"INFERENCE BY SIMILARITY" FROM "REPRESENTATIONS".

The general conclusion to be drawn from the two preceding

sections (123, 124) is that the Aristotelian and scholastic distinc

tion between two conditions of sense qualities in the external

domain, namely, their actual condition when being actually per

ceived, and ti\Q\r potential condition apart from perception (122),

must not be understood as implying that one set of these

qualities, the common sensibles, are less relative to the perceiver

and externally more real apart from perception, than the other

set, the proper sensibles (123); and much less as implying the

theory of mediate perception, and this other assertion which is

intelligible only on this theory, namely, that the external prim

ary qualities are like, while the external secondary qualities are

unlike, our conscious representations of them.

If we look for the origin, in modern philosophy, of this dis

tinction between the primary or quantitative characteristics and

the secondary or qualitative characteristics of the domain of our

sense experience ; of the notion that externally the former are

like and the latter unlike our internal representations of them ;

and of the consequent tendency to regard the secondary or

qualitative characteristics as belonging exclusively to the

internal domain of the perceiver, we shall have to go back

to Descartes (1596-1650) and his contemporaries. 1 Without

dwelling on the devious and doubtful method whereby Descartes

attained to certitude about any external reality by invoking the

Divine veracity (100), it will suffice to say that because he had

"clear and distinct ideas" of extension and motion and their

modes, and " obscure and confused " ideas of colour, taste, sound,

smell, temperature, and tactile qualities, he held the former to

1 C/. jEANNlfeRE, Op. Ctt., pp. 436-7.

139

1 40 TI/EOK V OF A JVO W LEDGE

be extramentally real, 1 and the latter to be extratnentally some

thing or other, vaguely apprehended as the cause of these ob

scure conscious representations. Meanwhile, in England, Francis

Bacon (i 561-1626) had fixed attention on the inductive study

of the external universe, and Thomas Hobbes (1588-1679)

re-echoed the view of Descartes by teaching that " All the

qualities called sensible are, in the object which causeth them,

but so many motions of the matter, by which it presseth on

our organs diversely"." He even went farther than Descartes

by continuing : " Neither in us that are pressed are they any

thing else but divers motions ; for motion produceth nothing

but motion ". 3 Such assertions, leaving consciously appre

hended qualitative differences unexplained, would not have been

so confidently made if it had occurred to their author to ask

himself on what sort of assumptions regarding the scope of con

scious perception he knew " matter" and " motion " and " organs "

to be not mere conscious states, and yet to cause conscious states

so entirely unlike them. But Locke (1632-1704), like Descartes,

did concern himself with the problem of sense perception ; and

like the latter too, he assumed both that the immediate objects

of all our knowledge must be ideas, or psychic, conscious states

of the knowing subject, 4 and that externally the primary qualities

are like, and the secondary qualities unlike, our ideas of them.

Berkeley (1685-1752) pointed out that if we are immediately

aware only of our own ideas, or psychic or conscious states, and

that if the nature of what is extramental is known only by in

ference, through similarity, from these ideas, we should in con

sistency hold that both the primary and the secondary extramental

correlates are equally like our ideas of them/ Finally Hume

1 And abstract, three-dimensional extension to be the essence of matter : extern

ally these primary qualities were assigned to resemble the " clear and distinct ideas "

which they produced in us, and which were regarded as the immediate objects of

our awareness.

-Leviathan, I., c. I. "Ibid.

4 Professor CASK thus pithily outlines the persistent progress of this assumption

in modern philosophy: "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 con

sciousness. But the hypothesis has not changed its essence, though the idealists

have changed their terms, Verbum, non animuni, mutant. They at least agree that

all sensible data are psychical objects of some kind or other." Of>. cit., p. 15.

Tl Unfortunately, instead of repudiating the gratuitous postulate of Idealism, and

investigating the merits of the alternative assumption, that what we are immedi-

observed that if the so-called primary qualities are apprehended

only by the co-operation of two or more external senses with the

internal faculty of association, co-ordination, and unification (i 14,

123), while the so-called secondary qualities are apprehended

each as the proper datum of some one separate external sense,

there is certainly more of the subjective or self factor in the ela

boration of the former qualities (as present to consciousness)

than of the latter ; and that therefore the former ought to be

less like their supposed extramental correlates than the latter. 1

Thus, under the overshadowing influence of the fundamental

gratuitous assumption of Idealism, that the mind can become

directly aware only of its own states, we find the pendulum of

scepticism about the nature of the extramental oscillating from

the one-sided inference that the primary but not the secondary

qualities of extramental reality may be inferred to be similar to

our ideas of them, to the opposite and equally one-sided inference

that if we are to infer similarity of the extramental at all we

should infer it in regard to the secondary rather than the primary

qualities.

And so the reflection is once more (104) forced upon us : If we allow

that in sense perception we are never directly aware of the extramental, but

always only of an intramental or psychic object, can we know intellectually

that this object is an " appearance " or " representation " produced in the

mind by an extramental cause, and can we have reasoned intellectual certi

tude of the existence and nature of this extramental cause ? In common with

realist supporters of the theory of mediate sense perception we hold it for an

undoubted fact that the knowing subject can intellectually transcend self and

attain to a reasoned conviction of the reality of a non-self domain of being.

Our reason, however, for holding that fact to be undoubted is because we

also hold it an indubitable fact that in normal sense perception sense directly

reveals a real non-self to intellect, or in other words that such normal sense

evidence of what a directly perceived reality appears, is identically valid intel

lectual evidence of what this reality really is (105).

ately aware of is extramental material reality, he rejected the latter altogether, and

held that our ideas are not representations of realities beyond themselves at all, but

are themselves the only realities, placed in our minds by the Divine Spirit. Cf.

supra, 123.

1 JEANNIERE, outlining the progress of Idealism, has this significant observation :

" Principio semel introducto participationis subjecti cognoscentis in objecto confici-

endo, aperta est via omni licentiae vel potius audaciae ". Op. cit., p. 437. Cf.

MAHER, op. cit., p. 154. In Kant s philosophy also, as MAKER remarks (ibid.), " the

objective significance of the two groups is similarly reversed ". The primary

qualities, extension and motion, are products of the a priori forms of perception, space

and lime, while the secondary qualities are somehow in the material that is supposed

to be " given from without ". Cf. infra, 130.

142 THEORY OF KNOWLEDGE

If, on the other hand, all the immediately apprehended conscious data

from which we abstract our concept of cause were phases of the self-reality,

it is difficult to see how inference from effect to cause would attain to the

existence of non-self reality for us (IO4). 1 Furthermore, the validity of the

logical process whereby we infer from the nature of an effect similarity in

the nature of the cause, is not self-evident or universal. It is experience alone

that can tell us, experience of causes producing effects similar in nature to

themselves, how far we can safely use such inference, or what degree of

similarity, univocal or analogical, we may infer in a particular case. J For

we have also experience of causes producing effects dissimilar in nature to

themselves. All that is self-evident is that the adequate cause of any effect

must pre-contain equivalently or eminently the perfections of its effect.*

But this gives us no certain guidance as to the kind or degree of similarity

we are justified in attributing to the extramental factor in perception, if we

can know this latter only by inferring it as partial cause of the " conscious

appearance " or " psychic representation " which alone we are supposed to be

capable of apprehending directly and immediately.

1 Kant himself in his " Refutation of Idealism T (Critique, tr. MULLER, pp.

779-80), arguing that " consciousness of [one s] own existence is, at the same time,

consciousness of the existence of objects in space outside [one s self]," and remark

ing that in his proof " the trick played by idealism has been turned against it," has

this observation : " Idealism assumed that the only immediate experience is the in

ternal, and that from it we can no more than infer external things, though in an un

trustworthy manner, as always happens if from given effects we infer definite causes :

it being quite possible that the cause of the representations, which are ascribed by

us, it may be wrongly, to external things, may lie within ourselves" . The observa

tion is just ; but, unfortunately, Kant s own improvement on the idealism he sought

to refute is not appreciable, inasmuch as in his own theory " space " and " outside "

are mere domains of the mind. Cf. vol. i., 61, 97 ; supra, 100 ; infra, chap. xxii.

2 The aphorism Omne agens agit simile sibi, is a rough inductive generalization

from our observation of the propagation of species in the domain of living organisms.

Cf. Ontology, 98, p. 372 (/). Nevertheless it is the principle upon which re-

presentationists must rely in vindicating a knowledge of extramental reality.

Thus JEANNIERE, in answer to the difficulty, " How can we, from an internal re

presentative state, come to know an external reality, especially a heterogeneous

reality? " replies, " We can do so quite intelligibly by the principle of causality :

there exists a proportion between effect and cause, so that the effect bears some re

semblance to the cause : agcns agit sibi simile " . Op. cit., p. 446. In the same con

text he refers to the view of some authors, that " by the apprehension of this re

semblance the external thing itself is immediately apprehended " ; and he rightly

observes that this is only confusing the issue, inasmuch as the " resemblance " is not

really identical with the " thing " . But the view referred to cannot be perception-

ism, for the perceptionist repudiates the view that what \ve apprehend is the " re

semblance," " image," or " mental state " .

3 " Whatever be the nature of efficient causality, actio and passio, or of the de

pendence of the produced actuality upon the active power of its adequate efficient

cause, the reality of this dependence forbids us to think that in the natural order of

efficient causation a higher grade of reality can be actualized than the agent is

capable of actualizing, or that the agent can naturally actualize a higher or more

perfect grade of reality than is actually its own." Ibid., p. 371. Is the truth of this

latter statement self-evident ? It has. been questioned by a scholastic writer,

Professor Laminne of Louvain, in the Revue neo-Scolastiqtte, Feb. 1904 ibid., n.

SUBJECTIVE IDEALISM AND INFERENTIAL REALISM 143

Hence the "transfigured" or "symbolic" realism of Spencer, that our

conscious states are mere symbols of an extramental but unknowable reality :

a realism almost attenuated to idealism. And hence, also, the phenomenism

of Kant, that we can know only mental appearances, but are forced to postu

late the reality of an unknowable extramental or noumenal cause of them.

Those idealisms are rejected by realist advocates of the theory of mediate

sense perception : but since according to these latter the domain of being

which is directly presented in sense perception and represented in intellectual

conception is a "self" domain of mental states, their refutation of such

idealisms will appear unconvincing to many. For instance, the difference be

tween their own form of critical realism and the " symbolic " or " trans

figured " realism of Spencer is not great. Distinguishing between the being

(" esse intentionale ") which the sense object has in the perceiver, and the

being (" esse physicum ") which it has outside the perceiver, Jeanniere says : 1

" It must at least be held that the former resembles the latter, not indeed in

the manner in which e.g. the colour of one rose resembles the colour of an

other, but by some analogy of proportion 2 founded in the fact that the one is

cause of the other (113). Hence we have a proper concept of formal or sub

jective colour or sweetness, but not, if we speak strictly, of extra-subjective

colour or sweetness." Explaining this he continues : 3 " There is a certain

proportion or resemblance (analogy) between the non-Ego and the impres

sions produced in the Ego by the non-Ego. Therefore the knowledge I

have of the things that cause my impressions is not purely symbolic, such,

for instance, as the light of the semaphore signalling a ship. Things are in

themselves what they must be in order to produce in the human organs all

that we are conscious of experiencing." On this we may remark (i) that

Spencer would distinguish between artificial and natural symbolism, and

would hold the percept to be a natural symbol of the extramental ; (2) that

it is not the effect produced in the sense organs by the external reality,

but the effect produced in consciousness by the extramental (organic and

extra-organic) reality that representationists hold to be the immediate object

of awareness ; (3) that while the perceptionist regards the whole perception

process to be, in scholastic terms, an imago, similitude, reproductio intention-

alis, of the extramental reality, he repudiates altogether the view that the

esse intentionale or esse mentale (20, 102) which the latter (the esse reale or

physicum} thus obtains in the perceiver, whether as species impressa or

species expressa (112), is itself an object of direct awareness, a prius cog-

nitum, from which the knower would infer the reality of a similar extramental

correlate. 4 The scholastic aphorism," Cognitum est in cognoscente secundum

aliquam sui similitudinem," 5 simply means that the mind, by virtue of the

whole cognitive process, becomes a similitudo of the known reality, or is

" assimilated or conformed " to the latter ; it does not mean that this mental

state, this " similitudo intentionalis," is first known, and its extramental cause

inferred from it ( 1 1 2). " Of, what, then [the author continues], it will be asked,

have we proper concepts ? (a) We have proper concepts of the facts of con

sciousness ; of toothache, etc. ; of green, red, sweet, bitter, resisting, etc. (b)

1 Op. cit., p. 425. 2 On the subject of analogy, cf. Ontology, 2, pp. 36-42.

3 Ibid,, n. 2, 4 Cf. infra, 129. JEANNIERE, op. cit., ibid.

1 44 THEOR Y OF KNO W LEDGE

Of all else we have only analogical concepts : of the soul, God, substance,

cause, etc. (<) At the same time things in themselves are said to be known

by proper concepts when they are known by their natural (per se) effects on

our organs, effects of which we have proper concepts. For example, I have

a proper notion of the cherry because of the proper notion I have of its sensible

effect. On the other hand when the thing in itself is not known to me by its

proper or natural sensible effects I have only analogical notions of it." The

perceptionist holds that we have proper concepts of all extramental (organic

and external) material qualities, causes, and natures or substances ; because

he holds that these are all directly given either to sense (scnsibilia per sc) or

to intellect with the data of sense (scnsibilia per accidens). For the re-

presentationist, however, the concepts under (c) cannot really be " proper," but

only " analogical " .

The progress of Idealism also forces upon us the reflection that if we once

admit the secondary qualities to be mere mental or psychic states or " sensa

tions," produced in us by external reality, we shall find it difficult to main

tain, as against the Kantian form of Idealism, that the primary qualities,

extensional or spatial determinations, are externally real. Of this we have

an interesting illustration in Prichard s otherwise very excellent criticism of

Kant s speculative philosophy.

Examining Kant s view that we are aware only of " appearances " pro

duced in us by things, he says, " To speak of appearances produced by things

is to imply that the object of perception is merely something mental, vis. an

appearance. Consequently access to a non-mental reality is excluded ; for a

perception of which the object is something belonging to the mind s own being

cannot justify an inference to something beyond the mind, and the result is

inevitably solipsism." : The principle here is that because an appearance is

"something mental," "belonging to the mind s own being," we are not

justified in inferring to the extramental. But he goes on immediately to allow

that the secondary qualities of bodies are not in the bodies, are not extra-

mental, but are " sensations " " produced in us by the extramental. Being

therefore only "sensations," or " something mental," neither can they "justify

an inference to something extramental " or give us any information as to the

real nature of the latter. What, therefore, can we know about the real nature

of the extramental, material universe ? Well, we have the primary qualities,

or " spatial relations " of this universe to fall back upon. These, Prichard

contends (against Kant), we can and do know to be extramentally real. So

the material universe is, then, (really and extramentally) merely a system of

homogeneous, space-filling, three-dimensional, spatially moving and interact

ing realities ? And how can we know even such a system of space-filling and

moving realities to be extramental ? Because the primary qualities which re

veal it are not sensations or sense-percepts, dependent on a perceiver, but are

conceived or intellectually apprehended thought-objects ; and being intellectu-

1 Op. cit., chap, iv., p. 76 (italics ours). 2 Ibid., p. 86.

3 Or, from the heterogeneous "sensations" or "mental realities" which the

secondary qualities are held to be, may we infer heterogeneity (and, if so, what sort

of heterogeneity ?) in the extramental causes of them ? Prichard holds, of course,

that the " sensations" are produced in us by extramental realities. Knowledge of

the latter he does not, however, seem to regard as inferred or infenible ftom the

former, but as attained independently of these.

SUBJECTIVE IDEALISM AND INFERENTIAL REALISM 145

ally judged to be extramentally real they must therefore be extramentally real,

inasmuch as " it is a presupposition of thinking that things are in them

selves what we think them to be V

It will be instructive to examine somewhat more in detail the line of

thought by which the author reaches these not very satisfactory conclusions.

He says we must allow the secondary qualities, even colour, which pre

sents special difficulties, to be merely facts of the mental or psychic order,

" sensations," 2 produced in us by external reality. They are not even ap

pearances of external bodies: "when once the issue is raised it is difficult

and, in the end, impossible to use the word appear in connexion with these

qualities. Thus it is difficult and, in the end, impossible to say that a bell

appears noisy, or that sugar appears sweet. We say, rather, that the bell and

the sugar produce certain sensations (not appearances ) in us." 3 The case

of colour he then proceeds to examine, and concludes that it too is a mental

state or sensation, that in respect of colour " things look what they never are " ; 4

but that the fact of their " looking " or " appearing " coloured implies that

they are (i) real, and (2) extended or spatial. 5 He then faces the Kantian

difficulty " that just as things may only look coloured, so things may only

look spatial ". {i This he meets by the contention that as a matter of fact what

things look or appear to be spatially they never are spatially ; that their real

(i.e. three-dimensional} spatial determinations are always other than, and

are in no limiting cases coincident with, what they appear spatially to be ;

that what they appear in point of spatial extension always implies " correlation

to a percipient," whereas what they really are in point of spatial extension

always is, nay means, what they are " independently of a percipient " ; 8 that,

therefore, " it is so far from being true that we only know what things look

and not what they are, that in the case of spatial relations we actually know

what things are, even though they never look what they are ". 9 But if we

have to admit " that we perceive things as they look, and not as they are," 1U

or that we perceive only what things look spatially, how, he asks (i) can we

have ever come to believe that things are really spatial, and (2) how can we

know that this belief is not illusory ? His reply is that this belief is implied

in our knowledge of what things look spatially, and that we know this belief

not to be illusory because in regard to spatial relations there is no transition

in principle, but only in respect of details, in passing " from knowledge of

what things look to knowledge of what things are " : n in other words, since it

is undeniable " that we can and do state what things appear " 12 in respect of

spatial relations, and since the possibility of knowing what things look or

appear spatially implies throughout the " consciousness " 13 or " belief" 14 that

things really are spatial, it must follow that things really are spatial.

I Op. cit., p. 100. *Ibid., pp. 85 sqq.

3 Ibid., p. 86. Cf., however, infra, 128. We agree with the author s refusal

to regard such produced mental states (if we admitted the secondary qualities to be

such) as " appearances ".

4 P. 87. 8 P. 88.

8 P. 89. All the primary qualities, being spatial determinations, are involved

in this charge.

7 P. 91. 8 Ibid. 8 P. 91. 10 Ibid.

II P. 92. l2 P. 93. i 3 P. 92. 14 P. 91.

VOL. II. 10

1 46 THEOR Y OF KNO W LEDGE

We must confess we are not convinced by this line of defending the ex

ternal reality of spatial determinations against Kantism.

(i) In the first place the author urges, against Kant s position, that "an

appearance, being necessarily something mental, cannot possibly be said to

be extended" ; that we cannot predicate of an "appearance" spatial deter

minations such as " convergent," 2 or divergent, etc.; that "an appearance

cannot be spatial"/ 1 Very well; but "sensations" are also "something

mental," and yet, on the author s view (that the secondary qualities are sen

sations), we can predicate of these "mental" facts or states that they are

" red " or " hot " or " bitter " : but if so, then, after all, why not predicate of

" appearances " that they are spatial ? It may, however, be urged that there

is, after all, a difference ; but then what about this other fact: Is not "ex-

tensity " or " voluminousness " a spatial " sensuous element " 5 or datum of the

tactual and organic senses, whose data, being secondary qualities, are pre

sumably regarded by the author as " sensations " ? When arguing that the

primary qualities, i.e. spatial determinations, are separable from colour, he says,

by way of confirmation/ 5 " moreover, if the possibility of the separation of the

primary qualities from colour is still doubted, it is only necessary to appeal to

the blind man s ability to apprehend the primary qualities though he may not

even know what the word colour means ". But can the blind man become

conscious of "the primary qualities," or "spatial relations," apart from his

tactual, organic, and motor perceptions ? And the data of these latter are

presumably secondary qualities, and therefore "sensations". " Of course,"

the author continues, " it must be admitted that some sensuous elements [i.e.

sensations, mental facts] are involved in the apprehension of the primary

qualities," but the case of the blind man shows that these may relate to sight

instead of to touch ". Yes, it shows that some of them may ; but are there

no " sensuous elements " in the blind man s tactual data ? Surely there are ;

and the point is, can he apprehend "spatial relations" apart from these?

Nay, are not the vaguely felt "extensity" and "voluminousness" of our

visual, tactual, organic, gustatory (and possibly auditory and olfactory) sense

data, themselves "sensuous elements" or at least inseparable from the

"sensuous elements" in these perceptions, if by "sensuous elements" the

author means the mental facts or data which he takes the secondary qualities

to be ? So far as introspection can discover, the " spatial determinations "

which we call primary qualities the felt "externality," "extensity," "volum

inousness," are inseparable from the secondary qualities which the author

calls " sensuous elements," in our conscious sense data. The author seems

to realize this, for he adds : " Moreover, it, of course, does not follow from

the fact that sensuous elements are inseparable from our perception of bodies

that they belong to, and are therefore inseparable from, the bodies perceived ."

Nevertheless, if we separate them we must show cause for separating them.

1 Op. dt., p. 76. 2 P. 81. 3 P. 93.

4 For these must be real predicates of something : if not of extramental realities,

then of mental states.

* Cf. p. 91 n. 6 P. 91 n.

7 So "primary qualities," or "spatial determinations" are "apprehended".

By sense or by intellect ? If by sense, then they " appear " to, and are " perceived

by " sense. If by intellect, cf. infra, p. 147.

8 Ibid.

SUBJECTIVE IDEALISM AND INFERENTIAL REALISM 147

Now, if they are separate or separable, and if we ask how do we appre

hend or become conscious of these spatial determinations at all, there seem

to be only two alternative answers possible. Either they too are sense data

and therefore :l relative to perception," J and dependent on the percipient

just in the same way as the secondary qualities or " sensuous elements " are.

This we believe to be the case. But if, this being the case, the secondary

qualities were also held to be mere sensations or mental facts it could no

longer be consistently argued against Kant s view that " an appearance can

not be spatial ". Hence the author, holding that the secondary qualities are

only mental facts or sensations, adopts the second possible alternative answer

to the question suggested, viz. that the primary qualities, or " spatial deter

minations," are not sense data at all, that they " cannot of their very nature

be relative to perception " ; 2 that they are exclusively concepts, objects of

thought, that they belong to what things are really, i.e. to what things are

for thought, for conception and judgment, and not to what things appear to

perception. 3

(2) Now this view is not supported by introspection. 4 It is, we believe,

an illustration of the dangerous tendency to which we have already called

attention (114), the tendency to lose sight of the primary qualities as concrete,

directly presented and felt sense percepts, and so to regard them exclusively

as abstract, intellectually represented concepts or thought-objects, a danger

increased by the fact that we can think and reason about sense data only

through concepts, i.e. by intellectually apprehending " what they are," so that

it requires a special effort of abstraction on our part to avoid reading into

sense data as such what belongs to them only as conceived (114). Moreover,

the view that real (three-dimensional) spatial characteristics must belong only

to what things are, and not at all to what they appear, (a) is based on the

mistaken assumption that what things really are must be wholly independent

of what they appear ; and (b] by leaving unexplained how we come to know

intellectually what things really are, it can yield no convincing refutation of

Kantism.

(a) If what things appear to sense is relative to the perceiver, neverthe

less, if at the same time we can by reflection discover the way in which they

are relative, vis. to the perceiver as organic, and not as mental and con

sciously perceptive (126), and if we can allow for such relativity in judging

what the perceived things are really and externally, then we can know what

they are really and externally even though we know that what they appear

to sense is partly relative to, and dependent on, the organic constitution of

1 Op. cit. 2 P. 91 n. 3 Ibid., pp. 99, 100.

4 Reflection on the data of sense consciousness fails to discover any difference

as to mode of presence or presentation, between directly apprehended " extensity,"

" externality," " solidity," " voluminousness," i.e. spatial characteristics, and

colour, sound, temperature, etc. That we become aware of the former with the

latter is undeniable. For consciousness they are inseparable. If it be contended

that, notwithstanding this apparent inseparability, the former are not sensuously ap

prehended at all, that they are known by intellect but do not appear to sense, the

contention can be supported only by showing that real spatial characteristics can be

proved to be of their very nature unperceivable. This the author tries to do, but,

as we think, unsuccessfully.

1 48 7HEOR V OF KNO WLEDGE

the perceiver, Of course, "our apprehension of what things 1 are is essenti

ally a matter of thought or judgment " ; - but we are by no means at liberty

to add, "and not of perception ". a For if our thought or judgment is not

an interpretation of what things appear to sense, or is not at least based upon

and motived by, what they appear to sense, what is it an interpretation of?

Moreover, it is implied, and rightly, that what things really are is what they

are for thought or judgment (i.e. for true judgment). But things cannot be

thought, conceived, judged, except in so far as they are " related " or " re

lative " to intellect, or in other words in so far as they " appear " to intellect.

What things really are (i.e. something at least of what they really are) is what

they are conceived, interpreted, represented, to be, by intellect (judging truly).

When, therefore, we contend that intellect can know things as they really are,

absolutely, this cannot mean that what they are when they are known, or

what they are known to be, is independent of their " appearing " to intellect,

but only that what they really are is not altered by this " appearing," and

that this " appearing" itself is a function of their reality. 3 This is what must

be established as against Kant. It is not established by asserting (what is

quite true) that " it is a presupposition of thinking that things are in them

selves what we think them to be " ; B and (what is true only in a certain sense)

that " from the nature of the case a presupposition of thinking not only cannot

be rightly questioned, but cannot be questioned at all ". It cannot, of course,

be really doubted, but it can be provisionally questioned and explored ; and

since, among others, Kant has explored it, he must be met by showing 8 that

the relation of reality to intellect interpreting it does not shut reality off from

all possibility of its being known as it is. But this cannot be effectually

shown, in regard to the spatial characteristics of material reality, by declaring

that these must really be as intellect conceives or apprehends them, so long

as no account is given of the way in which such characteristics become

cognitively related to, and are apprehended by, intellect.

(/ ) Prichard says " it is the view that what a thing really is it is, inde

pendently of a percipient, that forms the real starting-point of Kant s thought ". 9

But it is also an essential part of Kant s thought that what a thing really is it

1 This is true of all things, but it is only " material " or " spatial " things that

are under discussion here.

2 Op. cit., p. 99. *Ibid. 4 Cf. infra, 127, 128.

6 Prichard says " an appearance, as being ex hypotftesi and appearance to some

one, i.e. to a percipient, must be relative to perception " (p. 93). But if perception

implies an appearance of something to a perceiver does not thought imply an ap

pearance of something to a thinker ? Knowledge is an interpretation or representa

tion of something. But the " something " must appear to the knower : he must be

aware of it. Is not therefore this "appearance," or " something appearing " like

wise relative to the knower ? Cf. infra, 127, 128.

(i P. 100. " Ibid.

8 The fact that in this process of testing the capacity of intellect to know reality

as it is, we are supposing its capacity to reach true conclusions in the testing process

itself (i.e. its capacity to apprehend reality at least thus far) militates just as much

or as little, neither more nor less, against Kant s critics than against himself. The

fact, inevitable as it is, has not deterred men from undertaking the delicate and dif

ficult task of using thought to explore the springs of thought.

"P. 91.

149

is, independently of a conceiver, independently of thought, judgment, know

ledge. Moreover, both statements, as to what a thing really is, are tfue in

the sense that " what a thing really is " cannot be influenced or altered either

by perception or by thought ; but Kant defended the statements interpreted

in the sense that "what a thing really is" it can never "appear" either to

sense or to intellect. When, therefore, Prichard admits this in regard to

sense, and then tries to show, against Kant, that spatial relations really are

as we think them, or as they appear to thought, every step in his argument

can be countered by the contention that spatial relations are apprehended

(whether by sense or by intellect) in the same way as the secondary qualities

are apprehended (i.e. perceived and conceived), and that since he admits the

secondary qualities, whether in the concrete or in the abstract, whether per

ceived or conceived, to be only mental facts, so must spatial relations be only

mental facts.

For instance, answering the objection that our belief in the reality of space

(however it has arisen) may be after all an illusion, Prichard writes : " If

assertions concerning the apparent shape, etc., of things presuppose the con

sciousness that things are spatial, to say that this consciousness is illusory is

to say that all statements concerning what things appear, in respect of spatial

relations, are equally illusory. But since it is wholly impossible to deny that

we can and do state what things appear in this respect, the difficulty must fall

to the ground." 1 Admitting that " we can and do state what things appear "

in respect of spatial relations, Kant would simply reply that such statements

are illusory, and the consciousness implied by them is illusory, only if the

" things " to which they are understood to apply be regarded as extramental

things, not if the " things " be regarded as " mental facts ". 2 This implies,

of course, that it is possible to distinguish between the true or real 3 and the

false or apparent 4 within the domain of mental facts or " phenomena," 5 a

possibility which Kant proceeds to defend.

Again, in attacking Kant s defence Prichard writes : " We presuppose that

that quality is really, and not only apparently, a quality of a body, which we

and every one, judging from what it looks under various conditions (i.e. in

universal experience ), 6 must believe it to possess in itself and independently

of all perception ". 7 But how can a judgment formed " from what [the body]

looks under various conditions " tell us what or how the body really is, if the

body never under any conditions looks what it really is ? What right have

we to form such a judgment, or to entertain such a belief, if all the body s

appearing or perceptible qualities (" what it looks under various conditions ")

are, by being relative to a percipient, not real (because not " independent of

1 Op. cit., pp. 92-3.

2 We might also argue that the statement that a thing appears red implies the

consciousness that it is coloured. Yet Prichard holds this consciousness to be il

lusory unless " thing " be understood to mean a " sensation " or " mental fact ".

Hence he must either hold that all statements based on it are illusory, or else recog

nize the possibility of distinguishing between "real " or " true " appearances and

" deceptive " or " false" appearances within the domain of mental facts or sensa

tions, precisely what Kant does regarding spatial relations.

3 Erscheinung. 4 Schein. * Cf. ibid., p. 79.

6 Critique (MtJLLER), p, 37. 7 P. 99.

i 50 THEOR Y OF KNO W LEDGE

perception "), and if its real qualities (i.e. real three-dimensional extension,

real shape, real position, real motion, etc.) never appear, or are never per

ceived ? How does intellect apprehend these latter, if not in the data of

sense? "We do not perceive but thinks thing as it is." 1 But how or

whence does intellect apprehend the predicates whereby it thinks the thing, if

not in the data of sense ? Prichard, however, has admitted that the data of

sense, or " what things appear," are relative to the perceiver in the sense of

being themselves mental facts ; 2 and he sees that predicates or f/iought-ob-

jects, derived from data which are relative in that way could not yield know

ledge of what things other than mental facts can really be. The only alterna

tive is that intellect can apprehend what things really are, directly and of itself,

and apparently without any aid from sense. Of course what a thing really is

in respect of spatial relations is directly " correlated with thought," :i because

it is thought, not sense, that interprets what anything really is ; but even if

its correlation with thought did not involve its correlation with sense also, we

should have not merely to assert, but to show, as against Kant, that the former

correlation did not alter or transform the reality : and we should therefore

have to show how thought apprehends the extramentally real, a point to

which Prichard does not appear to have directed his attention. Correlation

of reality with thought docs, however, involve correlation of reality with sense.

Nor does this necessarily involve the confounding of thought with sense,

however Kant may stand in relation to this charge. 4 For our intellectual

knowledge of what material, spatial things are really and extramentally, is cer

tainly derived from what they appear to sense. And the only proper way of

showing, as against Kant, that nevertheless we can know intellectually what

things are really and extramentally, is by exploring the nature of the relation

in both cases, and showing that, when rightly understood, it does not screen

off the extramental reality either from intellect or from sense.

We shall have more to say later in criticism of the general position that

what we become directly aware of, whether in sense cognition or in intellectual

cognition, is and must be only a " representation " or " appearance," " pro

duced " in us by a something which this mental product is supposed to repre

sent. 5 But before doing so we have next to examine the difficulties which may

be urged against the form of Intuitive or Perceptive Realism so far outlined

in our inquiry.

Before passing from Idealism we may here note an objection

which is sometimes unthinkingly urged by idealists against real

ism in general. Vibrations of the air or the rcther, they argue,

are quite unlike our sensations of sound and colour : therefore

1 Op. cit., p. 99.

2 He admits that spatial characteristics which are de facto real and external ap

pear to sight and to touch (cf. p. 91 n.) ; but holds that they appear otherwise than

they really are : " in the case of spatial relations . . . things . . . never look what

they are". Hence predicates derived from " what they look" could never inform

us as to what they really are, any more than if " what they look " as regards spatial

relations were merely mental facts or " sensations," which he considers the second

ary qualities to be.

3 P. 100. 4 P. 99. " Cf. also vol. i., 92, ii. ; 93.

INTUITIVE REALISM 151

our sensations in general cannot give us any information about

the nature of external reality. Such an objection, coming from

an idealist, is suicidal, for, as Rickaby points out, 1 "he forgets

that it has been by the senses that the vibrations have been dis

covered, and that if the scientific result is worth anything, it

proves the ability of the senses to give us information about the

facts as they are in external nature ". In other words the objec

tion is based on an assumption which is the very negation of

idealism, viz. the assumption, common to physical science, that

at least motion and extension are really and externally identical

with, or similar or analogous to, that which we internally perceive

them to be.

The general conclusion to be drawn from the two preceding

sections (123, 124) is that the Aristotelian and scholastic distinc

tion between two conditions of sense qualities in the external

domain, namely, their actual condition when being actually per

ceived, and ti\Q\r potential condition apart from perception (122),

must not be understood as implying that one set of these

qualities, the common sensibles, are less relative to the perceiver

and externally more real apart from perception, than the other

set, the proper sensibles (123); and much less as implying the

theory of mediate perception, and this other assertion which is

intelligible only on this theory, namely, that the external prim

ary qualities are like, while the external secondary qualities are

unlike, our conscious representations of them.

If we look for the origin, in modern philosophy, of this dis

tinction between the primary or quantitative characteristics and

the secondary or qualitative characteristics of the domain of our

sense experience ; of the notion that externally the former are

like and the latter unlike our internal representations of them ;

and of the consequent tendency to regard the secondary or

qualitative characteristics as belonging exclusively to the

internal domain of the perceiver, we shall have to go back

to Descartes (1596-1650) and his contemporaries. 1 Without

dwelling on the devious and doubtful method whereby Descartes

attained to certitude about any external reality by invoking the

Divine veracity (100), it will suffice to say that because he had

"clear and distinct ideas" of extension and motion and their

modes, and " obscure and confused " ideas of colour, taste, sound,

smell, temperature, and tactile qualities, he held the former to

1 C/. jEANNlfeRE, Op. Ctt., pp. 436-7.

139

1 40 TI/EOK V OF A JVO W LEDGE

be extramentally real, 1 and the latter to be extratnentally some

thing or other, vaguely apprehended as the cause of these ob

scure conscious representations. Meanwhile, in England, Francis

Bacon (i 561-1626) had fixed attention on the inductive study

of the external universe, and Thomas Hobbes (1588-1679)

re-echoed the view of Descartes by teaching that " All the

qualities called sensible are, in the object which causeth them,

but so many motions of the matter, by which it presseth on

our organs diversely"." He even went farther than Descartes

by continuing : " Neither in us that are pressed are they any

thing else but divers motions ; for motion produceth nothing

but motion ". 3 Such assertions, leaving consciously appre

hended qualitative differences unexplained, would not have been

so confidently made if it had occurred to their author to ask

himself on what sort of assumptions regarding the scope of con

scious perception he knew " matter" and " motion " and " organs "

to be not mere conscious states, and yet to cause conscious states

so entirely unlike them. But Locke (1632-1704), like Descartes,

did concern himself with the problem of sense perception ; and

like the latter too, he assumed both that the immediate objects

of all our knowledge must be ideas, or psychic, conscious states

of the knowing subject, 4 and that externally the primary qualities

are like, and the secondary qualities unlike, our ideas of them.

Berkeley (1685-1752) pointed out that if we are immediately

aware only of our own ideas, or psychic or conscious states, and

that if the nature of what is extramental is known only by in

ference, through similarity, from these ideas, we should in con

sistency hold that both the primary and the secondary extramental

correlates are equally like our ideas of them/ Finally Hume

1 And abstract, three-dimensional extension to be the essence of matter : extern

ally these primary qualities were assigned to resemble the " clear and distinct ideas "

which they produced in us, and which were regarded as the immediate objects of

our awareness.

-Leviathan, I., c. I. "Ibid.

4 Professor CASK thus pithily outlines the persistent progress of this assumption

in modern philosophy: "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 con

sciousness. But the hypothesis has not changed its essence, though the idealists

have changed their terms, Verbum, non animuni, mutant. They at least agree that

all sensible data are psychical objects of some kind or other." Of>. cit., p. 15.

Tl Unfortunately, instead of repudiating the gratuitous postulate of Idealism, and

investigating the merits of the alternative assumption, that what we are immedi-

observed that if the so-called primary qualities are apprehended

only by the co-operation of two or more external senses with the

internal faculty of association, co-ordination, and unification (i 14,

123), while the so-called secondary qualities are apprehended

each as the proper datum of some one separate external sense,

there is certainly more of the subjective or self factor in the ela

boration of the former qualities (as present to consciousness)

than of the latter ; and that therefore the former ought to be

less like their supposed extramental correlates than the latter. 1

Thus, under the overshadowing influence of the fundamental

gratuitous assumption of Idealism, that the mind can become

directly aware only of its own states, we find the pendulum of

scepticism about the nature of the extramental oscillating from

the one-sided inference that the primary but not the secondary

qualities of extramental reality may be inferred to be similar to

our ideas of them, to the opposite and equally one-sided inference

that if we are to infer similarity of the extramental at all we

should infer it in regard to the secondary rather than the primary

qualities.

And so the reflection is once more (104) forced upon us : If we allow

that in sense perception we are never directly aware of the extramental, but

always only of an intramental or psychic object, can we know intellectually

that this object is an " appearance " or " representation " produced in the

mind by an extramental cause, and can we have reasoned intellectual certi

tude of the existence and nature of this extramental cause ? In common with

realist supporters of the theory of mediate sense perception we hold it for an

undoubted fact that the knowing subject can intellectually transcend self and

attain to a reasoned conviction of the reality of a non-self domain of being.

Our reason, however, for holding that fact to be undoubted is because we

also hold it an indubitable fact that in normal sense perception sense directly

reveals a real non-self to intellect, or in other words that such normal sense

evidence of what a directly perceived reality appears, is identically valid intel

lectual evidence of what this reality really is (105).

ately aware of is extramental material reality, he rejected the latter altogether, and

held that our ideas are not representations of realities beyond themselves at all, but

are themselves the only realities, placed in our minds by the Divine Spirit. Cf.

supra, 123.

1 JEANNIERE, outlining the progress of Idealism, has this significant observation :

" Principio semel introducto participationis subjecti cognoscentis in objecto confici-

endo, aperta est via omni licentiae vel potius audaciae ". Op. cit., p. 437. Cf.

MAHER, op. cit., p. 154. In Kant s philosophy also, as MAKER remarks (ibid.), " the

objective significance of the two groups is similarly reversed ". The primary

qualities, extension and motion, are products of the a priori forms of perception, space

and lime, while the secondary qualities are somehow in the material that is supposed

to be " given from without ". Cf. infra, 130.

142 THEORY OF KNOWLEDGE

If, on the other hand, all the immediately apprehended conscious data

from which we abstract our concept of cause were phases of the self-reality,

it is difficult to see how inference from effect to cause would attain to the

existence of non-self reality for us (IO4). 1 Furthermore, the validity of the

logical process whereby we infer from the nature of an effect similarity in

the nature of the cause, is not self-evident or universal. It is experience alone

that can tell us, experience of causes producing effects similar in nature to

themselves, how far we can safely use such inference, or what degree of

similarity, univocal or analogical, we may infer in a particular case. J For

we have also experience of causes producing effects dissimilar in nature to

themselves. All that is self-evident is that the adequate cause of any effect

must pre-contain equivalently or eminently the perfections of its effect.*

But this gives us no certain guidance as to the kind or degree of similarity

we are justified in attributing to the extramental factor in perception, if we

can know this latter only by inferring it as partial cause of the " conscious

appearance " or " psychic representation " which alone we are supposed to be

capable of apprehending directly and immediately.

1 Kant himself in his " Refutation of Idealism T (Critique, tr. MULLER, pp.

779-80), arguing that " consciousness of [one s] own existence is, at the same time,

consciousness of the existence of objects in space outside [one s self]," and remark

ing that in his proof " the trick played by idealism has been turned against it," has

this observation : " Idealism assumed that the only immediate experience is the in

ternal, and that from it we can no more than infer external things, though in an un

trustworthy manner, as always happens if from given effects we infer definite causes :

it being quite possible that the cause of the representations, which are ascribed by

us, it may be wrongly, to external things, may lie within ourselves" . The observa

tion is just ; but, unfortunately, Kant s own improvement on the idealism he sought

to refute is not appreciable, inasmuch as in his own theory " space " and " outside "

are mere domains of the mind. Cf. vol. i., 61, 97 ; supra, 100 ; infra, chap. xxii.

2 The aphorism Omne agens agit simile sibi, is a rough inductive generalization

from our observation of the propagation of species in the domain of living organisms.

Cf. Ontology, 98, p. 372 (/). Nevertheless it is the principle upon which re-

presentationists must rely in vindicating a knowledge of extramental reality.

Thus JEANNIERE, in answer to the difficulty, " How can we, from an internal re

presentative state, come to know an external reality, especially a heterogeneous

reality? " replies, " We can do so quite intelligibly by the principle of causality :

there exists a proportion between effect and cause, so that the effect bears some re

semblance to the cause : agcns agit sibi simile " . Op. cit., p. 446. In the same con

text he refers to the view of some authors, that " by the apprehension of this re

semblance the external thing itself is immediately apprehended " ; and he rightly

observes that this is only confusing the issue, inasmuch as the " resemblance " is not

really identical with the " thing " . But the view referred to cannot be perception-

ism, for the perceptionist repudiates the view that what \ve apprehend is the " re

semblance," " image," or " mental state " .

3 " Whatever be the nature of efficient causality, actio and passio, or of the de

pendence of the produced actuality upon the active power of its adequate efficient

cause, the reality of this dependence forbids us to think that in the natural order of

efficient causation a higher grade of reality can be actualized than the agent is

capable of actualizing, or that the agent can naturally actualize a higher or more

perfect grade of reality than is actually its own." Ibid., p. 371. Is the truth of this

latter statement self-evident ? It has. been questioned by a scholastic writer,

Professor Laminne of Louvain, in the Revue neo-Scolastiqtte, Feb. 1904 ibid., n.

SUBJECTIVE IDEALISM AND INFERENTIAL REALISM 143

Hence the "transfigured" or "symbolic" realism of Spencer, that our

conscious states are mere symbols of an extramental but unknowable reality :

a realism almost attenuated to idealism. And hence, also, the phenomenism

of Kant, that we can know only mental appearances, but are forced to postu

late the reality of an unknowable extramental or noumenal cause of them.

Those idealisms are rejected by realist advocates of the theory of mediate

sense perception : but since according to these latter the domain of being

which is directly presented in sense perception and represented in intellectual

conception is a "self" domain of mental states, their refutation of such

idealisms will appear unconvincing to many. For instance, the difference be

tween their own form of critical realism and the " symbolic " or " trans

figured " realism of Spencer is not great. Distinguishing between the being

(" esse intentionale ") which the sense object has in the perceiver, and the

being (" esse physicum ") which it has outside the perceiver, Jeanniere says : 1

" It must at least be held that the former resembles the latter, not indeed in

the manner in which e.g. the colour of one rose resembles the colour of an

other, but by some analogy of proportion 2 founded in the fact that the one is

cause of the other (113). Hence we have a proper concept of formal or sub

jective colour or sweetness, but not, if we speak strictly, of extra-subjective

colour or sweetness." Explaining this he continues : 3 " There is a certain

proportion or resemblance (analogy) between the non-Ego and the impres

sions produced in the Ego by the non-Ego. Therefore the knowledge I

have of the things that cause my impressions is not purely symbolic, such,

for instance, as the light of the semaphore signalling a ship. Things are in

themselves what they must be in order to produce in the human organs all

that we are conscious of experiencing." On this we may remark (i) that

Spencer would distinguish between artificial and natural symbolism, and

would hold the percept to be a natural symbol of the extramental ; (2) that

it is not the effect produced in the sense organs by the external reality,

but the effect produced in consciousness by the extramental (organic and

extra-organic) reality that representationists hold to be the immediate object

of awareness ; (3) that while the perceptionist regards the whole perception

process to be, in scholastic terms, an imago, similitude, reproductio intention-

alis, of the extramental reality, he repudiates altogether the view that the

esse intentionale or esse mentale (20, 102) which the latter (the esse reale or

physicum} thus obtains in the perceiver, whether as species impressa or

species expressa (112), is itself an object of direct awareness, a prius cog-

nitum, from which the knower would infer the reality of a similar extramental

correlate. 4 The scholastic aphorism," Cognitum est in cognoscente secundum

aliquam sui similitudinem," 5 simply means that the mind, by virtue of the

whole cognitive process, becomes a similitudo of the known reality, or is

" assimilated or conformed " to the latter ; it does not mean that this mental

state, this " similitudo intentionalis," is first known, and its extramental cause

inferred from it ( 1 1 2). " Of, what, then [the author continues], it will be asked,

have we proper concepts ? (a) We have proper concepts of the facts of con

sciousness ; of toothache, etc. ; of green, red, sweet, bitter, resisting, etc. (b)

1 Op. cit., p. 425. 2 On the subject of analogy, cf. Ontology, 2, pp. 36-42.

3 Ibid,, n. 2, 4 Cf. infra, 129. JEANNIERE, op. cit., ibid.

1 44 THEOR Y OF KNO W LEDGE

Of all else we have only analogical concepts : of the soul, God, substance,

cause, etc. (<) At the same time things in themselves are said to be known

by proper concepts when they are known by their natural (per se) effects on

our organs, effects of which we have proper concepts. For example, I have

a proper notion of the cherry because of the proper notion I have of its sensible

effect. On the other hand when the thing in itself is not known to me by its

proper or natural sensible effects I have only analogical notions of it." The

perceptionist holds that we have proper concepts of all extramental (organic

and external) material qualities, causes, and natures or substances ; because

he holds that these are all directly given either to sense (scnsibilia per sc) or

to intellect with the data of sense (scnsibilia per accidens). For the re-

presentationist, however, the concepts under (c) cannot really be " proper," but

only " analogical " .

The progress of Idealism also forces upon us the reflection that if we once

admit the secondary qualities to be mere mental or psychic states or " sensa

tions," produced in us by external reality, we shall find it difficult to main

tain, as against the Kantian form of Idealism, that the primary qualities,

extensional or spatial determinations, are externally real. Of this we have

an interesting illustration in Prichard s otherwise very excellent criticism of

Kant s speculative philosophy.

Examining Kant s view that we are aware only of " appearances " pro

duced in us by things, he says, " To speak of appearances produced by things

is to imply that the object of perception is merely something mental, vis. an

appearance. Consequently access to a non-mental reality is excluded ; for a

perception of which the object is something belonging to the mind s own being

cannot justify an inference to something beyond the mind, and the result is

inevitably solipsism." : The principle here is that because an appearance is

"something mental," "belonging to the mind s own being," we are not

justified in inferring to the extramental. But he goes on immediately to allow

that the secondary qualities of bodies are not in the bodies, are not extra-

mental, but are " sensations " " produced in us by the extramental. Being

therefore only "sensations," or " something mental," neither can they "justify

an inference to something extramental " or give us any information as to the

real nature of the latter. What, therefore, can we know about the real nature

of the extramental, material universe ? Well, we have the primary qualities,

or " spatial relations " of this universe to fall back upon. These, Prichard

contends (against Kant), we can and do know to be extramentally real. So

the material universe is, then, (really and extramentally) merely a system of

homogeneous, space-filling, three-dimensional, spatially moving and interact

ing realities ? And how can we know even such a system of space-filling and

moving realities to be extramental ? Because the primary qualities which re

veal it are not sensations or sense-percepts, dependent on a perceiver, but are

conceived or intellectually apprehended thought-objects ; and being intellectu-

1 Op. cit., chap, iv., p. 76 (italics ours). 2 Ibid., p. 86.

3 Or, from the heterogeneous "sensations" or "mental realities" which the

secondary qualities are held to be, may we infer heterogeneity (and, if so, what sort

of heterogeneity ?) in the extramental causes of them ? Prichard holds, of course,

that the " sensations" are produced in us by extramental realities. Knowledge of

the latter he does not, however, seem to regard as inferred or infenible ftom the

former, but as attained independently of these.

SUBJECTIVE IDEALISM AND INFERENTIAL REALISM 145

ally judged to be extramentally real they must therefore be extramentally real,

inasmuch as " it is a presupposition of thinking that things are in them

selves what we think them to be V

It will be instructive to examine somewhat more in detail the line of

thought by which the author reaches these not very satisfactory conclusions.

He says we must allow the secondary qualities, even colour, which pre

sents special difficulties, to be merely facts of the mental or psychic order,

" sensations," 2 produced in us by external reality. They are not even ap

pearances of external bodies: "when once the issue is raised it is difficult

and, in the end, impossible to use the word appear in connexion with these

qualities. Thus it is difficult and, in the end, impossible to say that a bell

appears noisy, or that sugar appears sweet. We say, rather, that the bell and

the sugar produce certain sensations (not appearances ) in us." 3 The case

of colour he then proceeds to examine, and concludes that it too is a mental

state or sensation, that in respect of colour " things look what they never are " ; 4

but that the fact of their " looking " or " appearing " coloured implies that

they are (i) real, and (2) extended or spatial. 5 He then faces the Kantian

difficulty " that just as things may only look coloured, so things may only

look spatial ". {i This he meets by the contention that as a matter of fact what

things look or appear to be spatially they never are spatially ; that their real

(i.e. three-dimensional} spatial determinations are always other than, and

are in no limiting cases coincident with, what they appear spatially to be ;

that what they appear in point of spatial extension always implies " correlation

to a percipient," whereas what they really are in point of spatial extension

always is, nay means, what they are " independently of a percipient " ; 8 that,

therefore, " it is so far from being true that we only know what things look

and not what they are, that in the case of spatial relations we actually know

what things are, even though they never look what they are ". 9 But if we

have to admit " that we perceive things as they look, and not as they are," 1U

or that we perceive only what things look spatially, how, he asks (i) can we

have ever come to believe that things are really spatial, and (2) how can we

know that this belief is not illusory ? His reply is that this belief is implied

in our knowledge of what things look spatially, and that we know this belief

not to be illusory because in regard to spatial relations there is no transition

in principle, but only in respect of details, in passing " from knowledge of

what things look to knowledge of what things are " : n in other words, since it

is undeniable " that we can and do state what things appear " 12 in respect of

spatial relations, and since the possibility of knowing what things look or

appear spatially implies throughout the " consciousness " 13 or " belief" 14 that

things really are spatial, it must follow that things really are spatial.

I Op. cit., p. 100. *Ibid., pp. 85 sqq.

3 Ibid., p. 86. Cf., however, infra, 128. We agree with the author s refusal

to regard such produced mental states (if we admitted the secondary qualities to be

such) as " appearances ".

4 P. 87. 8 P. 88.

8 P. 89. All the primary qualities, being spatial determinations, are involved

in this charge.

7 P. 91. 8 Ibid. 8 P. 91. 10 Ibid.

II P. 92. l2 P. 93. i 3 P. 92. 14 P. 91.

VOL. II. 10

1 46 THEOR Y OF KNO W LEDGE

We must confess we are not convinced by this line of defending the ex

ternal reality of spatial determinations against Kantism.

(i) In the first place the author urges, against Kant s position, that "an

appearance, being necessarily something mental, cannot possibly be said to

be extended" ; that we cannot predicate of an "appearance" spatial deter

minations such as " convergent," 2 or divergent, etc.; that "an appearance

cannot be spatial"/ 1 Very well; but "sensations" are also "something

mental," and yet, on the author s view (that the secondary qualities are sen

sations), we can predicate of these "mental" facts or states that they are

" red " or " hot " or " bitter " : but if so, then, after all, why not predicate of

" appearances " that they are spatial ? It may, however, be urged that there

is, after all, a difference ; but then what about this other fact: Is not "ex-

tensity " or " voluminousness " a spatial " sensuous element " 5 or datum of the

tactual and organic senses, whose data, being secondary qualities, are pre

sumably regarded by the author as " sensations " ? When arguing that the

primary qualities, i.e. spatial determinations, are separable from colour, he says,

by way of confirmation/ 5 " moreover, if the possibility of the separation of the

primary qualities from colour is still doubted, it is only necessary to appeal to

the blind man s ability to apprehend the primary qualities though he may not

even know what the word colour means ". But can the blind man become

conscious of "the primary qualities," or "spatial relations," apart from his

tactual, organic, and motor perceptions ? And the data of these latter are

presumably secondary qualities, and therefore "sensations". " Of course,"

the author continues, " it must be admitted that some sensuous elements [i.e.

sensations, mental facts] are involved in the apprehension of the primary

qualities," but the case of the blind man shows that these may relate to sight

instead of to touch ". Yes, it shows that some of them may ; but are there

no " sensuous elements " in the blind man s tactual data ? Surely there are ;

and the point is, can he apprehend "spatial relations" apart from these?

Nay, are not the vaguely felt "extensity" and "voluminousness" of our

visual, tactual, organic, gustatory (and possibly auditory and olfactory) sense

data, themselves "sensuous elements" or at least inseparable from the

"sensuous elements" in these perceptions, if by "sensuous elements" the

author means the mental facts or data which he takes the secondary qualities

to be ? So far as introspection can discover, the " spatial determinations "

which we call primary qualities the felt "externality," "extensity," "volum

inousness," are inseparable from the secondary qualities which the author

calls " sensuous elements," in our conscious sense data. The author seems

to realize this, for he adds : " Moreover, it, of course, does not follow from

the fact that sensuous elements are inseparable from our perception of bodies

that they belong to, and are therefore inseparable from, the bodies perceived ."

Nevertheless, if we separate them we must show cause for separating them.

1 Op. dt., p. 76. 2 P. 81. 3 P. 93.

4 For these must be real predicates of something : if not of extramental realities,

then of mental states.

* Cf. p. 91 n. 6 P. 91 n.

7 So "primary qualities," or "spatial determinations" are "apprehended".

By sense or by intellect ? If by sense, then they " appear " to, and are " perceived

by " sense. If by intellect, cf. infra, p. 147.

8 Ibid.

SUBJECTIVE IDEALISM AND INFERENTIAL REALISM 147

Now, if they are separate or separable, and if we ask how do we appre

hend or become conscious of these spatial determinations at all, there seem

to be only two alternative answers possible. Either they too are sense data

and therefore :l relative to perception," J and dependent on the percipient

just in the same way as the secondary qualities or " sensuous elements " are.

This we believe to be the case. But if, this being the case, the secondary

qualities were also held to be mere sensations or mental facts it could no

longer be consistently argued against Kant s view that " an appearance can

not be spatial ". Hence the author, holding that the secondary qualities are

only mental facts or sensations, adopts the second possible alternative answer

to the question suggested, viz. that the primary qualities, or " spatial deter

minations," are not sense data at all, that they " cannot of their very nature

be relative to perception " ; 2 that they are exclusively concepts, objects of

thought, that they belong to what things are really, i.e. to what things are

for thought, for conception and judgment, and not to what things appear to

perception. 3

(2) Now this view is not supported by introspection. 4 It is, we believe,

an illustration of the dangerous tendency to which we have already called

attention (114), the tendency to lose sight of the primary qualities as concrete,

directly presented and felt sense percepts, and so to regard them exclusively

as abstract, intellectually represented concepts or thought-objects, a danger

increased by the fact that we can think and reason about sense data only

through concepts, i.e. by intellectually apprehending " what they are," so that

it requires a special effort of abstraction on our part to avoid reading into

sense data as such what belongs to them only as conceived (114). Moreover,

the view that real (three-dimensional) spatial characteristics must belong only

to what things are, and not at all to what they appear, (a) is based on the

mistaken assumption that what things really are must be wholly independent

of what they appear ; and (b] by leaving unexplained how we come to know

intellectually what things really are, it can yield no convincing refutation of

Kantism.

(a) If what things appear to sense is relative to the perceiver, neverthe

less, if at the same time we can by reflection discover the way in which they

are relative, vis. to the perceiver as organic, and not as mental and con

sciously perceptive (126), and if we can allow for such relativity in judging

what the perceived things are really and externally, then we can know what

they are really and externally even though we know that what they appear

to sense is partly relative to, and dependent on, the organic constitution of

1 Op. cit. 2 P. 91 n. 3 Ibid., pp. 99, 100.

4 Reflection on the data of sense consciousness fails to discover any difference

as to mode of presence or presentation, between directly apprehended " extensity,"

" externality," " solidity," " voluminousness," i.e. spatial characteristics, and

colour, sound, temperature, etc. That we become aware of the former with the

latter is undeniable. For consciousness they are inseparable. If it be contended

that, notwithstanding this apparent inseparability, the former are not sensuously ap

prehended at all, that they are known by intellect but do not appear to sense, the

contention can be supported only by showing that real spatial characteristics can be

proved to be of their very nature unperceivable. This the author tries to do, but,

as we think, unsuccessfully.

1 48 7HEOR V OF KNO WLEDGE

the perceiver, Of course, "our apprehension of what things 1 are is essenti

ally a matter of thought or judgment " ; - but we are by no means at liberty

to add, "and not of perception ". a For if our thought or judgment is not

an interpretation of what things appear to sense, or is not at least based upon

and motived by, what they appear to sense, what is it an interpretation of?

Moreover, it is implied, and rightly, that what things really are is what they

are for thought or judgment (i.e. for true judgment). But things cannot be

thought, conceived, judged, except in so far as they are " related " or " re

lative " to intellect, or in other words in so far as they " appear " to intellect.

What things really are (i.e. something at least of what they really are) is what

they are conceived, interpreted, represented, to be, by intellect (judging truly).

When, therefore, we contend that intellect can know things as they really are,

absolutely, this cannot mean that what they are when they are known, or

what they are known to be, is independent of their " appearing " to intellect,

but only that what they really are is not altered by this " appearing," and

that this " appearing" itself is a function of their reality. 3 This is what must

be established as against Kant. It is not established by asserting (what is

quite true) that " it is a presupposition of thinking that things are in them

selves what we think them to be " ; B and (what is true only in a certain sense)

that " from the nature of the case a presupposition of thinking not only cannot

be rightly questioned, but cannot be questioned at all ". It cannot, of course,

be really doubted, but it can be provisionally questioned and explored ; and

since, among others, Kant has explored it, he must be met by showing 8 that

the relation of reality to intellect interpreting it does not shut reality off from

all possibility of its being known as it is. But this cannot be effectually

shown, in regard to the spatial characteristics of material reality, by declaring

that these must really be as intellect conceives or apprehends them, so long

as no account is given of the way in which such characteristics become

cognitively related to, and are apprehended by, intellect.

(/ ) Prichard says " it is the view that what a thing really is it is, inde

pendently of a percipient, that forms the real starting-point of Kant s thought ". 9

But it is also an essential part of Kant s thought that what a thing really is it

1 This is true of all things, but it is only " material " or " spatial " things that

are under discussion here.

2 Op. cit., p. 99. *Ibid. 4 Cf. infra, 127, 128.

6 Prichard says " an appearance, as being ex hypotftesi and appearance to some

one, i.e. to a percipient, must be relative to perception " (p. 93). But if perception

implies an appearance of something to a perceiver does not thought imply an ap

pearance of something to a thinker ? Knowledge is an interpretation or representa

tion of something. But the " something " must appear to the knower : he must be

aware of it. Is not therefore this "appearance," or " something appearing " like

wise relative to the knower ? Cf. infra, 127, 128.

(i P. 100. " Ibid.

8 The fact that in this process of testing the capacity of intellect to know reality

as it is, we are supposing its capacity to reach true conclusions in the testing process

itself (i.e. its capacity to apprehend reality at least thus far) militates just as much

or as little, neither more nor less, against Kant s critics than against himself. The

fact, inevitable as it is, has not deterred men from undertaking the delicate and dif

ficult task of using thought to explore the springs of thought.

"P. 91.

149

is, independently of a conceiver, independently of thought, judgment, know

ledge. Moreover, both statements, as to what a thing really is, are tfue in

the sense that " what a thing really is " cannot be influenced or altered either

by perception or by thought ; but Kant defended the statements interpreted

in the sense that "what a thing really is" it can never "appear" either to

sense or to intellect. When, therefore, Prichard admits this in regard to

sense, and then tries to show, against Kant, that spatial relations really are

as we think them, or as they appear to thought, every step in his argument

can be countered by the contention that spatial relations are apprehended

(whether by sense or by intellect) in the same way as the secondary qualities

are apprehended (i.e. perceived and conceived), and that since he admits the

secondary qualities, whether in the concrete or in the abstract, whether per

ceived or conceived, to be only mental facts, so must spatial relations be only

mental facts.

For instance, answering the objection that our belief in the reality of space

(however it has arisen) may be after all an illusion, Prichard writes : " If

assertions concerning the apparent shape, etc., of things presuppose the con

sciousness that things are spatial, to say that this consciousness is illusory is

to say that all statements concerning what things appear, in respect of spatial

relations, are equally illusory. But since it is wholly impossible to deny that

we can and do state what things appear in this respect, the difficulty must fall

to the ground." 1 Admitting that " we can and do state what things appear "

in respect of spatial relations, Kant would simply reply that such statements

are illusory, and the consciousness implied by them is illusory, only if the

" things " to which they are understood to apply be regarded as extramental

things, not if the " things " be regarded as " mental facts ". 2 This implies,

of course, that it is possible to distinguish between the true or real 3 and the

false or apparent 4 within the domain of mental facts or " phenomena," 5 a

possibility which Kant proceeds to defend.

Again, in attacking Kant s defence Prichard writes : " We presuppose that

that quality is really, and not only apparently, a quality of a body, which we

and every one, judging from what it looks under various conditions (i.e. in

universal experience ), 6 must believe it to possess in itself and independently

of all perception ". 7 But how can a judgment formed " from what [the body]

looks under various conditions " tell us what or how the body really is, if the

body never under any conditions looks what it really is ? What right have

we to form such a judgment, or to entertain such a belief, if all the body s

appearing or perceptible qualities (" what it looks under various conditions ")

are, by being relative to a percipient, not real (because not " independent of

1 Op. cit., pp. 92-3.

2 We might also argue that the statement that a thing appears red implies the

consciousness that it is coloured. Yet Prichard holds this consciousness to be il

lusory unless " thing " be understood to mean a " sensation " or " mental fact ".

Hence he must either hold that all statements based on it are illusory, or else recog

nize the possibility of distinguishing between "real " or " true " appearances and

" deceptive " or " false" appearances within the domain of mental facts or sensa

tions, precisely what Kant does regarding spatial relations.

3 Erscheinung. 4 Schein. * Cf. ibid., p. 79.

6 Critique (MtJLLER), p, 37. 7 P. 99.

i 50 THEOR Y OF KNO W LEDGE

perception "), and if its real qualities (i.e. real three-dimensional extension,

real shape, real position, real motion, etc.) never appear, or are never per

ceived ? How does intellect apprehend these latter, if not in the data of

sense? "We do not perceive but thinks thing as it is." 1 But how or

whence does intellect apprehend the predicates whereby it thinks the thing, if

not in the data of sense ? Prichard, however, has admitted that the data of

sense, or " what things appear," are relative to the perceiver in the sense of

being themselves mental facts ; 2 and he sees that predicates or f/iought-ob-

jects, derived from data which are relative in that way could not yield know

ledge of what things other than mental facts can really be. The only alterna

tive is that intellect can apprehend what things really are, directly and of itself,

and apparently without any aid from sense. Of course what a thing really is

in respect of spatial relations is directly " correlated with thought," :i because

it is thought, not sense, that interprets what anything really is ; but even if

its correlation with thought did not involve its correlation with sense also, we

should have not merely to assert, but to show, as against Kant, that the former

correlation did not alter or transform the reality : and we should therefore

have to show how thought apprehends the extramentally real, a point to

which Prichard does not appear to have directed his attention. Correlation

of reality with thought docs, however, involve correlation of reality with sense.

Nor does this necessarily involve the confounding of thought with sense,

however Kant may stand in relation to this charge. 4 For our intellectual

knowledge of what material, spatial things are really and extramentally, is cer

tainly derived from what they appear to sense. And the only proper way of

showing, as against Kant, that nevertheless we can know intellectually what

things are really and extramentally, is by exploring the nature of the relation

in both cases, and showing that, when rightly understood, it does not screen

off the extramental reality either from intellect or from sense.

We shall have more to say later in criticism of the general position that

what we become directly aware of, whether in sense cognition or in intellectual

cognition, is and must be only a " representation " or " appearance," " pro

duced " in us by a something which this mental product is supposed to repre

sent. 5 But before doing so we have next to examine the difficulties which may

be urged against the form of Intuitive or Perceptive Realism so far outlined

in our inquiry.

Before passing from Idealism we may here note an objection

which is sometimes unthinkingly urged by idealists against real

ism in general. Vibrations of the air or the rcther, they argue,

are quite unlike our sensations of sound and colour : therefore

1 Op. cit., p. 99.

2 He admits that spatial characteristics which are de facto real and external ap

pear to sight and to touch (cf. p. 91 n.) ; but holds that they appear otherwise than

they really are : " in the case of spatial relations . . . things . . . never look what

they are". Hence predicates derived from " what they look" could never inform

us as to what they really are, any more than if " what they look " as regards spatial

relations were merely mental facts or " sensations," which he considers the second

ary qualities to be.

3 P. 100. 4 P. 99. " Cf. also vol. i., 92, ii. ; 93.

INTUITIVE REALISM 151

our sensations in general cannot give us any information about

the nature of external reality. Such an objection, coming from

an idealist, is suicidal, for, as Rickaby points out, 1 "he forgets

that it has been by the senses that the vibrations have been dis

covered, and that if the scientific result is worth anything, it

proves the ability of the senses to give us information about the

facts as they are in external nature ". In other words the objec

tion is based on an assumption which is the very negation of

idealism, viz. the assumption, common to physical science, that

at least motion and extension are really and externally identical

with, or similar or analogous to, that which we internally perceive

them to be.