affected.

These conclusions apply both to outer and inner
sense. In sensation, I perceive not a mere quality, nor
a whole substance at once, nor an external substance ;
I perceive my nervous system, not so far as it is ner
vous structure moving, but so far as it is sensibly
affected in different parts, the optic nerve so far as it is
visibly white, the gustatory nerve so far as it is sweet
to taste, and so on. Similarly in consciousness, I perceive,
not mere thinking, nor the whole of myself, but myself
thinking, in the manner described in the last chapter.
The object of my sensation is myself as a physical sub
stance sensibly affected ; the object of my consciousness
is myself as a thinking substance. Descartes rightly said,
'I think.' He ought not to have deserted this prin
ciple. Locke ought to have returned to it, and have
applied it from consciousness to sensation. Modern
philosophy ought now to give up the sensation of
qualities and inference of substance, because there is a
direct sensation of my nervous system sensibly affected,
and a direct consciousness of myself thinking, both of
which are senses not of qualities, but of the qualified.
We have a sense of substances, in order to infer them.
Locke's complete theory is that alj_sense perceives
a^ simple idej^. Ttegjty sense always perceives a. sub
stancejgualified. It is doubtful whether the substance,
as perceived, is ever simply qualified ; for instance,
even when I feel simply pained, I doubt whether I do
not feel pained for a time. But, in any case, I do not

152 PSYCHOLOGICAL IDEALISM

PART II.

perceive anything simple in the sense of a simple
quality, which is only simple in the sense of abstract ;
but I perceive at least the simply qualified. Secondly,
I do not perceive anything simple in the sense of a simple
idea, which is really conceived, not perceived ; but I

• perceive, in sensation, my nervous system sensibly
affected, and in reflection, myself thinking. The object
of sensation, and the object of consciousness, so far
from being simple ideas, are not ideas at all. They
are two sets of materials of knowledge, of which
neither is a quality, and neither is an idea, but each
a substance. Locke's attempt to make the origin of
ideas determine the origin of knowledge breaks down
at the very outset by substituting abstractions for con-

^crete data of sense.

At the end of what he has to say on simple ideas,1
Locke comes to the operations which he supposes to
make other ideas out of them, and to the ideas thus
made.2 ' The acts of the mind,' says he, ' wherein it exerts
its power over its simple ideas, are chiefly these three :
First, combining several simple ideas into one compound
one, and thus all complex ideas are made. The second, is
bringing two ideas, whether simple or complex, together,
and setting them by one another, so as to take a view of
them at once, without uniting them into one, by which
way it gets all its ideas of relations. The third, is separat
ing them from all other ideas that accompany them in
their real existence. This is called abstraction, and
thus all its general ideas are made.'3 He then re
marks that ideas, made up of several simple ones put
together, he calls complex ; such as are beauty, grati
tude, a man, an army, the universe. Next he divides
complex ideas under three heads: modes, substances,

1 Essay, II. 2-11. 2 II. 11 scq. to the end of the Third Book. 3 II. 12. 1.

CHAP. VI.

LOCKE 15o

relations. Complex ideas of modes are ideas of affec
tions of substances, subdivided by him into simple,
or combinations of the same simple idea, e.g. a
dozen, formed of units, and mixed, or combinations
of simple ideas of several kinds, e.g. beauty, theft.
Complex ideas of substances are ' such combinations of
simple ideas as are taken to represent distinct particular
things subsisting by themselves, in which the supposed,
or confused idea of substance, such as it is, is always
the first and chief ; ' l they are subdivided into ideas of
single substances, e.g. a man, and collective ideas of
several substances, e.g. an army. ' The last sort of
complex ideas,9 he says, ' is that we call relative, which
consists in the consideration, and comparing one idea
with another,'2 e.g. father and son, bigger and less,
cause and effect.3 The consideration of all these com
plex ideas in their order occupies the remainder of the
Second Book ; while that of abstract ideas follows, along^
with general words, in a general treatment of language
in the Third Book.4

The whole discussion is full of variety. But it is
vitiated by two incurable errors. In the first place,
the objects of knowledge are complicated with their "
mere ideas. But many scientific objects are known to
exist, without being conceivable. Secondly, no thorough
analysis is attempted of the three acts of mind, which*
are supposed to be the sole causes capable of producing
out of simple ideas^ all other ideas. Locke calls them
composition, comparison and abstraction ; 5 making the
first to be the origin of all complex ideas of modes and
substances, the second the origin of all complex ideas
of relations, the third the origin of all general ideas.

Essay, II. 12, 6. 2 II. 12, 7. 3 II. 25, 2.

4 Of. II. 33, 19. 5 II. 11.

154 PSYCHOLOGICAL IDEALISM

PART 11.

He saw the foundation of these operations on sense ; but
he forgot to ask their relation to reason.

After sense, we conceive particular ideas in the
^reproductive imagination, and general ideas by abstrac
tion from sense. It does not follow that all general ideas
are thus formed ; on the contrary, it is impossible that the
idea of an insensible object should be either reproduced
or abstracted from sense, in which it has never been.
Again, we may compare and compound ideas. But at
the same time we also judge about sensible objects
and apprehend their relations. In judgment we use
ideas, particular and general. But, as Mill has pointed
out,1 we also judge about sensible objects in order to
apprehend their relations. I am in pain; this is a
judgment that I, who am real, am in pain, which is
real. Now, reasoning starts from such judgments about
the relations of sensible objects, and sometimes by
analogy, sometimes and better by induction and deduc
tion, infers rational judgments, no longer about simple
objects, nor about ideas, but about the relations of real
objects; on the principle, if the premises are true,
the conclusion is also true. That is, starting from
judgments of sense, we infer rational judgments on evi
dence about relations, as real as the sensible relations.
Nor is this all ; as I showed in the last chapter, reason,
having from sensitive concluded rational judgments,
forms indirect ideas, roughly corresponding to the objects
inferred, like to the ideas of sensible objects but not
the same, and only capable of being made by reason.
For instance, reason, having inferred that there are
particles in bodies, causes the idea of a corpuscle ; a
general idea of corpuscles, which is not a result of mere
abstraction, and particular ideas of this or that cor-

1 Mill, Logic, i. 5, 1.

CHAP. VI.

LOCKE 155

puscle, which are not results of composition and com
parison of ideas, but of inference from judgment to
judgment. Beyond sense and imagination, besides
composition, comparison and abstraction of ideas, there
are also judgments of sense about the relations of sen-/
sible objects, and reasoning from these judgments to
the relations of insensible objects, producing rational
conceptions of ideas, due to no other source but reason
ing. The narrow problem of the origin of ideas cannot
be separated from the whole problem of judgment, "^
reasoning, and the origin of knowledge.

Locke, in the Second and Third Books, saw only
one side of thinking, and that its weakest side :
imagination and abstraction, comparison and com-6'
position, of ideas from sense. Eational inference of
realities, beyond sense and ideas, he allowed to fade
into the distance of the Fourth Book. Consequently,
he found only the direct sources of ideas, and missed
their indirect source in reason. No doubt he was in
fluenced by the Cartesian logic of his day, which knew
only the order — idea, judgment, reason. But there is
a second order — reason, judgment, idea. As soon as
judgment begins to act on the senses, reason begins
with it, and, never stopping except to sleep and rise
again refreshed, constantly forms new judgments issuing
in new ideas. But Locke postponed reasoning, ignored
rational conception, and therefore always fell short even ^
of the origin of ideas.

Even in the ideas of simple modes, the very simplest
department of complex ideas, this defect is noticeable.
After sensations of motion, we may form ideas of motion
by imagination ; and the ideas of simple modes of
sensible motion by composition.1 But reason also

1 Essaij, II. 18, 1-2.

156 PSYCHOLOGICAL IDEALISM PART IT.

infers simple modes of insensible motion in nature,
such as electricity and magnetism, cohesion and chemi
cal attraction, which were never in sense, and frames
indirect ideas of these motions. Similarly, we may
imagine ideas of sensible duration and extension,
and compound ideas of these simple modes ; but when
Locke goes on to suppose that the mind extends itself
to infinity simply by repeating these ideas, he neglects
the rational evidences of the unbounded nature of time
and space. Unless men had thought they had reason
to infer infinity, no mere repetition of ideas of the finite
would ever have given the idea of the infinite, which is
always accompanied by a rational inference that the in-
* finite itself is beyond any idea we can possibly form
of it.

The mischievous consequences of omitting reason in
the formation of ideas are best seen in Locke's doctrine
of mixed modes and relations. Without reasoning, mere
composition and comparison, as soon as they go beyond
sense, would produce at most artificial ideas, the va
garies of imagination. Consequently, it is not sur
prising that Locke treats the ideas of mixed modes and
relations, which he supposes to be formed by pure com
position and comparison from and beyond sense, as
artificial, and even goes so far as to contend that not
merely the ideas, but mixed modes and relations them
selves, have no other reality but what they have in the
minds of men, and are real only in the sense of being
consistent, not in the sense of representing real things.
This paradox is a serious matter, for it affects the reality
not only of a mixed mode, such as beauty, or a rela
tion, such as father and son, but all moral modes and
relations. It reduces morality itself to an idea.1 ^

1 Cf. Essay, II. 22, 2 ; II. 25 ; II. 30, 4 5 ; II. 32, 10 ; III. 4, 2; III. 5.

CHAP. vi. LOCKE 157

But obligation is a mixed mode, which is real ; theft,
drunkenness, lying, are mixed modes which are only too
real, and the conformity of morality to law is a rela
tion, which is also real, though perhaps less common ; and
the complex ideas of these mixed modes and relations
are not artificial, but really, though inadequately, corre
spond to real morality and immorality. We may admit
that morality is not altogether immutable ; it is not
therefore unreal. We may admit that the ideas of the
beautiful, of the good, and of law, are differently com
pounded in ancient and modern morals ; they are not
therefore artificial. We may admit that actions of
virtue are uncommon ; but virtue is not an idea. By
reasoning, man finds out the moral relations suited
partly to humanity in general, and partly to the cir
cumstances of his time. By rational conception, he
apprehends ideas of moral relations, immutable and
mutable. Happy he who can also realise these ideas,
and be

Virtu tis verse custos rigidusque satelles.

There is even a certain fashion of ideas, which
Locke illustrates by the Greek idea of ostracism and the
Eoman idea of proscription. But these ideas were not
on that account artificial : they represented real mixed
modes at Greece and Borne : to be ostracised or pro
scribed was anything but a mere idea. The Greeks and
Romans inferred -that these institutions would serve
certain purposes, and thus both established the real
mixed modes and represented them by corresponding
ideas. The modern historian from his evidence infers
that these mixed modes existed in the past, and con
ceives the ideas in the present. Similarly, the relation
of paternity is not the idea of that relation, nor a mere
product of comparison. It is a real relation of generation,

158 PSYCHOLOGICAL IDEALISM TART n.

which from sensible data we infer really to take place,
and of which we afterwards form an idea, rational and by
no means artificial, though but superficially represent
ing the actual physical process of propagation. Mixed
modes and relations, and their ideas beyond sense, are
not always artificial constructions of composition and
comparison of sensible ideas ; but reasoning from judg
ments of sense discovers real mixed modes and relations,
and then forms indirect ideas, really, though inade
quately, corresponding to these realities, in science, in
art, and in morals.

The fallacy of omitting reason again appears in
Locke's treatment of universals in the Third Book.
He thinks that the sole source of general ideas is direct
abstraction from sense. The consequences he draws are
that all classes are abstract ideas, that no real essence
is knowable beyond ideas, that simple ideas are unde-
finable, and that universal truths are merely the agree
ments and disagreements of our abstract ideas.1 All
these consequences would follow if we had no higher
power than abstracting general ideas from particular
sensible objects. All classification would be artificial.
But there is a second source of general ideas. Eeason,
by discovering the numerous similarities of particulars,
infers real kinds or natural classes, which are not indeed
eternal but as constant as the similarities, and thereby
causes new, general, often very indirect ideas repre
senting these real classes, but not identical with them ;
e.g. the rational general idea of a corpuscle. Again,
a simple idea of sensible light is undefinable ; but light
in the universe is not, as Locke thinks, undefinable.
On the contrary, optical reasoning proves that the real
essence or fundamental similarity on which its pro-

•i Essay, IV. 3, 31 ; IV. 12, 7.

CHAP. vi. LOCKE 159

perties depend is an gethereal undulation, and defines
it accordingly. Lastly, whatever we may think of
essences and definitions, if Locke's theory that direct
abstraction is the sole source of general ideas, and that
classes are abstract ideas, were true, it would follow
that all uniformities would be universal relations of
abstract ideas ; and he accepted the consequence ; even
the variety of Locke's mind refusing to entertain a con-
ceptualism of classes along with a realism of natural
laws. If ships and liquids were abstract ideas, the laws of
flotation would be universal relations of abstract ideas.
These laws, however, are universal relations of real ships
and real liquids, inferred by reason. Therefore the
classes so related are realities beyond abstract ideas.
Abstraction of ideas from sense is not the sole source of
generality, as Locke thought : reason infers natural
classes and laws, and indirectly produces general ideas,
not identical with them, but representing them, not arti
ficially but really, though inadequately.

Curiously enough, Locke himself saw, through a glass
darkly, the interference of reason in the origin of one
complex idea, that of substance. If sense perceived
simple ideas of qualities, and composition united simple
into complex ideas, the only complex ideas we could
have would be complex ideas of qualities. We might
have, for example, a complex idea of a combina
tion of extension, solidity, motion, thinking, and no
thing more. But Locke saw that we have something
more. He, therefore, suddenly introduced, beyond
sense and over and above composition, a supposition ;
and says that ' not imagining how these simple ideas
can exist by themselves we accustom ourselves to ^
suppose some substratum, wherein they do subsist
and from which they do result ; which, therefore,

100 PSYCHOLOGICAL IDEALISM TART n.

we call substance' 1 Secondly, lie allowed that this
supposition causes an obscure and confused idea of the
supposed but unknown support of qualities. He re
cognised two such supposed and conceived substances :
body, the substratum to those simple ideas we have
from without ; and spirit, the substratum to those we
have from within.2 Finally, he regarded both these
substances as unknown, and neither of their ideas as
clear and distinct. Nevertheless, he thought that the
ideas of substance were real in a different way from those
of other complex ideas. The complex ideas of mixed
modes and relations were, according to him, real if con
sistent ; those of substances real only if agreeing with
things without us.3 It is the supposition of existence,
over and above the composition of ideas, which made
him allow this agreement with existence to ideas of
substances. Inconsistent as this supposition is with
his general theory of the composition of complex
ideas, it is nevertheless the truth, though in a very
imperfect shape. Let us then proceed to correct it,
by showing what is the real nature of this inference,
which Locke calls a supposition.

It is true that external substances are inferred. But
there are three views of what a substance is inferred to
be. Some say that it is only a combination of qualities.
But qualities are abstractions ; and a body is not ex
tension, solidity, motion, or any number of further
abstractions, combined, but the extended, solid, moving,
&c Locke went to the opposite extreme of supposing
a substance to be a substratum or kind of support on
which the qualities rest, and this is the ordinary view,
^ descended indeed from the compound, or ' concrete,'
substance of Aristotle, composed of matter and form.

1 Essay, II. 23, 1. ~ See II. 23, 1-5. 3 II. 30, 4-5.

CRAP. VL LOCKE 101

But here are two abstractions, the subject abstracted
from the qualities and the qualities from the subject.
If a body ceased to be extended, solid and moving, it
would cease to be ; there would be no substratum or
support left. Hence the third view, that a substance is
a qualified subject, the extended, solid, moving, &c. ; in
which the qualities are nothing except as characterising
the subject, and the subject nothing except as charac
terised by the qualities ; from which subject or sub
stratum, qualities or attributes are opposite abstractions.

Secondly, external substances must be inferred from
similar data. To infer qualified subjects beyond sense,
there must be qualified subjects in sense. If the data
were ideas, we could only infer other ideas. If the
data were qualities, we could only infer qualities. A
fortiori, if the data were ideas of qualities we could
never infer a real qualified subject, for which there
would be no analogue. Therefore, again we find that
Locke's sensible data were false. He thought that by
sense we perceive simple ideas of extension, resistance
or solidity, motion, &c., and then without rhyme or
reason suppose something totally different, a real sup
port in the external world. Eeally, sense perceives
qualified subjects, the extended, resisting, moving, &c.
within ; hence reason infers similar extended, resisting,
moving, qualified subjects without. It must not be
forgotten that muscular sense was not noticed in Locke's
day ; but the logic of reason had been known since
Aristotle's day, and he ought not to have neglected it.

Thirdly, substances are not unknown : they are the
only things that are known. Everything else is real/
and is known, only so far as it belongs to substance ;
and although qualities are abstracted and spoken of as
real and known for mere convenience, what is really

M

162 PSYCHOLOGICAL IDEALISM PART IT.

known in mathematics is not the quality of extension,
but the extended ; in physics, not the quality of gravity,
but the gravitating ; in morals, not the quality of good
ness, but the good. Substances would be unknown, and
uninferred, on Locke's data. But substances are known,
because sense perceives them within, and reason infers
them without, by parity of reasoning. They are the
data and conclusions of all our knowledge. Sensation
perceives the nervous system in different parts as sensibly
white, sweet, extended, moving, &c. Eeason infers
similar physical substances or bodies. Science goes on to
infer similar corpuscles. Nor does it stop till it infers the
body of the universe. Consciousness perceives myself
as thinking subject, partly body, partly soul. Eeason,
from the signs of bodily organs, language, actions, and
productions of others, infers similar thinking subjects.
Natural theology, not from bodily organs, but from
physical creations, infers God, not as a body, but as a
Creator. All this is knowledge of substance, logically
inferred from sensation and consciousness ; and only
because the objects of outer and inner sense are sub
stances, can reason logically infer substances, physical
and psychical. It does not follow, however, that reason
is infallible : it is fallible so far as not logical from sense.
Nor does it follow that we know substance completely.
We begin with sense, and perceive subjects only as
sensibly qualified. Eeason reveals subjects insensibly
qualified. But we never know the whole of any sub
stance whatever, not even ourselves, not even a crystal
which we seem to see through and through. This
imperfection of human knowledge misleads philosophers
into agnosticism. But the truth is, sense and reason
enable us to know substances not wholly but partially.
Finally, the knowledge of substance creates the idea

CIFAP. vi. LOCKE 103

of it. The original ideas are derived from my own sub
stance. From myself as sensible I derive my idea of a
physical subject ; from myself as conscious, my idea
of a thinking subject, partly physical, partly psychi
cal ; from both, my idea of a qualified subject. But
my ideas of all other known subjects are results of
reasoning, which first infers similar subjects, and then
forms ideas of them. Ideas of substance are right, so
far as they correspond to really known substances
sensible and inferred, and their correctness varies in
accordance with sense and reason. They are clear,
distinct, and adequate, in proportion partly to their
proximity to sense, and partly to the extent of reasoning
about any given substance ; but they are seldom or
never adequate to what is known of a substance.

Locke, though inconsistent, was justified in allowing
that the complex ideas of substances are not due to
mere composition of simple ideas ; and he ought to
have made the same admission in the case of other
ideas, because not all ideas of mixed modes are due to
composition, nor all ideas of relations to comparison,
nor all general ideas to abstraction. He was justified
in allowing that we infer substance, in order to conceive
the idea of it, beyond ourselves. He was justified in
allowing that ideas of substances are right, so far as
they represent real objects. But he was unable to
found a philosophy of substance, because, in the first
place, he failed to apprehend that sensation and reflec- *
tion both perceive substances within ; secondly, he was
accordingly, but falsely, constrained to reduce the in
ference of substances without to a mere supposition — a
supposition without any data, illogical, and impossible
to reason ; thirdly, he had to call all substances, all
qualified subjects, the only things in the world we

M 2

164 PSYCHOLOGICAL IDEALISM TART n.

know, unknown ; and all ideas of substances obscure
and confused, when really the clearest and distinctest
ideas we have are those of stones, waters, houses, plants,
animals, cats, dogs, men and other substances.

There are many sources of ideas. Sensation and
V reflection are not directly concerned with ideas, as
Locke thought, but with sensible objects. But after
sense, reproductive imagination without reasoning con
ceives particular ideas of the objects of sense, memory
refers the ideas to their objects, and abstraction con
ceives general ideas of the objects of sense. Eeasoning
infers insensible objects and forms their ideas. There is
a rational imagination of ideas. Eational ideas of known
objects are not artificial. Locke partly saw this in the
case of substance. But the ideas of modes and relations
are also rational and correct, so far as they agree with
modes and relations properly inferred as belonging to
external substances. While, however, rational ideas of
the insensible are not artificial, they are often inade
quate ; e.g. of a corpuscle, of infinite space and time,
of gravitation, of the universe, of God. Lastly, the
inventive imagination makes artificial ideas-, such as
those of a centaur, a fairy, ' The Iliad,' ' A Midsummer
Night's Dream.' But it has never yet been successfully
analysed. Perhaps even the comparison and composi
tion of artistic imagination are founded on reasoning, not
to the actual and real, but to the possible and ideal.

Let us now suppose that Locke's general account of
the origin of ideas is immaculate and superior to our
objections, that sensation and reflection perceive simple
ideas ; that comparison, composition and abstraction
are the three acts which form compound ideas ; and
that the introduction of a supposition of substance was
a momentary lapse of a philosopher from the consis-

CHAP. VI.

LOCKE 1G5

tency of philosophy. What will be the consequence ?
As he says himself, ideas will be the instruments and
materials of our knowledge.1 Then, by parity of reason
ing, all that we can know from such materials will be
other ideas, and, as he has said himself, ideas will be
also all the objects of our understanding.2 Locke you
would imagine to be the founder of pure idealism. We
should have expected him to go on to show that every
thing in the world of science is an idea. At the end
we should have been inclined to say-
There are more things in heaven and earth, Horatio,
Than are dreamt of in your philosophy.

But Locke had a various, though not a logical mind.
He was a student of Descartes ; he was also imbued
with the English devotion to nature. From the former
source he derived the theory of ideas, from the latter
the reality of things. Locke, after assuming that all
objects of understanding are ideas, admitted that ex
ternal realities exist. The Essay contains an undercur
rent of ontology, which comes up first in the famous
distinction of primary and secondary qualities,3 recog
nising external qualities as real, as external causes of
our ideas of sensation, and even as externally related
as cause and effect to each other :—

' The qualities then that are in bodies, rightly con
sidered, are of three sorts.

' First, the bulk, figure, number, situation, and
motion or rest of their solid parts ; those are in them,
whether we perceive them or no ; and when they are'
of that size, that we can discover them, we have by
these an idea of the thing, as it is in itself; as is plain
in artificial things : these I call primary qualities.

1 Essay, II. 33, 19. 2 I. 1, 8. 3 II. 8.

166 PSYCHOLOGICAL IDEALISM PART ir.

' Secondly', the power that is in any body, by reason
of its insensible primary qualities, to operate after a
peculiar manner on any of our senses, and thereby pro-
7 duce in us the different ideas of several colours, sounds,
smells, tastes, &c. These are usually called sensible
qualities.

' Thirdly, the power that is in any body, by reason
of the particular constitution of its primary qualities, to
make such a change in the bulk, figure, texture, and
motion of another body, as to make it operate on our
senses differently from what it did before. Thus the
sun has a power to make wax white, and fire to make
* lead fluid. These are usually called powers.' *

The same undercurrent of ontology reappears in the
admission of substances, and real essences, though
unknown. It becomes most marked in the Fourth
Book, where Locke adds to all his other entities, one's
own existence, the existence of God, and the existence
of other things, such as the clippings of our beards and
the parings of our nails. Finally it springs up into an
.elaborate picture of the insensible universe beyond the
reach of our ideas.2 It is a dangerous thing to be
an unconscious metaphysician. Locke's metaphysical
theory of existence is quite outside his psychological
theory of ideas. How does it agree with his logical
theory of knowledge ? If it be true to say, that beyond
ideas there is an external world of qualities, real and
causal, real substances and real essences, my own ex
istence, God's existence, the existence of bodies, and of
insensible corpuscles, what is truly said by a philoso
pher, who is after all but human, must be known to a
man. What then does Locke, the philosopher who says
all this, say about the knowledge of man ?

1 Essay, II. 8, 23. 2 IV. 3, 24.

CHAP. VI.

LOCKE 1G7

The Fourth Book, which is on knowledge and
opinion, starts with a theory quite consistent with the
previous Books, on the origin of ideas :—

4 Since the mind, in all its thoughts and reasonings,
hath no other immediate object than its own ideas
which it alone does or can contemplate, it is evident
that our knowledge is only conversant about 'em.

' Knowledge then seems to me to be nothing but the
perception of the connection and agreement, or disagree-*-
ment and repugnancy of any of our ideas. In this alone
it consists.' *

Locke proceeds to divide knowledge into intuition
and reasoning. He says that < sometimes the mind per
ceives the agreement or disagreement of two ideas
immediately by themselves without the intervention of
any others: and this, I think, we may call intuitive^
knowledge:* He adds that 'when the mind cannot so
bring its ideas together, as by their immediate com
parison, and as it were juxtaposition, or application
one to another, to perceive their agreement or disagree
ment, it is fain, by the intervention of other ideas (one
or more, as it happens), to discover the agreement or
disagreement, which it searches : and this is what we
call Reasoning! 3 Afterwards, he writes a whole chapter 4
on Reason, in which he again defines it as the percep- ^
tion of the agreement or disagreement of ideas by
intermediate ideas. At the same time he rejects the
syllogism, although the process which combines two
extremes by the intervention of a middle is clearly the
same process as his own. But the main point to be
observed is that, according to him, reasoning begins
with an intuitive perception of the relation of ideas
i IV. i, 1-2. 2 IV. 2, i.

3 iv. 2, 2. 4 IV. 17.

168 PSYCHOLOGICAL IDEALISM PART n.

and ends with a mediate perception of the relations of
ideas.

A theory of reasoning such as this must con
fine all reasoned knowledge, and therefore all science,
to relations of ideas. This actually is his view of
mathematics and morals. ' I doubt not,' he says, ' but
it will be easily granted, that the knowledge we have of
mathematical truths is not only certain, but real know
ledge ; and not the bare empty vision of vain insignifi
cant chimeras of the brain : and yet, if we will consider,
we shall find that it is only of our own ideas' l He
says the same of moral knowledge, which he also holds
to be as certain as mathematics.2 I admit that if all
objects of reasoning are ideas, mathematical knowledge
is only of our ideas. But, in this case, it is not of the
bulk, figure, number, structure and motion of bodies
and particles, which Locke himself recognises beyond
our ideas. Sir Isaac Newton, then, must have been
wrong in saying that all the particles of matter gravi
tate to one another with a force varying inversely to
the square of the distance ; for he was pretending to a
mathematical knowledge of the motions of particles
beyond ideas. What a curious contretemps, that in
1687 Newton should discover to mankind the Mathe
matical Principles of Natural Philosophy in every
particle of matter, and in 1690 Locke should publish
an Essay concerning Human Understanding to prove
that the knowledge of mathematical truths is only of
our own ideas !

We are relieved from further criticism of this pure
^dealism, however logical, because Locke himself deserts
it for realism, however hypothetical. At first he delibe
rately confines all knowledge to the perception of the re-

1 Essay, IV. 4, 0. 2

CHAP. VI.

LOCKE 1G9

lations of ideas, and throughout applies tliese limits strictly
to mathematics and morals. But all of a sudden he in
troduces us to a knowledge of real things in other de
partments of knowledge, and, as it were, writes a second
essay on another human understanding. The manner in
which he makes this abrupt transition is highly instruc-1^
tive. Having defined knowledge to be only conversant
about ideas, and to be nothing but the perception of
the agreement and disagreement of ideas,1 he reduces
these agreements and disagreements of ideas to four
sorts — identity or diversity, relation, coexistence, and
real existence.2 The knowledge of the first three sorts *
proceeds consistently enough, when suddenly, without
any previous preparation, much less argument, he lays
down the following dogma :-—

' Fourthly. The fourth and last sort is that of actual
real existence agreeing to any idea? 3

On, his original hypothesis that ideas are all the
objects of understanding, on his theory of the origin of
ideas in the Second Book, on his definition of know
ledge in the very same chapter of the Fourth Book, he
ought to have said, the knowledge of the idea of actual
real existence agreeing to any idea. But just as Des
cartes passed from the idea of God's existence to His
existence, so Locke passed from the knowledge of the^
idea of existence to the knowledge of existence agreeing
to any idea. But while Descartes had been inconse
quent, Locke to inconsequence added inconsistency ; he
had begun by saying that all objects of understanding
are ideas ; he afterwards admitted a knowledge of ' exist-*
ence agreeing to any idea.'

He afterwards divides this knowledge of existence
into three departments — an intuitive knowledge of our

1 IV. 1, 1-2. 2 IV. 1, 3. 3 IV. 1, 7.

170 PSYCHOLOGICAL IDEALISM PART IT.

own existence, a demonstrative knowledge of the exist-
"ence of a God, a sensitive knowledge of objects present to
the senses,1 — and devotes a chapter to each.2 In Locke's
philosophy, all three ought to have been knowledges
of ideas ; they are knowledges of the real and actual
existence of things. Again, the Fourth Book presents
us with two theories of a proposition to support this
inconsistency. First, he divides propositions into two
kinds ; mental, wherein ideas, and verbal, wherein words,
the signs of our ideas, are put together.3 Afterwards, he
says that there are two sorts of propositions; one, con
cerning the existence of anything answerable to an idea,
rand the other, concerning the agreement or disagree
ment of our abstract ideas.4 I am not referring to all
these places to criticise Locke for inconsistency, which
is a weakness of human nature, a weakness even of
philosophers, who are but men, and an amiable weak
ness, because one of two contradictories must be true.
My object is rather to show that Locke at last came to
the truth, that not all objects of knowledge, of proposi
tions, of understanding are ideas. But there is a further
question, How do we know these actual existences ? or, to
use Locke's own phrase, ' How shall the mind, when it
perceives nothing but its own ideas, know that they
agree with things themselves ? ' 5

The intuitive knowledge of our own existence is
settled in a single section, short but significant, in which
he gives up his original theory that we perceive nothing
but ideas :—

' As for our own existence, we perceive it so plainly,

^ and so certainly, that it neither needs, nor is capable of

any proof. For nothing can be more evident to us than

1 Essay, IV. 3, 21 ; IV. 9, 2. 2 IV. 9-11. 3 IV. 5, 5.

4 IV. 11, 18. 5 IV. 4, 3.

CHAP. VI.

LOCKE 171

our own existence. I think, I reason, I feel pleasure or
pain : can any of these be more evident to me than my
own existence ? If I doubt of all other things, that
very doubt makes me perceive my own existence, and
will not suffer me to doubt of that. For if I know I
feel pain, it is evident I have as certain perception of
my own existence as of the existence of the pain I feel :
or if I know I doubt, I have as certain perception of the
existence of the thing doubting as of that thought which
I call doubt. Experience then convinces us that we
have an intuitive knowledge of our own existence, and an
internal infallible perception that we are. In every act
of sensation, reasoning or thinking, we are conscious to
ourselves of our own being ; and, in this matter, come
not short of the highest degree of certainty! l

This passage breathes the very spirit of Descartes.
Cogito, ergo sum. I am conscious that I am a thinking
subject. This is the fact that never ought to have been *
deserted. Descartes deserted it for an inference of
substance, and Locke followed him out of the right path,
but he had to come back to it after all. Consciousness
reveals to me not thoughts but a thinker. This con
sciousness is indeed inconsistent with the previous state
ments of Locke ; first, that reflection perceives the ideas
of operations, whch is two removes from consciousness ;
secondly, that there is a supposition of a substance as
unknown substratum to those operations, which would
be a baseless inference from data containing nothing but
ideas of operations. Nevertheless, the direct conscious
ness of our existence is the fact. How then is it that it
is constantly disappearing out of philosophy, not only
in the seventeenth, but also in the succeeding centuries ?
Because philosophers are perpetually confusing abstract

1 IV. 9, 3.

172 PSYCHOLOGICAL IDEALISM PART n.

and concrete, forgetting that thoughts are abstract
qualities but the thinker the real being, and thus
concluding that consciousness and reflection reveal
thoughts, leaving the subject to inference and supposi
tion, when really consciousness and reflection tell me
that I am a thinking subject, from which I infer other
thinking subjects.

Unfortunately, in another part of the Essay, Locke
had exaggerated the truth, I am conscious that I am a
person, a thinking intelligent being, into the falsity, I
am that very consciousness.1 But in the first place,
I am conscious that I perform numerous operations
besides the operation of being conscious, that I am a
sensible, remembering, reasoning, desiring, willing sub
ject ; consciousness therefore itself tells me that I am
more than itself. Secondly, it is not my only source of
information about myself. I am conscious that I am
partly body thinking, but I also indirectly observe my
body. I reason from my consciousness and observa
tions, and infer that I am a permanent substance,
when I am asleep as well as when I am awake, when I
am conscious and when I am unconscious. Thirdly,
consciousness is interrupted ; if I were consciousness I
should have an intermittent existence. Finally, Locke
has confused the causa cognoscendi with the causa essendi.
Consciousness is necessary to tell me, I am a person ;
but it does not make me a person ; this am I made by
being a permanent substance, partly body and partly
soul, capable, when awake, of reasoning, and therefore
of the status of a rational being.

At the end of Butler's ' Analogy,' the Dissertation on
Personal Identity contains an excellent statement of its
relation to consciousness, as follows : —

1 Essay, II. 27, 9 seq.

CHAP. VI.

LOCKE 173

' But though consciousness of what is past does thus
ascertain our personal identity to ourselves, yet to say
that it makes personal identity, or is necessary to our
being the same persons, is to say that a person has not
existed a single moment, nor done one action, but what
he can remember, indeed none but what he reflects
upon. And one should really think it self-evident that
consciousness of personal identity presupposes, and
therefore cannot constitute, personal identity ; any more
than knowledge, in any other case, can constitute truth,
which it presupposes.'

Locke, then, is right in saying that consciousness
is an intuitive knowledge of oneself, wrong in saying ^
that it is oneself. Not from the false identification of
self and consciousness, but from the consciousness of
self, that is, from the intuitive knowledge we have of our
own existence, as cogitative beings, Locke deduces our
knowledge of the existence of a God by an argument,
which is an extension of the third argument in
the ' Principia Philosophise ' of Descartes.1 A finite
thinking subject requires an infinite thinking subject
to create it. Yes, but this argument holds only
if we are conscious of ourselves as thinking subjects.
God is not an idea, and consequently cannot be inferred
from mere ideas. Thus, if Locke had clung to his
ideas of reflection, he could not have proved a God :
the consciousness, not of mere thoughts, but of a think
ing subject, is necessary to natural theology. Simi
larly, it is necessary to infer any other thinking subject
but myself. If, then, I were conscious only of ideas of
operations, and even if I were conscious directly of
operations, I could not infer thinking subjects, and I
could not infer God. The object of consciousness, there-

1 IV. 10 ; cf. Descartes, Princ. i. 20.

174 PSYCHOLOGICAL IDEALISM PART IT.

fore, is not op 3rations, still less their ideas, but thinking
subjects. Here again, too, we find that not all objects,
and not all data, of understanding are ideas. Locke
was obliged to surrender his theory of ideas in order to

/prove his own existence, the existence of others, the
existence of God.

Next, we come to what Locke calls our sensitive
knowledge of objects presented to our senses. Here,
with Cartesian inconsequence, he tried to maintain his
theory of ideas, and yet show how we know external
realities, or originals, by inference. In the Fourth
Book he returns to this point again and again. He
begins by proposing this problem. 'There can be
nothing more certain,' he says, ' than that the idea we
receive from an external object is in our minds. But
whether there be any thing more than barely that
idea in our minds, whether we can thence infer the
existence of any thing without us, which corresponds
to that idea, is that, whereof some men think there may
be a question made ; because men may have such ideas
in their minds, when no such thing exists, no such
objects affect their senses.' l He answers the question by
saying that a man is conscious of a different perception
when he looks on the sun by day and thinks on it by
night, and concludes that this is a knowledge not intuitive

^nor demonstrative, but sensitive. Again, he divides
the problem by simple ideas and complex ideas of sub
stance ; and argues that, in the first place, simple ideas,
which the mind can by no meairs make to itself, must
necessarily be the product of things operating on the mind
in a natural way, and that the idea of whiteness in the
mind answers that power which is in any body to produce
it there ; 2 and, in the second place, the reality of our

1 Essay, IV. 2, 14. - IV. 4, 4.

CHAP. VI.

LOCKE 175

knowledge of substances is founded on our complex
ideas»of them being such as are made up of such simple
ones as have been discovered to coexist in nature.1
Finally, he devotes to the knowledge of objects without
us a chapter,2 in which he contends that its certainty is
as great as we are capable of concerning the existence
of anything but oneself and God, and that it deserves
the name of knowledge. He adds four arguments to
the preceding : first, that those who want the organs of
sense want the ideas of that sense ; secondly, that some
times I cannot avoid the having those ideas produced,,
in my mind ; thirdly, that many of these ideas are pro
duced with pain, which afterwards we remember with
out offence ; fourthly, that our senses in many cases
bear witness to the truth of each other's report. Finally,
he falls back on the practical argument that we have at
all events a knowledge of the external world by the
happiness and misery we receive from it. The whole of
these arguments are summed up in this one : I have
ideas of sensation, which I do not produce myself ; I
infer that they are produced by external bodies. It is
the Cartesian argument from the passivity or involun-
tariness of sensations and ideas.

Locke's admission of the reality and knowledge of
external bodies is right and honest, but completely
destructive of his original hypothesis of the objects
and data of understanding. It is true, as he admits,
that we know external bodies. But this admission de
stroys his original doctrine that knowledge is always^
concerned with ideas. Again, it is true, as he admits,
that we know bodies by inference. But this destroys
his doctrine that reasoning begins and ends with ideas.
Both admissions also destroy his original doctrine that

1 IV. 4, 12. - IV. 11.

176 PSYCHOLOGICAL IDEALISM PART n.

* ideas are all the objects of understanding. It is also
true, as lie says, that we infer external bodies from the
passivity of sensation. But is it true that we could

'have drawn this inference from sensible data, if sensa
tion had been a perception of nothing but ideas ? This
is what Locke makes no attempt to prove. It is con
tradictory to his own logic. Like Descartes, he recog
nised that real truth is the agreement of our propo
sitions with external reality.1 But unlike Descartes, he
has given up any special criteria of truth. The vera
city of God he uses only for revelation ; 2 and regards
the inherent clearness and distinctness of ideas not as
positive criteria, but only as conditions of truth.3 The
consequence is that he has no organon except the
rules of reasoning ; and he is aware that, as the data of
reasoning are, according to him, ideas with their agree
ment and disagreement, so the conclusions are logically
confined to the agreement and disagreement of ideas.4
Yet he expects us also to believe that reasoning starting
with ideas of sensation can be logically extended to ex
ternal bodies.

All logic demands that, as are the data, so are the
conclusions. I find that some of the sensible objects
I perceive are passive. I have a right to infer some
other cause. But I must by parity of reasoning infer a
cause similar to those already known. Now, what data
does Locke supply me ? Granting him every advantage

V and all his inconsistencies, I should have ideas of sensa
tion and reflection, and compound ideas from the Second
Book : from the Fourth Book, I should have conscious-

* ness of myself, and a demonstration of other thinking
subjects, and of God. These, ex hypothesi, are all the data,

1 Essay, IV. 5. - IV. 16, 14.

3 IV. 2, 15. 4 IV. 17.

CHAP. vi. LOCKE 177

direct and indirect, at the very best. What would be
the logical inference ? I could infer, from the passivity
of sense, that it resulted either from other ideas, or from
other thinking subjects, or from God. I could not, being
without bodily data, infer that it resulted from external
bodies. Locke saw the importance of the passivity of
sense, but forgot the rules of logic.

Newton and Locke were contemporaries. What,
then, was Locke's attitude to natural philosophy ? He
recognised its discoveries, and especially the corpus
cular philosophy revived by Bacon, developed by Des
cartes, and brought to its perfection by Newton. We
have followed him in his ' little excursion into natural
philosophy ' to distinguish primary and secondary
qualities. He there admits the existence of corpuscles,
real qualities, primary and secondary, though insensible,
and real powers between qualities, e.g. the power of fire
to make lead fluid. There is no fault to find here
except with his definition of a quality as ' a power to
produce any idea in our mind.' 2 A quality is really a
characteristic of a subject or substance. It has various
powers, and among them the occasional power of affect
ing our senses. For instance, motion is a characteristic
of every corpuscle, and has a power of affecting every
other corpuscle, and sometimes of affecting human senses
in the way of sensible motion, light, heat, sound, and so
forth. But to define it by its sensible power, would be
to convert a very occasional accident into the essence of
motion, forgetting that there are myriads and millions
of motions which come nowhere near the earth, much
less man, and are not powers of producing any ideas in
his mind.

Locke defined quality by a separable accident.
1 II- 8, 22. 2 ii. s, 8.

N

178 PSYCHOLOGICAL IDEALISM TART n.

Hence also a mistake in his definition of a secondary
quality as a power of an insensible primary quality
to produce in our senses a sensible idea. This is only
an occasional and accidental power ; and a secondary
quality is a specific modification of a primary quality,
which exists whether it produces a sensible effect or
not. Thus heat is a mode of motion transferred from
star to star, and before the origin of animals much of
it was exhausted without having the power of produc
ing sensible heat With these corrections, Locke ex
presses the scientific distinction of primary and second
ary qualities in the universe. He fully recognises the
existence of that part of insensible nature, which I have
called the imperceptible, to distinguish it from the in
sensible but inferentially perceptible originals of sensible
objects. He recognises corpuscles as well as masses, the
particles of this paper as well as the paper.

But when he came to give these imperceptible cor
puscles a place in the human understanding, he began
to vacillate. In the Fourth Book, he distinguishes
knowledge and opinion, as respectively the perception
and the presumption of agreements and disagreements
of ideas, as certain and probable. Strictly, he could put
natural philosophy in neither, because he admitted that
it was not about ideas, but things. But the alternative
to which he leaned was to draw the line between know
ledge and opinion, exactly between the paper and its
particles, between the mass and the corpuscle, between
the perceptible and the imperceptible ; and, therefore, to
call the first inferences from sense knowledge, and the
subsequent inferences of science opinion. On the whole,
according to him, knowledge1 includes mathematics and
morals because they are about ideas, knowledge of self
1 Essay, IV. 3, 5 ; IV. 10, 6 ; IV. 11, 9 ; IV. 11, 13.

CHAP. VI.

LOCKE 179

because it is intuition, knowledge of God because it is
demonstration from this intuition, and knowledge of ex
ternal originals directly inferred from sense : here ends
knowledge. In opinion falls natural philosophy. Why?
Because the further from sense the less, he thought, our
knowledge. Because we are not capable of the ideas
of things so remote and minute, and this defect, as he
thought, keeps us in ignorance of the things. Because
we merely make experiments which are not science
Because we can only guess and probably conjecture,
use hypothesis and analogy. ' Analogy,' says he, ' in
these matters, is the only help we have, and 'tis from
that alone we draw all our grounds of probability. Thus
observing that the bare rubbing of two bodies violently
one upon another, produces heat and very often fire
itself, we have reason to think that what we call heat
and fire consists in a violent agitation of the imper
ceptible minute parts of the burning matter : observing
likewise that the different refractions of pellucid bodies
produce in our eyes the different appearances of several
colours ; and also that the different ranging and laying
the superficial parts of several bodies, as of velvet,
watered silk, &c., does the like, we think it probable
that the colour and shining of bodies, is in them nothing
but the different arrangement and refraction of their
minute and sensible parts.' 1 But knowledge of these
insensible qualities he denies. He doubts ' that how
far soever human industry may advance useful and
experimental philosophy in physical things, scientifical
will still be out of our reach ; ' 2 and he suspects that
'natural philosophy is not capable of being made a
science.' 3 Yet this very Locke winds up his Essay by

1 IV. 1G, 12 ; cf. IV. 3, 16 ; IV. 3, 24-end ; IV. 6 ; IV. 12, 9 13.
~ IV. 3, 26. 3 IV. 12, 10.

180 PSYCHOLOGICAL IDEALISM PART n.

a triple division of science, one of which is Physics
or Natural Philosophy, the knowledge of things.1

It is a matter of deep regret that Locke should have
written thus of natural philosophy in the very time of
Newton. There is some truth in what he says, but
marred by exaggeration. There is a huge abyss of
ignorance, but it is not altogether an incurable igno
rance. Much of what is called science is opinion, but
fresh evidences convert opinion into science. Because
there are probabilities in natural philosophy, it does not
follow that there is nothing certain. We cannot have
a perfect knowledge of nature, but we can know some
thing without knowing everything. We cannot always
discover real essence, but there is a knowledge of co
existences and causes, of the conservation and correlation
of physical forces, as in electricity and magnetism, with
out always knowing their essences. Locke rightly saw
that there is more of the universe unknown than known,
and much which is only opined ; but he lost sight of
the main fabric of science. By the mere elimination of
chance such a concatenation of laws cannot but be true.

Locke was ignorant of the logic of science. The
two greatest men of science in his own country were
Bacon and Newton, of whom the former had shown
that there is an experimental science of nature, the
latter that natural science is capable of physical deduc
tions from mathematical principles. But Locke, like
Hobbes, was silent about Baconian induction, and
oblivious to everything except the old method of intui
tion and demonstration, which suits mathematics, but
not the whole of natural philosophy. Everything out
side demonstration, he calls hypothesis and analogy.
He did not recognise the variety of method, the ana-

1 Essay, IV. 21.

CHAP. VI.

LOCKE 181

logia demonstrationum pro natura subjecti, desiderated
by Bacon. He did not see that the corpuscular
philosophy is made independent of this hypothesis and
that analogy, by many different evidences in many
different departments— gravity, light, heat, sound, elec
tricity, magnetism, chemical attraction, nervous and
muscular motion — all of which point to corpuscles, their
motions, according to Newton's laws, their modifications
constituting secondary qualities, their convertibility
and indestructibility as motion. He did not recognise
that there is a circumstantial evidence, which in law is
sufficient to hang a man, in nature sufficient to prove a
fact; and an approximate certainty, by accumulation
of evidence, ever indefinitely approaching absolute
necessity.

But his greatest, though characteristic, blunder was
his attempt to carry inference beyond sense to the ex-
ternal original inferentially perceptible and then stop
short ; to allow us to know the paper and not the par
ticle, the mass and not the molecule. Such a logic
is arbitrary. If insensible modes of primary qualities
are truly said to be, as Locke allows, then they are
knowable. The same laAVS of reasoning which enable
us to infer from sensible effects an external cause, en
able us from that cause to infer another cause, and so on
till we have completely explained facts of sense by laws
of science. If it were not so, how could science correct
ordinary knowledge? Ordinary knowledge infers an
external object, like in secondary as well as primary
qualities. Science declares that the external world is
like in primary but not in secondary qualities to the
sensible effect. But if the former is knowledge and the
latter opinion, by the first principle of method ordinary
knowledge, as more certain, is to be preferred to the less

182 PSYCHOLOGICAL IDEALISM PART rr.

certain science ; so that the ordinary man is right in
his theory of external light and heat, and the natural
philosopher wrong ! Locke's line between knowledge and
science gives the ordinary man, with his inference of
bodies, knowledge, but the scientific man, with his
inference of corpuscles, opinion. He elevates ordinary
above scientific knowledge, which is absurd.

There is a standing difference between natural and
mental philosophy, and Locke has done much to pro
duce it. He would make theology and morals not only
sciences, which they are, but more scientific than natural
philosophy, and tells man, whose real function is to
know all and do all, that his proper business is his moral
duties and his future state.1 Newton had just written
the ' Optics ' and the ' Principia,' but Locke's theory of
science would reduce these works to mere opinions.
The whole history of science is against him. On the
foundation of Newton's mechanics of motion has been
gradually reared a system of science which has eventually
revealed to us the insensible and imperceptible causes
of our sensations in the external world. On the
other side stand the mental philosophers, philosopTiantes
secundum sensum, considering primarily their sensations
and ideas, and with difficulty extending their thoughts
even to the external originals, then gazing stupidly at the
perceptible world, and never dreaming that they have to
explain the knowledge of imperceptible nature. Locke's
hypothesis that we have a sensitive knowledge from
ideas of objects presented to sense, a mathematical and
moral science of ideas, and an uncertain opinion of the
physical universe, undervalues natural philosophy. It
immediately produced the false attitude of Berkeley and
Hume towards nature, but it has affected the whole

1 Essay, IV. 12, 11.

LOCKE 183

course of mental philosophy, which has unduly neglected
the problem of knowledge, presented to it by natural
philosophers. Hence, while natural philosophy has
shown that the insensible is the causa essendi of the sen
sible, mental philosophy has never yet shown how the
sensible is the causa cognoscendi of the insensible.

But let us suppose that the whole fabric of science
is opinion, the whole imperceptible world unknown.
Yet it is at least an object of understanding and reason
ing, because, as Locke himself says, ' not but that it is
the nature of the understanding constantly to close with
the more probable side,'1 and, as he admits, reason may
end either in certainty or in probability, either by
demonstration or an argumentum ad judicium.2 This
beino- so, imperceptible probabilities are objects of
understanding and reason, but are not all ideas ; there
fore not all objects of understanding and reason are

ideas.

Nor could they be reasoned from ideas as their data.
This want of consequence brings us to another defect in
Locke's theory of primary and secondary qualities ; his
false view of their sensible aspect. In his opinion, as
external they are powers, as sensible they are ideas.
But they are neither mere powers nor mere ideas.
If, as sensible, they were ideas, we could not logi
cally infer insensible primary qualities, which are ad
mitted not to be ideas, yet inferrible. Therefore, even
as sensible, primary and secondary qualities are not
ideas, but physical qualities in sense, from which to infer
physical qualities beyond. So universally, the inference
of imperceptible corpuscles with real qualities and
powers beyond sense, even if only probable, could not
be drawn from mere ideas of sensation. The natural
1 IV. 20, 12. 2 IV. IT.

184 PSYCHOLOGICAL IDEALISM PART IT.

philosophy of the physical world, whether it be know
ledge or opinion, demands physical data of sense.

How came Locke, having said that ' whatsoever the
mind perceives in itself or is the immediate object
of perception, thought, or understanding, that I call
idea, l immediately to conclude real primary qualities of
matter ? Through the Cartesian habit of surreptitiously
passing from the idea to the thing, and his own supposi
tion of a bastard sensation of the thing. His one argu
ment for the reality of primary qualities is that they are
' such as sense constantly finds in every particle of matter,
which has bulk enough to be perceived.' 2 But according
to him sense is of ideas. All then he could consistently
say would have been that primary qualities are such as
sense constantly finds in the idea of every particle of
the idea of matter, which has the idea of bulk enough
to be perceived. But this consistency with his hypothesis
would not have proved the fact of the material reality
of primary qualities beyond ideas. At the same time,
his lapse into a direct sense of matter is of interest,
because it is a distinct anticipation of intuitive realism.
It exhibits the constant tendency of the philosopher to
relapse into the ordinary man, and to fancy he directly
perceives the external thing, or, using the inaccurate ter
minology of modern psychology, after contending that
what we are conscious of is subjective affections, to sup
pose a consciousness of objective existence. As Locke
tried to bridge over the gulf from ideas of sensation to
qualities by a kind of bastard sensation of qualities,
so his modern followers try to bridge the gulf from
subjective affections of consciousness to objective exist
ence by an undefinable consciousness of objective exist
ence. But it is certain that sensation perceives not the

1 Essay, II. 8, 8. 2 II. 8, 9.

CHAP. vi. LOCKE 185

external thing, but its internal effect ; and the only way
in which we can reach external qualities of things is not
by sense but by inference from adequate internal data,
which cannot be mere ideas, nor any psychical states of
subjective consciousness.

The ' Essay concerning Human Understanding '
begins by assuming that all objects of understanding,*,
as well as all data of sense, are ideas : it ends by ad
mitting that things beyond ideas are objects of under-^
standing, reasoning, science. The end is better than
the beginning, though the conclusion does not follow
from the premises. External bodies are properly in
ferred by ordinary men, as Locke admitted ; and
imperceptible corpuscles and their qualities are known,
with more certainty than he admitted, by men of science.
Therefore, in the first place, not all objects of under- ^
standing, reasoning, science, are ideas. Secondly, the
data of sense are neither ideas of sensation nor externals
qualities of matter, but internal effects on the nervous
system, sensibly qualified as extended, moving, hot,
coloured, and by other primary and secondary qualities.
From internal, ordinary knowledge infers external, sub- t
stances. From these again science, correcting ordinary
knowledge, infers imperceptible corpuscles, qualities
primary and secondary as the modifications of primary
powers exerted between those corpuscles, and powers
of affecting our senses. Locke's ' Essay ' throughout, to
make it thoroughly correct, consistent, and consequent,
would need two fundamental alterations :—

1. Some objects of understanding are physical things. *•

2. Some data of sensation are physical effects

the nervous system.

186 PSYCHOLOGICAL IDEALISM PART n.

CHAPTER VII.

BERKELEY.

THE two philosophers hitherto discussed assumed
hypotheses, but admitted facts, and tried to explain
them. Descartes assumed that ideas are the data of
sense, but admitted the knowledge of physical objects,
and broke down on the inconsequence of reasoning
from psychical data in the premises to physical objects
in the conclusion. Locke made the same assumption,
the same admission, and the same failure. But he went
further into hypothesis, and to inconsequence added
inconsistency. He assumed that ideas are not only
all the data but also all the objects of understand
ing, and then admitted that physical objects are also
objects of understanding. The admission is true, and
therefore, while it contradicted, also destroyed the
double hypothesis. We now come to a philosopher
who, accepting the whole ideal hypothesis, consist
ently denied facts. Berkeley assumed, with Descartes,
that ideas are the data, and with Locke, that they are
the objects, of human knowledge, and consistently, but
falsely, deduced man's ignorance of a physical world.

The 'Principles of Human Knowledge,' after an
Introduction on Abstract Ideas, begin in the following
manner :—

' It is evident to any one who takes a survey of the
objects of human knowledge, that they are either ideas

CHAP. vii. BERKELEY 187

actually imprinted on the senses, or else such as are
perceived by attending to the passions and operations
of the mind, or lastly, ideas formed by help of memory
and imagination, either compounding, dividing, or
barely representing those originally perceived in the
aforesaid ways. By sight I have the ideas of light and
colours with their several degrees and variations. By
touch I perceive, for example, hard and soft, heat and
cold, motion and resistance, and of all these more or
less either as to quantity or degree. Smelling furnishes
me with odours, the palate with tastes, and hearing
conveys sounds to the mind in all their variety of tone
and composition. And as several of these are observed
to accompany each other, they come to be marked by
one name, and so to be reputed as one thing. Thus,
for example, a certain colour, taste, smell, figure and
consistence having been observed to go together, are
accounted one distinct thing, signified by the name
" apple." Other collections of ideas constitute a stone, a
tree, a book, and the like sensible things ; which, as
they are pleasing or disagreeable, excite the passions of
love, hatred, joy, grief, and so forth.' l

Here are most of the errors in the Second and Third
Books of Locke's Essay accepted as principles. With
out proof, ideas alone are supposed to be perceived ;
ideas of qualities without a qualified subject, and
ideas of operations without a thinking subject. Beason-
ing from the data of sense to their causes is entirely
postponed in favour of representing, compounding and
dividing ideas. Ideas, simple or complex, are consist
ently declared to be all the objects of human knowledge.
But these so-called principles are mere hypotheses.
There is not one word of proof that either the data or

1 Princ. i.

188 PSYCHOLOGICAL IDEALISM PART n.

the objects of human knowledge are ideas. Locke, not
human nature — and not even the whole of Locke — was
the oracle of Berkeley.

Berkeley, however, being a less various but a more
logical thinker than Locke, was truer to the data of his
predecessor. Locke, as we found, having assigned
comparison, composition and abstraction as the three
acts, which form new ideas from sense, suddenly, and
without any justification, introduced a fourth act of
supposition, which is a kind of reasoning, to account
for our idea of substance. Berkeley avoided the after
thought, and, at the same time, the truth, that reason
does intervene in the formation of ideas from sense.
Adhering to Locke's first thoughts, he perceived that
what his predecessor had allowed about other complex
ideas equally applied to complex ideas of substances.
If we start from ideas of sensation, such as those of
colour, taste, smell, figure and consistence, and merely
compound these ideas, we can construct a collection of
ideas and account it one distinct thing, called an apple ;
but we cannot, without introducing a qualified physical
substance into sense, and restoring its privileges to
reason, either perceive or infer an external physical
substance. Berkeley thus reduces Locke to logic ; nor
has mental philosophy ever recovered this purely hypo
thetical theory of substance.

Berkeley also made an important correction in one
of Locke's three acts, abstraction. Locke had supposed
that we can form a perfectly abstract idea of a triangle,
which is ' neither equilateral, equicrural, nor scalenon,
but all and none of these at once.' l Berkeley devoted
the Introduction of the ' Principles ' to a criticism of this
modern conceptualism, and founded modern nominalism.

1 Essay, IV. 7, 9.

CHAP. Til.

BERKELEY 189

He denied that he could abstract or conceive separately
qualities which cannot exist separately, or form a
general notion in Locke's sense.1 He admitted that
he could consider a figure merely as triangular, without
attending to its particular qualities, but not form an
abstract general inconsistent idea of a triangle.2 Simi
larly, Hume afterwards said, that all general ideas are
nothing but particular ones annexed to general terms.3
The essential truth at the bottom of this theory is
that abstraction is only a kind of attention. But, as
often happens, one extreme view begets another. We
cannot rise to a purely abstract idea, nor need we fall
to a purely particular idea ; we cannot form an idea of
triangle in general, nor need we think of a single
triangle. We can frame a general idea of a miscella
neous assemblage of similar individuals.4 Secondly,
the point about classes is, not what we conceive, but
what we infer and know. But, while correcting
Locke's exaggeration of abstraction, Berkeley left its
independence of reasoning. The consequence is that,
according to him, the limit of generalisation would be
some single simple idea or some single collection of
simple ideas of sense viewed generally. This narrow
ness pervades his whole philosophy. There is, indeed,
such a simple abstraction of ideas from sense, as we ad
mitted in the last chapter. But reason, at the same
time, starts from sense and first infers classes of in
sensible objects, and then constructs general ideas of
them in the rational imagination. Finally, this rational
imagination of general ideas accompanies a rational
abstraction ; like direct abstraction, attention, but atten
tion to objects of reason. We can abstract, in the sense

1 P-rinc. Introduction, x. 2 Id. xvi.

3 Treatise, ii. § 7. 4 Cf. Mill, Logic, iv. 2, 1.

190 PSYCHOLOGICAL IDEALISM PART n.

of attending to, an insensible object, not apart from the
qualities which belong to it, but apart from the quality
of being sensible, which does not belong to it. The idea
of an object will indeed contain some sensible qualities,
and usually some visible colour. But having inferred that
the invisible object is coloured only in the sense of reflect
ing sethereal undulations, by abstraction I consider the
object as so qualified, without attending to it as visibly
coloured. In short, I know by scientific reasoning that
objects exist apart from merely sensible qualities, and
I can attend separately to their existing apart. Berke
ley fell into the error of postponing inference about
classes, and therefore of limiting abstraction to direct
formation of ideas from sense. Really, there are objects
known by sense, and objects known from sense by
reason ; and there is an abstraction from sense, and an
abstraction from reason, though in both cases the ab
straction is but attention to sensible and rational objects
of knowledge.

According to Berkeley, then, starting from the
Second and Third Books of Locke's Essay, all the objects
of human knowledge are ideas of sensation and reflec
tion, and the collections of ideas made out of them by
memory and imagination, to which lie reduced abstrac
tion of ideas, and without reasoning about causes. But
it is impossible for errors to remain perfectly logical.
Though he had just said that all objects known to us are
ideas, he proceeds, like Locke, dogmatically to assert
that a thinking subject exists :—

6 But besides all the endless variety of ideas or objects
of knowledge, there is likewise something which knows
or perceives them, and exercises divers operations, as
willing, imagining, remembering about them. This
perceiving active being is what I call mind, spirit, soul,

CHAP. VII.

BERKELEY 191

or myself. By which words I do not denote any one of
my ideas, but a thing entirely distinct from them,
wherein they exist, or which is the same thing, whereby
they are perceived ; for the existence of an idea con
sists in being perceived.' 1

Berkeley was dogmatic, but right, in asserting the
existence of himself ; but he was wrong in calling this
thinking subject a thing entirely distinct from his ideas,
and in supplying no data for his knowledge of it.
I am a thinker, from whom the subject and the
thoughts are opposite abstractions. But, in spite of his
criticism of abstract ideas, Berkeley had already fol
lowed Locke's Second Book in supposing all the objects
of reflection to be mere ideas of operations. The ques
tion then arises, how he could possibly know that he
was also a thinking subject. Locke had said that the
thinking subject is a matter of mere supposition.
Berkeley went a stage further : he said that ' it cannot
be of itself perceived, but only by the effects which it
produceth.'2 But there are several difficulties in al
lowing him to take this view on his hypotheses.
In the first place, if it is true, there is something which
is known, though indirectly, without being an idea ;
therefore, not all objects of understanding, but only all
objects of sense, will be ideas. Secondly, if all the objects
of sensation and reflection were ideas of sensible qualities
and ideas of operations, as he supposes, the whole of
these data would contain no subject, not of course a
physical nor even a psychical subject, and nothing like
a subject, for a subject is, as Berkeley admits, not an
idea ; therefore, no subject, even no psychical subject,
could be logically inferred. We must choose, therefore,
between the original data and the illogical conclusion.

1 Princ. ii. ~ Id. xxvii.

192 PSYCHOLOGICAL IDEALISM TARTU.

But Berkeley was riglit in admitting the existence and
knowledge of a thinking subject. Therefore, the data
of sensation and reflection cannot be mere ideas. Even
if not sensation, at least reflection must be perception
of myself as a thinking subject, from which I infer
other thinking subjects, and God Himself.

Berkeley ought to have returned to Descartes, and
-sjiegun with the consciousness, ' I think.' But, although
he saw that we cannot abstract what cannot exist sepa
rately, he was so enthralled by Locke that he began by
supposing that we perceive ideas of qualities and ideas
of operations, when we cannot even abstract these ob
jects except in the sense of attending to them in their
subjects. The idea of colour and the idea of willing
are as much abstractions as the idea of a triangle. We
really perceive, by sensation, at least, the coloured, and
by consciousness, at least, the willing. But Berkeley,
like Locke, began all sense with abstract ideas of
qualities and operations. Though, unlike Locke, he saw
that he could derive no physical subject from the
former, he illogically thought he could derive a thinking
subject from the latter ideas, although, like Locke, he
had no data for a logical sequence from the conscious
ideas of operations to the thinking subject.

Curiously enough, he ended, like Locke, in after
all returning to Descartes, and in admitting, ' I know
or am conscious of my own being.' 1 This admis
sion that I am conscious of myself is quite incon
sistent with the original hypothesis that I perceive ideas
of operations directly, and the subsequent corollary
that I perceive myself only indirectly by my effects.
Nevertheless, the admission is true, and the hypothesis
and its corollary false. I cannot infer a thinking sub-

1 Hylas and PJiilonous, Third Dialogue.

CHAP. vii. BERKELEY 193

ject from mere operations. I am not conscious of
operations, still less of ideas of operations — an abstrac
tion, two removes from the truth. I am conscious of
myself, as thinking subject.

But Berkeley involved his admission of a thinking
subject with another hypothesis. He accepted the
Cartesian transition from self to soul without a word of
proof.1 As I have already shown, I am not conscious of
this identification, I am conscious of the very reverse.
The combined evidence of consciousness, observation,
and reasoning teaches me that I am a man thinking
partly by my body and partly by my soul. But, you
will say, Berkeley was a theologian, who, knowing that
God is a spirit, rightly inferred that man is a spirit.
The answer is that man is not God. It is true that
there is a resemblance, but there is also a difference.
When I infer that there are other men, I observe, by
direct inference from sense, two sorts of signs, bodily
organs and physical works, from both of which I infer
a man like myself, body and soul. But God only offers
me one of these signs, His works of nature, but no signs
of a body. Hence I have a right to infer that He is
similar to myself, so far as He by intelligence and will
produces works of order, beauty, and goodness, similar
to those of man, but I have no right to infer either that
He, like man, is also a body, or that man, like Him, is a
pure spirit. Nor have I a right to infer that-
All are but parts of one stupendous whole,
Whose body nature is, and God the soul.

Nature is to God as works are to man ; and as a man's
body is not his works, so neither is nature the body of
God. ' Hie omnia regit,' says Newton2 about the Deity,

1 Princ. ii.

2 Newton, Principia, Lib. III. Scholium Generate (sub fin.).

O

194 PSYCHOLOGICAL IDEALISM TART n.

' non ut anima mundi, sed ut universorum dominus.'
God lias no body ; for how could He have a body pro
portionate to His infinite intelligence and will, and show
it not ? God, then, is a spirit ; man is not.

Now, it is true that God, for a time, gave a bodily
sign, when He took upon Himself a body and made
Himself man. But the incarnation of Christ is a very
proof of the difference between God and man. Christ
ceased to be a pure Spirit, became flesh, and dwelt
among us. Berkeley cannot explain this union of the
Divine and the human in Christ. God is a spirit ; but
if man is also a spirit, what is the incarnation ?
Berkeley's only logical answer would be the gratuitous
hypothesis that Christ took upon Himself certain ideas,
called the human body. But Christ had the ideas already
from eternity. What He wanted was the very body, re
presented by those ideas, for a time. There is nothing
for it, but that God is a spirit, and Christ took upon
himself a body and became man, and man is both body
and spirit in one. The idealistic hypothesis that I am
a spirit is inconsistent both with philosophy and with
Christianity. Yet in our own time a false philosophy
of man as a purely spiritual subject is supposed to be
a justification of Christian theology.

Berkeley, in the Introduction and the first two sec
tions of his ' Principles,' furnished himself with his pre
mises. They are anticipations of human nature, mainly
derived from Descartes and Locke, with an occasional
assumption of his own. Let it be granted, from Des
cartes, that the thinking subject, myself, is a mind,
spirit, soul. Let it be granted, from Locke's Second
Book, that not only all data, but all objects of know
ledge, are simple ideas of sensation and reflection,
and ideas compounded by memory and imagination,

CHAT. VII.

BERKELEY 195

without taking any notice of reasoning ; and let us
avoid Locke's inconsistency of supposing an external
physical substance beyond a collection of ideas, and his
error of purely abstract ideas. Let the premises, which
he owes to Descartes and Locke, be granted to Berke
ley, without his proving them. What follows ? Why,
the purely hypothetical, fairly logical, wholly synthetic
deduction from false and unproved hypotheses, known
as the Berkeleian philosophy. He who is foolish enough
on the mere authority of this doctor to swallow the
hypotheses, like pills, will find that the deductions will
purge him of all knowledge beyond spirit and ideas.

Berkeley begins his deductions by explaining the
existence of what he calls sensible things, and denying
that what he calls unthinking things exist except as
perceived : — ' The table I write on, I say, exists, that
is, I see and feel it ; and if I were out of my study I
should say it existed, meaning thereby that if I was in
my study I might perceive it, or that some other spirit
actually does perceive it. There was an odour, that is,
it was smelled ; there was a sound, that is to say, it was
heard ; a colour or figure, and it was perceived by sight
or touch. This is all that I can understand by these
and the like expressions. For as to what is said of the
absolute existence of unthinking things without any
relation to their being perceived, that seems perfectly
unintelligible. Their esse is per dpi ^ nor is it possible
that they should have any existence out of the minds
of thinking beings which perceive them.' l

So far as this argument follows from its premises it
is hypothetically unanswerable. The esse of ideas is
per dpi-, if, then, all objects of human knowledge are
ideas, their esse will be per dpi ; and again, an unthink-

1 Princ. iii.

o 2

196 PSYCHOLOGICAL IDEALISM TART n.

ing tiling, which is not an idea, will not be humanly
known to exist. Berkeley was entitled to these hypo
thetical conclusions. But his argument conceals a
further false hypothesis, namely, that what is unknown
by man to exist, being unintelligible to him, is non
existent ; from which he concluded that a purely un
thinking thing is not only unknown by man, but also
non-existent. Thus to hypotheses and hypothetical de
duction Berkeley added dogmatism. He dogmatically
asserted the existence of mind and the non-existence
of matter.

The importance of the deductions which immediately
follow consists in their entire omission of reasoning
from the data of sense to their causes, and its conse
quences, when combined with Locke's premises.
Houses, mountains, rivers, and, in a word, all sensible
objects, are supposed to have a separate existence.
Now, says Berkeley, they are what we perceive by sense,
and what we perceive are ideas or sensations ; therefore
they are ideas or sensations.1 He adds that it is only
the doctrine of abstract ideas which makes us dis
tinguish the existence of sensible objects from their
being perceived.2 But it is not true that a house is
a sensible object which we perceive by sense ; sense
perceives only a sensible effect of an external house,
which is inferred by reasoning, and can be distinguished
from the sensible effect by the attention of abstraction.
But it is true that if We choose to omit reasoning about
causes, and suppose that sense perceives ideas or sen
sations, the only house we should know would be, not
the house now inferred, but only what we should then
perceive, a mere collection of ideas or sensations, in
capable of being abstracted from being perceived.

1 Princ. iv. 2 Id. v.

CHAP. VIT.

BERKELEY 197

This strict though hypothetical logic from Locke's
Second Book removed Berkeley into another arena of
philosophy. Descartes and Locke had admitted the
existence and knowledge of an external world, not
merely psychical but also physical ; that a house is an
external object causing our ideas ; and, in accordance
with the representative theory, that perception presents
ideas but represents external objects. Berkeley, agree
ing both with Descartes and Locke in the perception of
ideas, but aware that neither philosopher supplied data
from which to infer an external object, and following
Locke in postponing reasoning about it, logically con
cludes that the external object and the sensible object
are one, and that in perceiving an idea or sensation, we
are perceiving not a sensible effect of an external house,
but the house itself. His pure idealism produced the
metaphysical theory that objects, supposed to be ex
ternal, are nothing but ideas or sensations in the mind,
and the psychological theory of a presentative percep
tion of ideas or sensations, representing nothing.

Having hypothetically deduced that the esse of all
objects known to man is per dpi, and that what are called
external objects are really ideas or sensations, Berkeley
proceeds to the conclusion that ' all the choir of heaven
and furniture of the earth ; in a word, all those bodies
which compose the mighty frame of the world,' exist
in my mind, or in that of some created spirit ; or else
subsist in the mind of some eternal spirit.1 This con
clusion also follows from the premises. If all objects
of knowledge are ideas, and ideas subsist in the mind of
some spirit, it follows necessarily that the whole known
world subsists in the mind of some spirit. So far,
indeed, as the human spirit goes, we could only speak

1 Princ. vi.

198 PSYCHOLOGICAL IDEALISM PART IT.

of tlie whole known world. We saw above that Berke
ley, while speaking even of man,1 denied the existence
of what was not an object of human knowledge. He
now corrects this defect by the addition of the eternal
spirit,2 to whom whatever exists is known, while what is
not known does not exist. Of the Divine spirit at least
Berkeley could say, whatever exists is an object of His
knowledge; if, then, all objects _of knowledge are ideas,
and ideas subsist in the mind of a spirit, whatever exists
subsists in the mind of the eternal spirit of God. Even
so, however, it might be objected that, if ideas are the
objects of human, it does not follow that they are the
only objects of Divine knowledge. But in Berkeley's
' Principles ' there is a perpetual equivoque between the
sensible ideas of man and the intellectual ideas of God.
' From what has been said it follows that there is
not any other substance than spirit ; ' this is the next
hypothetical consequence.3 It is an immediate corol
lary. If there were only man, the only known substance
would be spirit, but add God and it would follow that
the only existing substance is spirit, so that there remains
no unthinking substance.4 Berkeley further proceeds to
deduce this denial of matter from the hypothesis of
ideas. He is perfectly logical. Ideas cannot exist in
an unthinking substance ; if then sensible qualities were
ideas, there would be no unthinking substance or substra
tum of those ideas or qualities.5 Again, he warns us
against those who maintained that, though unthinking
substance is not the substratum of sensible ideas, ideas
are nevertheless the copies or resemblances of unthinking
substance. ' I answer,' he says, ' an idea can be like no
thing but an idea.' G This memorable sentence marks

1 Princ. iii. 2 Id. vi. 3 Id. vii. 4 Id.

5 Id. 6 Id. viii.

BERKELEY 199

the return of the logic of reasoning into mental philo
sophy. Berkeley at this point begins to think about
reasoning, though too late ; for he had already fixed the
objects of knowledge without it. But he thinks about
it as a logician, and gives the answer to the illogical
attempt of Descartes and Locke to first enclose man
within psychical ideas, and then, without any clue in the
data, expect him to discover physical objects. In the
case of physical substances, if the data of inference were
sensible qualities as ideas, we could infer a similar col
lection of qualities as ideas ; if they were qualities with
out beim? ideas, we could infer a similar combination of
qualities^; but in neither case could we infer a physical
substance, for which we should have no analogue in

sense.1

This rigorous logic from Locke's hypotheses of ideas
enabled Berkeley to destroy Locke's theories of material
substance and its primary qualities at a blow :-

6 Some there are who make a distinction betwixt
primary and secondary qualities : by the former, they
mean extension, figure, motion, rest, solidity or impene
trability, and number ; by the latter they denote all
other sensible qualities, as colours, sounds, tastes, and so
forth. The ideas we have of these they acknowledge not
to be the resemblances of anything existing without the
mind or unperceived ; but they will have our ideas of
the primary qualities to be patterns or images of things
which exist without the mind, in an unthinking sub
stance which they call matter. By matter, therefore, we
are to understand an inert, senseless substance, in which
extension, figure, and motion do actually subsist. But
it is evident, from what we have already shown, that
extension, figure and motion are only ideas existing in

1 Cf. Princ. xxxvii.

200 PSYCHOLOGICAL IDEALISM

PART II.

the mind, and that an idea can be like nothing but
another idea, and that consequently neither they nor
their archetypes can exist in an unperceiving substance.
Hence it is plain, that the very notion of what is called
matter or corporeal substance involves a contradiction
in it.' 1

Yes ; if, and only if, qualities as sensible are ideas,
an idea is like nothing but another idea, and therefore
we could infer no external qualities of matter ; neither
insensible primary qualities like primary qualities as
sensible, nor insensible secondary qualities as modifi
cations of primary qualities and causes of secondary
qualities as sensible. Now, matter is nothing without
qualities ; therefore, we could not infer matter at all.
The argument is quite logical, if we once admit with
Locke, that, as sensible, all qualities are ideas. If with
modern idealists we should substitute sensations, it
would equally follow that we could infer no insensible
qualities of matter, and therefore no matter at all.

Berkeley added a second argument to prove that all
qualities exist only as ideas in the mind and not in
matter and its particles : —

' They who assert that figure, motion and the rest
of the primary original qualities do exist without the
mind in unthinking substances, do at the same time
acknowledge that colours, sounds, heat, cold, and such
like secondary qualities, do not, which they tell us are
sensations existing in the mind alone, that depend on
and are occasioned by the different size, texture, and
motion of the minute particles of matter. This they
take for an undoubted truth which they can demonstrate
beyond all exception. Now if it be certain that those
original qualities are inseparably united with the other

1 Princ. ix.

CHAP. VII.

BERKELEY 201

sensible qualities, and not, even in thought, capable of
being abstracted from them, it plainly follows that they
exist only in the mind. But I desire any one to reflect
and try whether he can, by any abstraction of thought,
conceive the extension and motion of a body without all
other sensible qualities. For my own part, I see evi
dently that it is not in my power to frame an idea of a
body extended and moved, but I may withal give it
some colour or other sensible quality which is acknow
ledged to exist only in the mind. In short, extension,
figure, and motion, abstracted from all other qualities,
are inconceivable. Where, therefore, the other sensible
qualities are there must these be also, to wit, in the
mind, and nowhere else.' l

This argument does not touch Locke, so far as it
depends on the admission that secondary qualities are
mere sensations ; for Locke said that, as sensible, they
are ideas, and, as external, powers. But it touches
later theories of secondary qualities, realistic and
idealistic. It is true that if secondary qualities are
sensations, primary qualities, as sensible, will also be
sensations, from which no external quality, and there
fore no matter, could be inferred. Moreover, the argu
ment is interesting as another instance of Berkeley's re
duction of the external to the sensible. He saw that on
the conjoint hypothesis that sense perceives qualities as
sensations, with abstraction of ideas, but without
reasoning to causes, we should only be able to infer and
attend to qualities, primary and secondary, as they are
fused in sensation. Hence his followers invariably re
gard primary and secondary qualities merely as various
kinds of sensations, and not as external qualities.

By this series of hypothetical arguments Berkeley

1 Princ. x.

202 PSYCHOLOGICAL IDEALISM PART n.

arrived at tlie following conclusions : all subjects are
spirits and all objects ideas of spirits. This absolute
universality logically applies only to the eternal spirit.
As far as the human spirit goes, Berkeley's conclusions,
so far as they are logical, must be put in a more
moderate form. If there are spirits, and all objects of
knowledge are ideas, then all known subjects are spirits
and all known objects are ideas ; a physical subject of
qualities is not known to exist, and qualities, primary
and secondary, are known as ideas or sensations in our
minds, but are not known to be external qualities of
physical subjects, bodies and corpuscles, in an external
world. What, then, is to become of the minute particles
of matter, their latent sizes, textures, and motions ; to
say nothing of their priority, and their production of
our sensations ? What, again, are the causes of the ideas
or sensations in the mind of a human spirit ? Berkeley,
like Locke, at last found himself face to face with the
problem of reasoning to causes. Given ideas of spirits
as all the data and objects of knowledge, what causes
can reason infer ?

We might feel tempted now to say that Berkeley,
having the universe of Divine ideas, as it were, in his
grasp, would at once say that the external world of
bodies, their corpuscles, and their qualities, which the
natural philosopher has discovered to be the insensible
causes of sensible qualities, even asther and its motions,
are Divine ideas, by which the Deity produces the sen
sations of man. But Berkeley no more than the modern
Berkeleian resorts to this Hegelian alternative. He
precluded himself from taking it, both by his identifica
tion of the external with the sensible object, and by
his doctrine of the inactivity of ideas. As the former
deprived him of the external world as a distinct object,

CHAP. VII.

BERKELEY 203

so the latter prevented him from regarding insensible
causes as ideas. ' All our ideas,' says he, ' sensations,
or the things which we perceive, by whatsoever names
they may be distinguished, are visibly inactive ; there
is nothing of power or agency included in them, so
that one idea or object of thought cannot produce or
make any alteration in another.' l So far from resolving
insensible scientific causes into Divine ideas acting on us,
he uses the theory of the inactivity of ideas to deny in
sensible scientific causes. ' Whence, it plainly follows,'
lie concludes, ' that extension, figure, and motion cannot
be the cause of our sensations. To say, therefore, that
these are the effects of powers, resulting from the con
figuration, number, motion, and size of corpuscles, must
certainly be false.' 2

Berkeley, having decided that the cause is not the
qualities of corpuscles, proceeded to infer that it is the
spirit of God :—

' We perceive a continual succession of ideas, some
are anew excited, others are changed, or totally disap
pear. There is, therefore, some cause of these ideas
whereon they depend, and which produces and changes
them. That this cause cannot be any quality or idea,
or combination of ideas, is clear from the preceding
section. It must, therefore, be a substance ; but it has
been shown that there is no corporeal or material
substance. It remains, therefore, that the cause of
ideas is an incorporeal active substance or spirit.' a

Berkeley, like Descartes and Locke, saw that there is
an involuntariness in our sensations which requires
some cause. They might have all stopped there, and
said that the nature of the cause is unknown ; but
they were too philosophical to be agnostics. Descartes

1 Princ. xxv. 2 Id. 3 Id. xxvi.

204 PSYCHOLOGICAL IDEALISM PART n.

and Locke, however, were not logical enough to see
what cause could be inferred from their data; but
guided by real facts rather than by their theories
illogically supposed that, without anything physical in
the data, we could infer a physical cause. Berkeley,
on the other hand, was the first of the psychological
idealists to see that the data and objects of knowledge
must determine the inference ; so that, if the data and
objects are mind and ideas, when we find ideas in
sensation, which are due neither to one's own ideas nor
to one's own mind, we cannot infer a corporeal or
material substance, but must infer that the cause is
either other ideas or another mind. He had elimi
nated other ideas by his doctrine of the inactivity
of ideas. There remained another mind. Now, proceeds
he, though we are conscious of being able to produce
some ideas by will, yet the ideas of sense have not
a like dependence on our will ; there is therefore some
other will or spirit that produces them, and in an order
which proves that this cause is the spirit of God.1 Thus,
the solution, which was suggested by Descartes, as a
possible alternative in his ' Principia Philosophise,' 2 and
which ought to have been taken by Locke in the
Fourth Book of his Essay, when he had deserted mere
ideas in favour of an intuition of oneself and a demon
stration of God, was at length adopted by Berkeley in
his ' Principles.' If all the data are ideas and minds,
created and eternal, and if ideas are inactive, the only
logical conclusion is that the sensible ideas of created
minds are direct imprints of the eternal Spirit of God.

This logical conclusion of psychological idealism,
evaded by Descartes and Locke, was accepted by
Berkeley, with all its hypothetical consequences. As

1 Princ. xxviii.-xxx. 2 Descartes, Princ. ii. 1.

CHAP. vii. BERKELEY 205

usual, he felt the double edge of his weapon, and was
prepared not only with what is, but with what is not.
On the one hand, he concluded that God is, and on the
other hand, that matter is not, the cause of our sensa
tions.1 Secondly, he concluded that ' the set rules or
established methods, wherein the mind we depend on
excites in us the ideas of sense, are called the laws of
nature' 2 Thirdly, he concluded that God is not merely
the prime cause, but the immediate and sole cause of
sensible effects, setting aside second causes, such as the
sun and the motion of bodies : —

'And yet this consistent uniform working, which
so evidently displays the goodness and wisdom of that
governing spirit, whose will constitutes the laws of
nature, is so far from leading our thoughts to him, that
it rather sends them a-wandering after second causes.
For when we perceive certain ideas of sense constantly
followed by other ideas, and we know that it is not
of our own doing, we forthwith attribute power and
agency to the ideas themselves, and make one the cause
of another, than which nothing can be more absurd and
unintelligible. Thus, for example, having observed
that when we perceive by sight a certain round luminous
figure, we at the same time perceive by touch the idea
or sensation called heat ; we do from thence conclude
the sun to be the cause of heat. And in like manner
perceiving the motion and collision of bodies to be
attended with sound, we are inclined to think the latter
an effect of the former.'3

Finally, he presents us with his complete theory of
real things, when second causes have been expunged :4 —

' The ideas imprinted on the senses by the author of
nature are called real things ; and those excited in the

1 Berkeley, Princ. xxvi. 2 Id. xxx. 3 Id. xxxii. 4 Id. xxxiii.

206 PSYCHOLOGICAL IDEALISM PART n.

imagination, being less regular, vivid and constant, are
more properly termed ideas or images of things, which
they copy and represent. But then our sensations, be
they never so vivid and distinct, are nevertheless ideas,
that is, they exist in the mind, or are perceived by it,
as truly as the ideas of its own framing. The ideas of
sense are allowed to have more reality in them, that
is, to be more strong, orderly and coherent than
the creatures of the mind ; but this is no argument
that they exist without the mind. They are also less
dependent on the spirit, or thinking substance which
perceives them, in that they are excited by the will of
another and more powerful spirit ; yet still they are
ideas, and certainly no idea, whether faint or strong,
can exist otherwise than in a mind perceiving it.'

This passage marks a turning-point in the history of
idealism. Hitherto, the line between ideas of sensation
and ideas of conception had not been so carefully
drawn as that between all ideas and the physical
realities which cause them. Now, Berkeley, having
deduced the destruction of physical realities, while still
preserving the hypothesis that ideas are the objects of
sensation, was puzzled to find some boundary between
the real and the ideal. He drew it between the ideas of
sensation and the ideas of imagination, partly by their
vividness and faintness, but mainly because the former
are directly produced by God. Hence, he identified
sensible ideas with real things, at the same time explain
ing that they are after all only ideas.

Sensible ideas he declared to be his rerum natural
He even admitted corporeal substances, 'taken in the
vulgar sense for a combination of sensible qualities,' not
' in the philosophic sense for a support of accidents or

1 Princ. xxxiv.

CHAP. VII.

BERKELEY 207

qualities without the mind.' 1 So sure was he that sensible
ideas are the real things, that he even said that ' we are fed
and clothed writh these things which we perceive im
mediately by our senses ; ' that is, by sensible ideas.2
Thus did he reduce reality to ideas imprinted on our
senses by God without the intervention of physical
causes, sense to the presentation of sensible ideas repre
senting no external bodies, and knowledge to collections
of ideas inferring no external cause except God. He
took the show of sense for the nature of things, and
thought that, if the veil were uplifted, we should see
nothing but God.

This doctrine of reality, much more logical, but
also far narrower than that of Descartes and Locke, is
the transition to Hume's distinction of impressions and
ideas, and has ended in the ordinary sensationalism of
modern Berkeleians, such as Mill, who do not indeed say
that God is the direct cause of our sensations, but give up
the problem and leave sensations in mid-air, nor dogma
tise about all reality but confine themselves to known
reality, in other respects differing in nothing but ter
minology from Berkeley. The fundamental character of
Berkeleianism is the theory that everything real is either
my sensations and combinations of sensations, or those
of other minds. ' I do not believe,' says Mill, ' that the
real externality to us of anything, except other minds,
is capable of proof.' 3

It is often said that Berkeley is unanswerable, in his
final position that the real world consists of ideas im
printed on our senses, not by nature, but by the spirit
of God. He cannot be answered by the hypothetical

1 Princ. xxxvii. 2 Id. xxxviii.

3 Examination of Sir W. Hamilton's Philosophy, chap, xi., note,
sub Jin.

208 PSYCHOLOGICAL IDEALISM PART n.

realism of cosmothetical idealists, sucli as Descartes and
Locke and tlieir modern successors, because they start
knowledge, like Berkeley, with nothing but psychical
data, from which nothing but the psychical could be
inferred, and only suppose it to infer physical causes,
by bad logic. Berkeley was the first logician of idealism.
Cosmothetic idealism is an inconsequence, which must
end in pure idealism at last.

Again, he cannot be answered by intuitive realism,
because it rests on the false identification of the sensible
and external world by common sense, instead of appealing
to the distinction of the sensible effect from the external
cause by science. It is no use to knock the stick on the
ground, when Berkeley resolves the ground and the stick
into ideas, and the agent into a spirit. It is no answer
to assert that the things immediately perceived are real
things ; for Berkeley admits it, but says that they are
also sensible ideas or sensations.1 It is no answer to
oppose a presentative perception of apples and houses
to a philosopher, who agrees but rejoins that the things
presented are only collections of ideas. If Berkeley is
equal to the intuitive realist on the ground of common
sense, he is superior on the ground of science and philo
sophy. The intuitive realist supposes that the real world
directly perceived is external ; science shows that it is
within ; Berkeley adds that it is within the mind. The
intuitive realist supposes that a secondary quality is
directly perceived as a mere sensation in the mind, a
primary quality as a real quality in the external world ;
Berkeley, in a far more philosophic spirit, shows that
they are directly perceived in the same manner, for, as
sensible it is impossible to separate extension, figure and
motion from other sensible qualities. Both confuse two

1 Princ. xxxiii. scq. ; Hylas and PJiilonous, Third Dialogue, sub fin.

CHAP. vii. BERKELEY 209

realities, distinguished by science, the sensible and the
external ; but, if this common confusion could be over
looked, it would be more scientific to make the real
object of immediate perception, with Berkeley, entirely
internal, than, with intuitive realism, partly internal and
partly external — as if I could perceive the light of a
candle within me, and its extension in the outside world.

The truth is that idealists and realists have had too
many errors in common with Berkeley to answer him.
Idealists share his error that the data are ideas, realists
that the real world is the object of immediate percep
tion. All of them, also, confine themselves too much
to perceptible bodies, to the neglect of imperceptible
corpuscles. Within that narrow circle Berkeley has
no difficulty in resolving apples and houses, and even
mountains and rivers, into sensible ideas. But we must
turn the corner of pure idealism. The question is not
what it makes of the sensible and the perceptible, but
what it does with the imperceptible. The true contra
dictory instance against Berkeley's position is the
natural philosophy of the imperceptible world of cor
puscles, which cause, but are not, and cannot be inferred
from, sensations or sensible ideas. This is the answer of
physical realism. Let us proceed to its details.

In one way God, in another way nature, causes our
sensations. There are two opposite extremes to be
avoided — the substitution of nature for God, and that
of God for nature : the former the temptation of the
natural philosopher, the latter that of the natural
theologian. The natural philosopher prolongs the
chain of physical causes, until at last he feels tempted
to believe that he has expelled intelligence from nature,
and say, ' I have swept the universe with my telescope
and cannot find God.' The natural theologian, dazzled

210 PSYCHOLOGICAL IDEALISM TART u.

by the universal cause, is apt to neglect the subordinate
agency of physical causes, and forget nature in the love
of God.

Natural philosophy is limited by the nature of its
evidence. God is inferred by combining the evidence of
outer and inner sense ; but natural philosophy reasons
only from sensation and observation, without conscious
ness and reflection. Of itself, nature can neither prove
nor disprove a deity. Even within its own limits natural
philosophy is limited. Evolutionists, for example, have
been more successful in dealing with organisms than in
the far larger problem of the inorganic world. Evolution
consists in the differentiation of homogeneous matter.
Now differentiation invariably requires one of two condi
tions : either one efficient cause must act on different
materials, as when the same kind of motion produces
molar motion in one body and molecular in another ; or
different efficient causes must act on one kind of material,
as when different lengths of undulation produce sensible
heat or sensible light in the nervous system. Both
alternatives presuppose difference ; the former difference
in the patient, the latter difference in the agent. There
is no known instance of one kind of cause acting on
one kind of material and producing different kinds of
effects. Hence, if we suppose matter, absolutely homo
geneous, universally diffused, and reciprocally acting in
its various parts, it would contain no difference either
of agent or patient to produce the different effects of
actual nature ; but all its particles, at equal distances,
would exert all forces equally in all directions, and
produce an exact balance, with no differences whatever.
The theory of evolution, therefore, is no explanation of
the beginnings of difference. But given a pre-existing
difference, even of two groups of particles with dif-

CHAP. vir. BERKELEY 211

ferent arrangements of their primary qualities, how
ever slight, evolution is the further differentiation,
not of the absolutely, but of the relatively homo
geneous into the more heterogeneous, arising from
different structures acted on by one kind of agent, or
different agents acting on one kind of structure, or
different agents acting on different structures, and so on
ad infinitum, not a parte ante, but a parte post. There
must, however, be something else to cause an original
difference in things. But limited as natural philosophy
is within, it is still more limited from without. Having
only reasoning from outer sense and observation, it dis
covers physical causes ; but it cannot tell what else they
may be.

Natural theology now steps in, to supplement sensa
tion by consciousness, observation by reflection, and to
reason from both outer and inner sense. To observa
tion, a workman and a product have the mere appear
ance of cause and effect ; but when we add conscious
reflection, we infer that he is an artist using means to an
end ; and, when we observe again a similar work, we still
infer an artist. So from His work, natural theology
infers a Divine Architect of nature, establishing the
original difference of things, and developing further
differences, by using physical causes of effects as means
to ends. ' Omnia qua? agunt in virtute primi agentis
agunt.' When science shows that evolution develops
living organs, this is no reason why this very evolution
should not be a Divine means of producing fresh life.
The growth of a tree has not been regarded as inconsis
tent with Divine agency ; why, then, should not Divine
power be exercised in the whole growth of the world ?
On the other hand, the natural theologian must not
forget that, after all, the existence of nature must be

r 2

212 PSYCHOLOGICAL IDEALISM TART n.

more certain than that of God, and that, indeed, without
the order of nature the main part of the evidence for a
God disappears. If God is the intelligent cause, most
certainly the means used are physical causes. AK
attempts to argue that because God is the cause of all
effects, insensible motions are not causes, or that there
can be no evolution, must fail, because nothing is more
surely established than the powers and laws of motion.
To convert God from an Intelligent Will using physical
means into the direct and sole cause of every effect,
even to the threshold of our senses, is the greatest
danger that can befall natural theology, which must
then yield to the laws of the communication and conser
vation of motions.

No reconciliation of theology and science will be
found superior to that of Bacon,1 which admits too of
being perpetually enlarged with every physical dis
covery : God having made nature uses it as a means ;
the more physical causes, the more means at His com
mand ; the more elaborate and indirect the physical
process, the more subtle the Divine Architect ; who,
having established a difference in corpuscular structure,
uses the evolution of one particle acting on another as
His further process of differentiation and His most in
genious plan ; and natural philosophy is always, how
ever unconsciously, prolonging the chain of physical
causes to the throne of God. ' Sic Dei sapientia effulget
mirabilius,' says Bacon, 'cum Natura aliud agit, Provi-
dentiaaliud elicit, quam sisingulis schematibusetmotibus
naturalibus Providentias characteres essent impressi.'

Berkeley for nature substituted God. By his hypo
theses and logical deductions he was compelled to say

1 De Augmentis Scientiarum, iii. 4, sub fin. (Ed. Ellis and Speddmg,
Vol. i. p. 570.)

CHAP. VII.

BERKELEY 213

that ideas are imprinted on our senses, not by the in
sensible motions of physical substances, but by the
direct agency of God Himself. Instead of an Intelligent
Agent, using nature as the means to produce effects on
our senses, God, without the intervention of insensible
nature, thus becomes the direct and sole cause of every
sensible effect. There is God, then, and no nature, but
the nature of man. The good bishop flattered himself
that he was thus serving the cause of his religion. But
how different is the doctrine of the Bible ! In the
beginning God created the heavens and the earth ; and,
only after nature, man. This is the meeting-point of
religion and science.

In substituting God for nature, and denying second
causes, Berkeley not only falsified religion but also
contradicted science. He said that God is, but nature
is not, the cause of our sensations. His followers have
deserted his theory of religion, but they have supplied
no adequate theory of science. Any mental philoso
pher, who says that real things are our sensible ideas or
sensations, whether he says that they are produced by
God, with Berkeley, or, with the modern Berkeleian,
gives up the knowledge of the causes of our sensa
tions, in either case he is following Berkeley in rejecting
the positions of natural philosophy that the external
sun is the cause of sensible heat, that the motion and
collision of particles of air insensibly proceed till at last
they produce sensible sounds, and that imperceptible
corpuscles, with their configuration, number, motion,
and size, cause our sensations.1

Psychological idealism had gradually brought mental
philosophy into this state of paradox by the very poverty
of its data. Descartes was a scientific genius, labouring to

1 Cf. Princ. xxv., xxxii.

214 PSYCHOLOGICAL IDEALISM PAET IT.

bring a narrow mental into harmony with a wider natural
philosophy. Locke, beginning to feel the difficulty, depre
ciated natural philosophy, because he could not explain
it. Berkeley, logically deducing the vanity of the
attempt at explanation, boldly wrote a polemic against
the natural philosophy of corpuscles and their motions.1
This sad, but inevitable, defect is generally omitted or
extenuated by historians of philosophy. But Berkeley
himself was well aware what were the logical conse
quences of idealism. One passage from his polemic will
be sufficient : —

' Some have pretended to account for appearances
by occult qualities, but of late they are mostly resolved
into mechanical causes ; to wit, the figure, motion,
weight, and such-like qualities of insensible particles :
whereas, in truth, there is no agent or efficient cause
than spirit, it being evident that motion, as well as all
other ideas, is perfectly inert. (See sect, xxv.) Hence
to endeavour to explain the production of colours or
sounds by figure, motion, magnitude, and the like,
must needs be labour in vain. Accordingly, we see the
attempts of that kind are not at all satisfactory.'2

But we have seen, since Berkeley's time, a sure
progress in the natural philosophy of mechanical
causes. A striking contrast to the passage just quoted
may be found in the following quotation from Professor
Tyndall's ' Fragments of Science ' 3 :—

' The domain in which this motion of light is
carried on lies entirely beyond the reach of our senses.
The waves of light require a medium for their forma
tion and propagation ; but we cannot see, or feel, or
taste, or smell this medium. How, then, has its exist
ence been established? By showing that, by the

1 Princ. ci. seq. 2 Id. cii. 3 Pp. 72-3.

CHAP. VII.

BERKELEY 215

assumption of this wonderful intangible cether, all the
phenomena of optics are accounted for, with a fulness,
and clearness, and conclusiveness which leave no desire
of the intellect unsatisfied. When the law of gravi
tation first suggested itself to the mind of Newton,
what did he do ? He set himself to examine whether it
accounted for all the facts. He determined the courses
of the planets ; he calculated the rapidity of the moon's
fall towards the earth ; he considered the precession of
the equinoxes, the ebb and flow of the tides, and found
all explained by the law of gravitation. He therefore
regarded this law as established, and the verdict of
science subsequently confirmed his conclusion. On
similar, and, if possible, on stronger grounds, we found
our belief in the existence of the universal aether. It
explains facts far more various and complicated than
those on which Newton based his law. If a single phae-
nomenon could be pointed out which the aether is
proved incompetent to explain, we should have to give
it up ; but no sucli phenomenon has ever been pointed
out. It is, therefore, at least as certain that space is
filled with a medium, by means of which suns and
stars diffuse their radiant power, as that it is traversed
by that force which holds in its grasp, not only our
planetary system, but the immeasurable heavens them
selves.'

Berkeley's idealism is unscientific. From this point
we must retrace our steps by the method of analysis.
By the falsity of the consequences we must destroy the
original hypotheses and find the real data of reasoning
from sense to science. By a chain of logic, he had hypo-
thetically deduced that, if all objects of human know
ledge are ideas, derived from outer and inner sense, and
by the help of memory and imagination variously com-

21G PSYCHOLOGICAL IDEALISM IMRT n.

pounded into collections of ideas, in the minds of created
spirits, then such a spirit will be able to infer nothing
but ideas and spirits, and to conclude that, if all ideas
are inactive, our sensible ideas, which are passive and
not caused by our own will, must be imprinted on our
senses by the will of the eternal spirit of God ; so that
real things, as distinguished from mere ideas of imagina
tion, will be the sensible ideas directly imprinted on our
senses by Divine, without the intervention of physical
causes. Now, the flaw in this chain is in its last link,
in the logical but false rejection, with which it ends, of
the bodies, corpuscles, and mechanical causes, discovered
by natural philosophy.

What is corpuscular science ? In brief, there are
bodies insensible and imperceptible, or corpuscles. They
possess primary qualities, various species of which are
secondary qualities ; especially, they possess motion, a
primary quality, whose secondary species are undula
tions of cether, vibrations of air, &c., and which also exists
in various forms, such as cohesion, gravitation, chemi
cal attraction, electricity, magnetism, &c. Corpuscles
have innumerable similarities and uniform relations or
laws of nature, and especially the laws of motion
and of the causation of motion by motion. They
are also the particles of masses, or larger bodies, which
are partly inorganic and partly organic. Among organ
isms are bodies containing nervous systems, which
consist, like other masses, of corpuscles having the
various motions of bodies in general and a peculiar ner
vous motion, combined with muscular motion. Lastly,
some of the other bodies, among their innumerable pro
cesses of cause and effect, produce in nervous systems
sensible effects, such as sensible motion, sensible heat,
&c. Such are the objects of corpuscular science.

CHAP. TIT. BERKELEY 217

Corpuscular science destroys Berkeley's idealism in
his logical conclusion from his original hypotheses. He
denied second causes ; but motions producing motions
are second causes. He said that God's will is the sole
cause of sensible effects ; but corpuscular motions,
acting on the corpuscles of the nervous system, also
produce sensible heat, colour, sound, &c. If God is
the prime cause, nature is the second cause, by means
of which He acts on man. He said that the rules wherein
God excites in us the ideas of sense are the laws of
nature. But the uniform relations of corpuscular mo
tions among themselves are an immense system of laws,
compared with which the laws of their action on the
nervous system and the senses are but a diminutive
fraction. What account would it be of the universal
law of gravitation, of every particle to every particle
in the universe, to say that it is merely a rule to
excite in us the sensible idea or sensation of weight ?
God, then, is not the only cause, but under Him nature
is also the cause and law of sensible effects. Again,
Berkeley said that sensible ideas imprinted on sense by
God are the real things, and external bodies are not :
the Berkeleian says the same thing of sensations, only
without dogmatism about the sole causation of God
and about the absolute non-existence of external bodies.
But the natural philosopher knows that external bodies
are not sensations, but the causes of sensations and
sensible ideas. For example, the gravitations of par
ticles are not sensations, but are the known causes of
sensible weight being felt by us. Therefore, so far
from being non-existent, or so far from not being known
to exist, external bodies and their motions are known
to exist as causes of sensible effects. To the Berkeleian,
then, we must answer, not all known realities are sensa-

218 PSYCHOLOGICAL IDEALISM PART n.

tions ; to Berkeley himself, not all realities are sensible
ideas imprinted on our senses by the Author of our
being ; but some known realities are external bodies and
their qualities producing sensible effects in us. There
is a known world of real bodies, intervening between
God and man, and used by God as a means to cause
effects in our senses.

Corpuscular science destroys Berkeleian idealism not
only in its hypothetical conclusions but also in its
original hypotheses of the objects and data of human
knowledge. Insensible corpuscles and their qualities are
not our ideas, but the causes of our ideas. They are objects
of natural philosophy, which, in the hands of Newton
and his successors, is a kind of knowledge. Therefore,
not all objects of knowledge are ideas, and some of the
objects are corpuscular causes of our ideas. Again, if
the original data were ideas, these corpuscular causes
could not be inferred, as Berkeley logically showed.
But they are scientifically inferred by natural philo
sophy. Therefore, neither the original objects nor the
original data are mere ideas. Corpuscular science deals
double death to logical idealism.

Berkeley had logically deduced from his hypothesis
that all qualities are only sensible ideas. But natural
philosophy has shown that insensible corpuscles have
the primary quality of insensible motion, obeying various
laws, and that insensible modes of corpuscular motion
are the secondary qualities of light, heat, and sound
in the universe. Sir Isaac Newton showed that, beyond
the sensible resistance or weight which we feel, there is
an insensible gravitation of particles which pervades
the universe, which connects parts of bodies inaccessible
to our senses, and which, in one of its myriad appli
cations, causes bodies to feel heavy in our hands.

CRAP. vii. BERKELEY

Qualities, then, primary and secondary, are known in
natural philosophy to belong to external bodies, as well
as cause sensible effects in us. Moreover, their range in
the insensible world of science is infinitely more extensive
than their perception by sense. Qualities, therefore,
are not mere sensible ideas or sensations, but are mainly
the external characteristics of masses and corpuscles in
nature. But, again, external qualities of bodies could
not be inferred from sensible ideas of minds. There
fore, qualities, even as sensible, are not sensible ideas.
Berkeley was compelled by the logic of his idealism to
reduce all qualities to sensible ideas, but he was doubly
wrong in point of fact. Primary and secondary qualities,
as known to corpuscular science, are neither reducible
to, nor inferrible from, sensations or sensible ideas.

Again, from Locke's hypothesis that sense always
perceives ideas of qualities, Berkeley consistently de
duced that we cannot suppose an unthinking substance,1
that Locke's substratum is an abstraction, like materia
prima^2 and that the only known substance is a combina
tion of sensible qualities, or ideas, with which we are
fed and clothed.3 But are these conclusions true and
scientific ? The matter, known to natural science, is
durable, extended, moving, causing and receiving mo
tion ; it is not, indeed, also something else distinct from
being these things ; nor, however, is it mere duration,
extension, motion, causation, and reception of motion,
distinct but combined. In other words, matter or body
is not the abstract substrate supposed by Locke, nor
the equally abstract combination of qualities substituted
by Berkeley, but a qualified subject, characterised
by a number of qualities. Now, besides all this, it is

1 Princ. vii. 2 Id. xi., xvi.

3 Id. xxxvii.-xxxviii.

220 PSYCHOLOGICAL IDEALISM PART TI.

not, though it sometimes causes, a collection of sensible
ideas. A drop of water contains the particles enume
rated in the first page of this essay ; but the sensible
effect of it on any of my senses, and the ideas I after
wards form of it, do not contain anything of the kind,
and are totally incapable of containing such a number
of units of any kind, which are only inferred by reason.
If there are so many particles in a drop of water, how
many in a river, and how many in the ocean ? The
truth is, that an analysis of a substance into particles is
not a division of the sensible object, sensation, or sensible
idea, but of the external object inferred. Corpuscles,
then, are a proof of external bodies. Hence it follows
that known substances are not abstract substrates of
qualities, nor abstract collections of qualities, nor still
more abstract ideas of collections of ideas of qualities,
but qualified subjects, some of which are thinking and
partly psychical, others unthinking and entirely phy
sical. Again, as physical substances are not qualities
nor ideas, so neither could they be inferred from such
data. If sense never perceived anything but spiritual
sensations or sensible ideas or qualities, science could
not infer durable, extended, moving bodies containing
corpuscles. But these substances are the very subjects
of the laws of motion and gravitation. It follows, then,
that the data of sense from which they are inferred are
not mere qualities, still less sensations, least of all
ideas, but the nervous substance sensibly qualified.

To return at last to Berkeley's first principle. He
said that all the objects of human knowledge are ideas
imprinted on the senses, or else such as are perceived
by attending to the operations of the mind or collections
of these ideas. This supposed principle is a false
hypothesis containing two fundamental errors ; an error

CHAP. vn. BERKELEY 221

about objects known, and an error about objects per
ceived. The insensible and imperceptible corpuscles
discovered by natural philosophers are not ideas of any
of these kinds, though they are causes of them. Not
all the objects of human knowledge, then, are ideas.
Secondly, if the objects imprinted on the senses were
ideas, the insensible corpuscles could not have been
inferred. Not all the objects of human perception, then,
are ideas. Insensible imperceptible corpuscles are
physical objects of knowledge inferred from physical
data of sense. Similarly their esse is not perdpi, as it
would be if they were ideas. The esse of ideas of sen
sation is per dpi. The esse of a sensible object is per-
cipi by sense. An accident of the esse of an external
body, e.g. water, is per dpi by inference. But the esse
of an imperceptible corpuscle, e.g. in a drop of water,
is not perdpi at all.

Berkeley, by a confusion of esse and perdpi, adopted
a presentative theory of perception, like the intuitive
realists ; by a confusion of the sensible object with a
sensible idea, his presentative theory is not realistic
but idealistic ; by a confusion of the sensible and the real,
it is a theory that we present sensible ideas as the real
things.1 He recurs again to this theory, as the very
kernel of his philosophy, at the end of the Third
Dialogue between Hylas and Philonous :—

' Phil. I do not pretend to be a setter-up of new
notions. My endeavours tend only to unite and place
in a clearer light that truth, which was before shared
between the vulgar and the philosophers ; the former
being of opinion, that those things they immediately per
ceive are the real things ; and the latter, that the things
immediately perceived are ideas which exist only in the

1 Cf. Princ. iv.

222 PSYCHOLOGICAL IDEALISM PART n.

mind. Which two notions put together do, in effect,
constitute the substance of what I advance.'

Each one of the propositions in this theory is false.
First, the things we immediately perceive are real things,
but not the real things. There is an immense multi
tude of real things known to science, but not im
mediately perceived. The apple, the table, the house,
the river, the mountain, cause sensible effects, which
are real enough ; but they are external bodies whose cor
puscles are known to have a like but different structure
from that of the sensible effects ; the particles of a table
are not the particles of my hand lying on it, nor of my
tactile nerves, still less of the operation of sensation.
Secondly, the things immediately perceived are not
ideas which exist only in the mind. It is true that
they are within me, and here is Berkeley's superiority
over the intuitive realist. But, apart from the absence
of direct evidence that the hard, or hot, or heavy felt
is an idea within my mind ; if it were so, I could never
infer the bodies and corpuscles, which, as we have
found, are too well established in natural philosophy to
be any longer denied. Therefore, things immediately
perceived are, not ideas which exist only in the mind,
but bodily effects of bodies on the nervous system.
Lastly, Berkeley wishes us to draw the conclusion that
ideas which exist in the mind are the real things, and
that physical objects are not real things. His premises
to prove it, however, are both false ; for, as we have seen,
the things immediately perceived are neither the real
things, nor ideas. Hence his syllogism only proves that
ideas are real things, and some real things are ideas ;
which is true enough, but also consistent with other real
things. Xow, corpuscular science proves bodies which
are real things in the external world ; and, to infer them,

CHAP, vi r. BERKELEY 223

n

logic requires bodily data, wliicli are real things in the
nervous system. Other real things, then, are known
and perceived, besides ideas.

Berkeley's idealism — and we may add all Berkele-
ianism — is false, metaphysically, psychologically, and
logically :—

1. His metaphysical theory of existence is false,
because not all real things are sensible ideas whose sole
cause is God ; but some realities are known to be
physical causes.

2. His psychological theory of immediate perception
is false, because we immediately perceive neither sen
sible ideas, nor sensations, nor the real things, but real
physical effects, representing real physical causes.

3. His logical theory of reasoning is false, because
from the first he prefers imagination and memory of
ideas to reasoning about causes, and reasoning synthe
tically from hypotheses to reasoning analytically from
facts.

Berkeley omitted nature, between sense and God.
Starting from Locke's hypothesis of the objects of know
ledge, he rejected discoveries of natural philosophy,
when he ought to have preferred the latter to the former.
He ought to have gone still further, and surrendered
not only Locke's hypothesis of the objects, but also the
hypothesis of Descartes that the data of knowledge are
psychical ideas. When Newton had shown what could
be done in natural philosophy, mental philosophy should
have reformed its data to explain his discoveries. But
how seldom philosophers realise that their theories of
man ought to explain a Shakespeare, a Bacon, a Newton !
To infer the Newtonian philosophy, the senses of man
must perceive, not ideas of qualities, but various parts
of the physical substance of the nervous system sensibly

224 PSYCHOLOGICAL IDEALISM TART n.

qualified as durable, extended, moving, as well as sound
ing, heated, coloured ; from which even an ordinary man
infers insensible bodies, a scientific man their imper
ceptible corpuscles and motions and laws. If all objects
of human knowledge were ideas of spirit, man could
infer nothing but spirit and ideas. But the antecedent
is an hypothesis, for which Berkeley had no authority
except Descartes and Locke : the consequent is false,
being contradictory to corpuscular science : therefore,
the antecedent hypothesis is also false, because from
true premises it is not possible to . draw a false conclu
sion. The real world includes, between the sensible
and the supernatural, the natural world of insensible
bodies and imperceptible corpuscles, which are physical
objects of scientific knowledge inferrible only from
physical data of human sense. Such is the answer of
physical realism.

225

CHAPTEE VIII.
BERKELEY'S THEORY OF VISION.

IN answering the objections which might be made
against his ' Principles,' Berkeley refers to his ' Essay
towards a New Theory of Vision ' as follows :—

' Thirdly, it will be objected, that we see things
actually without or at a distance from us, and which
consequently do not exist in the mind, it being absurd
that those things which are seen at the distance of
several miles should be as near to us as our own
thoughts. In answer to this, I desire it may be con
sidered, that in a dream we do oft perceive things as
existing at a great distance off, and yet for all that,
those things are acknowledged to have their existence
only in the mind.

4 But for the fuller clearing of this point, it may be
worth while to consider, how it is that we perceive dis
tance and things placed at a distance by sight. For
that we should in truth see external space, and bodies
actually existing in it, some nearer, others farther off,
seems to carry with it some opposition to what hath
been said, of their existing nowhere without the mind.
The consideration of this difficulty it was, that gave
birth to my " Essay towards a New Theory of Vision,"
which was published not long since. Wherein it is shown
that distance or outness is neither immediately of itself
perceived by sight, nor yet apprehended or judged of

Q

226 PSYCHOLOGICAL IDEALISM TART n.

by lines and angles, or anything that hath a necessary
connexion with it : but that it is only suggested to our
thoughts by certain visible ideas and sensations attend-
in o- vision, which in their own nature have no manner

o

or similitude or relation, either with distance, or things
placed at a distance. But by a connexion taught us
by experience, they come to signify and suggest them
to us, after the same manner that words of any language
suggest the ideas they are made to stand for. Insomuch
that a man born blind, and afterwards made to see,
would not, at first sight, think the things he saw, to be
without his mind, or at any distance from him. See
sect. xli. of the forementioned treatise.

'The ideas of sight and touch make two species,
entirely distinct and heterogeneous. The former are
marks and prognostics of the latter. That the proper
objects of sight neither exist without the mind, nor are
the images of external things, was shown even in that
treatise. Though throughout the same, the contrary be
supposed true of tangible objects : not that to suppose
that vulgar error, was necessary for establishing the
notions therein laid down ; but because it was beside my
purpose to examine and refute it in a discourse concern
ing vision. So that in strict truth the ideas of sight,
when we apprehend by them distance, and things placed
at a distance, do not suggest or mark out to us things
actually existing at a distance, but only admonish us
what ideas of touch will be imprinted in our minds at
such and such distances of time, and in consequence of
such and such actions. It is, I say, evident from what
has been said in the foregoing parts of this treatise, and
in section cxlvii. and elsewhere of the Essay concerning
Vision, that visible ideas are the language whereby the
governing spirit, on whom we depend, informs us what

CHAP. vin. BERKELEY'S THEORY OF VISION 227

tangible ideas lie is about to imprint upon us, in case
we excite this or that motion in our own bodies. But
for a fuller information on this point, I refer to the
Essay itself.' l

Here we find from Berkeley's own words that he
had more than one object in writing the ' Theory of
Vision.' It is an essay half physical, half psychological,
and this doubleness of purpose has ever since clung to
the subject. On the one hand, he wanted to destroy
the exaggerations introduced by mathematicians into
optics, by showing that the eye is not fitted to see any
thing, and therefore not any lines and angles, beyond
itself; on the other hand, he wanted to support the
idealistic theory, which he had already conceived, and
shortly intended to publish in the ' Principles,' by show
ing that, whereas we do not see things without, we do see
visible ideas and sensations. In its first purpose, the
main thesis of the ' Theory of Vision ' is a great optical
discovery, though exaggerated ; in its second purpose,
it is an excellent disproof of intuitive realism, but no
proof at all of psychological idealism. Perhaps no
treatise has ever evinced such a singular compound of
genius and confusion. The effect both of its truth and
its falsity persists to this very day, especially in the
hypothesis of ' local signs.'

What does Berkeley prove about the sense of vision ?
He divides the subject into four parts — distance, magni
tude, situation, and the difference between sight and
touch.2 On the first point, he says that ' distance being
a line directed endwise to the eye, it projects only one

1 Princ. xlii.-xliv.

2 Theory, i. Distance is discussed in i.-li. ; Magnitude, lii.-lxxxvii. ;
Situation, Ixxxviii.-cxx. ; The difference between sight and touch, cxxi.
to end.

Q 2

228 PSYCHOLOGICAL IDEALISM TAKT ir.

point in the fund of the eye, which point remains
invariably the same, whether the distance be longer or
shorter.' l This proves, according to him, that we do
not see distance at all, but really that we do not see
remote distance, in depth or the third dimension, verti
cally from the eye. As he says elsewhere, we see no
solidity or profundity.2 On the second point, he shows
that we do not see the real magnitude, greater and
smaller, of an external object. ' Thus, for instance,' he
very properly remarks, ' the very same quantity or
visible extension, which in the figure of a tower doth
suggest the idea of a great magnitude, shall in the figure
of a man suggest the idea of much smaller magnitude.' 3
On the third point, he relies on the inverted image in
vision to show that we do not see the real situation, as
high and low, of external objects. On the fourth point,
he makes the instructive remark that there is no vision
of resistance,4 and he has brought out more clearly
than any of his predecessors that ' there is no one self
same numerical extension perceived both by sight and
touch,5 and that 6 we never see and feel one and the
same object.' G

This conclusion is the great stumbling-block to the
ordinary man, who has so overlaid sense with inference,
and, we may add, had so many visible pictures of his
hand and other members visibly touching visible objects,
all within his sense of vision, that he finds himself almost
incapable, even when he becomes a philosopher, of
realising to himself that he is really seeing one set of
objects within the retina and feeling another within the
tactile nerves, while he infers an external object in re-

1 Theory, ii. ~ Ib. cxxxv., cliv.

3 Ib. Ivii. 4 Ib. cxxxv.

5 Ib. cxxi. 6 Ib. xlix.

CHAP. MIL BERKELEY'S THEORY OF VISION 229

lation to both. Nevertheless, Berkeley verified the pre
vious scientific discovery of the distinction between ex
ternal and sensible objects by his new discovery of the
invisibility of remote distance. Since we do not see
distance in the third dimension from the eye, we cannot
see, but only infer, a remote object. The visible object
might, indeed, still be an object touching the eye ; but
even this hypothesis is negatived by the further study
of the nervous system.

What did Berkeley not prove about the sense of
vision ? On the very first point, while he proved that we
do not see remote distance he did not prove that we do
not see distance at all. He did not prove that we do
not see a surface painted on the retina, with its distances.
There are three dimensions of extension or space ; in
each there is distance — distance from point to point of
a line, from line to line of a surface, from surface to
surface of a solid : in each dimension the parts or places,
which are distant, are out of one another. Now what
he proved was that there is no vision of the third
dimension, not that there is none of the other two ; that
there is none of distance in depth, not that there is
none of distance in length and width ; that there is none
of outness in the external world, not that there is none
of outness of parts on the surface painted on the retina ;
that there is none of solid, not that there is none of
superficial extension ; that there is none of distance
endwise to the eye, not that there is none of space and
its distances within the eye. In short, he concluded
more than he proved. ' It is,' lie says, ' I think, agreed
by all, that distance, of itself and immediately, cannot
be seen.' 1 It is still agreed by present psychologists ;
but we want something more than agreement to prove

1 Tlicory, ii.

230 PSYCHOLOGICAL IDEALISM PART IT.

that, because remote distance is not seen, therefore no
distance can be seen. ' From what we have shown,' he
says, 'it is a manifest consequence that the ideas of a
space, outness, and things placed at a distance, are not,
strictly speaking, the object of sight ; they are not
otherwise perceived by the eye than by the ear.' l But
he had proved only that we do not see external things
placed at a distance and their outness in space. It still
remained, and remains even now, to be proved that the
space, the outnesses, and the distances, within the sur
face of the picture painted on the retina, are not objects
of sight.

Therefore, he did not prove that we have no vision
of space. He vacillated ; sometimes allowing, some
times denying that the extended is visible,2 and finally
deciding that ' what we strictly see are not solids, nor
yet plains variously coloured ; they are only diversity
of colours.' 3 But the same evidence, which proves that
we do not see solid distance, proves that we do see a
plain, with its superficial extension and the distances on
its surface. ' There is, at this day,' as he says himself,
6 no one ignorant that the pictures of external objects
are painted on the retina, or fund of the eye.' 4 He began
then with the external object and the retina. Very well ;
but what is the external object, and what the retina?
Both of them have surfaces. Undoubtedly the- former
reflects what the mathematician abstracts as rays, but
what the physicist knows to be undulations, which ulti
mately impress the terminations of nervous fibres in the
retina of the eye. We do not see the sides of these
rays or undulations : hence we do not see distance in
the third dimension. But we do see the imprints of

1 Theory, xlvi. a Cf. ib. xliii., xlv.-xlvi., xlviii., civ.

3 Ib. clviii. 4 Ib. Ixxxviii.

CHAP. TIII. BERKELEY'S THEORY OF VISION 231

their ends. Now, in the first place, the end of any wave
of sether, however small, is a surface ; in the second
place, the end of every single optic fibre is a surface ;
thirdly, as a fact, no one undulation of aether ever
reaches the eye alone ; and, fourthly, no one nervous
fibre is excited alone, but the whole retinal surface by
a whole undulatory surface of aether. Though, there
fore, the visible picture painted on the retina by the
external world is not itself solid, it is painted by
the surface of one solid on the surface of another.
Not remote distance, but superficial extension is
visible.

It is unscientific in the extreme to arbitrarily select
one part of the optical evidence and reject the rest, or to
see through the mathematical abstractions of the line
and the angle, and then to confuse mathematical points
with the extended ends of physical objects. This is the
mistake of Berkeley. He knows that the rays of light
are not mathematical lines, yet he says at the opening
and often repeats, what has been repeated after him
again and again ad nauseam to the present day, that a
point is presented to the retina. Nothing of the kind ;
it is not a point, but, to say the very least, a physical
ray's extended end, which is a surface, presented to the
extended end of another physical object which is a
surface, the end of a nervous fibre. There is, no doubt,
a minimum visibile, which may be coextensive with the
end of a nervous fibre ; but it is not a point, it is a
surface. The whole point-to-point theory of vision is
nothing but a mathematical abstraction converted into
a physical reality.

It is true that the retina itself may not be sensible ;
but whatever part of the optic nerve or of the brain
itself is first sensible, that part is a surface. It is true,

232 PSYCHOLOGICAL IDEALISM PAKT IT.

again, that we have admitted a psychical element in
sensation, but we can only interpret its object by con
sciousness, observation, and reasoning. We have seen
the verdict of the two latter evidences : every physical
part without and within us has a surface. Now, what
does consciousness of vision say ? Why, I cannot help
being conscious that I am at this moment seeing an
extended surface. I confuse this picture within, I
admit, with what I infer without ; but the scientific
distinction between the external and the sensible only
shows that I was wrong in the supposition of the exter
nality or remote distance of the sensible, not that I was
wrong in being conscious that I see an extended sur
face, a plain variously coloured. The whole evidence,
scientific and conscious, is in favour of the visible
object being like a painting, or still more like a picture
in a camera obscura, flat to sense, inferred by a complex
process of reasoning to represent an external solid,
but confused, by a long-standing association, with the
external solid itself.

Again, on the second point, Berkeley proved that we
do not see the real magnitude of an external object.
That is no reason why we should not see the magnitude
of the visible object impressed on the retina, nor why it
also should not be a real magnitude, though distinct
from external magnitude. On this point, again, he
vacillated. First he admits a size of things seen, that
they grow greater and smaller, and that there is not
only a tangible but a visible magnitude ; then he says
that visible extension, though immediately perceived, is
nevertheless little taken notice of; and finally contends
that the ideas of visible magnitude are equally fitted to
bring into our minds ' the idea of small or great, or of
no size at all of outward objects,' like the words of a

CHAP. VTIT. BERKELEY'S THEORY OF VISION 233

language.1 The truth is, that we see an extended
coloured plain, as we have already said. We are not
able to alter its whole size on a single retina, because
the whole retina is used at once ; and this is a great
point of difference from touch, wherein we use a finger,
a hand, or our whole body to touch at pleasure. The
only variation we can get in the size of the whole
picture is the difference of magnitude between the area
of a single retina, and the whole field of vision covered
by both eyes. Usually, however, both eyes are used at
once, when the visible picture has a single fixed mag
nitude. But the parts of it have very varying degrees
of magnitude ; for example, the black spot made by a
blot of ink covers a trifling amount of the retinal mag
nitude, compared with that impressed by the white
paper before me. Hence within a single visible magni
tude, fixed on the retina, we see all sorts of sizes of the
parts not behind but beside one another, some greater,
others smaller, and therefore having various relations of
size to the whole retinal picture. It is on this sensation
of varying degrees of magnitude of the parts relatively
to the fixed magnitude of the whole of the superficial
picture on the retina, that the wonderful subtlety of the
sense of sight is founded. In itself, this vision of magni
tude within a magnitude carries us no further; but
when allied with data of other senses, it becomes the
basis of countless inferences about external size. For
example, the sight of Snowdon, when I am in the open
air, is smaller than the sight of my own room, when I
am indoors ; but knowing in other ways the real size of
Snowdon and of my room, I can from sight measure the
relative sizes of parts of each in a way possible to no
other sense. There is another element in the vision of

1 Theory, xxviii., 1., liv., Ixi., Ixiv.

234 PSYCHOLOGICAL IDEALISM PART n.

size about which we must be careful. We see tlie
magnitude of the parts relatively to the whole retinal
magnitude. There are minima, beyond which this de
duction of visible parts cannot go, supposed to be con
nected with the distinction of nervous fibres. But, as I
said before, a minimum visibile is not a point but an ex
tended end, like the end of a pencil. Secondly, it is
tempting, with Berkeley, to conclude that the minimum
is always the same size in vision.1 But it is not at all
impossible that the parts impressed on adjacent nervous
fibres may not be always visibly distinct. In looking at
an object of a single colour, as a white leaf of a book,
we do not so carefully distinguish small parts as when
the object is very varied, as in reading the printed
matter. The minimum, impressed on each fibre, may be
always the same, and yet the minima, distinctly visible,
greater or smaller according to the intensity and variety
of the excitations. On the whole, then, there is a visible
magnitude of the picture, always of the same size, deter
mined by the retina ; visible parts, greater and smaller,
in reference to the whole size ; minima visiUlia, beyond
which vision cannot go, but to which, perhaps, it does
not in every act of vision reach.

On his third point, Berkeley proved that we do not
see the real situations of external objects and, in especial,
that we do not see which is up and which down, but
an inverted image. He did not prove that sight sees
no places in its inverted picture ; nor has any of the
many philosophers, who have strangely attacked visible
places, ever disproved them. Berkeley, as before,
vacillated : he first denied them and then admitted
them. He first says that a blind man returned to sight
' would not at first think that anything he saw was

1 Theory, Ixxx.

CHAP. VIIT. BERKELEY'S THEORY OF VISION 235

high or low, erect or inverted.' l Afterwards, he says
that ' we denominate any object of sight, high or low, in
proportion as it is more or less distant from the visible
earth.'2 The latter is the truth. In the visible flat
extended picture, of a constant retinal magnitude, we
not only see some parts greater and some smaller, but also
some in one place, some in another, though all inverted.
Nor is there any occasion to suppose that the image is
ever re-inverted. It includes images of our own body
and of the earth. From the data of touch, we infer
that our feet are down towards the earth's centre, and our
heads erect as away from it ; next, we find, over and
over again, that these inferred objects, in this order,
have in the visible picture various parts corresponding
to them, in a corresponding order — one part for the
earth, another for our feet, another for our heads : con
sequently, from this combined evidence of touch and
sight, we do not see but infer that the part of the retinal
image answering to the head represents up, and the part
corresponding to the feet represents down, and so on
with all other visible places.

-Av-eL, 0R ^ie fourth point, Berkeley proved that we do
not see and feel the same object, and that the visible
picture is numerically distinct from the tactile impres
sion produced by the same external object; e.g. my
retinal picture of the paper before me is in my optic
nerve, my tactile impression in my nerves of touch.
But he asked the further question whether they are also
specifically distinct, or whether there is anything in
common, or similar, in the visible object and the tangible
object. After having, though in the vacillating manner
already stated, admitted in the visible picture a visible
extension, visible magnitude greater and smaller, and

1 Theory, xcv, 2 Ib. cxi.

236 PSYCHOLOGICAL IDEALISM PART IT.

visible situations high or low in relation to visible earth,
he answered the question by denying that there is any
thing common to sight and touch. \ The true answer is
that his previous admissions were better than his final
theory. Vision sees a picture visibly extended in the
above-mentioned ways : touch feels a tangible imprint
extended in the same ways. The visible and tangible
objects, so far as the former is coloured and the latter
heated, are dissimilar; so far as the former is in the
optic and the latter in the tactile nerves they are not
numerically the same ; but, so far as they are both ex
tended, they are similar. Aristotle was right in dis
tinguishing special and common sensibles, and in assign
ing the extended both to sight and to touch.1 Locke
was right in repeating the distinction.2 Berkeley was
wrong as well as inconsistent in rejecting it. But his
rejection has infected the whole subsequent course of
the science of vision, the metaphysics of space, and the
psychology of sense.

Berkeley's theory contains a double paradox. In
the first place, he supposed that we see no visibly ex
tended object, when all he had proved is that we see
no 'Visibly7 remote object. He used the action of the
external object on the retina, to prove that we do not
see a line endwise, but a point ; and then discarded it
when it would also prove that we do not see a point
but a surface broadwise presented to the retina.3 He
had no definite idea of what is meant by distance. He
evidently confused it at first with the third dimension
of space.4 Afterwards, he saw that there is a visible
distance between interjacent visible points.5 But he
never fairly faced the fact that distance is the interval

1 De Anima, ii. 6. 2 Locke, Essay, ii. 5.

3 Theory, clvii. 4 Ib. ii. 5 Ib. cxii.

CHAP. vin. BERKELEY'S THEORY OF VISION 237

between any places, that there is a distance in length
and width, as well as in depth, and that, though distance
in depth is invisible, distance in both the other dimen
sions is visible. He coolly rejected the constant appeal
of the geometer to visible figures.1 He supposed a
person without touch but with sight, and asked what
kind of geometry he would produce ; a useless question,
because man is an animal, and an animal without touch
impossible. This supposititious seeing geometer would
have certain limits, as Berkeley says : first, he would
have no sense of a solid,2 which requires distance in
the third dimension ; secondly, he would have no sense
of resistance, which requires touch.3 He would, there
fore, infer no external world. But he would have a
sense of an extended plain with its distances, the mag
nitudes of its parts, and the situations of its places. He
would, therefore, see a plain, and on that plain the
outnesses of the parts to one another, and their distances
from one another in length and width. He would have
a sense, not a science, of space. Yet Berkeley denied all
these consequences of his previous admissions, assuming
that an object presents a point endwise to the eye. In
short, throughout the ' Essay,' the same merit is con
stantly vitiated by the same defect, the discovery of
the invisibility of remote distance confused with the
assumption of the invisibility of extended space.

In the second place, he constantly asserts that what
we strictly see are not solids, nor yet plains variously
coloured ; they are only diversity of colours.4 He then
defies us to assign any similitude between the visible
and the tangible, and concludes that the objects of
vision are not similar to external objects, but mere

1 Theory, cl. scq. 2 Ib. cliv. 3 Ib. cxxxv.

4 Ib. xliii., Ixv., ciii., cxxix.-cxxx., cliii.-clviii.

238 PSYCHOLOGICAL IDEALISM TAUT n.

signs, like words ; so that ' the proper objects of vision
constitute an universal language of the Author of
nature.' l Meanwhile, all he has proved is, that we
do not see remote distance in the third dimension.
Secondly, he has forgotten his constant admission that
we see visible extension, visible magnitude, visible
situations. Thirdly, he is plainly under the dominion
of the abstraction of qualities. He says very truly that
we can neither abstract the idea of visible extension
from colour, nor that of colour from visible extension.
But the extraordinary thing is, that he thinks this
argument proves, not that we see something coloured
and extended, but that we see colour, not extension.2
He is evidently under the dominion of this simple
fallacy : colour is not extension ; what we see is colour ;
therefore it is not extension. But in reality, though
colour is not extension, what is coloured can be also
extended ; what we see is a picture at once coloured and
extended, and that is the reason why we cannot separate
colour and visible extension, but only attend the more
to colour, or the more to visible extension in the self
same picture.* Hence, the visible object is not an
arbitrary sign, but similar to the tangible object felt,
and, we may add, to the external object inferred, in
extension. / The visible figure in geometry is not, in
deed, the object of the science, but it is the best illus
tration of the object to the sense and imagination of the
geometer. The visible object is not like a word, and
vision not like a language, which may or may not be
like what is signified, but the former is the sensible
object, and the latter the sense, most correspondent to
the extended external world, though not the most
direct way of inferring it. In short, vision sees the

1 Theory, cxlvii. ; cf. li., Ixiv.-lxv.. cxliii. 2 Ib. cxxx.

CHAP. vin. BERKELEY'S THEORY OF VISION 239

visibly extended, touch feels the tangibly extended,
reason infers the externally extended; and all three
objects are similar, though not the same, in extension.

So far, we have seen how brilliant, and how delu
sive, was Berkeley's discovery of the invisibility of
remote distance, in its physical aspect. But, as we
said above, he wrote the ' Theory of Vision ' also in a
psychological interest. He certainly proved in it that
we_ do not see an object at a distance; and it is a
curious problem that, after this discovery, the intuitive
realists should have advanced their hypothesis that we
immediately perceive the external world. The reason
is that Berkeley buried his discovery under such a heap
of errors, that we can hardly be surprised if the truth
for a time lay hid. He found out that we do not see *
depth, and confirmed the theory that we do not see
anything without. But he also proceeded to infer that
we do not see visible space, but only various colours.
Not content with this double paradox, he proceeded to
another, which was indeed a main object of the ' Essay ; '
namely, that we see only visible ideas, visible ideas of
colours. There is no better instance of the extraordinary
way in which the assumption of idealism is made in books
of philosophy, than its sudden appearance in the ' Theory
of Vision.' After he has concerned himself with the
external objects, and the rays of light, and the retina
of the eye, we suddenly find ourselves transplanted into
quite a new world with the words : ' It is evident that
when the mind perceives any idea, not immediately and
of itself, it must be by the means of some other idea.' l

The invisibility of distance in the third dimension
proves that we do not see external objects at a distance
from the eye. The propagation of undulations to the

1 Theory, ix.

240 PSYCHOLOGICAL IDEALISM PART n.

retina and the consequent nervous motion prove that
we do not see external objects at all. But neither
evidence proves that we see something not merely
within ourselves, but also within our minds, or that the
visible object is a visible idea. Berkeley, however, falls
into this ordinary idealistic non sequitur, without any
evidence, either physical or psychological, throughout
the ' Theory of Vision.' For instance, he says that c a
man born blind, being made to see, would, at first, have
no idea of distance by sight ; the sun and stars, the
remotest objects as well as the nearer, would all seem
to be in his eye, or rather in his mind.' l Similarly,
lie assumes it as agreed on all hands that colours are
not without the mind, from which, of course, it
would follow that neither is visible extension.2 He
even uses the mere assumption that what is in the eye
is in the mind to argue that, as the objects of sight
do not exist without the mind, the pictures painted on
the bottom of the eye are not the pictures of external
objects ! 3 Meanwhile, the evidences, which are all
drawn from the way in which external objects affect
the retina, prove that there is a variously coloured
picture produced in the camera obscura of the eye upon
the retina, but prove absolutely nothing at all about
visible ideas within the mind.

Let us now shortly resume what Berkeley proved
and did not prove about vision as a sense. We see no
remote distance, no real magnitude, and no real situa
tions of external objects ; no solidity, no resistance, no
protrusion ; no outness of the world in external space :
this is what he proved. He vacillated about visible
extension, and finally concluded, but did not prove, that
6 what we strictly see are not solids, nor yet plains

1 Theory, xli. ; cf. xcv. 2 Ib. xliii. 3 Ib. cxvii.

CHAP. vin. BERKELEY'S THEORY OF VISION 241

variously coloured ; they are only diversity of colours.'
Nor did he prove that the visible object is not on the
retina, nor in the optic nerve, but is a visible idea in
the mind : this is a petitio principii committed very early
in the ' Essay.' l Consequently, he did not prove that
vision is a universal language, and that visible objects are,
like words, mere signs of extended objects without being
extended. The same optical evidence, which proves
that we do not see remote distance endwise, proves
that we do see the extended imprinted broadwise on the
retina of our eyes. The visible picture, with distances
not endwise but broadwise, magnitudes of parts, and
situations of places, though numerically different, is
specifically similar to the tangible imprint, and to the
inferred original of both, in physical extension.

But while Berkeley's psychological interest was en
ticing him to resolve optical effects into visible ideas,
his physical discovery was at the same time forcing
him to recognise external objects to cause them. At
the very outset he admits the existence of a distance
projecting an effect on the eye.2 In the sequel, he
allows that ' the object which exists without the mind
and is at a distance ' is different from the visible object,3
the former remaining the same while the latter alters
according to the remoteness of the eye from the external
object. He talks of our advancing forward to it so
many paces or miles.4 He considers that it affects not
only our bodies but even our minds. ' We regard,' he
says, 'the objects that environ us in proportion as they
are adapted to benefit or injure our own bodies, and
thereby produce in our minds the sensations of pleasure
and pain.' 5 No realist requires more admissions. Given
only the eye, and all the universe follows, bathed through-

1 Theory, ix. - Ib. ii. 3 Ib. Iv. 4 Ib. xlv. 5 Ib. lix.

R

242 PSYCHOLOGICAL IDEALISM TART IT.

out its mass and its molecules in that light which is re
flected from external bodies on the retina of the eye.
Optics requires external bodies to reflect and a sentient
body to receive light. As soon as Berkeley becomes a
natural philosopher, he deserts the pure idealism of
his ' Principles,' and admits between God and the ideas
within our minds the intervention of ' unthinking ' objects
projecting effects on the retina and causing visible ideas.

Nevertheless, he proceeded to misinterpret the ex
ternal object. In the ' Principles,' with much consist
ency though with no truth, he emphatically denies that
any sensible object, any primary or secondary quality,
is anything but an idea within the mind. But in the
'Essay,' while he thought that visible objects are
ideas within the mind, he identified the external and
the tangible, and supposed that tangible objects exist
without the mind./' For,' says he, ' all visible things
are equally in t]*e mind, and take up no part of the
external space ; and consequently are equidistant from
any tangible thing, which exists without the mind.' l
This view, which is entirely inconsistent with the
idealism of the ' Principles,' is curiously like intuitive
realism. But even if it were possible that colour and
extension could be wholly separated in this manner, at
any rate the identification of tangible and external ex
tension is a confusion of effect and cause. Eeally, the
externally extended object is the common original of
the visible and the tangible objects, both of which are
within ourselves.

Berkeley's identification of the external and the
tangible led him into two false consequences. In the
first place, it led him to deny any common cause of
sight and touch. ' It is a mistake,' he says, ' to think

1 Theory, cxi. ; cf. Iv.

CHAP. vni. BERKELEY'S THEORY OF VISION 243

tlie same thing affects both sight and touch. If the
same angle or square which is the object of touch be
also the object of vision, what should hinder the blind
man, at first sight, from knowing it ? ' It is true that
the square felt and the square seen are not the same :
one is in the tactile, the other in the optic nerves. It
is true, also, that a blind man, when first restored to
sight, would have a difficulty in comparing them. But
this is no proof, and it is not true, that the external
square object which causes the tangible square is
different from the square which causes the visible square.
Trafalgar Square is one object, though it is one thing to
look at it and another to walk round it. It is the same
crystal which presses the hand and dazzles the eye of
the natural philosopher, though the modes of motion,
by which it gravitates towards the hand, and by
which it reflects undulations towards the eye, are
different. Otherwise, science would be impossible, for
it would never be concerned with one and the same
object. Secondly, the identification of the external
and the tangible led Berkeley into a paradoxical theory
of the object of geometry. He proved, in the end of
the ' Theory of Vision,' that ' neither abstract nor
visible extension makes the object of geometry.' 2 But,
by a false disjunctive judgment, he thought himself
entitled to conclude that the object of geometry is there
fore tangible extension. This conclusion entailed the
corollary that a geometrical square is really a tangible
square, and is not even represented by a visible square,
which, according to him, has four parts rather than four
sides.3 But who ever heard of a geometer feeling a square
rather than looking at one ? Through his confusion of
the external and the tangible, Berkeley has entirely

1 Theory, cxxxvi. 2 Ib. clix. s Cf. ib. cxli.-cxlii.

K '2

244 PSYCHOLOGICAL IDEALISM TAUT IT.

overlooked tlie real square of the geometer, which is
neither an abstract idea nor visible nor tangible, but an
object of reasoning, capable of being partially repre
sented by a tangible square, much better by a visible
square, but perfectly by neither. In elementary geo
metry, a geometrical figure is better represented by
sight than by touch ; even sight fails adequately to
represent more complicated figures, such as a chiliagon,
while in the geometry of infinites a polygon with infinite
sides is a pure object of reasoning.

But Berkeley's confusion of external and tangible
objects needs no further criticism, for having published
it in 1709, he retracted it in 17] 0. In his 'Essay
towards a New Theory of Vision,' it was put forward
as the explanation of the external object, considered by
geometry, and required by optics. In his ' Principles,' 1
he calls it himself a ' vulgar error.' But he at once flew
to the opposite error, and confused the tangible object
with a tangible idea ; falsely identifying the physical
with the psychical, and logically but falsely resigning
the external object altogether. At the same time, he
insinuated that this oscillation between intuitive realism
and psychological idealism made no difference. In
reality, it spoilt his theory of external objects in both
books. In the ' Essay,' his confusion of the external
and tangible concealed from him that the external
object is the common original of touch and of vision,
distinct from the objects of both. In the 'Principles,'
his confusion of the tangible, as well as the visible,
with ideas made him omit the external object altogether.
Although the object at a distance directing a line end
wise to the eye had been the foundation of his discovery
of the invisibility of remote distance, he now proceeded,

1 Princ. xliv., quoted at the beginning of this chapter.

CHAP. vnr. BERKELEY'S THEORY OF VISION 245

in defiance of the science of optics, to make visible as
well as tangible ideas effects of ' the governing spirit,'
with not a single word about external objects without
our minds. At the price of the physical truth of
the ' Essay ' he saved the psychological idealism of the
' Principles/

Berkeley, in the ' Principles,' is a logical idealist ;
but Berkeley, in his works, is, like Locke, two philo
sophers in one. On the one hand, take the ' Theory of
Vision.' Here he is Locke, with his admissions. In
the same somewhat half-hearted way he recognises the
external objects of science : he has an undercurrent of
ontology : he is a cosmothetic idealist in visible ideas
which he supposes to be projected by external objects,
and an intuitive realist in tangible objects which he
supposes to be externally felt, as Locke, after limiting
sensation to ideas, had supposed primary qualities with
out to be objects of a kind of bastard sensation. On the
other hand, take the ' Principles of Human Knowledge.'
Here he is Locke, reduced to logic. He sees that mind
and ideas end in mind and ideas, and that if, as Locke
himself had at first said, ideas are all the objects of
knowledge, then, as Locke ought to have concluded,
not unthinking body but the Divine Mind is the only
external cause.

But Berkeley's optics were superior to his psychology.
We must appeal from the ' Principles ' to the ' Theory
of Vision.' ' There is,' says he in the latter treatise, ' no
one ignorant that the pictures of external objects are
painted on the retina or fund of the eye.' l Then,
these external objects are not tangible, nor visible,
nor sensible at all, but are causes of sensible objects,
or, as Berkeley would say, of ideas. Not all objects

1 Theory, Ixxxviii.

246 PSYCHOLOGICAL IDEALISM PART ir.

of knowledge, then, are sensible ideas. Again, these
external objects, whose pictures are painted on the
retina, are not God, and yet are causes of sensible
effects. God, then, is not the sole cause. In short,
optics require, between God and our ideas, an intervening
nature. The scientific admissions of the 'Theory of
Vision ' are sufficient to destroy the pure idealism of
the ' Principles of Human Knowledge.'

Besides the optical discovery of the invisibility of
remote distance, the psychological hypothesis that we
see visible ideas, and the ontological recognition of the
existence of an extended world without the mind falsely
confused with the tangible object, the ' Theory of
Vision ' finally contains a logical speculation on the
origin from vision of our knowledge of the extended
beyond vision. Like the main thesis, this speculation
contains much that is true, and especially that we do
not see but infer an external world. It is also most
suggestive, and in fact was the first hint of the hypo
thesis that association may be an account of, or rather
a substitute for, the origin of knowledge. But it does
not in the least explain the knowledge of extended
objects in the external world, required by optics and
admitted by Berkeley. We must not be led away by
the appearance of simplicity, but keep steadily before
us the known facts to be explained, and by them test
the hypothesis.

We do not see remote distance : we do judge and
infer it from sight : this is the essential truth in Ber
keley's theory.1 The question is, how this judgment and
inference are made. He answers that when the mind
perceives any idea not immediately and of itself, it must
be by the means of some other idea : 2 distance, then,

1 Theory, iii. 3 IL. ix.

CHAP. VIII.

BERKELEY'S THEORY OF VISION 247

is suggested to the mind by the mediation of some
other idea, which is itself perceived in the act of seeing.1
He finds three ideas, which arise according to the
different distances of objects : the first, the lessening or
widening the intervals between the pupils of our eyes,
attended with a sensation ; 2 the second, the more or
less confused appearance ; the third, the prevention of
this confusion by straining the eye, with its sensation.3
These are the ideas which he thinks will suggest the
idea of the distance of the external object : not that
there is any natural or necessary connection of those
ideas with distance, but that there is an habitual or cus
tomary connection between these ideas and the idea of
distance.4

This process implies that the idea of distance itself
has been acquired in some other way. This way,
according to Berkeley, is touch combined with motion,
lie gives the whole process in the following passage :—
' Having of a long time experienced certain ideas
perceivable by touch, as distance, tangible figure, and
solidity, to have been connected with certain ideas of
sight, I do upon perceiving those ideas of sight, forth
with conclude what tangible ideas are, by the wonted
ordinary course of nature, like to follow. Looking at
an object I perceive a certain visible figure and colour,5
with some degree of faintness and other circumstances,
which from what I have formerly observed, determine
me to think, that if I advance forward so many paces
or miles, I shall be affected with such and such ideas of
touch.' G

This process, by which visible ideas suggest tangible

1 Theory, xvi. 3 Ib. xvi. 3 Ib. xxvii.

4 Cf. ib. xxviii. with the above sections.

0 Note this admission of visible figure as well as colour. to Ib. xlr.

PART ir.

248 PSYCHOLOGICAL IDEALISM

ideas of distance, is what is ordinarily called association.
Berkeley describes this operation with great clear
ness. < That one idea,' he says, « may suggest another
to the mind, it will suffice that they have been observed
to go together, without any demonstration of the ne
cessity of their co-existence, or without so much as
knowing what it is that makes them so co-exist.' J There
is such an operation, and its recognition was not new
in philosophy. Plato and Aristotle, Hobbes and Locke
were aware of it. Aristotle, for instance, says that we
recollect from what is similar -or contrary or con
tiguous.2 But what was new in Berkeley's ' Theory of
Vision ' was the hypothesis, afterwards developed by
Hume, that this process of recovering ideas is some
times the analysis of what we call knowing objects. The
ordinary man, when he uses his eyes, supposes that he
knows, nay sees, that there is an external object at a
distance from him. Berkeley tells him that he is really
letting visible ideas suggest to him tangible ideas of
(^stance ; that is all.

There is a negative value in Berkeley's analysis. It
is greatly superior to the ordinary supposition that we
see a distant object. Berkeley, though he exaggerated
when he said that we do not see any distances at all,
showed conclusively that we do not see but infer a
distant object in the external world. It is also
superior to the supposition of ' Descartes and others ' 3
that we infer the distance of objects from the angles
they make with our eyes. Berkeley disposed of this
'jmmour of making one see by geometry/ when he
'showed that the lines and angles between the external
objects and the eyes are as invisible as the external

1 Theory, xxv. 2 De Mem. 2 = 451 B, 18-20.

3 Theory, iv. not

CHAP. viir. BERKELEY'S THEOLY OF VISION 249

objects themselves.1 Mathematical opticians had fallen
into the blunder of supposing that lines and angles,
known only to themselves by science, are sensible data
which ordinary men use in vision to infer an external
world.

There is also positive information in Berkeley's
analysis. It contributed some new truths on the senses.
In the first place, about sight ; he did not indeed show
that there is no vision of space, but he did show that
it is in a way unnecessary. He called attention to the
scientific observation of the misfortune of the blind,
who have no eyes, yet feel and infer space. Again, his
remark, that no resistance is perceived by sight, con
tains the true reason why from sight alone we could
not infer an external world, and therefore must appeal
to touch and motion. Lastly, though it is not the case,
it is possible that sight might, like hearing, or, at any
rate, like language, contain no apprehension of extension,
and yet enable us, when combined with a sense of exten
sion, to infer an extended object. On the whole, he has
not shown that the visible object is not an extended
picture ; but he has shown that, whatever the visible
object is, we can know an extended object in the
external world without it, and not by it alone.

Secondly, he has the great merit of having hinted,
however imperfectly, at what is now called the muscular
sense. When he speaks of ' the motion of his body
which is perceivable by touch,' 2 though he may be ex
aggerating its connection with touch, he is recognising a
sense of motion. He also saw that in vision there is
something more than seeing. What he calls ' lessening
or widening the intervals between the pupils,' which is
' attended with a sensation,' is the convergence or diver-

1 Theory, iv., xii., lii.-liii. 2 Ib. xlv.

250 PSYCHOLOGICAL IDEALISM PART n.

gence of tlie optic axes by the internal and external
recti muscles, with their muscular sense. When he
speaks of the confused appearance caused by objects
brought too close to the eye, and of preventing
the appearances growing more confused ' by straining
the eye,' and of its sensation,1 he is pointing towards
the increase of the convexity of the crystalline lens, for
the more rapid convergence of rays from near objects to
the retina, by the action of the ciliary muscle, with its
muscular sense. He also refers to the movements of
the eyes up and down, to the right and left, which are
performed by the four recti muscles.2 In none of these
cases did he analyse the muscular movements or assign
them a distinct muscular sense. Nevertheless, he called
attention to movement, to the sense of movement in
touch, and to the sense of some kind of action in sight,
connected with the knowledge of extension.

At the same time, these great achievements are
quite consistent with equally great blunders about our
senses, of which there are two, at opposite extremes.
On the one hand, he underrates the efficacy of vision
when he tends to confine it to visible ideas of colour ;
on the other hand, he exaggerates the efficacy of touch
when he tends to extend it to external objects. Eeally,
the former is the vision of the extended in the optic
nerves, the latter is the feeling of the extended in the
tactile nerves. However, these errors do not touch the
exact question before us. Whatever else it may be,
the object of vision is certainly not the external object
at a remote distance. Now, the question is whether,
when we say that there is such an external object, cor
responding to what we see, we are only letting visible
ideas suggest tangible ideas of remote distance.

1 Theory, xxi.-xxvii. * Ib. xcvii.- xcviii.

CHAP. viii. BERKELEY'S THEORY OF VISION 251

111 solving this problem, two concessions must be at
once made. In the first place, from vision, being no
sense of resistance, we do not infer the external world
directly, but only indirectly through touch and motion.
Secondly, visible ideas do suggest tangible ideas, and
other ideas also for that matter, by the customary tie
of association, which is a real fact of human nature.
But we must also ask ourselves whether this suggestion
is all that happens. If so, we should only have the
ideas ; we should not infer that, over and above the
ideas, the object seen corresponds to an extended object
in the external world. For example, at this moment,
my vision of the white would suggest my tangible idea
of the extended ; but I should not infer, as I really do,
that over and above the tangible idea there is an ex
tended paper in an external world, corresponding
both to. the object of touch and the object of sight.
Berkeley substituted the suggestion of ideas for the
inference of external objects.

Even in the c Theory of Vision,' in spite of having
admitted the existence of the external object, and its
action on the retina, Berkeley partly accepted the
consequence. ' Sitting in my study,' he says, ' I hear
a coach drive along the street ; I look through the
casement and see it ; I walk out and enter into it ; thus
common speech wou'd incline one to think, I heard, saw,
and touch'd the same thing, to wit, the coach. It is
nevertheless certain, the ideas intermitted by each sense
are widely different, and distinct from each other ; but
having been observed constantly to go together, they
are spoken of as one and the same thing.' 1 Similarly,
he afterwards says that, though the objects are different,
as they are called by the same name, he will, to avoid

1 Theory, xlvi.

252 PSYCHOLOGICAL IDEALISM PART n.

tediousness and singularity of speech, speak of them as
belonging to one and the same thing.1

Now, it is quite true that the audible, visible, and
tangible are different objects, and also that, if nothing
happened except that ideas of the audible and visible sug
gested ideas of the tangible, no real identification could
take place. But something further does take place. In
the first place, when I hear something sounding in my
auditory, see something coloured in my optic, and feel
something hard in my tactile nerves, and have often ex
perienced these sensible objects in a similar order, I infer
that there is one external object, which is the common
original of these sensible objects on any given occasion.
Secondly, I call this common original by one and the
same name, ' coach,' because I infer it to be one thing.
It is true that I have an habitual tendency to confuse the
one external object with the several and different audi
tory, visible and tangible objects within me. But it is
not true that there is no identity but an identity of
name. There is an identical external object, the coach,
which I infer, and which I can disengage from the con
fusion with its different sensible results, by means of
science. Now, if the auditory, visible and tangible
objects had been mere ideas in my mind, and if these
ideas merely suggested one another, I could never have
inferred the one external object, and it is most im
probable that I should have even called the different
ideas by one name. But I do infer one external object,
and am justified by optics and other sciences connected
with the senses. Therefore, in the first place, the pro
cess of this inference of one external object cannot be a
mere suggestion of different ideas ; and secondly, the
data of this inference of an external object cannot be

1 Theory, Iv.

CHAP. vin. BERKELEY'S THEORY OF VISION 253

auditory, visible and tangible ideas. In reality, from
physical data in the several nerves, I infer one physical
coach, and give it, not them, one name.

In the ' Theory of Vision,' however, Berkeley did
not fully realise the consequence of reducing the infer
ence of external objects to the suggestion of tangible
ideas, because he combined this association of ideas
with an intuitive touch of external objects. Hence,
later on, he says of a man, that ' when he has by
experience learned the connexion there is between
the several ideas of sight and touch, he will be able,
by the perception he has of the situation of visible
things in respect of one another, to make a sudden
and true estimate of the situation of outward tangible
things, corresponding to them. And thus it is, he
shall perceive by sight the situation of external
objects, which do not properly fall under that sense.' l
Such an estimate would require the impossible iden
tification of tangible ideas, tangible objects, and out
ward things. But, in the first place, touch does not
feel the outward thing. Secondly, a visible idea
suggests a tangible idea, but not a tangible object.
Thirdly, what we really do is to estimate the situation
not of a tangible idea, nor of a tangible object, but of
an outward thing corresponding on the one hand to the
tangible, on the other hand to the visible, object within
ourselves. We cannot bolster up the association of
ideas by an intuitive touch of outward things.

In the ' Principles,' when he had retracted the con
fusion of the external and the tangible, and the intuitive
touch of the external, the consequence of supposing
that the inference from vision is nothing but an associa
tion of ideas came out in its simple nakedness. He

1 Ib. xcix.

254 PSYCHOLOGICAL IDEALISM TAKT n.

then saw that, in this case, ideas of sight only admonish
us what ideas of touch will be imprinted on our minds,
and do not mark out to us things actually existing at a
distance.1 I freely admit that Berkeley was right in
retracting the tangible intuition of the external world,
and that if we start with visible ideas, and by the
suggestion of ideas let these visible suggest tangible
ideas, and have no tangible intuition of extended
objects in the external world, we shall begin and end
in ideas. But we do not end in ideas. His own optics
require that we know external objects, and that no one
is ignorant of their painting pictures on the retina of
the eye. His hypothesis of the suggestion of ideas does
not account for the knowledge of their causes. It is,
therefore, false.

The cause of Berkeley's error was that neglect of
logical inference which made its appearance in the
Second Book of Locke's ' Essay,' and led to the postpone
ment of reasoning to all kinds of lesser powers. Like
Locke, Berkeley was aware of the difference between
association and reasoning.2 But, like Locke, he kept in
the background, and to the last, reasoning, the one power
which will be heard and will not wait. Hence, in the
' Principles,' he supposes that ideas suggest ideas, until
reason at last infers a God. Hence, in the ' Theory of
Vision,' he substitutes for inference a false touch of out
ward things and an imperfect suggestion by visible of
tangible ideas. He overlooks in both the human, though
complex, inference of an external extended object
which causes both sight and touch.

The ' Theory of Vision ' contains the discoveries of

1 Princ. xliv., quoted at the beginning of this chapter.

2 Cf. Locke, Essay, ii. 33, 13 ; Berkeley, The Theory of Visual
Language Vindicated, § 42.

CHAP. vin. BERKELEY'S THEORY OF VISION 255

the invisibility of remote distance, and of the combina
tion of sight and touch with a sense of motion. It is
a very good answer to those who say that we see the
external world ; though even they could retort on
Berkeley that he says himself that we feel it. It is no
answer to those who say that we know the external
world. It is a good answer to those who say that we
infer it directly from sight by lines and angles, or by
any other direct inference, from sight, which feels no re -
sistance. But it is no answer to those who say that we
infer an external extended world first from the resistance
felt in the senses of touch and motion, and then from
the correspondence in extension between inferred, tan
gible, and visible objects. Finally, Berkeley's ' Theory
of Vision ' contains two fundamental errors of omis
sion : the first, that there is no vision of an extended
object within ; the second, no inference of an extended
object without, common to our senses of sight and
touch.

256 PSYCHOLOGICAL IDEALISM PART n.

CHAPTEE IX.

HUME.

THE academical or sceptical philosophy of Hume admits
of being summarised as follows. All the perceptions
of the mind are impressions and thoughts or ideas.1
All our ideas or more feeble perceptions are copies of
our impressions or more lively ones.2 Association is a
principle of * connection which, by resemblance, con
tiguity or causation, on the appearance of a perception
suggests thoughts or ideas.3 All the objects of human
reason or inquiry may naturally be divided into two
kinds, to wit, Relations of Ideas and Matters of Fact. Of
the former kind are the sciences of geometry, algebra,
and arithmetic, and, in short, every affirmation which
is either intuitively or demonstratively certain.4 The
origin of our beliefs, i.e. vivid ideas, of matters of
fact, is experience of a constant conjunction of impres
sions, and association which, from this constant con
junction, begets such a connection in the imagination
that, on the appearance of the antecedent, we have the
idea, i.e. belief, of the consequent and of their connection
as cause and effect.5 The mind lias never anything
present to it but the perceptions, and cannot possibly
reach any experience of their connection with objects.
The supposition of such a connection is, therefore, with-

1 Inquiry, § 2. 2 Ib. 3 Ib. § 3.

4 Ib. § 4. 5 Ib. §§ 4-7, esp. § 7, Part II.

CHAP. ix. HUME 257

out any foundation in reasoning.1 The great subverter
of Pyrrhonism, or the excessive principles of scepticism,
is action.2 There is, indeed, a more mitigated scepticism,
or academical philosophy, which may be both durable
and useful, and which may, in part, be the result of this
Pyrrhonism or excessive scepticism, when its undistin
guished doubts are, in some measure, corrected by
common sense and reflection.3

The point of this academical philosophy is that man
has the faculties to receive impressions and conceive
ideas or thoughts ; and by association to make vivid
ideas of causation, which are his only beliefs on
matters of fact ; but not by reasoning to infer exter
nal objects. Plume published it twice over, first in the
'Treatise of Human Nature,' afterwards in the 'Inquiry
concerning Human Understanding.' The earlier work is
more elaborate, containing in the First Part a fuller
discussion of the origin of ideas, modelled on the Second
Book of Locke's ' Essay,' but writh the stress laid on asso
ciation ; in the Second Part, a theory of the apprehen
sions of time and space, which hardly appears at all in
the later work ; in the Third Part, a longer but less
elegant exposition of his theory of association as the
origin of the belief in causation ; and in the Fourth
Part, a long discussion of the apprehension of objects,
answering to the last section of the ' Inquiry,' but com
prising a sceptical theory of substances, both material
and thinking, which he afterwards omitted but by no
means retracted in his later work. Since the ' Treatise'
was published when the author was a young man of
twenty-seven, the ' Inquiry ' ten years later in the prime
of life, the impartial critic must dwell mainly on the more

1 Inquiry, § 12, Part I. ~ Ib. § 12, Part II.

3 Ib. § 12, Part III.

258 PSYCHOLOGICAL IDEALISM TART n.

mature work ; especially as in his account of ' My own
Life ' Hume says himself, ' I had been guilty of a very
usual indiscretion in going to the press too early ' — a
useful warning to youthful philosophers. Nevertheless,
the essence of both ' Treatise' and ' Inquiry ' is the same :
it is a reduction of man to mere perceptions. Berkeley
had attacked natural science : it remained for Hume to
attack the human intellect. But we must take care not
to be argued out of our wits.

Hume's philosophy is founded on the following
distinction of perceptions into impressions and ideas,
which he identifies with thoughts :—

' Here, therefore, we may divide all the perceptions
of the mind into two classes or species, which are
distinguished by their different degrees of force and
vivacity. The less forcible and lively are commonly
denominated thoughts or ideas. The other species want
a name in our language, and in most others ; I suppose,
because it was not requisite for any, but philosophical
purposes, to rank them under a general term or appella
tion. Let us, therefore, use a little freedom and call
them impressions ; employing that word in a sense some
what different from the usual. By the term impres
sions, then, I mean all our more lively perceptions,
where we hear, or see, or feel, or love, or hate, or
desire, or will. And impressions are distinguished from
ideas, which are the less lively perceptions, of which we
are conscious, when we reflect on any of those sensa
tions or movements above mentioned.' l

The distinction between impressions and ideas is an
important discovery, or rather re-discovery. Aristotle
had, in the ' De Anima,' carefully distinguished between
cesthemata, or the objects in sense when an external ob
ject is present, and phantasmata, or their relics in the

1 Inquiry, § 2.

CHAP. ix. HUME 259

imagination when the external object is absent. But,
as we have already seen, Descartes afterwards confused
the object of sensation and conception under the name
' idea,' and Locke and Berkeley had followed him. So
long as it was admitted that some external object is
also known, this confusion had no very serious conse
quences ; for the reduction of sense to a purely psy
chical object at all was a far more fundamental error
than the reduction of this psychical object to an idea.
But when it began to be doubted whether any external
object could be known, it then became a serious ques
tion, how we can distinguish an adventitious idea im
printed on the senses from a fictitious idea generated by
the imagination.

Berkeley felt this difficulty,1 and got over it partly
by supposing that adventitious are more vivid than
factitious ideas, but mainly by his theory that the
former are directly inspired by God. Now, Hume
doubted our knowledge of any cause of our per-
ceptions^ natural or spiritual. Moreover, he saw that
' the word idea seems to be commonly taken in a very
loose sense by Locke and others, as standing for any of
our perceptions, our sensations and passions, as well as
thoughts.' 2 In these circumstances, he revived the
ancient distinction of cesthema and phantasma under
the new names ' impression ' and ' idea,' yet without
resorting either to matter or to God. As he says in the
' Treatise,' ' By the term of " impression," I would not be
understood to express the manner in which our lively
ideas are produced in the soul, but merely the percep
tions themselves.' 3 Consequently he had to look out for
some fresh criterion to distinguish the thing as well as
the term, and found it in the liveliness of an impres-

1 Princ. xxxiii. ~ Inquiry, § 2, note. 3 Treatise, i. § 1, note.

s 2

260 PSYCHOLOGICAL IDEALISM TART n.

sion as contrasted with an idea. In the 'Bhetoric,'
Aristotle had described, without meaning to define,
imagination as a kind of weak sense.1 Hobbes had
exalted this description into a definition — ' Imagination
being (to define it) conception remaining, and by little
and little decaying from and after the Act of Sense! 2
Berkeley had made faintness a partial test of an idea of
imagination : Hume exalted it into the sole criterion,
and committed himself to the consequences. ' The
most lively thought,' says he, ' is still inferior to the
dullest sensation.'

The hypothesis that there is no more distinction
between sense and imagination than between vivid and
faint perceptions, or states of consciousness, as they
now call them, has become a favourite with modern
idealists, simply because they have destroyed the real
criterion afforded by the presence and absence of exter
nal objects. But there is a difference in kind between
sense and imagination, of which ' different degrees of
force and vivacity' furnish no adequate criterion. The
faintest impressions would be undistinguishable from
the most vivid ideas. This objection Hume had noticed
himself in the ' Treatise,' and tried to evade it :—

• ' The common degrees of these are easily distin
guished : though it is not impossible but, in particular
instances, they may very nearly approach to each other.
Thus, in sleep, in -a fever, in madness, or in any very
violent emotions of soul, our ideas may approach to our
impressions : as, on the other hand, it sometimes hap
pens that our impressions are so faint and low, that we
cannot distinguish them from our ideas. But, notwith
standing this near resemblance in a few instances, they

1 Ar. Ehet. i. 11 = 1370 A, 28.

2 Hobbes, Human Nature, chap. iii. § 1.

CHAP. ix. HUME 261

are in general so very different, that no one can make a
scruple to rank them under distinct heads, and assign
to each a peculiar name to mark the difference.' *

The conclusion of this passage exhibits a common
practice of trying to get round a contradictory instance.
It is true that, on the whole, the livelier would be dis
tinct from the fainter perceptions, but there would still
be a margin between the lively and the faint, which, in
the absence of any other criterion, it would be arbitrary
to place among either impressions or ideas. But there
is a still more fatal objection : some ideas are livelier
than impressions, and would have, by the bare criterion
of lively and faint, to change places with them. Yet
Hume, to save his theory, has to say that c all ideas,
especially abstract ones, are naturally faint and ob
scure,' while ' all sensations, either outward or inward,
are strong and vivid.' 2 But abstract ideas of mathe
maticians are often brighter tfrarftheir concrete impres
sions, as in the case of the mathematician who in a fit
of abstraction held the egg in his hand while he boiled
his watch. Ideas of men in disease are often so vivid
as to be mistaken for impressions. The artistic imagina
tion is sometimes stronger than ordinary sensation, as
Handel, on being asked how he wrote the 'Hallelujah
Chorus,' said, ' I did see all heaven before me, and the
great God Himself.'

This superior vividness of imagination is finely de
scribed by Addison :—

' Words, when well chosen, have so great a force in
them, that a description often gives us more lively ideas
than the sight of things themselves. The reader finds a
scene drawn in stronger colours, and painted more to
the life in his imagination, by the help of words, than

1 Treatise, i. § 1. '2 Inquiry, § 2.

202 PSYCHOLOGICAL IDEALISM J>ART FT.

by an actual survey of the scene which they describe.
In this case the poet seems to get the better of nature ;
he takes, indeed, the landskip after her, but gives it
more vigorous touches, heightens its beauty, and so en
livens the whole piece, that the images which flow from
the objects themselves appear weak and faint, in com
parison of those that come from the expressions.' l

Hume, in distinguishing impressions and ideas,
rightly restored the word ' idea ' to its original sense,
from which Descartes had perverted it in making it
stand for all our perceptions. But he tried to put new
wine into old bottles. The Aristotelian distinction of
impressions and ideas does not accord with Hume's
distinctions of vivid and faint perceptions, and can only
be explained by the Aristotelian criterion of the pre
sence and absence of an external object, which was
repugnant to Hume's philosophy. Sensation is the ap
prehension of an object presented to the senses repre
senting an external object; while hallucination, or
' subjective sensation,' is a similar apprehension pro
duced by pressure on a sensory nerve : imagination is
the apprehension of an idea representing a sensible
object or something similar to it and inferred from it.
But the presentations of sense are often less vivid than
the afterthoughts of fancy.

Hume's second point is the empirical doctrine that
impressions are the originals of all our ideas, which, as
before, he identifies with our thoughts : —

'But though our thought seems to possess this
unbounded liberty, we shall find, upon a nearer exami
nation, that it is really confined within very narrow
limits, and that all this creative power of the mind
amounts to no more than the faculty of compounding,

1 Spectator, No. 416.

CHAP. IJ.

HUME 2G3

transposing, augmenting, or diminishing the materials
afforded us by the senses and experience. When we
think of a golden mountain, we only join two consistent
ideas, gold and mountain, with which we were formerly
acquainted. A virtuous horse we can conceive ; be
cause, from our own feeling, we can conceive virtue ;
and this we may unite to the figure and shape of a
horse, which is an animal familiar to us. In short, all
the materials of thinking are derived either from our
outward or inward sentiments : the mixture and com
position of these belongs alone to the mind and will :
or, to express myself in philosophical language, all our
ideas *or more feeble perceptions are copies of our im
pressions or more lively ones.' l

The conclusion of this passage is a neat summary of
the argument in the Second Book of Locke's ' Essay.'
Sense is the source of ideas, however indirect the process
of their formation. Locke had disposed of innate ideas.2
As Hume puts it, ' all our impressions are innate, and our
ideas not innate,' 3 meaning that the former are intuitive
and the latter derivative. But when we are told that
all our ideas are copies, direct or indirect, of our im
pressions, several questions present themselves. First,
what are those impressions which have to be the ori
ginals of all ideas ? Secondly, what are the processes
which enable us to copy the original impressions ?
Thirdly, what are the ideas and thoughts which we are
able to reach ? We shall have to ask ourselves about our
impressions, our processes, our ideas, and our thoughts.

What are impressions ? It is surprising how little
Hume condescends to tell us on this subject, incom
parably the most important in his philosophy. In the
'Treatise' he says that ' the examination of our sensations

1 Inquiry, § 2. ' Cf. Treatise, iii. § 14. 3 Inquiry, § 2. note.

264 PSYCHOLOGICAL IDEALISM PART n.

belongs more to anatomists and natural philosophers
than to moral ; and, therefore, shall not at present be
entered upon.' l It never is entered upon in the
' Treatise ' ; and in the ' Inquiry ' all that is said on the sub
ject is that impressions are more vivid than ideas, that
they are the origin of ideas, that there are impressions
of sensation and impressions of reflection, and that in
all cases the mind has never anything present to it
but perceptions, which are either impressions or ideas.
We are left to gather that the sensible object of the
impression, being a mere quality not distinct from the
operation, is, in short, the impression itself. Bare ab
stract impressions are the data of Hume's empiricism.
j'But it is one thing to admit that knowledge begins with
\sense, another to assume that it begins with a sense of
1 impressions.

When we reflect that these impressions are deliber
ately stated to be the premises of all our conclusions by
a philosopher who truly says that 4 one mistake is the
necessary parent of another,' the omission of an exami
nation of sensation strikes us with the greatest surprise.
But when we consider that all the idealists have taken
their data of sense with the same coolness, the wonder
ceases. As, to begin with, Descartes had attempted no
formal proof that a soul must perceive ideas, while
Locke and Berkeley simply accepted the hypothesis
that it does perceive ideas, so Hume assumed that the
mind has never anything present to it but perceptions,
and so, after him, Kant begged that the matter of sense,
and Mill begged that the information which the senses
give us concerning objects, is our sensations.2

1 Treatise, i. § 2.

2 Of. Hume, Inquiry, § 12 ; Kant, Critique (ed. Hartenstein, pp. 83,
55-6; Meiklejohn's translation, Bohn), pp. 1, 21; Mill, Examination of
Hamilton's Philosophy, chap. ii.

CHAP. IX.

HUME 2G5

111 order to correct Hume's theory of impressions,
and his followers' theory of sensation, it is necessary to
repeat, what we have already proved, that sense appre
hends neither itself nor abstractions. It is, in the first
place, always an operation of a subject apprehending
an object, internal, but not identical with the operation.
Secondly, its object is always a qualified substance,
internal but not resolvable into abstract qualities.
Thirdly, when it is outer sense, sensation, and sensi
tive observation, the substance apprehended is the ner
vous substance sensibly qualified in different parts as
coloured, heated, &c. Fourthly, when it is inner sense,
consciousness and conscious reflection, the substance
apprehended is the thinking subject, body and soul.
An impression without a substantial subject and object
is an abstraction, never perceived, never known, with
difficulty made an object of attention. A man is a sub
stantial subject impressed with a substantial object, and
can be conscious of himself being so impressed, as well
as conceiving, reasoning, and performing other opera
tions.

Hume's theory of impressions, when corrected by
being converted from the abstract into the concrete,
contains the valuable point, too often neglected, that,
even without judgment, a man's simple sensation of the
white, or the hot, is a beginning of knowledge, and no
mere abstraction. Nay, as we saw in the first part of
this essay, we can even trace knowledge to a still simpler
origin. I begin to know when I feel pained or pleased ;
not, be it remembered, when there is pain or pleasure,
which are afterthoughts. My first act of knowledge is
having a simple feeling in the concrete : my second act
of knowledge is having a simple sensation of a sensible
obiect in the concrete.

PSYCHOLOGICAL IDEALISM PART n.

There are two ways of criticising Hume's theory of
impressions. The wrong criticism is to accept it as a
complete account of sensation, conclude that pure sensa
tion is an abstraction which never occurs in conscious
ness, and yet assume these very abstractions to be the
elements of a psychological synthesis.1 Those who take
this view are too much tarred with the brush of Hume.
A pure sensation, or impression in Hume's terminology,
is an abstraction ; but so far from being an element of
knowledge, it is a subsequent result of concentration on
the mere operation of knowing, to the neglect of sub
ject and object, and is only put for the real elements of
knowledge by a convenient form of speech. The right
criticism is to point out that Hume substituted the after-
abstraction of sensation for the data of sense and the
elements of knowledge, which are always a substantial
subject sensibly perceiving substantial objects within
the nervous system and consciously perceiving himself.
There is no such a thing in rerum naturd as an im
pression and a consciousness, which are merely ab
stracted post rem, but there is such a thing as man
impressed with an object and man conscious of himself.
Sensation and consciousness, in this concrete shape, are,
moreover, not only the real elements of knowledge, but
are themselves knowledge ; for, as Aristotle saw, though
sense is not science (eVioTT?/^), on the other hand it is
knowledge (y^oicris).

Next, we come to the association of ideas, thus de
scribed by Hume :—

' It is evident that there is a principle of connection
between the different thoughts or ideas of the mind, and
that, in their appearance to the memory or imagination,

1 Cf. Wundt, Pliysiologisclie Psychologic, ii. 196.

CHAP. IX.

HUME 2G7

they introduce each other with a certain degree of
method and regularity.' 1

He also assigned three principles of connection among
ideas, namely, resemblance, contiguity in time or place,
and cause or effect ; for example, a picture leads our
thoughts to the original ; the mention of one apartment
in a building suggests a discourse concerning the others ;
and if we think of a wound, we can scarcely forbear
reflecting on the pain which follows it.2 Such is the
association of ideas on these three principles ; a pro
cess, which Hume did not exactly substitute for all
reasoning, for he regarded mathematics as intuitive and
demonstrative, and founded morals on sympathy and
reason ; but which he did make the substitute for
reason in all matters of causation, the organ of natural
philosophy, and, after sense, the main origin of our
ideas, which in the passage just quoted are, for the
third time, identified with our thoughts.

Association of ideas is a vera causa : this is the
great advantage of Hume over Kant. We are conscious
that when we have been sensible of two objects together,
and have a sensation or idea of one, we, in consequence,
have the idea of the other: we are not conscious that
we have an a priori idea, or any other apprehension of
an object which we have not apprehended beforehand,
either immediately or mediately. The simplest way, in
which this conscious suggestion of ideas acts, is when

Oo

simple feelings or sensations and their ideas introduce the
ideas of one another. That is not a bad instance given
by Hume ; the idea of a wound suggests the idea of
pain. The conditions of such a simple sensitive associa
tion, as we may name it, are, first, simple feelings or
sensations occurring together ; secondly, their being

* Inquiry, § 3. 2 Ib.

268 PSYCHOLOGICAL IDEALISM PART 11.

repeated together ; thirdly, imagination, or the power of
conceiving ideas; fourthly, the appearance of one of
the feelings or sensations, or of one of their ideas ; and
finally, the association itself, which consists in the con
sequent introduction of the idea of the other feeling or
sensible object.

A question may, indeed, be raised, which is evaded
by Hume and by many of his disciples. Need we only
have had the original impressions together, or must
we have also perceived their relation, or the relation
of their objects? For example, must we have been
sensible not only of the wound and of the pain, but also
that the wound was cause of the pain, before it would
suggest the idea of the pain ? Hume, starting as he
did from simple impressions, would no doubt have
answered this question in the negative. He would pro
bably have been right ; and, moreover, he was aware of
the reason, which is the anatomical connection of the
parts of the nervous system,1 though how, if everything
we know is perceptions, he could know about the nervous
system, which is the material cause of these perceptions,
he did not vouchsafe to explain. The unconscious con
nection of nervous centres, e.g. between those which
control the contraction of the iris and those which act
on the ciliary muscle to increase the convexity of the
crystalline lens in the eye, the facts of automatic action
in general, and those of unconscious cerebration in par
ticular, make it exceedingly probable, notwithstanding
the difficulty of isolating such a fact in consciousness,
that when a connection has been set up between ner
vous centres, through two simple sensations repeatedly
occurring together, then the occurrence of one will, by
an association founded on the nervous connection, in-

1 Treatise, ii. § 5 ; cf. Locke, Essay, ii. 33, 6.

x. HUME 1^09

troduce the idea of the other, without our having ever
perceived that the two sensations or their sensible
objects were connected. In accordance also with that
gradation of animal faculties first noticed by Aristotle,
it is not at all improbable that some animals, which
have got beyond feeling and simple sensation to the
phantasy of imagination, may possess this simple sensi
tive association of ideas, which also, through the power
of ideas over passions and passions over movements,
may be the guide of their lives.

It is a very different question how far simple sensa
tion, ideation, and association would carry a man on the
path of rational life. All association of ideas is an act
of reproductive imagination. It merely reproduces the
idea of something already known somehow or other.
We shall find, in the sequel, that as knowledge widens
association widens with it. But at present we are deal
ing with the simplest kind of association from simple
sensations, which is also the only kind which Hume
formally recognises. Now, his doctrine is that when we
have had simple impressions together, ideas are their
copies, and association introduces these ideas by the
laws of connection among the impressions. This can
only mean that association reproduces the idea of an
object already sensible ; for example, if having been
hurt I felt pain, being hurt again will reproduce the
idea of the previously felt pain.

An idea of a previously felt pain is quite a different
thing from the idea of a similar pain not yet felt ; the
former represents a previous impression, the latter a
future impression ; the former is an object of simple
reproductive, the latter of simple productive imagina
tion. Now, simple sensitive association reproduces in
the present the idea of a particular pain already actually

270 PSYCHOLOGICAL IDEALISM PART n.

felt ill the past ; but neither Hume nor any of his dis
ciples has shown that it will perform the very different
operation of producing the similar but new idea of a
similar but new pain to be possibly felt in the future.
j Here is the limit of association : it always reproduces
the idea of something already apprehended ; never pro
duces the idea of something not yet apprehended. I
have been hurt and felt pain : I am now hurt ; I imagine
the ideas of previous pain when I was hurt ; I also
imagine the idea of a pain now to follow the hurt, but
not yet felt. Association reproduces the former ideas ;
and why ? Because the particular impressions, of
which the particular ideas are copies, have occurred
together. But this reproductive act will not of itself
produce the latter idea, the impression of which has
never occurred at all. Association from simple im
pression reproduces particular ideas of particular ob
jects previously sensible : it does not produce a particu
lar idea even of a single particular object, not yet sen
sible ; a fortiori, it is powerless to generate a general
idea.

Hume's critic must constantly keep before him the
question — Is all the reasoning of the natural philosopher
nothing but a reproduction of sensible ideas by simple
association ? This question brings us to Hume's main
point, that while all mathematical reasoning is a process
from intuition through demonstration to relations of
ideas, all reasoning about facts is a process from ex
perience of constant conjunction of impressions through
association to ideas, i.e. beliefs, of cause and effect. The
discussion of this complicated theory compels the con
sideration of many points : belief or judgment and reason
ing or inference, intuition and demonstration, causation
and our knowledge of cause and effect. Judgment and

CHAP. ix. HUME 271

reasoning alone almost require a logic. I propose to
confine myself to these, leaving the remaining points
for subsequent discussion. We must not leave the most
precious of all man's gifts to be stolen from him without
striking a blow. What was wanted, and is still a de
sideratum, is, not a Critique of Pure Eeason, but a
Vindication of Logical Reasoning.

' Nature,' says Hume, ' by an absolute and uncon
trollable necessity, has determined us to judge, as well
as breathe and feel.' l He saw the importance of judg
ment or belief. He also set himself, both in the ' Treatise'
and in the ' Inquiry,' 2 to ' examine more accurately the
nature of this belief.' But in both cases he adopted the
same extraordinary paradox, that, as impression is only a
more vivid perception than an idea, so a belief is only a
more vivid idea than a fiction. ' I say that belief is no
thing but a more vivid, lively, forcible, firm, steady con
ception of an object than what the imagination alone is
ever able to attain ' : such is Hume's definition of belief.3

Why did he fall into this extraordinary confusion of
simple and complex apprehension, of conception and
judgment, of an idea and a belief? Because, wanting
to derive all beliefs from association, and being aware
that association terminates in ideas, there was nothing
for it but to reduce beliefs to vivid ideas. He cut his
coat according to his cloth — in a thoroughly iderjhstic
style of tailoring. ' Whenever,' he says, ' any object
is presented to the memory or senses, i^Jmmediately.
by the force of custom, carries the imagination to con
ceive that object which is usually conjoined to it ; and
this conception is attended with a feeling or^sentiment
different from the loose reveries of the fancy. In this

1 Treatise, iv. § 1. 2 Ib. iii. § 7 ; Inquiry, § 5.

3 Inquiry, § 5, Part II.

272 PSYCHOLOGICAL IDEALISM TART n.

consists the whole nature of belief. ' l All he really
shows is that, if association is the origin, of beliefs, they
would be mere ideas ; but he does not prove^that they
are so. It was to prepare for this confusion of concep
tion and belief, that he had said first tjb^all percep-
tipns__axfi impressions and ideas or thoughts, thatT'all
ideas or thoughts are copies of impressions, and that
association is a principle of connection of ideas or
thoughts ; as if all thought were ideas and ideas our
only thoughts !

Judgment, or belief, is, like conception, an appre
hension, but not like it in being the apprehension of an
idea ; nor can any accumulation of epithets added to a
conception make it a judgment. Judgment is the
apprehension of a relation. Hume entirely missed this
point, by which judgment is differentiated from all con
ception whatever. He was, no doubt, much deceived
by the conceptualistic theory of relation in Locke's
' Essay.' But Locke, though he had confused relations
with their ideas in the Second Book, changed his key
when he came to the Fourth Book, and regarded judg
ments as perceptions of the agreement and disagreement
of ideas, without resolving these relations into ideas, as
strict consistency would have demanded. It should be
noticed that this differentia of judgment holds even
when the things related are ideas ; when I judge that a
dragon is a serpent breathing flame, I have only ideas of
a dragon and of such a fiery serpent, but I judge that
the ideas actually have the relation of identity, which I
can express in a proposition by the copula, is. Locke,
then, might at least have taught Hume thajL .ajudgment
perceives, not mere ideas, but^ the_agreement _and dis
agreement^ of ideas. But, as usual, the Second Book

1 Inquiry, § 5, Part II.

CHAP. ix. HUME 273

attracted philosophers to the neglect of the Fourth
Book of Locke's ' Essay.'

But though Locke's theory of judgment is wider
than Hume's, it is not adequate ; not all judgments
apprehend relations of ideas ; for some judgments appre
hend relations of sensible objects. This point has been
excellently made by Mill, the logician of the school of
Locke and Hume, in his ' Logic,' when he says that

has for itsjsubject the facts

_

themselves, though a previous mental conception of the
facts is au indispensable condition. When I say that
fire causes heaT,~ do I mean that my idea of fire causes
my idea of heat ? No ; I mean that the natural pheno
menon, fire, causes the natural phenomenon, heat.
When I mean to assert anything respecting ideas, I give
them their proper name ; I call them ideas ; as when I
say that a child's idea of a battle is unlike the reality.' l
Hence, Mill's ' Logic' recognises judgments of relations
between phenomena as well as between ideas, from
which the founder of the modern distinction of im
pressions and ideas could hardly have escaped. Not
that even Mill's analysis is adequate. In the first place,
Mill's list of judgments is incomplete; there are judg
ments of relations between sensible objects, between
ideas, between insensible, and between imperceptible
objects, judgments of sense, of conception, of inferential
perception, and of transcendental inference ; secondly,
even if, so far as judgments are premises, conceptions
may be their conditions, so far as they are conclusions,
the judgment is often the condition of the conception,
as when we infer a corpuscle, and then conceive it.
But for our present purpose it is sufficient that, as Mill
sa}Ts, there are judgments of relations between phaeno-

1 Mil], Logic, i. 5, 1 ; cf. also Exam, of Hamilton's Phil. chap, xviii.

T

274 PSYCHOLOGICAL IDEALISM PART n.

mena, sensible objects, the objects of simple impressions.
Such a judgment is not an apprehension of a relation
of ideas, much less an idea of a relation.

It would not be difficult to distinguish judgments
from ideas by Hume's admissions. In the first place,
he recognises relations, reducing them in the ' Treatise ' 1
to seven general heads : resemblance, identity, those of
space and time, quantity or number, degrees of quality,
contrariety, and cause or effect ; and, in both ' Treatise '
and 'Inquiry,' admitting the relations of resemblance,
contiguity, and sequence of impressions, on which asso
ciation is founded. There are, then, relations to be
judged. Secondly, he held that mathematics are con
cerned with relations of ideas, and unwarily admits
Locke's doctrine that a mathematical proposition ex
presses a relation.2 Thirdly, and most curiously, in the
very chapter in which he had defined belief as an idea
made vivid by association, he goes on to allow that,
when a picture introduces the idea of a friend, the
association presupposes a belief in the friend's exist
ence. ' We may observe,' says he, ' that in these pheno
mena the belief of the correlative object is always
presupposed ; without which the relation would have
no effect. The influence of the picture supposes that we
believe our friend to have once existed.' What can this
belief in a friend's existence be, according to Hume,
except Mill's apprehension of a relation of phenomena
or impressions? And are not there such beliefs, not
only of existence, but also of the other relations of
phenomena, mentioned by Mill — coexistence, sequence,
resemblance, even if not of causation ?

Hume would, perhaps, reply that we have a vivid
idea of the relation between our friend and his exist-

1 Treatise, i. § 5. 2 Inquiry, § 4.

CHAP. ix. HUME 275

ence, in which the belief consists. It is true that we
have such an idea. But, in truth and in Hume's philo
sophy, this idea must be copied from an impression ;
there must, therefore, be a prior impression of the re
lation between our friend and his existence, in which
the belief consists. We first judge of a relation and
then conceive the idea of it, in consequence of the
judgment — an important source, by the way, of complex
ideas of relation.

If the judgment were merged in the mere idea of
the relation, it could not be distinguished from fictitious
ideas of relation. Hume, indeed, tries to distinguish
' ideas of the judgment ' from fictions of the imagination
by his usual criterion of vivacity, contending, for in
stance, that ' these ideas take faster hold of my mind
than ideas of an enchanted castle.' l This may be true
ofWalpole's ' Castle of Otranto.' But Hume wrote before
the appearance of Scott's historical romances, after
which he could not have failed to see that the mere
idea of a relation in belief is often very inferior in
vivacity to the idea of a relation in fiction. There are
few scenes in history so vividly painted in my imagina
tion as my idea of Quentin Durward conducting the
Countess Isabelle out of France to Liege, and from
Liege into Burgundy. But, in spite of the force of the
idea, I do not believe in the relation, and why not?
Because I do not judge or apprehend that the relation
ever occurred, and only conceive the idea of it in
imagination, stimulated by the genius of Sir Walter
Scott.

Belief, then, is not conception : a judgment is not
an idea, but the apprehension of a relation. Now, what
is the relation of judgments to association? According

1 Inquiry, § 5 ; cf. Treatise, iii. § 7.

T 2

276 PSYCHOLOGICAL IDEALISM PART IT.

to Hume, all of them are its results ; only, however, if
they are nothing but ideas, because the association of
ideas terminates in ideas. Well, as no judgment is an
idea, not one judgment, not even one which apprehends
a relation of ideas, is a result of association, which
never can give an apprehension of a relation. Secondly,
a judgment which apprehends a relation of sensible ob
jects, such as, I am hurt and feel pained, cannot be an
effect of association, because the judgment signifies, in
Hume's language, a relation of two impressions, wThile
in association, even when that which suggests is an im
pression, that which is suggested is an idea, and because
the judgment is prior to any association in which one
of the two impressions suggests the idea of the other.
Thirdly, those judgments which apprehend relations of
objects not now in sense are not results of association,
because they are not ideas, and are not concerned with
ideas.

Hence association is not an adequate origin of
memory and expectation, which are judgments of the
past and of the future. Memory, according to Hume, is
a more vivid idea. But ideas of fancy are often more
vivid than those of memory. Memory contains an
idea, but it is a judgment that the idea represents a
previously apprehended object. Now association can
reproduce the idea, but not produce the judgment of
memory. Still less is expectation a result of associa
tion. It contains an idea, but is a judgment that the
object will or may be apprehended. When the idea
represents an object already apprehended, as in the
case of memory, association reproduces the idea, but
does not produce the judgment of the object expected.
When, as in expectation, the idea represents an object
similar to previously apprehended objects, but not itself

CHAP. IX.

HUME 277

yet apprehended, association does not even produce the
idea of the expected object ; for, as we found before,
association only reproduces ideas. History and predic
tion are not results of association, because they consist
of judgments, because their objects have never been in
sense, because their ideas are ideas of insensible objects.
A fortiori, science, an apprehension of laws, or similar
relations between an indefinite number of insensible
objects, cannot be a result of association. The associ
ation of ideas could not make us conceive the idea,
much less judge that the cities of the plain once existed,
which we never saw ; nor that the earth will one day
be too cold for habitation, when we shall not be alive
to see that day ; nor that all fluids propagate their
motions equally in all directions, which we judge to
be universal, but cannot perceive, nor conceive univer
sally. The association of ideas does not produce the
judgments of history, prediction, and science. In short,
judgments are apprehensions of present relations in
objects of sense, of past relations in memory and his
tory, of future relations in expectation and prediction,
of universal relations in science, which, not being ideas,
are not results of association, but of sense and reasoning.
Association of ideas reproduces an idea : it does not
produce an idea : it neither produces nor reproduces a
belief. How, then, do we get these beliefs or judg
ments? That is the whole question. How does judg
ment apprehend present relations of objects in sense ?
That is the first and fundamental question, never faced
by Hume. I have admitted that, when his abstractions
have been interpreted, he was right in saying that we
have simple impressions in the sense of simply feeling
pleased and pained, simply perceiving sensible objects,
the white, the hot, &c., and simply being conscious of

278 PSYCHOLOGICAL IDEALISM TAKT IT.

ourselves operating, feeling, seeing, touching, &c. These
are simple acts of knowledge. A simple sensation re
quires, indeed, a subject apprehending an object, and
must not be resolved into a mere abstraction. It does
not follow that it contains a conscious judgment of the
relation of subject and object, as some philosophers
suppose. It requires also to be different from other
sensations, in accordance with the principle of Hobbes—
' Idem sentire et non sentire ad idem recidunt.' It does
not follow that it contains a sense of discrimination.
When light is presented to my retina, by simple sensa
tion, I see a visible object in my optic nerve, without
judging its relation to myself, or to other sensible objects,
and so far as I see it, know it. But though I have
simple sensations or, as Hume calls them, impressions,
without judgment, yet I also judge of the relations of
sensible objects. Hume rightly recognised simple sensa
tion, wrongly ignored sensitive judgment.

The source of sensitive judgment is synthetic sense.
Unless I actually had a sense of the succession of being
hurt and being pained I could not judge that the
succession occurred. Moreover, there is a synthetic
sense of other relations, on which sensitive judgment
is founded, of the relations enumerated in Hume's
6 Treatise,' of the relations regarded in his ' Inquiry ' as
necessary to association, of the relations truly regarded
by Mill in his ' Logic ' as part of the very import of a
judgment and a proposition. Hume should have dis
tinguished two kinds of impressions, simple and syn
thetic ; impressions of sensible objects and impressions
of relations of objects. 'Like simple tastes and smells,
or feelings of pleasure and pain,' as Professor Huxley
says, ' they are ultimate irresolvable facts of conscious
experience ; and, if we follow the principle of Hume's

CHAP. IX.

HUME 279

nomenclature, they must be called impressions of rela
tion. But it must be remembered that they differ from
the other impressions, as requiring the pre-existence of
at least two of the latter.' l In this way, when two sen
sible objects are presented to us, we are sensible of their
succession, their coexistence, their similarity, and so
forth.

The first point to notice about this sense of a rela
tion is, that as the sensible objects, so the sensible rela
tions, are not external but internal, yet not psychical.
When I feel a tangible effect in my tactile nerves, pro
duced by laying my arms on a table containing paper,
cloth, pens, &c,, I feel several tangible objects coexisting
with one another within my tactile nerves. Secondly,
this sense of a relation is as presentative as any simple
sense, and does not construct relations but apprehends
them, when they are present, between the sensible ob
jects. In a word, the immediate apprehension of a re
lation is not a psychological synthesis of abstract sensa
tions, but a synthetic sense of sensible objects. In the
books of idealists, sensation is an abstraction from a
substantial subject perceiving a substantial object of
sense ; and synthesis is a second abstraction, founded on
the first, from the receptivity of a sensible relation. But
in reality there are two apprehensions by a subject of
sensible objects, both equally sensitive ; first, simple
sensations of particular objects, and secondly, synthetic
sensations of particular relations of particular objects.
There are also two kinds of experience : the first, a sum
of simple sensations, e.g. of being pained ; the second,
a sum of synthetic sensations, e.g. of being pained at
repeated blows in a fight.

Curiously enough, Hume over and over again men-

1 Professor Huxley, Hume, chap. ii.

280 PSYCHOLOGICAL IDEALISM

TART II.

lions instances of synthetic sense and synthetic ex
perience intervening between impression and associ
ation, yet without formal acknowledgment. In the
'Treatise,' he says that 'when both the objects are
present to the senses along with the relation, we call
this perception rather than reasoning ; nor is there in
this case any exercise of the thought, or any action,
properly speaking, but a mere passive admission of the
impressions through the organs of sensation.' l Again,
in speaking of the data of the idea of causation, he
says : ' The nature of experience is this : We remem
ber to have had frequent instances of the existence of
one species of objects ; and also remember, that the
individuals of another species of objects have always
attended them, and have existed in a regular order of
contiguity and succession with regard to them. Thus
we remember to have seen that species of object we call
flame, and to have felt that species of sensation we call
heat. We likewise call to mind their constant conjunc
tion in all instances.' 2 In the ' Inquiry ' we find pas
sages close together, one implying synthetic sense

followed by another implying synthetic experience :

' Suppose a person, though endowed with the strongest
faculties of reason and reflection, to be brought on a
sudden into this world ; he would, indeed, immediately
observe a continual succession of objects and one event
following another. . . . Suppose again that he has
acquired more experience, and has lived so long in the
world as to have observed similar objects or events to be
constantly conjoined together ; what is the consequence
of this experience ? He immediately infers the exist
ence of one object from the appearance of the other.' 3

1 Treatise, iii. § 2. 2 Ib> ^ § 6 . cf. § 8>

3 Inquiry, § 5, Part I.

CHAP. IX.

HUME 281

Similarly, in a well-known passage, lie says : ' All
events seem entirely loose and separate. One event
follows another, but we never can observe any tie
between them. They seem conjoined, but never con
nected! l Whatever Hume may say about impressions,
he constantly admits an immediate observation and ex
perience of any relations, short of connection ; when he
says that events seem loose and separate, he does not
mean that they seem quite isolated ; and he allows a
power of apprehending constant conjunction, though
without causation, prior to association. How then
could he refuse to call this sensitive apprehension of a
relation belief, or contend that such a sensitive belief is
a result of association ?

Thus we find that belief or judgment is not an idea,
but an apprehension of a relation ; and not a result of
association, but originally derived from a synthetic sense
of relations. What are the consequences? In the
first place, synthetic sense and judgment are not asso
ciations, because the objects related are both sensible ;
neither is an idea ; and one does not suggest the other,
but their relation is presented. Secondly, synthetic
sense, experience, and judgment having apprehended a
relation in particular instances, cause a complex idea of
the relation ; thus forming a source of ideas unnoticed
by Locke and his followers. Thirdly, although simple
sensations and experiences sometimes, by the anatomical
connection of their nervous centres, without any judg
ment of their relation, produce an association of ideas ;
nevertheless, in an animal capable of judgment, it more
frequently happens that synthetic sense, experience, and
judgment apprehend the relation of the sensible objects,
and cause an association of the ideas of the objects and

1 Inquiry, § 7, Part II.

282 PSYCHOLOGICAL IDEALISM TART n.

of their relation, which I propose to call synthetic sen
sitive association. In these cases, so far from associa
tion producing belief, belief produces association. For
example, I judge that one object follows another in my
senses, and when one appears again, I consequently
have not only the idea of the other, but also the idea of
their sequence, which I could not get from simple sensi
tive association. Here, perhaps, is another stopping
place in animal intelligence. Fourthly, synthetic sense
and experience of relations, being the sources of sensi
tive and empirical judgments without association, supply
the original evidences of reasoning without association.
The want of a formal recognition of synthetic sense, at
the very time he was accumulating instances of its
action, concealed from Hume the true sources of reason
ing, and its independence of association.

The problem of inference or reasoning hinges on two
questions ; the origin of new judgments, and the origin
of new ideas. We have achieved some of the data for
solving this problem ; by showing that we have judg
ments of synthetic sense to start with, that no judgment
is an idea, and that association, in reproducing ideas of
objects already sensible, does not produce an idea of an
object not yet sensible, and does not produce a judgment
at all. These data of themselves indicate a difference
between the association of ideas and the inference of
judgments, and also point to an origin of ideas other than
association. Eeason starts directly from judgments of
synthetic sense, and, without passing through associa
tion, infers judgments, issuing in rational ideas.

' Man,' says Hume, ' is a reasonable being ; and, as
such, receives from science his proper food and nourish
ment.' 1 Hume did not deny reasoning, nor resolve it

1 Inquiry, § 1.

CHAP. IX.

HUME 283

all into association. He had no general theory on the
subject ; and this is one of the weaknesses of his philo
sophy. But he admitted, in mathematics, a species of
reasoning, not only distinct from association, but even
consisting of demonstration from intuition. However, in
spite of his distinction of impressions and ideas, in the
spirit of Locke, he thought that this mathematical
reasoning is limited to the relation of ideas. The point
of his polemic against reason was that it never reaches
matters of fact. He wanted to prove that judgments
of fact, being mere ideas, are mere products of associa
tion. He failed, because judgments are not ideas,
because association does not produce ideas much less
judgments, and because reasoning from sensitive judg
ments produces rational judgments of fact, and rational
ideas. In this part of his philosophy he shows a re
markable spirit of inquiry, and as remarkable a power
of missing the point of difference between one operation
and another.

All conclusions about facts, he thought, are about
cause and effect, all conclusions about cause and effect
are from experience. ' What,' he asks, ' is the founda
tion of all conclusions from experience ? ' 'I want,' he
says, ' to learn the foundation of this inference.' l ' All
inferences from experience, therefore, are effects of
custom, not of reasoning : ' 2 this is the starting-point of
his answer. ' All our reasonings concerning matters of
fact are founded on a species of analogy, which leads us
to expect from any cause the same events which we have
observed to result from similar causes : ' 3 this is his in
terpretation of customary inference. This interpretation
was required for his main point, that customary infer-

1 Inquiry, § 4, Part II. 2 Ib. § 5, Part I.

3 Ib. § 9 ; cf. § 5, Part I.

284 PSYCHOLOGICAL IDEALISM PART n.

ence is the same as the association of ideas ; for analogy
supplies the inference most like the association of ideas.
' When a sword,' he says, ' is levelled at my breast, does
not the idea of wound and pain strike me more strongly
than when a glass of wine is presented to me, even
though by accident this idea should occur after the
appearance of the latter object? But what is there in
this whole matter to cause such a strong conception,
except only a present object and a customary transition
to the idea of another object, which we have been
accustomed to conjoin with the former? This is the
whole operation of the mind in all our conclusions
concerning matter of fact and existence.' l That is,
judgments are strong conceptions, and inferences are
analogies, which are associations of ideas.

Analogous inference is like synthetic association in
data. Both start w^ith the same synthetic experience,
which, in Hume's example, is—

Swords levelled at me have already pained me ;

This sword is like previous swords.

This synthetic experience sets up three processes : (1)
this sword, being like previous swords, reproduces the
idea of having been already pained ; (2) the combina
tion of the two sensitive judgments produces the new
judgment that this sword may possibly pain me ; and
(3) this new judgment produces the idea of being
possibly pained. Of these processes, the first is associa
tion of ideas, the second is analogous inference, the
third is analogous conception. Now, analogous infer
ence is further like synthetic association in process.
Both are customary processes in obedience to certain
laws ; the laws of association, and the law of analogy.

1 Inquiry, § 4, Part II.

CHAP. ix. HUME 285

The law of analogy is expressed in logic as a general
axiom thus : what is related in experience to particu
lars in experience is possibly related to other particulars
like them in experience. This axiom, however, does not
appear in the premises of the inference, but as the laws
of association are laws we spontaneously obey in repro
ducing ideas, so the law of analogy is the spontaneous
law which, without knowing it, we obey in inferring
from particular to particular judgment. It is after
wards discovered by logicians, and then is expressed as
the law of the form of analogous inference ; but it would
be a sad confusion to suppose that because logicians know
it everybody who uses analogy knows it. The axiom
of analogy is a mechanical law of analogous inference ;
and the man who has not studied logic infers from the
above-stated premises that a sword may pain him, not
by reasoning from the axiom as a major premise, but
by the habit of using it as a mechanical law. The
nearest animals to man probably infer by the very same
habit of analogy, as Hume and Mill after him have re
marked. We have already suggested that some animals,
after the lowest stage of mere feeling and the stage of

o O O

mere sense, may stop at simple sensitive association,
while others may rise to synthetic sensitive association.
Perhaps the analogous inference, which we are now
describing, is the highest limit of brute reasoning.
Finally, analogous inference is like synthetic association,
not only in data and process, but also in result. Both
end in particulars.

Where, then, is the difference ? One ends in an
idea of the past, the other in a judgment of the
possible. The association of ideas terminates in an
idea of having been previously pained ; the analogous
inference concludes with a judgment that this sword

286 PSYCHOLOGICAL IDEALISM FART n.

may now possibly pain me. As a judgment is not an
idea, so there must be something different in the processes
which produce the one and the other. The difference
consists in the laws which the processes habitually
use : association, acting by the laws of the reproduction
of ideas from resemblance, contiguity and sequence,
&c., analogous inference by the law of the analogous
production of a particular judgment : what is related in
experience to particulars in experience is possibly re
lated to other particulars like those in experience.
Finally, besides terminating in a judgment, analogous
inference produces another effect, to which association
is incompetent; the conception of an idea of being
possibly pained, which is not to be confused, as Associa-
tionists confuse it, with the idea of having been formerly
pained. This further operation I propose to call analo
gous conception. It is an important operation. Thus,
having by analogy inferred that Mars, being like the
earth, may be inhabited, my analogous conception
pictures an idea of Martial men. If I mistake not,
analogous conception comes much nearer than associa
tion to the productive imagination of art. Analogous
inference, then, is custom, but not association. Asso
ciation is customary reproduction of ideas ; analogous
inference is customary inference from particular to par
ticular judgment; and analogous conception is the
conception in the productive imagination of an idea of
a new particular inferred by analogy.

Hume's reduction of inference from experience to
association breaks down at the very first touch of logic.
It would not be worth while to pursue the subject, had
he not made an audacious attack on induction, reducing
it to analogy, in order to identify it with association.
Moreover, a similar attempt appears in Mill's reduction

CHAP. ix. HUME 287

of induction to inference from particular to particular,1
though in a much more half-hearted fashion, partly be
cause he does not in his ' Logic ' further identify analogy
with association, and partly because, immediately and in
the sequel, he proceeds to treat induction in quite a differ
ent manner. We shall find that induction is not analo
gous inference, much less association of ideas. We must
retrace our steps from Mill, through Hume, to Bacon,
who says : — ' Aut enim defertur judicium ab experimentis
ad experimenta ; aut ab experimentis ad axiomata, quas
et ipsa nova experimenta designent ; ' 2 and to Aristotle,
who, as if foreseeing logical scepticism, warns us that
6 inference by example is neither as particular to general,
nor as general to particular, but as particular to
particular.'3

Induction is not analogy, because the aim of induc
tion is to arrive at a general judgment. By analogy
we infer, not a general but a particular conclusion : by
induction we infer not a particular but a general con
clusion. Hence induction does not contain the very
point of analogical inference, the analogy itself. To
judge that a particular sword is like previous swords is
necessary, if we wanT to reason about that one in par
ticular, but not if we want to conclude generally that
all swords whatever are painful, when levelled at one's
breast. It is true that there is a point in common be
tween the two processes — the judgment that sVords
levelled at me have already pained me, which is also
present in synthetic association. But the difference is
that, when a similar sword is levelled at me, by associa
tion I reproduce the idea of having been formerly

1 Mill, Logic, ii. 3, 7.

• 2 Bacon, De Aug. Scient. v. 2 (p. 622 ; ed. Ellis & Spedding) ; cf. Nov.
Ory. i. 103.

3 Ar. Prior. An. ii. 24 = 69, A 13-15.

288 PSYCHOLOGICAL IDEALISM PART ir.

pained, and by analogical inference I infer that tins

sword may pain me again ; while, without a similar

sword being present, by induction I conclude generally

that all swords, levelled at me, would be painful. As-

! sociation ends in a particular idea, and analogy in a

| particular judgment. Induction ends, not in particulars,

but in a general judgment, beyond the reach both of

association and of analogy.

There is, however, a difficulty in the superior claims
of induction, which did not escape Hume's inquiring
mind. How do we go from the particulars of experi
ence to the general conclusion of inference, from many
to all ; for in the vast majority of cases we cannot
experience all ? In the first place, particulars, which
at once prompt us to association and analogy, do not
justify logical induction. In order to draw a general
conclusion, we must not rest content with this or that
particular, but accumulate instances of three kinds, as
Bacon showed : instances of presence or agreement, of
absence or difference in similar circumstances, of com
parative degrees or concomitant variations.1 Secondly,
even then, we have only experience, albeit scientific, of
many, not of all. There may, in the remaining instances
not experienced, be an exception. ' Mox enim prod-
ibit,' says Leibnitz, ' qui negabit ob peculiarem quandam
rationem in aliis nondum tentatis veram esse.' 2 There
is a leap in induction from various members to the
whole class, from the particulars to the general, from
many swords to all. How do we effect this leap ?
By the axiom of generality : things so related as to be
always present, absent, and varying together in experi
ence are, with a probability proportionate to the extent
of the experience in time, place, and circumstance, so

1 Nov. Org. ii. 11 scq. 2 Leibnitz, DC Stilo Fliilosopliico Nizolii, xxxji.

CHAP. ix. HUME 289

related in all similar cases. This is the law of the form
of induction, distinguishing it from association and
analogy by the power of inferring a general judgment,
leaving it indeed a probable inference, yet with an
approximate certainty continually tending to absolute
certainty.

Three mistakes are often made about this axiom of
generality : first, it is stated most carelessly, as if it
were simply that what is true of many things of a class
belongs to the rest, omitting both the scientific cha
racter of the experience and the problematic character
of the conclusion ; secondly, it is frequently confined to
laws of causation, omitting inductions of coexistence,
&c. ; thirdly, it is often regarded as if it were known to
all men who induce, as an assumption involved in every
case of induction, and even as a major premise convert
ing induction into deduction. The first and second
mistakes I have just corrected by attempting a more
precise and general statement of the axiom. The third
mistake is really too absurd; overlooking, as it does,
that men, from time immemorial, however primitive,
have made, and at the present day, however savage,
do make inductions without dreaming of the axiom;
while Aristotle, the founder himself of the lo^ic of in
duction, even contradicted the law of uniformity by
holding that nature has only a uniform tendency, and
that there are exceptions to universality caused by
accident inherent in matter.1

This false view of the axiom of generality, by
which it is made a supposition involved in induction,
gave Hume his opportunity: he saw that it would
involve us in a circle. 'To endeavour, therefore,'
says he, < the proof of this last supposition by pro-

1 Cf. Ar. Met. E. 2 = 1027 A, 5-15.

U

290 PSYCHOLOGICAL IDEALISM TART n

bable arguments, or arguments regarding existence,
must be evidently going in a circle, and taking that for
granted which is the very point in question.' l This
difficulty has been often felt : we require an induction
to found the axiom, which is nevertheless supposed to be
the assumption involved in all induction. To surmount
it, some resort to the theory that the axiom is a priori,
though many, including Aristotle, have not even believed
it. This was not the alternative of Hume, whose plan
was to surrender universality, and renounce the inductive
inference from particular to general judgments, in favour
of the analogical inference from particular to particular
judgment, which he falsely, as we have seen, reduced
to association from particular impressions to particular
ideas :—

'What, then, is the conclusion of the whole matter ?
A simple one ; though, it must be confessed, pretty
remote from the common theories of philosophy. All
belief of matter of fact or real existence is derived
merely from some present object to the memory or
senses, and a customary conjunction between that and
some other object; or, in other words, having found,
in many instances, that any two kinds of objects, flame
and heat, snow and cold, have always been conjoined
together : if flame or snow be presented anew to the
senses, the mind is carried by custom to expect heat or
cold, and to believe that such a quality does exist, and
will discover itself upon a nearer approach.' 2

Hume was right in rejecting a quasi-inductive de
duction from the supposition of generality. It does
not follow that induction becomes mere analogy, still
less association. Such an alternative is inadequate to
the extent of general reasoning. Moreover, if induction

1 Inquiry, § 4. ~ Ib. § 5, Tart I.

CHAP. ix. HUME 991

were analogous inference and association, the analogy
must always be present, as Hume was well aware.
Wherever he mentions his analysis, he admits that
the analogous object is present, about which the in
ference is made. In the instance above, flame analo
gous to previous flames has to be £ presented anew to the
senses,' in order that we may expect its heat, and snow,
analogous to previous snows, in order that we may ex
pect its cold. In the instance of the sword, having
experienced that swords levelled at us have been pain
ful, I again experience that another sword is present,
in order to infer that this particular sword is painful.
If a new particular, similar to previous particulars, were
not present in the premises, how could analogy infer
an attribute of that particular in the conclusion, or
association use it to introduce an idea? Now, this
condition, though essential to association and analogy,
is unnecessary, or rather completely extraneous, to in
duction. Having experienced the relations of former
particulars, without any new particular being present,
it infers that all flame is hot, all snow is cold, all sword?
levelled at one's breast are painful.

We must find some other alternative, then, which
neither surrenders the inference of generality nor makes
the axiom of generality a supposition antecedent to
induction, whether by an inductive circle or by a priori
mysticism. We have already chosen such an alternative
in explaining analogous inference : it is also applicable
to induction. As the axiom of analogy is the law of the
form, without being a premise, of analogical inference,
so the axiom of generality is the law of the form, without
being a premise, of induction ; a law not known, but
mechanically and spontaneously obeyed by the ordinary
man. and only afterwards discovered by logicians. The

TJ -2

292 PSYCHOLOGICAL IDEALISM TART n.

reason why we induce from some flame is hot, some
snow is cold, some sword-thrusts are painful, is because
we have accumulated so many instances, in which the
related objects have been present, absent and varying
together in our synthetic experience, that, by the law of
generality acting on us without our knowing it, we
cannot but infer general judgments that all flame is hot,
all snow is cold, all sword-thrusts are painful.

Induction, then, like analogy, is the inference of a
judgment; but is distinguished from analogy, because
it proceeds from such an experience as will enable it
by the law of generality to infer a general judgment.
Induction, like association, is not a deduction from the
law of its form, but a customary process by that law.
But its custom is not association. First, association is
a reproduction of past ideas, induction an inference of
general judgments ; secondly, in order to suggest an
idea, synthetic association contents itself with experience
of any relation of objects ; but in order to produce a
general judgment, induction logically requires experi
ence of a relation of objects, present, absent and vary
ing together ; thirdly, the form of association is governed
by the spontaneous laws of the reproduction of ideas,
the form of induction by the equally spontaneous but
different law of generality. Finally, association and in
duction differ not only in themselves, their experience,
and their laws, but also in their result on conception :
association produces no new idea of a particular, much
less a general idea ; induction, having inferred a uni
formity, produces what we may call an inductive idea
of the uniformity, e.g. of the heat of flame in general,
of the cold of snow in general, of the painfulness of
sword- thrusts in general.

Deduction from induction must be discarded by

CHAP. ix. HUME 293

every philosopher such as Hume, who resolves induc
tion into analogy, because, in that case, the inference
from particular to particular usurps the double function
at once of the inference from particular to general, and
of the inference from general to particular. Suppose, as
Mill would say in imitation of Hume, this universal type
of all inference : swords have been painful ; this sword is
like previous swords ; therefore it is or may be painful.
Then there is nothing left for the double process up to
the general judgment about swords as a whole class,
and down to a particular judgment about a sword not
previously known. Accordingly, Hume banished reason
ing, by which he meant deduction, entirely from em
pirical conclusions ; and Mill declared syllogism to be
no inference, regarding the double inference from par
ticulars through a generality to a new particular as an
unnecessary circuit.1

But induction and deduction are integral and
complementary parts of a double process of inference,
from particular to general, from general to particular.
As we have seen, induction is not analogy ; it begins
with particulars, but ends, not with a new particular, but
with a general judgment about a class. Deduction from
induction, or empirical deduction, as we may call it, com
pletes the double process : it combines the general judg
ment with a particular judgment, that a new particular
belongs to this class, and infers that what belongs to the
class belongs to the new particular. Empirical deduc
tion differs from analogy in starting, not directly from
particulars, but from a general judgment, given by
induction ; it differs from induction, not only in using
this general judgment as major premise, but in adding
a minor, and drawing a particular or less general con-

1 Mill, Logic, ii. 3.

204 PSYCHOLOGICAL IDEALISM PART n.

elusion. It may be called the complement of induction,
needed to convert generalities into particulars, and
bring the double process of general reasoning to a
particular conclusion, like that of analogy, but reached
through a generality.

Deduction, as discovered by Aristotle, and disen
gaged from the mere schematism of Galen and later
logicians, consists of three orders or figures. They are
three different ways of thinking. Sometimes I want to
prove or disprove by means of a class ; for example,
belonging to the class of magnitudes whose angles are
equal to those of a perpendicular falling on a straight
line, proves that a triangle has its angles equal to two
right angles : this is the first figure or order of deductive
thinking. Sometimes I want to disprove by means of a
difference ; for example, as a demagogue differs from a
statesman in being a truckler, he is not a statesman :
this is the second order of deductive thinking, the figure
of difference. Sometimes I want to prove by an instance
or disprove by an exception ; for example, the genius
of Nelson is sufficient to prove what Englishmen were
capable of at the beginning of this century : this is the
third order of deductive thinking, the figure of instance.
Each of these figures has its own axiom ; that of the
first being the dictum de omni et nullo, discovered by
Aristotle ; those of the second and third being respec
tively the dictum de diverse, and the dictum de exemplo
et excepto, discovered by Lambert.1

I have a purpose in calling attention to these three
axioms of the three orders of deductive reasoning. They
are necessary laws of deduction ; yet they are not in the
premises. Moreover, as men, in deducing conclusions,
know nothing about them, they have not already been

1 Lambert, Neues Organon, i. 4, § 232.

CHAP. IX.

HUME 295

acquired by a previous induction, still less are appre
hended a priori. They are not presupposed, but used.
What is the explanation? Precisely the explanation
already given for the law of analogy and the law of
generality. They are spontaneous laws used by every
deducer, but discovered afterwards by the logician.
Hence they never appear in a syllogism, being not its
premises, but the laws of its form, each of the three
dicta being the law of its own order of deductive think
ing. As Aristotle said, ' the nature of a syllogism is
not premised in a syllogism.' 1

Deduction would not be an inference, if it were not
an advance in knowledge ; but it is an advance in
knowledge. If induction were founded on a complete
examination of all members of the class, there would be
no occasion for deduction. But usually we only examine
some members, from which induction leaps, by the
axiom of generality, to the class ; and this very fact is
what, according to Mill, makes induction an inference :
we need only know some, not all particular men, to say
that all men are mortal. Hence there is, so to speak, a
generality about induction which only says that every
body who may be a man is mortal : it does not, and
cannot, enumerate every particular man. The con
sequence is that the subsequent process of deduction,
which, by combining the generality in the major premise
with a new particular in the minor, enables us to dis
cover that a particular object, which we never appre
hended before to be a man, is mortal, must be an.
advance in knowledge, and therefore a process of
inference.

Mill would reply that, in this case, we are committing
a petitio prindpii by adducing in proof of a particular

1 Ar. Post. An. ii. G = 92 A, 11.

296 PSYCHOLOGICAL IDEALISM

PART IT.

a general judgment which presupposes it. This objec
tion can only mean that the general judgment, all men
are mortal, ought to have been inductively proved by
examining all men : otherwise it does not presuppose
every particular man. But, according to his own show
ing, the general judgment is not to be proved by every
particular : therefore it does not presuppose every par
ticular, but only the original particulars of induction ;
and therefore the process, which adds to the general
judgment a new particular, is not using a general judg
ment which presupposes that new particular, and is not
a petitio principii. Mill made the beauty of induction
the vice of syllogism : he first says that only some par
ticulars are presupposed to induce a universal, and
then that the universal presupposes every particular to
deduce a particular. Eeally, the justification of induc
tion is the justification of deduction from induction.
Induction from many of the particulars concludes all in
general : deduction adds the rest of the particulars.

Mill was deceived by another mistake : he thought
the inference was over when we get to the general judg
ment, and the remainder is deciphering our notes. But
the major premise is as powerless without the minor
premise as the minor without the major. < It is evi
dent,' as Aristotle says, c that a syllogism consists of
two premises, and no more ; for three terms make two
premises.' ] We therefore require two sets of notes in
order to decipher a conclusion, and their combination
in the two premises is the essence of syllogism. Mill
was further deceived by writing down two simple pre
mises, and thinking that, as the syllogism consists in
drawing the conclusion, which is contained in the pre
mises, it does not advance our knowledge. But a

1 Ar. Pr. An. i. 25 = 42 A, 32 3.

CHAP. ix. HUME 297

syllogism does not consist in merely drawing a conclu
sion ; and, when you have written down the two
premises, the essence of syllogism is over : the difficulty
is in combining the premises, and although the syllogism
does not discover each premise, it does combine the
two. So important an act is this, that, as Aristotle
says, ' a man may know that all B is A and all C is B,
and yet think that C is not A ; e.g. that every mule is
barren, and this is a mule, and yet think it is going
to foal, through not considering each of the two premises
in combination.' l

Syllogism, then, from induction is an inference,
because it is an advance in knowledge by adding par
ticulars not contemplated in induction ; a legitimate
inference, because it presupposes only the particulars
contemplated in experience, and the indefinite generality
inferred by induction, but not the new particular it is
about to prove ; a complex inference, which consists
neither in merely interpreting a major premise, nor in
merely drawing a conclusion, but in a new combination
of premises, or a direct comparison of two things with a
third thing, so as to draw an indirect conclusion about
their relation. In order to express the essence of syllo
gism as a process of inference, I propose to define it : a
combination of two premises so as to produce a conclu
sion, not presupposed in either separately, though con
tained in their combination. Hume's theory of inference
is inadequate, because it ignores this process of reason
ing from experience and induction ; and Mill's is false,
because it ignores the combination of premises, which
produces a new conclusion, advancing our knowledge.

There are two ways of inferring from particular to
particular ; directly by analogy, and indirectly through

1 M o-vvdfwp&v TO naff (Kartpov. Ar. Pr. ii. 21 = 67 A, 33-7.

298 PSYCHOLOGICAL IDEALISM TART n.

a generality by induction and deduction. Mill, follow
ing Hume, confuses them. 'The mortality of John,
Thomas, and others,' he says, < is, after all, the whole
evidence we have for the mortality of the Duke of Wel
lington. Not one iota is added to the proof by inter
polating a general proposition.' l Why, he asks, should
we not take the shortest cut ? We often do : we go,
like brutes, from particular to particular. But Mill
himself gives a very good reason why we should not ;
that to pass through a general proposition is « a security
for good reasoning.' 2 Now, surely the aim of every
honest man is, not reasoning, but good reasoning ; and
logic is the art of reasoning well. We must avoid the
shortest cut and go round the circuit of induction and
deduction to rational truth, as we must avoid the broad
and choose the narrow path to eternal life. We may be
sure also that there is something more than usual in
a security for good reasoning. That something more is
the evidence of induction. We saw that we want less
evidence for association and for analogy, which begin
directly after experience, than for induction, which
requires experience to be accumulated and sifted, by
finding things present, absent, and vary ing together, so as
to bring into operation the law of generality, by which
we spontaneously induce a general judgment. In order
to deduce a new particular we must have apprehended
not only the original particulars, but also that they are
sufficient to authorise a general judgment, which is the
same thing as inferring it. The beauty of induction is
the virtue of the syllogism. It is because analogy has
not, induction has, sufficient evidence to infer a general
judgment, that syllogism from induction is a security
for o-ood reasoning. I do not say that this security is

1 Mill, Logic, ii. 3, 3. 2 Ib. ii. 3, 8.

CHAP. ix. HUME 299

more than general probability. But analogy lias not
this general probability: it varies, indeed, in proba
bility, but directly its evidence guarantees general
probability, analogy becomes induction followed by
deduction. Induction is the inference of general pro
bability and empirical deduction the inference from
general probability ; and the probability of the double
process, induction and deduction, varies, according to
the original synthetic experience, from uncertainty to
approximate certainty.

Few modern logicians seem to have a sense of the
enormous importance of syllogism or deduction. They
do not feel the indefiniteness of the subject of a general
judgment, which signifies all whatever they may be I
do not know, the consequent imperfection of induction
without deduction, and the necessity of syllogism to give
defmiteness to our inferences from experience beyond
experience. It is but little use knowing that when the
earth intervenes between the sun and the moon there
will be an eclipse, unless we are prepared to combine
this mere generality with minor premises stating when
the earth will be in this position. It is by deduction
that we go back to the distant past: for example,
nations which have words in common, expressing a
degree of civilisation, too many to be explained by
nature, chance, or communication, lived together up
to that degree of civilisation ; the Greeks and Eomans
had a multitude of words in common up to the stage of
settled agriculture ; therefore they lived together to that
point. It is by deduction we dive into the imper
ceptible present : for example, perceptible bodies elastic
and compressible have parts and pores ; solid bodies are
elastic and compressible ; therefore they consist of parts
and pores, though imperceptible. It is by deduction that

300 PSYCHOLOGICAL IDEALISM

PART II.

we predict the future : for example, a planet deflected
from the path prescribed by its gravitation to the sun
gravitates to another planet in the direction of deflec
tion ; Uranus was found so to deflect ; therefore, a new
planet was predicted in the direction of deflection, and
the new planet, Neptune, was afterwards discovered in
that direction. Inferences of this kind are sometimes
analogical, but they are often deductive, and they are
so whenever induction has established a general judg
ment. They are sometimes confused with induction,
as when Mill calls the discoverer of a murderer by
circumstantial evidence induction.1 But when such an
inference is not analogy it is deductive, because it con
tains, besides the circumstances in the minor premises,
a number of major premises judging such circumstances
to be signs of murder, and a particular conclusion infer
ring a murderer.

Empirical deduction, like analogical and inductive
inferences, is not association, and for the same reason ;
it ends, not in conception, but in judgment. Even
syllogism is a customary inference ; but its custom is
not an association of ideas, but a habit of inferring
judgments by the three laws of the three figures. Asso
ciation, even of the more developed kind which starts
from synthetic experience, gets as far as reproducing
the ideas of the objects in that synthetic experience,
and there stops. At that point we have not even got at
the beginning of deduction : induction intervenes to
infer the general judgment, which, as we have seen, is
no result of association. Empirical deduction, then,
begins with this general judgment, which at once dis
tinguishes it from all sensitive synthetic association.
It proceeds frequently to ask synthetic experience for a

1 Mill, Logic, iii. 14, 7.

CHAP. ix. HUME 3Q1

minor premise, e.g. this is a sword ; and thus returns to
the arena of association. But association deals with
this new particular merely to reproduce ideas of former
pains already apprehended by experience, analogy, and
induction. Syllogism goes on to infer a new particular
judgment that this sword will also prove painful.

Nor is this all the difference : we must not deceive
ourselves by taking too simple an instance. Through
the power of general judgments we at last deduce par
ticulars not only beyond sense, but insensible and im
perceptible to us, e.g. the existence of insensible particles
or corpuscles of solid bodies. Association, from the pre
mises of this deduction and before deduction has drawn
the conclusion, will reproduce only the ideas of the
parts of bodies previously known : without deduction it
will not enable me either to judge that the particular
bodies in the minor premise consist of parts, or to con
ceive ideas of their particular parts. Deduction, on the
other hand, proceeds to draw the conclusion and then
conceive the idea. Not association of ideas, but deduc
tion, produces the judgment of the existence, and through
this judgment the deductive conception of the idea, of
a corpuscle.

Hume, in the ' Treatise,' said : ' I form an idea of
Borne, which I neither see nor remember, but which is
connected with such impressions as I remember to have
received from the conversation and books of travellers
and historians. ... All this, and everything else which
I believe, are nothing but ideas.' l This inadequate ac
count of my knowledge of Borne goes further than could
be justified by Hume, but not so far as is justified by
history. Association of itself would not even give me
an idea of Borne, which I have never seen ; history

1 Hume, Treatise, iii. § 9.

302 PSYCHOLOGICAL IDEALISM TART n.

infers the judgment that Rome has existed, perhaps
from the time of Romulus, certainly since the transition
from the monarchy to the republic. If association of
ideas were substituted for deduction, on being told
that there is a city like London and other cities I have
experienced, I could only reproduce my particular
idea of London, my particular ideas of other cities in
my experience, and, with induction, my general idea
of cities : I could not produce a new- particular idea
of Rome beyond my experience. But deduction from
the conversation and books of travellers and historians
enables me to produce a belief that Rome exists, and
has existed for centuries, which is not an idea, and,
moreover, besides the belief, an inferential idea of
the eternal city in my productive imagination. De
duction is not association of ideas, because it directly
produces deductive judgments about the existence, and
indirectly deductive ideas ; of objects beyond sense, such
as the danger of a sword which has not yet hurt me ;
of insensible objects, such as historic Rome ; of imper
ceptible objects, such as a physical corpuscle.

Hume, having falsely identified ideas with thoughts,
and resolved beliefs into ideas, could allow only one
succession of thoughts, the succession of ideas. But
judgments are not ideas but apprehensions of relations,
inferences are not successions of ideas, but successions
of judgments, and rational judgments are thoughts
which are not ideas. From synthetic sense, which
produces our first judgments of relations, there arise
two streams of thought, synthetic association, which is
a succession of ideas, and reasoning or a succession of
judgments. These two streams flow together, yet dis
tinctly ; but the stream of reasoning is the main river
of human thinking, compared with which the stream of

CHAP. ix. HUME 303

association is a mere rivulet. Hume and his followers
are like those explorers of the sources of the Nile who
have taken a mere tributary for the main river.

Inference and association are alike, not only so far
as both start from synthetic sense and experience, but
also in both being involuntary, spontaneous, custo
mary. Impressions involuntarily suggest ideas, though
we also recall them by voluntary reminiscence ; nor can
we help inferring judgments, though we also reason
voluntarily. Association and inference both sponta
neously use laws, neither inductively nor a priori, but
mechanically and without knowing it : as the laws of
association, by resemblance, contiguity, succession, &c.,
are spontaneously used to introduce ideas, so is the
law of analogy spontaneously used to infer from par
ticular to particular judgment, the law of generality to
infer a general judgment, the laws of the three figures
to infer from general to particular judgments. The ex
planation is probably the same in all cases, namely, the
evolution of an habitual tendency by the action of nature
on our organs without our knowing it. Again, analogy
and induction are not deductions from the laws of their
forms, but independent inferences from experience ; nay,
deduction itself is not a deduction from the laws of its
forms or figures, but from major and minor premises :
all three processes of inference use their laws to pro
duce judgments as habitually as association uses its
laws to reproduce ideas. But because inference is an
inevitable, spontaneous, customary use of laws, it is not
on that account to be confused with association.

Hume made two very great blunders about inference :
he confused custom with association, and limited reason
ing to deduction, or rather demonstration. But not all
custom is association : there are habits of conceivin^,

304 PSYCHOLOGICAL IDEALISM TART n.

of judging, of acting ; and analogy, induction, and de
duction are habits of judging by inference, not habits
of conceiving by association. Again, all inference is
reasoning, because it advances from judgment to judg
ment ; reasoning does not begin with demonstration
from axioms ; and there are three kinds of reasoning, all
ultimately founded on judgments of synthetic sense, all
inevitable, spontaneous and customary inferences bylaws
of their forms, used without being known, except to the
science of logic : — these three types of inference are
analogical, inductive and deductive reasoning.

Seasoning is an instinct. The premises are acquired
from experience, and the conclusion is inferred ; but the
process of inferring is instinctive. It was probably
gradually acquired by the action of natural uniformity
on our organs : but it is used without presupposing any
axiom of natural uniformity as a major premise. This
instinctiveness of reason escapes the notice of philoso
phers and even of logicians. Hume, for example, post
poned reasoning to association, because it is slow, be
cause it does not appear during the first years of infancy,
because it is liable to error, and because nature has im
planted in us the instinct of association.1 But, in the
first place, nothing is more rapid than reasoning, which
goes through its trains of judgment as quickly as asso
ciation through its trains of ideas ; secondly, it is an
adult prejudice to suppose that young infants are not
reasoning, because they are not talking, when they are
far better occupied in the sensible and rational discovery
of an internal and external world ; thirdly, reasoning is
liable to error, but association has no perception of truth;
fourthly, if association is an instinct, so is reasoning,
each spontaneously using its laws to proceed from expe-

1 Inquiry, § 5.

CHAP. ix. HUME 305

rience, but the former to ideas, the latter to judgments.
So closely related are the instincts of association and
reasoning, yet so different, that, if association were not
the vaguest term in the vague vocabulary of mental
philosophy, I should have proposed to distinguish the
two successions of thought as the association of ideas
and the association of judgments.

Hume allowed the psychology of association to
blind him to the logic of reasoning. The consequence
was that he missed the whole origin of rational judg
ments and of rational ideas ; thus defeating his own
object, which was to find the causes of ideas. The
origin of ideas is in reality a very complicated problem,
inseparable from that of judgments. We must distin
guish productive and reproductive conception. The
sources of productive conception, which we have
reached so far, are simple sensations of sensible objects
producing sensible ideas, synthetic sense and judgment
of sensible relations producing ideas of relations, and
reasoning to rational judgments producing rational
conceptions; moreover, we have distinguished three
kinds of rational conception, answering to three kinds
of inference — analogous, inductive and deductive ; and,
finally, deductive conception produces ideas not only of
the relation in the conclusion but of the insensible ob
jects of that relation, e.g. the idea of corpuscles as well
as of their cohesion. Eeproductive conception has two
main sources, both obeying the same laws — voluntary
recollection (am/^o-t?), analysed by Aristotle,1 and
involuntary association, analysed by Hume ; who pro
ceeded to elevate a mere reproduction of ideas into a
substitute for the inference of judgments, and, when it
does not produce ideas at all, and is only one way of

1 Ar. De Mem. ii.

306 PSYCHOLOGICAL IDEALISM PART u.

reproducing them, positively made it the sole source of
all belief in matters of fact. What a contrast there is
between the analytic genius of Aristotle, giving each
operation its due place, and the exaggerated scepticism
of Hume, exalting the weakest over the strongest force
in man's composition !

Hume invariably speaks as if all association of ideas
were of one kind ; so usually do his followers. It is
because they have become enamoured of one power to
the neglect of the rest. This kind I have ventured to
call ' simple sensitive association,' because it starts with
simple sensation and experience. But I have shown
that Hume covertly introduces another kind, which I
have called ' synthetic sensitive association,' because it
starts with synthetic sense and experience. To this
sort belongs the association used by him to explain the
apprehension of causation ; a process which, starting
from the sense of sequence, and passing through the
experience of constant conjunction, ends by the ante
cedent introducing the idea of the consequent, which
he falsely supposed to be our judgment of a cause
producing an effect. But, now that I have analysed
reasoning, I am prepared to take a further step and
say that reasoning, though never association, is the
foundation of a third kind of association, which I
shall call ' rational association.' When we have by
any kind of inference inferred a relation, and by any
kind of rational conception produced the ideas repre
senting the relation and its objects, then, and not till
then, rational association will enable us to reproduce
the ideas by its own laws. Thus the contempla
tion of eye, which suggests to the ordinary man
the idea of love or war, will to the optician repro
duce the rational ideas of aether, of undulation, of

CHAP. ix. HUME 307

reflection and refraction. But it would be mere con
fusion to merge the reasoning by which he dis
covered these facts in the association of the ideas,
when the rational conception of the ideas intervenes
between the rational inference and the rational asso
ciation. The optician first by reasoning judges the
existence of asther and its motions, then rationally
conceives what ideas of them he can, and finally is
reminded of them by association. Most associations
are post-rational.

The inference at the bottom of rational association
will be found to solve many unsolved problems. One
is the solubility of association. If we depended on as
sociation alone, an association acquired by a constant
experience could only be dissolved by one acquired by
a still more constant experience. But, as a fact, a
single instance will destroy the strongest association :
when the idea of the proverbial whiteness of swans was
dispelled by the discovery of a black swan, it was
reason which dissolved the association. Another pro
blem is the origin of complex ideas of substance. The
theory of Associationists is that, having by sensation
acquired together the ideas of yellowness from sight,
smoothness from touch, sweetness from taste, association
recalls these ideas so constantly as to form one complex
idea of an orange. In this analysis the main elements
of the simple ideas, and the process between them and
the complex idea, are omitted. By sight we already see
a yellow, by touch a smooth, by taste a sweet substance ;
hence the simple ideas of substances ; by reasoning, we
infer that all these correspond in our senses to one
complex substance outside, represented by the yellow
in sight, the smooth in touch, the sweet in taste ; and,
having thus inferred an external orange, we form a

x 2

308 PSYCHOLOGICAL IDEALISM TART n.

rational complex idea of it, which we then reproduce,
not produce, by rational association.

Another and somewhat different kind of problem is
the origin of fictitious ideas, of the ideas of art, and of
ideals. Hume truly said that an idea, such as that of
a golden mountain, is ultimately made out of im
pressions, but falsely thought that it is produced by
sensitive association, which by itself could only repro
duce the sensible ideas of gold and of mountain. The
reasoning of the possible intervenes. We infer that as
a mountain is made of one material it might be made
of another, and having judged the possibility, analogi
cally conceive the idea of a golden mountain, which is
only reproduced by association. Sometimes we infer
the possibility of more, sometimes of less, than sense per
ceives ; hence we multiply man and horse into centaur,
or diminish man into ghost. Sometimes we infer the
possibility of something better than ordinary, as Homer
did Achilles; sometimes worse, as Shakespeare did
Caliban. But in artistic idealisation there is always an
inference of possibility, which is the foundation of all
ideal conception. It is quite the same in philosophical
ideals. Plato thought of yhe possibility of men be
coming angels before he conceived his ideal state.

The final and most difficult problem is the influence
of the association of ideas beyond ideas. Locke started
this general problem in the 'Essay.' The following is an
often quoted instance from his chapter on Association :—

6 The ideas of goblins and uprights have really no
more to do with darkness than light ; yet let but a
foolish maid inculcate these often on the mind of a
child, and raise them there together, possibly he shall
never be able to separate them again so long as he
lives ; but darkness shall ever afterwards bring with it

CHAP. IX.

HUME 809

these frightful ideas, and they shall be so joined that
he can no more bear the one than the other.' l

Locke did not make so much of this effect of asso
ciation as the followers of Hume, who often suppose
that the association of the ideas of ghosts and the dark
produces a belief which produces a fear. But the fear
often follows the idea, without the belief. There are in
reality two different cases, in one of which there is no
belief, in the other a belief, but not caused by the asso
ciation of ideas. In simple sensitive association, where
there has been no judgment of the relation of a ghost
with the dark, the idea of the dark mechanically recalls
the idea of the ghost, and this the idea of pain which is
sufficient to generate fear. In synthetic sensitive asso
ciation, where there has been a judgment that a ghost
appears in the dark arising from a child's belief in the
narratives of its nurse, the association of ideas is accom
panied by a belief that a ghost may possibly appear,
which, however, does not arise from the association of
ideas, but by parallel inference from the same judgment
as that which produces the association. Sometimes this
judgment of possibility may arise, even when the person
is sceptical about the actuality of ghosts. Still more
often it is a vague inference of some dreadful possibility,
because the dark is mysterious to man.

Whenever, then, the association of ideas is of a
simple kind, which has not arisen from a judgment, it
is powerless to produce one ; and whenever it is ac
companied by a judgment, they are joint effects of an
original judgment, which produces on the one hand an
inference at least of possibility, and on the other hand
an association of ideas. At the same time there is an
effect of association on belief, like the effect of volition..

1 Essay, ii. 33, 10.

310 PSYCHOLOGICAL IDEALISM PART n.

These two reproductive causes of ideas, by constantly
promoting the same idea, challenge our attention not
only to the idea but also to the parallel judgments.
Thus a person, who constantly cherishes the idea of
being wiser than others, will at last come to think he is
so, not however from the association itself, but because
Ids attention is thereby called towards the evidences
which infer his superiority, and away from those which
disprove it.

Hume's empirical theory consists in three proposi
tions : (1) All perceptions are impressions and ideas or
thoughts ; (2) All ideas or thoughts are copies of
impressions ; (3) Association of ideas is the origin of
all beliefs of facts, that is, ideas or thoughts. But it is
one thing to assert an empirical theory in general, and
another to fill in its details. Impressions, as Hume de
scribed them, are not by the process of association, as
Hume described it, the origin of ideas, which are not,
as Hume described them, all our thoughts. In the
first place, the simplest sensation is merely an abstract
attribute of a substantial subject apprehending a sub
stantial object, and the simplest reflection an abstract
attribute of that substantial subject apprehending him
self. Secondly, sense is not only simple but synthetic ;
and synthetic sense is the immediate origin of sensitive
judgment, which is not an idea, but the immediate appre
hension of a relation of sensible objects. Thirdly, associa
tion is a reproduction, but it is not a production, of ideas,
still less of beliefs, which are not ideas but judgments
ultimately based on synthetic sense. Fourthly, reasoning
is not an association of ideas, but of judgments ; and
••there are three types of inference — analogical, inductive,
and deductive — all starting from synthetic sense, and by
their own laws instinctively inferring rational judgments

CHAP. IX.

HUME 311

which are not impressions nor ideas, yet are thoughts.
Fifthly, the productive origin of ideas is simple sense
forming the first ideas of qualified substances, synthetic
sense forming the first ideas of relation, and reasoning
analogical, inductive and deductive, which forms ideas
not only of what is inferred to be actual, but also of
what is inferred to be possible, fictitious, ideal : the re
productive origin of ideas is passive association and
active volition. Sixthly, there are three species of as
sociation, simple and synthetic sensitive association, and
rational association. A philosopher who, like Hume,
does not understand reasoning, cannot understand ideas
and their association. Logic is necessary to psychology.
Empirical philosophy must comprise reason. If all
knowledge is from experience, it is certainly not ac
quired by association.

Hume concludes his ' Inquiry ' with his Academical
Philosophy.1 He starts with what he calls the instinct
by which men c suppose the very images presented by
the senses to be the external objects ; ' on which he
makes the following comment:—

' But this universal and primary opinion of all men
is soon destroyed by the slightest philosophy, which
teaches us that nothing can ever be present to the
mind but an image or perception, and that the senses
are only the inlets through which these images are
conveyed without being able to produce any immediate
intercourse between the mind and the object.'

This most instructive passage shows, first, that ideal
ism has a real advantage over intuitive realism, which
falsely accepts the perception of an external object,
and secondly, that idealists tend to beg that the repre
sentative image perceived is a perception by confusing

1 Inquiry, § 12.

312 PSYCHOLOGICAL IDEALISM

PAKT II.

the object with the operation of sense. Idealism is the
scientific truth that sensible objects are effects on
the senses, misinterpreted into the hypothesis that they
are c perceptions in the mind,' as Hume calls them in
the same paragraph, without evidence.

Having now got himself into a self-made difficulty
about the data of sense, he proceeds to torture himself
with the following question : —

' By what argument can it be proved that the per
ceptions of the mind must be caused by external objects,
entirely different from them, though resembling them
(if that be possible), and could not arise either from the
energy of the mind itself, or from the suggestion of
some invisible and unknown spirit, or from some other
cause unknown to us ? '

This question is put with the logical power of
Berkeley, and is answered with even more logic :

' It is a question of fact, whether the perceptions of
the senses be produced by external objects resembling
them : how shall this question be determined ? By
experience, surely, as all other questions of a like
nature. But here experience is, and must be, entirely
silent. The mind has never anything present to it but
the perceptions, and cannot possibly reach any experi
ence of their connection with objects. The supposition
of such a connection is, therefore, without any founda
tion in reasoning.'

The fallacy of this argument consists in the assump
tion with which it begins. Eeally, we are conscious of
perceptions, or rather of ourselves perceiving ; but we
perceive not perceptions, but sensible objects, and not
in the mind, but in the nervous system ; and from these
physical objects within we infer physical objects with
out, different individually, but specifically similar to the

CHAP. ix. HUME 313

sensible objects from which they are inferred. But
though Hume's data were false, his conclusions were
logical. If all that we perceived were perceptions, they
would be entirely different from external objects ; and
experience, being confined to perceptions, would have
no data to prove anything at all about objects, internal
or external. Moreover, if the data both of sensitive
and reflective perception wrere perceptions, qualities as
ideas of sensation, and operations as ideas of reflection,
as Locke and Berkeley formally stated, we should only
be able to infer perceptions. Hume has the best of the
logic when he refuses to follow either Locke in sup
posing matter, or Berkeley in supposing mind, seeing
that neither of these philosophers allowed matter and
mind in the data. of sensation and reflection, when they
were delivering themselves ex cathedra on the subject of
sensible data. As Hume afterwards says, nothing re
mains but ' a certain unknown, inexplicable something,
?,s the cause of our perceptions.' Such is the false
though logical end of Hume's speculative philosophy.

He proceeds illogically to correct himself of his
Pyrrhonism by the old view of the Academy that c all
human life must perish, were his principles universally
and steadily to prevail,' which is no answer to the
Pyrrhonist or to Berkeley, who would immediately
resolve our bodies, our clothes, our food, our estates,
into perceptions. But Hume valued common life too
highly, and natural philosophy too little. We are not
committed to the dilemma of thinking in one way and
living in another. The answer to his 'mitigated scepti
cism or academical philosophy' is the physical dis
coveries of natural philosophy. If, indeed, the objects
of perception were perceptions, we should never infer
anything but perceptions, with an unknown, inex-

314 PSYCHOLOGICAL IDEALISM PART IT.

plicable something. But natural philosophy has dis
covered imperceptible objects, substances qualified,
causing and receiving motions, in accordance with uni
versal laws, and ultimately causing our perceptions.
Therefore, it is neither true that knowledge ends in an
unknown something, nor that the objects of perception
are perceptions, from which imperceptible objects of
science could not have been inferred. The slightest
philosophy teaches us that what is present to sense is an
image, but not that this image is a perception. Simple
sense perceives an object, internal but physical ; syn
thetic sense and experience perceive the relations of
these physical objects within, and reason infers the
relations and existence of physical objects without.

Hume's philosophy is a deductio ad absurdum of
idealistic hypotheses. It is what was sure to follow if
Locke and Berkeley were taken at their word, no re
gard being paid to their admissions. As soon as the
Cartesian consciousness of the thinking subject had
been forgotten, all the data of sense were reduced by
Locke and Berkeley to ideas, qualities as ideas of sen
sation, and operations as ideas of reflection ; and the
objects of understanding were logically inferred to be
also ideas. Locke illogically admitted the supposition
of substances, material and thinking ; Berkeley dog
matically asserted the existence of mind as gathered from
its effects ; and both ended by admitting the conscious
ness of one's own existence. Berkeley saw the incon
sequence of Locke's supposition of material substance
beyond mere ideas of sensation, but he did not see
that he was with equal inconsequence introducing
mind, soul, spirit, directly after mere ideas of reflection.
Hume acutely detected the half measures of Berkeley,
but took the wrong alternative. Instead of going be-

CHAP. ix. HUME 315

hind both Locke and Berkeley to show that both sensa
tion and reflection perceive qualified substances, he
banished the thinking to the limbo of the material
substance, and rigidly confined us to the abstract per
ceptions which form the sum of the data of perception
by the confession of both his predecessors. This con
clusion is argued out in the ' Treatise ' on the following
text : ' We have no perfect idea of anything but of a
perception. A substance is entirely different from a
perception. We have therefore no idea of a substance.' l
This logical syllogism, of which however the major is
quite false, is applied both to material and thinking
substance, in the ' Treatise.' In the ' Inquiry,' he became
silent on this point ; but ignorance of substance is a
necessary consequence from the perception of percep
tions, which is common to both books.

Hume may be said to have gathered the ideal theory
of perception into a focus which reveals to us its errors.
The supposition that sensible objects are psychical
operations deprives us of objects and physical objects
within, from which to infer physical objects without.
The supposition that sensible objects are qualities and
operations deprives us of the sense and inference of
substances ; of the sensation and inference of material
substances, and of the consciousness and inference of
thinking substances, partly physical, partly psychical.
On every side he paraded the mere logic of idealism.
He was particularly attracted by Berkeley's philosophy ;
for instance, by the theory of general ideas, and of
primary and secondary qualities. Berkeley's hypothesis,
in the 'Principles,' of the inactivity of ideas, antici
pated Hume's scepticism about power in causation ;
while, in the c Theory of Vision,' the hypothesis that

1 Treatise, iv. § 5.

316 PSYCHOLOGICAL IDEALISM PART u.

visible ideas suggest tangible ideas, without any inference
of an external object common to touch and vision, gave
the first hint for Hume's substitution of association for
reasoning. Hume's scepticism is the dark shadow of
Berkeley's theosophy, giving us the logical warning—
if no matter, then no spirit, and no God. He had no
suspicion that Berkeley's so-called principles were hypo
theses, any more than modern idealists have. Hence he
says of Berkeley's arguments that ' they admit of no
answer, and produce no conviction'1

Here Hume missed an opportunity, such as seldom
falls to the lot of a philosopher. Instead of being
merely logical from the original hypotheses of his pre
decessors, he ought to have used their subsequent
admissions for a new departure in philosophy. He
should have returned to the Cartesian consciousness of
a thinking subject. He should have shown that both
Locke and Berkeley, after beginning with a reflection of
mere ideas of operations, admitted at last a direct con
sciousness of one's own existence. He should have
pointed out that this means a reflective consciousness
of oneself as a thinking substance, and have similarly
recognised sensations of qualified substances within
oneself. From these data, together with the synthetic
sense of relations, he could have proceeded to explain
our inferences of external substances — bodies, thinkers,
God. But he preferred not to answer his predecessors,
to stick to the idealistic last, and to work on nothing
but impressions of sensation and reflection.

To this scepticism about sense Hume added a scepti
cism about reason. Logic, through the process of being
made into text-books for education, has been too much
schematised. For example, Aristotle distinguished

1 Inquiry, § 12, Part I., note.

CHAP. ix. HUME 317

simple from complex apprehension,1 and names from
propositions,'2 but did not co-ordinate reasoning with the
two other apprehensions. St. Thomas Aquinas schema-
tistically added reasoning as a third operation.3 The
moderns, by co-ordinating the three operations, have
tended to lose sight of the process of reasoning at the
back of conception and judgment, and many modern
logicians speak as if there were three independent pro
cesses, conducted quite independently, each with its
own independent laws. But reasoning is a process from
judgment to judgment, producing new conceptions.
Again, the conceptualistic view of logic intensified
the mischief, by regarding judgment as apprehending,
and therefore reasoning as inferring, relations of ideas.
At the same time, Descartes exaggerated the power of
ideas over knowledge.

These causes produced the exaggerated attention to
ideas and their origin, their arbitrariness, and the post
ponement of reasoning in Locke's 'Essay' and Berkeley's
'Principles.' The disease came to a head in Hume's
works. In the first part of his ' Treatise,' which is
directly modelled on Locke's Second Book, Hume takes
as his problem the mere origin of ideas. In the course
of the same work he animadverts on the distinction of
acts of the understanding into conception, judgment
and reasoning, and the definitions given of them. ' Con
ception,' he says, ' is defined to be the simple survey of
one or more ideas ; judgment to be the separating or
uniting of different ideas ; reasoning to be the separating
or uniting of different ideas by the interposition of
others.'4 But his animadversions on these purely con
ceptualistic definitions only end in his reducing all these

1 Ar. DC An. iii. 0. - Id. PeriJtcrm, i.

3 Aquinas in Perilierm, i. 4 Treatise, iii. § 7, note.

318 PSYCHOLOGICAL IDEALISM PART n.

acts to conceptions. Hence liis resolution of judgment
or belief into a vivid conception or idea, from which
the substitution of association of ideas for inference of
judgments immediately follows. The answer is that
judgment is an apprehension of relations, beginning
with the synthetic sense of the relations of sensible
objects, and reasoning an inference from sensitive to
rational judgments, culminating in the laws or uniform
relations of insensible objects. Judgment is not an
idea ; reasoning not an association of ideas.

Hume was misled by psychological idealism and
conceptualistic logic. Hence his scepticism about sense
and reason. His philosophy, after all, is only the most
conspicuous instance of four idealistic faults :' the con
fusion of the operation and the object of sense, the in
vention of all sorts of out-of-the-way sources of ideas
which are all the time due to sense and inference, the
postponement of reasoning, and the 'conceptualistic
supposition that conception, judgment, and reasoning
are all equally concerned with ideas, j The proper cor
rective is the study of Aristotle's ' Organon,' Bacon's
' Novum Organum,' and Newton's ' Principia.' ' The
fame of Cicero,' says Hume, ' flourishes at present ; but
that of Aristotle is utterly decayed.' Deservedly did
Aristotle's fame decay in natural philosophy. But his
logic of reasoning, widened by Bacon's theory of induc
tion and Newton's explanatory method, is necessary to
all mental philosophy. Logical reasoning from ade
quate data of sense is the main origin of knowledge,
and of ideas, and of their association.

319

CHAPTER X.
KANT'S 'CRITIQUE' AND NECESSARY TRUTHS.

KANT'S ' Critique of Pure Eeason ' l begins by assuming
Hume's theory of impressions ' : —

' That all our knowledge begins with experience
there can be no doubt ; for how should the faculty of
knowledge be awakened into exercise otherwise than
by means of objects which affect our senses, and partly
of themselves produce representations, partly rouse our
power of understanding into activity, to compare, to
combine, or to separate these, and so to convert the
raw material of our sensory impressions into a know
ledge of objects, which is called experience ? In
respect of time, therefore, no knowledge of ours is
antecedent to experience, and all begins with it.'

This passage contains the truth, which I have all
along admitted to lie at the foundation of psychological
idealism ; that sense perceives not external things in
themselves, but internal images representing them in
our senses. But, like his predecessors, Kant went on
to corrupt this truth by two assumptions. On the
one hand, he supposed the operation of sense to be
purely psychical ; on the other hand, he confounded the
representative image with the operation of represent a-

1 Critique of Pure Reason, ed. Hartenstein, p. 33 = Meiklejohn's
translation (Bohn), p. 1.

Ueberweg's summary of the Critique of Pure Eeason is printed in
an Appendix at the end of this essay.

320 PSYCHOLOGICAL IDEALISM TART n.

tjon — a confusion constantly favoured by tlie vague
abstractions of modern languages, in which representa
tion means indiscriminately both the operation of repre
senting and the representative object, sensation and
the sensible, or, in Aristotelian language, cesthesis and
cvstliema.

Hence, he started with the assumption that the
matter of sense is nothing but its own representations,
which do not exist out of the mind, and are not sensibly
apprehended as objects.1 This mere assumption vitiates
the whole work ; for, of course, if there is no sense of
objects within, reason cannot infer objects without,
and, to know objects, we must find some other origin
of knowledge. Hence, also, in the absence of adequate
data of inference, sense and reason are displaced and
divorced from one another by the intervention of an.
independent understanding, on which the main stress
is laid. Hence, finally, as understanding can act only
on sensible representations, which are not sufficient data
for a rational inference of external objects, knowledge
is limited to sensible representations converted by
understanding into objects of experience, or phenomena
of the mind. This would have been tolerable, if Kant
had started by proving that sense only apprehends its
own representations. But he did not even make it a
question. It never occurred to him that touch and
vision are operations, but the hot felt and the red
seen objects. He straightway begged that there is no
such distinction in sense, and founded the ' Critique '
on a petitio principii. Why ? Because, uncritically, he
accepted the hypothesis, that the matter of sense is
impressions, from Hume.

1 Cf. Hart. 111-20, 347 = Meik. 77-86, 307.

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 321

Of all the many errors of psychological idealism the
worst is its sequacity. Even critical idealism begins by
being uncritical. Kant seemed to delight in assuming
as data the unproved assumptions of his predecessors,
which have been already criticised in this essay. From
Descartes he accepted the confusion of subject and
soul, the imaginary power of eliciting ideas, and the
supposed psychical object of sense ; and from Locke the
deduction that all objects of understanding are psj-chical,
the hypothesis that outer sense is concerned with mere
qualities and inner sense with mere operations, the
neglect of logical reasoning, the consequent deduction
of the false conclusion that relations are a work of
understanding, and the unexplained supposition of an
unknown thing as cause of the data of sense. After
Berkeley, Kant surrendered the inconsequent deduction
by Descartes, and the inconsistent admission by Locke,
of a knowledge of physical objects, and accepted the
logical conclusion that the objects of human knowledge,
with all their qualities, primary as well as secondary,
are psychical objects of perception, and the consequent
but false identification of the perceptible and the real,
so far as known.

But Hume was Kant's main authority. They rightly
agreed in rejecting Berkeley's dogmatism about the ex
istence of mind and the non-existence of matter, and in
the revival of the real distinction made by Aristotle
between sensation and conception, in Hume's termin
ology between impression and idea, in Kant's between
intuition and conception. Along with these merits,
the critic, without a word of criticism, accepted
from the sceptic the extraordinary mass of paradoxes
about sense and the sensible, by which idealism had
become scepticism. What men call sensible objects,

Y

322 PSYCHOLOGICAL IDEALISM TART IT.

and believe to be external, what we have found to be
internal but not psychical objects, are supposed by
Hume and Kant to be not only internal but in the mind,
not objects distinct from the operation of sensation or
sensory ' representation,' as Kant would say, not sub
stantial, nor including any sensible relation of cause
and effect; in a word, impressions, nothing more.
Critical or transcendental idealism, and all the many
idealisms which have sprung from it, exist only under
the shadow and protection of Hume's scepticism ; for
all of them, without exception, start with a sense of
sensations, which has no authority except idealistic
hypothesis ending in Hume's paradox of impressions.
But we must go behind both Hume and Kant for the
data of sense.

Kant even went beyond Hume's scepticism about
the matter, which the senses receive from without,
The sceptic had doubted a sense of anything spatial or
temporal, and had denied a sense of connection ; but,
however informally, he allowed a sense of conjunction.
His critic, taking him at his word when he put forward
mere impressions as the data of sense, proceeded, logi
cally but falsely, to separate space and time from the
matter of sense, to obliterate the last trace of sensible
relation, and to reduce the matter of sense to sensible
representations or impressions, only lasting for an in
stant. Moreover, Kant was the author of the paradox
that ' the apprehension of the apparent manifold is
always successive,' and ' the manifold of appearances is
always successively produced in the mind,' 1 not allow
ing that even coexistence is sensible. According to
him, the matter of sense received from, without is

1 Critique of Pure Bcason, ed. Hartenstcin, p. 175 - Meiklejolm's
translation (Bohn), pp. 142-43.

CFIAP.X. KANTS 'CIUTIQUK' AND NECESSARY TRUTHS 323

nothing but a manifold or aggregate of unrelated im
pressions, a mere play of representations,1 a rhapsody
of perceptions.2 One wonders at last that he did not
say at once that nothing is sensible. Meanwhile, this
emasculation of the senses is not a result of any in
dependent examination, but simply the last step in the
imitation of one idealist by another. Yet it is necessary
to the argument of the ' Critique.' It is because the
matter of sense is presupposed to be mere impression
that our knowledge of objects is supposed to be due to
a priori sources. In short, Kant attempted a criticism
of pure reason without a previous criticism of the
matter of sense. After what I have said in this essay,
not against one but against all these idealistic assump
tions, I cannot be expected to enter even the vestibule
of this uncritical philosophy.

The opening of the Introduction to the ' Critique '
carries us insensibly back to the last section of Hume's
4 Inquiry ' : — ' The mind,' says Hume, ' has never any
thing present to it but the perceptions, and cannot
possibly reach any experience of their connection with
objects.'3 Hence we see the resemblance and difference
between the two philosophers. Both agree • that the
senses perceive impressions or representations. But
the point of Hume's philosophy is : given impressions,
we have not the faculties to experience objects of any
kind. The point of Kant's philosophy is : given repre
sentations, the objects of knowledge require faculties to
convert the raw material of our sensory impressions
into a knowledge of objects called experience. The
difference, however, is by no means so great as it
appears at first sight ; for Hume and Kant alike begin

1 Hart. 178 = Mcik. 145. 2 Hart. 152 - Meik. 118.

* Inquiry, 12, Tart I.

T 2

324 PSYCHOLOGICAL IDEALISM TART IT.

by assuming that the matter of sense is mere impres
sions, and end by denying a knowledge of objects beyond
experience.

How, then, from sensible representations, supposed
to be the matter of sense, do we arrive at a knowledge
of objects ? The answer of Kant immediately follows
the opening passage of the Introduction :—

' But, though all our knowledge begins with expe
rience, it by no means follows that all arises out of expe
rience. For it could well be that even our empirical
knowledge is a compound of that which we receive
through impressions, and that which our own power of
knowledge (merely occasioned by sensible impressions)
supplies from itself, an addition which we cannot dis
tinguish from the original element given by sense, till
long practice has made us attentive to, and skilful in
separating it. It is, therefore, at least a question which
requires close investigation, and is not to be answered
at first sight ; whether there exist a knowledge alto
gether independent of experience, and even of all im
pressions of sense ? Knowledge of this kind is called
a priori, in contradistinction to empirical knowledge,
which has its sources a posteriori, that is, in experience.' l

By a priori, as he proceeds to explain, he does not
mean merely deductive from the results of previous
experience, though this, or rather deductive from the
prior cause to the posterior effect, was the usual mean
ing of the phrase : what he calls knowledge a priori is
' absolutely independent of experience.' '2 It is nearly
related to what Descartes called innate. But the
novelty of Kant's theory is that even sense and expe
rience contain a priori forms. Given that mere repre
sentations are the matter of sense received from with-

1 Hart. 33 = Meik. 1. ~ Ib. 34 - Meik. 2.

CHAP. x. KANT'S ' CRITIQUE ' AND NECESSARY TRUTHS o20

out, sense requires a priori forms or pure intuitions of
space and time to receive representations in outer and
inner sense ; understanding requires a priori forms of
thought, pure notions, or categories, to convert repre
sentations into a perception and experience of objects ;
and reason requires us to conceive a priori ideas beyond
objects of sense, understanding, perception, experience,
knowledge, but cannot enable us speculatively to know
the unconditioned objects of those ideas. The Kantian
a priori theory differs from the Cartesian theory of
innate ideas by the assertion of a priori forms in
empirical knowledge and by the denial of a knowledge
through a priori ideas beyond experience.

I remarked in the first part of this essay that every
theory of the origin of knowledge is an hypothesis, which
must be tested by direct and indirect evidence ; and
that the indirect evidence must comprise both explana
tion of the known facts and elimination of other hypo
theses ; while of all things what must be avoided is
synthetical hypothesis, which, starting from the sup
posed verity of putative principles, arbitrarily dictates
and denies facts. It will be our task to apply these
logical rules to Kant's a priori theory, comparing it
with other theories of the origin of knowledge, as occa
sion may arise. In the treatment of this subject it is
too often supposed that the alternative lies between
Hume and Kant, and that an empirical origin of know
ledge means association, from which the only refuge is
transcendentalism. I shall avoid this danger, thinking
that in philosophy, as elsewhere, this is a pretty sale
rule : when opposite parties quarrel with one another
more hotly than usual, the truth lurks elsewhere.
Moreover, I have shown in the last chapter, first, that
sense is a very different thing from mere impression,

326 PSYCHOLOGICAL IDEALISM PART n.

and, secondly, that the empirical association of ideas is
quite different from, and insignificant compared with,
the empirical inference of judgments and the conse
quent conception of ideas, analogical, inductive, and
deductive ; so that there are at least two empirical
theories of the origin of knowledge and ideas, which
we may distinguish as the imaginative and the inferen
tial. Lastly, I pointed out that there are laws which
our operations mechanically obey without knowing
them, even in reasoning itself. It is evident that causce
cognoscendi are of a very complicated nature. The
choice does not lie between Hume and Kant.

Transcendentalism has no direct evidence. It sup
poses what may be called, perhaps, a self-informing
power, what Cudworth called a potential omniformity
of the mind. But, however we name it, it is a power
.of which one is not conscious. In this respect it is
inferior to all forms of empiricism, which assume only
conscious powers, such as sense, imagination, associa
tion, memory, judgment, and reasoning. Kant, on the
other hand, supposes a power of adding a priori to
a posteriori elements for reasons of his own, not on
account of, but rather in spite of, consciousness. I am
not conscious, for example, when I put my hand on the
table, that I apprehend something a posteriori as hard,
and a priori as extended : rather, I seem, as Berkeley
said, to be feeling the primary quality of extension
inseparably united with the secondary quality of hard
ness. Moreover, there is an absence of any anatomical
evidence for a self-informing power. Where is its ner
vous organ ? Not the brain in particular, which is the
general organ of sense, reasoning, will ; not the nervous
system as hereditarily adapted to perform its operations,
for quick is not a priori apprehension. When Kant

CHAP x. KANT'S ' CRITIQUE' AND NECESSARY TRUTHS 327

says that we know because we have an a priori power,
it is suspiciously like saying we know because we have
an occult power of knowing. Direct evidence, how
ever, is not absolutely necessary, and it may be urged
perhaps that the a priori stands on the same footing
as the aethereal hypothesis. But there is a decisive
difference. ^Ether is supposed to be moving according
to the known laws of motion of all bodies. But accord
ing to what laws does the supposed self-informing
power act ? The only laws at all like it are those of
pure fancy, which supplements the adventitious by the
fictitious. But the laws of fancy will not suit the a
priori hypothesis, which demands not fiction but know
ledge. The peculiarity of transcendentalism is that it
supposes a power and supposes it to obey laws of its
own. It is what Mill would call an hypothesis of both
cause and law.1

Transcendentalism really stands and falls on the
indirect evidence that the objects of knowledge cannot
be otherwise explained. Kant appeals, in the first place,
to necessary judgments. As experience examines only
many instances and not all, induction can conclude
only comparative universality, which is, after all, open
to exception. ' Necessity and strict universality,' he
concludes,, ' are, therefore, sure signs of a knowledge
a priori! 2 Now there are, according to him, necessary
judgments; for example, any proposition in mathema
tics, and the necessary connection of cause and effect in
the ordinary use of understanding. These necessary
judgments, then, will be not inductive but a priori.
Secondly, he argues that ' not only in judgments, but
even in universal conceptions, an a priori origin some
times discovers itself;' take away from the empirical

1 Mill, Logic, in. 14. 2 Hart. 35 = Meik. 3.

PSYCHOLOGICAL IDEALISM PART n.

conception of a body everything empirical, it disappears,
but the space it occupied remains ; take away from the
empirical universal conception of any object its em
pirical qualities, substance remains.1 Thirdly, he points
to universal conceptions, which have no object corre
sponding in experience, but belong to a suprasensible
sphere, where experience can, as he thinks, give no
guidance. ' These unavoidable problems of pure reason
itself are,' he says, ' God, Freedom, and Immortality.' 2

These three arguments for the a priori are stated in
the Introduction, and are the gist of the 'Critique.'
They have a common point : they all refer to objects,
supposed to require an a priori or self-informing power!
But the first appeals to necessary judgments about
objects of science, the second to objects of common ex
perience, and the third to objects beyond all experience
in a^suprasensible world. Again, the first challenges
the limits of induction, the second the limits of sense,
the third the limits of experience. To answer them'
we have to ask ourselves, indeed, whether induction'
sense, and experience are so limited ; but also, whether,
in each case, our apprehensions of the objects are
a priori. It should be noticed that there are always
two different questions to be answered, before we can
draw the transcendental conclusion; there is the
question what is not, and the question what is, the
origin of our knowledge and ideas. The negative
criticism of a given aspect of empiricism is not always
a positive proof of transcendentalism.

The three arguments require different answers. The
first is the most plausible. Induction is only probable ;
necessary judgments therefore are not merely inductive.
But it does not follow that they are therefore a priori ;

1nart.36 = Meik.4. * Ib. 37 = Meik. 4.

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS o29

on the contrary, as we shall presently find, they are
analytical judgments a posteriori. The second argument
depends, not on the logical limits of induction, but on
Hume's hypothesis of the limits of sense, uncritically
adopted by Kant. But sense is not limited to repre
sentations ; it perceives the extended, as we found in
examining Berkeley's < Theory of Vision,' and substance,
as we found in discussing Locke's ' Essay ' ; whatever
extended substance is in experience is previously in
sense, and what is not in sense is inferred by logical
reasoning from sense. The third argument depend^s on
the kind of experience which would be possible, if it
were made out of representations by a priori notions of
understanding, and were, therefore, confined to sensible
phenomena, as Kant supposes. In that case, there
would be no logical reasoning from experience of
phenomena to non-phenomenal objects. But sense,
outer and inner, apprehends internal but substantial
objects, unthinking and thinking ; experience is the sum
of sense ; and, not sense and experience, but logical
reasoning from them infers a posteriori similar "sub
stantial objects beyond experience ; God, nature made,
and man made, saved, and raised by Him. The whole
| Critique ' is a depreciation of sense and reason ; for,
if a philosopher denies the objects of sense, he destroys
the data of reason. Finally, to close this preliminary
sketch, even if we could give no positive answer to
Kant, we could at all events not accept his theory,
which confessedly limits our inferences of necessary
truths and extended substances to mere phenomena,
and our apprehension of God, freedom, and immortality
to bare ideas. He, at any rate, does not explain the
power, the extent, the grasp, of human reason, because
he has no adequate data of reasoning.

330 PSYCHOLOGICAL IDEALISM PART IT.

Of the three indirect arguments, which constitute
the proof of transcendentalism, the first is further de
veloped in the Introduction, and required throughout
the sequel of the 'Critique.' It was derived from
Leibnitz, who, in the Avant-propos of the ' Nouveaux
Essais,' had argued that necessary truths, especially
in pure mathematics, though they are occasioned by
the senses, do not depend on their evidence, but are
innate.1 Hume had, moreover, called attention to the
belief in the supposed necessary connection of cause
and effect, which he had explained away by experi
ence and association. Stimulated by the problem
of Hume, and prepared by the theory of Leibnitz,
Kant extended the hypothesis of an a priori origin of
necessary judgments from mathematics to natural
philosophy, with the special view of solving thereby
the problem of causation. At the same time, he did
more than extend the a priori theory ; he alterec^ its
character. Leibnitz had held an a priori analytical
theory of necessity, and thought that necessary truths
are innate in the sense of an analysis of our con
ceptions. Kant, agreeing that they are not inductive
but a priori, added the novel supposition that they
are not analytical but synthetical, and therefore pro
posed the question: How are synthetical judgments
a priori possible ?

At the present day, it is frequently supposed that
the question of necessary truths depends on a choice
between two synthetical theories, the a priori view of
Kant and the a posteriori view of Mill. Kant, in his
day, was at all events free from this defect. He knew
that he had to deal with Leibnitz as well as empiricists,
and directed his theory, so far as synthetical against

1 Leibnitz, Opera (ed. Erdmann), 195 A, 209 B.

CHAP. x. KANT'S ' CRITIQUE' AND NECESSARY TRUTHS 331

the former, and, so far as a priori against the latter.
There are, therefore, at least three alternatives about
necessary truths : that they are synthetical a posteriori ;
that they are synthetical a priori ; that they are analy
tical a priori. There is one more alternative : they are
analytical a posteriori.

Having, in imitation of Leibnitz, eliminated the in
ductive theory, Kant proceeded to eliminate the analy
tical a priori theory of Leibnitz, in order to establish
his own conclusion that necessary truths are synthetical
a priori judgments. An analytical judgment he defines
as one which analyses a subject into its constituent
notions, e.g. all bodies are extended ; while a synthetical
judgment is one which adds a predicate to our notion of
the subject, e.g. all bodies are heavy.1 Then he contends
that, though some necessary judgments are analytical,
all necessary principles are synthetical. He begins with
pure mathematics. From arithmetic, having selected the
sum 7 + 5 = 12, he points out that the universal con
ception of twelve is by no means thought by thinking
the union of seven and five. Pure geometry seemed to
him to contain the judgment, that a straight line is the
shortest between two points, which, as he contended, is
synthetical, because the notion of straight contains
nothing of magnitude, but only a quality, and the
notion of shortest is added to, not extracted from, the
notion of a straight line.2 ' Xatural science (Pkysica)
contains synthetical judgments a priori as principles in
itself;'3 this is his next point. Finally, he concludes
that ' metaphysics, at least as regards its end, consists
of merely synthetical propositions a priori ; ' and asks
the question, How are synthetical judgments a priori

1 Hart. 40 - Meik. 7. 2 Ib. 43 -4 = Meik. 10.

3 Ib. 44 = Meik. 11.

33^ PSYCHOLOGICAL IDEALISM PART n.

possible? He even commits himself to the extra
ordinary paradox that the solution of this problem
must determine whether metaphysics is to stand or
fall.1

Now, to resume his wThole argument from necessary
to synthetical a priori judgments : necessary judgments
are not inductive ; they are, therefore, a priori : but
there are necessary judgments, e.g. in mathematics and
natural philosophy ; they are, therefore, not inductive,
but a priori : again, they may be analytical or syntheti
cal ; now, analytical judgments are analyses of a subject
into its conceptions, and, though some necessary judg
ments a priori are of this kind, necessary principles,
e.g. in mathematics and natural philosophy, being
a priori additions of a predicate to a subject, are not
analytical but synthetical a priori. Such is the in
genious reasoning by which Kant tried to eliminate, first,
the inductive, and, secondly, the analytical theories of
the origin of necessary truths. It opens up a number
of questions ; but, as it admits the existence of ana
lytical judgments, and we have not as yet looked into
this aspect of analysis, our first anxiety must be to dis
cover what is the nature and value of analytical judg
ment, and what its limit.

Aristotle laid the foundation of the distinction
between analytical and synthetical judgments by his
investigations about simple and complex being and
intelligence (W^cris), about the axioms of being and
knowing, about the self-evident principles of demon
stration. In the ' Metaphysics ' he discussed, as axioms
of being, the principles of contradiction and excluded
middle,'2 and distinguished simple and complex being,
remarking that things simple (TO, aavvOeTa), such as a

1 Hart. 45 = Meik. 12, a Met. r. 3 seq.

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 333

unit, are objects about which we may be ignorant, but
not deceived, and either understand them altogether "or
not at all ; whereas about a combination, such as wood
being white, we may make propositions either true or
false.1 In the ' I)e Anima,' after distinguishing simple
and complex intelligence, he contended that the simple
apprehension of the essence of a thing is always true,
while the complex apprehension of something merely
belonging to it may be either true or false, e.g. a white
thing may or may not be a man.2 In the ' Posterior
Analytics ' he insists that the principles of demonstration
must be necessary, that is, self-evident,3 that the axioms
of being, though principles, are not the actual premises,4
and that the principles of demonstration are acquired
by a gradual process of sense, memory, experience,
induction, and are recognised by intellect (vovs\ of
which the obvious function is to apprehend their ne
cessity.5

I do not commit myself to the whole of this theory
of the self-evident principles of demonstration. Aris
totle did not successfully explain the power of intellect
to apprehend the self-evident, and, though he founded
the constituents, did not actually recognise the analytical
judgment. Especially I take exception to his doctrine
that the apprehension of an essence or definition is
always true. There are really two ways of arriving
at definitions, one of which I take to be on the whole
that described by Aristotle, and exemplified in the
simple definitions of mathematics ; but the other is far
more complicated, being an accumulation of facts,
followed by an explanatory hypothesis of essence ; a
way which is exemplified in the explanation of the

1 Met. O. 10. - DC An. \\i. 6. 3 Post. An. i. 46.

4 Ib. i. 11. 5 Ib. ii. 19.

334 PSYCHOLOGICAL IDEALISM PART IT.

foots of heat and light by defining them as undula
tions of aether. The omission of this second process
is a great blot in Aristotle's logic of science, which
is too much modelled on mathematics. It made him,
as Bacon remarked, fly to principles, think all scientific
principles simple and self-evident, and all science de
monstrative, or deductive from the self-evident, Hence
his anticipation of nature in natural philosophy ; for
example, his hypothesis that heat is a primary quality
of matter whose nature is simple and self-evident;
whereas it is a secondary quality, whose nature has
been discovered only after an indirect process of ac
cumulating its properties, and then explaining them by
sethereal motion. But at the bottom of these exaggera
tions Aristotle was the discoverer of a great truth. There
are self-evident truths about things, simple not synthetic,
in accordance with the principles of contradiction and
excluded middle, yet not deduced from them, discovered
a posteriori, but recognised by some power of intellect,
and forming principles of demonstration. Aristotle's is
a realistic theory of self-evident truths. It has, more
over, exercised an immense influence on modern philo
sophy, though it has become corrupted by conceptualism
and nominalism.

Even empirical philosophers admit self-evident truths,
and some of them even adopt the analytical theory of
mathematics. Among the conceptualists, Locke, at the
beginning of the Fourth Book of his ' Essay,' recognised
self-evidence under the name of c intuitive knowledge,'
which perceives the agreement or disagreement of ideas
bv themselves, e.g. that white is not black, that a circle
is not a triangle, that three are more than two and
equal to one and two ; he admitted that intuition is the
most certain kind of knowledge, and the foundation of

CHAP. x. KANT'S 'CRITIQUE1. AND NECESSARY TRUTHS 335

demonstration,1 in mathematics. But he gave no proof
that it is limited to ideas, nor any explanation of its
operation. Similarly, Hume in the ' Treatise ' admits
intuition. 'No one,' says he, 'can once doubt hut
existence and non-existence destroy each other, and
are perfectly incompatible and contrary."2 In the
1 Inquiry ' he regards pure mathematics as consisting
of propositions, which express relations of ideas, either
intuitively or demonstratively certain, and discoverable
by the mere operation of thought.3 'The conclusions,'
says he, ' which it draws from considering one circle
are the same which it would form upon surveying all the
circles in the universe.' 4 But he confined the self-
evident and demonstrative to mathematics.6 He
adopted from Locke the analytical theory of mathema
tics in a conceptualistic form, but neither of them
proved that self-evident truths express merely relations
of ideas. Mill differed on this subject from them in
two respects. In the first place, he adopted, from
Hobbes, a nominalistic view of self-evidence, regarding
all self-evident, analytical, identical, essential proposi
tions as purely verbal, stating the meaning of a name
but giving no information about a thing.6 Secondly,
he attempted to banish the self-evident entirely from
science, and went to a pitch of scepticism of which even
Hume hardly dreamt, by reducing mathematical neces
sity to probability, resulting from induction and asso
ciation. In this respect he at the same time departed
from Hobbes, who had taken up the extraordinary posi
tion that, while self-evident propositions are merely
nominal, they are principles of science, which would
make truth and falsity purely arbitrary. Meanwhile,

1 Essay, iv. 2, 1. • Treatise, iii. § 1. 3 Inquiry, § 4.

4 Ib. § 5. * Cf. ib. § 12, Tart III. 6 Mill, Logic, i. G, 4.

336 PSYCHOLOGICAL IDEALISM TART n.

Mill's nominalism did not rid him of self-evident pro
positions. He allowed that they are ' such as every
one assented to without proof the moment he compre
hended the meaning of the words.'1 Moreover, he
admitted 'the original inconceivability of a direct
contradiction,'2 without, however, seeing that it is a
negative instance entirely disproving the reduction of
all necessity to association.

There are, therefore, three theories of self-evidence,
all admitting the self-evident : the realistic theory of
Aristotle, the conceptualistic of Locke and Hume, and
the nominalistic of Hobbes and Mill; and there are
two theories among modern empiricists about necessary
truths in mathematics, the older empiricists holding them
to be self-evident, while Mill thinks them mere results
of induction and association. But before going any
further, we must first say something about Leibnitz,
whose views about self-evidence, and the self-evident
character of the necessary truths of mathematics, were
the immediate occasion of Kant's distinction of analy
tical and synthetical judgments a priori.

Leibnitz, being, even more than Locke, under the
influence of Descartes, adopted the conceptualistic
theory that self-evident truths are founded on ideas.
But his originality appeared in his attempt for the first
time to explain in detail how we apprehend their neces
sity. In opposition to Locke's criticism, Leibnitz con
tended for innate ideas, in the form that, on the
occasion of sensation, the mind by reflection finds
certain ideas in itself, and for innate principles formed
out of these innate ideas. The axiom of identity, that
which is is, of difference, that which is the same thing
is not different, of contradiction, it is impossible that a

1 Mill, Logic, i. 6, 4. - Examination of Hamilton's Phil, chap vi.

CHAP. x. KANT'S ' CRITIQUE' AND NECESSARY TRUTHS 337

tiling should be and not be at the same time, were re
garded by him as innate identical principles, from which
we deduce propositions such as, sweet is not bitter, and
a square is not a circle. To the objection that men
make such propositions without knowing the principles,
he answers that they are like the majors suppressed in
enthymemes. Finally, he regarded arithmetic and
geometry as purely innate, consisting of necessary prin
ciples analysing our innate ideas by the principles of
identity, difference, &C.1

Hence Kant's theory of analytical judgments. Ac
cording to him, an analytical judgment is obtained
a priori from an analysis of a conception by means of
the principle of contradiction, which he regards as the
supreme principle a priori of all analytical judgments.2
From Leibnitz he adopted the conceptualistic theory of
the nature, and the a priori theory of the origin, of an
analytical judgment. But he differed from his prede
cessor in thinking that the necessary principles of
mathematics are not included among such analytical
judgments, but are synthetical judgments a priori.
We have, therefore, now to find a way, if we can,
through a host of disputes, and to ask ourselves about
the nature, origin, and limits of analytical judgments.
Are they concerned with names, conceptions, or things?
Are they a priori or a posteriori ? Are they necessary
principles ?

To begin with the last point, mathematicians evi
dently use some analytical premises. The axiom, the
whole is greater than its part, is confessedly an analytical
judgment, which, according to Mill, would state the
meaning of the name, and according to Leibnitz and

1 Leibnitz, Nouveaiix Essais, i. 1 (Opera, p. 2,04 scq. ed. Erdmann).

2 Hart. 148-50 - Meik. 115-7.

338 PSYCHOLOGICAL IDEALISM PART 11.

Kant, the analysis of the conception, of a whole. Now,
to take one instance out of many, it is used as a major
premise in Euclid I. 7, twice over to prove that an angle
is greater than an angle contained in it. Again, Kant
confesses that the axioms of equality are analytical,1 and
the first of them is the major premise of the very first
proposition in Euclid, while the third is the basis of the
fifth proposition. The way of getting out of this ob
jection in the ' Critique ' is exceedingly lame. Kant,
having to admit the use of these analytical judgments
in geometry, maintains that they serve only ' for the
chain of method, and not as principles.' 2 But in
Euclid I. 7 the axiom, the whole is greater than its
part, is used as a primary major premise, and, when it
is combined with a minor premise, stating that a given
angle is a whole of which the contained angle is a part,
it produces the conclusion that the given angle is
greater than the contained angle. A confessedly ana
lytical axiom then is a primary major premise in a
geometrical deduction ; and it is a mere affair of words
whether it is called a principle or not.

Analytical judgments, being scientific principles, in
the sense of primary premises in mathematical reason
ing, are not mere analyses of conceptions, nor meanings
of names. Both Kant and Mill admit that mathematical
truths apply beyond conceptions and names to sensa
tions or phenomena, which they regard as facts, while
mechanics and all mixed mathematics prove that they
apply to the minutest particles, beyond our sensations,
conceptions, and names. But if any premise in a
mathematical deduction were about conceptions or
names, it would be a fallacy to conclude about any
thing else. The demonstration in Euclid I. 7 would be

1 Hart. 157-Meik. 124. 2 Hart. 44 = Meik. 11.

CIIAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 339

a paralogism with a quaternio terminorum, if it stood in
this form : the conception or the name, whole, is the
conception or the name, of something greater than its
part ; but the angle ACD is a whole, of which the
angle BCD is a part ; and therefore, the angle ACD
is greater than the angle BCD. The conception or
the name would never prove that an actual whole in
cludes its part, even among phenomena, much less that
a whole body is greater than its particles. Since, then,
this and other analytical axioms are principles, which
enable us to come to conclusions beyond conceptions and
names, they must themselves be concerned with some
thing more than conceptions and names. That something
more would, according to Kant and Mill, have to be
phenomena ; but really, it includes insensible things
beyond. The axiom of totality enables me, as I look at
this paper and its ink-marks, to infer that the whole
coloured surface must be greater than any one of its
black parts ; and it also enables science to infer that a
whole drop of water must be greater than any one of
its imperceptible particles. Every whole in the universe
is a case of this analytical law. Hence the conceptu-
alistic and nominalistic theories of analytic judgments
are miserably narrow ; for analytical judgments are
principles of sensible and of insensible objects. We
must return to Aristotle's realism of the self-evident.

The conceptualistic and nominalistic theories of
analytical judgments have each its peculiar error. The
former theory was caused by the Cartesian confusion of
the sensible and conceivable. Since the objects of sense
were supposed to be concerned with ideas, it followed
that analytical judgments, requiring no new experience,
could not go beyond our ideas. We have destroyed
this error from the foundation by separating sensible

z 2

340 PSYCHOLOGICAL IDEALISM PART ir.

objects from ideas. The latter theory, as it exists in
Mill's 'Logic,' is founded on a false disjunction. He
supposes that all propositions are either verbal or real,
and finding that analytical judgments, often expressing
the meaning of a name, are verbal, concludes that they
are not real. But the division of propositions into
verbal and real is defective. A verbal is not necessarily
opposed to a real proposition, a predicate does not cease
to be a characteristic of a thing by becoming the meaning
of a name, and there are some propositions which are
verbal and real, such as all bodies are extended, the whole
is greater than its part. Mill pokes fun at such a propo
sition as Omnis homo est rationalis, which expresses part
of the meaning of the name, man. But does that pre
vent men from being rational ? Again, his remark that
analytical judgments convey no information about the
thing, betrays a sad ignorance of human nature ; for most
men's simple apprehensions are miserably confused, as
you may find by asking them what is a substance, an
attribute, a body, a unit, a whole, a circle ; and one of
the main uses of analytical judgments is to make a con
fused apprehension distinct by dividing it into a subject
and the predicates contained in it. In short, the division
into analytical and synthetical does not correspond to
the imperfect distinction of verbal and real ; analytical
judgments are sometimes about names, sometimes about
conceptions, but also sometimes about objects distinct
from both ; and these latter are real. Sometimes the
same analytical judgment is at once real, notional, and
verbal, e.g. the whole is, is conceived, and means, that
which is the sum of its parts.

So far, then, we have ascertained that analytical
judgments, such as the whole is greater than its part,
are principles of science, and are accordingly not.

CHAP. x. KANT'S ' CRITIQUE ' AND NECESSARY TRUTHS 341

limited to names and conceptions, but are concerned
with sensible and insensible objects of science. Our
next step must be to find their origin. Mill has no
theory on the subject. Leibnitz and Kant have a theory,
the common point of which is that we deduce the
analysis of our conceptions from the principle of con
tradiction a priori.1 As Kant merely followed Leibnitz
in this respect, it will be best to criticise the original
authority, in accordance with the method of this essay,
which always contemplates the discovery of idealistic
errors at their first source.

Descartes had, as we found, a confused notion of
an innate power discovering ideas in ourselves, which
Locke showed to be nothing but inner sense or reflec
tion. It is an extraordinary thing that in the ' Nou-
veaux Essais,' which is an elaborate criticism of Locke's
'Essay,' Leibnitz knew Locke's theory of reflection,
and yet coolly repeats that the ideas derived from it
are innate, without taking any notice of the sensible and
presentative origin of such ideas from inner sense.
'Perhaps,' he says, 'our able author will not be entirely
removed from my sentiments ; for, after having employed
all his first book in rejecting innate lights, taken in a
certain sense, he avows at the commencement of the
second and in the sequel, that ideas which have not their
origin from sensation come from reflection. Now re
flection is nothing but attention to that which is in us,
and the senses do not give us that which we already
possess. This being so, can it be denied that there is
much that is innate since we are innate, so to speak, to
ourselves ; and that there is in us Being, Unity, Sub
stance, Duration, Change, Action, Perception, Pleasure,

1 Hart. 39-42, 148-50 = Meik. 79, 115-17.

342 PSYCHOLOGICAL IDEALISM PART ir.

and a thousand other objects of our intellectual ideas ? ' l
Locke's answer would have been simple and conclusive.
Admitting that we derive all these ideas, except that
of substance, by attention to what is in us, which is
reflection and not sensation, he had shown that this
reflection is a sense, which notices our being, unity,
&c., only because they are there to be noticed, and are
presented to it, precisely as sensation notices white or
hot when presented. To call these results of inner sense
innate is to confuse the intuitive and presentative with
the a priori and elicited.

Leibnitz made a second mistake about innate ideas,
in which Locke himself perhaps encouraged him. He
put ideas down to reflection which are not confined to
it. The correction of this mistake is of consequence,
because Locke's exaggeration of the sphere of reflection,
and the conversion of its ideas by Leibnitz into innate
ideas, gave occasion to Kant's hypothesis that time is
the mere form of inner sense and similar errors. Now, in
the list of ideas quoted above, and supposed by Leibnitz
to be innate, perception and pleasure are pure data of
reflection, but being and unity belong to all data of
sense, and to all things. Not being confined to reflec
tion, they are not innate ideas, in the Leibnitzian
meaning of this phrase. He made the same mistake
about numbers, which he supposed to be purely innate
ideas, giving rise to innate truths.2 But, as the very
hairs of our heads, so are the data of sense, and the
particles of matter, all numbered. His theory, there
fore, that number and its truths are innate, because
they are results of reflection, is not adequate to our
knowledge of universal number.

To come now to the bearing of the theory of innate

1 Leib. Opera (ed. Erdmann), 196 A. 2 Ib. 210 A, 212 A.

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 343

ideas on the origin of analytical judgments. If analytical
judgments were formed out of ideas, they would be
concerned with ideas, and, as we have already found,
they would not in that case be applied to the sensible
and insensible beyond ideas, and therefore could not be
principles of science. But Leibnitz admitted, or rather
contended, that they are the principles of science. It
follows that an analytical judgment, such as a square is
not a circle, cannot be formed, as Leibnitz thought,
purely from innate ideas, because it is applicable to
sense. In fact, it was the adoption of this theory of
analytical judgments from Leibnitz that made Kant
refuse to analytical judgments the title of principles.
But the right alternative would have been to conclude
that, since analytical judgments are universal principles
of science beyond conceptions, they are not derived
from mere conceptions.

But the most fundamental error of Leibnitz, which
Kant shared with him, was the supposed deduction
of analytical judgments from metaphysical principles
a priori. Leibnitz supposed that, in order to say a
square is not a circle, or bitter is not sweet, we must
already be in possession of the general axioms, A is A, A
is not non-A, A is not B, and so forth, which are there
fore innate principles of analytical judgments. It is
better to have no theory than a bad one ; and Locke,
though he did not probe the origin of an analytical
judgment, such as white is not black, at all events
divined that it cannot be derived by deduction from
principles, because men make such judgments in entire
ignorance of the principles. ' Who perceives not,' he
asks, ' that a child certainly knows that a stranger is
not its mother : that its sucking-bottle is not the rod,

344 PSYCHOLOGICAL IDEALISM

PART II.

long before lie knows that 'tis impossible for the same
thing to be, and not to be ? ' l

Leibnitz replies that ' one founds oneself on these
general maxims, as one founds oneself on the majors,
which are suppressed in reasoning by enthymemes.'2
But in an enthymeme we apprehend the major in
thought and suppress it in speech, usually because
the hearer will supply it himself, though sometimes be
cause we know it to be doubtful, and hope that it will
escape his notice ; moreover, we recognise the major
when expressed. On the other hand, it can hardly be
maintained that a young child apprehends but sup
presses the principle of contradiction : and it is cer
tainly false that he would recognise it when expressed.
It needs a considerable education to recognise such
principles ; and, indeed, they were rejected over and
over again by philosophers until the genius of Aristotle
established their metaphysical formulas. That same
genius established them without exaggerating them.
He pointed out that the principle of contradiction is
a condition, but not a premise of any deduction, unless
it has been denied in a particular case.3 Leibnitz, on
the other hand, and Kant after him, fell into the error
of confusing the man with the metaphysician, when
they supposed that we deduce analytical judgments
from the principle of contradiction a priori.

It does not follow that we must commit ourselves
wholly to Locke's view about the principle of con
tradiction. This and similar axioms put us in a kind
of dilemma. On the one hand, Locke shows that they
are not known a priori ; on the other, Leibnitz as
clearly shows that they are required to make any

1 Locke, Essay, iv. 7, 9. 2 Opera (ed. Era.), 211 A.

3 Ar. Post. An. i. 11.

CHAP. x. KANT'S ' CRITIQUE ' AND NECESSARY TRUTHS 345

analytical judgment. The way out of this difficulty
may be found by combining the hint of Aristotle,
that they are conditions, not premises, with the last
chapter, in which I pointed out that the laws of asso
ciation, and the axioms obeyed by reasoning, analogical,
inductive and deductive, are not premises of associa
tion and reasoning. Now, as, when the sight of a
dog recalls the idea of his master, I use the law of
association by contiguity ; as, when I reason from the
earth to Mars, I use the axiom of analogy ; as, when I
reason from dead men to the mortality of man, I use the
axiom of uniformity ; as, when I reason in the first
figure, I use the dictum de omni, in the second the dic
tum de diverso, in the third the dictum de exemplo ; but
in no case deduce either my idea in association, or
my judgment in reasoning, from the law, axiom, or dic
tum which governs the process ; so do I use the axioms
of identity, difference, contradiction, &c., when I make
an analytical judgment, such as the whole is greater
than its part, or white is not black, or a square is not
a circle ; but I do not deduce any of these analytical
judgments from these axioms, which are the sponta
neous laws of the form of analytical judgments, not
known premises to deduce them a priori.

The arguments of Leibnitz prove this and no more.
He admitted that they are not universally known, but
rejoined that ' one employs them without envisaging
them expressly,' that ' they are necessary as muscles
and tendons are necessary to walk,' and that they are
like veins in marble before they are discovered.1 But
these arguments and analogies only prove, not that
the principle of contradiction and similar axioms are
innate major premises, but that they are laws which

1 Opera (ed. Erd.), 207 B, 211 B, 213 A.

346 PSYCHOLOGICAL IDEALISM PART n.

regulate the operation of analytical judgment. The
ordinary man knows nothing about them : the meta
physician has often denied them, Plato only caught
glimpses of them, and they were never extended to the
whole universe of being and thinking, until Aristotle .
established them. In metaphysics, indeed, they are
themselves analytical judgments, and are a justifica
tion of the self-evidence of other analytical judgments,
but in ordinary thinking they are laws spontaneously
governing analysis, without being known.

If, then, we frame analytical judgments not from,
but only by, the axioms of identity, &c., from what
source do we derive them? Ultimately, by general
reasoning from sense, inductive and deductive. The
axioms alone, even if they were known a priori, would
be powerless : as it is, being only used, they are not
even major premises. Without sense and reasoning,
we should never know of anything being one and
many, whole and part, white or black, sweet or bitter,
square or round, or solid. By general reasoning we
infer that there are classes of these objects, and also
that a body moves its places in time, that a solid
body is of three dimensions, that things are one and
many, that a whole thing is greater than its parts. It
is thus we get the content of all our general judgments.
But I have confessed that induction and deduction from
induction are only probable. How, then, do we pass
from the probability of general reasoning to the neces
sity of analytical judgment? By the perfection of
rational abstraction.

There is another power in man, discovered by
Aristotle — abstraction. Abstraction has already been
mentioned in this essay. I have admitted, in the
chapter on Locke, that abstraction from sense may

CHAP X.

KANT'S ' CRITIQUE ' AND NECESSARY TRUTHS 347

conceive general ideas ; and in the chapter on Berkeley,
that abstraction is a kind of attention, which does not
form a merely abstract general idea. I have con
tended, against Locke and Berkeley alike, that it forms
a general idea of a miscellaneous assemblage of similar
individuals. But, without reasoning, such abstraction
is limited merely to a general idea of sensible objects ;
it is general, not universal. I added that there is a
rational abstraction, such that, when reasoning infers
a class of objects, e.g. of corpuscles, and rational con
ception forms a general idea of them, abstraction is
capable of attending to them. Now, because it is at
tention, abstraction is not limited to ideas, but attends
also to their objects. We may attend to names, to
ideas, and to objects of sense and reason ; and it is no
easier to attend to ideas than it is to objects. Abstrac
tion, like other powers, has suffered at the hands of
modern conceptualists.

Abstraction, as Aristotle was aware, neglects the
other characteristics of a complex object for the purpose
of isolating one characteristic, or rather the object as so
characterised. For example, there is no such thing as
a whole ; but we can neglect the other characteristics
of an object, which is, among other things, a whole,
and attend to it so far as it is a whole. Hence we often
use the formula, ' as a whole,' or ' qua whole ' — the
Latin 'qua 'being a translation of the Aristotelian "ry ."
The value of this operation, which the moderns ridi
cule as metaphysicians but use as men, is that we get
rid of the complexity of general reasoning, and are
able, by attention, to isolate a simple kind of ob
ject ; and all abstract sciences take advantage of this
isolation. Now, not in all cases, but in those objects
which are peculiarly susceptible of isolation, there is a

348 PSYCHOLOGICAL IDEALISM

PART 11.

further effect: we are able so far to isolate a simple
kind of object, that we get rid of the synthesis of
general reasoning, attend to a simple object in its com
pleteness, and apprehend its nature or essence. Thus,
general reasoning infers that a whole thing is greater
than its part ; but this conclusion is liable to exceptions,
for the thing may be absolutely simple, in which case
it has no part to be exceeded by the whole. Again,
general reasoning infers that one thing is undivided in
quantity ; but, if it is a complex body, it is also many
corpuscles in quantity, divided from one another. But
by rational abstraction we are so far able to isolate the
wholeness of a thing as to apprehend a thing qua whole
as that which is nothing but a sum of its parts ; and so
far able to isolate the unity of a thing as to apprehend
a thing qua one as the undivided in quantity, and
nothing more.

This perfect abstraction is the foundation of exact
science. The perfect abstractions of arithmetic have
just been given. In the same way in geometry,
general reasoning tells us that bodies are extended in
three dimensions, but perfect abstraction is required
to isolate the solidity of body and apprehend body qud
solid as that which is long, broad, and deep, and nothing
more. Similarly, in abstract mechanics, it is not till we
have regarded a body qud moving as simply changing
place during time, and not as possessing any particular
structure, that we can strictly apply to it the laws of
motion. There is, then, in exact sciences, a perfect
abstraction, not a priori, but founded on general reason
ing, inductive and deductive, from sense, consisting of
attention, not to an abstract idea, but to a simple object
in the abstract, and the apprehension of its nature, to
the neglect of its synthesis with other characteristics or

CHAP. x. KANT'S 'CRITIQUE- AND NECESSARY TRUTHS 349

with other objects. This power is sometimes called
intuition. But it is not intuitive any more than a priori.
It requires sense, general reasoning, and rational ab
straction ; nor is this rational abstraction always perfect ;
but when it is perfect it is a simple apprehension of
the nature of the object.

An analytical judgment is one which divides a
simple object of perfect abstraction into subject and
predicate. When we have thus got the entire content
from general reasoning, and have abstracted simple
objects, an affirmative analytical judgment simply
divides the same simple object into subject and predi
cate by, not from, the principle of identity — a thino- is
the same as itself. This operation must be carefully
guarded from misapprehension : there is no mystery
about it. In the first place, it is not merely concerned
with a common name, nor with an abstract idea, but
with an object in the abstract, discovered by reasoning,
isolated by perfect abstraction, and divided into subject
and predicate by analysis. Secondly, it is not, as
usually described, an analysis of the subject of the
judgment into the predicate, which would deprive the
latter of its content, but an analysis of the simple object
isolated by perfect abstraction into subject and predi
cate, as the object and its nature. Thirdly, it adds
nothing to the abstraction, but, as the abstraction iso
lates the simple object from the synthesis of general
reasoning, so the analysis divides this simple object into
subject and predicate. For example, having discovered
that things which are wholes contain their parts, and
having by perfect abstraction isolated a thing qud whole
as merely a sum of its parts, the analytical judgment
simply asserts this result of perfect abstraction in the
form of a judgment, for the purpose of making demon-

350 PSYCHOLOGICAL IDEALISM PART ir.

strations from it. Indeed, Aristotle was not wrong in
saying that there is a simple apprehension of simple
objects, though he ought to have added the analytical
judgment, because it is as a judgment that the appre
hension becomes a principle of demonstration. Fourthly,
the analytical judgment is made spontaneously by the
principle of identity, which is the law of its form, but
not deduced from the principle as a premise. It has
nothing a priori about it, being derived from sense and
general reasoning, through perfect abstraction, by ana
lysis, adding nothing but the division into subject and
predicate, not independent of experience, but only re
quiring no new experience ; in short, a priori only in
the old sense of indirectly a posteriori.

A negative analytical judgment is of the same kind, but
one decree more complicated. General reasoning from
sense infers that white objects are not black, that sweet
objects are not bitter, that square objects are not round,
and so forth. Perfect abstraction isolates the different
objects and causes a simple apprehension of their
natures as different. In the case of simple objects of
sense, such as sensibly white and sensibly black, perfect
abstraction is applicable, because the objects are so
simple, and the abstraction simply apprehends the sen
sibly white as containing nothing black, and rice versa.
In the case of other objects, such as things which are
square or round, the abstraction, to become perfect,
requires the neglect of many extraneous circumstances,
in order to apprehend a thing qua square containing
nothing round, and vice versa. A negative analytical
judgment, thereupon, divides the objects differentiated
in the abstract as subject and predicate of a negative
judgment, a sensible object qua white is never black, a
thins qud square is never round. Its principle is that

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 351

of difference, that which is the same thing is not different,
or two different things are not the same, or, in its more
developed form, the principle of contradiction. But
this law of the form of a negative analytical judgment
is not an a priori major premise from which any
analytical judgment is deduced, except in metaphysics
and logic as sciences.

Perfect abstraction and analytical judgments are not
unlimited. Quantitative objects are more capable of
abstract isolation than qualitative, in the narrow sense
of this word. Perhaps no precise limit can be marked
out, but we may lay down the general rules, that with
the power of isolating a simple kind of object and
apprehending its nature, abstraction ceases to be perfect,
and, when perfect abstraction fails, analytical judgment
is no longer possible. Thus we can perfectly abstract
a thing qua whole, and judge analytically that so far
it is greater than its part ; perfectly abstract the sensibly
white from the sensibly black, and judge analytically
that so far one is not the other. On the other hand,
when we come to so complicated an object as external
light, we can no longer apprehend in isolation what
light is as light, but must accumulate its facts and infer
that its nature is undulative by the method of explana
tion. Hence two origins of definition : perfect abstrac
tion in exact science, explanation of properties in other
sciences. An abstract science is one which attends to
an object, so far as characterised in some particular
manner : an exact science is one in which this abstract
attention is perfect.

An analytical is the same as a self-evident judgment,
and its necessity is self-evidence. If all other tilings
are possible, it is at least impossible that a thing should
not be the same as itself, or be the same as something

352 PSYCHOLOGICAL IDEALISM FAKT n.

different. Not metaphysics but perfect abstraction
gives this internal necessity to analytical judgments.
But metaphysics justifies it by analysing the analytical
axioms of identity and difference, and affords a technical
description, by which, if we are asked why a whole,
for example, is greater than its part, we can answer
because a thing qua whole is the same as the sum of
its parts, because otherwise it would not be a whole,
and because to deny it would be a contradiction in
terms. But such a deduction is purely metaphysical.
Nor is it a valid objection that the ordinary man could
not apprehend the necessity of his analytical judgments
unless he knows the axioms, for he is in the same
position about ordinary deduction, where he plainly
knows the logical necessity of the inference, without
knowing the axioms which it requires. Analytical
judgments, then, are self-evident, without being deduced
a priori from their axioms.

This self-evidence has several special characteristics.
In the first place, we have no apprehension of it till
we apprehend the objects, but directly we apprehend
them in the abstract we at once accept the analytical
judgment. Hence it is that there are many men, and
even nations, who have never heard of the very judg
ments which to others are self-evident, The former
have not, the latter have, performed the necessary
abstraction. A man who has not thought of a thing as
a whole has no acquaintance with the judgment, the
whole is greater than its part ; no sooner has he thought
of it qua whole, than he asks for no proof of the axiom.
The analytical theory of principles is the only one
which accounts for this extreme contrast between
ignorance and certainty. Secondly, self-evidence gives
to analytical judgments a universal applicability. They

CHAP. x. KANTS 'CRITIQUE' AND NECESSARY TRUTHS 353

are not liable to the difficulty of synthesis, that an
exception may be found to the combination of two kinds
of objects ; a difficulty which, Kant confesses, applies even
to a priori synthesis beyond objects of experience. In
an analytical judgment there is only one kind of ob
ject, which must be the same as itself and different from
other things, wherever it is found. Thus the synthe
tical judgment, a whole thing is greater than its part,
is liable to the exception that a thing may sometimes
have no parts ; but the analytical judgment, a thing so
far as it is a whole is greater than its part, can have no
exception, because qua whole it is only a sum of parts.
Thirdly, self-evidence makes analytical judgments con
vertible or coextensive ; so long as a thing is a whole it
is greater than its part, and as soon as it ceases to be
greater than its part it ceases to be a whole. We can
even say that such a judgment is of eternal application ;
for, even if things ceased to be wholes, it would still be
true that they would be greater than their parts if there
were wholes. Hence, there could not be another world
in which a whole would not be greater than its part,
for it could not be a whole ; nor can any really self-
evident or analytical judgment be reversed.

Such is the outline of a realistic theory of self-
evident analytical judgments a posteriori, of which the
points are, first, that such judgments are not always
about names and conceptions, but also about objects of
sense and reason ; secondly, that we discover the objects
by general reasoning from sense, by perfect abstraction
apprehend a simple kind of object, and analyse it into
subject and predicate by, not from, the principles of
identity and difference, or contradiction, a posteriori ;
thirdly, that analytical judgments are self-evident to
one who has abstracted the objects, universal without

A A

354 PSYCHOLOGICAL IDEALISM PART n.

exception, and convertible ; and, fourthly, that analy
tical judgments about objects of reason in the abstract
are sometimes principles of science.

As analytical principles are self-evident, conclusions
logically deduced from them are necessary, though not
self-evident, and the process of deduction from self-
evident principles is demonstrative. There are two
kinds of necessary truths : self-evident principles and
demonstrative conclusions. Again, there are two kinds
of deduction, which may be distinguished as empirical
and demonstrative, provided we remember that demon
stration is indirectly empirical. In the last chapter we
discussed empirical deduction from induction, which,
though formally necessary, is materially only as pro
bable as the induction on which it is founded. In the
present chapter we have added that deduction is not
always limited by the probabilities of induction, but,
when mediated by perfect abstraction, and starting
from analytical self-evident principles a posteriori, is
demonstrative of necessary conclusions. There are,
therefore, two kinds of knowledge : one consisting
of induction and deduction, combined together in cir
cumstantial evidence, with various degrees of proba
bility up to approximate certainty ; while the other starts
in the same manner, but by the perfect abstraction of
a simple, non-synthetic object, such as a thing qua
whole, a body qua solid, a body qua moving, &c.,
obtains self-evident analytical judgments, from which
deduction demonstrates conclusions, materially as well
as formally necessary. The former is science ; but the
latter is exact science.

Kant in the ' Critique,' and Mill in his ' Logic,' both
recognised analytical judgments and their self-evidence,
but the former was deceived by conceptualism and the

cirAr. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 355

latter by nominalism, and accordingly both fell into the
common error of excluding analytical judgments from
principles of science. In order to answer them, we
have only to remember that the axiom, the whole is
greater than its part, is confessedly an analytical judg
ment, and certainly a primary major premise in mathe
matical demonstrations. Hence it is not a mere analysis
of conceptions, still less the mere meaning of a name.
It is the analysis of an object of general reasoning iso
lated by a perfect abstraction of a thing qud whole as
a sum of its parts. This analytical a posteriori axiom,
being a real principle, is a sufficient contradictory in
stance to destroy both the theory in Kant's ' Critique '
that all mathematical principles are synthetical a priori,
and the synthetical a posteriori theory in Mill's ' Logic.'
Major est vis instantice negativce.

We found that Kant starts his argument by the
position that necessity and strict universality are not
inductive. This position is common ground. After and
beyond induction, Aristotle introduced an intelligent
understanding of principles, purposely to explain their
necessity. ' Neque tamen,' says Bacon, ' etiam in uni-
versalibus istis propositionibus exactam aut absolutam
affirmationem vel abnegationem requirimus.'1 Newton, in
the fourth ' Eegula Philosophandi,' with which he opens
the Third Book of the ' Principia,' acknowledges that in
duction is only valid 'donee alia occurrerint phenomena.'
Similarly, all that Mill contends is that ' whatever has
been found true in innumerable instances, and never found
to be false in any, we are safe in acting on as universal
provisionally until an undoubted exception appears ;
provided the nature of the case be such that a real
exception could scarcely have escaped notice.' More-

1 Nov. Org. ii. 33. 2 Mill, Logic, iii. 21, 4.

A A 2

356 PSYCHOLOGICAL IDEALISM *AKT n.

over, it is patent, from the limitation of human expe
rience to some instances out of all, that the induction of
all must end in probability, however great.

The difference between Kant and Mill begins with
the contention of the latter that there are no truths
more necessary than those mere probabilities of induc
tion which seem necessary to us only through insepar
able association. But, in the first place, Mill is not true to
his own position, because, as we saw before, he acknow
ledges ' the original inconceivability of a contradiction ' ;
though, like other philosophers, he passes lightly over
this negative instance destructive of his theory that
association is the origin of all ideas of necessity.
Secondly, he ought to have gone further than mere
inconceivability. Analytical principles of science are
such that the contradictory is not only inconceivable in
idea but impossible in belief, because it is incredible
that a thing should not be the same as itself. Now
Mill admits, on the one hand, that the impossible is
different from the inconceivable, and, on the other
hand, that association is limited to the inconceivable.
As, then, association is no origin of principles, whose
contradictions are impossible, and as self-evident ana
lytical judgments are such principles, it follows that
their necessity cannot be due to association of ideas.
Moreover, if the axiom, the whole is greater than its
part, were a synthetical a posteriori judgment, dis
covered by mere induction, with a mere idea of necessity
due to association, there would be two ideas, one of
which would suggest the other ; but there is only one
idea of one kind of object which is analytically judged
to be identically a whole and greater than its part,
Association, in fact, is no origin of the real and iden
tical necessity of an analytical principle, which is self-

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 357

evident. There are, then, necessary truths of which
the opposites are neither mere improbabilities of in
duction nor mere inconceivabilities of association, but
incredible impossibilities of existence ; namely, self-
evident analytical judgments.

Kant then was right in repeating after Leibnitz that
there are necessary judgments in the sciences ; thereby
he eliminated their synthetical a posteriori origin. But
he did not thereby eliminate their analytical a posteriori
origin. ' Necessity and strict universality are, there
fore,' says he, ' sure signs of a knowledge a priori'
That 'therefore' is a rash word. ' Baculus stat in
angulo ; ergo pluit.' There is another alternative.
Because the necessary is not inductive, it does not follow
that it is straightway a priori. Necessity is a soluble
and not an infallible sign, because there is another source
of necessity, namely, self-evident analytical judgments a
posteriori. But Kant was misled by Leibnitz into think
ing that analytical judgments were a priori. Hence
his non sequitur from the inductive to the a priori.
Hence also the importance of showing, as I have
attempted to do, that analytical judgments are a pos
teriori, real, and necessary principles. It is to found a
theory of necessity without mysticism.

Kant, in fact, eliminated analytical judgments from
the position of scientific principles, only in the concep-
tualistic a priori shape into which, under Cartesian
influences, they had been thrown by Leibnitz. He did
not eliminate them in the realistic a posteriori light in
which they were rightly regarded by Aristotle. Not
all necessary truths are a priori, because self-evident
necessary truths are a posteriori. Not all necessary
principles of science are synthetical judgments a priori,
because some analytical judgments a posteriori are

358 PSYCHOLOGICAL IDEALISM r.\ra- rr.

necessary principles of science. The analytical axioms,
the whole is greater than the part, if equals be added
to equals the wholes are equal, if equals be taken from
equals the remainders are equal, have a reality in
things, and an a posteriori origin, and a position among
Euclid's principle*2, which contradict the fundamental
hypothesis of Kant's ' Critique,' that all necessary
principles of science are synthetical judgments a priori.

Kant might reply that, though some analytical
judgments may be principles, they do not carry us far ;
and that most principles at all events are synthetical
judgments a priori ; such as 7 + 5 = 12 in arithmetic,
and a straight line is the shortest between two points.
But Kant was, to say the least of it, unfortunate in
his instances. The proposition, 7 + 5 are 12, is not an
arithmetical principle, but a demonstrative conclusion ;
and the shortest distance between two points is so far
from being the geometrical definition of a straight line
that it is not geometrical at all, being merely that
property of a straight line which is of most importance
in mechanics.

The definition of a straight line would require an
investigation of space and geometry. I will only remark
at present that Euclid's definition is at all events geome
trical, and it is unsatisfactory only because he attempted
to define a line without a superficies, committing a
blunder common with systematisers of previous dis
coveries, that of beginning too synthetically. A point
is only definable by abstraction from a line ; and simi
larly, a line from a surface, a surface from a solid, in
the manner indicated, though not completely developed,
by Dr. Simson in his Notes to the First Book of Euclid.
A straight line also requires this analytical treatment.
It has been for centuries perfectly abstracted ; but, as

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 359

often happens, it has been over-abstracted, and will
never be successfully denned until it is analytically
approached from its place in a superficies. But arith
metic comes before geometry : a unit is simpler than
a point, a number than a magnitude. As Aristotle
remarked, and Comte repeated, a science from fewer
data precedes a science which adds more.1 Accordingly,
the question of necessary truths ought to be contested
in the simpler and more universal science of arithmetic.

The arithmetical principle concerned with the number
12 is 11 + 1, which is its sole and sufficient definition.
If we were to take 7 + 5 for a definition, 12 would have
infinite definitions by the addition and subtraction of
other numbers, none of which would be of any further
use, because to use a number in a sum we must know
out of what number it is formed by the addition of a
unit. In the case of 12, 11 is that number which by
the addition of 1 makes 12, as 10 is the number which
by the addition of 1 makes 11, and so on till we come
back to 1 + 1 are 2. All those arithmetical principles,
which are definitions of numbers, are founded on the
units added together ; as the Greeks knew perfectly
well when they said that the unit is the origin of num
ber, and number is multitude composed of units.2

The discovery of abstract numbers is a good instance
of the process of abstraction and analysis I have been
describing in this chapter. By sense and reason we find
that objects are one and many and wholes, among other
of their attributes, and infer that one object is always
undivided, many are divided into units, and a whole is
greater than its part. We thus discover truths of num
ber. But how do we apprehend their necessity ? By
perfect, abstraction we isolate an object ^a one asundi-

] Ar. Post. An.i. 27. ' Eucl. VII. Def. 2.

3GO PSYCHOLOGICAL IDEALISM TART IT.

vided in quantity, objects qua two as one -\ one, &c., &c.
Tliis abstraction is necessary to the science of arithmetic.
As Plato, though lie did not understand abstraction,
long ago pointed out,1 concrete units are not altogether
undivided ; a man, for example, is many in his members
and only one on the whole ; but an arithmetical unit is
absolutely undivided. Why ? Simply because the thing
as divided is neglected, and attended to only as undi
vided, by perfect abstraction. On this abstraction of
the unit, not as a mere conception, but as a simple
object of attention, we have, not a priori, but by a
posteriori analysis, the analytical judgment, which is
the definition of a unit : not, be it remarked, the con
tingent proposition, one thing is the undivided in
quantity, which is not always true ; but a thing qua
one is the undivided in quantity, which is self-evidently
necessary. So far as a thing is one, it is undivided in
quantity, and so far as it is divided in quantity, it is no
longer one. This analytical definition is the foundation
of all arithmetical definitions, all of which are merely
analyses of numbers into units ; thus 1 + 1 are 2 ; 2 + 1
are 3, and so forth ; every one of which are analytical
definitions. Hence, though 7 + 5 is not, 11 + 1 is, the
analytical definition of 12. All things, qua 11 + 1 are 12,
and qud 12 are 11 + 1.

Mill, indeed, contends that there is a difference
between 2 + 1 and 3, because 'three pebbles in two
separate parcels, and three pebbles in one parcel, do
not make the same impression on our senses.' 2 But he
overlooks the fact that, when three pebbles are in two
separate parcels, if they give us the impression 2 + 1,
this is the impression 3 without any comparison with
three pebbles in one parcel ; and conversely, when three

1 Plato, Eep. vii. 525 D-G B. 2 Mill, Logic, ii. 6, 2.

CIIAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 301

pebbles are in one parcel, if they give us the impression
3, this is the impression 2 * 1, without any comparison
with three pebbles in two parcels. We do not require two
sets of three objects each to count 2 and 1 are 3. The
truth is that he was deceived by the formula 2 + 1 = 3,
in which, for mere convenience, we apply 'to number the
geometrical sign for equality of two magnitudes ; but
we must not allow this mere symbol to make us think
that we are always comparing different quantities on
each side of it; in arithmetic, equality means identity,
and the correct arithmetical formula is 2 + 1 are 3.

Kant, on the other hand, did not even take the
definition of the number 12, which, as we have seen, is

11 + 1, but one of its many properties, 7 + 5. He rightly
says that the proposition, 7 + 5 are 12, is not analytical :

12 is not the selfsame thing as 7 + 5, because it is
8 + 4, &c. But this proposition, though not analytical,
is also not a principle, but a demonstrative conclusion
from principles which are analytical, the definition of
the unit and the definitions of the numbers up to 12, as
11 4 1 ; and we are able from these analytical to demon
strate synthetical judgments, by that combination which
we found in the last chapter to be the essence of syllogism
or deduction. Kant's attempt to prove that the prin
ciples of arithmetical demonstration are not analytical
by the instance 7 + 5 are 12, is an ignoratio elenchi, be
cause this proposition is not a principle, but a demon
strative conclusion from analytical principles, including
11 + 1 are 12.

It is curious what a cursory attention is paid to
arithmetic in Kant's 'Critique' and Mill's 'Logic.' But by
looking a little more closely into this most fundamental
of all special sciences, we have found that it contains
analytical principles a posteriori both in the axiom, the

3G2 PSYCHOLOGICAL IDEALISM

PART IT.

whole is greater than its part, and in its definitions.
Thus we 'can destroy both the synthetical theories. On
the one hand, as these principles, being self-evident,
are such that the contradictory is impossible, Mill is
wrong in reducing arithmetic to the mere probability of
induction and association. He quotes, indeed, with
approval a supposition that there might be a world,
in which, whenever two pairs of things are contem
plated together, a fifth thing is brought within con
templation, and the result to the mind of contemplating
two two's would be to count five.1 But it is absurd
to suppose minds contemplating a fifth thing without
counting it in the enumeration, and yet to end the sum,
as if they had counted it, with the number 5. Either
one would count the fifth thing, in which case the sum
would be 2 + 2 + 1 are 5, or one would not, in which
case the sum would be 2 + 2 are 4. There can be no
world in which the result to the mind of contemplating
two two's would be to count five, because 2 + 2 are de-
monstrably 4, and 4 + 1 are identically the same as 5.
On the other hand, as necessary arithmetical principles
are a posteriori analytical judgments, we cannot follow
Kant in passing from the synthetical a posteriori to the
a priori synthetical theory ; for a definition, such as
11 + 1 are 12, is discovered by empirical reasoning, and
by perfect abstraction and analysis becomes a self-evi
dent principle, whereby 7 + 5 is 12 are demonstrated.

Finally, if we were to surrender entirely the analy
tical a posteriori origin of necessary truths, yet the
synthetical a priori origin is an untenable hypothesis,
because it does not explain the facts. Let us take for
granted the Kantian series of arguments : the neces
sary is not inductive, therefore it is a priori ; there are

1 Examination of Hamilton's Philoso2)Jiy, chap. vi. note.

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 3C3

necessary principles in the sciences, therefore they are
a priori ; analytical judgments are merely a priori ana
lyses of conceptions, but principles of science are true
beyond conceptions, therefore they are never analytical
judgments : but if they are neither synthetical a poste
riori, nor analytical a priori, all principles of science are
synthetical a priori. Now, everywhere throughout the
' Critique,' Kant confesses that the a priori is contri
buted by mind to mental representations, and that the
data of mental representations, without which the a
priori is mere conception, are sensations, which the a
priori converts into objects of knowledge. Hence he
concludes that perception, experience, understanding,
reasoning, knowledge, science are all confined to
sensible representations informed by a priori elements.
Hence, according to him, necessary principles of science,
being synthetical a priori, are necessary within, but
impossible without, the sphere of sense and experience.
Kant everywhere accepts this consequence : synthetical
principles a priori are necessary, and apply, only within
the limits of phenomena.1

This corollary of transcendentalism maybe illustrated
by its application to arithmetic. According to Kant,
arithmetic will contain analytical a priori axioms -
for example, the whole is greater than its part— which,
however, will not be principles ; and synthetical prin
ciples a priori, an example of which will be 7 + 5 are 12
He did not, indeed, leave a satisfactory theory of the
place of number in his system. There is a sentence in
the ' Critique ' 2 in which he says ' that number is no
thing but the unity of the synthesis of the manifold of
a homogeneous intuition in general, gained by my gene-

1 Hart. 57, 152-3, 208 = Meik. 44, 119, 177.
2 Hart. 144 = Meik. 110.

3G4 PSYCHOLOGICAL IDEALISM TART IT.

rating time itself in the apprehension of the intuition ; '
that is, apparently, by generating a successive addition
of units in time. The same view is confirmed by a passage
in the c Prolegomena to all Future Metaphysics.' ' Geo
metry,' says he, ' is based upon the pure intuition of
space. Arithmetic accomplishes its concept of number
by the successive addition of unities in time.'1 This,
however, is a conclusion so paradoxical, that we may
in charity suppose him to concede that we also apprehend
contiguous units in space. But even so, space and time
alike are regarded by him merely as a priori forms of
sense and of sensible phenomena. Moreover, the cate
gories, schemata, and principles of quantity are all
confined by him to phenomena. The consequence is
that number is strictly limited to phenomena, and even
the synthetical principles a priori of arithmetic are re
garded by Kant himself as necessarily true of phenomena
of sense, and no more. Hence his extraordinary state
ments, ' numerus est quantitas phenomenon,' and
' ceternitas, necessitas, phenomena,' &c.2

But is it true that the laws of number are limited
to the phenomena of sense ? The very first definitions
of Newton's ' Principia ' disprove such a narrow theory.
' The quantity of matter is the measure of the same
arising from its density and magnitude conjointly : ' this
is the first definition, which is immediately illustrated by
the arithmetical proposition that ' air of a double density
in a double space is quadruple, in a triple space sex
tuple,' while this quantity of matter is identified with
its mass. ' The quantity of motion is the measure of
the same arising from the velocity and quantity of
matter conjointly : ' this is the second definition, which

1 Prolegomena (translated by Mahaffy), p. 45.
2 Hart. 146 = Meik. 113.

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 305

is again elucidated by the arithmetical illustration, that
' in a body, double in quantity, motion with equal velocity
is double, with double velocity quadruple.' If, now,
arithmetic were limited to the phenomena of sense, the
laws of mass and motion would either have to be limited
to the phenomena of sense, or, beyond the phenomena
of sense, contain no quantity of matter or of motion, no
measure, no numerical proportion ; both of which alter
natives are absurd.

The truth is that the laws of mass and motion carry
us far beyond the phenomena of sense into a 11011-
phsenomenal yet scientific world of material particles,
and carry the laws of arithmetic with them. The law
of gravitation is a law of motion by numerical propor
tion. All the particles of matter gravitate to one an
other with a force directly as their mass, and inversely
to the square of the distance ; on the one hand, this
gravitation is inferred to be in numerical proportion
both to the quantity of matter and to the distances of
the particles; on the other hand, every particle of
matter in the universe is inferred to gravitate with this
numerical proportion, in times, places, and circumstances,
wholly inaccessible to any possible senses of living
beings. In the laws too of the structures and motions of
imperceptible particles, all the definitions and axioms
of arithmetic are employed. For example, in a drop of
water, every thousand of the imperceptible particles
with another particle makes one thousand and one, and
is a whole including every one of these particles as parts.
'In rebus enim,' says Bacon, ' qiue per numeros
transiguiitur, tarn facile quis posuerit aut cogitaverit
milleiiarium, quam imum ; aut millesimam partem
unius, quam uimni integrum.' l As Mill remarks, ' the

1 Bacon, Nov. Org. ii. 8.

3G6 PSYCHOLOGICAL IDEALISM TART. n.

great agent for transforming experimental into deductive
sciences is the science of number.' 1 We can neither
allow that these deductive sciences, which measure the
structures and motions of imperceptible particles, are
limited to phenomena of sense, nor that there is any
measure of quantity available except number.

But we need not go beyond the ' Critique' itself to see
the impossibility of limiting number, the quantitative
categories of unity plurality and totality, and arith
metical principles, to phenomena. One of the points
of the 'Critique' is that God is not a phenomenon.
But God is one. Therefore unity is not limited to phe
nomena. Kant also assures us that we have a unity of
apperception, an identical self, which is supposed by
him to be not a phenomenon, but that which unites
phenomena. Lest we should suppose that these objec
tions prove only unity and not number beyond phe
nomena, he distinguishes for us the human understand
ing with its unity of apperception to combine sensible
representations from the divine understanding, which
does not require it.2 There are, therefore, according
to Kant, who was innocent of the Hegelian identifica
tion of similars and confusion of divine and human,
two understandings, the divine and the human, numeri
cally different, yet neither a phenomenon. God, he
also tells us, does not make a whole with the world ;
there are, therefore, three things — God, the world, and
human understanding ; none of them phenomena.

All things are at least numbered, whether they be
material or spiritual ; hence the dispute, whether Kant
ought to have made number belong to space or to time,
is completely beside the mark, for it belongs to every
thing whatever. It is impossible, therefore, to confine

1 Mill, Logic, ii. 4, 7. 2 Hart. 119, 123 = Meik. 85, 89.

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 367

number, arithmetic, or arithmetical necessity to phe
nomena of sense. What is the consequence ? Not all
necessary principles are limited to phenomena. Con
sequently, again, they cannot be synthetical judgments
a priori, which, on Kant's own confession, would limit
them to phenomena. In other words, the synthetical
a priori theory does not account for arithmetical neces
sity, the simplest and best instance of scientific neces
sity, beyond phenomena in an imperceptible world.

Arithmetical principles apply to everything what
ever. After all, there is only one theory which can
account for this absolute universality of arithmetic, which
counts subjects as well as sensations, men, bodies, cor
puscles, and God Himself. There is not one arithmetical
judgment limited to phenomena any more than to ideas.
Now, this could not be, if they were analytical a priori,
which would limit them to ideas, nor synthetical a pos
teriori, which would make them contingent, nor syn
thetical a priori, which would make them necessary
only within the limits of phenomena. But it can be,
if they are analytical a posteriori judgments about
simple objects of reasoning in the abstract. Either,
then, this theory must be accepted, or some new theory
found. But where ?

When we look back on the whole discussion of this
difficult subject, we shall find that there is no evidence
for the Kantian hypothesis of a priori synthetical judg
ments, as the origin of necessary truths, except its
advantage over the synthetical a posteriori and the
analytical a priori theories. It has no direct evidence,
either from consciousness or from anatomy, and it is
not only that we are unconscious of a priori necessity,
but that we are unconscious of any a priori power, or
of anything like it. In indirect evidence it also fails. It

368 PSYCHOLOGICAL IDEALISM PART n.

does not attempt to eliminate an analytical a posteriori
theory, although such a theory, as I have shown, can
Teadily be developed from the works of Aristotle. But
what filially condemns it, and makes it quite impossible,
is its confessed inability to explain even the logical in
ference, much more the scientific knowledge, of neces
sary truths beyond the phenomena of sense ; when, as
a matter of fact, in all the sciences, and not only in
mechanics and all natural philosophy, but also in psy
chology and theology itself, insensible and impercep
tible objects are logically inferred and known to obey the
necessary laws of unity, plurality, and totality. Every
thing known is one ; not everything a phenomenon.

What makes so many philosophers at this moment
cling to an hypothesis so utterly wanting in verification,
elimination, and explanation? Partly, no doubt, its
superiority to the hypothesis of Mill. But two blacks
do not make a white ; and it makes little difference
whether we say that association makes us necessarily
conceive, or a priori synthesis necessarily believe, the
necessity of principles within the phenomena of sense,
when the real question is how we infer their necessity
in insensible and imperceptible nature, and in the super
natural world. Partly, it is thought that the Kantian
theory of necessity must be accepted, because other
parts of the ' Critique ' seem to support religion. But
we must beware of building the house of religion on
the sand ; and religion can hardly be supported by a
philosophy, which makes it a fallacy to say that God
is one. The main cause of the popularity of Kant's
philosophy, however, seems to be founded on the vague
use of the term ' phenomena,' which suggests to the
unwary all the facts in heaven and earth, sensible, in
sensible, and imperceptible. But this is not what Kant

CHAP. x. KANT'S ' CRITIQUE' AND NECESSARY TRUTHS 3G9

meant, nor what lie could mean, by phenomena, and
it would be a sad pity to rest the reputation of a
philosophy on an equivocation.

As Leibnitz before him spoke of 'phenomena sive
apparitiones quae in mente mea existunt,' l so Kant
always speaks of them as sensible representations which
cannot exist out of our mind ; opposing them to nou-
mena, or things of which we must form ideas, but
which as objects are unknown. He was aware that his
philosophy compelled him to make these sensible repre
sentations the limit of knowledge, not merely because
they are the matter of sense, but also because a priori
forms of mind cannot be valid beyond a posteriori data
of mind. Moreover, as we find from the Preface to the
Second Edition,2 he looked upon it as one of the ad
vantages of his ' Critique of Pure Eeason,' so to limit
speculative knowledge to phenomena that 'we can have
knowledge of no object as a thing in itself, but only so
far as it is an object of sensory intuition, i.e. as manifes
tation,' because thereby he thought to make room for a
practical proof of the freedom of the will beyond the
area of phenomena.

Kant, then, in limiting all speculative knowledge to
phenomena, meant that necessary truths, being synthe
tical judgments a priori, are only necessary about the
a priori forms of sensible representations and about
sensible representations converted into objects of know
ledge by these a priori forms. Such a limitation to
things of sense, Sinnenwesen, phenomena, is far too
narrow, because arithmetical necessity applies to every
imperceptible object of logical reasoning and scientific
knowledge. His fundamental position, that necessity is

1 Leibnitz, Op. (ed. Erdmann), p. 442, A.

2 Hart. 22 scq. = Meik. xxxii. scq.

B B

370 PSYCHOLOGICAL IDEALISM r.\ui n.

an infallible sign of a priori knowledge, must be tra
versed by this still more fundamental position : imper-
ceptibility is an infallible sign of a logical inference and
a scientific knowledge which is neither phenomenal nor
a priori.

Kant's ' Critique of Pure Eeason ' is a conspicuous
instance of the failure of the synthetic method, and
indeed of the impossibility of carrying it out con
sistently. He supposes himself to use the origin of
knowledge to determine the limits of the objects known.
Accordingly, on this synthetic method, he begins with
sense, at once begs a sense of sensible representations,
and thus founds his philosophy on an hypothesis which
dictates the conclusion that knowledge is limited to
phenomena. On the other hand, every one of his main
arguments takes a premise from the other end of know
ledge, its objects, and, by an analytic method, uses the
objects to infer an a priori origin of knowledge. Thus,
in the Introduction, necessary truths about objects of
science are used to deduce the theory of synthetic
judgments a priori-, in the 'Transcendental ^Esthetic,'
from their known properties space and time are inferred
to be a priori forms of sense ; in the ' Transcendental
Analytic,' a definition of the objects of knowledge is used
to prove that they contain a priori categories of under
standing. Nor is this all. Having taken as much about
an object as he wants for his a priori theory, Kant then,
by his synthetic method, uses his a priori theory to dis
pose of the rest of the object. Thus, he argues that
necessity requires synthetical judgments a priori, which
again prove necessity phenomenal ; that the properties
of time and space require a priori forms of sense, which
again prove time and space phenomenal ; that known
objects require a priori categories, e.g. substance and

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 371

cause, wliicli again prove known objects, e.g. substances
and causes, phenomenal.1

The ' Critique ' is a perpetual see-saw between two
methods ; the professed from the origin to the objects,
and the concealed from the objects to the origin of
knowledge. It is first synthetical, then analytical, and
finally synthetical. It assumes as a principle that the
matter of sense is representations. But this synthetic
beginning would not justify transcendentalism. It then
argues that the objects of knowledge require a priori
elements. Now, this analytical procedure gives trans
cendentalism a momentary plausibility. It finally con
tends that a posteriori representations converted into
objects by a priori elements are the objects of experience,
and that all objects of logical inference and knowledge
are phenomena. But this synthetical ending brings
transcendentalism into conflict with a characteristic
of the objects of knowledge, omitted in the analysis ;
namely, that they are not limited to phenomena. It is
as if a natural philosopher should show that the theory
of emission explains reflection and refraction, and then
deny the interference of polarised light. So Kant
shows that the a priori theory explains the necessity
of synthetical truths, and then denies their universal
applicability ; shows that it explains the properties of
time and space in, and then denies them beyond, sense ;
shows that it explains the experience of objects, and then
denies the knowledge of objects beyond experience.
He arbitrarily appeals to some of the characteristics,
but neglects the insensibility, of objects of science. His
whole method is ad placitum. He makes origin and
objects, objects and origin, origin and objects, recipro
cally determine one another, in a perpetual circle.

1 Hart. 22, 80, 123-4, 133-4, &c. = Meik. xxxiii. 44, 90, 100, &c.

B B 2

372 PSYCHOLOGICAL IDEALISM TART n.

The analytic method, used consistently, makes com
plete havoc of the ' Critique.' ' Tempus, spatium,
locus, et motus sunt omnibus notissima. Notandum
tamen, quod vulgus quantitates hasce non aliter quam
ex relatione ad sensibilia concipiat.' 1 In this passage,
Newton points out that the limitation of the objects of
science to the sensible is a vulgar error. Yet it is
a constant error of mental philosophers, who think that,
when they have considered only objects of sense, they
have solved the secret of the scientific universe. It
was the very error of Kant when he called time and
space forms of /sense, and therefore limited motion to
sense ; when hettnunciated the extraordinary series of
paradoxes : ' numerics est quantitas phenomenon, sensatio
realitas phenomenon, constans et perdurabile rerum
substantia phenomenon, ceternitas, necessitas, pheno
mena, &c.' 2 ; when he concluded that whatever is known
is a phenomenon, and what is not a phenomenon can be
conceived by pure reason, but neither inferred by logical
reason nor scientifically known. This philosophy, sccun-
dum sensum, was an hypothetical corollary from the
theory that all objects of experience are sensible repre
sentations informed by a priori intuitions and notions of
mind. But as certainly it is false, because it cannot
explain a millionth, nor even an infinitesimal, part of
the insensible objects of science.

Let us return from the ' Critique of Pure Eeason '
to the ' Philosophie Naturalis Principia Mathematical
and enlarge our thoughts, not to the immensity of the
unknown, but to the extent of the objects of science,
such as was made known by Newton,

Felix, qui potuit rerum cognoscere causas.

Insensible corpuscles of matter are scientifically inferred

1 Newton, Principia, Def. Scholium. ~ Hart. 146-Meik. 113.

CHAP. x. KANT'S-' CRITIQUE' AND NECESSARY TRUTHS 373

to be each one, many and numbered, and to obey the
necessary laws of number ; to be in insensible time and
space, and to obey the necessary laws of magnitude ; to
be insensible substances, and move according to the laws
of motion ; to be insensible causes of insensible motions,
and to be insensible causes of sensible objects. They
are actual, but they are not actual phenomena of sense,
but insensible external causes of internal sensible ef
fects. Nor, being actual, are they possible phenomena ;
for possible, which are not actual, phenomena are
nothing at present, whereas the insensible particles of a
drop of water, now gravitating towards my hand, are
actual at present, real because causal of effects in my
senses, and at the same time not only insensible but
imperceptible ; so far from possibilities, impossibilities of
sensation ; yet actual objects of science. The irresistible
conclusion of this consistent and thorough appeal to
the objects of knowledge, is that not all of them are
phenomena, actual or possible, but far the larger part
non-phsenomenal, noumenal in the sense of rationally
inferred, known things in themselves, as apart from our
senses, though not as apart from their relations to one
another.

The weight of natural philosophy is destined to
destroy all that mental philosophy of the present day
which begins from a sense of sensations, even if it
makes a vain effort to recover its false start by catching
at the shadow of a priori mysticism. When Kant pro
poses to convert sensible representations of minds into
objects of knowledge by a priori intuitions and notions
of minds, he all the more limits scientific inference to
phenomena. But, as we have just seen, the objects of
scientific inference include insensible and imperceptible
things, which are not phenomena, actual or possible.

374 PSYCHOLOGICAL IDEALISM

PART IF.

Therefore, in the first place, the ' Critique ' is incredibly
narrow and absolutely false in limiting scientific infer
ence to phenomena, actual and possible ; and secondly,
the data of scientific inference cannot be sensible re
presentations, even sublimated by a priori forms, which
all the more surely would condemn science to the nar
row limits of phenomena.

Finally, we must apply consistent analysis to the
Kantian arguments in detail. Thus, if necessary truths
were synthetical judgments a priori, they would be
limited to phenomena ; but science extends them to
all particles of matter ; therefore, they are not limited
to phenomena, and therefore are not synthetical judg
ments a priori. Secondly, if time and space were such
as to be necessarily a priori forms of sense, they would
be limited to phenomena ; but science infers that they
are forms of every particle of matter in the universe ;
therefore, they are not limited to phenomena, and
therefore they are not a priori forms of sense. Thirdly,
if objects were such as to require a priori categories of
substance and cause, they would be limited to phe
nomenal substances and causes ; but science infers that
all particles of matter are substantial bodies, causing
and receiving motions, acting and reacting on one an
other, inert until moved or resisted by one another,
and, among countless effects, producing sensible objects
in us ; therefore, objects scientifically inferred are not
phenomenal substances and causes, and, therefore, sub
stance and cause are not a priori categories. In short,
objects of scientific inference are not phenomena, and
could not be inferred from a posteriori sensations con
verted into objects by a priori forms. Critical idealism
is a false philosophy, both of the limits and of the origin
of knowledge.

CHAP. X.

KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 375

The one beacon of the present day is scientific dis
covery and invention : it was lighted by the principles
of the Newtonian philosophy. But though natural philo
sophy will reveal to us nature, and provide us external
goods, it will not alone produce philosophical wisdom
or constitute essential happiness. What we want is
principles in general philosophy. When such princi
ples have been found, it will be discovered that there
was a time when the details of nature were not so well
known, but the general relation of God, nature, and man
was much better understood than at present. We may
laugh at the want of knowledge, but we must never
forget the wisdom of the ancients. The stream of
human discovery has been like a river, part of which
escapes into marshes, while the main channel flows on
into the sea : so philosophy, the perennial sources of
which are to be found in Greek philosophy and sciences,
speculative and practical, has in modern times been
partly diverted into the marshes of idealism, while the
main stream has expanded into the natural philosophy
of Copernicus and Kepler, Bacon and Galileo, Descartes
and Newton, and perpetually issues in discoveries and
inventions.

Can we bring mental philosophy back into the main
stream of discovery ? We can, by using the discoveries
of natural philosophy as objects of science to discover
the data of sense. Idealism has failed because it has
used a wrong method, and begun at an unknown be
ginning. It has taken psychical data of sense for prin
ciples, which are really hypotheses, and has used them
to dictate the objects of knowledge. As it has found
new difficulties, it has feigned new hypotheses, until it
has culminated in the absolute idealism of Hegel, who,
by heaping hypothesis on hypothesis,— sensible repre-

PSYCHOLOGICAL IDEALISM

PART IT.

sentations, a priori categories, one spiritual subject
in God and all men, of which nature is a system of
objective thoughts — compiled a system of philosophy
which is as cumbrous a mass of hypotheses as the
Ptolemaic system of astronomy. But the truth is, that,
like the old hypothesis of planetary circles round the
earth, the modern idealistic hypothesis of a sense of
psychical data, whether called ideas or impressions,
representations or sensations, is a false beginning, and
could never lead to scientific knowledge.

Modern astronomers succeeded by reversing the
method of astronomy. They gave up reasoning synthe
tically from hypotheses of planetary circles to the de
tails of planetary motion, and began with the planetary
motions as facts. Copernicus found that the planets
, move round the sun, and Kepler that they move not in
circles but in ellipses: proceeding from these facts,
Newton inquired analytically what simple motions were
required to explain such elliptical motions : this was the
analytical method which ended in Newton's discovery
of astronomical principles. In the same way mental
philosophy should reverse its method. Instead of reason
ing synthetically from hypotheses of sensible data to
what objects we can and must know, we should find
what we do infer and know in the sciences, and then
inquire analytically what sensible data are required to
explain our inference and science. In this way, and no
other, as Newton by an analysis of elliptical motions
discovered the principles of astronomy, so may we by
an analysis of the objects of scientific reasoning dis
cover the principles of mental philosophy.

There is one characteristic of objects scientific,
which is at once a positive instance to bring us to
principles of mental philosophy, and a negative instance

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 377

to destroy all psychological idealism : it is the insen
sibility of corpuscles. Corpuscles, insensible and imper
ceptible substances in time and space, moving according
to laws of motion, are physical objects of science,
requiring physical data of sense. On the one hand,
V consider this analytical deduction destructively. In
/ ilrcH-ka^glace, it follows that these insensible and
(imperceptible objects of scientific inference are not
Sensible ideas, not perceptions, not phenomena, nor
unknown things. This consequence destroys the ideal
isms of Berkeley, Hume, and Kant. Secondly, these in
sensible and imperceptible objects of scientific inference
could not be logically inferred from ideas, nor from im
pressions, nor from sensible representations converted into
phenomenal objects by a priori forms. This destroys
the idealisms I have examined, from Descartes to Kant.
Nor could they be inferred from any sense of sensations,
however elaborated. This destroys the idealisms of
our own day. On the other hand, consider this analy
tical deduction constructively. These insensible and
imperceptible objects of scientific inference require a
sense, from which reason may infer them by parity of
reasoning. Hence, in the first place, sense is simple and
synthetic ; perceiving substantial objects, which are
internal but physical, durable, extended, and related
to one another, within our nervous systems. Secondly,
reason infers similar objects and relations in external
nature. There are three types of inference, analogical,
inductive and deductive, each mechanically obeying its
own laws, and primarily from judgments of synthetic
sense inferring judgments about objects insensible and
imperceptible. General reasoning, inductive and de
ductive, used circumstantially, produces a knowledge
and science with an approximate always tending to

378 PSYCHOLOGICAL IDEALISM TART ir.

an absolute certainty. After sense and general reasoning,
perfect abstraction by attention, mechanically obeying
the laws of identity and difference, makes us apprehend
a kind of object as simply the same with itself, and
frame analytical judgments of what it must be and not
be, in the abstract. These analytical judgments are self-
evident principles of demonstration, producing exact
science, but a posteriori. This is a general outline of
the analytical philosophy attempted in this essay.

Among many difficulties, which may occur to
others, I anticipate three main lines of objection to this
essay. In the first place, it may be thought that, what
ever value physical realism may have in dealing with
nature as an object of scientific knowledge, idealism
retains an advantage of its own in its treatment of man
as a spiritual subject. On the contrary, against Des
cartes and all his followers, but from the consilience of
consciousness, observation, and reasoning about myself,
I contend that man is an organism, partly body and
partly soul; who knows of himself, on the one hand,
that he is an animal, inhabiting the surface of no very
large planet in a considerable solar system, which is
only one among countless stellar worlds, in a stupen
dous immeasurable universe ; and, on the other hand,
that, infinitesimally little as he is in himself, by sense
and reason he is great, in his knowledge and power
over nature, which make him like even to God. But,
reasonable as is this realistic, but not materialistic, con
clusion, the idealists more and more tend to the hypo
thesis, that man is a purely psychical self, while his
own body is not an integral part of himself, as subject,
but is one among all other known bodies, which, as
objects, are either all alike inferred from, or all alike
identical with, his sensations or thoughts in general.

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 370

Now, the former of these alternatives leaves out half
the man ; the latter inverts man and nature ; while
both idealistic theories of personal identity, by draw
ing the line of self at spirit, or between soul and body
instead of between man and nature, contradict the con
silience of consciousness, observation, and reasoning :
a combined evidence not to be parted, because man
is a complex being, mainly imperceptible to himself,
who by night falls asleep and becomes oblivious of his
being, by day does not remember his infancy, never
can remember nor as yet be conscious of his future
career, and, therefore, is not aware of his personal iden
tity throughout life by retentive consciousness alone.
Why, then, does modern thought tend towards idealistic
spiritualism ? Partly from a want of simplicity and a
certain vanity of man, who in his rationality would
fain forget he is an animal ; mainly from a confusion
of idealism and spiritualism with Christianity. But we
have the best possible authority on the Christian doc
trine of man himself: the words of Christ incarnate.
'And He said unto them, why are ye troubled? and
wherefore do reasonings arise in your heart? See
My hands and My feet, that it is 1 Myself : handle Me,
and see; for a spirit hath not flesh and bones, as ye
behold Me having.' l

Secondly, it will be doubted whether my general
theory can be worked out in detail. I feel the full
force of this difficulty. When this essay was mapped
out it was to include many more details. Starting from
Newton's < Principia,' I had hoped to include a theory,
long cherished, that the properties of time and space,
enunciated by him in the ' Scholium to the Definitions,
can only be explained by defining time and space as

1 St. Luke xxiv. 38-39 (Revised Version).

380 PSYCHOLOGICAL IDEALISM PART n.

the continuance and continuity of tlie universe, the
former being the continual duration, the latter the
continuous extension, of the universe as one substance
including many substances, from a star to an atom.
From the same starting-point, I hoped to show that a
necessary physical cause of motion is any body which
displaces or resists any other body, as demonstratively
following from the impenetrability of matter, by which
two bodies cannot at the same time occupy the same
place in space, and that again from the analytical judg
ment that a body as solid is extended in three dimen
sions. Moreover, I had hoped to apply the analytical
a posteriori theory of necessity to geometry by starting
from the definition of a solid. I have also written
chapters, which are in print, on possible phenomena,
and on actual realities, in order to show at length that
scientific objects cannot be resolved into the former.
These were to be followed by a discussion of ideas,
including a criticism of Hegelianism, written but not
printed. I have in print chapters on touch and on
vision, directed against the doctrines of ' local signs,'
further developing the views in my criticism of Berke
ley's ' Theory of Vision,' and also based on the argu
ment that a sense of place is necessary to a sense of
motion. Finally, I meant to have revived the logic of
a method, which appears in Aristotle, but has fallen
out of logic. I mean the analytical deduction from
effect to cause, which appears to me to be the com
monest method of science, because man knows facts
before and better than causes. But seeing that all
these matters would have made at least another volume,
and fearful of becoming tedious, I also felt that I had
already claimed as much attention as could be hoped
bv an untried author.

CHAP. x. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 381

Thirdly, it will be said that I have exaggerated the
power of sense. This is a difficulty I do not feel or
allow. The emasculation of sense, which is the most
fundamental defect of modern philosophy, is a result of
a bygone anatomy. It was formerly thought that the
five senses were inlets, passages, or pores through which
sensible effects were received within us, according to
some, to the heart, according to others, to the head.
In these circumstances, it was excusable to suppose
that such poverty-stricken organs contributed nothing
but isolated data, which the soul worked up within into
all sorts of relations. But all that is changed now. It
has been discovered that the senses are highly compli
cated nervous structures ending in the brain, that the
brain is an integral organ of sense as well as of reason,
and that the whole nervous system has been for count
less generations hereditarily modified by its operations,
and, on the whole, better adapted to perform more and
more complex operations. Since these discoveries, I
submit that there is no bar to supposing that so wondrous
a sensitive structure, as a brain and a system of sensory
nerves has become, is an organ of simple and synthetic
sense of objects and relations, internal and physical,
as I have suggested. But I do not merely rely on
anatomy.

My main trust is in the philosophy of science.
Science proves the power of man to know nature. But
logic also proves the weakness of mere reason, which,
without adequate data of sense, is consistency, not
science. Eeason cannot logically infer insensible objects
and relations in external nature, unless there are sensible
objects and relations in our internal nature for sense
to perceive. Hence, to provide adequate data for the
parity of reasoning, I suppose a simple and synthetic

PSYCHOLOGICAL IDEALISM PART n.

sense of physical objects and relations within the ner
vous system. I hope, by this means, to have done
what I could to physic two diseases of modern idealism
—the separation of reason from nature, and the divorce
of reason from sense. The real problem of philosophy
is not how to form ideas, nor how to escape from them
to things ; it is not to start with sensations, and ask
how much, by association, we can conceive but not
know, nor how much, by a priori elements, we can
know, of mere phenomena. What are the adequate
data of sense, and what the logical processes of reason-
ing, which enable science to infer an insensible and im-
perceptible world. These are the questions for psycho
logy and Icgic to ask about sense and reason. ' Itaque,'
in the words of Bacon, ' ex harum facultatum, Experi-
mentalis scilicet et Eationalis, arctiore et sanctiore
foedere (quod adhuc factum non est) bene sperandum
est.' :

1 Nov. Org. i. 95.

APPENDIX.

UEBERWEG'S SUMMARY OF THE 'CRITIQUE:1

By the critique of the reason Kant understands the exami
nation of the origin, extent, and limits of human knowledge.
Pure reason is his name for reason independent of all experience.
The * Critique of the Pure Reason ' subjects the pure speculative
reason to a critical scrutiny. Kant holds that this scrutiny must
precede all other philosophical procedures. Kant terms every
philosophy, which transcends the . sphere of experience without
having previously justified this act by an examination of the faculty
of knowledge, a form of ' Dogmatism ' ; the philosophical limita
tion of knowledge to experience he calls ' Empiricism ' ; philoso
phical doubt as to all knowledge transcending experience, in so
far as this doubt is grounded on the insufficiency of all existing
attempts at demonstration, and not on an examination of the
human faculty of knowledge in general, is termed by him ' !
cism,' and his own philosophy, which makes all further philosophis
incr dependent on the result of such an examination, 'Criticism.
Criticism is < transcendental philosophy ' or 'transcendental idealism
in so far as it inquires into and then denies the possibility of i
transcendent knowledge, i.e. of knowledge respecting what lies
beyond the range of experience.

Kant sets out in his critique of the reason with a twof
division of judgments (in particular, of categorical judgments).
With reference to the relation of the predicate to the subject, he
divides them into analytical or elucidating judgments -where the
predicate can be found in the conception of the subject by simple
analysis of the latter or is identical with it (in which latter case t
analytical judgment is an identical one) -and synthcti

i Ucbcrwcg's Hist, of Phil. (English Trans.), vol. ii. pp. 154 58 « 122).

384 UEBERWEG'S SUMMARY OF THE 'CRITIQUE.'

cative judgments — where the predicate is not contained in the
concept of the subject, but is added to it. The principle of analy
tical judgments is the principle of identity and contradiction ; a
synthetic judgment, on the contrary, cannot be formed from the
conception of its subject on the basis of this principle alone. Kant
further discriminates, with reference to their origin as parts of
human knowledge, between judgments a priori and judgments a
posteriori ; by the latter he understands judgments of experience,
but by judgments a priori, in the absolute sense, those which
are completely independent of all experience, and in the relative sense,
those which are based indirectly on experience, or in which the concep
tions employed, though not derived immediately from experience, are
deduced from others that were so derived. As absolute judgments
a priori Kant regards all those which have the marks of necessity
and strict universality, assuming (what he does not prove, but
simply posits as self-evident, although his whole system depends
upon it) that necessity and strict universality are derivable from
no combination of experiences, but only independently of all ex
perience. All analytical judgments are judgments a priori ; for
although the subject-conception may have been obtained through
experience, yet to its analysis, from which the judgment results, no
further experience is necessary. Synthetic judgments, on the con
trary, fall into two classes. If the synthesis of the predicate with
the subject is effected by the aid of experience, the judgment is
synthetic a posteriori ; if it is effected apart from all experience, it
is synthetic a priori. Kant holds the existence of judgments of
the latter class to be undeniable ; for among the judgments whicli
are recognised as strictly universal and apodictical, and which are
consequently, according to Kant's assumption, judgments a priori,
he finds judgments which must at the same time be admitted to be
synthetic. Among these belong, first of all, most mathematical
judgments. Some of the fundamental judgments of arithmetic
(e.y. a=a) are, indeed, according to Kant, of an analytical nature ;
but the rest of them, together with all geometrical judgments, arc.
in his view, synthetic, and, since they have the marks of strict
universality and necessity, are synthetic judgments a priori. The
same character pertains, according to Kant, to the most general
propositions of physics, such as, for example, that in all the changes
of the material world the quantity of matter remains unchanged.
These propositions are known to be true apart from all experience,
since they are universal and apodictical judgments ; and yet they
are not obtained through a mere analysis of the conceptions of their
subjects, for the predicate adds something to those conceptions. In

APPENDIX ;;s-

liko manner, finally, are all metaphysical principles, at least in their
tendency, synthetic judgments a priori, e.g. the principle that
every event must have a cause. And if the principles of meta
physics are not altogether incontrovertible, yet those of mathematics
at least are established beyond all dispute. There exist, therefore
concludes Kant, synthetic judgments a priori or judgments of the
pure reason. The fundamental question of his Critique becomes,
then : How are synthetic judgments a priori possible ?

The answer given is : Synthetic judgments ^ jmm are possible,
because man brings to the material of knowledge, which he acquires
empirically in virtue of his receptivity, certain pure forms of know
ledge, which he himself creates in virtue of his spontaneity and
independently of all experience, and into which lie fits all given
material. These forms, which are the conditions of the possibility
of all experience, are at the same time the conditions of the possi
bility of the objects of experience, because whatever is to be an
object for me, must take 011 the forms through which the Ego, my
original consciousness, or the ' transcendental unity of apperception,'
shapes all that is presented to it ; they have, therefore, objective
validity in a synthetic judgment a priori. But the objects, with
reference to which they possess this validity, are not the things-in-
themselves or transcendental objects, i.e. objects as they are in
themselves, apart from our mode of conceiving them ; they are only
the empirical objects or the phenomena which exist in our conscious
ness in the form of mental representations. The things-in-them-
selves are unknowable for man. Only a creative, divine mind, that
gives them reality at the same time that it thinks them, can have
power truly to know them. Things-in-themselves do not conform
themselves to the forms of human knowledge, because the human
consciousness is not creative, because human perception is not free
from subjective elements, is not 'intellectual intuition.' Nor do
the forms of human knowledge conform themselves to things-in-them-
selves ; otherwise all our knowledge would be empirical and without
necessity and strict universality. But all empirical objects, since
they are only representations in our minds, do conform themselves
to the forms of human knowledge. Hence we can know empirical
objects or phenomena, but only these. All valid a priori knowledge
has respect only to phenomena, hence to objects of real or possible
experience.

The forms of knowledge are forms either of intuition or of
thought. The ' Transcendental ^Esthetic ' treats of the former, the
' Transcendental Logic ' of the latter.

The forms of intuition are space and time. Space is the form

C C

386 UEBERWEG'S SUMMARY OF THE 'CRITIQUE'

of external sensibility, time is the form of internal and indirectly
of external sensibility. On the a priori nature of space depends
the possibility of geometrical and on the a priori nature of time
depends the possibility of arithmetical judgments. Things-in-them-
selves or transcendental objects are related neither to space nor to
time ; all co-existence and succession are only in phenomenal
objects, and consequently only in the perceiving Subject.

The forms of thought are the twelve categories or original con
ceptions of the understanding, on which all the forms of our judg
ments are conditioned. They are : unity, plurality, totality, —
reality, negation, limitation, — substantiality, causality, reciprocal
action, — possibility, existence, necessity. On their a priori nature
depends the validity of the most general judgments, which lie at the
foundation of all empirical knowledge. The things-in-themselves or
transcendental objects have neither unity nor plurality ; they are not
substances, nor are they subject to the causal relation, or to any of
the categories ; the categories are applicable only to the phenomenal
objects which are in our consciousness.

The reason strives to rise above and beyond the sphere of the
understanding, which is confined to the finite and conditioned, to
the unconditioned. It forms the idea of the soul, as a substance
which ever endures ; of the world, as an unlimited causal series ;
and of God, as the absolute substance and union of all perfections,
or as the ' most perfect being.' Since these ideas relate to objects
which lie beyond the range of all possible experience, they have no
theoretic validity ; if the latter is claimed for them (in dogmatic
metaphysics) this is simply the result of a misleading logic founded
on appearances, or of dialectic. The psychological paralogism con
founds the unity of the I — which can never be conceived as a pre
dicate, but only and always as a subject — with the simplicity and
absolute permanence of a psychical substance. Cosmology leads to
antinomies, whose mutually contradictory members are each equally
susceptible of indirect demonstration, if the reality of space, time
and the categories be presupposed, but which with the refutation of
this supposition cease to exist. Rational theology, in seeking by the
ontological, cosmological, and physico- theological arguments to prove
the existence of God, becomes involved in a series of sophistications.
Still these ideas of the reason are in two respects of value : (1) theo
retically, when viewed not as constitutive principles through which
a real knowledge of things-in-themselves can be obtained, but as
regulative principles, which affirm that, however far empirical in
vestigation may at any time have advanced, the sphere of objects of
possible experience can never be regarded as fully exhausted, but that

APPENDIX 387

there will always be room for further investigation ; (2) practically,
in so far as they render conceivable suppositions, to which the
practical reason conducts with moral necessity.

In the ' Metaphysical Principles of Physics/ Kant seeks, by
reducing matter to forces, to justify a dynamical explanation of

o
nature.

PRINTED BY

srOTTISWOODE AND CO., NEW-STREET SQUARB
LONDON'

t • **

1NOING

JUK 1 5 1930

University of Toronto
Library

DO NOT

REMOVE

THE

CARD

FROM

THIS

POCKET