MASTER NEGA TIVE NO. 91 -80003 MICROFILMED 1991 COLUMBIA UNIVERSITY LIBRARIES/NEW YORK as part of the "Foundations of Western Civilization Preservation Project" Funded by the NATIONAL ENDOWMENT FOR THE HUMANITIES Reproductions may not be made without permission from Columbia University Library COPYRIGHT STATEMENT The copyright law of the United States ~ Title 17, United States Code - concerns the making of photocopies or other reproductions of copyrighted material... Columbia University Library reserves the right to refuse to accept a copy order if, in its judgement, fulfillment of the order would involve violation of the copyright law. AUTHOR: CASE, THOMAS TITLE: PHYSICAL REALISM PLACE: LONDON DA TE : 1898 Restrictions on Use: COLUMBIA UNIVERSITY LIBRARIES PRESERVATION DEPARTMENT BIBLIOGRAPHIC MICROFORM TARGET Original Material as Filmed - Existing Bibliographic Record Case, Thomas, 3844- «ans, Green, and co ,1188: ^"°*^"''' ^'^ Yorkf l2ng 3 p. I, 387 p. 23"^, 1. philosophy. 2. Realii^. ,. ^j tie. Title fro,„ St. Paul.Pab. ij , ., ^ '*"''' ^''"■- Printed by I,.a wmmfm^mJ Master Negative # TECHNICAL MICROFORM DATA FILM SIZE:_3S_02lIX1^- REDUCTION RATIO:, IMAGE PLACEMENT:! IAUIA) IB IIB DATE FILMED:_ll/ltSj INITIALS._jALi^ HLMEDBY: RESEARCH PUBLICATIONS. INC WOODBRIDGE. CT M ..XE Association for informntion and image Management 1100 Wayne' Avenue. Suite 1100 Silver Spring, Maryland 20910 301/587-8202 Centimeter 123456789 MlJllMllMillllllllllllllllllllllllllllllllllllllllllllllilllllllllllllllllllllllllll ill 10 11 12 iiiliiiiliiiiliiiiliiiilii 13 14 15 mm liiiiliiiiliiiiliiiiliiii rTTTT Inches TTT 1 I II II i 11 1 I 3 1.0 M 2.8 ilia m III 3-2 ■ 43 ^^ 3 6 If b£ 1.4 2.5 ?? I.I 2.0 1.8 1.6 1.25 TTT TTT MflNUFflCTURED TO RUM STflNDRRDS BY RPPLIED IMRGE, INC. 00 ^^^3 intI)f(CifpofiJmgork LIBRARY I: PHYSICAL EEALISM N X r PHYSICAL EEALISM BEING AN ANALYTICAL rilll-OSOl'IlY I'lllNlKD BY SrOTJliSWOODE AND CO., NEW^TnEET SQUAUE LONDON i'KOM THE PHYSICAL OBJECTS OF SCIExNCE TO THE PHYSICAL DATA OF SENSE BY rp THOMAS CASE, M.A. ■ FELLOW AND SKNIOlt TLTOU COHPLS CHKIsTl COLLKUE, AND LKCTLIIKIt AT •HHIST CIlLJltll; t'OlIVIKHLY FELLOW OF BHAHKNUSK AND TLTTOU OK BALLIOL ( OLLKCK, OXKolM) cVn d dnodfi^ut (XeyKTiKcos— Aiiistotlk LONDON L X G :\L A X S, G li E E X, A X I) AND NEW YOKK : 15 EAST 10"' STKEET 1888 CO. ,4 // , i ijhf .\ If SI r ml 00 ' Neque tamen illis nihil addi posse affirmamus : sed contra, nos, qui Mentem respicimus non tantum in facultate propria, Bed quatenus copulatur cum Rebus, Artem Inveniendi cum Inventis adolescere posse, statuere debemus.' Bacon, Noi\ Org. i. 180. )\ :> \K) TO WILLIAM S. SAYOEY, F.R.S. SURGEON-EXTIJAORDIXARY TO II.M. THE QUEEN I'KESIDEXT OF THE ROYAL COLLEGE OF SURGEONS OF ENOLANH SUltGEON TO ST ISAKTUOLOMEW'S HOSPITAL -^ C3> c::. ^ t O a5 S o 73'31 c^ CONTENTS. PART I. GENERAL PBOOF OF PHYSICAL BEALTSM. cirAr. I. THE PHYSICAL OBJECT OF SCIEXCE .... II. IDEALISM AXD REALISM • * • • . III. THE PHYSICAL DATA OF SENSE • • • • IV. THE HISTORICAL ORIGIN OF PSYCHOLOGICAL IDEALISM 3 13 40 82 PART II. PSYCHOLOGICAL IDEALISM. V. DESCARTES VI. LOCKE VII. BERKELEY .... VIII. Berkeley's theory of vision . IX. HUME X. KANTS 'critique' AND NECESSARY TRUTHS 101 141 18G 225 256 319 \ APPENDIX 383 Part I. GENERAL PROOF OF PHYSICAL REALISM. ♦ Itaque contemplatio fcrc desinit cum aspcciu ; adeo ut rerum invisibilium exigua atit nulla sit obscrvatio.' Bacon, Nov. Org. i. 50. B 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 sino-le drop may, as we have reason to believe, amount to somewhere about 10^^', 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 'Eecent Ad- vances in Physical Science' is a recognition of the reality of the insensible, and of its knowledge by the natural philosopher, as facts, ^o 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 PHYSICAL REALISM PART T. CHAT. I. THE PHYSICAL OBJECT OF SCIENCE explains liow these scientific objects of human know- ledge are known from the original data of sense. "xhe distinction between the sensible and the scien- tifi(5, 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 quaUties 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 exphiined— 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 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 cfoinfr round the sun, except by standing on tlie sun itself. No niiui 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 jDhilosophers latent structures and molecular processes to masses and tlieir 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 phaeno- mena. We are quite as certain of the existence of PHYSICAL REALISM PART I. that which cannot be brouglit 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 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 rether 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 connnuni- 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 w^ords, 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 pha3nomena, not only actual in nature, but known to science. There is a thing real and known which is not a sensible phsenomenon, because such things as imperceptible particles are known really to 8 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 phae- 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 aether 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 mif^ht mean a thinfij 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 ph^enomena 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 phaenomena ; and man, as a natural philosopher, knows things in themselves which are not phaenomena, when he knows imperceptible particles; knows not merely the phenomena which CHAP. I. THE PHYSICAL OBJECT OF SCIENCE 9 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 phaenomena, 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 phtenomena of sense. All this knowledge does man as a natural philosopher possess of things m 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 w^th reason to knowledge, without flying to faith. That we know invisible nature beyond sense in natural philosophy is a simple fact, expUcable 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, \\^ 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 wull conclude this chapter by quoting, from Sir John Ilerschel'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 phaenomena, 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 eye- lids, and in much less than a swift runner occupies in takinix a single stride ? Wliat mortal can be made to believe, w^ithout demonstration, that the sun is almost a million times larger than the earth ; and that, although CHAP. I. THE PHYSICAL OBJECT OF SCIENCE 11 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 comnmnication 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 amonf*- 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 our 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 astonishinir 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 12 PHYSICAL REALISM PART I. CHAT. II. 13 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 eighty- two millions of millions of times ; of vellowness, 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 ravinirs 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 trouljle 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.' CHAPTER II. IDEALISM AND RExVLISM. 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- ledtre. 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 J 14 rilYSICAL REALISM FART I. CHAP. II. IDEALISM AND REALISM 15 liow 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 ideale 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. n 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 phsenomena, Mill sensations. The most usual terms of the present day are sensations, feelings, psychical pha^nomena, 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, ini- / r 16 PHYSICAL KEALISM TART 1. CHAP. II IDEALISM AND REALISM 17 pression, ])hn2nome]ion, sensation, feelhifr or state of consciousness. There is a further difference among the idealists. Some of them, beginning witli 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 oidy 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 amon^r themselves about the physical world. Descartes held that a physical w^orld 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 unwaril}^ 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 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 consider-ation 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 + 1 = 2 ; the other, agreement in a common hypothesis. Now the proposition that all sensible data are psychical phaenomena is not a self-evident principle, but a de- batable hypothesis. Eealism is the philosophy of a reality beyond psy- chical facts. Th'fe 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 substances, which are concretions of matter and form. In 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 f? 18 14IYSICAL REALISM PART T. CHAP. rr. IDEALISM AND REALISM 19 conceptions. Those who adopted the latter view were afterwards called Conceptualists. It is not necessary to be either a Platonist or an Aristotelian. There is a third realism of universals possible ; and that, too, without falling into nominalism or conceptiialism. The theory of the reality of univer- sals, though overlaid with manv 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 Hving 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- ino- 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, w^hose similarities can be expressed in laws, thare would be no science ; and w^ithout 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, referrintj all the facts of colour back to the undulation of aether. A natural class, then, is the sum of individuals possessing an indefinite number of similarities. A real 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 aithereal 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 aether began before, goes on without, and will last after, our names and notions. If, on the otlier 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 tlie 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 reahsm of individuals rather of modern, philosophers. Not that Aristotle rejected the distinct reahty 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 psycliical 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 Kealism. Eealism is constantly misunderstood. It is some- times supposed to be a synonym for mere SensuaUsm, or the belief that physical things are as they appear to our senses. But sensualism is only a crude form of reaUsm. There is a reaUsm 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 MateriaUsm. 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 reahties, 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 reals 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 REAUSM 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 aU 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 ideahsts, 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 ideahsm. 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 oiignoratio ^Z^/zc/a', which 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 22 PHYSICAL REALISM PART 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 tlie 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 Realism. 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 notliing 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 estabhshed 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 Jiot 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 suflScient 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 liave seen, physical objects are scientifically inferred from sensible data. It follows that the sensible objects, which are these data, must also be pliysical. The similar can be inferred only from the similar, therefore the physical can be inferred only from the physical. This conclusion, liowever, places me in a dilemma. 24 PHYSICAL REALISM PART I. 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 w^orld, 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 w^hy 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 niYSICAL REALISM PA.RT I. similar data to similar objects too far. 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? J^ecause, 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 hypotlietical realism. As they are physical data within, to infer physical objects without, the realism which I advocate may be called Physical Kealism. 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 CHAP. II. 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 reahsm, 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 Eeahsm must be especially distinguished from intuitive, or, as it is also called, natural reahsm. 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.'^ The rejection of idealism, the reahty of the physical world, the belief in a phy- sical object of sense, and the possibihty that a psychical subject may apprehend a physical object, are all points in intuitive realism which find a place in physical reahsm. 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 thinkint? too much of the sensible data, and too Httle 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 * Reid, Essays on tlie Intellectual Powers. Essay IL, 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 natuix he contradicts science. But we must appeal from common sense to universal science, and from ordinary to scientific realism. The ideahst 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, tliey admit the ideaUstic 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. ir. IDEALISM AND REALISM 29 realists, having confused object and non-ego, supposed that whenever sense has an object it presents the non-ego. Eeally, sense always apprehends an object distinct from the operation, but never a non-ego distinct from the ego, that is, tlie man himself. Hence, also, their erroneous behef that in apprehending a primary quality, as an object, sense presents a quahty of the non-ego, and in not apprehending a secondary quahty as it is in the non-ego, it presents no object. Eeally, as sensible, botli 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 quahties are found by science to be only specific varieties of primary quahties. In the internal world, all qualities appear to sense to be equally elementary. As sensible, a primary quahty, 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 beincr, 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 quahties, 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 niYSICAL REALISM PART I. the combination of two causes. As active or efficient cause, the external world produces the sensible effect m the nervous system; as passive or material cause, the nervous system receives this effect according to its susceptibility. Hence the effect is hke 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- flectinfT the waves of the sea than the undulations of aether. 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 CHAT. II. IDEALISM AND REALISM ^1 'j1 physical reahsm 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 be found in the construction of the optic nerve, wliich, wlien acted on by a certain impercepti])le motion of asther, 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 Hke 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 quahties as they are in nature is being constantly ehminated by scientific reasoning. Thus, sense sometimes presents motion as motion, but cannot help presenting the hot, the red, ' 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 li^dit undulations of ffither. 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 uninq^ortant. 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 suflfice 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 a3thereal 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, ' -ound ' and ' hght.' Secondary quaUties 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 Light,' all the more valuable because it was not 8G niYSICAL REALISM PART 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 beUeve, 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 hi other lan'nia^'-es, it will be convenient to use the same word to desi<»'nate the accent considered in itself, and irrespectively of its capacity for exciting vision, a capacity which would be regarded as a mere accident of li^dit, in the technical logical sense of that word. Accordindv I shall now use the word " hght " 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 ordine ad universum. 1. Primary, original qualities; e.g. duration, extension, motion. 2. Secondary, specific modes of primary quali- ties ; e.g. sound, heat, hght, as modes of motion. II. Internal, in ordine 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 rSYCIIOLOGICAL 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 ahke in both cases. The difference entirely arises when tlie scientific man begins to infer that external secondary ai-e modes of primary qualities, because their sensil)le effects are so similar to those of primary quahties ; for instance, tliat the effects of external sound, heat and h'glit ai-e tlie effects of motion l)y the laws of motion. To the Atomists is due, not oidy the foundation of the theory of primary and secondary ([ualities, but also the discovery that the object of sense is not the extei-nal thing itself, but an effect produced by the external thing on the senses. They supposed that effluxes, contimially thrown off" from bodies, come into contact with our organs.^ They thus anticipated modern j^hysical in([uiiy on the senses, although their necessary ignorance of the laws of motion prevented them from reahsing the vibra- tions and undulations, which have taken the place of emissions, in the case of hearing, sight, and the perception of temperature by touch. The consequence of this sup- position to the theory of knowledge in Greek pliilosoi)hy was that its immediate object was henceforward ^^-ene- rally agreed to be not the thing at a distance, but a result of the thing on the organs of sense. In the Atomistic theory the immediate object of sense, though internal and representative, is neither im- material nor psychical : it is a physical object. This point has never been disproved. Modern physiology, as we have seen, has brought the motions of matter as for ] Arist. De Divin. per Somn. 2 = 4G4 A C (Berlin ed.) ; cf. Tlut. De Plac. Phil. iv. 8. 88 PHYSICAL REALISM PART I. as the physical substances of the nerves; but it has never shown tliat 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 pliilosophy. It even left a relic in the philosophy of Plato, who always represents sensation as a motion communicated from matter through body to soul.^ 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 inmiaterial ' 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 materiahsm. For atoms he substituted primary matter ; instead of figure, position, and arrangement, he regarded heat and cold, dry and liquid, as its primary contrarieties.^ The Atomists considered the external thing to be wholly corporeal ; Aristotle divided it into two heterogeneous substances — corporeal matter and incorporeal form ^ — 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 » Plato, Phil. 34 A ; Ti7n. 42 A, 64. ^ Arist. De Gen. et Corr. ii. 1. ^ Id. Met. Z 7 = 1032 B 14. cuAr. IV. ORIGIN OF PSYCHOLOGICAL IDEALISM 89 external thing is a corporeal efflux, 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 metaUic seal into wax.^ 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 tlie 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 alo-dybv €lSo9, 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 quahties ; 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 ^ Arist. De An. ii. 12. 90 niYSICAL REALISM PART r. not perceive the external thing, wrong in thinking that what it perceives witliin is an incorporeal form. Hamilton has misunderstood these Aristotelian errors.^ He says truly enough that Aristotle distinguishes proper from common objects of sense,^ and that the former agree with tlie secondary, the latter with primary quali- ties. Jhit he misses tlie 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 toucli ; 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 quaUties as heat and cold affective quahties, be- cause they produce affections in us.^ But we must not therefore infer that he meant either that they produce this effect through insensible primary quahties, or that they are themselves mere affections in us, or that, being qualities outside, the affections are not hke 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 ^ See Eeid's Works, ed. by Hamilton, Note D, on Primary and Secondary Qualities. « Arist. Dc An. ii. G. 3 i,|^ (.^^ 8 = 9 A 28 seq. CHAP. IV. ORIGIN OF PSYCHOLOGICAL IDEALISM 91 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 o])ject and the ol)ject perceived.^ 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.^ Hence, in sensible knowledire, he supposed that the external object propagates the form of the sensible quahty, 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 ^ Arist. De An. iii. 2 = 425 B 25-7. 2 Id. Met. Z 7 8, esp. 1033 B 29-1034 A 8. 92 PHYSICAL TJEALISM PART I. CHAP. IV. OUIGIN OF PSYCHOLOGICAL IDEALISxM 93 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 efflux, 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 ol)jects into the sensitive faculty, the latter by active intelligence into passive intelhgence. Aristotle has not explained this mysterious influence of intelligence on intelligence in the same soul ; nor is it 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 Avitli many matters of different members of the same kind, but also to be something different from matter, even when so closely conjoined with matter in f^ict, and so inseparable from it in definition, as concavity with nose in snubnoye, 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 witli matter in a material substance. The form of God Him- self is pure, not in tlie sense of being less material than other forms, but only in the sense of never being con- joined with matter. Hence, sensible and intelhgible species or forms arc nnmaterial, 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, wliicli 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 pliilosophy, 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, hke 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 rilYSICAL REALISM PAUT 1. vestige of resemblance to tlie pliysical cause by identi- 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 w\as 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.^ Descartes thought that it is only a psychical idea.'^ 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. He might also, especially since Galen's discoveries in the nervous system, have restored the Atomistic theory that the object of sense is a pliysical effect on our organs, and have added that it is an effect on the nervous svsteni. 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 ' Nov. Org. i. 51 ; ii. 17, 20, 52. ^ Princ. i. 58. CHAP. IV. GRICUN OF rSYCIIOLOGICAl. IDEALISM 95 himself as if the idea itself were not distinct from the nervous imprint. But he finally and dehberately sepa- rated it from the physical effect in the brain in his Eephes to the Objections raised ao^ainst his Medita- tions. The ' Eesponsio ad Secundas Objectioiies ' 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 : — ' 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 nothin^% conscious of the supposition, Avhich 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 bodv I should still be thinking. 104 PSYCHOLOGICAL IDEALISM PART II. 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- scious of the former, not of the latter. I am conscious that I am a thinking subject : I am not conscious that this thinking subject is not body but soul. Secomlly, 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 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 principle, cogito, ergo sum, but is vitiated by the hypo- theses combined with it. Thus does Descartes lead his reader to confuse the thinker and the soul, and transfer the conscious certainty of being res cogitans to the hypo- thesis of being res a corpore plane distincta. That I 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: the 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 for 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 I 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 Avho 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- 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. I am con- scious of using my bodily senses in touch, taste, vision, hearing, and smell. I do not consciously first feel the 106 PSYCHOLOGICAL IDEALISM TART IT. sensation and then refer it to the l>odily member ; I am not conscious of these two stei)s. Reasoning is the highest kind of thinking ; I am conscious of doing it m my liead, and by no force of abstraction can I get it out of my liead. Similarly I am conscious that I will in my head, and I am conscious that my head may ache with reasoning, and dehberating, and resolving. My body is not a mere companion but a conscious partner of all my thoughts. But consciousness is a superficial power. In speak- ing of the data of sense I remarked tliat by an illusion, arising from the confusion of sense and inference, we cannot help seeming to be conscious that sense per- ceives an external object, though we can make ourselves independent of the illusion by science, which dis- tniguishes 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, which shows that what performs^ the operation is not 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. J^ow, thouoh we are sensible of the optic nerve so for as it is sensibly affected with red, we are neither sensible nor conscious of it as nervously constituted. But from very early inf^mcy 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 what sees is not our 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. ^"ow, 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 PART II. the body reveal tlie 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. It 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 to external observation, I am conscious of the pain of hunjxer somewhere in the re^^ion of the stomach ; science reduces this indefinite verdict to definiteness by proving the 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 tli3 nervous structure ; if we were not already conscious of reasoning in the head, anatomy 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, liowever wrong CHAP. T. DESCARTES 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- ence between a sensation of an internal sensible object and the inference of its external cause, as we have 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 cause 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, a prize-fighter may be made by a blow to see stars ; a drunkard under the influence of delirium tremens may no PSYCHOLOGICAL IDEALISM TART II. have a vision of tlie devil. Such sensations are excel- lent instances to show 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 that the soul is the sole subject of vision. A prize- fitditer seeinf? 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 the 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 body. 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. Now, ordinary and scientific observation being limited to the body, 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 111 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 spirituahsm. It does not follow that the bodv 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 hmited. 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 stdl remain to prove that nervous motion is completely identical with 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 diffi- 112 PSYCHOLOGICAL IDEALISM PART II. 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 light, 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 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 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 pperations 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- mg ? ^ This is the second question, suggested by tlie consciousness that I think, but not answerable without further argument. Descartes assumed that all the immediate objects are psychical 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 objects of sense and all other innnediate objects are not only internal but psychical, are ideas. Descartes never proved this ideal theory. In the Third Meditation we find the foUowin^r passage : — ^ ' Nevertheless I before received and admitted many things as wholly certain and manifest, M^iich 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 mnid. And even now I do not deny that these ideas are found in my mind. But there was yet another thuig which I affirmed, and which, from having been accustomed to beheve it, I thought I clearly perceived although, in truth, I did not perceive it at all ; I mean the existence of objects external to me, from which ru PSYCHOLOGICAL IDEALISM PART II. those ideas proceeded, and to wliicli tliey 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.' ^ 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. Science 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 ol)ject 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, he defined soul as a purely thinking substance, and body as a purely extended substance. From these definitions he deduced tlie heterogeneity of mind and matter, 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 Cai^tesian evidence is this : ^ ' Ex vi meae perceptionis,' in the Latin edition. CHAP. V. DESCARTES 115 tlie subject is soul, the soul is such as to apprehend only /^ K^as; therefore all immediate objects arfideas. B T neither premise is proved. / It is not true that the wliole subject is the soul Descartes as we have seen, exaggerated the soul from a part to he whole thinking subject. The man is the whole subject : the body is part of that by which le hinks ; and, being a factor in thinking as well as ^^, It is not a purely extended substance. The assumntJon at the bottom of the Cartesian definition of b^C tnnking and extension are different. So they are h wfTtf ' '''' """' ^''"°" ™'^J' P°^«««« ^^^^^ attri- butes in he concrete. Number is not extension, but it rt ""? " "r ''''''' '-'"^ <^^^^"^«^ ■' ^-tensi;n is Zl tS"'wr " """ '"'^' '"^^' ^^ ^°''^ -tended and think. When we appeal from abstractions to con- sciousness, we find it does think. The body, therefore >s not purely extended substance, but also thinkin '' Again the soul is a factor in thinking, and is in othe^r respects latent : it does not follow that it is nothing e e Eather such a supposition is impossible ; for, as Locke wittily remarked, men think not always, .nd iV the soul -ere purely thinking substance, either 'it must a W tlnnk in order to be, or it must have an intermittei xistence, both of which alternatives are impossible d absurd Descartes resolved body and soul Lo the wo opposite abstractions of extension and thinking. Bu^ he did not thereby prove that body is purely ex'tended nor that soul is purely thinking, nor their heferoge^ y' nor hat the body is no factor in thinkin., nor fhTit whole thinking subject is the soul. " "" But if we concede that the soul is the whole thinkin<. subject, and that thinking is accordingly a pure v psychical operation, whether it be feehng, perceiving I 2 IIG rSYCIIOLOGICAL IDEALISM PART II. reasoning, willing, or wliat not, wliat do we know abont its nature ? On this point we have a dictum of Sir W. IlamiUon, 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 liope to attain any insight into the question.'^ The most we know of the soul is that it thinks, whatever else it is ; we camiot enter further into its secret nature to 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 pliysical 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 infer pliysical things without. I conclude, therefore, that, as we apprehend the pliysical 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. It is not true, then, that the soul is such as only to per- ceive 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 perceive ideas ; for all we know of its thinking is that, as a matter of fact, it immediately perceives pliysical, tliou<^h internal, effects on the nervous system. Thiidiing 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 * Hamilton's Mctaj;)hjj8ics, Lect. xxv. p. 122. CHAP. V. DESCARTES 117 still be physical. Granted, then, that the subject is the soul, it is a 7io?i sequitur that it perceives only psychical ideas. A fortiori, if the subject is the man, he can perceive the physical in himself. A certain conditional plausi- bihty 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 ideal theory of the sensible object. The physical can appre- hend the physical, the extended the extended, within the nervous system. The thinking subject, body and soul, does apprehend the physical: it therefore can. Wliat we apprehend as a fact is better known than what we are to apprehend it. Kiiowable objects must be explained, not denied, by knowing subjects. Descartes was a clear and distinct writer; 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 PART II. 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- sation. But why an idea ? Because he merged sensati^^n 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 alad-qai^ and <^ar- rao-ta, and their ol)jects as aLo-Orjfxa and (pdvTao-fia. 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 tlie 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 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.^ 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, ' Cf. Princ. i. 9. CHAP. V. DESCARTES 119 from which mental philosophy has never quite recovered itself. So much for the evidence of the Cartesian theory 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 reasonin^p ; and if anv 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 not sensible. There are, according to him, three / 120 PSYCHOLOGICAL IDEALISM PART II. sorts 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, under the head of innate ideas, it recognises ideas which are not sensible. He remarks that ' the philoso- phers of the schools accept as a maxim that there is nothing in the understanding which was not pre- viously in the senses, in which, however, it is certain that the ideas of God and of the soul have never been.' ^ 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 ' Eesponsiones Tertia3 ' 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.' '^ This doctrine of ideas, innate in the sense of elicited from one's own faculty of thinking, is developed at length in the ' Notes on the Programme of Eejects, liow do we know them? How do we know when our iudf^ments acrree with them ? What is the criterion of truth ? Objects are twofold, internal and external. About the internal we judge immediately by sensation and consciousness, tlie ol)jects 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 judo-e 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 tlie conscious within, and tliereupon 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 1; '19 rSYCIIOLOGICAL IDEALISM TART II. truth. Knowledge is tlie apprehension of reahty, imme- diately by sensation and consciousness, mediately by logical reasoning therefrom. To know by reasoning requires at least two conditions ; sensible data and logical 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- clusions, 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 all be, as he supposed, a soul and its ideas. Descartes was a man of sul)tle genius, retiring, as it were, within the chamber of his own soul to survey his own ideas, and trying to think wliat they could reflect of a world without. Let us follow him into tliis retire- ment, and imagine ourselves each to l)e a pure soul contenq^lating pure ideas. A man must have a diffi- culty in performing this feat, because he neither is the one nor does the other. He cannot rid himself of his body, nor fail to contemplate the effects on his nerves. Pliilosophers take advantage of this superhuman diffi- culty ; feigning a psychical man, but knowing all the CHAP. V. DESCARTES 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- 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 w^ould 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 incommunicabihty, 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 voUtions, 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 knew 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 1DEALIS.M PART II. conclusion and contradict science. The Cartesian Scliool denied that the body is either cause or effect of ideas. Instead, Geuhnx invented occasionahsni, or the hypo- tliesis 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 liave 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 bodv 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 of 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 fire as part of myself. It may be objected that I should find it always with me. But so I do the earth and the atmosphere. It would seem with ClIAP. V. DESCARTES 1*^ no them 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.' ^ 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 inconsequently and inconsistently admitting that he is body and soul. 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 ^ Discourse on Method, Part IV. ^ MeJltation VI. p» 13G rSYCIIOLOGlCAL IDEALISM PART 11. 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,^ 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 by 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- natives : another soul ; this would be Berkeley's Divine Spirit : a cause unknown ; this would be Hume's inex- pUcable something : another idea ; this would be Hegel's absolute idea. A logical ideahsm 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 * Princ, ii. 1. % \} CUkV. V. DESCARTES 137 such a thinix, must follow from another similar system of my ideas. Such a logical deduction escaped Des- cartes, but it has not escaped Mill, who only sub- stitutes sensations for ideas : ^ — ' 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 data * Examination of Sir William Hamilton's Philosophy, pp. 244-5 cf. Lotze, Metaphysics, Book III. chap. iv. 138 rSYCIIOLOGICAL IDEALISM TART H. CHAP. V. DESCARTES 139 enable us to infer physical bodies as causes beyond sense and conception in tlie external world, that these bodies consist of insensible particles, that the external world is like the sensible in some qualities, unlike in others, 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 pre\'ious quotations one passage as a sample of this profound scientific spirit : — ' But to the insensible particles of bodies, I assign determinate fii^ures 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, considered generally, what could be the principal differ- ences among the magnitudes and figures and positions of bodies, insensible only on account of their smallness, and 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.' ^ 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 premises of such a science of bodies and their insensible particles. If all immediately perceived effects and causes had been soul and ideas, there would have been * Princ. iv. 203 t 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 physical data of sense. After the dogmatism of media3val philosophy, Des- 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 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 140 PSYCHOLOGICAL IDEALISM PART ir. 141 o])jects are not ideas but pliysical effects on tlie nervous system, and because soul and ideas would not enable man to infer pliysical 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.'^ This seclusion in a hot room is an admirable w\ay of distilling thoughts, provided only these vapours of the heated brain can be condensed into a knowled^fe of the outside world. * Diss, de MetJiodo, ii. I CHAPTER VI. LOCKE. Locke, at the outset of the ' Essay concerning Human Understanding,' states tliat 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.^ Tliat is, he rejects the Cartesian method of using the nature of the thinking subject to "" deduce our knowledge ; and rightly, because it was a ' method from tlie less to the more certain. But he leaves the Cartesian deduction, that the data of the under-* standing are ideas, simply removes the hypothesis from the premises to the conclusion, and nowhere throughout gives any new evidence that ideas are the data of know-| ledge. ITlie hypothesis of the soul is thus replaced by the hypothesis of ideas, as a principle. -.JSTow, there had been some plausibility in tlie argument — the subject is the soul, tlierefore its immediate objects are ideas.t There was notliinghut j^etitioprincipii in the hypotliesis — >the immediate objects of understanding are ideas, * Yet this hypothesis in one form or otlier has remained ever since Locke's time as the putative principle of all ideahsm. Many a philosopher, wlio 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- ^ Essay, I. 1, 2. U2 rSYCriOLOGlCAL IDEALISM PART If, 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. h First, I shall inquire into the orhjinal 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. )f Secondly, I shall endeavour to show what knoiv- 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 wdiose truth vet we have no certain knowled<]^e. And here we shall have occasion to examine the reasons and degrees of assent' ^ From this passage we 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 LockS^it was necessary that ideas should be all the materials of knowledge ^ for the next question * E^say, I. 1, 3. CHAP. vr. LOCKE 143 is— what 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 ^ w^itli an understanding of ideas and end with a know-/ ledge of things. The fu'stfruits 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: : — ^ V ' 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 fmd 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, I have used it to express whatever is meant by phantasm, 7iotion, species, or w^hatever it is wliich the mind can be employed about in thinking; and I could not avoid frequently using it.' ^ V These words, wliicli, if anywhere, ought to have come as a proved conclusion at the end, occur as an undoubted principle at the entrance of the Essay. Tliey contain a double hypothesis ; first, that ideas are the innnediate, secondly, that they are all, the objects of understanding, and therefore of knowledge. The first part is the ideal lu'pothesis of Descartes, the second is * Essay, I. 1, 8. 144 rSYCIlOLOGICAL IDEALISM TART II. CHAr. VT. LOCKE 145 Locke's corollary. It is a logical corollary, not however scientific, but hypothetical from. an hypothesis. Three hypotheses started modern 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, witli the following peroration : — ' This seems to me the first and most general, as w^ell 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 ccelo different, they seemed to be the three great provinces of the intellectual world, wholly separate and distinct one from another.' ^ In the same chapter he has already told us what he . includes under things and signs. On the one hand, by si- . ^ perceive anything simple in tlie sense of a simple quality, wliicli is only simple in tlie sense of abstract ; but I perceive at least the sim2:)ly 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/AVTlie 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 jL sVibstance. ^ 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,^ Locke comes to the operations which he supposes to make oJ,her"^ideas out of them, and to the ideas thus made.^ ' The*act^of themind,' sayshe, ' wherein it exerts^^ y its power over its simple ideas, are chiefly these three : ( First, combining several simple ideas into one compound V 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.'^ He then re- marks that ideas, juade 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, > Essaij, IT. 2-11. 2 II. 11 scq. to the end of the Third Book. « II. 12, 1. CHAP. VI. LOCKE 1r ^ relations. Complex ideas of modes are ideas of affec- tions of substances, subdivided by him into sinqjle, 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; ' ^ 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,' 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.^ The consideration of all these com- plex ideas in their order occupies the remainder of the Second Book ; while that of abstract ideas follows, alonir with general words, in a general treatment of language . in the Third Book.^ ^ The whole discussion is full of variety. But it is » vitiated by two incurable errors. Li the first place,\ the objects of knowledge are complicated with their J \ * mere ideas. But many scientific objects are known to exist, without being conceivable. Secondly, no thorougli*^' 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 r composition, comparison and abstraction ; ' 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, G. « H. 12, 7. * Cf. II. 33, 19. ' II. 25, 2. » n. 11. 134 rSYCIIOLOGICAL IDEALISM PART 11. He saw tlie 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 jjeneral 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,^ 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, startingf from judgments of sense, we infer rational judgments on evi- dence about relations, as real as the sensible relations. Kor 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, hke 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- ^ 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- 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.^ But reason also * Essay, II. 18, 1-2. 150 PSYCHOLOGICAL IDEALISM PART IT. infers simple modes of inseiisiljle 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 simjjle 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 reahty 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 morahty itself to an idea.^ ' Cf. Essay, II. 22, 2; IL 25; IL 30, 4-5 ; 11. 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 immorahty. 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 reahse these ideas, and be Virtutis 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 Eome : to be ostracised or pro- scribed was anything but a mere idea. The Greeks and Eomans 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 PART II. 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 tliat 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. ^ 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;ssa//,lV. 3, 31; IV. 12, 7. CHAr. VI. LOCKE 159 perties depend is an astliereal undulation, and defines it accordingly. Lastly, whatever M-e 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 witli 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 sei se 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 IGO rSYCIIOLOGICAL IDEALISM PART II. we call substance.' ^ 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 suj^posed and conceived substances : body, the suljstratum to those simple ideas we have from without ; and spirit, the substratum to those we have from within.'^ Finally, he regarded both these substances as unknown, and neither of their ideas as clear and distinct. Nevertheless, he tliought 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 liim, real if con- sistent ; those of substances real onlv if a^^-reeinir with things without us.^ 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, movino- &c Locke went to the opposite extreme of supposino- a sulistance to be a siil)stratura or kind of support on wliich the qualities rest, and tliis is the ordinary view, descended indeed from the compound, or 'concrete,' substance of Aristotle, composed of matter and form. « Essay, 11. 23, 1. « See II. 23, 1- 5. ' II. 30, 4-5. cirAP. VI. LOCKE 161 But here are two abstractions, the subject abstracted Irom 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, movino-, &c • in which the qualities are nothing except as characterising the subject, and the subject nothing except as charac- terised by tlie qualities ; from which subject or sub- stratum, qualities or attributes are opposite abstractions, becondly, 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 y,e 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, resistin- yo^' V 18G PSYCHOLOGICAL IDEALISM PART IT. CHAPTER VII. BERKELEY. The two plillosopliers 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 onlyA all the data but also all the objects of. understand- I 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, i<^nsist- ently denied facts Vj3erkeley assumed, with Descartes, that ideas are the data, and with Locke, that they are the oyects, 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 : — M| ' 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 forme'd 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 hght 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 furnislies 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.' ^ 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 quahties without a quahfied subject, and ideas of operations without a thinking subject. Eeason- ■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 * Princ. i. •• » 188 rsYCIIOLOGICAL IDEALISM PART II. the objects of Immau knowledge are ideas. Locke, not human nature — and not even tlie whole of Locke — was the oracle of Berkeley. Berkeley, however, being a less various but a more logical thinker than Locke, w^as 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* lias 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.' ^ Berkeley devoted the Litroduction of the ' Principles ' to a criticism of this modern conceptualism, and founded modern nominalism. » Essay, IV. 7, 9. CHAP. VII. 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. ^ 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.^ Simi- larly, Hume afterwards said, that all general ideas are nothing but particular ones annexed to general terms^^ 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.^ 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. Tlie 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 pliilosopliy. 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 * Princ. Introduction, x. 3 Treatise, i\ § 7. ^ Id. xvi. ' Cf. Mill, Logic, iv. 2, 1. 190 PSYCHOLOGICAL IDEALISM TART II. 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 tliat the invisible object is coloured only in the sense of reflect- ing aethereal 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 o])jects 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. Eeally, 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, tlien, starting from the Second and Third Books of Locke's Essay, &,11 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 he reduced abstrac- tion of ideas, and without reasoning about causes.) But it is im]>ossible for errors to remain perfectly logical. Though he had just said tliat all objects known to us are ideas, he proceeds, like Locke, dogmatically to assert tliat a thinking subject exists : — ' 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. Tliis perceiving active being is what I call 77iind, 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, wlierein they exist, or which is the same thing, whereby they are perceived ; for the existence of an idea con- f sists in being perceived.' ^ Berkeley was dogmatic, but right, in asserting the \ existence of himself; but he w^as 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.'^ 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 ; tlierefore, 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. * Princ. ii. ^ Id. xxvii. 102 PSYCHOLOGICAL IDEALISM TART II. ( 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.. J Berkeley ought to have returned to Descartes, and begun 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 tiiangle.^ We really perceive, by sensation, at least, the coloured, and by consciousness, at least, the willing. vBut Berkeley, like Locke, began all sense with abstract ideas of qualities and operations. ^ Tliough, 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 j or am conscious of my own being.' ^ 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- ^ Hylas and Philonous, Third Dialogue. criAP. 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.^ 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 f^ir 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 Newton ^ about the Deity, ^ Princ. ii. 2 Newton, Princijpia, Lib. III. Scholium Generale {sub fin.). O 194 PSYCHOLOGICAL IDEALISM TART II. ' non ut aniraa mimdi, sed ut universorum dominus.' God has no body ; for how could He have a body pro- portionate to His infinite intelHgence and vdll, 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 verv 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. CHAr. 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 explainhig 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 percipi^ nor is it possible that they should have any existence out of the minds of thinking beings which perceive them.' ^ 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 percipi ; and again, an unthink- * Princ. iii. 2 196 PSYCHOLOGICAL IDEALISM PART II. ing tiling, which is not an idea, will not be humanly known to exist. Berkeley was entitled to these liypo- 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. Tims to hypotlieses 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.^ 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.^ 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. * Princ. iv. 2 Id. V. CHAP. VII. BERKELEY 197 This strict though liypothetical 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 tliat neither philosopher suppHed data from whicli 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 percipi, 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. ^ 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 ^ Princ. VI. 198 rSYCIlOLOGICAL IDEALISM PART II. of tlie Avliole known world. We saw above that Berke- ley, while speaking even of nian,^ denied the existence of wluit was not an object of human knowledge. He now corrects this defect by the addition of the eternal spirit,^ 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 I not any other substance than spirit ; ' this is the next hypothetical consequence.^ It is an immediate corol- j 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.^ 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 quahties.^ 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- thinix but an idea.' ^ This memorable sentence marks * Princ. iii. ^ Id. vi. ^ Id. vii. 5 Id. « Id. viii. Id. CHAP. VII. 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 beinsT 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. ^ 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 : — ' 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 * Cf. Princ. xxxvii. r~ 200 rSYCIIOLOGICAL IDEALISM PART II. tlie 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 •. * 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 lie 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.' ^j:^ 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 II. ^ arrived at the 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, wdiat 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 a3ther 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 w^ith 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.' ^ 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,' he concludes, ' that extension, figure, and motion cannot be the cause of our sensations. To say, therefore, that these are the efiects of powers, resulting from the con- figuration, number, motion, and size of corpuscles, must certainly be false.' ^ 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.' ^ Berkeley, Hke Descarles 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 ^ Princ. XXV. Id. ^ Id. xxvi. 204 rSYCnOLOGICAL IDEALISM PART II. CHAP. VII. BERKELEY 205 and Locke, however, were not logical enougli 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.^ Thus, the solution, which was suggested by Descartes, as a possible alternative in his 'Principia Philosophise,' ^ 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 ogical 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, wdtli all its hypothetical consequences. As Ki * Princ. xxviii.-xxx. ^ Descartes, Princ. ii. 1. 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.^ 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.' ^ 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.'^ Finally, he presents us with his complete theory of real things, when second causes have been expunged :^ — ' The ideas imprinted on the senses by the author of nature are called real things ; and those excited in the * Berkeley, Princ. xxvi. ^ Id. xxx. ^ jj^ xxxii. ■* Id. xxxiii. 206 PSYCHOLOGICAL IDEALISM TART 11. CHAP. VII. H) BERKELEY 207 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. LTlie 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 mindji 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 natura} 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 * Princ. xxxiv. \ qualities without the mind.' ^ So sure was he that sensible ideas are the real things, that he even said that ' we are fed and clothed with these things which we perceive im- mediately by our senses ; ' that is, by sensible ideas.^ 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 doo-ma- tise about all reality but confine themselves to known reality, in other respects differing in nothincr 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.' ^ 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 j^ xxxviii. 3 Examination of Sir W. Hamilton's Philosophij, chap, xi., note, 8ub fin. 208 PSYCHOLOGICAL IDEALISxM PART II. realism of cosmotlietical idealists, such as Descartes and Locke and their 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.^ 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 reahst supposes that a secondary quahty is directly perceived as a mere sensation in the mind, a primary quality as a real quahty 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 * Frinc. xxxiii. scq.; Hylas and Philonous, Third Dialogue, sub Jin, 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 j)ercep- 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.' Tli£ natural theologian, dazzled 210 PSYCHOLOGICAL IDEALISM PART II. 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 Ibeen 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. VII. 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 quae agunt in virtute primi agentis at^unt.' 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 foro-et that, after all, the existence of nature must be P2 212 rSYClIOLOGlCAL IDEALISM PARI II. 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 inteUigent cause, most certainly the means used are physical causes. Ali 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 reconcihation of theology and science will be found superior to that of Bacon,^ 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 estabUshed 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 si singulis schematibusetmotibus naturalibus Providentias characteres essent impressi.' Berkeley for nature substituted God. By his hypo- theses and logical deductions he was compelled to say ^ De Augmentis Scientiarum, iii. 4, sicb fin. (Ed. Ellis and Spedding, Tol. 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, w^ithout 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 rehgion. 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, w^ho 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 knowdedge 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.^ 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 ^ Cf. Princ. xxv., xxxii. ■ . 214 rSYCEIOLOGICAL IDEALISM PART II. bring a narrow mental into harmony witli 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.^ 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 ideahsm. 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-hke 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 hke, must needs be labour in vain. Accordingly, we see the attempts of that kind are not at all satisfactory.' ^ 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 ' ^ : — '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 * JPrinc. ci. sej. ^ Id. cii. » Pp. 72-3. CHAP. VII. BERKELEY 215 assumption of this wonderful intangible either, all the phsenomena 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 such phienomenon 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- 210 PSYCHOLOGICAL IDEALLSM PART ir. 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 ov»^n 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, w^ithout 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, wliose secondary species are undula- tions of either, 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, w^hich 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. VIT. 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 law^s, 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*nust answer, not all known realities are sensa- 218 rSYCnOLOGICAL IDEALISM TART II. CHAP. TIT. BERKELEY 219 tions ; to Berkeley liiraself, not all realities are sensible ideas imprinted on our senses by the Author of our bein^T ; 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 orile 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 tlie possiljihty of more, sometimes of less, than sense per- ceives ; hence we multiply man and horse into centaur, or diminisli man into gliost. 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 the possibihty of men be- cominfT 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 : — 'The ideas of (johlhis and sprights 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.' ^ 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. Li 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. Li 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 gliost 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. ^ Essatj, ii. 33, 10. 310 PSYCHOLOGICAL IDEALLSM PART II. These two reproductive causes of ideas, by constantly promoting the same idea, challenge our attention not only to tiie 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 his attention is thereby called towards the evidences w^hich 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- liension of a relation of sensible objects. Thirdly, associa- /ion 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 TTTiot 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. IIUMR 311 which are not impressions nor ideas, yet are thoughts. Fifthly, the productive origin of ideas is simple sensed forming the first ideas of qualified substances, synthetic sense forminj^c the first ideas of relation, and reasoninjj: 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 a^^^ 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.^ He starts with what he calls the instinct by which men ' 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 percej)tion, 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 ' Inquiry, § 12. 312 PSYCHOLOGICAL IDEALISM TART II. tlio object with the operation of sense. Ideahsiu is the sc^ieiitific trutli tliat sensible objects are efTects on tlie senses, misinterpreted into the hypothesis that tliey are ' 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 possil)le), and could not arise eitlier 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 j no data to prove anything at all about objects, internal or external. Moreover, if the data both of sensitive and reflective perception were 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 sa^^s, nothing re- mains but ' a certain unknown, inexplicable something^ as the cause of our perceptions.' Such is the false though logical end of Hume's speculative philosoph}^ He proceeds illogically to correct himself of his Pyrrhonism by the old view of the Academy that ' 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 j^erceptions, we should never infer anything but perceptions, with an unknown, inex- 314 PSYCHOLOGICAL IDEALISM PART II. plicable sonietliing. But natural pliilosopliy lias 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 dedudio ad absurdum of idealistic hypotheses. It is what was sure to follow if Locke and Berkelev 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 su2)position 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 wroni^ alternative. Listead of e:oin£^ 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 tlierefore no idea of a substance.' ^1 This logical syllogism, of which however the major isy 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 ideahsm. He was. particularly attracted by Berkeley's philosophy ; for instance, by the theory of general ideas, and of primary and secondary quaUties. Berkeley's hypothesis,^ in the 'Principles,' of the inactivity of ideas, antici- pated Hume's scepticism about power in causation; while, in the ' Theory of Vision,' the hypothesis that * Treatise, iv. § 5. o o IG rSYCIIOLOGICAL IDEALISM TAUT 11, visible ideas su^L^i^est tangiljle ideas, without any iiifereuee bf an external object common to touch and vision, gave ^he 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'^ 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 * Inquiry, § 12, Part I., note. ciiAr. IX. HUME 317 simple from complex apprehension,^ and names from propositions,^ but did not co-ordinate reasoning with the two other apprehensions. St. Thomas Aquinas schema- tistically added reasoning as a third operation.^ The moderns, by co-ordinating the three operations, have tended to lose sight of the process of reasoning at the l)ack 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 thne, 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. Li 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. Li 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.' "^ But his animadversions on these purely con- ceptualistic definitions only end in his reducing all these ^ Ar. De An. iii. 6. ^ Aquinas in Periherin, i. "^ Id. Periherm, i. * Treatise, iii. § 7, note. 318 rSYCIIOLOGICAL IDEALISM PART II. acts to conceptions. Hence liis resolution of judgment or belief into a vivid conception or idea, from wliicli the substitution of association of ideas for inference of judgments immediately follows. The answer is tliat 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/ 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 CHAPTEE X. If KAXTS 'CRITIQUE AND NECESSARY TRUTHS. Kant's ' Critique of Pure Eeason ' ^ 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 tlian 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 sensor}^ 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 representa- * Critique of Pure Beason, ed. Hartenstein, p. 33 = Meiklejohn's translation (Bohn), p. 1. Ueberweg's summary of the Critique of Pure Reason is printed in an Appendix at the end of this essay. 320 rSYCIIOLOGICAL IDEALISM TART II. tiou— a confusion constantly fiivoured 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 seiisible, or, in Aristotelian language, cesthesis and cestJiema. 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.^ 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 plijenomena 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 ps^x^hical, the hypothesis that outer sense is concerned with mere qualities and inner sense witli 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 ideahsm had become scepticism. What men call sensible objects, 322 rSYCIIOLOGICAL IDEALISM PART II. and ]x4ieve 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 sensor}^ ' 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 tlie 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. ]3ut 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,' ^ not rJlow- ing that even coexistence is sensible. Accordmg to him, the matter of sense received from without is ^ Critique of Pure Reason, ed. Hartenstein, p. 175 = Meiklejohn's translation (Bohn), pp. 142-43. ciiAi\x KANT'S ' CRITIQUE ' AND NECESSARY TRUTHS 323 nothing but a manifold or aggregate of unrelated im- pressions, a mere play of representations,^ a rhapsody of perceptions.^ 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 tlie last step in the imitation of one idealist by another. Yet it is necessary to the argument of the ' Criticjue.' It is because the matter of sense is presupposed to be mere impression that our knowledge of objects is suj)posed to be due to a priori sources. In short, Kant attempted a criticism of pure -reason witliout a previous criticism of the matter of sense. After what I have said in this essav, 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 ' Inquiry ' : — ' The mind,' says Hume, ' has never any- thing 2)i'esent to it but the percej)tions, and cannot possibly reach any experience of their connection with objects.' ^ 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 haye 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 experiencej The difference, however, is by no means so great as it appears at first sight ; for Hume and Kant alike begin 1 Hart. 178 = Meik. 145. ^ jjart. 152 = Meik. 118. ^ Inquiry^ 12, Part I. T 2 i ^..^ .? / 324 rSYCIlOLOGICAL IDEALISM PAET 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 oz«f ^/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.' ^ 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 hidependent of experience.' ^ It is nearly related to what Descartes called innate. But the noveUy 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- » Hart. 33 - Meik. 1. 2 lb. 34 - Meik. 2. CHAP. X. KAxrs ' cnn iQUE ' and necessary truths 325 L^ 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 thouglit, pure notions, or categories, to convert repre- sentations into a perception and experience of objects ; and reason requires us to conceive a p)riori ideas beyond objects of sense, understanding, perception, experience, knowledge, but cannot enable us speculatively to know the unconditioned objects of those ideas. Tlie Kantian a priori theory differs from the Cartesian theory of innate ideas by the assertion of a priori forms m./ 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 tlie 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 ori^^^in of knowled«^e, 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 safe rule : when opposite parties quarrel with one ar\pther 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 II. and, secondly, that tlie 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 reasonino" itself. It is evident that causce coynoscendi 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 irriori 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 'CKITIQUE' 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 ptriori stands on the same footing as the a3thereal hypothesis. But there is a decisive difference, ^ther 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.^ Transcendentalism reall}^ 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 sims of a knowledire a priori' '^ 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, iii. 14. '^ Hart. 35 = Meik. 3. 328 PSYCHOLOGICAL IDEALIS>r PART II. 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.* 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.' '^ 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 ; 1 Hart. 36 = Meik. 4. '' lb. 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 hmits 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 depends 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 phaenomena, as Kant supposes. In that case, there would be no logical reasoning from experience of phsenomena to non-pha^nomenal 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 immortaUty 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. ooU PSYCIIOLOGirAT. IDEALISM PAKT II. Of the three indirect arguments, which constitute the proof of transcendentahsm, the first is furtlier 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/ 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 jmori 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 altered its character. Leibnitz had held an a jwiori 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 » 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 estabhsh 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 Ijodies are extended ; while a synthetical judgment is one which adds a predicate to our notion of the subject, e.g. all bodies are heavy.^ 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.^ ' Natural science {Physica) contains synthetical judgments a priori as principles in itself;'^ 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 ^ Hart. 40 = Meik. 7. '^ lb. 43-4 = Meik. 10. 3 lb. 44 = Meik. 11. PSYCHOLOGICAL IDEALISM PARI II, 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.^ Now, to resume his whole argument from necessary to synthetical a joy wz 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 w^e 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 (j^oT^crts), 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,'"^ and distinguished simple and complex being, remarking that things simple (rd do-w^cra), such as a ^ Hart. 45 = Meik. 12. ^ ^^i^ y. 3 seg. oo o CHAP. X. KANTS ' CRITIQUE ' AND NECESSARY TRUTHS 666 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 beino- white, we may make propositions either true or false.^ In the ' De 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.'^ In the ' Posterior Analytics ' he insists that the principles of demonstration must be necessary, that is, self-evident,^ that the axioms of being, though principles, are not the actual premises,^ and that the principles of demonstration are acquired by a gradual process of sense, memory, experience, induction, and are recognised by intellect {vov<;\ of which the obvious function is to apprehend their ne- cessity.^ 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 exemphfied in the explanation of the ' Met. e. 10. ^ lb. i. 11. - Dc An. iii. 6. ^ Post. An. i. 4-6. ^ lb. ii. 19. O O I oo4 rSVCIlOLOGICAL IDEALISM TART II. facts of lieat and liglit by defiiiiiig them as undula- tions of aetlier. 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 liacon 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- cunudating its properties, and then explaining them by icthereal motion. ]3ut at the bottom of these exajiirera- 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 conceptuahsm 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 Fourtli Book of his ' Essay,' recognised self-evidence under the name of ' intuitive knowled<»-e ' which perceives the agreement or disagreement of ideas by 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 o o rr CHAP. X. KANT S 'CRITIQUE' AND NECESSARY TRUTHS ooo demonstration,^ 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,' savs he, ' can once doubt but existence and non-existence destroy each other, and are perfectly incompatible and contrary.''^ In the '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.^ '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.' * But he confined the self- evident and demonstrative to mathematics.^ 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. ^ Secondly, he attempted to banish the self-evident entirely from science, and went to a pitch of scepticism of which even Hume hardlv dreamt, bv 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, ' Essay, iv. 2, 1. ^ lb. § 5. - Treatise, iii. § 1. ^ Inquiry, § 4. * Cf. ib. § 12, Tart III. " Mill, Logic, i. 6^ 4. Y 000 PSYCHOLOGICAL IDEALISM TART II. 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 meaninpr of the words.' ^ Moreover, he admitted ' the original inconceivability of a direct contradiction,'^ 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 conceptuahstic theorv 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 ^ Mill, Logic, i. 6, 4. ^ Examinaiion of Hamilton's Phil, chap vi. CHAP. X. KAKT'S 'CRITIQUE' AND NECESSARY TRUTHS 337 thing 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.^ 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.^ From Leibnitz lie adopted the conceptuahstic theory of the nature, and the a priori theory of the origin, of an analytical judgment. But he difl^ered 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 liost 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 » Leibnitz, Nouveaux Essais, i. 1 {O^cra, p. 204 sea. ed. Erdmann) 2 Hart. 148-50 ^ Meik. 115-7. 338 rSYCIIOLOGICAL IDEALISM PART II. Kaut, 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,^ 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, liaving to admit the use of these analytical judgments in geometry, maintains that they serve only 'for the chain of method, and not as principles.""^ 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 phaenomena, 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 » Hart. 157 = Meik. 124. 2 Hart. 44 = Meik. 11. ! I CHAP. 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 phaenomena, 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 phaenomena; 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 reahsm of the self-evident. The conceptuahstic and nominahstic theories of analytical judgments liave each its pecuHar 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 11 I 340 PSYCHOLOGICAL IDEALISM PART II. objects from ideas. The latter tlieon^ 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 verl^al, 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 thincr, 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 I 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.^ 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 origm 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, » Hart. 39-42, 148-50 = Meik. 7-9, 115-17. I ii 342 PSYCHOLOGICAL IDEALISM PART It. and a thousand other objects of our intellectual ideas ? ' ^ 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 thincfs. Not beincr 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.^ 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. 0;pera (ed. Erdmann), 196 A. ^ lb. 210 A, 212 A. CHAP. X. KANT'S 'CrJTIQUE' 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 tlie 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 prol)e 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 IT. long before he knows that 'tis impossible for the same thinj? to be, and not to be ? ' ^ Leibnitz replies that ' one founds oneself on these general maxims, as one founds oneself on the majors, which are suppressed in reasoning by enthymemes.'^ 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 formula. 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.^ 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 tlie 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 * Locke, Essay, iv. 7, 9. ^ Opera (ed. Erd.), 211 A. * Ar. Post. An, i. 11. CFAP . X. KANT'S ^CRITIQUE' AND NECESSARY TRUTHS 345 analytical judgment. The way out of this difficulty mav 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 doer 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 ficTure, I use the dictum de omni, in the second the die- turn de diverso, in the third the dictum de exemplo ; l)ut 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.^ 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 » Opera (ed. Erd.), 207 B, 211 B, 213 A. 316 PSYCilOLOGlCAL IDEALISM PART ]I. 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 tliinking 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 w^liole 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 NEOESSAKY TRUTHS 347 conceive general ideas ; and in the chapter on Berkeley, that abstraction is a kind of attention, wdiich 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 ; l3ut 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 " y ." 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 IT. 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 a])stractions of arithmetic have just been given. In the same way in geometry, general reasonin<]^ tells us that bodies are extended in three dimensions, but perfect abstraction is required to isolate the solidity of body and apprehend body quel solid as that which is long, broad, and deep, and notliing more. Similarlv, in abstract mechanics, it is not till we have regarded a body qua 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 prioi^i, 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 CHAT . 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 creneral reasonincr 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 thing 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 havincT by perfect abstraction isolated a thing qua 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- 850 PSYCHOLOGICAL IDEALISM PART ir. strations from it. Indeed, Aristotle Avas 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- hen'sion becomes a principle of demonstration. Fourthly, the analytical judgment is made spontaneously by the l)rinciple 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. Anegative analytical judgment is of the same kind,but one deg'ree more complicated. General reasonhig from sense infers that white objects are not black, that sweet objects are not bitter, that square objects are not round, and so forth. Terfect 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 appUcable, because the objects are so simple, and the abstraction simply apprehends the sen- sibly white as containing nothing black, and vice 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 hi the abstract as subject and predicate of a negative judgment, a sensible object qiid white is never black, a thing qua square is never round. Its principle is that cnAP . X. KANTS '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 unhmited. 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 things are possible, it is at least impossible that a thing should not be the same as itself, or be the same as something ^^9 PSYCHOLOGICAL IDEALISM TART II, 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 ifniversal applicability. They I CHAP. X. KANTS 'CRITIQUE' A^'I) NECESSARY TRUTHS 353 are not liable to the difficult}^ of synthesis, tliat an exception may be found to the combination of two kinds of objects; a difficulty which, Kant confesses, applies even to a priori syntliesis 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 w^hole it is only a sum of parts. Thirdly, self-evidence makes analytical judgments con- vertible or coextensive : so lonjj^ as a \\\m^ 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 theorv 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 anah^tical judgments are self-evident to one who has abstracted the objects, universal without A A Sol rSYCII )LOGICAL IDEALISM PART II. 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 lo'^ically 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 probabiUties 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 ; wdiile the other starts in the same manner, but by the perfect abstraction of a simple, non-synthetic object, such as a thing qud whole, a body qud solid, a body qud 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 N? I CHAP. X. KANT'S 'CRITIQUE* AND NECESSARY TRUTHS 35d 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 qua 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 instanticc 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.'^ 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 pli£enomena.' 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. "" Mill, Logic, iii. 21, 4. A A 2 35(1 rSYCIlOT.OGICAL IDEATJSM PVRT TT. 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- ledcres ' 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 beUef, 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 hmited 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 inconceivabihties of association, but incredible impossibiUties 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 ehminated 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 sequitar 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 p)Osteriori are 358 rSYCHOLOGlCAL IDEALISM PART ri. necessary principles of science. Tlie 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, haye a reality in things, and an a posteriori origin, and a position among Euclid's principles, 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 EucUd'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 cuAP. X. KANT'S 'CRITIQUE' AND NECESSARY TRUTHS 859 often happens, it has been over-abstracted, and will never be successfully defined 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.^ 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 sufiicient 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 w^hich 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.^ 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 qua one as undi- ' Ar. Post. An. i. 27. "" Eucl. VII. Def. 2. 3G0 PSYCHOLOGICAL IDEALISM PART II. vided in quantity, objects qua two as one -| one, &c., &c. This abstraction is necessary to the science of arithmetic. As Plato, though he did not understand abstraction, long ago pointed out,^ 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 j^riofH, but by a jiosteriori 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 4- 1 are 3, and so forth ; every one of which are analytical definitions. Hence, though 7 + 5 is not, 11+ I is, the analytical definition of 12. All things, qua 11 + 1 are 12, and qua 12 are 11 + 1. MiU, 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.' '^ 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, w^lien three » Plato, Bep. vii. 525 D-6 B. Mill, Logic, ii. 6, 2. CHAP. X. KANTS 'CRITIQUE' A^D KECESSARi TRUTHS oGl pebbles are in one parcel, if they give us the impression o, this is the impression 2 -i- 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 tw^o 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 + 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 PAKT II. whole is greater than its part, and in its definitions. Thus we can destroy both tlie 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.^ 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 tliey 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 p)osteriori 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 » Examination of Hamilton's Philosophy, chap. vi. note. .^-j CHAP. X. KANTS 'ClUTIQUE' AND NECESSARY TRUTHS 3C3 necessary principles in the sciences, therefore they are a prion ■ analytical judgments are merely a prion ana- lyses of conceptions, but principles of science are true beyond conceptions, therefore they are never analytical iudcrments: 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 prion 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, understandmg, reasoning, knowledge, science are aU confined to sensible representations hiformed by a prion elements. Hence, according to him, necessary principles of scieiice, beinnpli- 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 382 rSYCIlOLOGICAL IDEALISM PART II. sense of pliysical 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 ideahsm — 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 plia3nomena. 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- mentahs sciHcet et Eationahs, arctiore et sanctiore fcodere (quod adhuc factum non est) bene sperandum est.' 1 ' J^iov. Org. i. 95. APPENDIX. VEBEBWEG'S SUMMAIiY OF THE ' ClilTIQVE.'^ By tlie 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. Tlie ' Critique of the Pure Reason ' suhjects 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 ' Skepti- cism,' and his own philosophy, which makes all further philosophis- ing 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 a 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 twofold division of judgments (in particular, of categorical judgments). With reference to the relation of the predicate to tlie 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 the analytical judgment is an identical one) —and synthetic or amplifi- 1 Ueberweg's Hist, of Phil. {English. Trans.), vol. ii. pp. 154-58 (§122). 384 UEBERWEG'S SUMMARY OF THE 'CRITIQUE.' Ccative 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 j^ 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 which are recognised as strictly universal and apodictical, and w^hicli are consequently, according to Kant's assumption, judgments a priori, he finds judgments which must at the same time be admitted to be svnthetic. Amonc: these beloncj, first of all, most mathematical judgments. Some of the fundamental judgments of arithmetic (e.g. a=a) are, indeed, according to Kant, of an analytical nature ; but the rest of them, together with all geometrical judgments, are, 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 385 like manner, finally, are all met aphysi cal principles, at least in their tendency, synthetic judgments a j>riori, 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 a pyriori 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 he 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 on the forms through which the Ego, my original consciousness, or the ' transcendental unity of apperception,' shapes all tliat 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 phsenomena 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 p from subjective elements, is not ' intellectual intuitiojii' 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 phjenomena, but only these. All valid a j^riori knowledge 1 has respect only to pha3nomena, 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 UEBERWEQ'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 tlie possibility of geometrical and on the a prior'i 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 phrenomenal 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 phtenomenal 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 — wliich 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 wliich 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 nature. PRIKTED BY SPOTTISWOODB AND CO., NEW-STBBBT 6QITABB LONDON COLUMBIA UNIVERSITY LIBRARIES P This book is due on the date indicated below, or at the expiration of a definite period after the date of borrowing, as provided by the library rules or by special arrangement with the Librarian in charge. 1 DATE BORROWED DATE DUE DATE BORROWED DATE DUE MAR 3 1952 1 1 i I 1 1' 1 1 C28(25I)100M i 1 I , 1010653503 p i ■ '■r ■ ;i '1 \oc MAY 1 9 1932