Bu 594 1^59 Hicliaraspn -^rzirrtTT^T^l. plural - University of California At Los Angeles The Library Form L I 2)54- This book is DUE on the last date stamped below ^hk ^ tB3$ OCT 2 ^95(r MAN 1 6 1954 i£ fflpy'f '^ ^'^ FonnL-9-10iii-3,'27 SPIRITUAL PLURALISM AND RECENT PHILOSOPHY CAMBRIDGE UNIVERSITY PRESS C. F. CLAY, Manager LONDON : FETTER LANE, E.G. 4 NEW YORK : G. P. PUTNAM'S SONS BOMBAY 1 CALCUTTA ImACMILLAN AND CO., Ltd. MADRAS j TORONTO : J. M. DENT AND SONS, Ltd. TOKYO : MARUZEN-KABUSHIKI-KAISHA ALL RIGHTS RESERVED SPIRITUAL PLURALISM AND RECENT PHILOSOPHY BY C. A. RICHARDSON, M.A. (Cantab.) J333J3 *>* 3^ 3,3 , 3 3 3 > , CAMBRIDGE AT THE UNIVERSITY PRESS 1919 9 TO JAMES WARD FROM THE INSPIRATION OF WHOSE TEACHING THIS BOOK DERIVES ITS MAIN THEME N V PREFACE ^ '^^ ALTHOUGH my main purpose in this volume has Jr\ been to present and defend the theory that the texture of the Universe is through-and-through spiritual, it will be evident to the reader that the particular form of presentation has been considerably influenced by the teachings of a school of thought which appears at first sight consistently to oppose any kind of idealistic or spiritualistic interpretation of the world. I refer to that school which is represented in America by the neo- realists, and in this country by logical atomists of the \ \ '^ type of Mr Bertrand Russell. Differing widely from them on certain fundamental points, I cannot but agree with a portion of the results of their critique. Never- theless I am convinced that in the main (except on some questions of epistemology) these results are not only not incompatible with that metaphysic which ac- knowledges the reality of the subject of experience and takes spiritual pluralism as its starting-point, but are actually complementary to it. On the other hand, I am no less convinced that discussion will continue at cross-purposes, and conse- v> quently no satisfactory progress will be made, until the neo-realists or their English allies proclaim in a precise and unambiguous fashion what they are going to do about the subject of experience. I am aware that their present tendency is rather to give the impression that they think that the latter may be ignored or dispensed with altogether. But I am by no means satisfied that R. s. p. /!> "0 ^ viii Preface they really «'<9 ignore or dispense with it in their own argu- ments. There is a suggestion of camouflage about the whole affair. For not only do they state that there are such things as sense-particulars, images, etc., but they seem plainly to admit that there is also such a thing as the experiencing of these objects, the experiencing of an object being by no means identical with the object itself. Very well, but will they please tell us a little more plainly exactly what they believe about this "experiencing" as distinct from the object experienced, and whether they think that no such individual as a subject is involved in it, and if not, why not? For after all, the theory which takes the existence of the subject for granted gives a perfectly clear and concise account of what it means by the experiencing of an object, and it cannot be said that neo-realism has as yet provided a satisfactory equivalent for this account. It is true that Mr Russell, in the last of a series of lectures recently delivered in London \ states that there is an empirically given relation between two experiences which constitutes their being what is commonly called "experiences of the same person," and says that we might therefore regard the person simply as the particular series of experiences between which this relation holds, dispensing with him altogether as a metaphysical entity. But think of all that may perhaps be hidden beneath such an apparently innocent phrase as "empirically given!' And there is still the distinction between the series of objects and the series of experiences to be made clear. ^ These lectures have now been published in the Monist, and the particular reference here is to pp. 373 ff. of the July 19 19 number of that journal. Preface ix I believe that a definite statement from the neo-realists on this point, is one of the most pressing needs of current philosophy. As regards the contents of this book, the greater portion now appears for the first time. The first essay, however, and the substance of the fourth, were published in the numbers of The Philosophical Review for May 19 1 8 and Jan. 1919 respectively, while the third essay was published in Mi7td of ]a.n. 191 9. I should like to express my gratitude to the editors of those journals for permission to include in this volume the essays referred to. I have made slight alterations in the latter to avoid needless repetition, though in certain cases, where a point has appeared to me particularly import- ant, I have allowed the repetition to remain, regarding it as justifiable in the circumstances. To my former teacher, Dr James Ward, I owe a debt of gratitude the measure of which I cannot ade- quately express. From both his written and spoken word I have received invaluable help, while throughout my work his advice and encouragement have contin- ually sustained me. Among others too numerous to mention individually, I am greatly indebted to the works of Mr Bertrand Russell ; while for the materials on which are based the preliminary descriptions of the various types of abnor- mal phenomena discussed in the eighth essay, I have largely to thank Sir W. F. Barrett's two books Psy- chical Research and On the Threshold of the Unseen. My grateful thanks are also due to Prof. James Seth (who, indeed, first suggested to me that I should publish in a book the results I had so far reached) for b2 X Preface valuable comments and advice ; and to my wife for the very considerable labour she has expended in the preparation of the manuscript for the press. In conclusion, I should like, if I may, to thank the Syndics of the Cambridge University Press for their generosity in helping me on my first philosophic ven- ture by the publication of this volume. I am only too well aware of its many shortcomings. But my excuse must be in part that it is only meant to be preliminary and tentative. I hope later to put forward a fuller ex- position of my beliefs, in the development of which I shall be greatly aided by whatever reception may be in store for the present book. Hence I shall be extremely grateful for any criticism, however adverse, that the latter may call forth. C. A. R. 2 2 August 1 9 19. CONTENTS PAGE INTRODUCTION xiii I. SCIENTIFIC METHOD IN PHILOSOPHY AND THE FOUNDATIONS OF PLURALISM 1—64 I. Introductory. II. Outline of Scientific Method. III. Outline of Pluralism and the Genetic Method. IV. Points of Conflict between Pluralism and the "New Realism." V. The Existence of the Self VI. Continuity. VII. Causality. VIII. Other Cate- gories of Experience. IX. Summary and Conclusion. II. ON CERTAIN CRITICISMS OF PLURALISM 65—83 I. Introduction. II. Externality. III. Conscious- ness. IV. The Evolution of Law. V. The "Bare" Monad. VI. Summary and Conclusion. III. THE PHILOSOPHICAL PROBLEM RAISED BY THE WEBER-FECHNER LAW. . . 84—113 I. Introduction. II. Examination of Conclusion drawn from the Weber- Fechner Law. III. The Nature of Sense-Experience. IV. Summary and Conclusion. IV. THE NOTION OF A DETERMINISTIC SYSTEM 114— 154 I. Introduction. II. The Case for Determinism. III. Analysis of Determinism. IV. Determinism and the Mind. V. The Problem of Free-Will. VI. Summary and Conclusion. V. THE INTENSITY OF SENSE-DATA . . 155— 161 VI. IMMORTALITY 162—180 I. Introduction. II. The Logical Statement of the Problem. III. Concrete Significance of the Problem for the Individual. IV. The Unity of Experience. V. Summary and Conclusion. xii Contents PAGE VII. THE RELATION OF MIND AND BODY . 181—243 I. Introduction. II. True Meaning and Implica- tions of the Problem. III. The Pluralistic Inter- pretation of the Matter. IV. Application of Pluralism to the Solution of the Problem. V. Memory and Imagination and their Dependence on the Body. VI. Personality. VII. Summary and Conclusion. VIII. SUBCONSCIOUSNESS AND CERTAIN AB- NORMAL PHENOMENA 244—328 I. Introduction. II. Ultraliminal Impressions. III. Subliminal Images. IV. The (so-called) Sub- liminal Self. V. Abnormal Phenomena. VI. Con- clusion. CONCLUSION 329—332 INDEX 333—335 INTRODUCTION WHEN a man sets out to write a book on philo- sophy, he is faced with a difficulty at the very outset. Nor is this difficulty one to be lightly overcome, for it consists in a doubt as to what the book ought to be about if it is to merit the title of "philosophical" at all. Many, and often widely diverging, are the defini- tions that have been given of the philosopher's task. While on the one hand it is asserted that philosophy is the enquiry into the ultimate nature of reality^ (what- ever an ultimate nature may be as distinct from any other kind of nature), on the other it is maintained that the only possible knowledge is that set forth in the results of the so-called "positive" sciences, to the unifying and co-ordination of which the philosopher must accordingly be content to confine his attention. Yet another view is that, even with regard to the body of knowledge constituting positive science, philosophy should merely investigate the form thereof without reference to any particular content". Opposed to this vision of an arid desert of abstractions, we have the more human and idealistic tendencies to deal with life and experience in its concrete fullness and variety ; to enquire into the goodness or evil, the beauty or ugliness of our own familiar world ; to determine the meaning of Man's struggles, and the goal of his endeavour. ' Cf. Dr McTaggart's definition of philosophy. - This view is apparently held by Mr Russell. xiv Introduction Probably the persistency with which every one of the above views has ever and anon been urged, is ac- counted for by the fact that each contains a kernel of truth, for each envisages the high aim of philosophy from one particular point of view. The special task to which one school after another would urge all efforts to be bent, represents in every case a philosophic end secondary to the ultimate end of all such thought. General agreement as to the nature of the latter is, as we have seen, unlikely to be achieved ; but it is prob- able that most people would agree that the philosopher's search might be deemed to have ended satisfactorily if he should succeed in formulating an hypothesis which could truly be said to explain and interpret the world as we know it, and not merely to describe in the simplest possible way the relations and sequences of phenomena. Now the origin of knowledge is experience. Juggle it as he may, cloaking it under a hundred different disguises, it is with experience — and his own experience at that — that the philosopher starts, and must start. The results of his analysis of experience, experience considered with reference to its subjective and its ob- jective factors, its phenomenal aspect and its noumenal implication, and the point of view to which his results lead, will run, like a dominant theme in a symphony, throughout the structure which he subsequently erects. The stages in the philosopher's pursuit will then be threefold. The first consists in the consideration of experience, and closes with the arrival at an ontological hypothesis to which that consideration leads. A neces- sary condition which this hypothesis must satisfy, is Introduction xv that it shall interpret and explain as opposed to merely describing. The next stage is the development of the hypothesis, a consideration of its claims as against rival hypotheses, and an investigation of its general bearings. In the third and final stage, the hypothesis is tested by applying it to the solution of particular philosophic problems. Complete demonstration or verification can never be attained, but a continually increasing degree of probability may be progressively exhibited, as the applications become at the same time more numerous yet more co-ordinated. The following series of essays marks a consistent attempt to advance a few steps along a road such as that indicated in the foregoing. In them a particular view of experience — idealistic in its general features — is adopted, and an endeavour made to establish a certain ontological hypothesis — spiritual pluralism. Finality has not been aimed at. Indeed, it is expressly recognized that while this hypothesis is perhaps the only one from which we can make a satisfactory advance, and while it will take us a long way towards the solution of the difficulties which beset us, it yet needs to be supplemented (though by no means sup- planted), by a further postulate. The writer's aim has thus been largely tentative. He will be amply satisfied if he has succeeded only in expressing his convictions lucidly, in clearly exhibiting the results to which he has so far been led, and in indicating the path along which, as it appears to him, further progress must be sought. The essays, though each is more or less complete in itself, form the development of a single coherent line of thought. In the first essay, the essential features xvi Introduction of the view of experience to be adopted are explained, and the general bearing of this view on questions of space, time, and the more important categories, is indi- cated. In the course of this exposition, the hypothesis of spiritual pluralism is reached inductively, and some of its most general applications are considered. To bring out the points involved more clearly, and inci- dentally to give additional support to the position maintained, there is throughout a reference, by contrast and comparison, to current schools of thought, particu- larly to that associated with the names of Mr Bertrand Russell and the New Realists, whose position differs in some of its most fundamental features from that taken up by the writer. The second and third essays are attempts to strengthen the general thesis. The former deals with the most important direct criticisms of spiritual pluralism which havebeenadvancedin recent philosophic writings. A distinction is made between that portion of the criticism which is aimed at the fundamentals of plu- ralism on the one hand, and that which, on the other, points out the necessity under which the pluralist ulti- mately finds himself of supplementing his hypothesis by a further principle. It has been found necessary to analyze carefully the exact meaning of the criticisms considered, and to disentangle, in at least one respect, two factors which appeared as a result, but one of which bears directly upon pluralism, the other having rather a general bearing upon all philosophic hypotheses. In the third essay, the view of experience adopted throughout (which, as mentioned above, is idealistic in most of its essential features) is maintained as against Introduction xvii its chief contemporary rival, the view held by Neo- Realism. The avenue of approach to the discussion is somewhat unusual, namely through a well-known psychological law. It becomes apparent in the upshot, that the latter has extremely important philosophical bearings. The chief issue between Realism and Ideal- ism is found to depend on the consequences we deduce from it. Having led up to the issue in this way, the discussion is continued on other grounds. In the course of an investigation of the nature of experience, it is pointed out that the analysis leads to results which necessitate a view idealistic in its essentials as opposed to realistic, while in addition certain inherent contradic- tions developed in views of the latter type are exposed. The essay closes with a brief discussion of pluralism as the hypothesis to which the conclusions reached in regard to experience point. The position to be taken up having been explained in this way, and the reasons in support of it defended, it becomes necessary to test the results so far obtained by applying them to the solution of particular problems. For this purpose, the classical problems of Free-will, Im- mortality, and the Relation of Body and Mind, have been selected for discussion. The fourth essay deals with the first of these. The question is approached by a statement and an analysis of determinism, for the latter is commonly accepted as the contradictory of the liber- tarian position. It appears later, however, that each view is really significant only of one particular type of entity, and the types concerned in the two cases are quite different. An elucidation of this point suffices to prove that the subject of experience is not a type of entity to xviii Introduction whom determinism is applicable, so that the universe, whatever else it may be, is not a deterministic system. So far as the major part of the discussion is concerned, the conclusions are demonstrated by the investigation of the experience of any single individual, without reference to a particular ontological hypothesis. The latter becomes necessary when it is required to extend the results obtained to the case of the universe as a whole. The fifth essay is merely an addendum to the fourth, being the detailed discussion of a psychological point raised in the latter. Although a settlement of the question involved is unnecessary so far as the solution of the free-will problem is concerned, it is yet of con- siderable general importance as regards both psychology and certain philosophical issues. It has therefore seemed worth while to consider it in some detail. In the sixth essay, the most general aspects of the problem of immortality are examined. These are so often entangled with particular questions such as those relating to the nature of a future life (if any) beyond this body, which have nothing whatever to do with the general problem of immortality, that it is necessary carefully to distinguish the issues involved. The inves- tigation hinges largely upon the position adopted with regard to the nature of time. The latter, already briefly discussed in the first essay, is now examined more fully. Just as is the case with the problem of free-will, the question as to the immortality of the individual is found to admit of an answer, at any rate as regards its wider aspects, capable of being deduced from the con- sideration of the experience of the individual alone, without calling in the help of a metaphysical hypothesis Introduction xix of the structure of the universe as a whole. Recourse to such an hypothesis only becomes necessary if we wish to estimate the consequences to other individuals of the facts arrived at in connection with any particular individual. It appears finally that the results in regard to the latter are summed up implicitly in the position adopted throughout with reference to one aspect of experience, its unity. The essay accordingly concludes with a development of this aspect. The relation between mind and body is dealt with in the next essay. To no other problem does the atmo- sphere of the out-of-date Cartesian dualism cling so closely and persistently. It is apparent in the treatment of the issue by nearly every school of thought. Conse- quently before any attempt can be made at a solution, it is found necessary to clear the problem of all its ancient dualistic cobwebs, and to re-view it from the standpoint of individual experience. In the light of this, such terms as "mind," "matter," "body," are re-interpreted, and the problem is then capable of being stated in more exact terms. To attempt its solution, however, it is found requisite to advance from the consideration of a single individual experience to an investigation of the relations subsisting between individual subjects. 1 1 thus becomes necessary to make use of the hypothesis of spiritual pluralism, and by its application a solution of the problem is advanced, so far as it appears to be soluble ; that is, the matter is pushed until we are left with indefinables whose nature is actually realized, though it defies descriptive statement. The conclusions reached are employed in the further discussion of a question of great theoretical and practical interest, XX Introduction namely the nature of memory and imagination, and their dependence on the body. This involves a state- ment of the noumenal implications of experience, in the light of our hypothesis. Finally, we are led on to the question of personality, and hence by a natural tran- sition to a brief review of the possibilities of a future life beyond this body, a topic which was not touched upon in the previous essay. At various points in the thesis developed in the essays so far briefly outlined, it necessarily happens that the chief types of ordinary phenomena are analyzed and interpreted. But there remains a large class of phenomena which calls for special consideration, not only because definite knowledge about it is continually increasing and pressing for co-ordination, but also because it has up till now been found impossible to bring such phenomena into satisfactory connection with the principles of existing sciences as at present stated. The phenomena are those commonly described by that term of not very happy associations — "spiritualistic." In spite of the sneers of certain bigoted scientists and others of narrow outlook, the labours of psychical researchers have established abundant and indisputable evidence of the main facts, and their results show that we have here a matter of high importance calling for very careful consideration. In particular, it is necessary to bring this new body of knowledge into relation with that already existing. For these reasons, the interpreta- tion of phenomena of the kind mentioned, has been chosen as the task to be attempted in the final essay. We are here brought most closely face to face with the necessity of supplementing spiritual pluralism by some Introduction xxi principle of unity. Such a principle is therefore postu- lated, although the determination of its exact nature is unnecessary for the matter in hand. With our hypo- thesis thus supplemented, it is found possible to co- ordinate the various types of abnormal "spiritualistic" phenomena, and to offer an explanation of them which may perhaps be claimed to be as satisfactory a one as can be obtained with the data at present available. It seems reasonably certain that these data, so far from meriting nothing but ridicule on account of the frauds frequently practised in connection with them, should give, if properly handled, a deeper insight into the real nature of the universe than that provided by any other group of facts. With the position thus established, strengthened, and tested, the writer has halted for the moment. But it is evident in what direction the next advance must be attempted. When pluralism has done its utmost, we are left to search for the nature of that concrete uni- versal principle or entity, whereby subject is linked to subject, and the many made one. I SCIENTIFIC METHOD IN PHILOSOPHY AND THE FOUNDATIONS OF PLURALISM I. Introductory The opening years of the twentieth century have witnessed, among other things, the gradual emergence of a new school of philosophic thought, and a new philosophic method. To the former, which originated in America, the name of ' New Realist ' has been given. The exponents^ of the latter designate it the ' Scientific Method.' New Realism is largely imbued with the spirit of the scientific method, and, for the purposes of criticism, the two may be treated together. The new method claims" to make an entirely fresh start in philo- sophy, to discover what problems are capable of solution, and to open the road to that solution to those who are willing and able to make the requisite effort. The reason for this new departure in philosophy is not far to seek. The last half of the nineteenth century gave birth to changes in the world of science of a revo- lutionary nature. To take two examples alone, Darwin's ^ Whitehead and Russell. ^ B. Russell, Our Knowledge of the External Worlds Lect. I and conclusion of Lect. VIII. Mr Russell designates his particular type of reahsm, " logical atomism." R. s. P. I 2 Scientific Method in Philosophy doctrine of Natural Selection and Maxwell's Electro- magnetic Theory brought about an entire change of outlook in their respective domains of biology and physics, with a consequent overhauling of all the tra- ditional concepts and principles. It is a commonplace that science and philosophy invariably act and react on one another. The scientific revolution was followed by an upheaval in philosophy. Everything went into the melting pot. There emerged a confused mass of opinion following upon the attempt to fit new facts to old systems, which finally settled down into two main tendencies of thought, the one clinging as far as possible to a modi- fied Absolutism based on Kant and Hegel, the other launching out more boldly in an attempt to apply Evolu- tion as a comprehensive metaphysical formula. The latter, though now generally discredited even by its posterity, was the forerunner of the modern evolution- ary and pluralistic schools. While much of the old remained in the new, there was, of necessity, a vagueness both in ideas and in method. Philosophy was cumbered with a mass of useless metaphysical cobwebs. Inevitably there arose an increasing demand for a general clearing up with a view to a fresh start. In this demand the new^ scientific method finds its motive force, and history is repeating the story of Descartes over again. Schools of philosophy may be classified according as they differ in method or in system. Whichever mode of classification is adopted, the various schools fall, broadly speaking, under two main headings. In method 1 ' I 'New,' that is, in its particular mode of application of scientific principles to philosophy. Introductory 3 they are Empirical or Rationalist ; in system, Pluralist or Singularist, Empirical method and pluralistic belief tend to go together, for if we appeal for the most part to the crude data of sense, we are confronted by a mani- fest plurality. On the other hand, the craving for unity has constantly caused men to mistrust the world of sense with its eternal diversity and flux, and led them to seek the characteristics of the totality of existence by pure thought alone. Hence the method which leads to the singularist or absolutist view of reality is essentially rationalistic. We can best trace the path of progress if we observe the development and interaction of the two classical methods of attacking the problem. Most great advances in philosophy have consisted in a partial synthesis of Rationalism and Empiricism. Kant's work is a supreme example of such a synthesis. The type of advance is analogous to that of the Hegelian dialectic. In emerging into antithesis after each successive synthesis, Rational- ism ever tends to include more empirical data in the material with which it works, while Empiricism is in- clined to stray ever further from the surface of things by building up its data into abstract intellectual construc- tions, and by framing more or less abstract hypotheses to account for these data. Pluralism, with its genetic method, is the modern outcome of Empiricism. The position of Rationalism is not, at first sight, so clear. One's thoughts turn natur- ally to the idealism of the Absolutist school; but although the beliefs of this school are upheld by some of the foremost philosophers of the day, they represent an influence which is rapidly on the wane. The true pro- 4 Scientific Method in Philosophy gressive product of Rationalism, despite the fact that its data are mainly empirical, is the New Realism, for its scientific method purports to deal with the form and structure of existence as opposed to its concrete con- tent. The final synthesis of the two points of view consists not in an amalgamation, but in a recognition of the fact that each is necessary to the complete fulfilment of philo- sophic purpose, and in a determination of the particular function, domain, and limitations, of each associated method. II. Outline OF Scientific Method The scientific method^ attacks the problem of the Universe piecemeal. A problem is selected, isolated as far as possible, and an attempt made to clarify our con- ceptions relating to it, and to determine, by continued analysis, the true source of the perplexities underlying the question. The final step, and the most difficult, is to formulate an hypothesis which will resolve these perplexities. The difficulty of this last process lies in the fact that the necessary hypothesis is inevitably of a peculiarly abstract nature, for at each successive stage of the analysis the matter under consideration becomes more abstract. In any particular investigation, the initial data con- sist in the generally accepted body of knowledge on the subject. This knowledge will almost invariably be vague and confused, and the task of analysis is to resolve ' B. Russell, op. cit., beginning of Lect. VIII. Out line of Scientific Method 5 it into a number of definite propositions. The latter, when the ultimate premises have been reached, are arranged in logical sequence. The premises should be stated with the minimum of redundance. They must then be scrutinized in turn with a view to ascertaining the particular degree of doubt or certainty attaching to each. This provides us with a criterion of the doubt- fulness or certainty pertaining to each proposition of the sequence, and to the initial data in particular. In general the scientific method makes use of three types of data^: sense-data, the testimony of others, and certain primitive logical truths. In making use of testi- mony, the existence of other people must be tacitly assumed. It is impossible entirely to justify this assump- tion. On the other hand, its falsity cannot be established, and it is indispensable in opening up a relatively im- mense fieldof knowledge, whereas solipsism is practically barren. One of the most important applications of the scientific method is to the analysis of the meaning of the concepts of physical science", and the investigation of their validity as representative of the world of fact. Evidently, the material from which the start must in this case be made, is the data of sense. Since all scientific observation consists in perceiving sequences of sense- data, and since the verification of a physical prediction lies in an appeal to the occurrence of some sense-datum, it follows that if the entities of the physicist are to be valid conceptions, they must be capable of representa- tion as logical functions of sense-data. Even if they be 1 B. Russell, op. cit, Lect. III. 2 Ibid., Lect. IV. 6 Scientific Method in Philosophy so represented, it does not follow necessarily that they exist concretely, nor does it matter. The importance of such a critique of physical science is great, for the physicist is often apt to consider his entities as the true realities of the universe, regarding them as inferences whereas they are merely constructions. The data of sense are the indubitable concrete facts. The exponents of the scientific method claim that it is capable of ascertaining all that is soluble in the problems of philosophy, and of effecting the solution. They regard it as giving the method of research in philosophy just as mathematics gives the method of research in physics. It alone is capable of yielding whatever objective scientific knowledge it is possible to attaint This is certainly true enough so long as we regard merely the objective side of experience as such. For the task of science is twofold. It has first to record and correlate the particular concrete facts of experience. This is the function of the specialized sciences. But the forms of the particular facts and the general aspects of experience have also to be investigated. This in- vestigation falls to the lot of the philosopher. The things he deals with in performing it are therefore essentially abstract. The function of the scientific method, then, though in part constructive, is mainly critical. Its field is the objective side of experience, and its scope is comprised in the determination of the validity of the concepts we apply to this objectivity, and in the solution of certain problems by the construction of other valid concepts and hypotheses. This construction it performs by analyzing ^ Op. cit, Lect. VIII, conclusion. Oittline of Scientific Method 7 the general forms of experience. AH questions such as, for example, ethical problems, are therefore regarded as outside the scope of philosophy, for they are con- sidered to deal with the particular characteristics of the particular things composing the world\ Although the results obtained by the scientific method may be mathematically accurate, and therefore a com- plete solution of the type of problem with which it deals, it will be seen later that the data from which it starts introduce an element of inadequacy. Moreover, the question arises as to whether the problems considered are the only ones with which philosophy may legiti- mately deal. It has commonly been considered that one of the supreme tasks of philosophy is to provide an explanation of the facts of experience. The hypotheses yielded by the scientific method are evidently purely descriptive in type. Can any hypothesis, however, be considered explanatory, or are all hypotheses descriptive, differing merely in the entities in terms of which the description is expressed.'* It will shortly be seen that one type of hypotheses may be considered as truly ex- planatory in an entirely unique sense. In this connection it is of the utmost importance to remember that there is both an objective and a subjective side to every question. Just as much light will be thrown on a problem by an investigation of the manner in which we have arrived at our relevant concepts on the subject at our present level of experience, as by an investigation of what the precise meaning of the concepts must be, if they are to ^ All New Realists do not take up this attitude with regard to ethical and analogous problems — e.g.^ R. B. Perry in Present Philo- sophical Tendencies, Ch. xiv. 8 Scientific Method in Philosophy be validly representative of objective experience. The concepts we have formed may involve unwarrantable assumptions if applied uncritically to objective experi- ence as such, but if their true meaning for us, and the manner in which that meaning has been acquired, are analyzed, not only shall we realize our own nature more clearly, but there is a possibility of the suggestion of fruitful hypotheses as to the nature of existence gener- ally. We are thus led naturally to the consideration of pluralism and the genetic method. III. Outline of Pluralism and the Genetic Method Pluralism is based on the existence of the self All philosophers do not believe in the existence even of one self. Reasons will be stated hereafter, however, which seem to render doubt on the subject not only logically impossible, but inherently meaningless. The word ' self ' will be used as synonymous with ' subject of experience.' All confusion with the various meanings of the empirical self, which relate essentially to the self as conceived, will thus be avoided. For the class of selves or subjects considered in relation to their experience, the term ' mind ' will be used. This meaning appears to approximate most closely to the general usage of that somewhat vague term. Starting from matter, i.e., from matter as generally conceived by physics and by the main body of common sense, it is impossible to bridge the apparent gulfs be- tween the inorganic and the organic, and between Life Outline of Phiralism and Genetic Method 9 and Mind. Herbert Spencer's work bears eternal wit- ness to this fact. The only alternative, therefore, is to start from Mind and endeavour to work back. This is the endeavour of pluralism. The pluralistic hypothesis is briefly as follows : " Reality comprises selves [i.e., active subjects of experience) alone, differing simply in degree or in kind of mental development, though the diversity is indefinitely various." Experience, then, con- sists in action and reaction between self and other selves, described by Professor James Ward in the expressive phrase ''muhmni commercitim!' The meaning of 'acti- vity ' is considered to be fundamentally realized by everybody, but to this point we shall return. The comparative hopes of a solution of the problems of philosophy held out respectively by Pluralism and by any form of Materialism, are sufficiently indicated by comparing the start made from the existence of selves as a basis, with that based on the existence of ultimate material particles — atoms, corpuscles, electrons, or what- ever they may be considered to be. We know that some selves exist'. Strictly, each of us knows that one self exists; but, as we have seen, if we are to philosophize to any extent worthy of the name, we must take a further step and assume the existence of other selves, nor can this assumption in any way be demonstrated to be false. It is therefore justifiable to make it. On the other hand, even should the material particles of physics actually exist (and this seems very doubtful), we could not know of their existence. The scientific method demonstrates this sufficiently. Moreover, if, as that method shows to be extremely probable, the entities of physics are simply ^ See Section V below. lo Scientific Method in Philosophy constructions of sense-data, we cannot conceive a self in terms of these entities; for evidently a self cannot simply be a logical function of sense-data. To sense- data we apply the term ' phenomenal/ i.e., ' presented to a subject,' thereby implying that we realize the funda- mental distinction in kind between the subject and the sense-data or phenomena which it perceives. It is impossible, therefore, to imagine that we ourselves can be analyzed into sense-data ; in fact, the supposition involves a contradiction in terms, for sense-data are * given ' or ' presented ' by the very meaning of the term, and it is the self to whom they are presented. Admitting, then, the existence of at least some selves, is it possible to explain the facts of experience entirely in terms of selves? As a matter of fact, the explanation has already been partially accomplished, and modern pluralists are engaged in applying it to the remaining difficulties. While at the present stage there seems to be no reason for supposing their attempts will be un- successful, many pluralists are of opinion that the bare hypothesis as stated above is incomplete. It seems probable that, for the complete solution of the prob- lem, pluralism must be supplemented by some form of Theism^ However that may be, the pluralistic hypo- thesis is admissible until disproved by fact, and there- fore it is justifiable to continue to apply it as far as possible. It is no part of our present purpose to analyze in detail the application of pluralism to the solution of philosophic problems, but the type of method adopted may be briefly illustrated by a consideration of its * E.g., J. Ward, Pluralism and Theism. Out Hue of Phiralism and Genetic Method 1 1 application to the case of what is commonly called ' inorganic matter.' The chief feature of that class of sense-data from which we construct the concept of in- organic matter, is the uniformity of the sequences manifested therein. There seems to be no expression of individuality observable. The opposite is true of selves. A self is essentially individual, for it is unique. In fact it is only to a self and its particular experience that we can correctly apply the term ' individual.' Animals manifest individuality and we have every reason of analogy to regard them as selves or subjects of experience. The animal merges insensibly into the vegetable world, and there is little difficulty in applying the pluralistic hypothesis to the latter. Now one funda- mental characteristic of mind is its plastic retentiveness, which is manifested in the formation of habits. The lower we go in the scale of life, the more habit do we find, and the less spontaneity. The latter, however, and consequently individuality, never entirely disappear. Remembering how narrow is the line dividing the organic from the inorganic, we are led to regard the latter as constituted by individual agents of extremely inferior mentality, whose behaviour is therefore suffi- ciently habitual to 2iAm\\.,forthe most part, of description in general terms. The reason for the apparent absence of individuality is that we are here probably dealing with individuals in bulk, so that our results are statistical. These results will be even more uniform than the majority of statistics, on account of the nature of the individuals concerned; but there is no reason to suppose that, if we could observe the behaviour of one of these individuals in isolation, we should be unable to observe 1 2 Scientific Method in Philosophy any traces of uniqueness. The above is but a broad outline of the pluraHst argument as appHed to the in- organic world, and if the question is examined in greater detail, many other reasons can be adduced which show that there is nothing incompatible with experience in this view of apparently lifeless matter. The first stage in the growth of a pluralistic philosophy is analytic. It consists in the analysis of experience, perceptual and conceptual, and of those particular concepts which we apply to experience under the name of categories. The investigation leads in all cases to results which suggest a pluralistic hypothesis, although they do not lead to it as a strict logical necessity. The second stage in the process consists in the application of the hypothesis to the solution of the particular problems of existence. In the first stage the investigation takes the form, for the most part, of an analysis of \\\^ growth of individual experience, and of the transition by intersubjective intercourse to universal conceptual experience. Hence the method employed is genetic. In this way we determine the process by which we have arrived at such knowledge as that of space and time, for example, and at such conceptions as Causality, Quality, and Relation. Thus abundant light cannot fail to be shed on the time-honoured problems associated with these names. Pluralism is an hypothesis, and it therefore stands under the universal limitations inherent in the nature of hypothesis in general. For a just appreciation of values, then, it is necessary that these limitations should be precisely stated and clearly borne in mind. We usually look upon hypotheses as put forward to Outline of Plitralisni and Genetic Method 1 3 * explain ' certain groups of facts. Let the case of physical science serve as an illustration. As a rule, men of science are content to dismiss the data of sense as merely ' subjective.' They look upon them as due to the action of certain hypothetical entities on our senses. The function of such an hypothesis, however, is not really explanatory, but simply an attempt to describe the facts of existence in the simplest possible terms. The immediate facts of existence are confused, complex, and loosely ordered. Any attempt to deal with them as they stand, for the purpose of calculated interference in the course of events, will be foredoomed to hopeless failure. Consequently, physics introduces such conceptions as those of a material particle and a luminiferous ether, in order to unify and co-ordinate the phenomena, so as to render them amenable to mathematical treatment. The majority of hypotheses are merely descriptive in this way. They are attempts to describe the facts of existence in simpler terms than the immediately given data. It might therefore be urged that pluralism is also a merely descriptive hypothesis, the 'explanation' being simply taken back one step, and expressed in terms of different things. Yet it is just in this difference of terms that the root of the essential disparity between pluralism and other hypotheses is to be found. It implies a difference of type. For pluralism is expressed in terms of active selves. We all realize what it is to be active — it is just living and doing. We all realize what a self is. This realization is far more than knowledge in the ordinary sense. It is something of what the older thinkers were trying to express when they said that 14 Scientific Method in Philosophy for perfect knowledge, knower and known must be one. It is a unique, supremely intimate fact, and therefore stands in a class of its own. It cannot be subsumed under one of the three types of knowledge proper — - knowledge by acquaintance, knowledge by description, and knowledge of logical truths^ It is this last fact which so often causes the realization of the nature and existence of self to be passed by, with the inevitable consequence that doubt is expressed of the existence of self at all. With a clearer view of the facts, such doubt is seen to be inherently meaningless. Moreover, it follows from the above that pluralism, being expressed in terms of active selves, is truly explanatory y(7r such active selves, i.e., for us. It thus differs in type from all hypotheses which are not expressed in such terms. Although pluralism differs in type from other hypotheses, it isyet bound by certain limitations common to all hypotheses. An hypothesis passes from neces- sarily partial observations of a system to a description of the system as a whole, and is therefore inevitably fallible unless the system be assumed capable of com- plete description in general terms. Any system, however, comprising subjects of experience, is quite incapable of being so described, for the subjects, and the experience of each, are essentially individual and unique. It follows from the above that, as we have at our disposal but limited observations of the world, it would be possible to find an infinite number of hypotheses descriptive of the world, which would sufficiently fit the narrow range of the observed facts. We could not ^ Evidently the subject or knower cannot be an object of knowledge. Outline of Pluralis^n and Genetic Method 15 form a unique and infallible hypothesis^ unless we knew all the facts, past, present and future, and then it would no longer be an hypothesis, but a mere recital of those facts. The fallibility of hypothesis is sufficiently illustrated by one fact alone, namely, that we have no reason at all to assume that laws which have held in the past will continue to hold in the future, unless we also assume some principle, such as that of induction, which depends on a priori principles of probability. Hence, though we may know an hypothesis to be false if it is contra- dicted by any fact, we can never certainly know it to be true. All that can be said is that it is more or less prob- able, the degree of probability depending on its applic- ability to the facts observed up to that time. Thus any final beliefs as to the constitution of the universe cannot depend on knowledge alone, but must be based on faith. In selecting an hypothesis, then, we have a very great range of choice, for no hypothesis is ruled out of court till it fail to account for some fact, or rather, till it be definitely disproved by some fact. This being so, we naturally turn, in the first instance, to hypotheses which are truly explanatory. For our purposes an explanatory hypothesis may be formally defined as * an account of a system which can be formulated symbolic- ally in terms of active subjects of experience.' The conceptual formula as such, which sets forth the hypo- thesis, is, of course, descriptive, but its concrete meaning is explanatory in a sense in which that of a formula in terms of objective things is not. ' I.e., we could not be sure that it was infallible nor that it was the only hypothesis which would fit the facts. 1 6 Scientific Method in Philosophy It should be noted with reference to the type of result likely to be obtained from the scientific as opposed to the genetic method, that logical constructions of sense-data can never give a self Hence, as selves certainly exist, no hypothesis in such terms can explain the universe nor even completely describe it. Pluralism, on the other hand, is not only explanatory, but it also complies with the condition demanded by Occam's razor. Far from multiplying entities, it is expressed in terms of entities certain examples of which we know to exist, and which any hypothesis must therefore take into account. IV. Points of Conflict between Pluralism and THE ' New Realism ' The supporters of the new scientific method hold that pluralism cannot be true because the conceptions on which it is based conflict with their results, and are therefore invalid. An analysis of the problem seems to show that the supposed conflict springs in the first instance from two main points of difference. These are the existence of the self, and the true meaning and validity of the categories of experience, particularly those of continuity and causality. The scientific method lays stress on the objective side of experience. It investigates the object of ex- perience, not in relation to the subject, but considered per se, and therefore in abstraction from the subject. It considers what meaning certain concepts must have if they are to be valid when applied to the object of Points of Conflict 1 7 experience thus isolated from the subject. Hence it fails to take account of the fact that the orrowth of experience consists in action and reaction between subject and object, manifested in an ever-increasing complexity and differentiation of the object, and that the latter is therefore determined in part by the activity of the subject. This activity is a fundamental realiza- tion \ but when considered from the conceptual stand- point of Empirical Psychology, it appears to consist essentially in attention. If, then, we are fully to realize the concrete meaning of the concepts we apply to experience, we must examine them in the light of this mutual interaction of subject and object. The consideration of the subject implied in ex- perience brings in its train certain ethical and teleological concepts which are meaningless except in application to such a subject. The failure to take the existence of the subject fully into account in the analysis of experience, thus leads to the inevitable result that certain most important characteristics of existence are entirely overlooked or regarded as invalid conceptions. It is owing to their preoccupation with the objective side of experience that the New Realists look upon the notion of teleology, for example, with such doubt and suspicion. Bertrand Russell" regards it as possible for a system to be both mechanical and teleological, according to the point of view. Such a supposition evi- dently entirely invalidates the generally accepted notion of teleology, and we shall therefore examine it hereafter. ^ See p. 32 below, note on 'activity.' - "On the Notion of Cause/' Scientia, Vol. XIII (1913), N. XXIX, 3. P- 333- R. S. P. 2 1 8 Scientific Method in Philosophy When concepts applied to experience are analyzed genetically, the meaning of them as thus determined is invariably found to contain more, and to strike deeper, than that determined by the scientific method. The former seems to throw considerably more light on the true nature of existence than the latter. This is illustrated particularly clearly in the case of causality. Experience is a unity, comprising a duality of subject and object, and we cannot fail to get more and more out of touch with its true inwardness, if we lay stress on one side of it to the exclusion of the other ; for all separation of subject from object, though neces- sary to a certain extent for purposes of analysis, is to that extent artificial. The problem of continuity brings out most clearly, perhaps, the difficulties raised by this artificial separation. V. The Existence of the Self In the course of his analysis of our knowledge of the external world, Bertrand Russell makes the assertion that the (bare) self, if it exists at all, is an inference\ This sentence defines very precisely the general attitude of the new realists. It is somewhat as follows : In any case, the reasons we may have for stating that the self exists can only be arrived at by inference, but even then, it is doubtful whether the inference can be made. As already pointed out, it appears that any such doubt of the existence of self is really meaningless. In the first place, from what may be called the concrete ^ Our Knowledge of the External World, Lect. Ill, p. 74. The Existence of the Self 19 point of view, we certainly cannot knoiv the self from the very nature of the case ; but we have instead the central and unique fact of the ' realization ' of our own existence. Evidently no general term can adequately express the full nature of a fact so essentially particular ; but this is no reason for ignoring the fact — perhaps rather the reverse. As will shortly be seen, we have in addition abundant data from which we can infer the existence of self, but the concrete realization of its existence is of infinitely greater importance. From the more abstract point of view, psychology traces the gradual growth of the concept of self, from the primitive idea of the body-self, through ever more refined and spiritual stages. Eventually we seem to be coming in sight of the bare active subject of experi- ence, as distinguished from the empirical self in all its phases. By proceeding in this way, however, we can never quite reach the subject of experience (though we may come very near it), for we are here dealing with self as conceived, i.e., as an object of knowledge ; whereas the concrete self is the knower. Knowino- is a relation between two entities, so that evidently the subject cannot know itself. It simply realizes its own existence, though the formulation in conceptual symbols of the fact of this realization, is itself a piece of know- ledge. As Kant pointed out, it follows from the foregoing that the only course open to epistemology is lo postulate the existence of the pure Ego, or subject of experience, as a regulative idea. Although we cannot, by continually modifying, and, as it were, centralizing the concept of self, arrive at the concrete subject of experience, we may yet infer 20 Scientific Method in Philosophy its existence. The inference is necessarily immediate, for all such terms as 'knowledge,' 'experience,' 'per- ception,' etc., imply the existence of a subject in their very meaning. Without it they have no significance whatever. In fact, all psychological discussion inevit- ably assumes the existence of an individual subject. We cannot speak simply of the existence of thoughts and feelings. There is always the implication of ' one who feels and thinks.' Knowledge and consciousness only possess meaning at all in so far as they are referred to something knowing and conscious of something else\ Experience implies presentation of an object to a subject, thus comprising a duality in a unity. The existence of the subject in this duality is just as much a fact as the existence of the object. The Cogito ergo sum of Descartes was one of the most conclusive inferences ever stated. Some philosophers, following Huxley, have regarded the self as being merely the series of mental phenomena constituting the individual mind. This supposition implies the existence of the very entity which it is attempting to dispose of. For, in the first place, what is meant by the ' individual mind ' } Why should the series be individual at all ? What gives it its essential characteristic of unity ? The fact of presentation to an individual subject is the only possible reason. Moreover, the very word ' phenomenon ' implies appearance or presentation to something — to what we call the subject. It is useless to state, as some have done, that even if this be so, the subject may still be merely a logical construction, for this is to lose sight of the fact that the ^ See also J. Ward, Art. " Psychology," Ency. Brit. The Exis fence of the Self 21 agent which constructs can be no other than that subject which is supposed to be a logical abstraction. Finally, it should be noted that the exponents of the new scientific method continually use the word * sense-data.' By so doing, they not only assert the existence of experience, but they also, by the very term, tacitly acknowledge that one element of experience is something which is 'given.' But if there be something given, there must be something else to which it is given. To sum up : The existence of experience is a fact, and as such, cannot significantly be doubted. But experience consists essentially in the presentation of an object to a subject, and has no meaning which does not involve the existence of both the latter. The existence of the subject being once granted, propositions can be asserted concerning it. These constitute pieces of knowledge of a kind which has been termed ' know- ledge by description.' Such knowledge is fundamentally distinct from that concrete realization of our own existence, which can in no sense be termed knowledge in itself, but which is, for us, the central and most abiding fact of all. The existence of at least one self being granted, we proceed to assume the existence of other selves. This assumption is in accordance with the pluralistic hypothesis and is justifiable, for it in no way conflicts with the facts. It cannot be proved by the latter, however, so that it is an assumption ; but it must be remembered that no philosophy can proceed without it. Solipsism is logically irrefutable, but quite barren. A man who is not a solipsist can prove a solipsist to be wrong to his own satisfaction, for he knows that he 22 Scientific Method in Philosophy exists ; but he cannot prove to the solipsist himself that he is wrong. To all such attempts, the latter simply replies that the whole thing is merely part of a particularly vivid dream of his own. On the other hand, solipsism is equally unable to prove its case, so that we are at liberty to assume the existence of other people. This assumption is a most valuable one, for it at once opens to us an immense fresh store of knowledge by description, in addition to the knowledge we have through our own immediate sense-experience. VI. Continuity Turning to the objective side of experience, we find in what is called the ' continuity ' of experience, a source of difficulty which has been keenly realized by philo- sophers of all periods. The new realists claim to have disposed of the difficulties by means of their new scientific method, but the sort of continuity they are led to, is not the sort of continuity we find in experience, although, for most purposes, it may represent it ade- quately enough. An examination of the disputes which have always centred round this question of continuity renders it clear that they are almost invariably rooted in the ambiguity of the term. Analysis shows that it is used with two very different meanings. There is first of all what might be called the older, common-sense meaning ; and then the modern, mathematical use of the term. The first may be best illustrated by considering the conception of a continuous material substance. Such a conception has appeared at various times and in various Continuity 23 connections in physical science, as opposed to the atomic view of material substance. A continuous substance is structureless in the sense that it is not built up by the aggregation of a number of small elementary substances. Such a substance, though it seems para- doxical at first sight, would be indivisible ; for the separation of an ordinary substance into two parts consists in overcoming the mutual forces between certain of its elementary particles. But in the case of our so-called ' continuous ' substance, where there are no elementary particles, a moment's thought suffices to show that the operation of division could not actually be performed at all — all of which simply comes to this, that in such a case, when we say that our substance is continuous, we really mean that it is one—nol relatively, but absolutely one and indivisible. In fact, the use of the word ' continuous ' in this way is both arbitrary and unnecessary. Such continuity is just unity. Nothing more nor less than this is meant by the continuity of experience. The individual experience is an indivisible unity. The use of all such words as ' interpenetration ' is simply the groping after the expression of that one fact — experience is one and indivisible. And, after all, what more do we need ? There is no great difficulty in the conception of such a unity. It is one, because it is presented to one subject. The introduction of the ad- ditional notion of continuity is entirely gratuitous, and at once raises fresh and irrelevant difficulties. Much dispute and confusion would be avoided, if people would stop talking about the continuity of experience, and simply speak of its unity. The modern tendency is rightly to restrict and make 24 Scientific Method in Philosophy definite the use of the word ' continuity,' by employing it with one meaning only, viz., that of mathematical continuity. In this sense continuity is essentially a property of ordered series. The new realists suppose continuity of this kind to be typical of experience. Mr Russell \ for example, asserts that the particular degree of continuity known as 'compactness' is sufficient to describe the continuity of experience. A compact series is one in which to any term there is no next term, that is one in which, if any two terms be selected, it is possible to find other terms between them. The number of terms of such a series is, of course, infinite. The view we are considering regards the objective side of experience as a compact series of sense-data. The correctness or falsity of the view just stated hinges entirely upon the fact that a series consists of terms, and that however many terms there are, and of whatever magnitude they may be, they are discrete, each existing per se. Hence, if sense-data form a compact series, we must consider them to consist in an infinite number oi separate members, each of indefinitely short duration. So much seems to be admitted by Mr Russell. Yet again the point is overlooked that sense-data, though absolute and objective for the individual to whom they are presented, are relative and subjective from a universal standpoint. The separation of subject and object is still artificial. All that the theory under consideration has any right to assert is that the introduction of the notion of compact series is one of the most adequate ways of dealing with the unique continuity of experience considered ^ Op. cit, Lect. V. Continuity 25 objectively in abstraction. No doubt results based on analysis on these lines will be sensibly accurate when tested by experience ; but this simply follows from the fact that the original constructions of compact series are sensibly accurate to the same order. It cannot be true, however, that experience is really composed in part of such a series of sense-data, for, as we have seen, the members of a compact series, in spite of their infinitude, are each a definite separate entity. The question might be raised as to whether such a series could have any- thing but an abstract existence. For example, we may write down any member of the compact series of rational fractions, but it is difficult to see how the complete series could exist concretely. Apart from this, however, we cannot look upon sense-experience as a compact series, for its continuity really consists in its unity. The totum objectivum of perceptual experience is one and indivisible, for it is presented to one subject. Hence if we analyze the former into a number of separate entities, we imply that the latter is also a series of separate existences. Again the appeal must be to concrete experience. The subject is one, persisting through change. This much we realize, though the notion of separate instantaneous existences may not be a logically impossible one. Moreover, it should be noted that the object of experience is in part determined by the activity of the subject. Again it is difficult to see how this could be so if there were sense-data existing independently as separate entities, and merely passively perceived in appropriate conditions. It would probably be replied that all that can be said is that certain motor sensations are followed by changes in the 26 Scientific Method in Philosophy other sense-data, the motor sensations themselves being part of the data. But this assertion ignores the fact that the ground of the motor sensations is the activity of the subject. The source of the whole difficulty, then, lies in the distinction between perceptual experience and its sym- bolic representation. Individual experience is unique, particular, and incommunicable. In describing it symbolically, therefore, our description cannot be entirely adequate, for it is conceptual, and the conceptual must always contain some element of the general. Hence the essential privacy of concrete individual experience cannot be comprehended in a descriptive formula. Moreover, in reflecting on experience and its implications, we are bound to attack it piecemeal, and to analyze it by abstraction, on account of our intellectual limitations. This inevitably entails a certain artificial immobilizing and dissection of experi- ence, which we effect by means of concepts. Experience is dynamic and continuous, but concepts are static and discrete, even though they be concepts of things which are themselves dynamic. The above point is illustrated particularly clearly by what Mr Russell calls the logical answer^ to the objections raised against the application of the mathe- matical theory of continuity to experience. Change, he urges, is a fact. But change involves relations, and relations are fundamental. Thus change demands analysis. Now we may grant that relations are funda- mental, but what exactly does this mean ? Simply that so soon as we come to analyze experience reflectively ^ Op. cit., Lect. V, p. 150. Co7itimuty 27 in abstraction, we find that we cannot proceed at all without the concept of what we call 'relation,' in addition to the concept of what we call ' thing.' Yet in the actual concrete individual experience there is no question of ' relation ' or ' thing.' There is just a presented whole perceived by the subject, a whole which simply exists and is given as a whole. For example, let us try to imagine what may be called an * instantaneous ' section of experience. At any instant we perceive in fact but one object, the presented whole. No spatial series of separate parts (however great in number and however small in magnitude the latter may be) enters into the actual experience itself. The same is the case when we include time within our purview, and consider individual experience as a whole. There is here no temporal series of sense-data. Experience in its actuality is not a series. Considered in its entirety (which is the only adequate way of considering it) it is simply 'subject perceives object.' The object is an individual whole, and therefore, by its very uniqueness, cannot be characterized, as stick, by such a general term as relation, for the latter implies the existence of more than one distinct individual. It is only when we come to reflect upon experience that we are bound to consider it piecemeal, and to introduce such general terms as * parts ' and ' relations.' To whatever closeness we may in this way approximate to the actual experience, we can never entirely get rid of that element of the general, which necessarily renders inadequate the conception of what is essentially particular. Mr Russell makes the further statement\ that the ' Op. cit., Lect. V, p. 150. 28 Scientific Method in Philosophy type of objection we have raised against regarding the continuity of experience as being of a mathematical kind is a particular example of a more general doctrine, which, broadly stated, amounts to saying that there can never be two facts concerning the same thing. He points out that the latter is evidently untrue. This may certainly be granted in general ; but how does it apply to the particular thing we are examining, viz., concrete individual experience .'* Strictly speaking, there is only one fact about such an experience in its actuality, which fact may be stated in the proposition ' It exists.' The * it ' of this proposition is the totwn objectivwn, or presented whole, of individual experience. Its actual nature is only realized fully by the particular perceiving subject. In actual perception, before reflec- tion follows, it is perceived as an indivisible unity. It is often called the ' presentational continuum,' but it is more correct to call it the ' presentational unity. ' All other propositions asserting facts (so-called) about experience, are simply attempts to express as adequately as possible in conceptual form the nature of private experience. They are inadequate, for the proposition expressing the sole and particular fact of the existence of the perceived object in its peculiar intimacy and uniqueness realizable only by the percipient, is replaced by a number of propositions, expressing our attempts to deal, by general characterizations, with something which is essentially particular. The attempts furnish us, for the most part, with a sufficient approximation, but in dealing with ultimate questions, it is of the first im- portance to remember that of necessity they are but approximations. Continuity 29 It is evident, then, that the results of analysis by the scientific method cannot be fully adequate. This does not detract from the value of the former in practice, so far as it goes, for it is the most adequate conceptual method of dealing with experience. We could have no better conceptual way of representing what is called the ' continuity ' of experience than by mathematical continuity. But this representation must not be regarded as a final complete solution, for the reasons we have given. It is necessary and sufficient for purposes of calculation, and for the establishment of the validity of certain physical conceptions, but the final solution ot the difficulties which have been raised, lies in realizing that the so-called ' continuity ' of experience is actually its unity, being, as it is, the experience of one subject. What exactly is meant by 'one' thing ? For example, we talk of a chair or a table as being one object ; but the physicist regards the chair or the table as made up of a number of other single objects, viz., molecules. The molecules again, are made up of atoms, and so on. It appears that there are two distinct meanings to be attached to the word ' one ' when applied to things. Thefirstissubjective or teleological, the second objective and absolute^ It is in the former sense that a chair or a table is one object. In other words, we speak of a thing as being one, when it functions as one in relation to our purposes, or to the purposes of other subjects of experience. On the other hand, we can only regard a thing as being one in the absolute sense, when it is a ^ We might, perhaps, distinguish a third use of the phrase ' one thing,' viz., as applied to a body which moves as a unit. [See Perry's Fresetit Philosophical Tendencies, p. 52.] 30 Scientific Method in Philosophy true individual. Where do we jfind such individuals ? Evidently physicsdoes not deal with them. Its molecules and atoms are not unique. Subjects and their experi- ences are the only true individuals. The self may be considered an absolute unit, for it is unique and indivisible. For this reason alone, pluralism, being expressed in terms of selves, would have an advantage over any description of existence in terms of sense-data, for the latter are, as we have seen, purely artificial units. Finally, it should be pointed out that the consideration of the meaning to be attached to the concept of ' one thing ' is in no way connected with the meaning of the number * one.' VII. Causality Among the chief categories which are commonly regarded as applicable to experience, is the category of causality. This category has been a source of difficulty and confusion owing to the unfortunate vagueness and ambiguity with which the term ' causality ' is frequently used, especially in its scientific application. Analysis by the new scientific method has done much to clear up this confusion^ As a result of this analysis, it is pointed out that with regard to the objective side of experience, we can only say that the sequences observed in it are characterized by sufficient similarity to admit of more or less adequate description in general terms. Hence we can enunciate certain propositions in virtue of which the occurrence of some events can be inferred from the occurrence of other events. To these pro- ^ B. Russell, op. cit., Lect. VIII. Causality 31 positions we may give the name of ' causal laws.' Therefore, from the purely scientific point of view, we should go no further than the mere statement that such causal laws do subsist. This is evidently true, for if we take objective experience as it stands, there is simply the fact that certain sense-data are invariably followed by certain other sense-data. The sequence contains in itself neither hint as to the reason for this invariance, nor warrant that it will continue to hold in the future. Scientific observation alone, then, can do nothing more than formulate descriptions of these sequences, together with the statement that it seems probable that they will continue to hold in the future as they have invariably done in the past. From this point of view, any further extension of the principle of causality is both unnecessary and unjustifiable. If the exponents of the scientific method were content to stop at this point all would be well ; but they go further, and assert that the meaning of causality considered above is the only valid one. Yet the roots of the concept of causality go far deeper than this. If we trace the development of this concept during the growth of experience, we find that it is inseparably bound up with the notion of efficiency or activity. We ourselves, as active agents, initiate changes in our environment, and we realize our activity to be the ground of which these changes are the consequence. Many of the sequences which occur in experience independently of us, we can reproduce at will. Thus we arrive at the conception of efficient causality as distinct from merely descriptive causal laws, ourselves being efficient, and, for the most part, self-determined 32 Scientific Method in Philosophy causes. Inevitably we come instinctively to consider efficient causality as the ground of those sequences which we observe in experience. Logically, as we have seen, mere observation only gives us the right to assert that certain sequences do recur, and to state the fact in a general proposition. Seeing, however, that we actually realize self-activity to be the ground of many sequences — sequences which we can always render essentially similar — there is no reason why we should not adopt, at any rate hypothetically, the view that all observed sequences have their ground in the activity of experiencing subjects. In this way, causality, as applied to perceptual experience, comes to have a definite concrete meaning, namely, the efficiency of active individuals. No doubt the concept is anthropo- morphic, but that simply means that it is based on the nature of the subject, as distinguished from the object, of experience. Hence we realize that efficient causality certainly exists, and we are therefore justified in attempting to find a satisfactory interpretation of the sequences which occur objectively in experience, by the application to them of this concept of efficiency. Activity is fundamental. Everyone realizes what it is to be active. Yet certain modern representatives of the traditional idealist schooP, dismiss activity as ^ E.g., F. H. Bradley in Appearajice and Reality. The New- Realists also reject activity (cf. R. B. Perry in Present Philosophical Tendencies, pp. 70, 99 and elsewhere). It is stated that all that is perceived is certain muscular sensations, etc., but no 'power.' This is not denied, but the fact (too often overlooked), which lies at the root of the question, is that activity is not an object of perception or knowledge at all. It is not presented to the subject, for it is the sub- ject who is active. But we realize that we are active, although our Causality 33 pure illusion. It is difficult to see what the assertion ' All activity is illusion ' can possibly mean, if it mean anything at all. When I think or do, I say that I am active. All that is meant by activity is a living and doing. If the idealist asserts that living and doing are illusions, the reply is simply that the illusion at any rate exists, and therefore it is the illusion itself that we mean by activity, if it be an illusion. There is no meaning at all in the term 'illusion' as applied to direct experience. It is only when wrong judgments are based on experience that illusion can be said to exist. When we talk of being active, it is simply a way of specifying a certain fact. We may draw wrong conclusions from the fact, and in that case it is correct to say that we are the victims of illusion. Yet, however that may be, the fact remains. Returning to the new realists, we find then that they deny the validity of the concept of efficient causality. They maintain that the only meaning which causality can have is that which we have seen it possesses in descriptive science, namely, that in objective experience certain essentially similar sequences recur, which fact is expressed in a number of propositions (one for each set of similar sequences) termed 'causal laws.' This activity is not presented to us. The realists and others might just as well deny the existence of perception, because we only find certain things given, of which our own 'perceptivity' is not one. We do not perceive our perceptivity — it is not an object of knowledge — but we realize that we perceive things, and the proposition asserting this fact is of course a piece of knowledge (by description, not perception). Hence there is no more ground for denying the existence of activity in general, than for denying the existence of perception, in which the subject is active in a particular way. R. s. p. 3 34 Scientific Method in Philosophy they consider is all that can be said on the subject. As an example of a causal law the law of gravitation is frequently cited\ The latter contains no reference to * cause' or 'effect.' The expression of the law, as applied to a material system, simply takes the form of certain differential equations. From these it follows that the configuration of the system at any given instant is a function of that instant, and of the configuration at two given instants. This is true enough, but the fact remains that such differential equations, and the function which is their integral, are purely descriptive. They contain no hint as to 'how' and 'wherefore.' They simply tell us what does occur, without suggestions as to why it occurs. Moreover, there is still the question as to what deter- mines the particular ybr?;^ of the equations from which the configuration at any instant can be deduced. It is not determined by logic, for logic and mathematics can give no answer to the question. As already suggested, the ground of the motions of such a system lies in activity. The particular nature of the motion, with its corresponding typical descriptive function, is determined by the particular type of activity of the agents concerned. The fact that our differential equations are shown by experience to hold for past and future as well as for present, simply means that the activities of certain in- dividuals are sufficiently habitual to admit almost com- pletely of description in general terms. We have seen that the introduction of the notion of active subjects does more than shift the descriptive formula one step ^ See, e.g., B. Russell, "On the Notion of CdMse,'' Scientia, p. 327. [See p. 17 above, footnote.] Causality 35 further back, for it provides an explanation as opposed to a mere description. In this connection Mr Bertrand Russell discusses the possibility that the universe may be a deterministic system\ He considers a system of this kind to be one whose state at any time is given by a functional relation involving certain data which specify the state of the system at certain times, time being the independent variable. Such a functional relation is clearly equivalent to a description in general terms. But the universe contains unique individuals whose nature cannot there- fore be exhaustively described in general terms. Even were such exhaustiveness possible in a relation of the type considered, the data contained in that relation would consequently have to comprise every individual in the universe, and the experience of each at every instant of his history. It is true that Mr Russell admits that the relation may be of strictly infinite complexity, but if it must necessarily be of the order of complexity we have indicated, it would simply be a recital of the whole history of the universe. That is, it would have to contain explicitly all the information which it might have been hoped to contain implicitly. There would, in fact, be no room in it for a variable at all, for it would contain all values of such a possible variable in its data. Hence, no such relation can be significantly applied to the universe in any ultimate sensel It is then urged against the notion of efficient causality that the future determines the present to the ^ Cf. "On the Notion of Cause," Scientia, pp. 331 ff. ^ This point is considered in detail in the essay on " The Notion of a Deterministic System." 3—2 36 Scientific Method in Philosophy same extent as does the past ; in other words, cause does not ' compel ' effect, in some sense in which effect does not compel cause \ But, again, this determination of the present by the future is only logical and descrip- tive. Even assuming for the moment that certain functional relations actually subsist which are signifi- cant as a complete or partial description of such a Universe as ours, there is still a difficulty in the way; for although the formulation of the relations may work either way in time, we cannot xgnor^th^ one-directionality of time in concrete experience. The relations, though symbolic of a dynamic process, are themselves static. They simply assert that on given assumptions such as uniformity, there is logical dependence of the present on the future, just as there is logical dependence of the present on the past; but they fail to comprehend con- crete experience fully, in that they ignore the actual fact that time progresses in one direction only. Thus, if all that these relations imply were true, there is no reason why the crime should not sometimes follow the punishment which is its due, nor why the determination to build a house should not follow the appearance of that house on the scene. Perhaps the existence of purpose and consequent action leading to realization illustrates best of all the hopelessness of the attempt to replace the notion of causal efficiency by the notion of mere logical dependence. For there is certainly a sense in which it can be said, for example, that the house was built because Jack determined to build it, in which it cannot be said that Jack determined to build the house because the house was built. Moreover, we are indu- ^ B. Russell, Our Knowledge of the External IVortd, Lect. VIII, p. 220. Causality 37 bitably aware that our actions determine their ends in a sense altogether different from that in which the ends determine the actions. This could not be so if the relation between them were purely logical. The matter may therefore be summed up somewhat as follows : The true meaning which causality has for us is rooted in the realization of our own efficiency, as active individuals. The active individual is the 'cause.' The end which his (generally purposive) activity accom- plishes is the 'effect.' The scientific method, however, takes the sequences which occur in experience as they stand and determines what may truly be said of them per se. In the first place, it finds that sequences con- tinually recur sufficiently similar in nature to admit of a considerable degree of general characterization. Secondly, it follows that a general proposition may be affirmed with regard to each recurring sequence, where- by the occurrence of one event may be inferred from the occurrence of another event. Thirdly, there is no guarantee (except the rather doubtful one of probability) that such propositions will continue to hold in the future. Finally, it is seen that we can go no further than this from the objective standpoint of science. It might also be pointed out that, strictly speaking, the term ' causal law ' ought not to be applied at all to such propositions as we have been considering. For, in view of the con- crete meaning which 'cause' has for us, the word 'causal' implies that the sequences to which the propositions refer, have their ground in the activity of individuals^ The results of this analysis by the scientific method ^ If this implication is granted, however, the term 'causal law ' is of course appropriate. i ^ J nJ ^ 38 Scientific Method in Philosophy are valuable for the philosopher, for they make clear the exact nature of the assumptions he is making in applying the pluralistic hypothesis to the sequences observed in experience. Still more valuable are they for the physicist, seeing that they warn him from un- warrantable applications of causality, and point out the only valid way, from the scientific (and therefore de- scriptive) point of view, of looking upon the succession of phenomena with which he deals. There is no doubt that physicists of all times have been strongly influenced by the notion of causality based on subjective activity. One fact alone is sufficient to show this, namely, the curious reluctance which has always been shown to accept the idea of action at a distance. Attempts are invariably made to reduce everything to terms of con- tact action. The reason is that our own interference with the environment is conditioned by the contact of our bodies. Had we been endowed with powers of levitation and removal without contact, the notion of action at a distance would probably have been adopted as a matter of course. Thus far, and in this application, we may recognize the truth and value of the results due to analysis by the scientific method. Pluralism, on the other hand, approaches the question in a different way, and with a different purpose in view. It is concerned not simply with the phenomena as such, but with an explanation of them which shall satisfy such beings as we are. On the basis of our own existence as efficient individuals, and of the fact that sequences observed independently of our activity can often be essentially reproduced by that activity, it proceeds to explain all sequences by the Causality 39 activity of individuals. This, of course, it is required to do if its hypothesis is to hold, and this it is success- ful in doing while no facts can be brought forward to disprove its case. VIII. Other Categories of Experience Although the consideration of continuity and caus- ality brings out most clearly, perhaps, the distinction between the aim, method, and scope of pluralism and the new scientific method respectively, incidentally making clear the value to be attached to the criticism of the former by the latter, it is of great importance to examine the other categories of experience if a clear conception is to be framed of the basis on which plural- ism rests. The attention may be directed in the first place to the categories of Substance and Attribute. A review of the classical attempts to deal with the notion of substance makes it clear that the problem resolves itself into an endeavour to reconcile the principles of permanence and change. Heraclitus, who was the first to bring out more or less plainly the nature of the diffi- culties involved, held that only change is permanent ; but closer examination shows that, with any significant meaning which can be attached to the term ' change,' the truth of the matter is that change implies perman- ence. For, in the first place, it is apparent on general grounds that if there is a change, there must be a thing which changes, the said thing maintaining its identity throughout the change. Otherwise, there is simply one thing and then another thing, that is, mere succession and not change at all, properly so-called. From the 40 Scientific Method in Philosophy scientific standpoint we certainly do consider mere alteration alone, that is, simply a succession of different presentations. But from the subjective point of view, if I have first A and then B before me, I can in no significant sense be said to have apprehended a process of change; at most there has been a change in myself, and this, since it is I who have perceived both A and B, assumes my permanence. As a matter of fact, we do only perceive a process of change, as such, at a high level of experience; yet, when we have reached this level, we feel impelled to look for a permanent basis as a ground of the ceaseless flux of experience, whether it be logically necessary or not. If we analyze the meaning of a process of change from a conceptual point of view, it would seem to be somewhat as follows : At a given time certain true pro- positions may be asserted of a given individual. At another given time, certain other true propositions, wholly or partly incompatible with the former set, may be asserted of the same individual. If we consider the propositions as particular values of certain propositional functions, the particular value considered of the argu- ment of these functions remains the same throughout. This is the symbolic counterpart of the fact that the individual considered maintains his identity. Evidently, from this point of view, it would be difficult, and perhaps impossible, to formulate in words the reconciliation of the principles of permanence and change. The reason for this difficulty is that, conceptu- ally, we necessarily consider experience as a time-series. Let us attempt to estimate the true bearing of this. And here, perhaps, an analogy may be of use. We do other Categories of Experience 41 not consider the identity of an individual at any given time to depend upon his position in space. At a given time, I should not be a different person if I were in London and not in Edinburgh. That is, identity is not conditioned by the spatial series\ Why, then, should it be conditioned by the time-series? It would probably be answered that the nature of the individual is different at different instants of the time-series. He develops (or the reverse) in time; and it certainly seems, at any rate piama facie, that time is more closely bound up with existence than space. But what is the time-series referred to ? Not the conceptual or universal time- series, for that is a mental construction. The private time-series of the individual concerned, then '^ But his time-knowledge is based on change and the existence of the memory-perspective, which implies maintenance of identity. Hence this line of thought bids fair to end in a vicious circle. The truth is that we can never entirely resolve the difficulty conceptually for reasons we have considered in another connection, namely, that actual existence is particular and cannot therefore be comprehended in a conceptual formula. The conceptual formulation of the facts, if pressed too far, necessarily gives rise to diffi- culties which do not admit of complete solution. Never- theless, it is possible to indicate to a considerable extent a method of viewing- the facts which bringrs us as near as possible to a complete comprehension. In the first place, it is necessary to get rid of the time-bound view of experience. Just as in forming as adequate a con- ^ I cannot, of course, be in two places at once, but that does not mean that I am what I am because of my position. 42 Scientific Method iri Philosophy ception as possible of the object of experience it is necessary to consider it as an indivisible whole in respect of space-time, so also must the individual subject of experience be regarded as a unity beyond space-time. In other words, we must try to conceive some such world as the universe of Minkowski. The latter applies his conception to the problems of physical science. In such a universe as he imagines, the entire existence of a physical system is specified by means of three space- and one time-coordinate, and is presented as a whole. In a somewhat analogous way we must look upon the individual subject as an entity transcending space and time. His existence can only be specified as a whole ; it is neither punctual nor instantaneous. From a logical standpoint, the proposition 'He exists' must not be supposed to imply any spatial or temporal reference ; that is, there is no real meaning in the notion of existence at a given point or at a given time, though we may adopt the idea conventionally. The point is brought out still more clearly if we consider non-existential pro- positions which may be asserted of the individual. In examining this point previously, we pointed out that one set of propositions might be true at one time, and another partly or wholly incompatible set at another time. If, however, the propositions are modified by the insertion of date and place, their truth is independent of space and time. The date and place referred to may be considered as uniquely determined. For example, if they be specified by position in conceptual space and time, they will yet be connected by a one-one correlation with the private space and time of each individual. Hence there will be a set of propositions concerning other Categories of Experience 43 the individual which will be true once for all. Regarding the propositions as particular values of certain propo- sitional functions, the particular value of the argument of these functions is the individual considered. Since the propositions, however, are not limited as to their truth-value by space or time, the particular value of the argument cannot be dependent upon space or time. Thus the individual is a unity transcending space and time. From this conceptual standpoint, such notions as ' process ' and ' development ' lose nothing of their meaning or value, but, like all concepts which refer to matter-of-fact, their inadequacy leads to the difficulties we have been analyzing. Yet although the solution of the problem of identity and change is attended by such difficulty when looked at from an abstract point of view, the concrete solution is more easily realized. The self combines, in a particularly complete way, the principles of identity and change. In spite of change, I realize myself to be the same individual that I once was. Even if we cannot formulate in words, on account of its uniqueness, the exact nature of this reconciliation of change and permanence in the subject of experience, it is, to say the least, almost as satisfactory to realize its existence. This being so, we are encouraged to apply the pluralistic hypothesis by regarding the per- manent ground of the changing flux of experience as consisting in individual subjects. It is important to comprehend clearly the general bearing of the view of space and time here adopted. Subjects of experience cannot be considered to be in any sense * in space and time.' The latter phrase, indeed, 44 Scientific Method in Philosophy signifies an abstract concept. Analysis distinguishes in the object of perceptual experience a succession of elements each of which has a sensible duration. The qualities and relations here implied are those to which we give the name ' temporal.' On them we base the concept of a time ' in ' which the elements have their being. In a similar way, to the characteristics implied in the co-existence and sensible extension of the elements, we give the name 'spatial,' and hence pass to the abstract conception of space. But in the individual subject no such parts or elements, whether co-existing or succeeding one another, can be distinguished. Hence the subject is neither spatial nor temporal. It may per- haps be urged that we surely have a sense of enduring, which would imply time; but even if this be so, duration alone cannot give time. Our so-called sense of enduring simplycomes back to the realization of our own existence, apart from any spatial or temporal implications. This is borne out by the consideration of any proposition, in which the subject is one term, having a spatial or temporal reference. In all such cases, a brief analysis shows that, so far as space and time are concerned, the assertion refers entirely to the object of experience. For example, the proposition : ' I remembered what happened to Smith, when I went to London last Friday,' simply asserts that there is in my object of experience (considered as a whole) a certain complex of motor, sensory, and ideational presentations exhibiting such- and-such spatial and temporal characteristics and rela- tions. It would be hopelessly prolix, however, to attempt to make all our statements about the subject more exact by putting them in some such form as the other Categories of Experience 45 above. For discursive exposition it is necessary to employ the usual shorter form, even though it appears to imply that the subject is himself in space and time. It may be well to remark, in order to anticipate any charge of inconsistency, that In all that follows the foregoing view is maintained\ and wherever statements (and they are unavoidably many) appear to relegate the subject to the realm of spatial and temporal entities, their use is simply an inevitable necessity if intolerable circumlocution is to be avoided. All such statements, however, are capable of exhibition in a form (as ex- emplified above) in which it is evident that any spatial or temporal reference is to elements in the object of experience alone. Before considering the notion of ' attribute,' it may be of interest to make a short digression at this point, by referring back to Minkowski's conception of a space- time world and its bearing on philosophy. The con- ception arose in the first instance out of difficulties similar to those we meet with in analyzing change. Recent researches in physical science have brought to the fore, with increasing insistence, the question as to what meaning, if any, is to be attached to such notions as absolute velocity and absolute position. The con- troversies to which these problems gave rise culminated in the enunciation of the well-known Principle of Rela- tivity, There are several ways of stating the latter, but each amounts to this: 'Different descriptions of the same system will be given by different observers.' A description depends on the motion of the observer ^ Cf. especially the essay on "Immortality," for a detailed discus- sion of the points raised here. 46 Scientific Method in Philosophy relative to the system. There is no criterion which may be appHed to a set of descriptions, by means of which a single ti^-ue description may be determined. All the descriptions are true. The reason is that if we carry our analysis far enough, we are bound to consider the fact as a whole, namely, not only is there an object, but the object is seen by an observer. Consequently each perception is a different fact, and even admitting the object to be the same, for the purpose of argument, the descriptions, though all true, will be different, for each actually involves the observer and the observation as well as the object observed. Hence physics, which purports to describe things independent of any particular perceiving subject, is compelled in the end to take account of that subject. This is inevitable, seeing that the concepts of physics are constructions based in the first place on individual perceptions. In fact, the prin- ciple of relativity, as applied to physical science, is a particular example of the more general philosophic fact that while the experience of the subject is objective and absolute for him, it is subjective and relative from the universal conceptual standpoint. In considering existence, then, from the conceptual point of view, we are continually brought face to face with its relativity. This is the root of the difficulty in the problem of change. As regards physical science, Minkowski succeeded in transcending the difficulty of relativity by introducing this idea of a space-time world. In this way he not only made clear the source of the trouble, but also indicated how it might be eliminated in analysis. It is simply a question of taking a wider view of existence; and in considering an individual who other Categories of Experience 47 changes and yet maintains his identity, we shall get rid of the difficulties to a great extent if we proceed on similar lines. In specifying an individual, reference must be considered to be made to a space and time- transcending unity. In such a proposition as 'A went to London on Saturday,' A must not be supposed to be specified by any time or place. A is an entity whose existence is considered as a whole. The proposition, though it contains a spatial and temporal reference, is asserted of this individual whole, which transcends both space and time. In the existence of the self, then, the principles of permanence and change are reconciled ; therefore the self is a concrete actuality corresponding to the concept of substance. A distinction is sometimes made between substances and Substance. The latter is regarded as some unity which is the ground of all existence. With the questions as to whether such a unity exists, and if so, what is its exact nature, we are not here concerned ; but enough has been said to indicate that we can only form a satisfactory idea of such an all-pervading sub- stance, by considering it to possess all the general characteristics of a self or subject of experience. Keep- ing, however, to the selves which we know to exist, and which we have identified with substances, what meaning ought we to attach to the term ' attribute ' as applied to such individuals? In the first place, the term should be strictly limited. In particular, the fact that A stands in a certain relation to B must not be held to constitute an attribute of A. Propositions assigning attributes to an individual are of the subject- predicate form. For our purposes, such propositions 48 Scientific Method in Philosophy may be regarded as falling into two main types. These types may be illustrated by the two propositions : ' He is just,' and ' He is French.' The first makes a statement about the nature of the individual to whom it refers. The second, as such, asserts nothing directly about the nature of the individual, but is rather a specification of certain relations in which he stands. It should be noted, however, that this proposition may be held to imply a number of other propositions of the first type, namely, those assigning to the individual the characteristics he shares in common with all Frenchmen. The term ' attribute ' might well be limited to the predicate of propositions of the first type. A brief consideration suffices to show that such propositions invariably imply something about the mode of activity of the individual concerned. For example 'He is just' really means *He acts justly.' We base our judgments about the individual on observations of his actions. Thus the attributes of the individual are the ways in which he acts. The fundamental proposition about an individual A is ' A exists,' which is equivalent to 'A acts.' A is 3. unique particular who cannot be further specified symbolically. What we call his attributes consist simply in his mode of activity. Two categories which seem interwoven particularly closely with the fabric of experience are those of Quality and Relation. Quality and attribute are often used as synonymous terms, but to maintain precise definition we may distinguish between them. Just as we haye taken the conception of attribute as appro- priately applicable to the individual subj ect of experience, other Categories of Experience 49 so may we appropriately apply the concept of quality to the object of that experience. For example, the sense-data presented through different organs differ in quality, e.g., sensations of colour differ from those of touch. There are also qualitative differences between sense-data presented through the same organ, e.g., red differs from blue. Differences of intensity are really qualitative too, although expressed quantitatively, thus implying a relation of more or less between the sen- sations\ It is correct to say that Quality and Relation are fundamental in the object of experience, if the exact implication of the statement is clearly comprehended. In concrete experience, as such, there is no question of quality or relation. There is simply a given indivisible unity. This unity is particular, and can only be referred to by such words as 'it' or 'this.' Its characteristics cannot be specified conceptually with adequacy. We cannot take a single step in analyzing experience, however, without introducing the concepts of quality and relation. It is this which should be meant by the statement that quality and relation are fundamental. They are fundamental to the extent that we cannot reflect upon experience at all without introducing them ; but into the actual experience as such, they do not enter. This is evident when we remember that quality and relation are general conceptions, whereas experience is essentially particular. All we can say is that when attempting to represent experience conceptually {so far as it can be thus represe7tted), by hypothetically con- ^ For a detailed discussion of this point, see the essay on "The Intensity of Sense- Data." R. s. p. 4 50 Scientific Method i7i Philosophy sidering it to exhibit certain general characteristics\ we find that two of the most indispensable of such characteristics are quality and relation. Failure to realize the foregoing has been a fruitful source of objections levelled at the concepts of quality and relation. For example, consider Mr Bradley's criticism of the concept of relation" on the ground that it implies an indefinite regress, seeing that a relation between terms requires further relations to relate it to its terms, and so on. There would be some point in this criticism if we asserted that the concept of relation adequately represents experience. But, admittedly, such general conceptions as quality and relation cannot adequately comprehend the essentially particular. All that is claimed is that in representing experience as adequately as possible by general characteristics, the introduction of the conception of objects between which certain relations subsist, is, for the most part, perfectly satisfactory for the purpose of calculated prediction and interference in the course of events. No such complication as the introduction of fresh relations between the relation and its terms is needed to carry on the reasoning based on our premises, and this reasoning is justified, so far as it goes, by empirical verification. Thus Mr Bradley's objection cannot hold good, for we do not suppose our conceptual system of terms and relations to comprehend experience fuljy, though on the other hand it is sufificiently adequate to describe it and to render possible sufficiently accurate ^ These are hypothetical in so far as we consider them to be ab- solutely identical elements in every individual experience. '^ Vide Appearance and Reality, 2nd ed., Ch. in, pp. 30 ff. other Categories of Experience 51 prophecy and successful interference in the course of events. Therefore the objection has no significance as appHed to perceptual experience as such, nor can it be urged against our conceptual apparatus; for we construct the latter ourselves, and find it sufficiently competent to perform its task, which is the only significant test. We may conclude the investigation of the categories of experience by examining two of a somewhat different type from those already considered. They are the categories of Means and End, or Purpose. These categories are only significant in application to a universe containing individual subjects of experience. The categories we have been analyzing are applied in the first place to the object of experience, though the origin of the concept of the category is in some cases subjective, but the category of purpose is primarily applicable to the subject of experience, for it is a characterization of activity. It seems probable that all activity is originally purposive, though oft-repeated actions become less and less consciously purposive and more and more reflex and habitual. As regards those individuals whose nature we realize most clearly, namely, selves at our own level of development, the ground of activity is in most cases evidently pur- posive, and not purely material in the scientific sense. For in science, 'material' means 'phenomenal,' whereas the ground of our own activity is the very opposite of phenomenal. Certainly phenomena in part determine the purposes which guide the activity, and the latter may itself be limited by material conditions, but the ground of its initiation is subjective or real as opposed to objective or phenomenal. 4—2 52 Scientific Method in Philosophy We may, however, attempt to apply the category of purpose to the ground of what we observe in the object of experience. In such observation we at once notice actions which may be regarded as purposive by analogy with our own. In fact all organic life appears to exhibit this purposive character. We might perhaps describe the activity of an organism in terms of mole- cular action, that is, in terms of the purely objective constructions of physics, though it is by no means certain that organic activity could even be completely described \}i\\!&. In any case, the description, if complete, could not be general, for every organism is unique. Each would therefore require a separate description. On the other hand, we may explain the organism by the organized collective activity of individuals, thus changing the terms from purely mental constructions to concrete entities whose nature we can all more or less realize. The fact that organic activity is thus apparently teleological strongly suggests the applicability of the pluralistic hypothesis, at all events to organic matter. We say 'apparently teleological,' for it is not certain that the existence of such teleology can be conclusively proved from a logical point of view. Could it be so proved the fact would be of enormous significance, for pluralism would immediately be verified as regards organic matter, since the existence of purpose implies the existence of experiencing subjects. However that may be, the attempt to describe organic life in purely physical terms invariably leaves an inexplicable residue of spontaneity, whereas its explanation in terms of individuals differing only in degree from ourselves, is other Categories of Experiejtce 53 'enabled to take the latter fully into account. The pluralistic hypothesis is therefore to that extent justified. Inorganic matter may be treated by a similar, but somewhat modified, theory. It is analogous to organic species which have become stationary at some period of their evolution. It approaches the lower limit of development. It may be regarded as comprising in- dividuals of an extremely low order of mentality, who therefore exhibit the minimum of spontaneity and the maximum of habit in their reactions. They are thus particularly susceptible of an almost complete description in oreneral terms. It may be concluded that wherever the category of end or purpose can be successfully applied, so also may the pluralistic hypothesis be applied to the same extent ; for the existence of purpose implies the presence of mind, that is, of subjects of experience. We have seen that pluralism is in this way applicable not only to the ground of what we term the organic activity observed in the object of experience, but also (with certain modifications which yet conform to the necessary conditions required by an explanation in terms of mind) to the ground of inorganic activity. The whole field of experience may therefore be covered by pluralism ; though we are here concerned, not with the details of the application of that hypothesis, but only with its basis. 54 Scientific Method in Philosophy IX. Summary and Conclusion We may end our examination of the two most important tendencies in the trend of modern philosophic thought, by summarizing the results to which we have been led. In this way, the scope and limitations of each school of thought, and the results which each may hope to obtain, will be set out concisely and in brief compass. The scientific method, as expounded by philosophers of the new realist type, is embodied in an analysis of the object of experience, with a view to ascertaining the form of the facts concerned. With the particular content of any set of facts it is the business of one of the particular sciences to deal. Philosophy aims at determining form, without reference to any particular content. This philosophic analysis has an important application in the investigation of those concepts which we ordinarily apply to the object of experience, notably the concepts of physical science. Since all observation, avowedly scientific or otherwise, must start from sense- data, and since all verification of calculation based on such observation must lie in an appeal to sense-data, it follows that if our concepts are valid, they must be capable of being exhibited as logical functions of sense- data. The analysis of the concepts, therefore, finally resolves itself into an attempt to^ build up such constructions of sense-data as may be considered satisfactorily to represent the concepts involved. Hence the method, at any rate in the last stages of its applica- tion, is constructive. Nevertheless, its function is evidently to a considerable extent critical. Its field of application consists in the whole of the objective side Summary and Conclusion 55 of experience, and we may willingly admit the claim of its exponents, that it is the only method of obtaining accurate objective scientific knowledge, provided it is clearly recognized that it is subject to two limitations springing from a common root. In the first place, all individual experiences are essentially particular, and the assumption that they exhibit certain general characteristics of form must therefore be regarded as an approximation which is only justified by the fact that it works satisfactorily in practice so long as we are not concerned with a final complete adequacy, and by the still more cogent fact that we are bound to it by our intellectual limitations. In the second place, the units with which the scientific method works are sense- data, and the sense-datum is a purely artificial and conventional unit. The object of experience is an indivisible unity, and (whatever convention we may be compelled to adopt for the purposes of calculation) cannot be considered to consist in a series of members termed ' sense-data,' compact or otherwise. Keeping in mind these limitations of the method, its critical and constructive value in its own field is apparent. In any case, however, its results are purely descriptive. Its exponents claim that the determination of such results as their method affords is the only business with which philosophy ought to concern itselP. As the opinion of philosophers of all ages, with very few exceptions, has differed widely from this, the claim must be regarded as a purely arbitrary one. In making it, its supporters are proposing their own definition of philosophy, a definition which is not accepted by the ^ But see note on p. 7 above. 56 Scientific Method in Philosophy majority of philosophers. In addition to the critical investigation of the form of facts, it is the further business of philosophers to provide an hypothesis which may be said to explain those facts to the satisfaction of such beings as ourselves, while remembering that, although no hypothesis can be regarded as infallible, it may be invested with a very high degree of probability, in virtue of its ability to fit the facts already known, and to furnish explanations of those new facts which are constantly forthcoming. In applying the scientific method to the various problems of philosophy, the new realists have little or nothing to say about the subject of experience. Such brief references as are made imply that the subject, if it exists at all, is merely an inference. But, as we saw, doubt of the existence of the subject is without signifi- cance ; and, moreover, although the existence of the subject may certainly be inferred — immediately inferred, indeed, from every single fact of experience — there is, in addition, the far more important central and unique fact of our experience, namely, the concrete realization of our own existence. There are several important consequences of this ignoration of the subject. In the first place, certain problems are considered to be outside the scope of philosophy. Such, for example, are the problems of ethics. Mr Russell says : " The difference between a good world and a bad one is a difference in the particular characteristics of the particular things that exist in these worlds. It is not a sufficiently abstract difference to come within the province of philosophy^" Again ^ Op. cit., Lect. I, p. 26. Summary and Conclusion 57 this limitation of the philosopher's task is a purely arbitrary one. The terms ' good ' and * bad ' are only significant in a universe containing such individuals as ourselves. In their fundamental application they refer to the acts of an individual considered in relation to other individuals. Judging the individual by his acts, we may conventionally use the terms in reference to the individual himself. Further, we may also apply them to the object of experience considered in relation to the subject, thus introducing the categories of value. The latter, being categories, are necessarily concerned with the form of facts, thus coming within the scope of philosophy even as limited by the new realists. More- over, even in application to the acts of the individual, the terms 'good' and 'bad' might be taken to refer to the form of the acts. For example, we might define ' good ' as the class of all acts which have as their motive the benefit of others. In any case 'good' and 'bad' are general characterizations, and it is the business of philosophy to define their meaning precisely, and to determine their application. The ignoration of the subject also leads to the consequence that the results of the scientific method are purely descriptive and not explanatory. It does not seek the ground of the object of experience. Men have always felt that there must be such a ground, regarding sensations, /£?r^^, as flimsy and ever-changing manifestations of a more substantial reality. In dealing with sense experience, we find it easy to distinguish and to compare, and generally to construct a complicated network of terms and relations. The facility with which we perform such conceptual gymnastics tends to make 58 Scientific Method in Philosophy us lose sight of the fact that the object of experience, as given, is an indivisible unity. When we turn to the subject, however, the case is different. We come down to bedrock almost at once. Any attempt to analyze the subject into parts and relations, at once shows the futility of regarding it in any other light than as a single unity. In the case of the individual subject, we are therefore concerned with content rather than with form. We find in it a substantiality which the object of experience seems to lack, for we are ourselves individual subjects of experience. We are thus led, with pluralism, to look for the ground of the object of experience in the activity of individuals differing only in degree from ourselves. Our own existence is for us the central fact of the Universe, and any attempt to limit philosophy to enquiry into matters where the existence of the subject may be safely ignored, on the ground that we must deal only with form and not with content, is both arbitrary and highly unsatisfactory. After all, the facts of the universe are particular, and it surely lies with philosophy to explain those facts so far as it can. Every philosophic theory must necessarily assume certain logical axioms in accordance with which its reasoning is to be carried on. It is one of the principal tasks of modern logic to reduce such axioms to a mini- mum. Taking these principles of reasoning for granted, the theory will proceed to start from certain definite facts as data. The more incontrovertible and immediate the facts, and the more fully realizable, the more satis- factory is the theory likely to prove. Pluralism starts from the existence of the self. It makes the assumption Summary and Conclusion 59 of the existence of other selves. Thus it is based on the existence of entities at least one example of which we know to exist, and whose nature we actually realize. It is therefore superior at the outset to theories which start from entities such as sense-data that are objective for the individual. For, in the first place, all such objects are purely artificial units, whether they be sense-data, or the constructions of sense-data which constitute the units of the world of physics. On the other hand, a self is a true unit, a true individual. I n the second place, we realize what a self is. We perceive a sense-datum, but we cannot realize what it is, in itself. Moreover, there is the further point that selves cannot be resolved into sense-data, whereas it may be possible to explain sense-data in terms of selves. The next step in the development of pluralism is the analysis of the growth of the experience of the individual subject by the genetic method. It is not sufficient to enquire what certain concepts ought to mean, but also what they do actually mean for us, and how they come to acquire that meaning. If we proceed on these lines, particularly with reference to the chief categories of experience, we arrive at results which in each case, while not leading to it as a logical necessity, strongly suggest the pluralistic hypothesis. The part played by the subject in experience is not a purely passive one. We find that we are able to in- terfere in the course of events, and, within limits, to guide the latter to the fulfilment of our ends. The reali- zation of this ability is the basis of the notion we form of efficiency, and in it the root of the concept of caus- ality is grounded. The concrete meaning of causality 6o Scientific Method in Philosophy for us is therefore the efficient determination of one thing by another. This relation of efficient determina- tion is one-directioned ; it is not reciprocal. This follows from the fact of the one-directionality of time in actual experience. No doubt if we formulate symbolically this sequence of cause and effect, there is /(^^zVi^/ dependence of one on the other. Such a logical dependence, how- ever, is descriptive, and does not alter the fact that in actual experience our activity determines its con- sequences in an entirely different sense from that in which it is determined by them. In the course of his development, man comes to apply this idea of efficient causality to the sequences which he observes in the object of experience. The question then arises as to whether this application is valid. If we take the sequences simply as they stand, it is certainly not valid. We can only say that certain sequences do occur, and that we are able to formulate propositions in virtue of which the occurrence of some events can be inferred from the occurrence of others. From this purely scientific standpoint, causality is merely a logical and descriptive dependence of one event upon another. We may, however, wish to go beyond the mere existence of the sequences in an endeavour to find some satisfactory explanation of their existence as sequences. We know that some efficient individuals exist, and we also know that some of the sequences observed are initiated by the activity of these individ- uals. Hence we take as an hypothesis the proposition that all sequences have their ground in the activity of efficient individuals. This hypothesis is not logically proven, but it covers the facts by explaining them in Summary and Conclusion 6 1 terms of entities whose nature we can realize. It is therefore justifiable. It is necessary for the purposes of reasoning to formulate our ideas in terms of sequences of sense-data. But the artificiality of the latter is brought out by the consideration of the problem of continuity. In fact, the problem itself is due to this artificiality. That funda- mental characteristic of the object of experience which is commonly termed 'continuity,' is really unity. This consideration alone is sufficient to show that any theory which purports to give a final answer to any of the problems of the Universe in terms of such things as sense-data, may be ruled out at once. On the other hand, the unity of the object implies the unity of the subject, for this is its ground. It thus emphasizes the fact that selves are single entities that may be taken as true units in terms of which to express our explana- tion of the objective facts of existence. This idea of the self as a unity which, in its com- pleteness, transcends space and time, though for most purposes we conceive it as developing in space and time, is the conceptual representation of something realized concretely in actual experience, namely, the persistence of our identity through change and develop- ment. This reconciliation in the self of the principles of permanence and change provides us with a concrete example of that which we endeavour to conceive when we talk of 'substances.' It is impossible to formulate the reconciliation adequately in words, but it is there, and we realize its existence and its nature. We cannot rest content with regarding the object of experience as mere change (whatever that may be) based on no 62 Scientific Method in Philosophy elements of permanence, so that we come to look upon experience as interaction between self and other selves, following the pluralistic hypothesis. Accordingly, the attributes of these selves or substances are their modes of activity. In many cases this activity seems to lack spontaneity and to conform more or less completely to general laws, being due, as we suppose, to selves of extremely in- ferior mentality, and so, for the most part, the slaves of habit in their reactions. In many other cases the activity is only completely explicable with reference to the end which it achieves. Possibly we might be able to describe the activity completely in terms of the ordinary objective conceptions of physical science. This alone, however, leaves us far from satisfied. We can no more be content with it than we could be content with a mere description of the acts of other people accompanied by no statement nor understanding of their reasons for those acts. But if we regard all activity as being due to purposive individuals, we not only observe and describe the activity but we understand it. It acquires meaning, where before it was meaningless. The new scientific method is, then, in its own field and for its own purposes, a most powerful weapon of research. For the ends it has in view, the ignoration of the subject of experience is justifiable; but this only so long as we remember that the results obtained must not be regarded as giving a fully adequate account of things, even on the objective side of experience alone, but simply an account which, in its proper application, is the most satisfactory that can be obtained, owing to the limitations of the conceptual standpoint. The Summary and Conclusion 63 ignoratlon of the subject has, however, the important consequence that the results obtained by the scientific method may not vaHdly be used to criticize such an hypothesis as pluralism, for they stand on an altogether different ground. On the one hand we are investigating the logical form of facts, on the other we are out to explain the facts, and unless in doing so we describe the facts wrongly, we cannot lay ourselves open to criticism of the kind indicated. The type of result afforded by the scientific method leaves most of us unsatisfied. We wish to go further than mere description. The pluralistic hypothesis is an endeavour to satisfy this wish. It attempts to put every- thing in terms of things whose nature we actually realize, and which may therefore be simply indicated without the necessity of formal conceptual specification. This is all the more important because such a specification can never adequately comprehend the object to which it refers. Pluralism is, of course, an hypothesis, and therefore subject to the limitations of hypothesis in 'general, but it is based on no assumptions save in so far as it makes use of the logical canons of reasoning, if those can be called 'assumptions.' The assertion of the existence of the self is not an assumption ; and although we have referred to the assertion of the existence of other people as being an assumption, it is not so, strictly speaking, but rather the first step in the application of the pluralistic hypothesis to the explana- tion of the facts of experience. So far as we are able to explain the facts by it, pluralism is therefore an eminently satisfactory hypo- thesis; for, while it avoids the introduction of unknowns. 64 Scientific Method in Philosophy it brings home to us the nature of existence in general in an entirely unique way. We have the assurance that where it is successfully applied, the result will be, not merely to shift the problem back a step, thereby creating a new problem of the same type, but to provide a final explanation — an explanation which is capable of fully satisfying such beings as ourselves, in the search for the true nature and meaning of reality. II ON CERTAIN CRITICISMS OF PLURALISM I. Introduction It is incumbent on anyone who attempts to establish and develop a pluralistic view of the universe, to consider, and, if possible, to meet certain vital criticisms which have been urofed aofainst such a view. The answers to these criticisms must be prefaced by a brief indication of the standpoint from which they are approached. The present writer regards a spiritualistic pluralism (essentially such, for example, as that maintained by Dr James Ward) as the most satisfactory hypothesis on which to base a system of philosophy. It is satis- factory, in the first place, on account of the fundamental conceptions from which it starts. These are perfectly definite and easily realized. Secondly, it affords a most promising method of attacking and of partially or completely solving some of the outstanding problems of philosophy. In the course of the development of this hypothesis, however, it becomes clear that alone it is incomplete. This is to be expected, for the history of philosophy shows that no system can hope to approach within measurable distance of its object which lays undue stress on either of the dual aspects of the universe (its oneness and its manyness) to the neglect or exclusion of the other. R. S. P. C 66 On Certain Criticis^ns of Phiralisin We find, accordingly, that criticisms of pluralism fall mainly into two classes, those which demonstrate its incompleteness as a final answer to the questions which it seeks to resolve, and those which are aimed at supposed flaws radically inherent in the hypothesis itself As has been indicated, the former may be regarded as justified, but the latter call for an answer, and it is with certain of them that we are here concerned. Of the great philosophic systems of the past, the Monadology of Leibniz is perhaps the most remarkable for the logical skill with which it is sustained, and for the keen insight manifested in the fundamental principles on which it is based. From it all modern pluralisms derive their central theme. But two centuries of criticism have ensured the evolution of systems in which the more prominent weaknesses of the original monad- ology find no place. These later systems drew inspiration afresh from the great biological advances of the last century, advances made in the light of the doctrine of the evolution of species, a doctrine already foreshadowed in Leibniz' celebrated Principle of Continuity^ Yet there remained in pluralism certain vulnerable points which its opponents were not slow to attack. With all the criticisms thus put forth it is both impossible and unnecessary to deal at length. The most important of them are to be found in the writings of two men : Prof. Pringle-Pattison" and Dr Bosanquet^ If the ' The doctrine of pre-established harmony shows, however, that evolution, as we now understand it, did not enter into Leibniz' con- ception of the universe. ^ In The Idea of God in the Light of Recent Philosophy. ^ In The Principle of Individuality a fid Value. hitroduction 67 objections there urged can be successfully countered, the chief difficulties which block the path of the modern pluralist (not necessarily as regards philosophy in general, but as regards pluralism in particular) will be swept away. Accordingly, it is with the criticisms put forward by Prof. Pringle-Pattison and by Dr Bosanquet that we are called upon to deal. II. Externality For the pluralist, the environment of the self or subject of experience consists in other selves or subjects whose mentality differs from his only in degree. This belief is attacked by Dr Bosanquet in a criticism which may be summed up essentially somewhat as follows : "[Selves] as inward centres in the popular sense [cannot] form the circumferences for each other'," and again, "Even if there were, de facto, a psychical some- thing underlying matter, yet it is only as definite exter- nality that it plays a part in our life. We have no use for it as inwardness"." Now the true implication of these sentences is by no means evident if we inspect them as they stand. The spatial metaphor involved in the use of such words as "centre," "circumference," "inwardness," "externality," tends rather to obscure the issue, though the introduction of that metaphor may be very convenient and to a certain extent necessary. But what does this distinction between "inwardness " and "externality" really imply ? ^ See, e.g., op. cit., pp. 75 ff. - Ibid., p. 194, note. These quotations summarize the idea in- volved and explained at length. 5—2 68 On Certain Criticisms of Plnralism Evidently "inwardness" is something which essentially characterizes the individual subject, at least for that subject, whereas " externality " is something which characterizes (for him) the not-self. Hence the distinc- tion between "inward" and "external" refers ultimately to the fundamental distinction within each individual experience of subject from object. Consequently, if the pluralist asserts that the object of experience of one subject consists of other subjects, Dr Bosanquet's criticism becomes in effect, " How can a subject of experience be, in any circumstances, an object of experience .'*" In this form the criticism is justified, and the pluralist is wrong 2/" he asserts that to any subject other subjects are presented as objects of experience. Before con- sidering the latter point, however, it should be noticed that in any case the criticism only applies to pluralism incidentally. At the root of it is the fact that no existent entity can be an object of experience. No entity other than myself can be given to me as an object of knowledge in such a way that I realize what it is in its actual essence\ We cannot in experience know anything else as it really is in itself. What, then, of the sense-data which form for each individual his object of experience ? They are objects of acquaintance-knowledge. Are we to say that they do not exist ? Strictly, it is neither true nor false to say that they exist. It is meaningless. There is no significant sense in which existence can be asserted of ^ I do not mean to imply here that even the self is given as an object of immediate knowledge in experience. This point is dealt with more fully in the previous essay, Section V. Externality 69 the immediate data of perception. There they are, and that is all that can be said of the matter. Accordingly we must regard the object of experience not as one or more existent entities, but as the " appearance " to the subject of existent entities other than himself. This fact of "appearance" or "presentation," being ultimate in nature, defies satisfactory definition. It might be provisionally indicated somewhat as follows : Given a percipient subject and certain other existent entities, under suitable conditions, of which the existence of these other entities is the most necessary and important, the given subject will perceive an object which may be defined as the "appearance" to him of the other entities. It is important to notice that this "appearance" is neither the given subject nor the other entities, though its being is dependent on the existence both of the subject and of the other entities^ Prof. Pringle-Pattison' also makes a brief reference to the point under consideration. He remarks that " internality is impossible without externality." This, as we have seen, is equivalent to saying that a subject of experience is inconceivable apart from a presented object of experience. But the latter is simply the appearance to the subject of other existent entities. It is not itself to be classed as an existent entity, though it has being in the sense that it is there. A subject, however, to whom no appearance is presented is just as inconceivable as an appearance presented to nobody. ^ For a fuller treatment of this point see the next essay, end of Section III, and also the essay on " The Relation of Mind and Body," end of Section II. - Op. cit., pp. i78ff. yo On Certain Criticisms of Pluralism It follows, then, that Dr Bosanquet's criticism does not apply in any special way to pluralism, but is really an expression of the fact that an existent entity cannot be an object of knowledge. In particular, an experienc- ing subject cannot be an object of knowledge. But pluralism is in no way bound to assert this impossibility. For pluralism, the living experience of the subject con- sists actually in his interaction with other subjects. This interaction is manifested in the ever-increasing differen- tiation of a presented indivisible whole or object of experience, namely, the appearance to the subject of other subjects. We are not acquainted in sense-experi- ence with other individuals in their actuality. Selves cannot be reduced to sense-data. The latter are but what we have termed the "appearance" to us of other selves. We may conclude our reply to this type of criticism by briefly considering another quotation from Dr Bo- sanquet. In pan-psychism, he asks, "what becomes of the material incidents of our life } . . . Is it not obvious that our relation to these things is essential to finite being, and that if they are in addition subjective psychical centres their subjective psychical quality is one which so far as realized would destroy their function and character for us^ } " Now the nerve of this criticism is destroyed, as before, when it is realized that for a given subject the object of experience does not consist in a number of other "subjective psychical centres," but in the appear- ance to the given subject of these other subjects. Moreover, the function of material incidents in our life consists in the determination and limitation of our ^ Oj>. «'/., Lect. X, p. 363. Externality 7 1 purposive activity. It is simply the manifestation of our interaction with other subjects. In fact, it is here that the fundamental ambiguity of Dr Bosanquet's term "inwardness" as a characterization of subjective centres becomes completely evident. For the activity of the subject is essentially "outgoing" as it were. It is not directed in upon itself (if that could have any definite meaning), but out towards others. How, then, is it possible that the development of this psychical quality can destroy the function of the subject with regard to other subjects ? The growth of experience, in the pluralistic view, does not and cannot consist in a gradual withdrawal into itself of the subject, culminating in a complete isolation, but in continuous interaction with other subjects which, so far from leading to individual isolation, aims rather at mutual co-operation in ensuring the interests of the society as a whole. III. Consciousness Dr Bosanquet's conception of consciousness is in entire conflict with the position which pluralism takes up. But his view is largely vitiated by the fact that he adopts on this point an attitude which appears to tend very strongly to that Cartesian dualism of mind and matter, which for so long clogged the progress of philosophic thought. This tendency is particularly evident in his treatment of the relation of body and mind. The pluralist, on the other hand, recognizes that the fundamental fact from which the start must be made, is not a dualism of matter and mind, but the unity of the individual experience, which comprises a 72 Oil Certain Criticisms of Pluralism duality of subject and object. For the pluralist "mind" is a generic term denoting the class of subjects of experience. According to Dr Bosanquet "organic regulation is natural and immanent, but independent of conscious- ness'." Consciousness is a " perfection " granted by the Absolute in certain circumstances^ Such statements imply that matter is given as prior, while mind only supervenes at a certain stage of the development of matter. This seems to approach perilously near to the epiphenomenal view. Moreover, if we grant with Dr Bosanquet that organic regulation is " natural and immanent," what evidence have we that it is "indepen- dent of consciousness " .^ Apparently the reference here is to the fact that the behaviour of an organism (especially of a lower type) consists largely in reflex action. The question is then whether the establishment of reflex action presupposes mind or not^ Now we have an abundance of examples of such a presupposition — a simple case being a man learning to ride a bicycle. In fact the formation of habits is a fundamental character- istic of mind. On the other hand, there are no cases in which we observe the establishment of a reflex action where we can infallibly assert the absence of mind. It is the essence of the pluralistic position to recognize that the start must be made from individual experience, ^ Op. cii., Lect. V, p. 195. ^ Ibid., p. 189. ^ Of course it is a well-known fact that established reflexes occur without the intervention of the dominant consciousness of the organ- ism, but it by no means follows that the latter played no part in the original establishment of the reflex, nor that, even when established, the reflex is independent of any consciousness. On all these points see also J. Ward, The Realm of Ends, 2nd ed., pp. 462 ff. Conscioicsness 73 which impHes mind. It is the task of the pluraHst to interpret matter from this standpoint. On the other hand, if we start from matter, how can we interpret mind ? There is nothing in what Dr Bosanquet says on the subject which provides a satisfactory answer to that question. But from the standpoint of mind there is no such difficulty in interpreting organisms, at least. The striking feature of an organism is the fact that it exhibits "behaviour" analogous in every way to our own. Hence, what the subject distinguishes within its objective experience as organisms are, for the pluralist, the appearance to the subject of other subjects differing from himself only in degree or in kind of mental development. Speaking again of consciousness, Dr Bosanquet says that "conscious process is meaning (or appreciation), not effect, of physical process^" — and in another place : " Mind is the meaning of externality, which under certain conditions concentrates in a new focus of meaning, which is a new finite mindV It is not easy to assipfn a definite sig-nificance to these assertions. In the first place "meaning" and "appreciation" are by no means synonymous terms. They apply respectively to the objective and the subjective aspects of the process which consists in the interpretation of an object by an individual subject. In other words, we regard the subject as "appreciating" the "meaning" of the object. It is difficult to see in what sense, if any, consciousness may be considered as " meaning." For the latter term implies both an object and a subject for whom the ^ Op. cit., pp. 196 ff., margin. " Ibid.^ App. II to Lect. V, p. 220. 74 On Certain Criticisn^is of Pluralism object has meaning. We cannot regard the subject as being a "meaning." If we attempt to do so, we are bound to imply a further subject\ and are thus led into a continuous regress. Moreover, Dr Bosanquet fails apparently to distinguish clearly between sensations and the mind of which they are the sensations. It is not clear whether the mind or the sensations constitute the meaning of physical process". But, at all events, we cannot suppose mind to be simply the "meaning" of something else. " Meaning," though it implies a subject, is not itself that subject. Nor does it help us to adopt the term "appreciation" instead. For the subject is not the appreciation, but the individual who appreciates. The conception of a mind as a "focus" of externality also appears to have no valid significance. As we have seen, the only legitimate meaning that can be given to the term "externality" is "the objective side of ex- perience." But we cannot possibly conceive the subject as consisting in the " concentration " of sense-data into a "focus." To use Dr Bosanquet's terminology, in- ternality can in no way be constructed out of externality. The term implies the fundamental distinction in ex- perience between subject and object. We might perhaps speak (very loosely) of the subject as concentrating externality, by his unifying activity, into a focus. But externality thus focussed would be the product of the subject's activity and not the subject himself. ^ Even here there is a difificulty. For, as we have seen, a subject cannot be an object of knowledge, and anything which has "mean- ing " for anybody must in some sense be an object of knowledge. 2 Op. cit, p. 197. The Evolution of Law 75 IV. The Evolution of Law In the type of pluralism advocated by Dr James Ward, the laws of inorganic matter, commonly called the " Laws of Nature," are regarded as having evolved in time, only reaching their present fixed and stable form after a long process of development. Prof. Pringle- Pattison raises objections to this view. According to him we cannot suppose the possibility of action without environment, nor can we conceive the interaction of monads, even in the beginning, apart from laws in accordance with which that interaction takes place\ And again : " A system of unvarying natural order is demanded, it may be pointed out, in the service of the higher conscious life itself, as the condition of reasonable action-. Now, in the first place, it may be admitted that action is impossible without environment. But pluralism does not deny this. The environment of a rnonad is constituted by the other monads, with which it interacts. And, coming to the further point, Prof Pringle-Pattison is evidently right in so far as he asserts that the monads must always have had some nature. But by the evolution of natural laws, the pluralist simply means that the laws of nature did not always exist in their present relatively fixed form. It must be remembered that such laws are not, as it were, imposed upon things from without, but are merely descriptions of the way in which things behave. Consequently, if the behaviour is modified, the descriptions or laws are correspondingly 1 Op. cit., pp. 183 fif. 2 ji,!^^ p J 8^^ 76 On Certain Criticisms of Pluralism modified also ; though in certain cases behaviour may tend to a comparatively fixed system of habitual reac- tions, in which cases we may speak of a fixed law. The attitude of pluralism on this point may, perhaps, be made clearer by an illustration. In the first place it must be noted that, for the pluralist, there is no absolute gap between organic and inorganic matter. Now if we survey the realm of organic matter, past and present, we find that whereas some species continue to develop into more and more complex types, others have, after a long period of development, eventually approached a stationary condition in which their actions have become practically entirely habitual and relatively fixed in nature. Inorganic matter may be regarded as an extreme form of such stationary species. Hence there is no difficulty in supposing that inorganic matter has evolved into its present condition, and it is in this process that the evolution of the so-called "laws" of matter consists. There is obviously no reason to suppose that a limit must be placed on the number of these laws. Hence we may consider that originally each monad, while dis- playing the general characteristics of mind in a low degree, was yet, in its particularity, a law unto itself Only as interaction proceeds is there a tendency for individuals en masse to behave in similar ways. This tendency proceeds from the characteristic, which must be present in some degree in each individual, of learning by experience. As to what Prof Pringle-Pattison says of the neces- sity for a system of unvarying law as the condition of reasonable action in higher conscious life, it certainly seems probable that the tendency of the individuals The Evolution of Law 77 composing inorganic matter to develop a system of habitual reactions has greatly aided the process of evolution of other individuals to higher and more com- plex types. Yet it must not be forgotten that each of us has to deal not only with material objects but also with persons. Although the behaviour of the latter does not admit of description to a degree of precision in any way comparable with such principles as the law of gravitation, for example, yet we do not find it impos- sible to live a rational social life on that account. In dealing with individuals whose behaviour is subject to continuous modification and development, the only necessary conditions of success are that the process of development should not be too rapid, and that we should have a knowledge at least of the general trend of that process. Such knowledge would itself be em- bodied in a law, but of a different type from those we consider in general under the conception of the evolution of law. For it would be the description of a dynamic process and not of a static form of behaviour. It is evident, then, that the notion of the laws of nature as evolving gradually into their present stable form is not a contradictory one. For the evolution of law means nothing more nor less than the gradual modification of behaviour. We have examples in plenty of such modifications, and we find that in many cases the process tends asymptotically, as it were, to a limit, and we have species, which, after developing through count- less ages, become relatively fixed. Relatively, we say, for there is no guarantee that even the laws of inorganic matter will, after the lapse of future vast periods of time, remain in their present form without sensible alteration. yS On Certain Criticisms of Plnralisjn V. The "Bare" Monad All mental life of which we appear to have clear evidence is associated in every case with an organism. The pluralist conceives the organism as a system of monads in association with a dominant monad, the latter constituting the self of which the organism is the body. But if we press the pluralistic hypothesis far enough, we seem bound to postulate, somewhere or somewhen, the existence of "bare" monads, i.e., monads unassociated with any body or organism. Prof. Pringle-Pattison points out objections to this view\ Leibniz endeavoured to avoid the difficulty by assuming that every monad was associated with an organism composed of relatively inferior monads. For him, a piece of inorganic matter was a mere collection of organisms. In this way he piled infinity on infinity. We cannot be satisfied with such an endless regress. Nor does it really clear away the obstacles in any very definite manner, for it is difficult to see how, in con- sidering the relations of organisms external to one another, we can entirely avoid the notion of the inter- action of bare monads. But, in any case, there seems to be no intrinsic difficulty in the conception of a bare monad. There is apparently no inevitable reason why that peculiar com- plex of presentations" which constitutes what we call "the body" should enter as an element in every experi- ence. A bare monad would simply be a subject from 1 Op. cit., p. 1 8 8. ^ Not only of sight and touch, but also that mass of organic sensations which constitutes what is called "general sensibility." The ''Bare'' Monad 79 whose object of experience this element was absent, and there is no way of showing that its absence is an impossibiHty. No doubt there is a difficulty of another kind, if we try to hark back to the monads as they originally were. For there is bound to be a difficulty here, but it lies, not in the notion of a bare monad, but in the inherent incompleteness of the pluralistic hypo- thesis. We are faced, in short, with the problem of Creation, which pluralism alone is powerless to solve. Yet one word of warning is necessary. Prof. Pringle- Pattison seems, in one place, to identify the bare monad w^ith what lies behind the atom, or whatever the ultimate physical particle may be\ This is quite unjustifiable. Physical objects, whether they be common-sense objects such as chairs and tables, or entities such as atoms and electrons, are conceptual constructions based on sense- experience, and therefore have a purely formal exist- ence". If the truth be told, the bare monad is not the real root of the trouble at all; the latter must be sought rather in the conception of interaction between the monads — and this applies just as much when the monads are members of one organism as when they are not. We need some concrete ground of this interaction, which shall serve as a principle of unification whereby the existence of selves forming a plurality, and yet entering into relations with one another, may be rendered 1 Op. cit, p. 180. ^ "This table" and "an atom" are alike capable of being ex- hibited as logical constructions of sense-data, though the latter is a more complex construction than the former. See B. Russell, Our Knowledge of the External World, Lects. Ill and IV. 8o On Cert am Criticisms of Pltiralisni intelligible \ Although the start must be made from a plurality, and although the pluralistic hypothesis will carry us a long way in the understanding of the world, we must take account at the latter end of that other aspect of the world — its unity. With the further con- sideration of this question we are not here concerned. Suffice it to say, as in the introduction above, that such limitations of pluralism as are implied in this matter may be freely admitted. VI. Summary and Conclusion It would appear, then, that the most important criticisms recently directed against pluralism fail of justification. We saw, in the first place, that there is no more difficulty in accounting on the pluralistic hypo- thesis for what Dr Bosanquet calls "externality," than on any other hypothesis, provided that we interpret that term correctly. It can only mean the object as dis- tinguished from the subject of experience. For pluralism, the object of experience does not consist of other subjects (as Dr Bosanquet's criticism implies), but of the appear- ance of these other subjects to the individual subject considered, where "appearance" is defined in some such way as we have indicated. These " appearances " cannot be said to exist, for no existent thing can in itself be an object of knowledge, though they have being in the sense that they are there. Secondly, Dr Bosanquet's account of consciousness does not agree with the facts. We have no reason / Cf. below, pp. 203 f., and p. 250. Summary and Conclusion 8 1 whatever to assert that organic regulation is indepen- dent of all and every kind of consciousness. On the contrary, wherever we can observe the formation of a habit culminating in reflex action, it is associated with mind. Thus, whereas we have instances of reflex action presupposing the existence of mind, we have no instances of such action where mind can be certainly asserted to be absent. We cannot construe consciousness merely as the meaning of externality. Such an interpretation is in- herently contradictory. For, using the term legitimately, we speak of the " meaning " of an object y0\ The fundamental fact from which we start must not be adualism of "mind" and "matter" such as was conceived by Descartes, for " mind " and " matter " are both secondary conceptions, but this unity of individual experience comprising the duality of subject and object standing in the relation of presentation /. Of the terms ^ Cf. James Ward, Psychological Principles, p. 371. The Nature of Sense-Experience 93 thus related, doubt is sometimes cast upon the existence o{ S. With this point the writer has dealt elsewhere \ Suffice it to say that doubt of the existence of the subject is without significance. In any case, however, since our problem deals with differences in sense-data, and since in perceptual experience O is composed of sense- data, the question hinges on the nature oi O. In particular, we must enquire whether O is conditioned in any way by the fact that it stands in the relation/ to 6", and, if so, in what way. The nature of the problem underlying the difficulty raised by the Weber- Fechner law now begins to be apparent. It is, in fact, no less than the crux of the issue between realism and spiritualism. The realists contend that O exists quite independently of S, and that its nature is not conditioned in any way by the fact that it stands in the relation/ to 5. Its entry into that relation is, so to speak, "accidental." It exists independently of that relation, and, upon entering into it, the characteristics it already possesses are not altered in any way". On this view there is evidently no difficulty, at any rate prima facie, in supposing that differences of sense-data exist even though unobserved by the subject. For the spiritualist, on the other hand, the esse of the object of experience is necessarily percipt\ For him there cannot be a thing of such a nature as (9, which is yet not perceived by anybody. Hence, if ^ In " Scientific Method in Philosophy and the Foundations of Plurahsm," Section IV. ^ See, e.g.^ R. B. Perry vcvPresent Philosophical Tetidencies, Ch. XIII. ^ It is not, of course, necessary for spirituaUsm to assume that the being of O depends only on its being perceived, but merely that the latter is 07ie necessary condition of its being. 94 The IVeber-Fechner Law spiritualism is to be thoroughgoing, it cannot counten- ance the assertion that there may be differences of sense-data which are unperceived. The reaHst conception of experience is based on the theory of the externaHty of relations. According to that theory, the being of an object is quite independent of any relations in which it may happen to stand to other objects. Although a relation may add something to the nature of its terms, the characteristics already possessed by the latter are entirely unaltered by it\ Hence the being of O is regarded as quite independent of the fact that it stands in the relation p to S. The only addition which this relation brings is the awareness of the subject. In examining this view it is important in the first place to recognize the difference between relations distinguished within the object of experience, that is, relations between the sense-data composing the latter, and the relation of presentation in which that object stands as a whole to the subject. The relation p is evidently unique". One consideration alone makes this sufficiently clear, namely the fact that 5" cannot be ' See, e.g.^ R. B. Perry, op. cit., Ch. XIII, p. 319. - This is a very important point. For example, R. B. Perry (in his essay in The New Realisjii) enumerates types of dependence, of which the presentational relation is not one, and then has no diffi- culty in showing that the latter is not included under any of the types enumerated. Obviously not, since it was excluded by definition, for, owing to its uniqueness, it cannot be subsumed under any other type of relation. All Mr Perry's types refer to relations between objects^ whereas we have here a relation between an object and a subject. Mr Perry's case could only hold if he gave a general criterion which all relations of dependence must satisfy, and then showed that the presentational relation did not satisfy it. The Nature of Sense- Experience 95 considered to exist out of the relation /. For there could not be a subject to whom nothing was presented, in other words, a subject without experience. Hence the existence of S is conditioned thus far by its standing as one term in the relation /. This fact gives us sufficient warrant for refusing to assume, without the most searching enquiry, that the being of O is not conditioned in any way by the relation/. Seeing that such an entity as a subject of experience to whom nothing is presented cannot exist, it is conceivable, to say the least, that such an entity as an object of ex- perience that is presented to nobody cannot exist either. This is not meant to suggest, of course, that it is impossible for miything to exist which neither knows nor is known, but that if such things should exist, they cannot be considered to be in any way akin either to subjects or to sense-data. Mr Russell states that the existence of a sense-datum cannot be logically dependent upon that of the subject, for the only way in which the existence of A can be logically dependent upon the existence of B, is when B is part of A^. But the matter does not appear to be quite so simple in the unique case of subject and object standing in the presentational relation. For let us denote the individual experience SpO by E. Sup- posing it turns out, as we have suggested, that such an entity as O can by its very nature only exist as part of an experience. In that case E would be logically dependent upon O, for (9 is a part of E, but the existence of O would also be logically dependent upon ' In " The Relation of Sense-data to Physics," Section III, which originally appeared in No. 4 of the 1914 volume of Scientia. 96 The Weber-Fechner Law the existence of E. Moreover, as we have seen, the existence of 5 is logically dependent on that of E and vice versa. In such a case then, it would surely not be straining the point to say that the existence of O is logically dependent upon that of S. Leaving aside, however, the question of logical dependence, there is yet another difficulty to face if we refuse to admit that the existence of O is conditioned in any way by the relation /. For if we apply the Weber-Fechner law to the limiting case of sense-data of very small intensity, we find that a finite stimulus is needed before any corresponding sense-datum is perceived at all. From this, as we have seen, it would be commonly concluded that the sense-datum does exist, although unperceived. In other words, we should have to grant that there must be parts of O which are not perceived at all. But if this is so, what exactly is the difference between such parts of O and what Mr Russell calls "sensibilia^" (that is, entities similar to sense- data in every way except that they do not necessarily form part of an object of experience), and why are we justified in including the ones within O while excluding the others'" from O ? And if we are not justified, how can we escape the conclusion that the being of O must depend in part upon its being perceived .-* Has the phrase " an unperceived sense-datum " any real meaning? By a sense-datum we must obviously mean something that is "given," and "given" in this sense means "presented to a subject." Can we signifi- cantly say that an object is presented yet not perceived .'* ^ Op. cit., Section III. ^ I.e.^ " sensibilia" not forming a part of an object of experience. The Nature of Sense-Experience 97 It seems doubtful whether we can, when we take into account the pecuHar nature of presentation, but the point need not be insisted on. It is noteworthy, how- ever, that Mr Russell asserts that sense-data "probably never persist unchanged after ceasing to be data\" Yet he maintains the existence of unperceived sensibilia which " resemble " sense-data in every way except that they are not perceived. But if two things resemble one another, we must be able to state that they possess a common characteristic. Now what is this characteristic common to sense-data and unperceived sensibilia alike ? Is such a thing possible at all ? For the only charac- teristic common to all sense-data is that of being pre- sented to a subject, and this is here ruled out. Other considerations make it evident that the object of experience depends in part upon the subject. In the first place, the order of our sense-data is to a certain extent determined by our own activity manifested in movements. To this it would probably be replied that we have only the right to assert that certain motor presentations are followed by changes in sensory pre- sentations. This is true, but a motor presentation follows the idea of the movement, and the latter is con- sequent only upon the active attention of the subject. Psychological analysis thus reveals that the order of sense-data is determined in part by the activity of the subject. Yet this is not all. Not only the order, but also the nature of the sense-data which make up the object depends partly on the activity of the subject. The growth of perceptual experience consists essentially in an ever-increasing differentiation of the object by ^ Op. cit., Section IV. R. s. p. 7 98 The Weber- Fechjter Law the subject \ The realists would contend that these differentiations were present in the object all the time, but were only gradually observed by the subject. This statement really brings us to the heart of the whole matter. For the realists, if they are to maintain their view, must evidently look upon the object as made up of a number of mutually external though related units termed sense-data. But this is only a necessary and very close approximation to the facts. The object of experience is actually an indivisible whole, albeit differ- entiatedl If the realist contention were true, that at a low level of experience the object is characterized by the fact that it comprises many elements not as yet observed, our experience at such a level would be full of " gaps," as it were. This, however, is not the case. The difference between a high and a low level of perceptual experience is simply the difference between a more or a less differ- entiated unity. But in each case this unity is a given indivisible whole. In each case the subject perceives the whole object, and the latter is only what is per- ceived by the subject. The fact that the differentiation is much greater in the one case than in the other is due in large measure to the activity of the subject; due also, as we shall see, to other factors ; but these other factors are not sense-data which form part of the object from the beginning though unperceived by the subject. ^ Hence the so-called "subjectivity" of the object is not merely physiological as Mr Russell maintains. ["The Relation of Sense-data to Physics," Section III.] - See the essays on "Scientific Method in Philosophy, etc.," Section VI, and "Immortality," Section IV. The Nature of Sense-Experience 99 Although the object is determined by the subject to the extent we have seen, it is by no means entirely so determined. Evidently the nature of our sense-data is largely independent of us. In fact we may say roughly that whereas our activity plays a large part in deter- mining the form of our experience, the matter of that experience is independent of us to a very great extent. But although its nature^ maybe thus independent, it does not follow that its being is also independent. There still remains the question as to the most satisfactory interpre- tation of the facts. This question we shall now consider. The realist speaks of the " appearances " presented by "things," but on his theory the appearances are the ultimate entities, and the things are simply logical con- structions. On grounds of continuity and resemblance we correlate certain aspects perceived by ourselves and others as being appearances of one thing. The "thing " is simply the class of these correlated aspects". We may grant this view so far as the term " thing " refers to such common-sense objects as tables, chairs, etc. ; but there are difficulties in the way of accepting the view that the appearances or aspects are ultimate entities and independent of their presentation to a subject. Mr Russell states that continuity makes it not unreason- able to suppose that things present some appearance at places where there is no percipient subject^ But it ^ This word of course implies "form," but the distinction between "matter" and "form" is not meant to be considered absolute here. The distinction between "nature" and "being" is in this case not very clear for reasons which will be fully considered later. - B. Russell, "The Relation of Sense-data to Physics," Section V. ^ Op. cit, Section III. 7—2 loo The Weber-Fechner Law is difiBcult to see how an appearance can be presented to nobody, unless " appearance " is defined in some very unusual way. Mr Russell's definition does not avoid the difficulty. Such correlated sensibilia as we have mentioned, he defines as '' appearances of one thing." Such a definition evidently begs the whole question of the nature of sense-data and their dependence on a percipient. Moreover the difficulties become still more insistent if we try to deal with the question of change. If we define the thing as the class of all its appearances, the thing will change if one of its appearances changes; for instance, if I close my eyes when I am looking at it. But in this case the change is in my eyes, and the realist would maintain that there still existed appear- ances of the thing nearer to it than my eyes, which were unchanged. Mr Russell deals with the point as follows: "We may say as a matter of definition, that a thing changes when however near to the thing an appearance of it may be, there are changes in appearances as near as, or still nearer to, the thing. On the other hand we shall say that the change is in some other thing if all appearances of the thing which are at not more than a certain distance from the thing remain un- changed, while only comparatively distant appearances of the thing are altered \" Now throughout this state- ment the same " thing " is referred to. But how exactly is it defined? Not as the class of all its appearances, for in that case it would not remain unchanged in the second instance given by Mr Russell. It would there- fore have to be defined as the class of all its appearances at not more than a certain distance from it. But we ^ Op. cit., end of Section VIII. The Nature of Sense- Experience loi have still to say what this distance is. Evidently in the case mentioned it would have to be less than the distance of the observer. Hence we cannot define the class of appearances which is to be considered to constitute one thing, without reference to a percipient subject, if any percipient subjects exist. The case is evidently still more complicated when there is a number of percipient subjects at varying distances from the thing. Supposing, however, while retaining the realist's terminology, we reverse his hypothesis by regarding the object of experience as the appearance to the sub- ject of entities other than the subject. These entities should be by no means identified with the objects of common sense. The latter may still be regarded as logical constructions of sense-data actually perceived by different people at different times and at different places. Passing over for a moment the actual nature of the entities which appear, let us consider what the term "appearance" really means. It should first of all be noted that an existent being cannot be an immediate datum \ It is impossible for anybody other than such a being to know it in its actu- ality and completeness. If it is not another being, that is, if it is the knowing subject himself, then it evidently cannot be an object of knowledge for that subject. In this case the fact can only be termed " realization" Returning, however, to the case of the relation between the subject and other beings, we may proceed some- what as follows: " Given a subject 5" and certain other entities, under appropriate conditions, of which the ex- ^ Cf. the previous essay. Section II, and the essay on "The Rela- tion of Mind and Body," end of Section II. I02 The Weber-Fechner Law istence of these other entities is the most necessary and important, 5" will perceive an object O which may be defined as the ' appearance ' of the other entities to 5," It is obvious that appearance in this sense implies the existence of at least two entities. It must be the appear- ance ^something to something else. The appearance of Axo El's, neither A nor B, but it has no being apart from the existence oi A and B. Moreover, it is possible that A may change and yet appear the same to B. If it appears the same, the appearance is the same ; that is, if the object of experience is perceived as the same, it is the same. In this way, the philosophical questions raised by the Weber-Fechner law maybe answered. I n the first place, there is no thing similar in nature to an object of experi- ence which is not perceived by any subject, although the being of an object of experience is not entirely depen- dent on the existence of the percipient subject. And in the second place, if the object of experience is per- ceived by the subject as unchanged, it is unchanged. It must not be supposed that the term "appearance" is in any way connected with " illusion." " Illusion " is only a significant term when a judgment has been passed on the appearance. Appearances, being immediate data, cannot strictly be said to exist, as we have seen. In fact, such words as "existence," "real," and "unreal," cannot be significantly applied to them at all. Yet they have some sort of being. We can say of them that they are there, and that ends the matter. It is important to notice that Dr Whitehead and Mr RusselP have been led from logical considerations to the same conclusions as those just stated with regard ^ Principia Mathematica, Vol. I, 14. The Nature of Sense-Experience 103 to the applicability of such words as "existence" to immediate data. Yet Mr Russell says that we may legitimately enquire as to the existence or reality of other sensibilia inferred from such data\ But according to him, sensibilia which are unperceived are essentially similar to sense-data, differing only in the fact of not being perceived. It is difficult to see how terms which are quite insignificant in application to an entity standing in a certain relation (the terms having no reference to that relation), may yet be significant when it does not happen to stand in that relation ; and this more especially if we should by any chance adopt Mr Russell's view of the externality of relations I There remains for philosophy the further question as to the probable nature of the entities which appear. We are not here concerned with a detailed attempt to answer that question, but the possibilities may be briefly indicated and weighed. There are four such possibilities. The entities may be wholly material, or wholly spiritual; they may be partly material and partly spiritual, or they may be neither. If by " material " we mean to indicate the molecules, atoms, and electrons of physics, it may be said at once that we have no reason whatever to suppose that they ^ In "The Relation of Sense-data to Physics," Section XII. ^ It is true that in the second passage quoted, Mr Russell qualifies his assertion by saying that "^ is my present sense-datum" and "my present sense-datum exists" are both significant, though ".t exists" is not. Clearly "existence" is here used in a special technical sense. But the point is that that very immediacy, which renders such a pro- position as ^^x exists" insignificant, is so essential a characteristic of sense-data, that apart from it they would not be sense-data at all, nor in any way akin to sense-data. I04 The Weber-Fechner Law actually exist. They may all be exhibited as logical functions of sense-data\ and any attempt to assign to them in addition a concrete existence is both gratuitous and unnecessary. With regard to the possibility that the beings which appear to us are neither material nor spiritual, it need only be remarked that such a supposition should be avoided if we can do without it, in accordance with the principle of Occam's razor. We are now left only with the spiritualistic hypo- thesis, namely that the beings whose appearance we perceive are other subjects of experience. This hypo- thesis constitutes pluralism. Certainly some subjects exist. Strictly speaking each of us knows that one sub- ject exists; but, to avoid a barren solipsism, we may make the further assumption of the existence of other people, in common with most schools of thought, in- cluding realism. Selves or subjects of experience cannot be resolved into sense-data, therefore let us suppose that our sense- data are the appearance of other selves to us, thus securing a philosophical economy. On this hypothesis, then, the object of experience is the appearance to the subject of other subjects. The individual experience consists in the interaction of the subject with other subjects, manifested in the increasing differentiation of the object. The object of experience is a presented whole, one and indivisible. The other subjects are many, but their appearance to a given subject is one. As it is not intended, however, to indicate here any- ' See, e.g.^ B. Russell's Otir Knowledge of the External World, Lectures III and IV. Summary and Conclusion 105 thing more than the relative probabiHties of the possible hypotheses to which our interpretation of the Weber- Fechner law gives rise, we may now conclude by summarizing the results already arrived at. IV. Summary and Conclusion Our enquiry commenced with an examination of certain conclusions commonly drawn from the experi- mental facts on which the Weber- Fechner law is based, and of the reasons generally assigned for those con- clusions. The latter assert that unobserved differences of sense-data must sometimes exist even when the subject is on the lookout for such differences. A par- ticular case of great importance occurs in connection with sense-data of a comparatively low degree of inten- sity. In such cases it is asserted that sense-data must sometimes exist which are not perceived at all. Analysis showed that these assertions must be considered as not proven by the experimental facts to which appeal is made. For no allowance is granted for the possibility of the existence of other factors than those generally taken account of in determining the result. Such possible factors, for example, are the par- ticular temporal order of the stimuli, and the peculiar nature of the presentational relation of sense-data to the subject. The latter of these appeared in the sequel as by far the more important. In any case, however, should such factors be effective, their action may be in the nature of a compensation for the effects of the other factors. Taking account of this possibility, it was seen that the experimental facts, which are generally con- io6 The Weber- Fechier Law sidered with reference to the latter factors alone, while compatible with the conclusions commonly drawn, are equally compatible with the opposite conclusions. In accordance with the latter it would be asserted that no objects of a nature essentially akin to sense-data can exist unperceived, and that sense-data which are per- ceived as the same, are the same. Hence it became evident that further investigation was necessary before a definite statement on the matter could be made. In any philosophical enquiry of this nature, we must start, not from a dualism of mind and matter, but from the unity of individual experience comprising the duality of subject and object standing in a relation of presenta- tion. The effects of the traditional dualism of matter and mind still linger in philosophical thought, and are manifested, for example, in the vague use of such terms as " mental." This word is used in at least two distinct ways. In the first place, it is used of such things as dreams or hallucinations which are supposed to exist only "in the mind" of the subject, whatever that may mean. The only way in which we could modify this usage so as to give a precise meaning to the term " mental," would be by defining it as applying to those portions of the object of experience which only exist in so far as they are perceived by the subject. With this meaning of the term not only dreams and similar things but also ordinary sense-data might be mental. Such a use of the word is, however, highly unsatisfactory. In the second place the term "mental" is sometimes supposed to refer to anything which is part of the sub- ject — as if a subject could have "parts." Probably what are here indicated, are such things as thinking, willing, Sttminary and Conchtsion 107 and desiring. These are essentially connected with the feeling and activity of the subject, and it is here that we have the strictly correct use of the word. " Mental " should only be used to characterize the feeling and activity of the subject. Mr Russell supposes that sense-data are not mental but physical, that is, part of the subject-matter of physics^ Granted this, he appears to assume that it follows that the being of sense-data is not dependent in any way upon their being perceived. Such a conclusion, how- ever, is by no means necessary. We may grant that sense-data are not mental according to our definition of the term, and also that ultimately physics deals with sense-data. But this does not shed any light on the question of their dependence upon perception. It is sometimes urged in support of the view that they are independent of perception, that physics can describe the object of experience and make verifiable predictions about it without any reference to the subject or to per- ception. But even if we grant this, it is quite irrelevant. One example is sufficient to show that this is the case. I can observe the positions and movements of the hands of my watch, and make true predictions as to their future positions, without any reference whatever to the mainspring. Yet the latter is the sme qua non of all that I have observed and inferred. It is evident that we continually describe things and make successful inferences about them, without any reference to the necessary conditions of their existence. To solve our problem we must endeavour to deter- mine what is the real nature of the object of experience, ^ "The Relation of Sense-data to Physics," Section III. io8 The IVeber-Fechner Law or, if we cannot obtain a definitive solution, to determine the hypothesis which fits the facts most satisfactorily. The real issue at stake is that between realism and spiritualism. If sense-data can exist unperceived, there is no reason why objects essentially akin to sense-data should not exist yet without forming part of an individual experience at all. In such a case the theory that reality is wholly spiritual goes by the board. On the other hand, if it turns out that objects such as sense-data are essentially characterized by presentation to a subject, realism, at all events in its modern form, cannot be true. Our first step in the analysis of the question was to comment on the theory of the externality of relations by calling attention to the importance of making the necessary distinction between relations holding within the object of experience, and that relation of which the latter is one of the terms, namely the presentational relation. The presentational relation is evidently unique. There cannot be a subject to whom nothing is presented ; that is, the subject is essentially characterized by always standing as one term of the relation p. Evidently, then, it is at least possible that the object may be essentially characterized by always standing as the other term of the relation p. The point demands further investi- gation. A realism of the type maintained by Mr Russell is faced with a dilemma. For if we accept the conclusion commonly drawn from the Weber-Fechner law it follows that there must be three classes of sensibilia. Those which do not form part of an object of experience; those which do, but are not perceived; and those which do. Summary aiid Conclusion 109 and are perceived. What, then, is the criterion of difference between members of the first two classes? What difference can there be? And if we admit that there is no difference, it follows that all sensibilia which form part of an object of experience must be perceived. On the other hand, if we disagree with the common conclusion drawn from the Weber- Fechner law, the same result follows at once. If we accept this result, we are under no necessity of asserting the existence of particu- lars essentially akin to sense-data, which yet form no part of an object of experience, nor is their existence even rendered probable. A similar dilemma arises in the case of small differ- ences in sense-data. For the only possible criterion by which we can certainly decide whether two sense-data differ or not, is the judgment of the percipient subject. According to this we should have to say that if two sense-data appear the same to the subject, they are the same. On the other hand, if no judgment is made, not only have we no means of deciding whether two sense- data perceived by the given subject differ or not, but we can obtain no information whatever about the sense- data. Evidently the object of experience is partly de- pendent upon the subject. The movements of attention and, in fact, the whole growth of experience exemplify this. On the other hand, it is no less evident (provided we reject solipsism) that the object is not wholly de- pendent on the subject. The activity of the latter plays a part which is largely formative. The further back we go in experience, the cruder does the object appear. Yet it is never entirely formless. We may therefore I lo The JVeber-Fechner Law conclude that whereas the form of the object of experi- ence is largely determined by the shaping activity of the subject, the matter thereof is no less determined by entities other than the subject. There still remains, however, the question as to the conditions upon which the being of the object depends. As we have seen, the view that particulars essentially resembling objects of experience may exist without forming part of an experience is beset with great difficulties. We therefore attempted to frame a more satisfactory hypothesis. Let us suppose that two entities A and B exist, A being a percipient subject. Then given certain conditions, of which the existence of B is the most necessary and important, A perceives a certain object. Evidently this object is not A. But neither is it B, seeing that an existent thing cannot be an immediate datum of perception, for clearly the thing in its actual essence cannot be an object of acquaintance to anything else. Hence we defined the object which A perceives as the " appearance " of B to A. For the reasons stated, it does not exist in the sense in which A and B exist. Yet it has being in the sense that it is there. On the other hand, its being is dependent on the existence of both A and B. We may therefore regard the object of experience as the appearance to the subject of entities other than the subject. The entities maybe many, but the appearance is an indivisible unity, being presented to one subject. On this view it would appear that realism and the traditional idealism are half truths. Realism is riorht o in asserting that the being of sense-data is not entirely dependent on their being perceived, but wrong in so Stmmiary and Conclusion 1 1 1 far as it asserts that they are a type of objects whose being is quite independent of perception. Idealism, on the other hand, is right in maintaining that if sense- data are to be at all, they must be perceived, but wrong in maintaining that their being perceived is the only condition of their being. The reconciliation is effected by regarding both the existence of a percipient subject and also the existence of some other entity or entities as necessary conditions of the being of sense-data. The latter are the appearance of something to something else. There remains for the philosopher the problem of the nature of these entities which appear to the subject. In selecting a working hypothesis we should be guided by two principles. In the first place the introduction of unnecessary entities whose real nature remains a mystery should be avoided ; and in the second place we should choose if possible beings whose nature we can satisfactorily realize. In accordance with the first principle we may rule out objects of common sense such as chairs and tables, for these are capable of being exhibited as functions of sense-data, and can lay no claim to concrete existence as objects apart from, and in addition to, sense-data. On the same count, physical entities such as atoms, molecules, electrons, etc., may also be ruled out. Physics certainly carries its con- structions one step further than ordinary common sense, but many of the objects with which it deals have already been shown to be simply logical functions of sense-data, and analogy leaves little doubt that all physical objects are of a similar nature. The second principle evidently points the way to 1 1 2 The Weber-Fechner Law the hypothesis of pluralistic spiritualism. According to that hypothesis the beings which appear to the subject are other subjects differing from himself only in degree of mental development. The individual experience thus consists in the interaction of the subject with other subjects. The object of experience is the appearance to him of the other subjects, and his interaction with them is manifested in the ever-increasing differentiation of the object. With the development of this hypothesis, the present writer has dealt elsewhere^ It is sufficient here to indicate it as a possibility, and to point out the facts we have mentioned as appearing to favour its inception. We have seen then, that an examination of the facts on which the Weber-Fechner law is based, and of the general nature of individual sense experience, leads to the following conclusions : Firstly, the reasons generally asserted for supposing that sense-data sometimes differ when perceived as the same, and that there must be sense-data which are not perceived at all, are not conclusive ; in fact they are quite compatible with the opposite view. Secondly, there cannot be such things as unperceived sense-data nor objects essentially re- sembling sense-data and yet not forming a part of any object of experience ; moreover, sense-data perceived as the same must actually be the same according to any applicable criterion and in any sense in which the word "same" is significant of sense-data. Thirdly, the diffi- culties can be resolved most satisfactorily if sense-data are considered to be the appearance to the subject of ^ In "Scientific Method in Philosophy and the Foundations of Pluralism." Summary and Conclusion 1 1 3 entities other than the subject. Finally, while the facts we have had under review point to no absolutely definite conclusion as to the nature of these entities, certain indications contained in those facts, together with the principle of philosophical economy, make it probable that the most satisfactory hypothesis is to be found in the assertion that those beings which appear to the subject are also subjects of experience, differing only in degree of mentality. Such an hypothesis is in con- formity with the views of modern pluralists, and can be shown to be capable of providing a satisfactory solution of many of the traditional problems with which philo- sophy is called upon to deal. R. s. p> 8 IV THE NOTION OF A DETERMINISTIC SYSTEM I. Introduction 1 HE question as to whether the Universe is completely determined, either eternally or for a very long period of time, by unchanging laws, has provided abundant material for discussion to thinkers of all ages. The prominence given to it in the great speculative systems of the past is attributable to the fact that it has for every man a deep personal interest. Physical science has been largely successful in reducing the world we perceive to a system governed by fixed laws on the basis of which we are enabled to make successful calculations as to the state of the material universe past, present, and future. In face of this reduction to a calculable order, we are compelled to reckon with the possibility that those things with which physics has not hitherto dealt may themselves be reducible to an orderly system concerning which confident predictions based on calcu- lation are, at any rate theoretically, possible. Yet it has always come hard to men to admit that the subjective impulses to which human actions owe their inception, are objectively determined in such a way that an observer provided with sufficient data could foretell the future behaviour of any living being. The reluctance to make such an admission is by no means unjustified, for the question is a very subtle one, and, in spite of endless Introduction 1 1 5 argument, it can hardly be said that a definite and satisfactory solution has ever yet been put forward. Verbal ambiguity is to be found at the root of very many of the difficulties which have arisen in connection with the problem of determinism. It will therefore be necessary to make quite clear what is meant by a *' deterministic system," by rendering precise the notion which is at the back of men's minds when they speak of thingsas being eternally "predestined "or "determined." Words and phrases such as these are used in various meanings which tend to become inextricably confused with one another. In the first place there is the idea of physical compulsion. It is suggested that the be- haviour, not only of the inorganic world, but also of the individuals composing the organic world, is the necessary and inevitable outcome of an endless chain of purely physical causation rigidly fixed by a number of unalter- able laws. In other words, the mental impulses from which our actions originate are held by this view to be entirely due to conditions which are purely material. Now it is evident that our actions are not unrestricted physically. They must be adapted to our material environment, and are therefore governed, at least in part, by the laws which hold in that environment. Yet it is no less evident that the mind which originates these actions, and which in fact takes account of the limitations imposed by matter, is not itself necessarily determined by matter. Consequently determinism in this sense of physical compulsion is now accepted by very few people. Whether there is logical interdepend- ence of mind and matter is, of course, an entirely different question. 1 1 6 The Notion of a Deterministic System There is another sense in which everything is sometimes said to be determined, namely that it is what it is. The past has been what it has been, the present is what it is, the future will be what it will be — all are by this fact equally determined. Closer examination of such statements as these, however, seems to indicate that all they assert is that everything has a "'what^' i.e., a definite nature. This is quite true. Since the time of Hegel, at any rate, it has been recognized for the most part that Pure Being is identical with Non-Being. All Being is determinate. But all this sheds no light whatever on such questions as, for example, that of free-will. A volition is what it is, but it may be free. Moreover, the statement that the future will be what it will be, is quite irrelevant to the question as to whether the future is determined now. All it asserts is that the future, when it comes, will be deter- minate, and this is very different from asserting that it is in fact determined at the present time. Evidently the distinction should be made clear between the meanings of the words "determinate" and "determined." In any case, however, a "deterministic system" is not generally taken to mean simply a system in which everything is what it is. Although we may know, then, that the state of a system at any time is determinate, this does not give us sufficient ground for describing the system as " deterministic " in the ordinary sense of that word. The additional condition is required that the state of the system at any time, past, present, or future, within a given interval, should be susceptible of calculation under suitable conditions. We thus arrive at the notion Introduction 1 1 7 of a system the course of which during any given interval is subject to perfectly definite laws capable of precise statement. If these laws are to be precise enough to render possible calculation and prediction to a given order of approximation, they must evidently be such that they can be enunciated in quantitative terms. If they are known, and if there are also certain data describing the state of the system at given times, a functional relation may be obtained whereby we can determine at once the state of the system at any time within the interval considered, by assigning appropriate values to the variables in the functional relation. The process thus consists in combining the laws to which the state of the system must at all times conform, with certain data, and hence obtaining a functional relation which is equivalent to a propositional function describing the " generalized " state of the system, as it were. By assigning suitable values to the arguments in the propositional function, it is transformed into a proposition descriptive of the state of the system at a particular time. A deterministic system is therefore one the state of which at any time is theoretically calculable given its states at certain definite times. In such a system the future, no less than the past, is irrevocably determined. Moreover, it should be noticed that it is unnecessary for the purpose of defining a deterministic system conformably with what appears to be the essence of the vague notions generally formed of such a system, to go into any discussion of "interaction" or "causation" within the system. Whetherany processescorresponding to such terms exist or not, is irrelevant to the question. 1 1 8 The Notion of a Deterministic System Whatever the inner workings of the system, if it be in fact amenable to calculation and prediction in the manner described, it thereby satisfies any condition that might be laid down by even the most rigid determinist. In the definition given, the state of the system was considered as " theoretically " calculable. The number and complexity of the laws and other data might be so great (perhaps infinite) that the necessary computation could not be carried out in practice, at any rate by human beings. In spite of this, however, it is evident that the system is yet deterministic, provided the necessary data do not comprise alt actual states of the system. If the latter were the case, there would no longer be any other assignable values of the arguments in the propositional function. In other words, this function would lapse, with the disappearance of the variables, into a proposition consisting merely in a theoretical description of the whole history of the system in the interval considered. If, however, the necessary data are not of this order of complexity, the system is deterministic, although exact calculation of its states might not be capable of practical accom- plishment. The present state of the problem of determinism is characterized chiefly by the fact that only recently has a stage been reached in which the problem itself, quite apart from attempted solutions of it, is set forth without ambiguity. It is recognized that in deciding whether a system is deterministic or not it is quite unnecessary to know what may be called the " inner mechanism " of the system. Empirical observation alone is sufficient to decide whether the behaviour of the system, no Introduction 119 matter from what it originates, is such as to render successful calculated predictions of its future course based on general laws derived from that observation. It is, of course, an assumption that the system will continue to follow the same laws in the future, but, granted that assumption on the grounds of probability, it follows that the fate of the system is rigidly determined. For philosophy there arises the problem as to whether the universe is a deterministic system in the sense we have defined. If it should turn out that any parts of the universe are free, the universe as a whole cannot be a deterministic system, for parts which might be in themselves determined are liable to interference from those which are free. It may of course happen that some regions of the universe are so remote from the vicinities in which freedom exists that they are practically unaffected by it. Such regions, considered in isolation, would approximate very closely to deter- ministic systems, although the universe as a whole would not be deterministic. It is evident that it is important to make clear the meaning of the term " free." This task, however, will be postponed for later consideration. Meanwhile we may proceed to consider a statement of the case for determinism, which puts the whole matter very clearly, and which suggests very forcibly the arguments in favour of a deterministic view of the universe. II. The Case for Determinism The account just referred to of a deterministic system is due to Mr Bertrand Russell, and to it all 120 The Notion of a Deterministic System the foregoing has been intended to lead up\ Mr Russell's definition is as follows: "A system is said to be 'deterministic' when, given certain data, e^, e^,...e^^y at times t^, 4» ••• 4 respectively, concerning this system, if Ei is the state of the system at any time t, there is a functional relation of the form The system will be deterministic throughout a given period if /, in the above formula, may be any time within that period, though outside that period the formula may be no longer true." In this definition, all unnecessary lumber, such as the notion of causation, is cleared away. As Mr Russell points out, the common view is that inference of the future from the past is made possible by the principle of causality. But the explicit introduction of that principle, bringing, as it does, in its train all sorts of problems as to the exact nature of causality, is superfluous. Granted our functional relation, obtained by empirical observation, any further postulate as to the inner working of the system is irrelevant in deciding whether the system is deterministic or not. The very fact of the existence of the functional relation is sufficient to establish the determinism of the ^ In all that follows, the reference is to an essay on "The Notion of Cause," by B. Russell, which originally appeared in the Proceedings of the Aristotelian Society for 191 2 — 13. It was then published in Sciefttia, Vol. 19 13, N. xxix — 3. Recently it has appeared in a volume entitled Mysticism and Logic and other Essays. The re- ferences are to pp. 199 ff. of the latter, or to pp. 331 ff. of the number of Scientia mentioned. It should be said at once that Mr Russell does not himself come to a definite conclusion here as to whether the Universe is deterministic or not, though he inclines to the former view. The Case for Determinism 121 system. Mr Russell calls the data e^, e,, ... e„ "deter- minants " of the system, remarking that it is evident that a system having one set of determinants will in general have many. In view of the fact that his account of a deterministic system is perfectly clear and un- ambiguous, and comprises all that is generally contained in the notion of such a system, we shall take it as a basis for criticism and discussion. The fact that the future will be what it will be is regarded by Mr Russell as being of considerable im- portance. For reasons already given we do not take this view. It is true, as Mr Russell says, that we cannot make the future other than it will be, but this is very different from saying that the future is in fact determined now. For evidently what the future will be is in part determined by our actions, and we cannot decide as to whether the future is determined now, unless we know whether our actions between the present and any given future date form part of a deterministic system or not. It is true that those actions themselves will be what they will be, but here again the reference is to the future. To say that any future event whatever is determined now by the fact that when it comes it will be determinate, is simply equivalent to saying that everything is something, a true enough statement but hardly to be urged as an argument in favour of however lax a determinism. Mr Russell does not insist on the point, recognizing that it is not what people usually mean by "determinism," but confines his attention for the most part to a deterministic system as defined in the quotation given above. Two important illustrations are given by Mr Russell 122 The Notion of a Deterministic System as bringing out clearly the conception of a deterministic system. Both refer to the possible nature of the universe. In the first the hypothesis of psycho-physical parallelism is introduced — that is, it is assumed that to a given state of brain a given state of mind always corresponds. The highly probable assumption is also made that to a given state of a certain brain a given state of the whole material universe corresponds, since the recurrence of exactly the same brain-state is extremely unlikely. Hence, if n states of the material universe are deter- minants of the material universe, they would also be determinants of the whole universe, mental and material, as would also the corresponding n states of a given man's mind. Evidently, if the above holds, the universe, including man, forms a deterministic system, and conversely if the universe is determined it must be in some way which is closely represented by the above. Moreover, it should be noted that psycho-physical parallelism is not an assumption essential to the latter. For psycho- physical parallelism is rather a methodological principle than an hypothesis, and the fact of correspondence which it asserts would exist equally if there were any form of interaction between mind and brain. It is true, as Mr Russell remarks, that the correspondence between mind and brain may not be one-one, but many-one, or one-many ; but in that case the universe would still be deterministic (though its determinants might be more complex) provided the scope of the correspondence on the multiple side was determined. The second illustration relates to the dispute between the teleological and the mechanistic views of the world. The Case for Determinism 123 A " mechanical " system is reasonably defined as one having a set of determinants which are purely material, such as the positions of certain pieces of matter at certain times. But if some account of the universe such as that, for example, given in the first illustration were true, all mental facts, including purposes and desires, as well as the universe of matter, would be determined by such a set of material determinants. Hence purposes, whether realized or not, could exist in a mechanical system, so that the latter might also in that case be fairly designated " teleological." Thus if the view taken is correct, the terms " teleological " and " mechanical " are not incompatible. There might be a mechanical system which was also teleological and vice versa. Clearly these two illustrations are particularly valu- able, for the first shows us under what form we must conceive the universe if it is actually determined as a whole, while the second indicates one important conse- quence which would necessarily follow in a universe of that nature. Evidently if we form part of such a world we cannot be content to regard ourselves as " free " in any satis- factory sense ; for given (say) the positions of certain pieces of matter at certain times, all our actions would be theoretically calculable, past, present, and future. Even now our destiny would be irrevocably fixed by the laws of mathematics. Could a sterner necessity, a more unbending taskmaster, be imagined .'* Determinism possesses as its chief advocate the success of physical science. Nobody pretends that our knowledge of the material universe is all-inclusive. 1 24 The Notion of a Deterministic System But science has dealt so remarkably with the limited portions in space and time at our disposal, by weeding out, in any given case, the superfluous (because negligibly effectual) accompaniments so as to temper the problem to our intellectual capacity, that we feel little difficulty in thought in extending the process, by analogy, to the performances of a Laplacean calculator who, given certain data, would derive the knowledge of all things. For a mind thus capable of grasping the infinite com- plexity of the determinants involved, the universe would be an open book. The success of science seems to render it highly probable that the material universe is completely determined by a limited (though perhaps infinite) number of material data, and hence, granted the existence of the psycho-physical correspondence, that the mental universe is also completely determined by those data. If we are not to pass unchallenged a view of such far-reaching significance, it will be necessary to analyze fully the grounds on which it is based. Only thus can we arrive at a true estimate of the credibility to be attached to it. III. Analysis of Determinism The first step in the analysis of determinism in the sense which has been defined, consists in theexamination of the functional relation H f^j \€^,) ^i> ^2> ^2> ••• ^Ji> ^m ^/J Upon the probable existence of which the case for determinism rests. If the state of a system at any particular time is given by such a relation, what exactly is it that thereby determines the system ? The deter- Analysis of Detennmism 125 minants e^, €.,,...6^ with the corresponding times are not sufficient. From them alone we can derive no information about the system. Given the function, they fix its value for particular values of the variable ; that is, they may be considered as necessary and sufificient determinants of the values of the function in particular cases. But for the system to be determined, not only must the data e^, t^, -..^rt, 4 be given, but also the relations between them, that is, the way in which they enter into the function f In other words, the form of that function must be given, and it is therefore a determinant of the system equally with the data ^1, e.,, etc. Summing up then, we must regard ^i, e.,,...e,^, together with the form of the function connecting them with one another and with the variable t, as the determinants of the system. Let us now consider the course we should have to pursue, had we the intellectual grasp of the Laplacean calculator, in order to discover from observation whether the universe is a deterministic system or not, i.e., whether there exists for it a functional relation such as Ef =f. At least one possible type of the necessary determinants ^1, ^2, ...^,j would probably be discoverable from em- pirical inspection alone, as is generally found to be the case in scientific observation. But the verification of this possibility and the number of determinants thus required, together with the form of the function into which they enter as constituents and' which is the remaining determinant, would only be certainly demon- strated in that final synthesis whereby the functional relation is constructed in its completeness. In the first place, we must not assume that the 1 26 The Notion of a Deterministic System course of the material world is entirely independent of mind. Prima facie it is not so, and should the appear- ance be misleading, that fact could only be demonstrated in the course of our calculation. Consequently our first step, after a course of exhaustive experiment and observation of mind and matter and the interaction (real or apparent) between them, would be to collect all the results of that empirical procedure under a number of general laws. This number should be the mini- mum possible in accordance with the results so far obtained. The next step would be the inspection of these general laws with a view to reducing their number. If our process had been carried on in a manner more or less analogous to the evolution of science, the laws would have been obtained as a result of many relatively independent lines of enquiry. An hypothesis fitting all the facts must now be sought by means of which the phenomena can be brought under a common heading, and the laws to which they conform synthesized if possible into one all-embracing formula. This is the ideal to which all science turns its endeavour in dealing with the material world. Examples of it on a com- paratively small scale are common enough. For instance the kinetic theory of matter enables us to reduce many of the facts observed in the study of Heat and Light to mere manifestations of an under-lying process which is purely mechanical. The science of Heat proceeded originally as an enquiry quite independent of the science of mechanics, yet in the end it has been reduced to a common basis with the latter. The theoretical investi- gation we are considering would, of course, be far more Analysis of Determinism 127 complicated, for in it we are dealing not only with all material phenomena, but with all mental facts as well. But in any case one important point is evident, namely that a necessary condition of the possibility of reducing the laws we have obtained to a single formula, is that those laws should be capable of precise statement, i.e., of being put into exact quantitative form. If this con- dition is fulfilled, and if we can relate all the quantities to which the different laws refer, the way is clear for the con- struction, by numerical calculation, of that single functional relation which shall determine the whole universe. In the course of our calculation it might appear that we should be led to a functional relation capable of statement in different ways according to our choice of the determinants e^, e., ... e,,. At this stage of the calculation we should have a number of data at our disposal all of which would not be found to be necessary in achieving the final result. Our decision as to which set of independent data to select would then depend on the particular form we wished our function to take in virtue of the nature of its determinants. Possibly we might be able to make the latter all material, or all mental, or partly material and partly mental. If, how- ever, it turned out that by appropriate manipulation we could eliminate all the mental determinants and yet arrive at a single functional relation, the universe would be a deterministic system whose history, including the history of every living being in it, would be fixed by a set of purely material determinants. Given some such data as the positions of certain pieces of matter at certain times, we could predict with absolute certainty the future behaviour of any man. Discredited astro- 1 28 The Notion of a Deterministic System logers may perhaps draw some comfort from this con- sideration ! It follows from the foregoing that if a system is to be deterministic in the sense we have been considerino-, it must be one of which quantitative notions are signifi- cant, that is, one whose state at any time is capable of being described in terms of quantities which are theo- retically measurable. For example, qualities, as such, cannot enter as constituents into the functional relation which gives the state of the system, seeing that qualities cannot be exactly specified but only indicated — -to be comprehended they must be experienced. Yet it may be possible to indicate the qualities by quantitative concepts, just as, for instance, we correlate " red " with a certain wave-length or range of wave-lengths. If this is so, the quantity may enter into the functional relation, thus " representing," as it were, the quality, and results deduced from the relation will be valid and capable of being re-interpreted, where necessary, in terms of qualities. If, however, it is impossible to make precise quantitative notions in any way significant of the system, the latter cannot be deterministic. Another possibility suggests itself. There might be a system the state of which at any time is capable of description in terms of measurable quantities, and yet for which no functional relation exists. Mr Russell makes the following statement in this connection : " If formulae of any degree of complexity, however great, are admitted, it would seem that any system, whose state at a given moment is a function of certain measur- able quantities, must be a deterministic system. Let us consider, in illustration, a single material particle, Analysis of Determinism \ic) whose co-ordinates at time / are x^, y^, Zf. Then, how- ever the particle moves, there must be, theoretically, functions yi,y".,y^, such that But let us take another example. Consider two material particles attracting one another with a force which is some function of the distance between them. Now suppose this function itself varies, also that the law of its variation varies, and so on. If at any stage of this regress (which may be infinite) the law of variation were known, we could construct our functional relation. This, apparently, is what Mr Russell means when he makes the proviso that formulae of any degree of complexity, however great, should be admitted. But it is conceivable that we should never (even after infinite regress) come to a laiv of variation. The variation might conceivably be purely haphazard, or at least containing a haphazard element which renders any precise statement of a law impossible. Whether such a system could exist is not the question. At any rate we can imagine it to exist. Its state at any time could be exactly described in terms of measurable quantities, such as the co-ordinates of the particles and their velocities and accelerations ; but no functional relation could be constructed giving its state at any time\ Such ^ It may be granted that Mr Russell's contention is true if it merely implies that for any moving particle there is a one-many relation between its positions and the instants of time. This assertion, however, is simply equivalent to saying that such-and-such an entity (the particle) exists in a space-time continuum. But the point is that in a case such as that considered, it would be impossible to construct (even theoretically) equations of the type .r^ =/i (/), etc., from which, given the positions at some (not all) of the instants in the interval considered, we could infer the positions at the other instants. That R. s. p. 9 130 The Notion of a Deterministic System a system would be called "non-deterministic" or, to use Mr Russell's word, "capricious." If, however, it is not even possible to describe the state of the system in quantitative terms, because quantitative notions are not significant of it, then the words "deterministic" and " non-deterministic " are not significant of it either. Let us now consider what must be the essential characteristics of a system of which quantitative notions are significant, and to which in consequence numerical calculation may be applied — calculation which will be successful, at least within limits, unless the system is wholly "capricious," Quantity is expressed bymeans of number. Number is a property of classes. A given number is the common property possessed by all classes having that number of members. Now a class is a collection of objects (using the last term in its widest sense), and the latter may be considered as units. Replacing, permissibly for formal purposes, the common property defining members of a class by that class as a whole, we have as the definition of the number 7i, the class of all classes of n units \ This definition in its first intention applies only to positive integers, but the concept can be extended with- out great difficulty to negative, fractional, and irrational numbers. The quantities which spring naturally to mind at once are those termed "extensive" quantities, i.e., those having a nature such that a given quantity may be is, no matter how many elements of the motion were known, they could not be made the basis of inference to unknown elements. The inference of the position at any instant would presuppose knowledge of that position, thus reducing the process to a vicious circle. ^ See, e.g., Dr Whitehead and B. Russell's Principia Mathematica, Vol. I, Part n. Analysis of Determinism 131 regarded as the sum of smaller quantities of the same kind, which we may term "parts" of the given quantity. It is common to divide such quantities into two types, viz. those associated with a finite number of discrete objects, and those which cannot ultimately be regarded as made up of a finite number of parts. The former are used in statements such, for example, as that re- ferring to the number of legs of a certain species of insect, or, to give another instance, in dealing with phenomena depending on the number of molecules in a given volume of a gas. In such cases, any one of the discrete parts forms a natural unit of measurement. To the latter type belong such quantities as distances, which cannot be considered to be composed tdtimately of 2i finite number of parts. For our purposes, however, it is not the difference between the two types which is important, but the characteristic which they have in common, namely that any given finite quantity^ may be regarded as the sum of a finite number of smaller yf;^2V^ parts, these parts being themselves quantities of the same nature as the given quantity. In addition to quantities of extensive magnitude, science also has to deal with those having "intensive" magnitude, such as density and temperature. We can- not regard a density as the sum of other densities without great ambiguity. But it is important to notice that, in any case, the measurement of such quantities is only effected by correlating them with quantities possessing extensive magnitude. Thus, if the notions of quantity and calculation are to be significant at all, we must ultimately deal in every case with extensive quantity, ^ In the case of quantities of the first type, the given quantity must evidently comprise at least two of the discreta. 9—2 132 The Notion of a Deterministic System i.e.y with things which may be considered as made up of parts similar in nature to themselves. Intensive quantities are similar in one respect to certain qualities, such as those of colour, seeing that they may be specified by correlation with extensive quantities. I n fact, strictly speaking, the term "quantity " might well be restricted to the latter. For if there were any actual entities corresponding to things such as temperature and density, having what we call "inten- sive" magnitude, they would be really more akin to abstract qualities or states. Probably the only reason we call them "quantities" at all is by an illegitimate transference of idea, because we can correlate them with true quantities. The last point is made clear by the consideration of objects of a certain kind which are supposed to be intensively quantitative, namely sense- data. For example, people commonly regard a bright yellow light as having in some way a greater magnitude than a dim yellow light. But the difference is purely qualitative. Certainly both lights are yellow, but the difference between bright and dim is qualitative; just as red and yellow are both colours, but the difference between them is entirely qualitative. That the differ- ences of sense-data in respect of intensity are purely qualitative is shown by the fact that there could be no objective standard of measurement for them'. To assign the number 10 (say) to a sense-datum of a certain in- tensity, would be arbitrary and meaningless. Fechner made an ingenious attempt to construct a system based on a "least perceptible" difference of intensity. Not only, however, would this difference vary with different ^ The nature of the intensity of sense-data is a point of consider- able interest, and is dealt with at length in the next essay. Analysis of Determinism 133 people and probably also with the same person for different total situations, but it is itself qualitative. It is meaningless to talk of a difference in intensity of sense-data as being so many times the "least perceptible" difference. All such attempts reduce ultimately to cor- relation with true {i.e., extensive) quantities, viz. the physical stimuli concerned. Moreover, it should be remembered that all physical concepts, whether those such as mass and energy, or those such as density and temperature, are really constructions of sense-data. Hence the possibility of applying quantitative notions to what is perceived, will depend finally on whether the object of experience may be regarded as made up of parts (sense-data) standing in certain relations, or not. To this we shall return shortly. Quantity, then, is expressed by a number of units, one important condition being that while dealing with a fixed type of quantity the units must be homogeneous. From the above it is evident that quantity is actually significant only of things which can be considered to be made up of parts, these parts constituting the units. Evidently the number expressing a fixed quantity will depend on the scale chosen, that is on the part selected as the unit of measurement. The foregoing is made clear by considering its ap- plication to the world of physics. Evidently the latter is deterministic, if the fundamental postulates of physics be granted. In any case, quantitative notions are signi- ficant of it. Let us endeavour to work back to the ultimate reason for this. The unitary entities constituting the universe as conceived by physical science are points, instants, and particles. Such entities as these are capable of being 134 ^-^^ Notion of a Deterministic System exhibited as logical constructions of the immediate data of sense^; they are not inferences from the latter. The physical conceptions which are psychologically primitive are those of force, duration, and distance. The notion of mass is derivative. In ordering our ideas, however, it is common to make a re-arrangement by taking the concepts of mass, time {i.e., lapse of time), and length as logically prior, and making that of force derivative. On these fundamental conceptions the science of Me- chanics is based, and all Physics is based on Mechanics. It is true that as physics has developed it has been found necessary to introduce two more fundamental quantities, namely (as the most convenient choice) temperature, and either magnetic permeability or spe- cific inductive capacity. The dimensions of the two latter in terms of mass, length, and time are not known, but the dimensions of their product are known. They are those of the inverse square of a velocity. Hence it is not improbable that the dimensions of the separate quantities may ultimately be discovered. In any case, however, no difficulty arises in practical calculation, for the two quantities mentioned enter into our equations merely as numerics, namely as the ratios of their values for any substance to their values for air. Hence the question of their expressibility in terms of mass, length, and time does not arise. Temperature, permeability, and inductivity are intensive. Hence it has been proposed to replace them by entropy and electric charge, both of which have extensive magnitude. Resuming our discussion, it should be observed that in measurements of mass we are always determining ^ See, e.g.^ B. Russell's Our Knowledge of the External Worlds Lect IV. Analysis of Determinism 135 mass-ratio. We select any convenient standard of mass and find how many such parts would make up the mass of the body we are considering. Now the relations between force and mass (or inertia) are expressed in the fundamental postulates of mechanical science, com- monly known as Newton's Laws of Motion. From the second law it appears that the ratio of the masses of two bodies is inversely as the rates of change of their velocities produced by equal forces acting on them, or, if the forces are impulsive, inversely as the sudden changes of velocity produced. The third law states that exactly equal forces (though in opposite directions) act on the bodies in the case of any interaction between them. In particular, in the case of impact between the bodies there exist equal, opposite, impulsive forces. Hence the theoretical measurement of mass-ratio (which is all that can be measured) depends on the observation of changes in velocity. Hence measurements of mass reduce to measurements of time and distance. Thus the significance of quantitative notions in physical science depends ultimately on the fact that we are here dealing with things which may be considered as made up of parts, namely times and distances. The parts selected will be the units of time and distance in terms of which we measure. We are thus led to the conclusion that physical cal- culation in connection with the world we perceive is rendered possible by the fact that the nature of the object of sense-experience is such that it may be con- sidered (at least to a close approximation) as made up of parts standing in spatial and temporal relations. Hence quantitative notions are to this extent applicable to wha twe perceive in sense-experience, and conse- 136 The Notion of a Determmistic System quently the terms "deterministic" and "non-deter- ministic" are significant of what is thus perceived. It does not follow, of course, that sense-data form a deterministic system. As we have previously stated, the physical system is deterministic provided the funda- mental postulates of physics are granted. But the physical system, based as it is on the supposition that the object of experience is made up of parts standing in certain relations, is only an approximation. In actual sense-experience the object is a presented whole, one and indivisible. The object perceived by each subject is unique; but in reflective analysis, which is discursive, we are compelled by the limitations of intellect to regard all objects of experience as having at least one common characteristic, namely that they are made up of parts (termed sense-data) standing in spatial and temporal re- lations. That this conception is a close approximation to actuality is shown by the fact that under suitable circumstances we can successfully predict, by adopting it, what we shall perceive at future tim^s, provided we a7'e attending appropriately at those times. But we are only approximating, and the perception by which we verify our calculation only approximates to the predicted result to the same order as our oris^inal data for calcu- lation approximate to the perceptions on which they are based. Moreover, it must be remembered that the object of experience is qualitative. Qualities cannot be dealt with by calculation directly, though it may be possible to deal with them indirectly by correlating them with quantities. But even in this case we have no guarantee that the quality which one person correlates with a certain quantity can be considered as similar to that Analysis of Determinism 137 which another person correlates with the same quantity. The appearance which I call "red," for example, may be qualitatively quite different from that which you call "red." If such were the case, it would be possible at most to predict approximately one's own sense-data. One could not predict in imagination those of other people. And there is also the further point that it seems probable that even for one person each per- ception is qualitatively unique. This being so, any prediction of one's own sense-data would be of the roughest character. Evidently, then, there are strong presumptions against the view that sense-data form a deterministic system. The question will not be pressed, however, for we are not here concerned to come to a definite conclusion on that particular aspect of the problem. But we may note that the material world can only be regarded as a deterministic system, if it be isolated from mind. Prima facie, however, mind interferes with the course of matter, and it remains to decide whether that interference is determined or not. This brings us to our next point, the application of determinism to the mind. IV. Determinism and the Mind In investigating the problem as to whether the notion of determinism is applicable to the mind, we must first of all be quite clear on the meaning of the latter term. In other words, what types of facts are to be in- cluded under the heading "mental"? The most satis- factory and unambiguous definition of "mental" is that which pertains to the subject as distinguished from the object of experience. This limits us to feeling and the 138 The Notion of a Detennmistic System various forms of subjective activity, such as thinking, willing, desiring, etc. All such modes of activity may probably be reduced to the single activity of attention, the differences between them consisting in the different types of objects respectively attended to. There seems, however, to be an exception in the case of volition, which appears incapable of reduction to attention alone in so far as it implies motives^. Let us approach the problem before us by referring back to the first example given by Mr Russell as illus- trating a deterministic system. We saw that it repre- sented a world which the actual universe must resemble more or less closely if it is deterministic. In that example the probability is considered of there being a number of mental determinants of the world. "Given n states of a given man's mind," it might be possible to calculate the state of the whole universe, mental and material, at any given time, past, present, or future". Now there are two points in the phrase "Given n states of mind" which demand criticism. In the first place, what exactly is a state of mind ? It is evidently impossible to differen- tiate precisely one state of mind, which might accordingly be "given." State of mind, if it means anything, must mean the subject acting and feeling. But the subject is one — an individual entity. We cannot significantly suppose a section of his activity and feeling "cut out," so to speak, as one state of mind. His active existence is an indivisible whole. We cannot even conceive of sections of it. Nor can it be objected that we can at least say that some actions are before others, thus making a time-basis on which such a conception might ' See, e.g., Dr James Ward's article "Psychology" in the Ency. Brit. (Sec. 9). ^ Seep. 122 above. Determinism and the Mind 1 39 be founded, for in all cases the temporal sign attaches not to the activity itself, but to the changes in the object of experience of which the activity is the ground. We only arrive at the perception of temporal relations in the object of experience through our activity in dif- ferentiating that object; and we cannot conceive of a "section" of that activity by attempting to correlate it with a temporal section of the object, for it is the activity of the one individual subject in whom it is evidently meaningless to try and distinguish temporal relations. We might just as well (and just as meaninglessly) attempt to conceive a spatial section of activity on the ground that we attend to objects situated in different places. A second point now arises. Even if it be granted that there is something actually corresponding more or less vaguely to the notion of a state of mind, what is meant by saying that it is "given"? Evidently it cannot be given as an immediate datum, that is, as an object of acquaintance. For clearly one subject cannot be acquainted with the feeling and activity of another subject. Nor can he be acquainted with his own feeling and activity. The latter would imply that the subject (not as conceived, but in his actuality) was object of his own knowing, which is impossible. Yet we certainly have knowledge about activity and feeling. How does this arise? It is based on what may be called realization. We realize our own activity and feeline, for it is we who feel and are active. Such realiza- tion is not itself knowledge, for it implies no object, though the proposition asserting its existence is, of course, a piece of knowledge by description. Thus a 140 The Notion of a Deterministic System "state of mind" cannot be given as an immediate datum, nor can it be described with any adequacy. Let us even grant, however, that there is some sense in which a "state of mind" may be supposed to be given. We must then enquire as to the form in which it is given. In considering the attempts of the Laplacean calculator to discover whether the universe is deterministic or not, we saw that it would be neces- sary to formulate general laws referring to matter, mind, and their interaction (real or apparent), based on ex- haustive observation. From these it might be possible to construct, by calculation, a functional relation of the type considered, and thus to establish the determinism of the universe. If it were found possible, in the course of the calculation, to eliminate mental factors, the uni- verse would have at least one set of purely material determinants. As, however, we cannot avoid introducing mental factors at the outset, it is evident that if the calculation is to be possible at all, a "state of mind" must be capable of being given in a quantitative form. Otherwise, we are debarred at the very beginning from attempting to construct our functional relation ; de- barred, not by practical difficulties of computation, but by the fact that nothing can be inferred from such a relation as to things of which quantitative notions are not significant, and conversely, that if such things exist in the universe, the attempt to construct a functional relation which shall be significant of the whole universe, is meaningless. For it implies calculations involving things to which calculation is not applicable at all. Moreover, if material determinants are sufficient, we ought to be able to predict future mental facts Determinism and the Mind 1 4 1 simply by determining the material state of the universe at that time. But the material state could only be speci- fied quantitatively, and how would it be possible to correlate it with mental facts if quantitative notions are in no way significant of the latter ? We might conceiv- ably be able to predict fairly accurately the sense-data which a given man would perceive at that tim^, provided he should be attending appropriately. But how are we to foretell whether he will be attending appropriately, unless attention is susceptible of calculation, that is, is quantitative in nature? It does not help us in the least to assume that to a certain state of brain, a certain "state of mind" corresponds, unless we can state pre- cisely the nature of the correspondence. Nor can it be urged that just as we correlate a sense-datum such as "a patch of red" with something quantitative such as a wave-length, so may we perhaps be able to correlate mental facts with something quantitative. For the con- cept of a wave-length is itself merely a construction of sense-data, and ultimately depends on the fact that it is possible, at least approximately, to apply quantitative notions to the object of sense-experience by regarding it as made up oi parts standing in spatial and temporal relations. Unless some such approximation is also pos- sible in the case of activity and feeling, it is hopeless even to attempt to correlate them with objects of which quantitative notions are significant. The crucial test, then, lies in discovering whether quantity is significant of feeling and activity or not. Let us consider feeling first. Certainly we talk as if the vaguer quantitative notions might apply to it. We speak of being more or less pleased or pained. But 1 42 The Notion of a Deterministic System obviously we cannot get a certain pleasure by addition of other pleasures. A state of "being pleased" is one indivisible thing, not a collection of parts which are themselves states of "being pleased." It will prob- ably be urged that pleasure and pain are analogous to quantities having intensive magnitude. We saw, how- ever, that the latter are not strictly quantities at all, and, in fact, the only reason we ever apply the term to them is that we can correlate them with true quantities. But the basis of this correlation is the fact that both the terms in it are reducible ultimately to sense-data, of which they are constructions, and sense-data may for most purposes be regarded as parts of the object of ex- perience standing in spatial and temporal relations. No such correlation as that mentioned is possible in the case of feeling. For what is the necessary extensive quantity ? Certainly not something physical, for there would then be no common basis such as we get in the case of two correlated quantities which are both physical. Feeling is not a sense-datum. We do xiO\. perceive feeling. We feel. Probably we only use even the vaguer quantitative terms "more" and "less" of pleasure and pain, which are purely qualitative, because we instinctively try to objectify them by comparing them to physical stimuli possessing intensity. Thus if we are to specify feelings quantitatively at all, it must be by correlating them with some other mental factor which is quantitative. This brings us back to the original question. For the only other mental factor is activity, and this we must now consider. When we come to deal with the various modes of subjective activity, we find that the hopelessness of Determinism and the Mind 143 the attempt to make quantitative notions significant of mental facts is more clearly demonstrated than ever. Again we use the vaguer quantitative terms. We " concentrate " our attention. We are "more absorbed" in some things than in others. But here the quanti- tative reference is evidently to the objects to which we attend. For example, by concentration we simply mean that we confine the portion of the object of experience termed " the focus of attention " to very narrow limits. Hence the quantitative reference is strictly to that portion of the object, and not to the attention itself. Moreover, attention depends on in- terest. Clearly interests are not quantitative. They are not made up of parts which are themselves interests. Similar considerations make it evident that what we have just said is true of all mental activity. A willing is not the sum of parts which are themselves willings. A thinking is not made up of thinkings, nor a desiring of desirings. And in general, we may say that the notion of an act of attention as being made up of parts which are acts of attention, is quite meaningless. Any attempted analogy of mental activity with quan- tities having intensive magnitude is of no help what- ever. As we have seen, if there are entities actually corresponding to the concept of such quantities, they must really be more like abstract qualities. But this is not the most serious objection. To insist on a pre- vious point, the idea of a quantity only becomes applied to these concepts because we can correlate them with true or extensive quantities. On what is this correlation based ? An intensive magnitude is always a function of extensive magnitudes. Density is the 144 T^^^ Notion of a Deterministic System ratio of mass and volume ; temperature, of energy and en- tropy; permeability, of magnetic induction and magnetic intensity. In each case the ratio is that of two extensive quantities, or of two quantities which may be simply and immediately correlated with extensive quantities. Now extensive quantities are logical functions of sense-data. Hence all quantities we thus deal with start from the same basis. From sense-data we construct extensive quantities, from extensive we construct intensive quan- tities. That is the principle of the correlation. It is not a correlation of two things springing from utterly independent sources. It follows from the fact that we may consider the object of experience as composed of parts standing in spatial and temporal relations. We cannot, however, arrive at feeling and attention by constructions of sense-data, especially as the con- structing process itself involves attention. Therefore we cannot possibly correlate mental facts with any material quantities, whether the latter possess intensive or extensive magnitude. Hence, if quantity is to be significant of any mental factor which is not itself directly quantitative, it must come about by correlating it ultimately with some mental factor of which the notion of extensive quantity is significant. But we have seen that this notion is not significant of any mental factor. Thus it is impossible to effect the necessary correlation in any way whatever. Evidently the foregoing is summed up in the state- ment that the feeling, acting subject of experience is an absolute, indivisible unit. The notion that the subject is made up of parts (themselves subjects) is without any meaning. Nor is it possible by any means Determinism and the Mind 145 to correlate his feeling and activity with something possessing magnitude. Therefore quantitative notions are utterly without significance in application to the facts of mind. Hence it is impossible to construct a functional relation of the type considered which shall take account of mental factors, and it is impossible, not because of any practical difficulties of calculation, but because the existence of a functional relation in such circumstances is contrary to the very idea of such a relation. The subject, then, is not determined. Strictly, it is neither true nor false to say that the subject is deter- mined. It is meaningless. But the universe comprises subjects, and so no functional relation can exist which is descriptive of the state of the universe as a whole. Therefore the universe is not a deterministic system. V. The Problem of. Free-Will The application of the results of the preceding section to the problem of free-will is more or less obvious. There are one or two points of importance, however, which deserve notice. In the first place, it is necessary to give a clear meaning to the term " free." Probably confusion has often arisen in past discussions on the subject by regarding this term as the opposite of " determined." That view is incorrect. The opposite of " determined " is "undetermined." Both these terms apply to systems of which quantitative notions are significant, i.e.^ to systems whose states at any given time can be described in terms of quantities which are theoretically measurable. As we have seen, if, from observations on the state of such a system at certain R. s. p. 10 146 The Notion of a Defennimstic System times, it is theoretically possible to construct a func- tional relation giving its state at any time (at least, within a given interval), the system is "deterministic," or "determined." If, on the other hand, the construction of such a relation cannot be performed, the system is "undetermined," Neither of the terms thus defined, however, is applicable to systems of which quantitative notions are not significant ; and it is to these systems that the term " free " may properly be applied. Hence, since quantity is not significant of volition, the will is free, or rather we are free in willing. Clearly, though, volitions are not utterly chaotic. There is a very definite sense in which they are intellig- ible and coherent. But the coherence is not of a logical kind. When we say that a man's actions are intelligible, we mean that we understand them. The basis of this comprehension is not formal and abstract, but concrete. It is not the laws of logic, but the nature of the self. The self is purposive; its striving is towards betterment by entrance into a completer harmony with the active beings that surround it. Thus a man's actions are intelligible to us, when we realize that they are the expression of purposes analogous to our own. In that Realm of Ends (to use Kant's expressive phrase) which constitutes the world as we know it, we find, not logical determinism, but teleological guidance. The category of End or Purpose is subjective. It cannot be reduced to any other category, nor can the things to which it applies be subsumed in any way under the notion of quantity. Purposes, intentions, and motives are not measurable. They are not capable of description in quantitative terms, nor can they be cor- The Problem of Free-Will i^j related with quantities. Yet their introduction into our explanations of certain facts is inevitable. We cannot explain the coming together of the parts of a watch or of a motor-car simply in terms of the motions and configurations of the molecules composing the brain and nervous system of each person taking part in the manufacture. For such creative work is the very opposite of what is implied in the laws of molecular physics. There we have a constant breaking down and levelling, not a building up. Any attempted explanation of a work of the kind considered, by the help of purely physical categories, inevitably leaves us dissatisfied, with an irreducible minimum which cannot be thus explained. This residual factor is the purpose for which the watch or the motor-car is designed ; and the fact that quantity is not significant of purpose emphasizes the truth of the statement that the subject, in the exercise of the power he possesses of guiding the course of phenomena, is free, in the fullest sense of that term. VI. Summary and Conclusion At the outset of our enquiry into the problem of determinism, we saw that it would be necessary to elucidate a clear meaning of that term, owing to the fact that many of the difficulties which have arisen in the past have been purely verbal in nature. Consequently we decided to attempt to formulate a satisfactory definition of a "deterministic system," and one which should correspond as far as possible to the essence of the generally accepted meaning of that phrase, by discover- ing the bare necessary and sufficient conditions which lO- 1 48 The Notion of a Deterministic System such a system would have to satisfy. The notion of purely physical compulsion was rejected ; for although our actions, in so far as we are part of the material world, must conform to the laws which hold in that world, yet the mental factors which give rise to those actions are not prima facie governed by purely physical conditions. In fact, the reverse appears to be the case. Even should it turn out, however, that mental "events" are determined in some way by physical events, it seemed probable that the facts would be capable of more abstract statement than would be involved in the assumption of some definite type of "compulsion" — whatever that may mean. The statement that "the future will be what it will be " was next considered. Analysis showed that all such propositions reduce to the assertion that being is determinate. Pure being is identical with Nothing. Accordingly, it appeared necessary to call attention to the distinction between the terms "determinate" and " determined." To say that the future will be what it will be, is to say that the future when it comes will have a definite nature, i.e., will be determinate. This is very different from saying that the future is in fact deter^nined now. The latter implies that if only we had sufficient data, we should be able to predict the state of the universe at any future time. The last consideration leads at once to the necessary condition which a system must fulfil if it is to be deter- ministic. The formal statement of that condition is due to Bertrand Russell. If a system is deterministic throughout a given interval, there must exist a functional relation giving its state at any time during that interval, Sttmmary and Conclusion 1 49 involving as data certain events at certain times, these events being defined as the "determinants" of the system. A system might have more than one set of determinants. If one set is purely material, the system is "mechanical." The mere existence of the said functional relation is the one necessary and sufficient condition that the system should be deterministic. The particular type of inner nature or "mechanism" to which the behaviour of the system described in the functional relation is due, is not referred to in the formal definition, nor is it necessary to refer to it. The illustrations given by Mr Russell, expressed in terms of the methodological principle known as psycho- physical parallelism, depicted a world of a type to which ours must closely conform if it be deterministic. It appeared that the state of such a world at any time might be given by a functional relation having purely material determinants, or purely mental determinants, the latter consisting in a number of " states of a given man's mind." One important characteristic of a world of this kind, pointed out in the second illustration, is that it might be called " mechanical " or " teleological," according to the point of view — " mechanical," because one of its sets of determinants is purely material, "teleological," because purposes exist in it which may be realized. It was seen that the view which regards the universe as a deterministic system, receives powerful support from the fact that the conception of such a system has been applied most successfully to the world as dealt with by physical science. Analysis of the foregoing statement of determinism showed, in the first place, that \ki^ form of the functional 150 The Notion of a Deterministic System relation should be included as a determinant of the system. The determinant data of the functional relation alone could give us no information about the state of the system. We must also know the exact relation between them, i.e., the form of this function into which they enter as constituents. Assuming that we had the intellectual grasp of Laplace's imaginary calculator, we saw that in order to construct a functional relation descriptive of the state of our universe at any time, we should have to start by drawing up a number of general laws, based on exhaustive observation and experiment, dealing with mind, matter, and the interaction between them. No initial assumption about the latter is permis- sible. We should then have to try and combine our general laws by calculation, in such a way as to be left with the required functional relation. 1 1 might be possible to effect this combination in more than one way, the final function involving correspondingly different sets of determinants. In particular, one appropriate elimina- tion might give a set of purely material determinants — another might give asetof purely mental determinants. In any case, it is evident that if the construction of the functional relation is to be conceivable, the general laws originally drawn up must be capable of expression in precise quantitative terms. Hence the words "deter- mined" and "undetermined" can only be significant of systems of which quantitative notions are significant, i.e.^ of systems whose state at any time can be described in terms of theoretically measurable quantities. If a functional relation can be constructed for a system of this kind, the system is determined ; if not, it is undetermined. But unless the state of the system can be described Summary and Conclusion 151 thus quantitatively, neither of these terms is significant of it. The quantities with which we have to deal are of two kinds, those having extensive, and those having intensive, magnitude. It appeared, however, that the term " quantity " might well be applied only to the former. It becomes applied to the latter, because in every case such a quantity may be specified numerically by correlation with extensive quantities. If, however, we try to imagine something actually corresponding to these so-called "intensive" quantities, we should have to conceive them rather as abstract states or qualities. Temperature may be taken as a typical example. Accordingly it follows that quantitative notions are only applicable ultimately to systems the constituents of which are such that any finite portion of them may be considered as made up of a finite number of finite parts. Whether the process of dividing can be carried on to an ultimate conclusion is another matter which is irrelevant to the point here considered. It is now apparent why the purely physical world may be regarded as deterministic. Physics is based on Mechanics, and the logically prior concepts of Mechanics are mass, distance, and time-interval. Analysis shows that all physical measurements (including that of mass) reduce finally to measurements of distance and time. Hence the fact that quantitative notions are significant of the physical world depends ultimately on the fact that the object of experience may for most purposes be considered as made up of parts (sense-data) standing in spatial and temporal relations. This view, however, is only a very close approximation, for in actual sense-ex- 152 The Notion of a Deterministic System perience the object is one and indivisible. Hence it does not follow necessarily that perceived phenomena form a deterministic system, and this more especially as the qualities perceived cannot be specified, but merely indicated. It remained to apply the test for determinism to the facts of mind. It was agreed to restrict the term "mental" to its strict meaning of "that which pertains to the subject as distinguished from the object of experience." Accordingly we had to consider two factors — feeling and activity. The different modes of the latter reduce essentially in every case to attention, any particular mode being characterized by the particular type of object attended to. Reverting to Mr Russell's illustration, we selected for criticism the phrase "given n states of mind," which expresses the possibility of the existence of a set of mental determinants of the universe. In the first place, there appeared a difficulty in assigning any definite meaning to ''a state of mind." The oneness of the subject is so absolute as to render meaningless any attempt to split up his feelings and activities into theoretical actions. Secondly, even supposing this obstacle to be surmounted, a fresh difficulty appears when we try to discover how a " state of mind " can in any sense be "given." For it cannot be given in acquaintance even to the subject to whom it refers, seeing that it is not object of knowledge for that subject, although realized by the latter. It must therefore be supposed to be given by description based on that realization. A third difficulty then arises as to the form which the description must take. If it is to be of any Summary and Conclusion 153 use in leading to one of the general laws from which a functional relation may be constructed, or if, given that functional relation, it is to be capable of substitution therein, it must evidently be described in precise quan- titative terms. For this to be possible, quantitative notions must be significant of it, and hence we are finally led to the test which lies in discovering whether feeling and activity can in any sense be considered to be made up of parts. In the case of feeling, i.e., pleasure and pain, it is true that we use the vaguer quantitative terms such as "more" and "less." This, however, is because of a fancied analogy with intensive quantities. Feeling- difference in respect of intensity is purely qualitative. It can in no way be correlated with true, or extensive, quantity, as can intensive quantities. For in the latter case the principle of the correlation is the fact that the conception of both types of quantity springs from a common root, namely sense-data. Quantities such as volume, mass, and energy, having extensive magnitude, are constructions of sense-data. Quantities such as temperature, density, and magnetic permeability, having intensive magnitude, are functionsof extensive quantities, and hence ultimately constructions of sense-data. But we cannot construct feelings from sense-data, and the same consideration applies to subjective activity. Hence there is no basis whereby we can correlate mental factors with any of those intensive or extensive quantities which may be measured by observation. Nor are these mental factors themselves quantities. A pleasure is not the sum of parts which are themselves pleasures. The case is, if anything, clearer still in the case of activity. 154 ^■^^ Notion of a Deterministic System Here the use of the vaguer quantitative terms evidently has a reference to the object attended to, and not to the attending itself, as when we speak of "concentrating" the attention. Moreover, it is obvious that an act of attention is not made up of parts which are themselves acts of attention. Willing is not composed of willings, desiring of desirings, nor thinking of thinkings. Even granted, then, that such a thing as a "state of mind" may exist in any definite sense, the act or feeling in which the state consists is an absolute indivisible unit. Therefore quantitative notions are not significant of mind, and, consequently, the terms " determined " and ''undetermined" are not significant of mind either, nor of any system in which minds exist, considered as a whole. It follows immediately that it is strictly neither true nor false to say that the will is determined. It is mean- ingless, in view of what the will is. In willing we are therefore " free " in the fullest sense of the word. But that freedom is not manifested in chaos. The acts of a man are intelligible to others in virtue, not of a logical, but of a teleological coherence. We have, in fact, not logical determinism, but purposive guidance of pheno- mena. Nor can these purposes and interests be weighed out quantitatively and numerically ear-marked for calculation of the future. Given the end which it is purposed to bring about, we may approximately predict the probable actions whereby the end will be realized. But how to foretell the purpose } It is hidden in the individuality of the man. Man is free — free in his thoughts and aspirations, free in his intercourse with his environment, free to make the best or the worst of what he finds therein. V THE INTENSITY OF SENSE-DATA »!>ENSE-DATA are commonly regarded as possessing both types of magnitude which pertain to quantities, namely intensive and extensive. Not only do we speak of sense-data as differing from one another in their quali- tative aspect, but as regards their quantitative aspect, we say they differ in respect of extensity and of intensity. Now it is an undoubted fact that sense-data possess extensive magnitude. It is on this, indeed, that all the measurements of physical science depend. The latter reduce ultimately to the observation and measurement of spatial and temporal features ; and accordingly the possibility of physical measurement depends on the fact that, for most purposes, the presented object of individual experience can be analyzed into parts (sense-data) standing in spatial and temporal relations, and possessing spatial and temporal characteristics such as size and duration, which are quantities possessing extensive magnitude. The commonest illustration of this is the (relative) measurement of the spatial dimensions of an object by visual comparison with another standard object. Thus extensive magnitude is directly given as a quantitative aspect of visuaP sense-data, though the ^ Sense-data of other types, e.g.^ tactical, auditory, etc., also possess extensive magnitude, though the measurement of this mag- nitude has not been developed to any degree of precision comparable with that attained in the measurement of visual magnitudes. 156 The Intensity of Sense-Data number expressing any particular magnitude will of course depend on the standard or unit chosen. On the other hand, it appears at first sight equally certain that sense-data possess intensive magnitude. We speak of one patch of light as being brighter than another, or of one sound as being louder than another ; and this implication of "more" and "less" is apparently a sufficient criterion of the existence of magnitude. It would therefore follow that the intensity of a sense- datum is a quantity, and should thus be measurable. But at this point difficulties arise. We can correlate the intensity of any given sense-datum with a certain physical quantity having intensive magnitude not directly given, without difficulty. Strictly, however, this affords us no information as to the intensive magnitude (relative or absolute) of the directly given sense-datum itself. The problem is, then, to effect the direct measurement of the latter. Evidently we can only expect to determine the relative intensity, and for this it is necessary to carry out a comparison of different sense-data in respect of theirintensity. With thenumerousobstacles encountered by such an attempt it is unnecessary to deal at length — they are only too well-known. It is not simply the impossibility of fixing a standard which shall be the same for everyone that is the chief difficulty, nor even the impossibility of fixing a standard which shall be the same for the same person at different times. Such obstacles might be surmounted by methods analogous to those adopted in fixing a standard of extensive magnitude which shall be the same for all persons at all times. Methods of this kind consist essentially in certain correlations of sense-data. They must not be The Intensity of Sense-Data 157 confused with methods such as that attempted by Fechner. The principle adopted by him was simply the expression of intensity in terms of a standard intensity. He tried to avoid the difficulty attending a choice of the latter by fixing on the minimum perceptible difference as unit. But not only is this complicated by the fact that the minimum is different for different people ; it also involves the assumption that it is the same for the same person quite independently of the rest of the presented field — quite independently, for example, of the total intensity of the particular sense- datum^ whose change is being observed. For such an assumption there is no warrant whatever. The great difficulty lies, however, not in fixing a standard, but in making use of it when fixed. We cannot say, for example, that one sense-datum is twice as intensive as another merely by comparing them. In the case of intensity there is no ground for such a judgment apart from a standard, but when the latter is chosen the difficulty recurs in attempting to compare other sense-data with it. There seems to be no definite meaning to be attached to such commonly heard state- ments as " This is ten times as bright as the other." If we try to make an exact statement, there appears nothing to go on, nothing to get hold of, as it were, in estimating that the intensity of one sense-datum is so ^ The minimum change in stimulus necessary to give a percep- tible difference in the corresponding sense-datum is proportional, as Fechner discovered, to the total stimulus employed. But he could only infer a quantitative relation between stimulus and sense-datum by assuming the least perceptible change in the latter to be a constant, independent of the actual intensity of the sense-datum. 158 The Intensity of Sense-Data many times that of another. Only in judging sameness of intensity are we on ground that is at all safe. It appears, then, that we have two facts to account for which seem to imply more or less contradictory results. In the first place, we use quantitative phrase- ology in speaking of the intensity of sense-data ; yet, secondly, the direct measurement of the latter is im- possible. Moreover, not only is it a fact that we cannot judge (say) that the brightness of one patch of light is double that of another, it is doubtful whether any real meaning can be attached to such a statement at all. To realize this, it is only necessary to try to think out such a case and to arrive at some definite meaning of the statement ; and, be it noted, a comparative statement of this kind involves no reference to any particular unit. Now most extensive quantities and all intensive quantities with which we deal are functions of two directly given extensive quantities. The latter are length and duration. With these we start, and from them we construct first of all other extensive quantities (area, volume, etc.) and later intensive quantities (density, etc.). The latter are thus without exception functions of extensive quantities. Indeed, physically speaking, we cannot conceive of an intensive quantity which is not a function of extensive quantities. Yet we cannot say this of the intensity of sense-data, for it is something immediately given in experience. True, the intensity of the physical stimuli is a function of extensive quantities, but that is a very different thing. We are therefore faced with two alternatives ; either the intensive magnitude of sense-data is an ultimate given fact, or else what we call the "intensity" of sense- The Intensity of Sense-Data 159 data is not a quantity at all. Now the first alternative is quite hopeless. For the intensive magnitude of a sense-datum cannot be given in itself, nor yet by com- parison with other sense-data. As regards their extensity we can compare sense-data immediately. Not so, how- ever, as regards their intensity; in fact, this, as we have seen, is the root of the whole difficulty and contradiction. It remains, then, to consider the second alternative, namely that the intensity of sense-data is not a quantity at all. Acceptance of this is conditional upon our being able to account for the use of such terms as "more" and " less " in this connection. Yet there seems to be no difficulty about this. We use quantitative phrase- ology simply because, as regards intensity, we can arrange sense-data in a certain order. But this does not imply that intensity is a quantity. Consider the case of a man who should see for the first time. Suppose he is shown two screens of equal size, one dully, one brightly illuminated. Provided associated organic sensations could be eliminated (and this is an important point, though quite impossible to carry out in practice of course), would he arrange the two screens in any particular order of "greatness," being simply told to do that .-* It is doubtful whether he would even realize in respect of what they were supposed to differ in greatness. If, on the other hand, he were shown three screens, he would be able to arrange them in order, though probably without making any assertion as to whether the order were ascending or descending. But upon what, then, is this order based ? We are i6o The Intensity of Sense-Data now committed to regarding the intensity as a quality and we must accordingly suppose the sense-data to be arranged in order of qualitative similarity as regards their intensity, just as we arrange patches of light in order of qualitative similarity as regards their colour. But the basis of such judgments of similarity probably rests on the movements of attention \ In greneral we arrange lights in a certain order because we have previously attended to them in that order, e.g., when the gas is turned up. Or if the differences are not continuous (as in the case of the three illuminated screens), our judgment is based on the amount of continuous change in the accommodation of attention in passing from one to the other. Perhaps the best illustrative analogy of the foregoing is that of the pitch of a note. We talk of higher and lower pitch. Yet pitch is not a quantity although we can correlate it with a pure quantity, namely frequency of vibration. We talk of higher and lower notes simply because in respect of pitch we can arrange musical notes in a certain order of similarity. Now according to the results we have arrived at, exactly the same holds of the loudness of notes. Just as notes of the same loud- ness may differ in their pitch-quality, so may notes of the same pitch differ in their loudness-quality. In other words, brightness, loudness, and all similar "intensities," though commonly regarded as quantities, are really quali- ties. The order in which we arrange them is the order of their similarity, though probably the use of quantita- tive terms depends also to some extent on the occurrence of associated sensations having extensive magnitude. ^ Cf. James Ward, Psychological Principles, p. 329. The Intensity of Sense-Data i6i We have seen, then, that the paradox arising from the impossibility of measuring the intensity of sense- data and from the difficulty of giving a definite meaning to the quantitative statements involved, leads to the raising of the question as to whether this intensity is a quantity at all. Since all the intensive quantities of which we know are ultimately functions of certain immediately given extensive quantities, we are driven to suppose either that the intensive magnitude of sense- data is something directly given, or else that the intensity of the latter is not a quantity at all. Rejecting the former alternative, and accepting the latter, we found that the application of quantitative terms depends here simply on the fact that we can arrange sense-data as regards intensity in a certain order of qualitative similarity based on the movements and accommodation of attention. The recognition of the fact that the intensity of sense-data is a quality, and not a quantity, would do much to clear up certain branches of so-called "psychical" measurement, besides having an important bearing on certain philosophical questions. R. S. p. II VI IMMORTALITY I. Introduction A CLOSE inspection of any one of the traditional prob- lems of philosophy reveals in nearly every case a re- markable amount of confusion. This confusion is generally due in part to verbal ambiguities, but the latter by no means give rise to all the difficulties con- cerned. Usually it turns out that several distinct issues are involved in what appears at first sight to be but a single question. Further consideration shows that as regards some of these issues there is no real problem at all. Others of them are strictly meaningless as generally stated, while to yet others a more or less definite answer can be given. It is intended here to investigate the general aspects of one of these long-standing problems — the problem of immortality. An attempt will be made to state that problem clearly, to disentangle the various issues at stake in it, and thereafter to show where the real diffi- culties lie, which of them are insoluble, and to solve as far as possible those which are capable of solution. Particular points, on the other hand, will not be con- sidered. We shall not be concerned, for example, with speculations as to the probable nature of a " future Introduction 1 63 life " beyond this present bodily one, should there be such a life. The question before us, as popularly stated, takes some such form as: " Do we live for ever? " This state- ment would of course include in its scope not only life after death, but also life prior to birth. From the general point of view precisely the same issues are in- volved in both these cases. But before going further it is necessary to point out a possible difficulty. What of the "we" referred to in the question ? Is it to be supposed that after the lapse of considerable (not to say eternal) periods of time, a surviving individual could in any sense be considered to be identical with the individual as he now is ? Evidently we must hereinafter take further account of this matter, for to all appearances the hand of Time lies heavy upon every man, and change (whether it be growth or decay) is one of the common- places of existence. A further point now arises. Is the question " Do we live for ever .'* " to be identified with the question " Do we exist for ever ? " In other words, is existence for individuals such as we identical with living, that is with experiencing } This may be answered in the affir- mative. We are essentially experiencing subjects. Our existence consists entirely in our experience. Apart from the latter we are nothing. More, however, will be said on this point later. For the present we may be content simply to emphasize the identity of our existence with our experience and to go on to the consideration of what our problem really amounts to when stated as clearly as possible. II- 164 Immortality II. The Logical Statement of the Problem Consideration of the question " Do we live for ever ? " shows that the crux of the whole matter lies in the final phrase — "for ever." A definite meaning must be assigned to this phrase. Before this can be done, how- ever, it will evidently be necessary to make some in- vestigation of the nature of time. In fact, the latter is clearly the master factor in the problem. If we can come to a satisfactory conclusion as to the nature of time, a key will have been found wherewith the gate to the solution of the riddle of immortality may be unlocked. By far the best method of approaching such a task is to carry out a psychological analysis of the growth in the individual experience of that particular type of knowledge which is termed "time-knowledge." It is not proposed to carry out that analysis here. It will be sufficient to give an account of the results at which it arrives and then to estimate the bearing of these results on the main question which confronts us. In the first place, analysis shows that it is impos- sible to conceive of the existence of time apart from objects which are, as we say, "in time." That is, time is itself conditioned by those very phenomena the de- scription of which involves temporal ideas. Although we may construct the concept of time as something which subsists of itself, we cannot suppose that there is really such a thing as empty time in which objects come to have their being. What we find as actually given in experience, is the fact that objects have certain peculiar characteristics, such as duration, whichare signi- The Logical Statement of the Problem 1 65 fied by the name " temporal," and that certain types of relation hold between objects, these types also being sig- nified by that name. In other words, what actually exists is not a self-subsistent time, but sense-data wiikin experience possessing temporal qualities and standing to one another in temporal relations. Time-knowledge is then simply the recognition of these qualities and relations within the totum objectiviim of our experience. The individual subject of experience thus arrives at the notion of time by the perception of temporal qualities and relations. But the time to the knowledge of which he in this way attains is purely private. The peculiar characteristics of the sense-data in each in- dividual experience are incommunicable as such, so that each subject has his own private time. We pass, however, by intersubjective intercourse to the con- struction of the concept of a public or universal time. This is carried out by establishing a one-one corre- lation between the sense-data of any two individuals. We can then speak of an event in A's experience as being before, after, or simultaneous with an event in ^'s experience. We must, however, carefully distinguish the relation thus established between the two events from the relation between two events both of which occur in the experience of the same individual. In what has so far been said, there appears no ground whatever for regarding the former as temporal in any sense similar to that in which we regard the latter as temporal ; nor have we so far any grounds for supposing that the experience as a whole of one individual stands in any temporal relation to that of any other individual. For one thing, it must not be forgotten that the supposed 1 66 Immortality establishment of the one-one correlation between the experiences of two individuals really consists for each in the establishment of a correlation between events in his own experience, and the inference which he makes to the establishment of a correlation between his own experience and that of the other is by no means logic- ally necessary, however irresistible it may be practically. The existence of the subject implies the existence of the object of experience. We meet here a point re- ferred to in the introduction', namely the possibility of the existence of an experiencing individual out of the presentational relation to an object implied in experi- ence. The case of dreamless sleep is sometimes cited as an example of such existence. This particular case will be dealt with in the sequel. For the moment we may confine our attention to generalities. In the first place it is evident that the individual, if he exists at all out of the presentational relation, cannot be described during that phase of his existence as a subject. It thus becomes a question as to whether it is of the essence of the nature of such individuals as ourselves to be subjects of experience. This question may be answered in the affirmative. For us, existing means nothing more nor less than living or experiencing. Take away that presented whole of experience to which all our activities are directed, and those activities (which are themselves partly conditioned by the objects of them) must cease, and ourselves with them. We realize in fact that to abstract a subject from the pre- sentational relation is to make such a fundamental change that he could no longer be considered to be in ' See page 163 supra. The Logical Statement of the Problem 167 any sense the same individual. This is a theoretical way of stating the fact that we are, essentially and existentially, experiencing individuals. The fundamental fact is, then, the unity of the in- dividual experience comprising the duality of subject and object, each of which implies the other. The question "Shall I Hve for ever?" thus becomes "Will there be a time when I shall not exist, i.e., shall not have experience ? " The latter, as we have seen, implies and is implied by the question "Will there be a time when no object is presented to me ?" To assign a definite significance to these questions, we must be quite clear on what is meant by the "time" referred to. Evidently it cannot be my private time, without contradiction. For my private time is con- ditioned by the objects presented to me, in fact it only subsists at all in virtue of these objects. Hence there cannot be an instant in my private time when nothing is presented to me. The two ideas " instant of private time " and " absence of presented objects " are con- tradictory in this context. The questions we have mentioned are thus only significant if the time referred to is public time. We must accordingly ask " Are there instants of public time when no object is presented to me ? " A reply to this in the negative might not rule out immortality in a certain sense. For it might be said that perhaps there are "gaps" in our existence, as it were. The idea of such "gaps" could, of course, only be significant if interpreted in terms of public time. A further question would arise as to the duration (in terms of public time) of these gaps. Are they in any cases infinite ? These considerations need 1 68 Immortality not detain us, however, for the general characteristics of the problem are not in any way modified by them. A final step remains. The concept of public time is but a construction based on the correlation of in- dividual private times. The latter, or rather those characteristics of sense-data which are termed "tem- poral," are the actualities. Our question therefore, in the last analysis takes the form "Are there events in the experience of others which cannot be correlated with any event in my own experience by the one-one correlation referred to above ? " This is the logical statement of the problem in terms of the actual facts, when that problem is viewed from a universal standpoint. The problem propounded in this form would appear to be quite insoluble. We cannot determine from logical considerations whether the required correlation would always be capable of being established (at any rate theoretically) or whether it would break down in certain cases. Nor can we solve the problem empiri- cally. For the failure of the correlation could never be demonstrated in this way. It would not be known to the individual concerned seeing that it implies his own non-existence, and it could not be investigated by other individuals. For, in the first place, since experience is incommunicable we cannot establish the correlation directly, but must always carry out the process indi- rectly by correlating different events in our own experi- ence ; and, in the second place, we cannot assume that an individual does not exist simply because we are not able to communicate with him, that is because certain sense-data are lacking from our experience which wehave been accustomed to regard as manifesting his existence. The Logical Statement of the Problem 169 The insolubility of the problem as stated above is, however, a fact of but little practical importance. For, as the last considerations have made clear, the signi- ficance for the individual of the question put thus, and its answer (whatever that may be), is purely formal and abstract. The absence of the correlation between the subject's own experience and that of other people, could ipso facto not be brought home to him in any concrete way. For that matter, the fact of its presence could not be evidenced concretely to him either, seeing that every correlation which he makes is actually between events in his own experience. To estimate the practical bearinof of the matter we must therefore abandon the universal and conceptual standpoint we have adopted up to this point, and come back to the private stand- point of individual experience. Returning then to the question of immortality in the form " Shall I exist for ever ? " we shall endeavour to bring out the concrete significance of the question and its answer for the in- dividual. It will be found that the results obtained in this direction are more definite. III. Concrete Significance of the Problem FOR THE Individual Pursuing our enquiry in this new direction, we must first of all emphasize a fact which is continually recur- ring throughout the investigation. This is the fact that in experience subject and object each imply the existence of the other. Both begin, continue, and cease to be, together, though whether such terms are here strictly applicable at all is a question for later considera- tion. 1 70 Immortality Now the cessation of the existence of an individual could evidently not be a given fact for any other in- dividual. We can give a meaning to the idea of a broken intercourse between subject and subject, for events occur in our experience which can be interpreted in that way. But no event in our experience gives (nor could it give) us certain information of the fact that an individual with whom we no longer have intercourse has ceased to exist. In other words, the ceasing to exist of another individual cannot possibly be given as an object of our acquaintance any more than that individual himself can be given as an object of our acquaintance. There is a further and far more important point. Not only could the ceasing to exist of an individual not be a fact for other people, but it would also not be a fact for that individual himself For his experience, in which consists all those events which are facts for him, comes to an end with himself. What is the status of a fact which is not a fact for anybody, and whether there could be such a fact, we need not pause to enquire. Meanwhile, the conclusion so far reached may perhaps be most easily realized by taking up in imagination a solipsistic position. Such a position is a perfectly possible one to adopt theoretically ; it cannot be logically refuted. The events in the experience of an individual take place y?^^^ as if\\^ were the only existing subject^ This brings out the difficulties involved in assigning a" definite meaning to the phrase "ceasing to exist." For ^ The apparent manifestation of the existence of similar subjects can be interpreted on the analogy of dreams, if we stick consistently to the solipsistic position. Cojicrete Significance of the Problem 1 7 1 it would have no meaning for the individual subject concerned, and if he were the only subject existing it would therefore have no meaning at all. It follows from the necessary co-existence of subject and object that there are no "gaps," as it were, in the object of experience. Forthe subject cannot beconscious of nothing. Any such gap in the object of experience would imply a corresponding gap in the existence of the subject. We have no reason whatever to postulate these gaps, for in any case their presence could not possibly be known. They would not be a fact for anybody. The case of dreamless sleep (if there be such a thing) illustrates the point. Dreamless sleep for the subject concerned means nothing more nor less than the occurrence of an unusual sequence in the object of experience. But there is and can be no absolute break in the sequence. We can draw no inferences of any kind from dreamless sleep as regards ceasing to exist, without making some assumption as to the existence of time as an actual entity, for which we have no grounds whatever. For, as we have seen, the actual fact is the existence of temporal qualities and relations within experience and not of some distinct entity called Time. This last point contains the crux of the whole matter, namely in actual fact it is only objects within the individual experience which can in any concrete sense be said to exist at times (or places), these being conditioned by those very objects. The idea cannot be applied to experience as a whole. For the individual the fact is simply his existence — not his existence at such-and-such a time or at such-and-such a place. The 1 72 Immortality latter has noconcrete meaning forhimatall. Accordingly, all such terms as "begin," "continue," "cease," which have a temporal reference, are really applicable only to things within the object of experience and not to experience as a whole. We arrive, then, at the result that for the individual the question "Shall I exist for ever?" is devoid of any concrete significance. We might have expected some- thing of the sort from the fact that whatever interpre- tation we attempt to put on the idea of ceasing to exist, the latter would be nothing for the individual concerned. Putting aside, therefore, the question " Shall I exist for ever?" we can only replace it, if at all, by the question " Do I exist ? " — which is answered in the very asking of it. The conclusion we have come to makes it clear that all particular questions commonly connected with the problem of immortality (such, for example, as the question of a future life out of this body) should strictly be put in the form " Are there such-and-such elements in my experience ?" no temporal mark of any kind being attached to the proposition. In other words, ultimately we must consider experience as the unity it actually is. Ultimately we cannot regard experience as made up of parts termed "sense-data" or "events." Experience is not a series, but an indivisible unity. What is com- monly represented and referred to as its " unique " continuity, is really its tmity. Before summarizing, therefore, it will be worth while to consider this point briefly, for, in the first place, it is of the utmost import- ance in all philosophic thought directed to final problems, and, secondly, it draws together all the threads in the Concrete Significance of the Problem 173 particular problem we have been considering. In the light of it our results receive their fullest interpretation and justification. IV. The Unity of Experience The employment of analysis in any given case implies that the ideas of whole and part can be adequately applied in the matter under consideration, for analysis consists essentially in the discovery of the " parts " (whether terms or relations) of a given "whole." In the course of philosophic enquiry we sometimes have to deal with unities to which the ideas of whole and part are not ultimately adequate. Such unities are by their very nature indivisible entities. We find among unities of this type, the subject and the object of individual experience. It is frequently granted that the subject is an indivisible entity, but the same is far more rarely admitted of the object. We shall therefore confine our attention chiefly to the latter. It is necessary to mention at the outset that no attempt is being made to demonstrate that an utter falsification is committed by analysis of the object of experience. But we must recognize degrees of validity ; and the analysis of experience cannot be entirely valid. The ideas of whole and part apply with different degrees of adequacy to different entities, and the degree in which they apply to the object of experience is not perfect. Now it is sometimes said that the object of experience is complex and therefore demands analysis. But surely 2l petitio principii is involved here. For what is meant 1 74 Immortality by saying that an object is complex ? Simply that so soon as we attempt to reflect critically on it, or, in other words, to analyze it, we find that we cannot make a single step without introducing the concepts of whole and part. Now it may be granted that this is true of reflection on the object of experience, but in actual experience apart from any judgments referring to it (except perhaps the existential judgment "It is" the "it" being the totum objectivum) there are no "whole" and " parts," but just a given indivisible unity. Con- sidering for the moment, in imagination, that group of objects which we class as " simultaneous parts " of ex- perience (what is sometimes, but perhaps illegitimately, called an " instantaneous " section of experience), we have in the actual experience not a whole of parts but an indivisible unity. Taking a further step, and including within our view what analysis calls the " time-series " of events, we must again conclude that there is not actually a whole consisting of parts which are events, but an indivisible unity. Accordingly, the individual experience must ultimately be taken account of as just such a unity. The point may be made clearer by asking what is meant by the term "part." It is said^ that the whole is an entity and the parts are other entities which exist or subsist independently of their discovery by analysis, and are real in quite the same sense as the whole of which they are parts. This seems a fair statement of the matter. But if the parts are definite entities which are real in quite the same sense as the whole, they must ^ See the introductory section to "A Defence of Analysis" by E. G. Spaulding in T/ie New Realism. The Unity of Experience 175 surely be capable (at any rate in conception) of existing or subsisting apart from the whole and at the same time maintaining their nature unchanged. This, however, is the root of the whole difficulty. For in the case of a unity such as the object of experience, we find, on considering the supposed "parts," that we have to conclude that the parts are "coloured," as it were, by the whole and their relations to it ; that it is essentially of their nature to be parts of that whole ; or that the whole exists in and through the parts, while the parts exist in and through the whole. To such statements it would be difficult to assign a definite meaning while still maintaining the clear significance of the ideas of whole and part. The truth is, of course, that we are discovering the inadequacy of those ideas to the indi- visible unity of experience. I f we press them far enough in this connection, contradictions are bound to emerge. What we have said applies most obviously to the case of sense-experience. But we must beware of thus limiting the application. The indivisible unity we have been considering is really what may be called the objective situation which we analyze into objects of sense, objects of thought, etc., though strictly, as we have seen, we must regard it as in no way divisible. Much of the disputation over the matter we are dealing with has arisen in connection with what is called the " continuity of experience." The exponents of analysis^ regard the latter as akin to mathematical continuity, analyzing experience into continuous (or, at least, compact) series of elements termed "sense-data." ' See, e.g., B. Russell in Our Knowledge of the External World, Lect. V. 1 76 Immortality Their opponents, while attacking the analysis, admit some sort of continuity in experience, and in endeav- ouring to define it have to assume what they are trying to disprove. This illustrates our point excellently, for the cause of the trouble is that experience does not possess continuity at all (seeing that it has no parts) but unity, and the common reference to the " unique " continuity of experience is nothing more nor less than a tacit admission of its unity\ Finally, it would be well to insist again on the point mentioned at the beginning of our consideration of analysis. Analysis of experience is by no means entirely invalid. It is a necessary and justifiable mode of procedure, provided its limitations and presuppositions are kept in mind. But we must recognize that there are degrees of validity governed by the adequacy of the ideas of whole and part to the matter in hand. In the case of the object of experience, analysis into parts (sense-data, etc) standing in certain relations in virtue of which they form continuous series, provides a close approximation to the actuality. This follows from the fact that the results of calculation based on the analysis can be verified empirically, within limits. But it must not therefore be assumed that the ideas of whole and part are perfectly adequate here, for in any case the results of calculation would approximate to actuality in the same way and to the same order as do the data (yielded by analysis) from which the calculation starts. If we press on far enough, however, difficulties arise ^ The writer has dealt with this question in the essay entitled "Scientific Method in Philosophy and the Foundations of Pluralism," Section VI, The Unity of Experience 177 which make it clear that the conception of the object of experience as a whole of parts consisting of terms and relations in which the terms stand, is not an entirely adequate one. Ultimately therefore, we must consider the object of experience as the indivisible unity it actually is. The bearing- of these facts on the problem of immortality now begins to be apparent. The unity of the individual experience comprises the duality of subject and object each of which is an indivisible unity. Not only then do time-relations apply in any case simply to parts within the object of experience as opposed to that object of experience as a whole, but the judgments on experience which yield analytically these parts and their time-relations are not ultimately adequate. Hence we have absolutely no grounds whatever for postulating a Time in which experiences actually exist — and this in addition to the fact that even were our analysis quite valid, we should have no reason at all to assign to experience as a whole the temporal characteristics and relations of the parts. Consequently neither of the terms "begin" and "cease" has any valid significance when applied to experience considered as the indivisible unity it actually is. Such questions as "Shall I exist for ever?" need therefore no longer trouble us, as they have ultimately no meaning, and the related question which replaces them, namely "Do I exist?" and which is certainly valid, provides its answer, as we have said, in the very asking of it. R. s. p. 12 1 78 Immortality V. Summary and Conclusion The problem of immortality, no less than the other traditional problems of philosophy, exhibits when analyzed a number of distinct issues which have become confused in most treatments of the question. Accordingly it was considered necessary to obtain a clear statement of the problem to find which of the supposed issues at stake are really significant, and to determine those points on which a definite solution may be reached. In the first place, we found that in order to state the problem clearly it was necessary to assign a definite meaning to the phrase "forever," in the question "Shall I exist for ever } " The temporal implications of the phrase required a brief investigation of the nature of time, which led to the result that in actual fact there is not a definite independent entity, "Time," in which objects have their being, but certain unique character- istics (named ' ' temporal ") of the objects within individual experiences, and certain unique (temporal) relations subsisting between those objects. On these temporal qualities and relations the concept of the private time of the individual is based. By one- one correlations between "simultaneous" events in the private times, public or universal time is constructed. We therefore found it possible to obtain a logical state- ment of the problem of immortality involving the idea of public time, and running as follows : " Are there events in the experience of others which cannot be correlated with any event in my own experience by the one-one correlation referred to.'*" Now since the public time implied in this statement is simply a construction Summary and Coitclusiojt 1 79 based on the individual private times which are the actuahties, the statement is necessarily formal and abstract in significance. Consequently the conclusion to which we were compelled, namely that the problem as thus stated is insoluble, loses much of its importance. It remained to come back to the standpoint of the individual and thence to determine the concrete signifi- cance for him of the question in its original form : "Shall I exist for ever?" Now the subject and object of experience imply each other, so that any assertions regarding the existence of the former apply also to the latter. This being so, together with the fact that the actual entities on which the private time of the individual is based are the temporal qualities and relations of the parts within his object of experience, we were led to the resultthattheabove question has no concrete significance for the individual at all. For there are no g^rounds whatever for asserting temporal characteristics of ex- perience as a whole. Such terms as "begin," "continue," "cease," are therefore meaningless when applied to that whole. Finally the central principle which binds together the various aspects of the question, was seen to lie in the unity of experience. The analysis of the object of experience into parts consisting of terms and relations, though a close approximation and a necessity in thought and discussion, is not altogether valid. The ideas of whole and part are not altogether adequate to the object of experience. For the latter is by its very nature an indivisible unity, being, as it is, presented to one subject, that subject itself being an indivisible unity. A fortiori, then, we have ultimately no reasons whatever for 12 — 2 1 80 Immortality assigning temporal characteristics to the indivisible unity of experience. The question "Shall I exist for ever?" must be replaced, if at all, by the question "Do I exist ? " We may conclude then, that questions having the most practical significance for the individual are not those relating to his own existence, which concern his experience as a whole, but those dealing with events within his experience. Such, for example, is the question (usually connected with immortality) which refers to the possible meeting beyond this bodily life with other personalities to whom we are attached, and who have passed on before us. Formally, this question would be of the same type if it referred merely to the possibility of meeting again in the future friends who have gone abroad, though the particular determining conditions would of course be different — for example, we should have in the first case to take into account the objective factor of bodily death. No such questions are touched by the general aspects of the problem we have been considering, nor can these aspects possibly shed any light on them. But the consideration of the generalities of the case at least serve a clarifying purpose, and rid the particular problems of elements which, although commonly interwoven into the investigation of those problems in a more or less vague manner, are really quite irrelevant to them. VII THE RELATION OF MIND AND BODY . I. Introduction Ihe exact nature of the relation of mind and body is a problem which, although it has occupied the attention in a greater or less degree of the speculative thought of all ages, has come into particular prominence during recent years. The behaviour of those remarkable and unique complexes (unique, that is, in the physical order of Nature) to which the name "organisms" has been given, has been studied in the past by two schools of scientists from two distinct points of view. On the one hand there are the biologist and the physiologist, the latter of whom investigates the individual organism, while the former considers rather the behaviour of species — organisms more or less en masse. Both physiologist and biologist regard behaviour from the objective standpoint of physical science. The material with which they deal is essentially phenomenal. On the other hand stands the psychologist. He is concerned with what is known as the mental "aspect" of the organism. Accordingly his point of view is largely subjective, and this notwith- standing the attitude and methods of recent experimental psychology. Moreover, many (indeed most) of the facts which it is his business to investigate cannot be regarded as phenomenal in the valid sense of that term. 1 82 The Relatiojt of Mind and Body A large part of physiology's task consists in the ob- servation and description of the functions of the sense- organs, brain, and central nervous system. It traces the path of afferent nerve-currents from the periphery to the sensory centres, and the connection of the latter with the higher centres. Conversely, it observes the passage of efferent impulses from the higher centres to the motor centres and thence to the muscles. Or again it investigates the reflex arc through sensory and motor centres without intervention from the higher centres. But, in any case, at the nerve centres of the brain and central nervous system the physiologist comes to a standstill. He does not proceed to consider how impulses "pass over" into the mind and become content of con- sciousness, nor how impulses originate in the mind and "pass over" into the brain as the source of bodily acti- vity. For him, at least in so far as he is concerned with his own particular affair, consciousness is epiphe- nomenal — a mere accompaniment of neural and cerebral phenomena. The psychologist, on the contrary, approaches the question of behaviour from the mental side. His busi- ness is to analyze the individual experience, and to elucidate in particular the subjective factors which determine the growth of that experience. As long as he keeps strictly to his appointed task the concepts of physiology do not come within his scope at all. So far from being epiphenomenal, consciousness is for him the primary concern, while its neural and cerebral concomi- tants have no interest for him, as psychologist. Evidently, then, there is a somewhat ill-defined region lying between the realms of physiology and Introduction 183 psychology with which neither is strictly concerned, a region whose fundamental fact is the mysterious nexus of mind and body. Consequently, this region has been at all times the hunting-ground of the philosopher. As a hunting-ground, however, it can hardly be said to have been happy, for philosophy has at best made but a poor show in its attempt to arrive at a satisfactory solution of the problem of the relation of body and mind. Until comparatively recently, however, this failure has been of no great importance, except, perhaps, to philosophers. But in the course of time both physio- logist and psychologist have pushed their respective investigations so far that the categories now employed cannot hope to be interpreted adequately without a satisfactory clearing up of the debatable region referred to. Moreover, such a clearance would possess the ad- ditional importance that it could not fail to shed light on the question as to which is the most fundamental standpoint — that of psychology or that of physiology. Physiology and biology, in alliance with physics, would, if possible, interpret the Universe ultimately in terms of matter; while psychology, when it becomes meta- physical, seeks to interpret the Universe in terms of mind. The issue is fundamental for the philosopher. If we enquire how it is that the efforts of philosophy to solve this problem of body and mind have not been attended by greater success, we find one of the chief obstacles to that success to be of a type which occurs only too commonly in philosophical speculation, namely the lack of a precise statement of the problem. A perusal of the works of various writers on the subject reveals the fact that the terms "body" and "mind" are 184 The Relation of Mind ajtd Body used in many different senses, senses which are fre- quently vague and ambiguous to the last degree. It is therefore essential, in the first place, to arrive at an exact definition of these terms. Now it is doubtless true that "body" and "mind" are used with more than one meaningr to which a rea- sonable significance may be attached. The problem and its solution will vary accordingly. This calls for con- sideration, but it will appear that there is one interpre- tation of the problem which can claim to be the most definite and significant. After attempting to make this clear, it is intended to approach the problem from the standpoint of a pluralistic spiritualism with a view to obtaining as satisfactory a solution as possible. The general grounds for such an ontological hypothesis will not be advanced, but the latter will be left to be judged by its fruits in this particular case. II. True Meaning AND Implications OF THE Problem It was formerly an almost invariable custom to regard the relation of body and mind from a frankly dualistic point of view. Nor is this custom by any means a characteristic only of discussions in the past. The dualistic standpoint is adopted, whether tacitly or avowedly, in a considerable portion of the current writing on the subject\ The body is regarded as a material thing in an environment of other material ^ See, e.g., Dr Bosanquet's Principle of Individuality and Value, Lect. V, in which, although the general position is idealistic, the concepts employed and the corresponding terminology are full of implications which are thoroughly dualistic. True Meaning of the Problem 1 85 things, some organic like itself, others inorganic, while the mind is apt to be identified with the content of consciousness. The question then becomes one of the relation between a particular portion of organic matter, and the particular content of consciousness connected with it. Sense-data are thus regarded as mental, and an important part of the investigation consists in an attempt to determine the connection between sense-data and their material stimuli. Now for those who take up this attitude the concept of matter differs little from that framed by Descartes. Material things are those which obey the laws discovered by physics, and matter is vaguely conceived as a "substance" possessing such qualities as inertia, im- penetrability, and extension. The conception of mind in this connection is also characterized by considerable vagueness. The individual mind is regarded more or less as a bundle of sensations, thoughts, feelings, and volitions. The development of this view is commonly supposed to lead to certain difficulties. Two of these stand out pre-eminently and are worth a brief consideration. In the first place it is urged that mind and matter as thus conceived are such utterly "disparate" entities that the notion of their interaction is inconceivable. Now it may be granted that mind and matter are disparate, but it does not follow necessarily that their interaction is inconceivable. For there seems to be no reason why any two entities should not interact, unless "action" is defined in some particularly narrow way. But it will perhaps be pointed out that whenever a material body is observed to change its state the change is found to 1 86 The Relation of Mind and Body be due to another material body. This evidently limits the notion of action to the ground of the sequences which occur in the material world. Moreover the state- ment is not strictly true, and in fact begs the very question at issue. For it cannot be said, for example, that the movements of organic bodies are invariably observed to be due to purely material causes. In the second place, it is sometimes said that the view we are considering involves a violation of the principle of the conservation of energy. The haziness of the notion entertained by many philosophers of the conservation of energy is truly remarkable. As a matter of fact it is a physical principle which only holds within extremely strict limitations, namely that the systems within which energy is conserved are isolated material systems. Now the body of the organism is not an iso- lated system, in so far as the organism has a mind which, at least privta facie, influences its body; and if we consider the organism as a whole, it is not a material system, for it comprises mind as well as matter. The principle of the conservation of energy is thus utterly irrelevant to the question at issue. Even could it be shown (which it cannot) that the energy of the body, after allowing for the effect of external material influ- ences, conforms to the principle of conservation, nothing of interest would follow. Controversy rages round the idea of mind as the di7^ector of bodily energy, and it is objected that the direction of energy involves in practice the performance of work, which would be manifested as additional energy in the body. But we can only assert that changes in the distribution of energy in a system by material influences imply the performance of work by True Meaning of the Problen^ 187 the latter. As to the possibility of changes in distribution of energy due to mental influences, and whether if there are such occurrences expenditure of work is necessary, we can arrive at no conclusion at all. We are thus bound to conclude that objections such as the above cannot apply to this view of the interaction between body and mind. But the theory fails rather for negative reasons. It does not carry us one step to- wards the solution of the problem, for it is quite unable to give any intelligible account of the interaction it talks of. It is compelled to fall back on the assertion of the probability of an exact one-to-one correspondence be- tween events in what it calls the mental series, and events in the neural or cerebral series. As to the ground of this correspondence, it is silent. In such a psycho- physical parallelism all views based on dualism find their logical end ; and whatever may be said of psycho- physical parallelism as a methodological principle, as a metaphysical theory it is an utter failure. The reason for this failure lies in the hopeless artificiality of the dualistic standpoint. Our next concern must therefore be to take up a position in which we shall escape this artificiality. The life of any being such as ourselves consists in the sensations, feelings, desires, thoughts and acts which go to make up the routine of its existence. All these we group together under the term "experience." The individual experience is a unity, but it is a unity which comprises a duality, for in it are distinguished two fun- damental factors. A subject who attends or cognizes, and an object which is attended to or cognized. This duality in unity of experience is the only natural 1 88 The Relation of Mind and Body position that can be taken up when we begin to philo- sophize. Not only is it natural, it is also inevitable. For each of us is one of the factors in an individual experience, each of us is an experiencing subject. The necessity of adopting this position has been called "the ego-centric predicament^" Certainly the position is ego-centric; but it can hardly be called a pre- dicament in the usual somewhat deprecating sense of the term, so long as we keep clearly in mind the fact that our position is ego-centric, and refrain from making invalid inferences from the necessity we are under of adopting it. The realist urges us to try to adopt an "external" point of view, as it were, in our attitude towards the entities with which science deals. He regards our cling- ing to the ego-centric position as a fruitful source of fallacies. So it is, if we forget the limitations of the position, but not of the kind of fallacies which the realist usually imagines. For what, in fact, do sciences such as physics, chemistry and biology, actually dis- cover ? Simply the occurrence of certain sequences in the sense-data presented to us. The discovery of a physical fact lies in perceiving that (say) A is invariably followed by B, and the verification of hypotheses built on discovery is always an appeal to the perception of some sequence of sense-data. Now, according to the realist, we must drop saying "' B is 2XsN2.ys perceived to follow A',' and say simply "i9 always follows A" If the realist is right, the existence is implied of entities essentially akin to sense-data, except that they are not perceived. With the possibility of this we are not here particularly concerned. But at least we are on surer ^ See R. B. Perry in Present Philosophical Tendencies, pp. 129 ff. True Meaning of the Problem 189 ground in making the statement " Because B has always been perceived to follow A, it is extremely probable that, with appropriate attention, B will always be/^r- ceived to follow A,'' than in making the statement "Because B has always been perceived to follow A, therefore B will always follow A\ A, B, and their se- quence being objects quite independent of perception." Now of course we cannot deny the possibility of the existence of things which have no influence whatever on us, and whose existence can therefore never be manifested to us in perception. The realist seems to feel that in taking up and maintaining the ego-centric position, we shall certainly deny the possibility of the existence of such things. But why should we? In any case, such things could have no concrete significance for us, and provided we are willing to admit the possi- bility of their existence, and provided also we refrain from denying (since it will not here concern us) that such objects as sense-data may exist unperceived, no exception can be taken to our ego-centric position, nor shall we be committed to any fallacies thereby. In fact, the reverse is the case. For if, with the realist, we as- sume the existence of unperceived objects otherwise essentially like sense-data \ and proceed to act on that assumption, the burden of proof is on us, and a very heavy burden it is. It is strange, in view of these con- siderations, to reflect that the realist is the first to urge that facts must be ascertained empirically and not a priori. ' It is doubtful whether such a phrase as this has any real mean- ing. See the essay on "The Philosophical Problem raised by the Weber-Fechner Law," Section III. 1 90 The Relation of Mind and Body The ego-centric position is, then, the standpoint we must take up ; for not only is it the position in which we actually are, but also so long as we are dealing with entities which are in fact perceived (whether the per- ception of them is necessary to their being or not) the position is an unexceptionable one, from which it be- hoves us to discover as much as possible. And, indeed, the adoption of any other point of view carries with it certain assumptions, as we have seen, which constitute, if not insuperable, at least very great difficulties. We must now proceed to consider what we really mean when we talk of "material things." The objective field presented to each percipient subject comprises certain shapes, patches of colour, etc., which are at rest or in motion, and between which certain spatial and temporal relations subsist^ They are termed "sense- data." As we have seen, the investigations of physical science really consist in determining the sequences which occur among sense-data. But as soon as the naive stage is passed, physics leaves the immediate data of sense, and proceeds to work with more abstract con- ceptions such as "material body," "molecule," "atom," etc., and, in so far as space and time are concerned, "point" and "instant." The physical concept which is psychologically primitive is that of a material body, and ^ The writer holds that the presented object of experience, as actually given, is an indivisible unity to which the concepts of whole and part are not entirely adequate. But in any reflection on experi- ence whatever, we are necessarily bound to use those concepts, and we can in fact approximate very closely to actuality by a proper employment of them. This is dealt with more fully in "Scientific Method in Philosophy and the Foundations of Pluralism," Section VI, and in "Immortality," Section IV. True Meaning of the Problem 191 the physicist now regards sense-data, such as colour and shape, variously as properties of material bodies or as effects on us of which material bodies are the causes, — nor is he always quite clear as to what he really believes in this respect. At all events, when the physicist talks of "this square, brown body," he does not generally be- lieve that the body itself is the square, brown patch in his object of experience, but rather something to which the squareness and the brownness are somehow attached as qualities or effects. In other words, he tends to conceive the "matter" of which material bodies are es- sentially composed as being some mysterious substratum which causes our sense-data or of which the latter are the qualities. Now at this point the fact must again be insisted on that physical science consists fundamentally in the empi- rical observation of sense-data and their sequences, and in the construction of hypotheses which are tested by appeal to the perception of certain sense-data. Hence it follows that the laws and hypotheses of physics, no matter what terms be employed in them, must really be statements about sense-data. Thus concepts such as point, instant, material body, and material particle, must actually be logical functions of sense-data if they are to be valid, while matter is itself such a logical func- tion, namely the class of material things. I n other words, the material entities spoken of in physics must be re- garded not as inferences from sense-data (namely the causes or substrata of the latter), but as constructions of them\ ^ The writer considers this to have been unanswerably demon- strated by Bertrand Russell in Our Knowledge of the External World, 192 The Relation of Mind and Body Care must be taken to avoid a certain confusion here. As will be seen later, it seems necessary to postulate a ground of our sense-data. But the point to be urged is that this ground is not what is meant by the matter and material bodies of science and common-sense. The propositions of physics, when analyzed, contain no indications of the existence of entities beyond the immediate data of sense. The terms of those propositions, though they are not in general the data themselves, are functions of them ; and the assumption of concrete entities corresponding to these functions, in addition to the data of which they are functions, is a gratuitous and unnecessary one for which there is no reason whatever. This fact, however, is quite irrelevant to the possibility of the existence of a ground of the data which is in no sense akin to the entities with which physics deals, and to which, indeed, physics has no reference at all. Using the term "matter" in the physical sense, then, it appears that the body as a particular material object, is a certain class of sense-data. This class includes among its members both the sense-data of the individual to whom the body belongs, and those of other individuals by whom that body is perceived. Now there is a certain similarity and a certain difference between these two sets of sense-data. The sight of our own body (as regards those parts which we can see) is essentially similar to the sight which other people have of it. The same holds of touch, with the reservation Lects. Ill and IV, and The Ultimate Constituents of Matter xwd^voXwrnt. of essays entitled Mysticism and Logic. The above is a condensed statement of the view therein set forth. True Meaning of the Problem 1 93 that in exploring our own body we are sensible of that double touch whereby we distinguish it from other bodies. But when we come to musculo-motor and organic sensations, together with that vague presenta- tional mass which goes by the name of "general sensibility," we are concerned with sense-data to which, in connection with oitr body, there is nothing corre- sponding in the presentations of other people ; though, of course, similar objects are presented to them in con- nection with their own bodies. On the other hand, there are certain sense-data (corresponding, for example, to our brain) which may be perceived by other people under suitable conditions, but which are never perceived by us. It appears, then, that the class of sense-data which make up the body of an individual in the physical sense of that term comprises three sub-classes : ( i ) Sense-data which, though of course peculiar to the individual, resemble essentially the perceptions which other people have of his body. (2) Sense-data perceived only by the individual concerned. (3) Sense-data perceived only by other individuals. Sensations of sight, organic sensations, and observations of the brain, are typical respective examples of members of these three sub- classes. Evidently it is members of the first two only that enter concretely into the presented complex which constitutes the body as a physical object for the individual concerned. But the latter in constructing the concept of his body includes in addition the third type, basing his construction on his observations of the brain and internal arrangements of other people, whether he has actually perceived these or merely representations of them. R. s. p. 13 194 The Relation of Mind and Body We must nowleavefor the moment our consideration of the body, to enquire what, from the standpoint we are adopting, can be meant by "the mind," By re- garding sense-data as the elements of which material things, the organic body among them, are composed, we have ruled out from the concept of mind, either actually or analogically, all that is contained in the objective side of experience. We are thus left simply with the purely subjective factors in experience, of which there are two — feeling and activity, the latter consisting chiefly and perhaps entirely in attention. In determining the relation between body and mind, where body is interpreted in the sense we have been con- sidering, we must therefore identify the individual mind with the individual feeling, acting subject of experience. By this is meant the concrete Ego, not the empirical Ego or Me\ The latter is but a concept constructed by the concrete Ego — a presented object among other presented objects. What, then, is the relation between the subject and his body regarded as a physical object, when the latter is interpreted in the only valid way ? To this question a definite and simple answer can be given. As we have seen, for the individual, his body, as a definite material object, consists in a certain presented complex of sense-data which differs in some important qualitative respects from other presented objects. Now the relation of this group of sense-data to the individual subject whose body it is, is a presen- tational relation. Thus, for the individual concerned, the relation of body and mind, where the former is regarded as a physical object and is interpreted as such ^ See William James, The Principles of Psychology, Chap. x. True Meaning of the Problem 195 while the latter is identical with the subject of experience, is simply a relation of presentation, the latter being an ultimate relation, the nature of which we all realize. But if we take a more objective standpoint, and include in the body of the individual the sense-data connected with it of other individuals, we are faced with the further question as to the relation of these other sense- data to the first individual, who is the "mind" considered. It seems doubtful whether a satisfactory answer can be given to this question so long as we continue to consider the body simply as a physical object. We shall now, however, proceed to give a more concrete meaning to the term " body" than it possesses when it refers (as it must do in the so-called "objective" sciences) to a certain class of sense-data. It is not an easy matter to settle the exact onto- logical status of the objects presented to us in perceptual experience. In attempting to get beyond these im- mediate data, we must beware of employing the categories which we use in connection with events occurring within the object of experience. For while, if we consider the way in which knowledge has grown, we find that many of these categories in the fullest significance they have for us are applicable more widely than to the object of experience alone, this is not always the case. The category of causality, for example, as used in physical science, is strictly applicable only within the object of experience, although in the deeper meaning that we find it to possess for such individuals as ourselves if we investigate its basis as a concept (namely, efficient activity), it implies a reference beyond the immediate data of sense, the latter being 13—2 196 The Relatio7i of Mind and Body in certain respects capable of interpretation in terms of it'. The matter we are now considering may perhaps be best approached indirectly. In the first place, it would appear that no entity which can be said to be existent in any really concrete sense of that term, can be an object of immediate acquaintance. For the essential nature of such an entity — the entity as it is in itself'^ — could never be realized by any other entity, just as an individual like ourselves, for example, cannot know any other individual as he really is in himself. It follows that sense-data, being objects of immediate acquaintance, cannot strictly be regarded as existent. In particular, they are not existent in the sense in which the percipient subject is existent. But they have being of some kind — they are there, that is, they are presented to a subject. Secondly, we must enquire whether sense-data, though not themselves existent entities, have their ground in existent entities. If we deny this to them we are forced to regard them as purely subjective modifi- cations or states (in which case we might of course say that their ground was the subject itself and no other entity) and are logically carried on to solipsism. The latter position can no doubt be maintained in such a way as to be logically irrefutable. But it is to be rejected on grounds of a priori improbability and philosophical ^ See the essay on "Scientific Method in Philosophy and the Foundations of PluraHsm," Section VII. - This is dealt with more fully elsewhere : ibid., Section V, and in "The Philosophical Problem raised by the Weber-Fechner Law," end of Section III. True Meaning of the Problem 197 sterility. We must therefore postulate a ground of our sense-data, in existent entities other than ourselves. Adopting this position, then, we infer that the existence of certain entities in certain relations to a subject, involves the perception by that subject of certain objects termed "sense-data," which may also be called the "appearance" of those entities to that subject, thereby defining the term "appearance." Hence we must so far regard appearance as an ultimate mode of being pertaining to the data of perception and to be carefully distinguished from existence, which is another ultimate mode of being. The existence of the entities referred to is manifested to us in the sense-data which make up the object of perceptual experience, and our interaction with them is manifested to us in the pro- gressive differentiation of that object as experience advances. We are left with the question as to the nature of the entities concerned, and to the consideration of that we must now turn. III. The Pluralistic Interpretation of the Matter In rejecting solipsism, and postulating accordingly the existence of entities other than ourselves as the ground of our sense-data, we are faced at once with a possibility which is, to say the least, discouraging. It is that the essential nature of these other existents is utterly unlike that of anything about which we have knowledge. They may be quite unknown and un- knowable to us. Many philosophers have arrived at some such con- clusion in the past. Herbert Spencer's "Unknowable," 1 98 The Relation of Mind and Body and the " things-in-themselves " of Kant, are striking examples of this. And it must be admitted that there seems no way of disproving this alternative, at any rate a priori. On the other hand, we are by no means logically compelled to it provided we can justify em- pirically the other alternative of interpreting sense-data as the appearance of entities similar in nature to existents about which we have knowledge. Now each of us indubitably realizes the existence and nature of one such entity only — himself, and has knowledge about this entity contained in propositions based on that concrete realization. But we may proceed to assume that sense-data are the appearance of individuals funda- mentally akin to us in that they are subjects of experience, and differing from us only in the indefinitely various levels of experience to which they have attained ; and we may apply this hypothesis not only to the existence of other people, not even merely to the realm of organic matter in general, but to inorganic matter as well. It is in such an ontology that pluralistic spiritualism consists, and its justification must be looked for in empirical verification. It is no part of our purpose here to advance the general arguments which support this hypothesis\ Suffice it to say that pluralism accounts satisfactorily not only for the behaviour and development of organisms in general from the amoeba up to man himself, but also for the apparent lack of spontaneity and the prevalence of fixed routine observable in the inorganic world. To pluralism we shall therefore have recourse as an hypo- ^ See "Scientific Method in Philosophy and the Foundations of PluraUsm." Pluralistic Interpretation of the Matter 199 thesis which may help us towards the solution of the problem of body and mind, and the first step must lie in the restatement of the problem in terms of that hypothesis. The object of experience, as actually given in perception, is an indivisible unity. But in any reflection on experience we have to regard this object as a whole analyzable into parts, the synthesis of which does in fact approximate very closely to the given indivisible unity. Now there must be an interpretation of the fact that this analysis, whose results approximate so closely to actuality, yields parts (sense-data) of various kinds and standing in various relations. There must be something corresponding to each kind of sense-data or group of sense-data — something which determines that analysis, though not entirely adequate in this case, should, so far as it goes, distinguish this particular kind and no other. Among the sense-data presented to each one of us there is one group which is marked by the fact that there are always some of its elements given in perception. This complex is made up of certain sensations of sight and touch, together with organic sensations which can be correlated with them and which are said to make up what is called "general sensibility." We come to regard this complex as peculiarly our own, and, as we have seen, the concept of our own body is based on it. Not only does it differ from other groups of sense-data in the fact that some or other of its elements are invariably presented to us, but also in the fact that certain of its elements (organic sensations, for example) have a peculiar character which is quite unique. 200 The Relation of Mind and Body To the subjects of experience which constitute the reahn of existent entities according to the pluraHstic hypothesis, we may give the Leibnizian name of " monads." The facts we have just been considering can then be interpreted by supposing that among the monads whose appearance constitutes the object of experience of each of certain particular monads, there exists a group (perhaps even " society " may prove to be an appropriate term) which stands to the latter in a unique relation. The members of this group may be considered to have attained a level of experience inferior to that of the monad to whom they are related in this way. Moreover the members of the group are related to one another in at least some ways in which they are not related to other monads who are not members of the group. An individual organism thus consists of a dominant monad in a certain relation to a group of subordinate monads. The former constitutes what is commonly called the " mind " of the organism. The latter, albeit themselves inferior " minds," constitute the " body " of the organism. In this way the concepts of body and mind can be interpreted in terms of existent entities as opposed to the appearances of such entities ; and the determination of the relation between them consists in the determination of the relation between a dominant monad and its attendant group of subordinate monads. As will shortly be apparent, the necessary investigation involves an enquiry into the nature of certain subsidiary relations closely connected with the one in question. We have thus at last given a more concrete signifi- cance to our problem — a significance of a type which is Phiralistic Interpretation of the Matter 201 probably at the back of most men's minds when they consider the relation of mind and body, though all would not apfree as to the nature of the existent entities involved. Let us now see what can be said on the question when the nature of these entities is such as has been indicated. IV. Application of Pluralism to the Solution OF THE Problem Before proceeding with our main theme, let us pause a moment to consider what type of result may be regarded as truly a solution of the problem we are examining. It may be remarked in the first place, that no statement of the relation of body and mind in terms of relations of the kind distinguished in the object of experience can be considered a satisfactory solution. For not only are propositions asserting the subsistence of such relations purely descriptive as opposed to explanatory, but they subsist in all cases between phenomenal objects, whereas we are here dealing with existent entities. In particular, the causal relation, as it is understood in physical science, is no exception to this. Empirical observation only warrants the assertion of the subsistence of causal laws among phenomena, i.e., general propositions whereby the occurrence of certain events can be inferred from the occurrence of certain other events. No such object, for example, as a "causal nexus," as it is sometimes called, which evidences the presence of some principle of efficiency, is presented in immediate experience. The existence of causal efficiency in the subjective interpretation is, 202 The Relation of Mind and Body of course, another and entirely different question'. But the point is, that no type of "action " observed in the phenomenal world can be applied to the case of the action of mind on body or of body on mind. To attempt such an application would bring us no whit nearer the solution we require. On the other hand, it must not be forgotten that the pursuit of our enquiry will inevitably lead, sooner or later, to ultimate indefinables. Provided, however, the latter are such that their essential nature is fully realizable by us, the fact that it cannot be described explicitly with any adequacy matters not a jot. Indeed, the state of affairs which precludes such explicit descrip- tion is the very goal to be sought after in all enquiries possessing any degree of finality. For if the result of enquiry be stated in terms of entities whose nature we clearly realize ^ the mere indication of those entities provides a solution which can be called explanatory'^ in the only true sense. The matter which primarily concerns us is the relation of the dominant monad to the subordinate monads associated with it. Only second in importance to this, is the relation between the various subordinate monads which are members of one organism. Evidently a clear appreciation of the latter relation will depend upon ^ See "Scientific Method in Philosophy and the Foundations of PluraHsm," Section VII. ^ See ^94 Forgetting, 228 Free will, xvii, 116, 145 ff., 154, 330 f. God, 331 Gurney, Edmund, 266, 275 Haeckel, Ernst, 207 n. Hallucination, 272 ; collective, 318 f. Haunting, 294 f. Hegel, 116 Heraclitus, 39 Hume, 221 n. Hypnotism, 265 ff., 311 ff., 315 f- Hypothesis, 12 ff. Idealism, xvii Illusion, 32 f. Imagination, xx, 220 ff., 241 f., 245 f., 252 ff. Immanence, 212 ff., 240, 313 ff. Immortality, xviif., 162 ff., 330 f. Immunity, from pain, etc., 270 f. 334 Index Impressions, of sense, 221 fF., 241 f., 245 f., 247 ff. Individual, 29 f., 331 Insanity, 276, 279 Intellection, 233 Intensity, of sense-data, see Sense- data Interaction, 79, 185 f., 203 f., 250, 285 ff., 291, 319 ff., 329 James, William, 194 n. Knowledge, xiv, 21, 90 Laplace, I24f, 140, 150 Laws, causal, 31 ; evolution of natural, 75 ff., 81 f. Leibniz, 66, 78, 200, 216 n. Levitation, 317 ff., 327 Locke, 206, 226 Lotze, 213 n. Lourdes, miracles of, 270 Materialism, 9 Materialization, 317 Matter, xix, 8, 9, 11, 53, 72 f., 76 f., 92, 126 f. McDougall, W., 282 McTaggart, J. M. E., xiii n., 221 n. Meaning, 73 f., 210 Means, 51 Mechanism, 122 f, 149 Medium, psychic, 282, 304 ff., 326 f. Memory, xx, 220 ff. Method, scientific, i, 4 ff., 16, 29, 54 ff., 62 f. ; genetic, 12, 16, 59 Mind, xvii, xix, 8 f, 71 ff., 92, io6f, 126 f, 137 ff., 181 ff, 194 f., 200 Minkowski, 42, 45 Monads, 75, 78 ff., 82, 200 ff., 323 f. Newton, 135 Number, 130 Object, of experience, 6, 16 f, 28, 52, 58 f., 68 ff , 92 ff., 102, 108 ff., 166 ff., 169 ff., 173 ff, 250, 260, 329 Occam's razor, 16, 104 Organism, 72 f, 78, 207 ff., 240, 329 Parallelism, psycho-physical, 122, 124, 149, 187, 238 Part, 173 ff Perception, abnormal, 205, 219, 257, 261, 264, 283 ff. Permanence, 39 ff., 330 Perry, R. B., 7 n., 29 n., 32 n., 93 n., 94 n., 188 n. Personality, xx, 232 ff., 305, 310 Phantasms, 288 ff. Philosophy, nature of, xiii f Physics, objects of, 79, 103 f., in, 133 ff., 190 f, 238 Planchette, 275 Pluralism, vii, xvf., 3, 8ff., 53, 58 ff., 63f., 65ff., 75 ff., 104, ii2f., 198 ff., 206, 239, 329 ff. Poltergeist, 317 Presentation, relation of, 92 ff., io8f., 194 Prince, Morton, 282 Pringle-Pattison, A. S., 66, 69, 75 ff., 81 f. Psychical research, xx, 244 ff., 328 Purpose, 51 ff., I46f. Quality, 48 ff., 128, 132, 136 f., 160 f. Quantity, I27f, 130 ff., 151 ff., 155 ff. Rapport, 213 f. Raps, psychic, 317 ff. Realism, New, vii f., xvi, i, 4, 16 ff., 33, 85, 93 f, 98 ff, iiof., i88f. Realization, 13 f., 19, loi, 139, 202 Recognition, 222 f. Reflexes, 72, 81, 207 f. Relation, 26 f., 48 ff. Relativity, principle of, 45 f. Russell, Bertrand, viif., xiii n., xvi, in., 4n., 17, 18, 24, 26 ff., 35, 56, 85 n., 90, 95 ff., 99 f., 102 ff., 107 ff., iigff, i28f, i3on., I48f., 152, 175 n., 191 n. Science, concepts of physical, 5 f., Ill, 133 ff., 191 Self, as subject, 8ff., 13 f, 43, 62, 67, 146, 258 f.; existence of, 18 ff., 44f., 58, 63, 163; empirical, 19, 194, 232, 259f, 278 ff., 326; con- tent of, 234 ; subliminal, see Sub- liminal self Index 335 Self-consciousness, see Conscious- ness Sensation, 84 ff., 211 fif. Sense-data, intensity of, 49, 132, 155 ff.; differences in, 86 ff, 105 f ; series of, 24 fif. ; being of, 95 fif., loi f., 195 ff., 225, 239; general characteristics of, 5 f-, 10, 59 Sensibilia, 96, 100, 102, 108 f. Sensibility, general, 193, 199 Sleep, 166, 171, 262 ff., 265 f, 267, 290, 291 n. Solipsism, 5, 21 f., 170, 197 Space, xvi, 42 ff., 330 Spaulding, E. G., 174"- Spencer, Herbert, 9, 197 Stigmata, 271 Stimulus, 84 ff., 157, 210 Stout, G. F., 85 n., 88 n. Subconsciousness, 85, 234, 244 ff. Subject, of experience, vii f , 8, 9, 14, 17, 19, 43> 56 ff., 68 ff., 92 ff., loif, io8ff., 137 ff., 144 f., 166 ff., 169 ff., I73> 194, 223ff., 249 ff., 256, 258 f, 277, 285, 329 ff. Subliminal images, 246, 252ff., 326; self, 256 ff., 263, 326 Substance, 39 ff., 47 f Suggestion, 265 ff. Teleology, 51 ff., 122 f, 146 f, 149, 207, 210 Telepathy, 283, 284, 295, 297 ff., 311,326 Tendencies, subjective, 255 ff., 260, 266 ff. Theism, 10, 203 n. Time, xvi, 36, 40 ff., 43 ff., 164 ff., 178 f, 330; appreciation of, 273 Trance, 265 ff, 283 f , 305 ff., 311 f. Unity, of experience, see Experi- ence ; of reality, xxi, 79 f , 82 f , 203 f., 250, 285, 286, 299 ff., 325, 329 ff- . . Ultraliminal impressions, 245, 247 ff., 285 f , 326 Volition, 211 ff., 219 f., 288, 289 f Ward, James, 9, ion., 20 n., 65, 72n., 75, 85n., 92n., i6on., 203n.,207n., 209 n., 212, 213, 215 n., 222 n., 223 n., 228 n., 229 n., 232 n., 234, 245 n., 246 n., 278 n. Weber, 84 Weber-Fechner Law, 84 ff. Whitehead, A. N., i n., 102, 130 n. Whole, 173 ff. Wundt, 253 CAMBRIDGE : PRINTED BY J. B. PEACE, M.A., AT THE UNIVERSITY PRESS. UNIVERSITY OF CALIFORNIA LIBRARY Los Angeles This book is DUE on the last date stamped below. Form L9-Series 444 H' ™mm\\mi 3 Tr5ro64S''Tif' UC SOUTHERN REGIONAL LIBRARY FACILITY AA 000 184 997 5 m^T^^F 'ORNIA U£^^.. :'6